BAF312 (Siponimod) : 2.7.2 Summary of Clinical Pharmacology Studies in Multiple Sclerosis

Download as pdf or txt
Download as pdf or txt
You are on page 1of 143

Novartis Institutes for BioMedical Research

BAF312 (siponimod)

BAF312A

2.7.2 Summary of Clinical Pharmacology Studies


in multiple sclerosis

Document type: CTD Clinical summary document

Document status: Final

Release Date: 14-Aug-2018

Property of Novartis
Confidential
May not be used, divulged, published or otherwise disclosed
without the consent of Novartis

Anonymized Page 1
Novartis Confidential Page 2
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table of Contents

Table of Contents................................................................................................................... 2
Abbreviations......................................................................................................................... 8
1 Background and Overview ................................................................................................. 12
1.1 General overview ........................................................................................................ 12
1.2 Summary of overall conclusions ................................................................................. 14
1.2.1 Pharmacokinetics ............................................................................................. 14
1.2.2 Drug interactions ............................................................................................. 17
1.2.3 Pharmacodynamics .......................................................................................... 19
2 Summary of Results of Individual Studies ......................................................................... 22
2.1 In vitro, ex vivo, and in silico studies using human biomaterials ............................... 22
2.1.1 Plasma protein binding .................................................................................... 22
2.1.2 Hepatic metabolism and drug interaction studies ............................................ 23
2.1.3 Studies using other human biomaterials .......................................................... 23
2.2 Studies in healthy subjects .......................................................................................... 24
2.2.1 Single ascending dose study ............................................................................ 24
2.2.2 Multiple ascending dose studies ...................................................................... 25
2.2.3 Dose titration study .......................................................................................... 26
Table 2-1 Study A2107 Dose Titration Schedule ..................................... 27
Figure 2-1 Mean daily minimum heart rate after once daily
dosing administered over 12 days as a fixed
dose or uptitration to 10 mg ................................................. 28
2.2.4 Treatment interruption and re- initiation .......................................................... 29
2.2.5 Human ADME study ....................................................................................... 30
2.2.6 Bioequivalence and relative bioavailability..................................................... 30
2.2.7 Drug-drug interaction studies .......................................................................... 31
Figure 2-2 Emax heart rate (bpm) by time (0 to 24 hours)
and treatment group for co‑ administration of
siponimod and beta blocker propranolol .............................. 34
2.2.8 Absolute bioavailability and pharmacodynamic/cardiac safety of
intravenous formulations ....................................................................... 35
2.2.9 Thorough QT/QTc study ................................................................................. 36
Figure 2-3 Estimated mean difference and 2-sided 90
percent CI for change from placebo-adjusted
time-matched Baseline in QTcF (msec) by
time point and siponimod treatment (2 and 10
36
mg) .......................................................................................
2.2.10 Studies in special populations........................................................................ 37

Anonymized Page 2
Novartis Confidential Page 3
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.3 Studies in patients........................................................................................................ 38


2.3.1 Multiple sclerosis ............................................................................................. 39
2.3.2 Polymyositis and dermatomyositis .................................................................. 40
3 Comparison and Analysis of Results Across Studies ......................................................... 41
3.1 Pharmacokinetics – General characteristics ................................................................ 41
Figure 3-1 Effect of intrinsic and extrinsic factors on siponimod PK
(geometric mean ratios and 90 percent confidence
intervals) ................................................................................................42
3.1.1 Absorption and absolute bioavailability .......................................................... 43
Figure 3-2 Geometric mean plasma concentrations of
siponimod after single doses of siponimod (0.1
to 75 mg) (linear scale)......................................................... 43
3.1.2 Effect of food ...................................................................................................45
3.1.3 Dose proportionality ........................................................................................ 45
Table 3-1 Individual study estimation of siponimod dose-
pharmacokinetic relationship after a single
46
dose.......................................................................................
Table 3-2 Pooled study estimation of siponimod dose-
pharmacokinetic relationship after a single
46
dose.......................................................................................
Figure 3-3 Pooled study estimation of dose normalized
Cmax dose proportionality after a siponimod
single dose ............................................................................47
Figure 3-4 Pooled study estimation of dose normalized
AUC0-24h dose proportionality after a
siponimod single dose .......................................................... 48
Table 3-3 Estimation of siponimod dose-pharmacokinetic
relationship after multiple doses in healthy
49
subjects .................................................................................
Figure 3-5 Plasma concentrations of siponimod normalized
to dose as a function of the dose in RRMS
49
patients .................................................................................
3.1.4 Pharmacokinetics of single doses .................................................................... 50
Table 3-4 Main PK parameters after siponimod single dose
administration (0.1 to 75 mg) to healthy
51
subjects .................................................................................
3.1.5 Pharmacokinetics of repeated administration .................................................. 52
Table 3-5 Main pharmacokinetic parameters after
siponimod once daily administration (0.3 to
20 mg) over 28 days to healthy subjects –
53
Day 28 ..................................................................................

Anonymized Page 3
Novartis Confidential Page 4
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-6 Daily geometric mean plasma concentrations of


siponimod after once daily dosing
administered over 12 days as a fixed dose or
uptitration to 10 mg .............................................................. 54
3.1.6 Distribution and protein binding...................................................................... 54
55
3.1.7 Metabolism ......................................................................................................
Figure 3-7 General scheme of siponimod biotransformation
pathways in human ............................................................... 57
Table 3-6 M3 unbound AUCtau,ss and EC50 on human
S1P1 receptor ....................................................................... 59
Table 3-7 Summary of M17 pharmacokinetic parameters in
Study A2126 Part 1 .............................................................. 60
Table 3-8 Unbound M17 AUCtau and EC50 on human
S1P1 and S1P5 receptors...................................................... 60
3.1.8 Elimination/Excretion ...................................................................................... 60
3.1.9 Population pharmacokinetics ........................................................................... 62
3.2 Pharmacokinetics in subpopulations ........................................................................... 63
3.2.1 Patients versus healthy subjects ....................................................................... 63
Table 3-9 Geometric mean (percent geometric mean
coefficient of variation) siponimod plasma
trough concentrations by treatment and visit in
MS patients........................................................................... 64
Table 3-10 Geometric mean (percent geometric mean
coefficient of variation) trough concentrations
by visit in MS patients (2 mg qd) ......................................... 64
Figure 3-8 Box plot by dose for comparison of steady state
pharmacokinetic trough concentrations
between healthy subjects and patients with
multiple sclerosis .................................................................. 65
3.2.2 Hepatic impairment ......................................................................................... 66
3.2.3 Renal impairment............................................................................................. 67
68
3.2.4 Ethnic origin ....................................................................................................
68
3.2.5 Genotype ..........................................................................................................
Table 3-11 Genotype frequencies of CYP2C9
polymorphisms in Caucasian, Asian and
African subjects .................................................................... 69
Table 3-12 Summary statistics of siponimod
pharmacokinetic parameters per CYP2C9
genotype after a single 0.25-mg dose in Study
A2128 Part 1 (pharmacokinetic analysis set) ....................... 70

Anonymized Page 4
Novartis Confidential Page 5
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-13 Number of individuals (percent) with different


CYP2C9 genotypes across studies (population
pharmacokinetic analysis sets) ............................................. 71
Table 3-14 Estimated siponimod CL/F per CYP2C9
genotype in the two population
pharmacokinetic analyses for a typical 69 to
70 kg subject......................................................................... 72
Table 3-15 Recommended siponimod maintenance doses
per CYP2C9 genotype.......................................................... 73
73
3.2.6 Gender..............................................................................................................
3.2.7 Body weight ..................................................................................................... 73
3.2.8 Elderly (greater than or equal to 65 years old) ................................................ 74
3.2.9 Children (less than 18 years old) ..................................................................... 74
3.3 Drug interactions ......................................................................................................... 74
Figure 3-9 Effect of concomitant medications on siponimod PK and
effect of siponimod on comedication PK (geometric mean
ratios and 90 percent confidence intervals) ........................................... 75
3.3.1 Pharmacokinetic interactions ........................................................................... 76
Figure 3-10 Schematic description of the drug disposition
pathways for siponimoda ..................................................... 77
Table 3-16 Summary statistics of pharmacokinetic
parameters for co‑ administration of
siponimod and CYP2C9/3A4 inducer rifampin
- Days 12 and 24 (pharmacokinetic analysis
79
set) ........................................................................................
Table 3-17 Main pharmacokinetic parameters for
co‑ administration of siponimod and CYP2C9
inhibitor fluconazole ............................................................ 80
Table 3-18 Siponimod pharmacokinetic parameters on Day
1 (siponimod alone) and Day 19 (siponimod
co-administered with itraconazole) ...................................... 81
Table 3-19 Predicted siponimod single dose inhibition area
under the curve ratios in the presence of
CYP3A4 and CYP3A4/CYP2C9 inhibitors in
various CYP2C9 genotypes ................................................. 84
Table 3-20 Predicted siponimod steady state induction area
under the curve and maximum concentration
ratios in the presence of CYP3A4 and
CYP3A4/CYP2C9 inducers in various
CYP2C9 genotypes .............................................................. 85

Anonymized Page 5
Novartis Confidential Page 6
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-21 Net inhibition effect ratios (inhibition x


CYP2C9 genotype exposure ratios) ..................................... 86
Table 3-22 Net induction effect ratios (induction x
CYP2C9 genotype exposure ratios) ..................................... 86
3.4 Pharmacodynamics...................................................................................................... 87
3.4.1 Efficacy assessments ....................................................................................... 87
Table 3-23 Lymphocytes categorical outlier analysis –
multiple dose studies ............................................................ 88
89
3.4.2 Safety assessments ...........................................................................................
Table 3-24 Incidence of AEs (at least 2 percent) by SOC
and PT in single dose studies in healthy
90
subjects .................................................................................
Table 3-25 Incidence of AEs (at least 2 percent) by SOC
and PT in multiple dose studies in healthy
91
subjects .................................................................................
Table 3-26 Heart rate categorical outlier analysis (Holter
data) – multiple dose studies ................................................ 93
Table 3-27 Bradyarrhythmia events categorical analysis
(Holter data) - multiple dose studies .................................... 96
Table 3-28 QTcF events categorical analysis (Holter data) -
multiple dose studies ............................................................ 98
3.4.3 Exposure – lymphocyte relationship ............................................................... 104
Table 3-29 Effect of covariates on the lymphocyte response
model parameters ................................................................. 105
Figure 3-11 Simulated median of absolute lymphocyte
count profiles following administration of
2 mg siponimod once daily for 40 days in
typical SPMS patients, by gender, ethnicity
and population type .............................................................. 106
Table 3-30 Median lymphocyte count at the end of the last
day of dosing in patients with secondary
progressive multiple sclerosis following
administration of 2 mg siponimod once daily
for 40 days (steady-state conditions).................................... 108
Table 3-31 Time in days needed for absolute lymphocyte
count recovery to 1.0 x 109/L for 2 mg
siponimod once daily for 40 days (steady-state
108
conditions) ............................................................................

Anonymized Page 6
Novartis Confidential Page 7
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-32 Median (90 percent prediction interval)


simulated trough siponimod plasma
concentration and trough lymphocyte count at
steady state in healthy subjects by dose ............................... 109
Figure 3-12 Median simulated steady-state trough
siponimod concentration versus trough
lymphocyte count in healthy subjects by dose ..................... 109
4 Special studies .................................................................................................................... 109
5 Appendix............................................................................................................................. 110
Table 5-1 Clinical Pharmacology studies in healthy subjects and special
populations and studies in patients with MS, PM and DM ...................................110
Table 5-2 In vitro, ex vivo, and in silico studies using human biomaterials .....................118
Table 5-3 Summary of pooled analyses included in the Summary of Clinical
Pharmacology ........................................................................................................123
Table 5-4 Summary of in vitro and ex vivo plasma protein binding of
siponimod and metabolites....................................................................................128
Table 5-5 Siponimod, M3 and M17c acting as a perpetrator drug ...................................129
Table 5-6 Siponimod acting as a victim of drugs..............................................................130
5.1 Literature references....................................................................................................133
5.2 Post-text tables and figures .........................................................................................134
Figure 11-1.1 Dose normalized BAF312 PK parameters for primary
interest for siponimod single oral doses(0.1, 0.3, 1, 2.5, 5, 10,
17.5, 20, 25, 75 mg) PK analysis set ....................................................... 135
Figure 11-3.2 Box plot by dose for comparison of steady state PK
trough concentration between Healthy subjects and MS patients
PK analysis set ......................................................................................... 137
Table 11-1.1 Dose proportionality of BAF312 PK parameters of
primary interest for siponimod single oral doses(0.1, 0.3, 1, 2.5,
5, 10, 17.5, 20, 25, 75 mg) PK analysis set ............................................. 138
Table 12_5.1 Lymphocytes categorical outlier analysis - Single dose
studies Safety analysis set........................................................................ 139
Table 12_5.2 Lymphocytes categorical outlier analysis - Multiple dose
studies Safety analysis set........................................................................ 141
Table 12_5.3 Lymphocytes categorical outlier analysis – PM/DM
studies Safety analysis set........................................................................ 143

Anonymized Page 7
Novartis Confidential Page 8
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Abbreviations
ABCB1 ATP-binding cassette sub-family B member 1/P-glycoprotein
ABCG2 breast cancer resistant protein
ABPM ambulatory blood pressure monitoring
ADME absorption distribution metabolism excretion
AE adverse event
ALC absolute lymphocyte count
ALT alanine aminotransferase
AP alkaline phosphatase
AST aspartate aminotransferase
AUC area under the curve
AUC0-t area under the curve to a defined point in time
AUCinf area under the plasma (or serum or blood) concentration-time curve extrapolated to
infinity
AUClast area under the curve calculated to the last quantifiable concentration point
AUCMABP area under the mean-arterial-blood-pressure curve
AUCtau area under the curve calculated to the end of the dosing interval, tau
AUCtau,ss area under the curve calculated to the end of the dosing interval, tau, at steady state
AUEC area under the effect curve
AV atrioventricular
BCRP breast cancer resistant protein
bid twice a day
BMI body mass index
BP blood pressure
BSEP bile salt export pump
Cav,ss average steady state concentration during multiple dosing
Cb/Cp ratio of blood-to-plasma concentrations
CM central memory T cells
CI confidence interval
Clast last observed quantifiable concentration
CL systemic (or total body) clearance from plasma (or serum or blood) following
intravenous administration
CL/F apparent systemic (or total body) clearance from plasma (or serum or blood) following
extravascular administration
CLr renal clearance from plasma (or serum or blood)
Cmax observed maximum plasma (or serum or blood) concentration following administration
Cmax,ss observed maximum plasma (or serum or blood) concentration following administration
at steady state
Cmin,ss observed lowest plasma (or serum or blood) concentration during a dosing interval at
steady state
CNS central nervous system
CSF clinical service formulation (capsule)
CUAL combined unique active [MRI] lesion
CV coefficient of variation

Anonymized Page 8
Novartis Confidential Page 9
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

CYP cytochrome P450


D1 duration of zero-order absorption
DBP diastolic blood pressure
DC dendritic cells
DDI drug-drug interaction
DIBD development international birth date
DLCO diffusing capacity for carbon monoxide
DM dermatomyositis
DT dose titration
ECG electrocardiogram
EE ethinylestradiol
Emax maximum effect
EoP2 End-of-Phase 2
EOS End-of-Study
F bioavailability of a compound
F10 modified release formulation which releases siponimod from its matrix over 10 hours
F16 modified release formulation which releases siponimod from its matrix over 16 hours
FDA Food and Drug Administration
FEF25-75% maximum mid-expiratory flow
FEV1 forced expiratory volume in 1 second
FMI final market image
FSH follicle stimulating hormone
fu unbound fraction
FVC forced vital capacity
GGT gamma-glutamyl transferase
HEK293 human embryonic kidney
HLM human liver microsomes
HR heart rate
IC50 Concentration at which 50% of the maximum effect/inhibition of response is attained
IIV inter-individual variability
Ig immunoglobulin
Imax maximum inhibitory effect
IR immediate-release
iv intravenous
ka absorption rate constant
LA long acting
LFT liver function test
LH luteinizing hormone
LLC-PK1 porcine kidney cell line
LVG levonorgestrel
MABP mean arterial blood pressure
MAD multiple ascending dose
MATE multidrug and toxic extrusion

Anonymized Page 9
Novartis Confidential Page 10
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

MDCKII Madin-Darby canine kidney II cells


MDR1 multidrug-resistant protein 1/ P-glycoprotein
MF market formulation
MR modified-release
MRI magnetic resonance imaging
mRNA messenger ribonucleic acid
MRP2 multidrug resistance-associated protein 2
MS multiple sclerosis
MTD maximum tolerated dose
MXR mitoxantrone resistant protein
N naïve T cells
NK natural killer cells
OAT organic anion transporter
OATP organic anion transporting polypeptide
OC oral contraceptive
OCT organic cation transporters
od once a day
PBPK physiologically-based pharmacokinetic(s)
PD pharmacodynamic(s)
PEM peripheral effector memory T cells
PEmax maximum static expiratory pressure
PFT pulmonary function test
PG pharmacogenetic
P-gp P-glycoprotein
PI prediction interval
PImax maximum static inspiratory pressure
PK pharmacokinetic(s)
PM polymyositis
po oral(ly)
PopPK Population pharmacokinetics
PopPKPD Population pharmacokinetics/pharmacodynamics
PT preferred term
PXR pregnane X receptor
qd once daily
QTc corrected QT
Racc accumulation ratio
rhCYP recombinant human cytochrome P450
RRMS relapsing-remitting multiple sclerosis
S1P sphingosine-1-phosphate
SAD single ascending dose
SAE serious adverse event
SBP systolic blood pressure
SD standard deviation

Anonymized Page 10
Novartis Confidential Page 11
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

SE standard error
SHBG sex hormone binding globulin
SLC solute-carrier
SOC system organ class
SNP single nucleotide polymorphism
SPMS secondary progressive multiple sclerosis
SVE supraventricular ectopy
T1/2 half-life
TEmax time to maximum effect
Tlag lag time
Tmax time to reach peak or maximum concentration
Tmax,ss time to reach peak or maximum concentration following drug administration at steady
state
ULN upper limit of normal
UPLC ultra performance liquid chromatography
WHO World Health Organization
Vc/F apparent volume of central compartment
VE ventricular ectopy
Vp/F apparent volume of peripheral compartment
Vz apparent volume of distribution during the terminal elimination phase following
intravenous administration
Vz/F apparent volume of distribution during the terminal elimination phase following
extravascular administration
Vss apparent volume of distribution at steady state following intravenous administration
Vss/F apparent volume of distribution at steady state following extravascular administration

Anonymized Page 11
Novartis Confidential Page 12
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

1 Background and Overview


1.1 General overview
Siponimod (BAF312) is a new chemical entity proposed as an oral therapy for the treatment of
secondary progressive multiple sclerosis (SPMS). Siponimod is a new member of a class of oral
compounds referred to as sphingosine-1-phosphate (S1P) receptor modulators and selectively
targets 2 (S1P1 and S1P5) of the 5 known S1P receptors. Overall, non-clinical observations
provide evidence that siponimod has immunomodulatory and neuroprotective properties
required for therapeutic approaches in multiple sclerosis (MS).
T cells selectively require S1P1 activation for emigration from the thymus, and both T and
B cells require this receptor for egress from peripheral lymphoid organs (Matloubian et al 2004,
Brinkmann et al 2004). Siponimod promotes a marked and long-lasting internalization and
degradation of S1P1 receptors, thereby acting as a functional antagonist on S1P1. This makes
lymphoid cells unresponsive to S1P signaling, thus depriving them of their capacity to egress
from lymphoid organs (lymph nodes and Peyer’s patches) and thereby preventing the
recirculation of lymphocytes to blood and tissues including the central nervous system (CNS).
Similar to other drugs in this class, siponimod does not deplete lymphocytes but rather results
in redistribution away from the intravascular compartment (Matloubian et al 2004). This effect
results in a dose-dependent reduction of peripheral blood absolute lymphocyte count (ALC).
Additionally, siponimod readily crosses the blood brain barrier and evidence from preclinical
models demonstrates that siponimod targets S1P1 on astrocytes and S1P5 on oligodendrocytes
(Bryan et al 2008), suggesting a potentially beneficial direct neurobiological effect in the CNS
independent from effects on peripheral lymphocytes.
A total of 25 clinical studies of siponimod have been completed to date: 20 Clinical
Pharmacology studies in healthy subjects (18) or special populations (2), 2 studies in patients
with MS and 3 studies in patients with polymyositis (PM) or dermatomyositis (DM) (Table 5-1).
As of 31-May-2017, a total of approximately 3278 unique subjects (1249 healthy subjects, 24
hepatic impaired subjects, 8 renal impaired subjects, 49 PM/DM patients and 1948 MS patients)
have been enrolled into the siponimod clinical program (inclusive of Phase 1, Phase 2 and
Phase 3). The cumulative exposure of clinical trial MS patients to siponimod since the
development international birth date (DIBD) 05-Mar-2009 and until the 31-May-2017 cut-off
is estimated at 4048.6 patient-years.
In the 20 Clinical Pharmacology studies, a total of approximately 1281 subjects have been
enrolled, of which approximately 880 healthy subjects, 24 hepatic impaired subjects and 8 renal
impaired subjects have received siponimod and approximately 434 healthy subjects have
received placebo. This includes subjects who sequentially received placebo and siponimod. In
addition, approximately 363 healthy subjects were exposed to other study drugs either alone or
in combination with siponimod.
Healthy subjects have received siponimod as single doses (0.1 to 75 mg) or multiple doses (0.25
to 20 mg) daily up to 38 days. Patients with MS have received siponimod (0.25 to 10 mg) daily
up to 73 months during the clinical development program. Siponimod was generally well
tolerated and showed a safety profile which was in line with that of fingolimod, an S1P receptor
modulator approved for the treatment of relapsing MS, and comparable to that of other S1P

Anonymized Page 12
Novartis Confidential Page 13
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

modulators (Subei and Cohen 2015). The maximum tolerated dose (MTD) for a single dose of
siponimod was determined to be 25 mg based upon the occurrence of symptomatic bradycardia
after a single dose of 75 mg. Safety and tolerability of the highest investigated multiple dose of
20 mg for 28 days was satisfactory, so that the MTD in the multiple dose setting could not be
established.
The effects of siponimod cease more rapidly after discontinuation in comparison to fingolimod
(recovery of ALC within days, compared to months) due to its short half-life (T1/2: 30 hours).
The short T1/2 has particular relevance in clinical situations requiring a fast wash-out and re-
establishment of normal immune function, including patients in need of vaccinations, surgical
interventions, becoming pregnant or with specific (infectious) adverse events (AEs) requiring
treatment discontinuation.
Similar to other S1P modulators acting through S1P1, negative chronotropic effects were
common at treatment initiation of siponimod. In the single ascending dose (SAD) and multiple
ascending dose (MAD) studies, a transient, dose-dependent decrease in heart rate (HR) was
observed during the initial dosing phase that plateaued at doses ≥ 5 mg.
Occasional asymptomatic bradyarrhythmic events (atrioventricular (AV) blocks and sinus
pauses) were detected at a higher incidence under siponimod treatment compared to placebo
and were typically evident after the first dose. A specifically designed dose titration (DT)
[Study A2107] demonstrated that the effects on HR and AV conduction observed under non-
titrated conditions could be attenuated by DT. A DT regimen was employed in subsequent
multiple dose Clinical Pharmacology studies and in the Phase 2 and 3 studies in MS patients,
in which it efficiently mitigated the known effects on HR and AV conduction.
During the US FDA End-of-Phase 2 (EoP2) consultation in September 2011, Novartis agreed
with the agency on a panel of 19 Clinical Pharmacology studies which were considered to
constitute a complete package that supports the submission of the siponimod NDA (Table 5-1).
In addition, [Study A2126] was conducted to evaluate the absolute bioavailability of a single
oral dose of siponimod.
At the time of conduct of the DT study the highest anticipated therapeutic dose in the ongoing
Phase 2 [Study A2201] in MS patients was 10 mg, which was also investigated as the highest
dose in [Studies A1101, A2104 and A2107] conducted after the SAD/MAD studies (N = 3).
During the course of clinical development, the doses investigated in the Clinical Pharmacology
program were adjusted to the highest predicted therapeutic maintenance dose level as
determined by pharmacokinetic (PK)/pharmcodynamic (PD) analyses on magnetic resonance
imaging (MRI) assessments (combined unique active lesion; CUAL) in the Phase 2 dose
selection (Study A2201) in relapsing-remitting MS (RRMS) patients ([SCE-Section 4.1]). The
highest therapeutic dose was adjusted to 4 mg, as investigated in the Clinical Pharmacology
[Studies A2119, A2108, A2110, A2111 and A2121] and finally to a therapeutic maintenance
dose of 2 mg qd, as investigated in the Phase 3 [Study A2304] in SPMS patients and in the
subsequent 9 PK or PD Clinical Pharmacology [Studies A2118, A2116, A2122, A2129, A2130,
A2128, A2126, A2125 and A2124].
The clinical benefit (efficacy), safety and tolerability of siponimod have been mainly
investigated in patients with RRMS and SPMS. Siponimod was also investigated in 3 clinical
studies in patients with PM and DM (Table 5-1). PM and DM comprise a heterogeneous group

Anonymized Page 13
Novartis Confidential Page 14
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

of chronic inflammatory muscle diseases suggested to be mediated by autoreactive lymphocytes.


This program was prematurely terminated in May 2016 for non-safety related reasons (limited
efficacy, slow recruitment). Clinical investigation of intravenous (iv) and oral siponimod as a
treatment for intracerebral hemorrhage has been started in December 2017 and the first patient
was treated in January 2018.
This Summary of Clinical Pharmacology Studies summarizes the PK (absorption, distribution,
metabolism and excretion (ADME), population PK (PopPK) analyses and influence of intrinsic
and extrinsic factors, such as CYP2C9 genotype and food effect), PK and PD drug-drug
interactions (DDIs) and PD (analyses related to efficacy, safety and exposure-response
relationships) of siponimod in healthy subjects, special populations and patients with PM/DM.
In addition to results of the 25 clinical studies listed in Table 5-1, these summaries include
results from pooled PK and pooled categorical analyses of primary and secondary PD effects
of siponimod in Clinical Pharmacology and PM/DM patient studies (detailed in Table 5-3), and
2 Population PK analyses and an exposure-lymphocyte relationship analysis (which include
data from MS patients in (Studies A2201 and A2304). The results of in vitro, ex vivo and in
silico studies for siponimod and its metabolites M3, M5 and M17 using human biomaterials
(presented in Table 5-2) are also summarized in this module.

1.2 Summary of overall conclusions


The PK and PD characteristics and DDIs of siponimod in humans are summarized as follows.

1.2.1 Pharmacokinetics
An overview of intrinsic and extrinsic factors with and without influence on the PK of
siponimod is provided in Figure 3-1 (Section 3.1).

1.2.1.1 Single and multiple dose pharmacokinetics


Siponimod PK after administration of a single oral dose is characterized by a moderate
absorption, small oral clearance, moderate apparent volume of distribution and moderate
apparent terminal half-life (Section 3.1.4). In the SAD study (siponimod 0.1 to 75 mg),
siponimod was measurable in the plasma as early as 0.25 hours post dose and plasma
concentration peaked between 3 and 6 hours (median). The decay of siponimod plasma
concentration was mono-exponential with an apparent terminal T1/2 between 27 and 57 hours.
Following multiple once daily dose administration of siponimod (0.3 to 20 mg) over 28 days in
the MAD study, siponimod PK was consistent with that after single dose administration
(Section 3.1.5.1) thus demonstrating time-independent PK. Median Tmax,ss ranged between
3 and 4 hours. Siponimod trough concentrations indicated that steady state was reached for all
subjects after approximately 6 days. The mean accumulation ratio (Racc) was between 1.88 and
2.72 on Day 28. The effective T1/2 (based on drug accumulation at steady state) ranged between
22 and 36 hours. Mean apparent clearance (CL/F) at steady state was 3.06 to 3.89 L/h,
comparable to the values observed after a single dose.
The exposure (Cmax and AUC) of siponimod rose in an apparent dose-proportional manner
following single and multiple doses of siponimod (Section 3.1.3).

Anonymized Page 14
Novartis Confidential Page 15
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

1.2.1.2 Absorption, distribution, metabolism, elimination/excretion

1.2.1.2.1 Absorption
In vitro, siponimod was classified as a highly permeable compound. Luminal membrane
permeability most likely occurs by a passive permeation process (Section 3.1.1).
Across clinical studies of single oral administration of various solid oral dosage forms, the
median time to reach maximum concentration (Tmax) ranged between 2 and 6 hours
(Section 3.1.1). At equivalent doses, exposure to siponimod is generally similar between oral
formulations. The absolute oral bioavailability of siponimod (0.25-mg dose) was 84%.
Despite a slight delay in Tmax, food intake had no effect on the systemic exposure (Cmax,
AUC) of siponimod (Section 3.1.2). Therefore, siponimod can be taken without regard to meals.

1.2.1.2.2 Distribution and protein binding


Overall, siponimod was moderately distributed within the human body following oral
administration (apparent volume of distribution (Vz/F) of 143 L after a single 0.25 mg-dose;
Section 3.1.6). Siponimod and its metabolites were mainly confined within the plasma
compartment (mean blood/plasma AUC ratio of radioactivity approximately 0.67) and very
highly bound to human plasma proteins (≥ 99.4% each), mainly plasma lipoproteins.

1.2.1.2.3 Metabolism
In vitro it was estimated that CYP2C9 and CYP3A4 contribute to 79.3% and 18.5%,
respectively, of the oxidative metabolism of siponimod (Section 3.1.7).
In human, no active or toxic siponimod metabolites were identified
([Toxicology Written Summary]). Consistent with in vitro findings, the primary phase I
metabolic pathway identified in the human ADME study was hydroxylation of siponimod to
form M5, M6 and M7; phase II glucuronidation of M5 yields M3, 1 of the 2 main circulating
metabolites of siponimod (M3 AUCinf amounted to 27.6% of siponimod exposure).
The cholesterol ester M17 was identified as the most prominent systemic metabolite in the
absolute bioavailability study (M17 AUCinf amounted to 81% to 97% of siponimod).
Additional metabolites detected in the human ADME study are described in Section 3.1.7.2.
Further details of the PK of the main metabolites M3 and M17 are discussed in Section 3.1.7.3.
Overall, unbound Cmax and EC50 values suggest that M3 and M17 systemic exposure are
unlikely to translate into a significant increase in pharmacological activity on S1P1 or S1P5
receptors.

1.2.1.2.4 Elimination/excretion
The apparent clearance of siponimod was low (typical CL/F value: 3.11 to 3.15 L/h).
The effective T1/2 ranged from 22 to 36 hours (Section 3.1.8). For M3, M5, M6 and M7 the
apparent terminal T1/2 ranged between 29.3 and 35.2 hours, compared to 150 to 155 hours for
M17. Fecal/biliary excretion was the major route of elimination (mean: up to 86.7% of
radioactive dose). Unchanged drug accounted for 9.2% of radioactive dose in feces. M5 and
M4 were the major metabolites in feces and M3 was the major metabolite in urine.

Anonymized Page 15
Novartis Confidential Page 16
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

1.2.1.3 Pharmacokinetics of subpopulations

1.2.1.3.1 Patients with multiple sclerosis or polymyositis/dermatomyositis versus


healthy subjects
The PopPK analyses of the 2 clinical studies of siponimod in patients with RRMS (Phase 2)
and SPMS (Phase 3) suggest that PK of MS patients and healthy subjects do not have significant
differences (Section 3.2.1.1). PK assessments across the 3 clinical studies of siponimod in
patients with PM or DM were also consistent with healthy subjects (Section 3.2.1.2).

1.2.1.3.2 Organ impairment: hepatic and renal impairment


In clinical studies of siponimod (0.25 mg single dose) in subjects with mild, moderate and
severe hepatic impairment (Section 3.2.2) and severe renal impairment (Section 3.2.3), a similar
total siponimod plasma exposure was observed compared to demographically matched healthy
subjects. Thus, no dose adjustment of siponimod is warranted in subjects with impaired hepatic
or renal function.

1.2.1.3.3 Ethnic sensitivity


The evaluations of siponimod PK in the clinical study in Japanese subjects and the
Phase 1/Phase 2 and Phase 3 PopPK analyses (in Caucasians, Blacks, Japanese and subjects of
another race, and in Japanese and Chinese subjects, respectively) suggest that race/ethnicity
does not significantly affect siponimod PK (Section 3.2.4).

1.2.1.3.4 CYP2C9 genotype


CYP2C9 is the major metabolizing enzyme for siponimod. CYP2C9 is polymorphic and the
CYP2C9 genotype has been shown to have a significant impact on siponimod metabolism.
In the Caucasian population, the prevalence of the 6 most prominent CYP2C9 genotypes ranges
from 62% to 65% for CYP2C9*1*1, 20% to 24% for CYP2C9*1*2, 1% to 2% for
CYP2C9*2*2, 9% to 12% for CYP2C9*1*3, 1.4% to 1.7 % for CYP2C9*2*3 and 0.3% to
0.4% for CYP2C9*3*3 (Section 3.2.5).
The PK of siponimod varied between CYP2C9*1*1 (wild type) extensive metabolizers and
subjects with other CYP2C9 genotypes in the dedicated clinical study (Section 3.2.5.1) and the
PopPK analyses (Section 3.2.5.2). Overall, regarding siponimod elimination, CYP2C9*1*1 and
*1*2 subjects behave as extensive metabolizers, *2*2 and *1*3 subjects as intermediate
metabolizers and *2*3 and *3*3 subjects as poor metabolizers. Compared to CYP2C9*1*1
subjects, individuals with the CYP2C9*2*2, *1*3 and *2*3 heterotypes had 20%, 35% to 38%
and 45% to 48% smaller CL/F values, respectively. Individuals with the CYP2C9*3*3
genotype had a 74% smaller CL/F. The siponimod exposure is therefore approximatively 25%,
61%, 91% and 284% higher in CYP2C9*2*2, *1*3, *2*3 and *3*3 subjects, respectively, as
compared to *1*1 subjects.
Recommended maintenance dosing adjustments (from 2 mg to 1 mg of siponimod) in
CYP2C9*1*3 and *2*3 subjects are discussed in Section 3.2.5.3. Novartis recommends
exclusion of *3*3 subjects (due to expected significantly higher chronic exposure with potential
safety implications) from a marketing authorization.

Anonymized Page 16
Novartis Confidential Page 17
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

1.2.1.3.5 Other intrinsic factors: gender, body weight, age


The effect of gender, body weight and age were evaluated in the 2 PopPK analyses. There was
no evidence that gender and age influence the PK of siponimod (Section 3.2.6 and Section 3.2.8,
respectively). Additionally, hepatic and renal impairment studies support a minimal difference
in siponimod PK between elderly and younger subjects.
Body weight was identified as a significant predictor of siponimod PK. However, while body
weight influences siponimod CL/F and Vc/F, no dose adjustment based on body weight is
warranted considering the limited effect on siponimod exposure (Section 3.2.7).

1.2.2 Drug interactions


An overview of the effect of concomitant medications on siponimod PK and the effect of
siponimod on comedication PK is provided in Figure 3-9 (Section 3.3).

1.2.2.1 Pharmacokinetic interactions

1.2.2.1.1 Siponimod as a causative agent of interaction


In vitro assessments of siponimod and its main metabolites M3 and M17 did not show any
clinical relevant DDI potential at the therapeutic dose of 2 mg qd for all investigated CYP
enzymes and transporters (Section 3.3.1.1).

1.2.2.1.2 Siponimod as an object of interaction


In vitro studies demonstrated that siponimod is mainly metabolized by the polymorphic enzyme
CYP2C9, followed by a lower contribution of CYP3A4 (Section 3.3.1.2.1).
PK DDI with CYP2C9/3A4 inhibitors and inducers
Three (3) comedications were evaluated in 3 clinical studies (Section 3.3.1.2.2).
Rifampin (CYP2C9/3A4 inducer): Co-administration with siponimod (uptitrated to 2 mg qd) in
CYP2C9*1*1 subjects decreased siponimod Cmax and AUC by 45% and 57%.
Fluconazole (CYP2C9 inhibitor): Co-administration with siponimod (4-mg doses) in
CYP2C9*1*1 subjects increased siponimod AUC by 2-fold.
Itraconazole (CYP3A4 inhibitor): Co-administration with siponimod (0.25-mg doses) led to a
slight decrease (9% to 24%) in siponimod AUC, rather than an increase as predicted by
physiologically-based PK (PBPK) modeling (Section 3.3.1.2.3). The underlying mechanism
leading to this observation is not understood. However, it is considered that the PBPK modeling
appropriately describes the CYP3A4 pathway. These results therefore cannot be extrapolated
to other strong and specific CYP3A4 inhibitors (such as clarithromycin).
Considering the variable CYP3A4 and CYP2C9 metabolic contributions to the overall
clearance of siponimod in the different CYP2C9 genotypes, it is expected that the CYP2C9
genotype influences the effects of CYP3A4 and CYP2C9 inhibitors and inducers
(Section 3.3.1.2.3). PBPK modeling indicated that with decreased CYP2C9 metabolic activity
in the respective genotypes, a stronger effect of the CYP3A4 perpetrators on siponimod
exposure is anticipated.

Anonymized Page 17
Novartis Confidential Page 18
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

When considering dosing adjustment recommendations based on CYP2C9 genotype, DDI


management based on CYP2C9 genotype is discussed in Section 3.3.1.2.4 for CYP2C9/3A4
inducers and inhibitors. Briefly, siponimod may be combined with all types of CYP3A4 and
CYP2C9 inhibitors without relevant implications on safety or efficacy in all patients except for
CYP2C9*2*2 patients. Particular caution should be applied in patients with a CYP2C9*2*2
genotype for combination treatment with moderate CYP2C9/CYP3A4 inhibitors (e.g.
fluconazole). as significant increase of siponimod exposure is expected. Dosage adjustment to
1 mg daily may be considered. Siponimod may be combined with most types of CYP2C9 and
CYP3A4 inducers. However, because of an expected significant reduction in siponimod
exposure, caution should be applied when siponimod is combined with strong
CYP3A4/moderate CYP2C9 inducers (e.g. rifampin, carbamazepine) in all patients regardless
of genotype and with moderate CYP3A4 inducers (e.g. efavirenz, modafinil) in patients with a
CYP2C9*1*3 or *2*3 genotype.

1.2.2.2 .Pharmacodynamic and pharmacokinetic interactions


The following 3 studies primarily evaluated PD effects of siponimod co-administration with
medications frequently used in the target patient population. In the 2 studies (with monophasic
oral contraceptive (OC) or propranolol comedications) which also evaluated PK effects, the PK
of siponimod and of the co-administered medications were not affected to a clinically relevant
extent.

1.2.2.2.1 Vaccinations: T cell-independent (PPV-23) and T cell-dependent (influenza)


Based on the results of a dedicated study to investigate potential effects of siponimod on the
immune response/immunogenicity of selected vaccines (Section 2.2.7.2.1), it is recommended
that siponimod treatment is paused 1 week prior until 4 weeks after vaccination. Concomitant
siponimod treatment does not compromise the efficacy of a PPV-23 vaccination
(T cell-independent response). The efficacy of quadrivalent influenza vaccination
(T cell-dependent vaccine) is not compromised if siponimod treatment is paused 1 week prior
until 4 weeks after vaccination. Shorter treatment pauses (10 days prior to 14 days after
vaccination) and concomitant treatment showed only modest impairment of influenza
vaccination efficacy, with responder rates approximately 15% to 30% lower than on placebo.

1.2.2.2.2 Monophasic oral contraceptives: combined EE and LVG


OCs represent a frequently prescribed comedication in female MS patients. An effect of
siponimod on the efficacy of OCs is not expected, based on the results of a dedicated study with
a monophasic OC (combined ethinylestradiol (EE) and levonorgestrol (LVG);
Section 2.2.7.2.2). Co-administration with siponimod (4 mg qd) did not reveal clinically
relevant effects on the PK of the OC components or on the PD of the OC, as determined by the
PD markers estradiol, follicle stimulating hormone (FSH) and luteinizing hormone (LH),
ovarian follicle sizes, Hoogland scores of ovarian activity and sex hormone binding globulin
(SHBG), demonstrating that the efficacy of the tested monophasic OC is maintained under
siponimod co-administration.

Anonymized Page 18
Novartis Confidential Page 19
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

1.2.2.2.3 Beta blockers: propranolol


Beta blockers represent a frequently prescribed drug class in the target patient population.
Patients with MS, and in particular patients with SPMS who are more advanced in average age,
have co-morbidities requiring beta blocker therapy (e.g., for hypertension or angina pectoris).
Based on the results of a dedicated study with propranolol (selected as a prototype beta blocker
with both beta 1 and beta 2 receptor-blocking activity; Section 2.2.7.2.3), beta blocker treatment
can be safely initiated in patients receiving stable doses of siponimod; caution should be applied
when siponimod is initiated in patients receiving beta blockers (Section 2.2.7.2.3). The addition
of propranolol on top of siponimod PK/PD steady state had less than additive effects on
lowering HR, in comparison to an additive effect on lowering HR observed with addition of
siponimod on top of propranolol PK/PD steady state. Based on the absence of new or different
safety findings in this study the use of beta blockers (previously prohibited) was permitted in
the Phase 3 study in SPMS patients, using a specific decision algorithm dependent on the
Baseline HR prior to initiation of concomitant therapy (Section 2.2.7.2.3).

1.2.3 Pharmacodynamics

1.2.3.1 Effects on lymphocytes/leukocytes

1.2.3.1.1 Lymphocyte count and leukocyte/lymphocyte subpopulations


Acting as a functional antagonist on S1P1 receptors, the resulting ALC reduction in peripheral
blood due prevention of lymphocyte recirculation from lymphatic tissue to target organs
represents the desired PD effect of siponimod. Across studies, siponimod reduced ALC levels
in a dose-dependent manner (Section 3.4.1.1). Pooled analyses of multiple dose studies in
healthy subjects support the effectiveness of the selected therapeutic dose of 2 mg in reducing
ALC to a therapeutically relevant level, while at the same time avoiding exaggerated ALC
reduction to levels < 0.2 × 109/L. The short T1/2 of siponimod (30 hours) results in a fast
recovery of ALC to normal values and re-establishment of normal immune competence
(Section 1.2.3.1.2). This has particular relevance in the management of patients with infections
and in need of vaccinations, surgical interventions or any other condition requiring fast re-
establishment of immune function ([Study A2130]; Section 2.2.7.2.1). The effect of chronic
treatment with siponimod on ALC in patients with MS is summarized in Section 2.3.1 and
further discussed in the [SCS-Section 3.3]. The effect of siponimod on reduction of peripheral
counts of leukocyte and lymphocyte subsets is described in Section 2.2.2.

1.2.3.1.2 Exposure-lymphocyte relationship


In [CBAF312A Phase 3 PopPKPD], 8 studies in either healthy subjects or RRMS or SPMS
patients were pooled to characterize the dose-exposure-response of ALC. The dose-response
and time profiles of ALC were very well captured by the model, in which the percent decrease
of circulation rate was a saturable maximum inhibitory effect (Imax) function of drug
concentrations (Section 3.4.3). The computed Imax was 79% (95% CI: 78 to 80%), with an
IC50 of 1.9 ng/mL (95% CI: 1.4 to 2.4 ng/mL) in a typical female SPMS patient. After 2-mg
chronic dosing, the median time for ALC to recover to 1.0 × 109/L was predicted to be 5 days
(90% prediction interval (PI): 2 to 12 days) for a typical female SPMS patient. In healthy
subjects where the widest range of dose was tested, the predicted inhibition for circulating ALC

Anonymized Page 19
Novartis Confidential Page 20
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

for 2 mg was 62% and was plateauing at doses ≥ 5 mg. At the therapeutic dose of 2 mg qd, it is
expected that for the majority of subjects siponimod treatment will not result in a reduction of
ALC values below 0.2 × 109/L. Among covariates that were tested (age, body weight, gender,
ethnicity (Chinese and Japanese) and disease status/population type), the only predictors of
lymphocyte response were population type (healthy subjects different from RRMS patients
different from SPMS patients), gender and Japanese ethnicity.

1.2.3.2 Cardiac/hemodynamic effects

1.2.3.2.1 Bradycardia/bradyarrhythmia (atrioventricular blocks and sinus pauses)


In the early single and multiple ascending dose studies of siponimod, a transient,
dose-dependent decrease in HR was observed during the first dosing days that plateaued at
doses ≥ 5 mg (Section 3.4.2.2.1). The daily peak effect occurred at approximately 2 hours post
dose, followed by rapid recovery of HR values to near Baseline levels.
Overall, across the Clinical Pharmacology studies (single doses up to 75 mg, multiple doses up
to 20 mg qd) and studies in PM/DM patients (highest dose 10 mg qd), detected AV blocks were
mostly first degree AV blocks, second degree AV blocks type Mobitz I (Wenckebach) or
occasional second degree AV blocks with 2:1 conduction or 3:2 conduction; incidence rates for
each type of event in siponimod-treated subjects/patients in multiple dose study pooled analyses
were ≤ 5%, ≤ 3.5%, ≤ 2.3% and ≤ 0.2%, respectively (Section 3.4.2.2.2). Second degree AV
blocks resolved spontaneously (i.e., without medical intervention), were mostly asymptomatic
and only in very few cases accompanied by mild AEs, such as headache or dizziness, with
uncertain causal relationship to the AV block occurrences ([Studies A1101, A2101, A2102,
A2105, A2107, A2108, A2110, A2111, A2116, A2118, A2119, A2126 and X2206 and SCS
Appendix 1-Listings 16.2.7-1.1p, 16.2.7-1.2p and 16.2.7-1.3p]). No second degree AV blocks
of Mobitz II or higher degree were observed.
In the pooled analyses most AV blocks and sinus pauses were detected at dose levels above the
therapeutic dose of 2 mg and their incidence was notably higher under non-titrated conditions
compared to DT conditions, demonstrating the efficient mitigation of bradyarrhythmic effects
under DT of siponimod (Section 1.2.3.2.2). AV blocks and sinus pauses predominantly
occurred during resting or nocturnal hours, which represent episodes of increased vagal tone
known to be associated with negative chronotropic and dromotropic effects. Irrespective of the
dosing regimen, detected sinus pauses were mostly asymptomatic and not considered to be of
clinical relevance.

1.2.3.2.2 Dose titration and treatment interruption


A DT [Study A2107] investigated the efficacy of 2 different DT regimens in attenuating the
observed negative chronotropic and bradyarrhythmic effects of siponimod (Section 2.2.3). Both
of the 8- and 9-day DT regimens up to 10 mg of siponimod robustly attenuated the effects on
HR and AV conduction observed under non-titrated conditions. The Fibonacci DT regimen was
employed in subsequent multiple dose Clinical Pharmacology studies and in the Phase 2 and 3
MS studies, where it efficiently mitigated the known effects on HR and AV conduction. The
effectiveness of the DT regimen in attenuating the bradycardic/bradyarrhythmic effects
observed under non-titrated conditions is described across individual studies (Section 2.2 and

Anonymized Page 20
Novartis Confidential Page 21
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Section 2.3) and based on pooled categorical analyses in healthy subjects (Section 3.4.2.2). In
addition, a dedicated [Study A2128] investigating DT in specific CYP2C9 genotypes (CYP2C9
*1*1 extensive metabolizers and CYP2C9 *2*3 and *3*3 poor metabolizers) showed that there
were no clinically-relevant bradyarrhythmic effects of siponimod in both genotypes. Average
of HR and arrhythmia analysis did not show any clinically relevant difference between CYP2C9
genotype groups (Section 2.2.10.3).
A dedicated treatment interruption and re-initiation [Study A2110] identified conditions
requiring siponimod re-titration after variable treatment interruption periods at different dose
levels (Section 2.2.4). In accordance with the results of this study, re-titration was only
warranted in patients missing 4 or more consecutive daily doses during maintenance dosing in
[Study A2304] in SPMS patients (Section 2.3.1.2). In addition, re-titration was required for
patients who missed 1 dose or more during the DT phase.

1.2.3.2.3 Cardiac repolarization/QT effects


A moxifloxacin- and placebo-controlled thorough QT [Study A2118] investigating the effects
of therapeutic (2 mg) and supratherapeutic (10 mg) doses of siponimod on cardiac
repolarization demonstrated no significant direct QT prolonging effect of siponimod and did
not suggest an arrhythmogenic potential related to QT prolongation (Section 2.2.9). The pooled
analyses across studies were consistent with these results (Section 3.4.2.2.2).

1.2.3.2.4 Hemodynamic effects


Overall, the effects of siponimod on systolic blood pressure (SBP) and diastolic blood pressure
(DBP) in the pooled categorical analyses under multiple dose conditions in both healthy
subjects (0.3 to 20 mg, including DT up to 10 mg, maximum treatment duration 38 days) and
PM/DM patients (up to 10 mg, titrated and non-titrated) were not considered to be clinically
relevant (Section 3.4.2.2.3). The effect of chronic treatment with siponimod on blood pressure
(BP) in patients with MS is discussed in the [SCS-Section 4.1.2].
Orthostatic hypotension AEs were infrequently reported and no clear dose dependency or
relation to siponimod treatment could be identified (Section 3.4.2.2.4).

1.2.3.3 Pulmonary effects


Analyses of the pulmonary function tests (PFTs) in studies in healthy subjects and PM/DM
patients indicated minor reductions in forced expiratory volume in 1 second (FEV1) during
siponimod treatment, but without signs of clinically significant bronchoconstriction at single
doses up to 75 mg and multiple doses up to 20 mg qd over 28 days (Section 3.4.2.3).

1.2.3.4 Other pharmacodynamic effects

1.2.3.4.1 Hepatotoxicity/liver transaminase elevations


Across Clinical Pharmacology studies in healthy subjects and PM/DM patient studies,
occurrences of elevated liver enzymes (e.g., alanine aminotransferase (ALT), aspartate
aminotransferase (AST) and gamma-glutamyl transferase (GGT)) were mostly asymptomatic
and not exceeding CTCAE Grade 3. There were no cases of Hy’s law detected. In the pooled
categorical analyses of select liver function test (LFT) parameters (ALT, AST, GGT, alkaline

Anonymized Page 21
Novartis Confidential Page 22
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

phosphatase (AP) and bilirubin), elevations were mainly observed for ALT, AST and GGT
enzymes (Section 3.4.2.4.1). In healthy subjects these elevations were mainly increases < 3-fold
above the upper limit of normal (ULN) and mostly occurred at doses > 2 mg. Elevations in liver
enzymes in MS patients are described in [SCS–Section 3.4].

1.2.3.4.2 Suicidal ideation and behavior


Safety assessments in healthy subjects and PM/DM patients did not reveal any indication for
an increased risk for suicidal ideation or behavior with siponimod treatment (Section 3.4.2.4.2).

1.2.3.4.3 Infections
Considering the primary PD effect of siponimod on lymphocytes, potential translation to an
increased rate of infections was explored. Pooled analyses of AEs of healthy subjects revealed
a slightly higher incidence of upper respiratory tract infections in siponimod-treated subjects
(up to 2.0%) compared to placebo subjects (up to 1.0%); for uptitration to the therapeutic dose
of 2 mg qd, the incidence was 1.7% (Section 3.4.2.4.3). Other infection-related AEs reported
in > 1 subject were urinary tract infections (incidence rates up to 0.3% in siponimod-treated
healthy subjects, compared to no events for placebo treatment).

1.2.3.4.4 Abuse and dependence potential (8-factor drug abuse liability assessment)
Chemistry, nonclinical and clinical data with siponimod did not indicate any signals of abuse,
misuse or dependence potential in animals or humans or demonstrate any potential
pharmacological similarities to existing drugs of abuse or subjective effects that may be related
to drug abuse, such as reinforcing, mood-elevating, sedative, stimulant or hallucinogenic effects
(Section 3.4.2.4.4). Therefore, it can be concluded that siponimod has no abuse or dependence
potential and it is not expected to be subject to abuse, misuse or diversion in the community, or
result in harm to the public health as a result of abuse, misuse or diversion.

2 Summary of Results of Individual Studies


2.1 In vitro, ex vivo, and in silico studies using human biomaterials
The results of 33 in vitro and ex vivo studies of siponimod, M3, M5 and M17 using human
biomaterials are summarized in the across-study summaries throughout Section 3.
These studies are listed in Table 5-2, which includes details such as study objectives, matrix of
human biomaterial, and concentrations of siponimod. This section provides an overview of the
sections in which the results of in vitro and ex vivo studies are described.

2.1.1 Plasma protein binding


The results of in vitro and ex vivo plasma protein binding of siponimod, M3 and M5 in
[DMPK R1300334], [DMPK RCBAF312A2129-01] and [DMPK R1400021] are discussed in
Section 3.1.6.

Anonymized Page 22
Novartis Confidential Page 23
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.1.2 Hepatic metabolism and drug interaction studies

Metabolism by CYP450 enzymes


The results of the metabolism of siponimod in human liver microsomes (HLMs) and
21 recombinant human CYPs (rhCYPs), conducted at siponimod concentrations considerably
higher than systemic concentrations achieved at a therapeutic dose of 2 mg qd (Cmax,ss,u ≈ 0.6
nM), in [DMPK R0500432] are discussed in Section 3.1.7 (metabolism) and Section 3.3.1.2
(assessment of DDI risk concerning hepatic metabolism).

CYP450 induction and inhibition


CYP450 induction and inhibition by siponimod and its main metabolites was assessed in
8 studies: siponimod ([DMPK R0500496], [DMPK R0500497-01], [DMPK R1200710] and
[DMPK R1300188]), M3 ([DMPK R1300932] and [DMPK R1300933]) and M17
([DMPK R1500795] and [DMPK R1500796]). Potential clinical implications for DDIs from
these in vitro results for siponimod, M3 and M17 acting as a perpetrator drug are discussed in
Section 3.3.1.1.1 (details in Table 5-5).

Drug interaction studies


The results of PBPK modeling conducted in [DMPK 1600759-01] to predict the DDI potential
of siponimod as a substrate in the presence of typical CYP2C9/3A inhibitors and inducers for
the 6 different CYP2C9 genotypes are discussed in Section 3.3.1.2.3 and Section 3.3.1.2.4.
Hepatocyte investigations to better characterize the inhibition/induction properties of
itraconazole (strong CYP3A inhibitor) in vitro in [DMPK 1701078] are also summarized in
Section 3.3.1.2.3.

2.1.3 Studies using other human biomaterials

Cellular uptake
The results of [DMPK R0800531] and [DMPK R1300921], for which the mechanism of
intestinal uptake of siponimod was investigated in Caco-2 cell monolayers in the presence and
absence of inhibitors of efflux transporters, are discussed in Section 3.1.1 (absorption) and
Section 3.3.1.2.1 (details in Table 5-6) (assessment of DDI risk concerning transporters
involved in absorption).

Transporter inhibition
Efflux, uptake and SLC transporter inhibition by siponimod, M3 and M17 was assessed in
19 studies (detailed below). Potential clinical implications for DDIs from these in vitro results
for siponimod, M3 and M17 acting as a perpetrator drug are discussed in Section 3.3.1.1.1
(details in Table 5-5).
For efflux transporters, inhibition by siponimod is described in [DMPK R1200722] (P-gp,
BCRP), [DMPK R1300847] (BSEP) and [DMPK R1300849] (MRP2). Inhibition by M3 is
described in [DMPK R1300852] (BSEP) and [DMPK R1300853] (P-gp, BCRP). Inhibition by
M17 is described in [DMPK R1500825] (P-gp, BCRP) and [DMPK R1500826] (BSEP, MRP2).

Anonymized Page 23
Novartis Confidential Page 24
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

For uptake transporters, inhibition by siponimod is described in [DMPK R1200723] (OATP1B1,


OATP1B3), [DMPK R1200724] (OAT1, OAT3) and [DMPK R1200725] (OCT1, OCT2).
Inhibition by M3 is described in [DMPK R1300855] (OCT1, OCT2), [DMPK R1300856]
(OAT1, OAT3), and [DMPK R1300857] (OATP1B1, OATP1B3). Inhibition by M17 is
described in [DMPK R1500827] (OATP1B1, OATP1B3), [DMPK R1500828] (OCT1, OCT2)
and [DMPK R1500829] (OAT1, OAT3).
For SLC transporters, inhibition of MATE1 and MATE2K by siponimod, M3 and M17 is
described in [DMPK R1300848], [DMPK R1300854] and [DMPK R1500830], respectively.

2.2 Studies in healthy subjects


The results of the 20 Clinical Pharmacology studies are summarized individually in this section.
All Clinical Pharmacology studies are listed in Table 5-1, which includes details such as study
objectives, design, doses and formulations of siponimod used and number of subjects for each
treatment group.

2.2.1 Single ascending dose study


The first-in-human [Study A2101] was designed as a 2-part, randomized, double-blind,
placebo-controlled trial to explore the safety, tolerability, PK and PD of ascending single doses
of oral siponimod (0.1 to 75 mg) including fed and fasted conditions in healthy subjects. Single
doses of siponimod were safe and well tolerated but associated with acute, negative,
chronotropic effects. Reported AEs were mostly transient and either mild or moderate in
severity and there were no serious AEs (SAEs).
The MTD was determined to be 25 mg based upon symptomatic bradycardia observed at the
75-mg dose (37.5-fold the highest therapeutic dose). A dose-dependent decrease in mean
ventricular HR was observed over the first 24 hours post dose (0.1 to 75 mg), starting to plateau
at the 5-mg dose. Decreases were transient and peaked approximately 2 hours post dose. No
events of bradycardia were reported at doses ≤ 1 mg. Corresponding increases in PR and RR
intervals were also observed, with peak effects observed at the 2.5- and 5-mg dose levels
between 4 and 8 hours post dose. These increases in PR interval declined in magnitude at doses
≥ 10 mg and no clear dose-response relationship could be detected. Occasional asymptomatic
first and second degree AV blocks on Day 1 were observed in all dose groups (including
placebo). Changes in mean arterial BP (MABP), SBP or DBP were not clinically significant or
statistically different between dose groups. Siponimod did not show clinically significant
effects on pulmonary function.
AUC increased dose proportionally in the investigated dose range. Median Tmax ranged
between 3 and 6 hours and an apparent terminal T1/2 of 27 to 57 hours was determined.
A delayed median Tmax was observed after food intake (8.00 versus 3.00 hours); however,
mean AUC and Cmax were similar under both fasted and fed conditions (tested at 5-mg dose).
Subjects with heterozygous CYP2C9*3 genotype tended to have higher siponimod exposure
compared to those who did not carry CYP2C9*3, which was more pronounced at higher doses.
A dose-dependent decline in ALC was observed beginning at the 0.3-mg dose
([Study A2101-Table 11-2 and Figure 11-1]). At doses above 2.5 mg the area under the effect
curve (AUEC) from time zero to 24 hours (AUEC0-24) for mean ALC values started to show

Anonymized Page 24
Novartis Confidential Page 25
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

a flooring effect. Only small increments of further decreases in the mean ALC AUEC0-24 were
observed at doses above 5 mg (i.e., ≥10 mg). This flooring effect is consistent with observations
for the HR effects (discussed above). An ALC reduction of at least 80% was reached at the
10-mg dose or higher. Mean values for ALC changed by ≥ -1.0 × 109/L for all doses ≥ 5 mg.
Recovery was not observed until the End-of-Study (EOS) Visit for subjects receiving the
highest dose (75 mg). The nadirs of the decline in ALC (median TEmax0-24h values) ranged
from 3.00 to 7.00 hours with the exception of doses ≥ 10 mg.
After single doses above 1 mg the mean ALC did not show a clear trend for recovery at 24 hours
post dose. At 48 hours post dose mean ALC returned to normal values after single doses of 2.5
mg and showed a clear trend of recovery after single doses of 5 mg, 10 mg and 17.5 mg. Based
on the observed ALC recovery dynamics and supported by an apparent terminal T1/2 of 27 to
57 hours, a qd dosing regimen was considered appropriate to maintain reduced ALC levels in a
multiple dose setting.

2.2.2 Multiple ascending dose studies


[Study A2102] was a single-center, randomized, double-blind, placebo-controlled and
time-lagged trial designed to evaluate the safety and tolerability of ascending multiple doses of
siponimod (0.3 to 20 mg, i.e. up to 10-fold the highest therapeutic dose) in healthy subjects.
The study as conducted did not meet the study objectives. Several healthy subjects in the 10- and
20-mg cohorts had higher than expected siponimod plasma levels. Investigation of the route
cause revealed that the observed exposure differences were a result of dosing irregularities at
the site and a decision was made to repeat the study ([Study A2105]). Some subjects were likely
treated with a dose of 200 mg daily for a duration of 3 to 4 days. No SAEs or major safety issues
were reported. There were transient elevations in LFT values. The maximum recorded
individual Cmax and AUC0-24h values (1740 ng/mL and 27400 ng*h/mL, respectively) did
not exceed the cap values set by the FDA and incorporated into the protocol (2700 ng/mL and
35000 h*ng/mL, respectively).
As a repeat of the above MAD study, [Study A2105] was conducted using the same study design
and objectives. Siponimod was well tolerated, demonstrating a favorable safety profile for both
standard safety and special safety assessments including 24-hour ambulatory blood pressure
monitoring (ABPM), telemetry, Holter monitoring, pulmonary function tests, Snellen scores,
ocular coherence tomography and neurological exams. Day 1 PK parameters were comparable
to those in the SAD study. Bradyarrhythmic events were all benign, transient and did not require
medical intervention. The highest investigated multiple dose of 20 mg over 28 days was well
tolerated, and therefore no multiple dose MTD was formally established.
Transient, negative chronotropic effects of siponimod were observed on Day 1. The maximum
dose-dependent decrease in mean HR was observed approximately 2 hours post dose for
siponimod 0.3 to 20-mg doses; subsequently, the HR started to recover and all subjects had a
mean HR ≥ 60 bpm approximately 5 hours post dose. The maximum effect reached a plateau at
siponimod 10-mg dose on Day 1. Negative chronotropic/dromotropic effects of siponimod,
including bradycardia and occasional asymptomatic first and second degree AV blocks (Mobitz
I) were observed under siponimod treatment, and to a lesser extent in the placebo group. All
were benign, transient and did not require intervention. There were no notable effects of
siponimod on SBP and DBP. Four (4) of the 6 subjects that had elevated ALT reported as an

Anonymized Page 25
Novartis Confidential Page 26
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

AE were in the siponimod 20-mg dose group (the other 2 subjects in the 0.3 mg siponimod and
placebo groups), suggesting a dose effect relationship.
Siponimod steady-state concentrations were achieved in all 5 cohorts following approximately
6 days of multiple dosing for most of the subjects, and comparable PK parameters were defined
on Days 7 and 28. The mean Racc ranged between 1.88 and 2.72 on Day 28, with similar values
determined on Day 7 (between 1.82 and 2.42). The effective T1/2, estimated after multiple
administrations based on drug accumulation at steady state (Boxenbaum and Battle 1995),
ranged between 22 and 36 hours and was comparable to the apparent terminal T1/2 determined
in the SAD study. Systemic exposure was not to exceed the cap values set by the FDA (Cmax
of 2700 ng/mL or an AUC0-24h of 35000 h*ng/mL). The maximum individual Cmax/Cmax,ss
and AUC0-24h/AUCtau values were well below this limit (498 ng/mL and 9235 h*ng/mL,
respectively).
A dose-dependent reduction of mean ALC, when compared to Day -1, was observed for all
doses of siponimod ≥ 1 mg ([Study A2105-Figure 11-7]). Dose-dependent effects were
observed for ALC %Emax0-12h, Emax0-12h and AUEC0-12h. The means of Emax0-12h on
Day 1 and Day 28, as adjusted to Day -1, ranged from 1.71 to 0.62 × 109/L and 1.09 to
0.37 × 109/L, respectively, across dose groups (compared to 1.70 and 1.71 × 109/L for placebo).
For all dosing groups, the maximal reduction in peripheral lymphocyte counts on Day 1 was
observed at approximately 4 to 6 hours post dose and was maintained throughout the 28-day
treatment. At the EOS Visit (14 to 21 days post treatment completion), recovery to Baseline of
ALC was observed for subjects receiving 0.3 to 10 mg siponimod and incompletely at the dose
of 20 mg.
The effect of siponimod on leukocyte/lymphocyte subsets (T cells, monocytes, natural killer
(NK) cells, dendritic cells (DCs) and B cells) was also assessed. Relative levels of both T and
B lymphocytes were reduced in a dose-dependent manner. The 2 major T cell subpopulations,
CD4+ and CD8+ cells, were affected to different degrees by siponimod
([Study A2105-Figure 11-8]). While the CD4+ cells showed a clear dose-dependent reduction
among all T lymphocytes, the relative proportion of CD8+ T cells was increasing. The most
pronounced increase was observed for siponimod 2.5 and 10 mg, on Days 14 and 28, when
compared to the placebo group. Central memory (CM), peripheral effector memory (PEM) and
naïve (N) subtypes of both CD4+ and CD8+ T cells responded differently to siponimod. CD4N
T cells were reduced for siponimod 10 and 20 mg. CD8N T cells showed a pronounced
reduction for siponimod 2.5 mg and an even stronger response at the highest doses. In contrast,
relative proportions of CD4PEM and CD8PEM T cells were increasing in the siponimod
10- and 20-mg dose groups and in the siponimod 2.5-mg dose group for CD8PEM T cells. PEM
T cells are present primarily in peripheral tissues and are endowed with immediate effector
functions (Sallusto and Lanzavecchia 2009). The observed preferential effect on naïve T cells and
the less pronounced effects on PEM T cells may indicate a preservation of the relative capacity of the
immune system to mediate recall responses to known antigens or to vaccinations. Further detail can
be found in [Study A2105-Section 11.4.5.2].

2.2.3 Dose titration study


[Study A2107] investigated the efficacy of 2 different DT regimens (DT1, DT2) in attenuating
the observed negative chronotropic effect of siponimod. Subjects were randomized equally to

Anonymized Page 26
Novartis Confidential Page 27
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

the DT1 or DT2 groups (both 0.25 to 10 mg), a fixed-dose group (10 mg qd), or a placebo group
(Table 2-1). The starting doses of both tested DT schemes were selected based on the results of
[Study A2101] showing that a single dose of 0.3 mg was devoid of any relevant
bradyarrhythmic effect. DT1 applied an arithmetic algorithm starting with stable dosing over
the first 3 days and subsequent daily dose increases using a factor of 2 for dose increments,
while DT2 used incremental dose increases based on the Fibonacci algorithm. The arithmetic
DT regimen is based mainly on PK-based considerations, while the Fibonacci-based dose
titration was designed mainly based on empirical observations of adaptive processes in
biological systems (i.e., PD-based considerations).

Table 2-1 Study A2107 Dose Titration Schedule


Study Day Day Day Day Day Day Day Day Day Day Day Day
Period 1 2 3 4 5 6 7 8 9 10 11 12
DT regimen 0.25 0.25 0.25 0.5 1 2 4 8 10 10 10 10 mg
#1 mg mg mg mg mg mg mg mg mg mg mg
DT regimen 0.25 0.25 0.5 0.75 1.25 2 3 5 8 10 10 10 mg
#2 mg mg mg mg mg mg mg mg mg mg mg
Placebo Pbo Pbo Pbo Pbo Pbo Pbo Pbo Pbo Pbo Pbo Pbo Pbo
regimen
Non-titration 10 10 10 10 10 10 10 10 10 10 10 10 mg
regimen mg mg mg mg mg mg mg mg mg mg mg
Pbo = placebo.
Source: [Study A2107-Figure 9-2]

The primary PD evaluations of this study were the daily chronotropic effects of the 2 DT
regimens compared to the chronotropic effect of the 10 mg qd regimen and of placebo over
12 days, and the frequency of AV blocks and sinus pauses (RR > 2 seconds) over the course of
each DT regimen using daily 24-hour, 12-lead digital Holter electrocardiogram (ECG)
recording over the entire dosing period. Overall, DT2 was identified as the most robust DT at
mitigating negative chronotropic effects and was used in subsequent multiple dose studies and
in Phase 2 and 3 studies.
Both DT regimens effectively attenuated the initial bradycardia observed on Day 1 of treatment
with siponimod 10 mg. HR in the non-titration regimen showed considerable separation from
placebo throughout the study (Figure 2-1). There was no statistically significant reduction in
HR versus placebo on Day 1 in either DT regimen. On Days 3 to 7, subjects in DT1 and DT2
experienced minor reductions in HR. There were no subjects in the DT groups that displayed
significant bradycardia during the trial, in contrast to the 3 subjects in the 10 mg group that had
bradycardia ranging from 34 to 54 bpm on Day 1. The 2 DT regimens only induced a decrease
of 4 bpm over 7 days; this effect on HR is not considered to be clinically relevant. By Day 9,
HR in both DT regimens was comparable to placebo. This effect was maintained until end of
treatment on Day 12.

Anonymized Page 27
Novartis Confidential Page 28
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 2-1 Mean daily minimum heart rate after once daily dosing administered
over 12 days as a fixed dose or uptitration to 10 mg

PBO = placebo.
Source: [Study A2107-Figure 11-3]

There was no difference between treatment groups in the number of subjects with
supraventricular ectopy (SVE) and ventricular ectopy (VE) ([Study A2107-Table 11-3]).
With regards to their ability to attenuate bradyarrhythmic effects of siponimod both DT
regimens were similar, with slightly more favorable data seen with the Fibonacci algorithm
(DT2). The results of the standard analysis of the Holter data did not show any significant
difference between the 2 DT schemes. In DT2 only 1 subject experienced at least 1 sinus pause
compared to DT1 and the 10 mg treatment groups, where 4 subjects experienced at least 1 sinus
pause ([Study A2107-Table 11-4]). No sinus pauses were observed in the placebo group.
Second degree AV blocks (Mobitz I) were reported in 4 subjects (2 subjects in DT1 and 1
subject each in the DT2 and placebo groups) and a 2:1 AV block was noted in 1 subject (10 mg
non-titrated group). Based on the placebo-like occurrence of AV blocks and sinus pauses in
DT2, this Fibonacci regimen was implemented in the Phase 2 and 3 studies in MS patients as
well as in the Phase 2 studies in PM/DM patients where it successfully attenuated the known
bradyarrhythmic effects of siponimod.

Anonymized Page 28
Novartis Confidential Page 29
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.2.4 Treatment interruption and re-initiation


[Study A2110] investigated the impact of re-initiation of siponimod treatment on HR and
cardiac rhythm after variable periods of treatment discontinuation at different dose levels.
The study was aimed at identifying conditions under which siponimod can be safely re-initiated
at the dose of drug at the time of drug discontinuation and conditions requiring re-titration.
Four (4) fixed-dose levels (0.5, 1, 2 and 4.0 mg) and placebo were evaluated in combination
with 5 drug discontinuation periods (48, 72, 96, 120 and 192 hours) corresponding to 1, 2, 3, 4
and 7 missed doses ([Study A2110-Figure 9-1]). In each period, subjects went through
3 treatment phases, sequentially: 10-day drug treatment phase (A), drug discontinuation phase
(B) and 1-day drug re-initiation phase (C) ([Study A2110-Figure 9-3]). Overall, the results of
this study suggested that treatment re-initiation at the discontinued therapeutic dose level
without the need for re-titration is possible under certain conditions. Consequently, in the
Phase 3 [Study A2304] in SPMS patients treated with a therapeutic target maintenance dose of
2 mg preceded by a 5-day DT (Section 2.3.1), a re-titration was only warranted for patients
re-initiating siponimod treatment after missing 4 or more consecutive daily doses during
maintenance dosing (i.e., no re-titration was needed for up to 3 consecutive missed daily doses).
The magnitude of the negative chronotropic effects at siponimod re-initiation appeared
to be dependent on both the dose and the duration of treatment discontinuation ([Study A2110-
Figures 11-3 and 11-4]). The most pronounced HR decreases (> 30 bpm) were observed after
drug discontinuation periods of 120 hours or more (1 mg/120 hours, 2 mg/120 hours and
2 mg/192 hours). Re-challenge at the highest investigated dose (4 mg) after the longest
investigated discontinuation period of 192 hours (7 consecutive missed daily doses) induced
the strongest decrease in pooled placebo-adjusted HR of a magnitude similar to that at treatment
initiation with an adjusted mean difference (standard error (SE)) of 14.53 (1.664) bpm.
Irrespective of the dose, 96 hours and 120 hours treatment gaps mimicking 3 or 4 missed doses
respectively induced a mean maximum decrease in pooled placebo-adjusted HR of < 10 bpm.
The upper bounds of the 2-sided 90% CI for the mean maximum HR effect at nearly all
conditions up to 96 hours were below 10 bpm, except for the upper bounds at the conditions
1 mg/72 hours (10.230 bpm) and 2 mg/96 hours (12.069 bpm). At the 120-hour conditions, the
upper bounds of the 2-sided 90% CI ranged between 10.098 bpm (0.5 mg) and 12.096 bpm (2
mg). Except for a single outlier of HR decrease under the condition placebo/96 hours, all other
outliers of HR decrease were detected at re-challenge after 120 and 192 hours of drug
discontinuation.
Sinus pauses in 4 subjects were asymptomatic and not considered to be of clinical relevance.
No trend for the incidence or duration of sinus pauses could be detected in relation to dose and
number of missed doses. All observed first and second degree AV blocks (Mobitz I in 2 subjects
and 2:1 AV conduction in 1 subject) were asymptomatic and not considered to be of clinical
relevance.

Anonymized Page 29
Novartis Confidential Page 30
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.2.5 Human ADME study


In [Study A2104] of [14C]-siponimod (10 mg; N = 4), the extent of oral absorption was
estimated to be > 70% of the radioactive dose. The 3 main biotransformation pathways were:
1) phase I C-hydroxylations (M5, M6 and M7), 2) phase I cleavage/hydrolysis at the oxime
ether bound (M1, M2) and further reduction (M8), and 3) phase II sulfation (M4a, M4b and
M4c) and glucuronidation (M3 and M12) of hydroxylated metabolites ([Study A2104d]).
The most abundant radioactive component in plasma was unchanged siponimod (57.1% of the
radioactivity AUC0-120h). M3 was the main metabolite (18.4% of the radioactivity
AUC0-120h, 27.6% of the exposure to siponimod). M5, M6, M7, P29.6 and P30.5 each
accounted for 1.5% to 3.7% of the radioactivity AUC0-120h (11.43% combined).
The apparent distribution volume of siponimod was moderate (mean 291 L), the apparent
clearance of siponimod was low (mean CL/F = 4.0 L/h). Siponimod and its metabolites were
mainly confined with the plasma compartment. Siponimod was eliminated from the systemic
circulation mainly due to metabolism and subsequent biliary/fecal excretion. The apparent
terminal T1/2 of total radiolabeled components and siponimod in plasma were 171 and
56.6 hours, respectively. For M3, M5, M6 and M7, the apparent terminal T1/2 ranged between
29.3 and 35.2 hours. Renal excretion of radioactivity was mainly in the form of M3 (2.1% of
the radioactive dose). Excretion of radioactivity was close to complete after 13 days (88.2% to
93.3% of the dose).
The [14C]-radioactivity levels in final drug product and human whole blood, plasma, urine and
feces samples by means of validated liquid scintillation counting methods are discussed in
[DMPK RCBAF312A2104a]. Metabolism is summarized in [DMPK RCBAF312A2104b].
Investigations of the isomers M4a, M4b and M4c, previously summarized as M4 in
(DMPK RCBAF312A2104b), are summarized in [DMPK RCBAF312A2104d].

2.2.6 Bioequivalence and relative bioavailability


In [Study A2111] of siponimod (0.25 and 4 mg), FMI and MF tablet formulations were
compared. The median Tmax for both formulations under fasting conditions was 4 hours for
both doses. Bioequivalence between MF and FMI was demonstrated for Cmax, AUClast and
AUCinf for both doses. Similarly, FMI fasted and FMI fed fulfilled the bioequivalence criteria
at both doses. Food intake had no effect on the systemic exposure to siponimod (Cmax and
AUC). A delay in Tmax was observed for fed versus fasted conditions (6 to 7 hours across
both doses, versus 4 hours) but was not considered to be clinically relevant. The
bioequivalence of the 2 formulations and the food effect results of this study are discussed in
detail in [SBP-Section 3.3] and [SBP-Section 3.5], respectively.
[Study A2119] assessed the potential of MR formulations F10 and F16 of siponimod (4 mg)
(which release siponimod from their matrix over 10 or 16 hours, respectively) to attenuate
negative chronotropic effects. The incidence of changes in ECG morphology, conduction and
HR were in line with previous clinical experience for the IR formulation and similar for the
2 MR formulations, suggesting that use of a MR formulation does not attenuate the negative
chronotropic effects observed with the IR formulation. Consistent with the MR formulation
release profiles, the median Tmax was delayed 3 to 4 hours with F10 (6.98 hours) and F16
(8.0 hours) compared to IR (3.98 hours). Compared to IR the Cmax of F16 and F10 were 55%

Anonymized Page 30
Novartis Confidential Page 31
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

and 16% lower and the geometric mean AUC0-inf of F10 and F16 were 16% and 51%.
The relative bioavailability of the 2 MR formulations is discussed in [SBP-Section 3.2.2].

2.2.7 Drug-drug interaction studies

2.2.7.1 Pharmacokinetic evaluation


CYP2C9, the major metabolizing enzyme for siponimod, is polymorphic and the CYP2C9
genotype has been shown to have a significant impact on siponimod metabolism (Section 3.2.5).
The following 3 studies evaluated the effects of CYP2C9/CYP3A4 inducers and inhibitors on
siponimod PK, the results of which are detailed in Section 3.3.1.2.2.
In the DDI [Study A2125] with rifampin (a moderate CYP2C9 and strong CYP3A4 inducer),
CYP2C9*1*1 subjects received siponimod uptitrated to 2 mg qd in Period 1 (12 days) and
rifampin 600 mg qd and siponimod 2 mg qd in Period 2 (12 days). Overall, co-administration
of rifampin decreased exposure of siponimod (Cmax, AUC) by 45% to 57% but not M3 or M5.
During combination with rifampin in Period 2 the mean ALC increased to levels consistent with
those in previous studies in healthy subjects at the 1-mg dose level, with levels below
1.0 × 109/L until the end of combination treatment. Mean ALC recovered to near Baseline levels
by the EOS Visit (approximately 7 days after the last dose), consistent with ALC recovery
patterns in previous studies at the same dose level.
In the DDI [Study A2108] with fluconazole (a moderate CYP2C9/CYP3A4 inhibitor),
CYP2C9*1*1 subjects received a single 4-mg dose of siponimod on Day 1 in Period 1 (14 days);
in Period 2 (20 days), subjects received fluconazole 200 mg from Day 1 to 19 and a single 4-mg
dose of siponimod on Day 3. In presence of fluconazole, siponimod AUC was increased by 2-
fold and apparent terminal T1/2 was prolonged by 50%. Observed chronotropic effects were
minimal and resolved spontaneously.
In the DDI [Study A2124] with itraconazole (a strong CYP3A4 inhibitor), CYP2C9*1*2 and
*1*3 subjects received a single 0.25-mg dose of siponimod on Day 1 in Period 1 (14 days) and
itraconazole 100 mg bid (twice a day) in Period 2 (Days 15 to 18). In Period 3, subjects received
a single 0.25-mg dose of siponimod on Day 19 and 100 mg itraconazole bid until Day 31
(CYP2C9*1*2) or Day 35 (CYP2C9*1*3). Varying degrees of decreases in AUC observed in
CYP2C9*1*2 (by about 10%) and *1*3 (by about 25%) subjects suggest a possible influence
of CYP2C9 genotype on the effect of itraconazole co-administration (Section 3.2.5.2).

2.2.7.2 Pharmacodynamic and pharmacokinetic evaluation


The combined PD or PK effects of siponimod co-administration with medications frequently
used in the target patient population were investigated in 3 DDI studies, discussed below.

Anonymized Page 31
Novartis Confidential Page 32
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.2.7.2.1 PD DDI with T cell dependent and T cell independent vaccines


To support clinical recommendations for the timing of vaccinations or other medical
interventions requiring an intact immune system (e.g., surgical procedures), [Study A2130]
evaluated the modulation of the immune response to T cell-dependent and T cell-independent
antigen stimuli when co-administered as interrupted, preceding or concomitant treatment of
siponimod (2-mg dose, preceded by 5-day DT; [Study A2130-Figure 9-1]).
A quadrivalent influenza vaccine recommended for use in North America during the 2014/2015
season was used as a T cell-dependent vaccine, containing the following 4 antigens:
Influenza-A/California/7/2009 (A-Cal), Influenza-A/Texas/50/2012 (A-Tex),
Influenza-B/Brisbane/60/2008 (B-Bri) and Influenza-B/Massachusetts/2/2012 (B-Mas).
Non-inferior responder rates with respect to each of the 4 antigens were determined for the
preceding treatment group (siponimod treatment paused from 7 days prior to 28 days after
vaccination). For the interrupted (siponimod treatment paused from 10 days prior to 14 days
after vaccination) and concomitant treatment groups, treatment with comedication showed only
modest impairment of influenza vaccination efficacy, with responder rates approximately 15%
to 30% lower, respectively, than on placebo (except for B-Bri with similar responder rates to
placebo). Post-vaccine treatment pauses between 2 and 4 weeks (not investigated in this study)
may be associated with no or only modest impairment of influenza vaccination efficacy, as
observed for interrupted and concomitant siponimod treatment. The proportions of responders
at 4 weeks after vaccination by treatment group for each antigen are presented in
[Study A2130-Table 11-7]. These results demonstrated that the immune response and efficacy
of the quadrivalent seasonal influenza vaccine was not compromised after short treatment
interruption of 1 week prior until 4 weeks after vaccination as tested in the preceding treatment
group.
After PPV-23 (Pneumovax®) vaccination (T cell-independent vaccine), non-inferior responder
rates were determined for each of the 3 treatment regimens with the PPV-23 vaccine (p < 0.001).
Comparisons of responder proportions 4 weeks after vaccination (Day 49) based on IgG
concentration data is presented by treatment group in [Study A2130-Table 11-8]. The PPV-23
vaccine can be co-administered with siponimod without compromising immune response and
vaccination efficacy.
As a conservative approach, siponimod treatment discontinuation 1 week prior to until 4 weeks
after a planned vaccination is recommended. The proposed short siponimod treatment
interruption of 1 week prior to the vaccination is supported by the short T1/2 of siponimod
(approximately 30 hours). This allows for a rapid PK washout (on average within a week of
treatment discontinuation) and therefore a fast re-establishment of normal ALC values in
peripheral blood and of normal immune function that allows the immune system to mount a
vaccination response leading to sufficient seroprotection. The continued treatment pause of
4 weeks after an influenza vaccination is considered appropriate based on PD considerations of
the time required for an intact immune system to mount a sufficient serological response to a T
cell-dependent vaccine. Under the same considerations, the use of live attenuated vaccines
should be avoided during siponimod treatment and for up to 4 weeks after treatment with
siponimod.

Anonymized Page 32
Novartis Confidential Page 33
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.2.7.2.2 PK and PD DDI with a monophasic oral contraceptive


Although siponimod has not been described as an inducer or an inhibitor of CYP3A4 in vitro,
[Study A2121] was conducted to confirm the lack of PK and PD DDI between siponimod
(uptitration to 4 mg; 4 mg qd in Period 2) and a common monophasic OC (30 µg of EE and
150 µg of LVG: qd administration in Periods 1 and 2) in healthy female subjects.
Progesterone levels remained below 5 nmol/L, indicating that no ovulation occurred during the
combination treatment. All ovarian follicle sizes remained below 10 mm on Day 21 of the
combination treatment, indicating their lack of activity. Hoogland scores calculated on Day 21
to differentiate between various stages of ovarian activity demonstrated ‘no activity’ for all the
subjects in the combination treatment. The minor elevation in SHBG levels during the
combination treatment was not considered to be clinically significant.
Siponimod had no effect on the PK of EE (AUCtau and the Cmax,ss levels for EE were
comparable when given OC alone and when combined with siponimod). When the OC was
given alone, the AUCtau of LVG was 28% lower and the Cmax,ss levels were 18% lower than
the corresponding levels that were reached when OC was combined with siponimod. While
there are no historical data for steady state siponimod PK at a dose of 4 mg, considering the
linear PK of siponimod, the steady state PK of siponimod is expected to be similar when given
with OC as compared to siponimod alone.
Together the PK and PD (estradiol, FSH, LH, follicle size, Hoogland score, SHBG) results
demonstrate that the efficacy of the OC is maintained with siponimod co-administration (i.e., no
clinically significant changes in PK/PD in the combination treatment in comparison to OC pill
alone).

2.2.7.2.3 PD and PK DDI with propranolol


Beta blockers are commonly used for treatment of cardiovascular disorders (hypertension,
angina pectoris), migraine and other disorders. Patients with MS, and in particular patients with
SPMS who are more advanced in average age, have co-morbidities requiring beta blocker
therapy. Typical PD effects of beta blockers are negative chronotropism and
bronchoconstriction by blockade of beta adrenergic receptors.
In the dedicated PD/safety [Study A2116], propranolol was selected as a prototype beta blocker
with both beta 1 and beta 2 receptor-blocking activity, allowing for evaluation of both cardiac
(via beta 1 antagonism) and pulmonary (via beta 2 antagonism) DDI effects in combination
with siponimod. The negative chronotropic effect of co-administration of siponimod
(uptitration, over 5 days, to 2 mg qd) and propranolol (80 mg long acting (LA) preparation) was
evaluated using 4 treatment groups: propranolol added on top of siponimod PK/PD steady state
(Group A), siponimod added on top of propranolol PK/PD steady state (Group B), placebo
(Group C), and propranolol alone (Group D). Groups A and B were designed to mimic typical
clinical scenarios in which propranolol treatment would be initiated in patients already treated
with siponimod and vice versa.
The addition of propranolol on top of siponimod PK/PD steady state had a less pronounced
negative chronotropic effect (Emax HR effect less than additive) in comparison to addition of
siponimod on top of propranolol PK/PD steady state (additive Emax HR effect).
The combination treatment of siponimod and propranolol (Groups A and B) led to a mean Emax

Anonymized Page 33
Novartis Confidential Page 34
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

HR decrease of 6.21 bpm compared to propranolol alone, while the mean minimum hourly
average HR remained above 50 bpm.
The HR data for each treatment group over 24 hours is plotted in Figure 2-2. The combination
treatment at steady state (Groups A and B combined) showed an additional decrease of mean
Emax HR by 6.21 bpm (95% CI: 2.32, 10.11; p = 0.002) when compared to propranolol alone
(Group D). Propranolol treatment introduced on top of siponimod steady state (Group A) led to
a less than additive additional mean Emax HR decrease of 5.04 bpm (95% CI: 0.52, 9.56;
p = 0.0292) and siponimod treatment introduced on top of propranolol steady state (Group B)
led to additional mean Emax HR decrease of 7.39 bpm (95% CI: 2.87, 11.90; p = 0.0016),
in comparison to propranolol alone at steady state. The mean (range) Emax HR decreases from
Baseline were 15.44 (6.4, 23.6) and 7.77 (2.0, 29.9) bpm with propranolol (Group D, Day 20)
and placebo (Group C, Day 20), respectively. The maxima in mean (range) Emax HR decrease
in Groups A and B occurred on Day 16, and were 21.34 (10.5, 40.6) and 28.14 (15.9, 37.7) bpm,
respectively. The corresponding decreases in the propranolol (Group D) and placebo (Group C)
groups were 17.96 (2.1, 31.8) and 9.99 (-0.3, 46.0) bpm, respectively.

Figure 2-2 Emax heart rate (bpm) by time (0 to 24 hours) and treatment group for
co-administration of siponimod and beta blocker propranolol

Group A = propranolol added on top of siponimod, Group B = siponimod added on top of propranolol,
Group C = siponimod-placebo/propranolol-placebo, Group D = siponimod-placebo/propranolol
Source: [Study A2116-Figure 11-1]

All detected bradyarrhythmic events were asymptomatic and there was no second degree AV
block or sinus pause of > 3 sec duration noted in the study in subjects receiving siponimod
(during combination treatment or while receiving siponimod alone).
Changes in pulmonary function (FEV1) with the individual drugs or with the combination
treatment were not clinically significant in view of the magnitude of change, placebo effect and
the week-to-week variability (Pellegrino et al 2005).
A slight decrease in Emax MABP (2.93 mmHg, p = 0.0734) in the combination treatment
(Groups A and B combined, Day 20) in comparison to propranolol alone and a trend for slight
increase in PR interval (2.45 msec at 2.5 hours post dose, p = 0.5341; 7.06 msec at 6.5 hours

Anonymized Page 34
Novartis Confidential Page 35
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

post dose, p = 0.0485) with the combination treatment (Groups A and B combined) were noted,
which were overall not considered to be clinically relevant.
Based on the cardiac safety results (Emax HR, MABP, PR interval, AV block, sinus pause) of
this study the use of beta blockers (previously prohibited) was permitted in the Phase 3
[Study A2304] in SPMS patients, using a specific decision algorithm dependent on the Baseline
HR prior to initiation of concomitant therapy. The results of this study indicated that due to the
less than additive HR effects beta blocker treatment can be safely initiated in patients on stable
siponimod maintenance treatment. In patients receiving a stable dose of beta blocker, the resting
HR should be considered before introducing siponimod treatment. If the resting heart rate
is > 50 bpm under chronic beta blocker treatment, siponimod can be introduced. If resting heart
rate is ≤ 50 bpm, then beta blocker treatment should be interrupted until the Baseline HR
is > 50 bpm prior to initiating siponimod treatment. The beta blocker treatment can be re-
initiated after siponimod has been uptitrated to the target maintenance dose.
Combination treatment did not significantly alter the exposure of siponimod. Mean siponimod
AUCtau,ss and Cmax,ss were approximately 7% lowered when co-administered with
propranolol. However, the test/reference ratio 90% CI remained within bioequivalence limits
(0.80 to 1.25). No change in siponimod Tmax,ss was observed.
Decreases in propranolol exposure in the presence of siponimod were considered not to be
clinically relevant (Cmax,ss and AUCtau,ss decreased by 15% and 18%, respectively).
Propranolol Tmax,ss was similar in presence of siponimod.

2.2.8 Absolute bioavailability and pharmacodynamic/cardiac safety of


intravenous formulations
The absolute bioavailability of siponimod was investigated in Part 1 of [Study A2126], in which
subjects were randomly assigned to 1 of 2 different sequences to sequentially receive single
0.25-mg doses of siponimod administered either orally (Treatment A) or intravenously as a
3-hour infusion (Treatment B) (Section 3.1.1); in Part 2 of the study a 1-mg iv infusion over
24 hours (Treatment C) was investigated. In addition, this study investigated the ALC and the
cardiac safety of iv doses of 0.25 mg/3 hours and 1 mg/24 hours.
The absolute bioavailability of siponimod was 84%. The PK profiles of the metabolites M16
and M17 were also evaluated: M16 was not detected and concentrations of M17 displayed a
median Tmax of 96 hours and geometric mean T1/2 of 150 hours. The geometric mean
metabolite-to-parent molecular ratios for M17 ranged from 0.687 to 0.974 for AUCinf. The
absolute bioavailability results of this study are discussed in further detail in
[SBP-Section 3.2.1].
Mean ALC and mean ALC change from Baseline were comparable after a single 0.25 mg oral
and iv dose of siponimod ([Study A2126-Tables 14.2-5.7a and 14.2-5.7b]). No clinically
relevant or symptomatic effect on mean hourly average HR was observed. Mean hourly average
HR remained above 50 bpm during the entire 25-hour Holter ECG recording period following
Treatments B and C. Bradyarrhythmic events were asymptomatic and their frequency, duration
and diurnal pattern of occurrence was consistent with observations from previous clinical
studies and yielded no new safety signals.

Anonymized Page 35
Novartis Confidential Page 36
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.2.9 Thorough QT/QTc study


A dedicated thorough QT [Study A2118] investigated the effects of therapeutic (2 mg) and
supratherapeutic (10 mg) doses of siponimod on cardiac repolarization, as assessed by the time-
matched, baseline- and placebo-corrected QTcF (ΔΔQTcF). Subjects were randomized to 1 of
3 treatments (18 days each): 1) siponimod 5-day uptitration to 2 mg, followed by 2 mg qd (Days
6 to 10), followed by 3-day uptitration to 10 mg and 10 mg qd (Days 14 to 18), 2) placebo and
3) moxifloxacin 400 mg on Days 10 and 18 (positive control).
Overall, the results of this study do not show an arrhythmogenic potential related to QT
prolongation in healthy subjects. Consistent with these findings were the pooled categorical
analyses of QTcF across studies in healthy subjects and PM/DM patients (Section 3.4.2.2.2)
and the pooled QTcF analyses in MS patients, as described in [SCS-Section 4.3.2.2] (for dose
initiation in [Study A2304]), [SCS-Section 4.2.2] (for the controlled and long-term safety pools
for [Studies A2201 and A2304] for times other than the treatment initiation monitoring period)
and [SCS-Section 4.3.3.2.1] (for the titration pool for (Studies A2201 and A2304).
The upper bounds of the 2-sided 90% confidence intervals (CI) for ΔΔQTcF at both siponimod
dose levels were within the regulatory threshold of 10 msec at all predefined on-treatment time
points, with the absence of any dose-related effects (Figure 2-3). The highest observed upper
limits of the 2-sided 90% CI were 9.86 and 9.69 msec for the therapeutic and supratherapeutic
dose, respectively (both at 3 hours post dose). Siponimod increased the mean ΔΔQTcF by more
than 5 msec, with a maximum mean effect of 7.8 and 7.2 msec at therapeutic and
supratherapeutic doses, respectively, at 3 hours post dose. Assay validity was demonstrated on
both on-treatment assessment days (Day 10 and Day 18) and both the magnitude and the time
pattern of the QT effect of moxifloxacin were consistent with published data (maximum mean
ΔΔQTcF of moxifloxacin of 11.69 and 12.62 msec on Day 10 and 18, respectively, at 3 hours
post dose).

Figure 2-3 Estimated mean difference and 2-sided 90 percent CI for change from
placebo-adjusted time-matched Baseline in QTcF (msec) by time point
and siponimod treatment (2 and 10 mg)

Siponimod therapeutic dose (2 mg) on Day 10 Siponimod supratherapeutic dose (10 mg) on Day 18

Source: [Study A2118-Figure 11-17]

Anonymized Page 36
Novartis Confidential Page 37
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

The frequency of categorical QT/QTc outliers in the siponimod group was low and could not
be differentiated from placebo. Categorical analysis revealed no treatment-emergent QTcF
values above 480 msec and no QTcF increases from Baseline of more than 60 msec on any of
the on-treatment assessment days. No corrected or uncorrected QT/QTc value exceeded
500 msec ([Study A2118-Tables 11-4 and 11-9]). Correlations between ΔQTc and plasma
concentrations of siponimod, M3 and M5 were characterized by prediction lines of negative
slopes with the upper bound of the 2-sided 90% confidence band remaining below 10 msec
within the investigated exposure range ([Study A2118 Addendum-Figures 2-3, 2-4 and 2-5]).
Positive correlations were found between the siponimod, M3 and M5 plasma concentrations
and ΔHR, which translated into corresponding negative correlations between the respective
plasma concentrations and ΔQT due to the interdependence of HR and QT
([Study A2118 Addendum-Figures 2-12, 2-13 and 2-14]). There was no delayed PD response
on ΔΔQTcF, ΔΔQTcI and ΔΔHR for any of the investigated analytes (siponimod, M3 and M5).
No notable repolarization changes or arrhythmias occurred under siponimod treatment and the
incidence of T-wave abnormalities was low in all treatment groups with no trend between the
treatment groups and assessment days.
All observed episodes of second degree AV blocks and sinus pauses were asymptomatic and
considered not to be of clinical relevance.
Plasma exposure and steady-state PK were consistent with previous studies in healthy subjects:
geometric mean Cmax,ss of 30.4 ng/mL and AUCtau,ss of 558 h*ng/mL on Day 10, the fifth
day of 2 mg qd dosing. From Day 10 to Day 18, the fifth day of 10 mg dosing, dose-proportional
(5-fold) increases in siponimod Cmax,ss and AUCtau,ss (to 152 ng/mL and 2680 h*ng/mL,
respectively) were observed; thus the 10 mg supratherapeutic dose of siponimod was
considered appropriate to investigate the effects of siponimod on QT/QTc at substantial
multiples of the anticipated maximum therapeutic exposure. The PK results for M3 and M5 and
the additional PK/PD analyses are reported separately in an addendum to this study
(Section 3.1.7.3).

2.2.10 Studies in special populations

2.2.10.1 Hepatic and renal impairment


The PK of siponimod and metabolites M3 and M5 (prior to identification of M17 as a main
metabolite in humans) following a 0.25-mg oral dose was assessed in subjects with mild,
moderate and severe hepatic impairment ([Study A2122]) and severe renal impairment
([Study A2129]). Overall, the results suggest that no dose adjustment of siponimod is warranted
in subjects with impaired hepatic or renal function (Section 3.2.2 and Section 3.2.3,
respectively). PD assessments of lymphocyte, cardiac and liver function did not reveal any
safety signal or indication for a different PD response in subjects with hepatic or renal
impairment compared to healthy subjects.
Ex vivo clinical plasma protein binding of [14C]siponimod in hepatically and renally impaired
subjects is discussed in Section 3.1.6.

Anonymized Page 37
Novartis Confidential Page 38
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.2.10.2 Ethnic origin: Japanese subjects


Overall, the PK profile of Japanese subjects was similar to the SAD [Study A2101] in
non-Japanese subjects, indicating that there was no influence of Japanese ethnicity on the
disposition of Siponimod (Section 3.2.4). PD assessments of lymphocyte, cardiac, pulmonary
and liver function yielded similar observations as in non-Japanese subjects (Study A2101).
There was no clinically significant effect on pulmonary or liver function. The dose-dependent
decline of ALC observed was similar to that in (Study A2101). PD assessments of cardiac
function (HR, PR interval, AV blocks, QTcF and BP) did not reveal any safety signal or
indication for a different PD response in Japanese subjects compared to healthy subjects.
A comparison of the exposure-lymphocyte relationship between Japanese and non-Japanese
subjects, investigated in a Phase 3 population pharmacokinetics/pharmacodynamics (PopPKPD)
analysis, is described in Section 3.4.3.

2.2.10.3 Genotype: CYP2C9 poor and extensive metabolizers


In [Study A2128], CYP2C9 extensive and poor metabolizers were assigned to receive a
0.25-mg dose of siponimod in Part 1; in Part 2, poor metabolizers from Part 1 were assigned to
0.25 mg siponimod (Days 1 and 2) and 0.5 mg siponimod (Day 3), which represents the first
part of the [Study A2107] DT2 regimen (Table 2-1), which was adopted for the clinical program.
Overall, siponimod exposure was increased by about 2- and 4-fold in poor metabolizers,
CYP2C9*2*3 and *3*3 genotypes, respectively, compared to extensive metabolizers
(CYP2C9*1*1 genotype; Section 3.2.5.1).
Assessments of lymphocyte, cardiac and liver function as well as HR and arrhythmia analyses
did not reveal any safety signal or indication for a different PD response among subjects with
different CYP2C9 genotypes. HR effects observed in the first hours after siponimod
administration or during the nocturnal hours were asymptomatic in all subjects and not
considered clinically relevant. Few detected AV blocks and sinus pauses observed in a single
subject were asymptomatic and not considered clinically relevant. There were also no clinically
significant changes in BP. The results indicate that the first part of the established DT2 regimen
with siponimod (Legangneux et al 2013) efficiently attenuates bradyarrhythmic effects in
CYP2C9*2*3 and *3*3 poor metabolizers.
A comparison of the exposure-lymphocyte relationship between subjects with different
CYP2C9 genotypes, investigated in a Phase 3 PopPKPD analysis, is described in Section 3.4.3.

2.3 Studies in patients


Summaries of key PK, PD and safety results of the 5 studies in MS (2) and PM/DM (3) patients
are discussed individually in this section. Each study is listed in Table 5-1, which includes
details such as study objectives, design, doses and formulations of siponimod and number of
subjects for each treatment group.

Anonymized Page 38
Novartis Confidential Page 39
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

2.3.1 Multiple sclerosis


The efficacy of siponimod in patients with RRMS in the Phase 2 [Studies A2201 and A2201E]
and SPMS in the Phase 3 [Study A2304] (and available data, as of 31-May-2017, from the
ongoing extension) is discussed individually in the [SCE-Sections 2.2 and 2.3]; the safety data
from these MS studies is discussed in the [SCS-Sections 2, 3 and 4]. Overall, no relevant
difference in the PK of siponimod is observed between healthy subjects and patients with MS
(detailed in Section 3.2.1.1). Lymphocyte count reductions were observed in both studies; for
the Phase 2 study, which evaluated multiple doses of siponimod, the reductions were
dose-dependent, consistent with previous multiple dose studies in healthy subjects.

2.3.1.1 Relapsing-remitting multiple sclerosis


In the adaptive Phase 2 dose-ranging (Study A2201), patients were randomized to siponimod
(0.5, 2 or 10 mg) or placebo for 6 months (Period 1); after an interim analysis, a second group
of patients was randomized to siponimod (0.25 or 1.25 mg) or placebo for 3 months (Period 2).
Due to bradyarrhythmic effects observed in patients in Period 1, the DT2 regimen (Table 2-1)
as investigated in [Study A2107] was implemented by protocol amendment for initiation of
treatment in Period 2 or re-initiation of treatment after interrupted drug intake. Thus, Period 1
patients started treatment with full dose without titration, and Period 2 patients were uptitrated
to a target maintenance dose 0.25 or 1.25 mg, respectively. Patients who completed the core
study were eligible to enter the extension [Study A2201E1] that allowed all placebo patients to
be switched to siponimod treatment (0.5, 2 or 10 mg if on placebo in Period 1 of core study;
0.25 or 1.25 mg if on placebo in Period 2 of core study) and patients already on siponimod to
continue the same dose of active treatment during the dose-blinded phase. In the open-label
phase all patients received siponimod 2 mg qd, except for patients who had confirmed
lymphocyte levels < 0.2 × 109/mL who were required to undergo dose reduction to 1 mg qd.
Overall, mean siponimod plasma concentrations in the core and extension studies suggest the
PK of siponimod in RRMS patients is comparable to that in healthy subjects receiving
equivalent doses ([Studies A2105 and A2118]). This is supported by the Phase 1/Phase 2
PopPK analysis (Section 3.1.9.1), where there was no difference in the PK of healthy subjects
and RRMS patients.
In the core study a dose-dependent reduction in ALC was observed over time (Day 7 to
Month 6), with maximal reductions observed at the 10- and 2-mg dose levels, close to maximal
effects at the 1.25-mg dose level and submaximal effects at the 0.5- and 0.25-mg dose levels.
Patients receiving 10, 2 and 1.25 mg siponimod experienced a decrease in ALC to > 0.2 × 109/L
on at least 1 occasion during the course of the study, at frequencies of 36.2%, 17.0%, and 2.4%,
respectively. The highest change from Baseline in mean ALC was similar for the 10-, 2- and
1.25-mg doses: 9.88% (Month 6), 8.19% (Day 7) and 8.11% (Month 3), respectively. In the
dose-blinded phase of the extension study, the 10-mg group had the greatest proportion of
patients with an ALC < 0.2 × 109/mL (54.4%) compared to the other dose groups (range: 0% to
17.2%).

2.3.1.2 Secondary progressive multiple sclerosis


In [Study A2304], a siponimod target maintenance dose of 2 mg was selected based on the
results of [Study A2201]. The target maintenance dose was preceded by a 5-day uptitration,

Anonymized Page 39
Novartis Confidential Page 40
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

which demonstrated efficacy in attenuating bradyarrhythmic effects similar to the DT regimens


investigated in [Study A2107] in healthy subjects and in (Study A2201) in RRMS patients.
Dose reduction to 1 mg qd was permitted in patients with ALC reduction to < 0.2 × 109/L.
Based on the results of [Study A2110], a re-titration was warranted for patients re-initiating
siponimod treatment after missing 4 or more consecutive daily doses during maintenance
dosing (i.e., no re-titration was needed for up to 3 consecutive missed daily doses). In addition,
re-titration was required for patients who missed 1 dose or more during the DT phase.
Overall, mean siponimod plasma concentrations following a 2 mg qd dosing suggests the PK
of siponimod in SPMS patients is comparable to that in healthy subjects ([Study A2118]).
Additionally, in the Phase 3 PopPK analysis (Section 3.1.9.2), there was no evidence of
difference in the PK of SPMS patients and healthy subjects or RRMS patients.
The mean ALC was 0.57 × 109/L for siponimod patients on Day 28 and this reduction was
sustained throughout continued treatment. The incidence of patients with new or worsening
Grade 4 lymphocyte values (< 0.2 × 109/L) at any time was 2.7% (29 patients) in the siponimod
group and 0.2% (1 patient) in the placebo group (due to lab error). Eight of the patients in the
siponimod group had low lymphocyte counts at 2 consecutive visits; of these, 5 had dose
reduction to 1 mg, following which lymphocyte counts started to recover within approximately
2 to 4 weeks.

2.3.2 Polymyositis and dermatomyositis


Three (3) studies ([Studies A2202, X2205 and X2206]) of siponimod in patients with PM or
DM were performed before the clinical program of siponimod in PM/DM indications was
terminated in 2016 for non-safety related reasons. Overall, the PK of PM/DM patients was
generally comparable to observations in healthy subjects (Section 3.2.1.2).
In Period 1 of (Study A2202), PM/DM patients received placebo or 10 mg siponimod using the
9-day DT 2 regimen in (Study A2107) (Table 2-1) followed by stable 10 mg maintenance
dosing. In Period 2 (open-label extension phase), patients on active treatment continued to
receive siponimod; patients previously on placebo were converted to a 10-mg dose following
the 9-day DT regimen used in Period 1. There were no events of bradycardia or other significant
conduction abnormality reported during the uptitration period, demonstrating that DT regimen
of siponimod effectively mitigated the risk of bradyarrhythmia. No liver enzyme elevations
were observed. The primary PD effect of siponimod, decrease in ALC, was observed in all
subjects.

Anonymized Page 40
Novartis Confidential Page 41
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

In Period 1 of (Study X2205), PM patients received siponimod (uptitrated to 2 mg (over 5 days)


or 10 mg (over 9 days)) or matching placebo. In Period 2, patients on active treatment in Period
1 continued to receive siponimod at the same dose; patients randomized to placebo in Period 1
were randomized to receive siponimod 2 or 10 mg (uptitrated as in Period 1). Following
2 unplanned interim analyses the study was terminated prematurely, and the efficacy results
were considered inconclusive. Besides the expected ALC decreases, no other hematologic
findings were noted. There were also no unfavorable cardiac or pulmonary effects in this small
study. Mean GGT of patients in the 2 mg/2 mg sequence appeared to have increased over time
without any other signs of liver damage. However, all values were within the normal range for
the central laboratory (5 to 36 U/L). Safety data from this study remain inconclusive due to the
small number and uneven distribution of patients between the groups.
In Period 1 of (Study X2206), DM patients received siponimod at 0.5, 2 and 10 mg (preceded
by uptitration over 2, 5, or 9 days, respectively) or placebo. In Period 2, all subjects received
siponimod uptitrated to 2 mg (as in Period 1). The study was terminated based on an interim
analysis due to a lack of efficacy and the absence of a dose-response relationship. The majority
of AEs reported in the study were mild to moderate in severity and the incidence increased
during Period 1 with increasing dose of siponimod, specifically in the highest dose group of 10
mg/day. Besides the expected ALC decreases, isolated laboratory abnormalities were noted in
the majority of patients. There were also no unfavorable cardiac or pulmonary effects in this
small study. Two (2) patients showed more pronounced decreases in lymphocyte count which
were recorded as AEs (resulting in study discontinuation for 1 patient). Overall, the safety data
from this study remain inconclusive due to the small number of patients enrolled.

3 Comparison and Analysis of Results Across Studies


3.1 Pharmacokinetics – General characteristics
Figure 3-1 provides an overview of intrinsic and extrinsic factors with and without influence on
the PK of siponimod. Geometric mean ratios and 90% CIs for siponimod Cmax and AUC are
displayed for various comparisons. The clinical relevance of these results is translated into
recommendations also mentioned in the figure. The results of the respective investigations are
summarized in Section 3.1.1, Section 3.1.2, Section 3.2.5.1, Section 3.2.2, Section 3.2.3,
Section 3.3 and [SBP-Sections 3.2.2, 3.3 and 3.5]. The influencse of additional covariates on
siponimod PK (i.e., gender, age, body weight, race and health status (healthy subjects or RRMS
and SPMS patients) that were investigated in 2 PopPK evaluations (discussed in Section 3.1.9
and Section 3.2) are not represented in the figure.

Anonymized Page 41
Novartis Confidential Page 42
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-1 Effect of intrinsic and extrinsic factors on siponimod PK (geometric mean ratios and 90 percent confidence
intervals)

Source: [Study A2128-Table 11-5], [Study A2129-Tables 11-5 and 11-8], [Study A2122-Tables 11-5 and 11-13], [Study A2101-Table 11-7],
[Study A2111-Tables 11-4 and 11-5], [Study A2119-Table 11-9].

Anonymized Page 42
Novartis Confidential Page 43
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.1.1 Absorption and absolute bioavailability

Absorption
In vitro, siponimod was classified as a highly permeable compound. Intestinal uptake data of
siponimod in [DMPK R0800531] and [DMPK R1300921] (Table 5-2; membrane permeability
characteristics presented in [DMPK R0800531-Table 2-1] and [DMPK R1300921-Table 1-1])
suggest that luminal membrane permeability occurs most likely by a passive permeation process
without an involvement of drug efflux transporters (Table 5-6). In vitro assessments of DDI
risk concerning transporters and enzymes involved in absorption is summarized in
Section 3.3.1.2.1.
A representative figure of siponimod absorption in clinical studies is presented in Figure 3-2.
The median Tmax ranged between 2 and 6 hours following single oral administrations of
various solid oral dosage forms including the clinical service formulation (CSF) capsule, market
formulation (MF) tablet and final market image (FMI) tablet ([Studies A2101 and A2111]).
Similarly, the median Tmax,ss after multiple doses ranged between 3 and 4 hours ([Studies
A2105 and A2125]). In individual subjects, the median Tmax ranged from 1.5 to 24 hours
(Study A2101). Further comparison of siponimod absorption between different oral
formulations is discussed below.

Figure 3-2 Geometric mean plasma concentrations of siponimod after single


doses of siponimod (0.1 to 75 mg) (linear scale)

Source: [Study A2101-Figure 16.2.5-2.3]

Anonymized Page 43
Novartis Confidential Page 44
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Formulation effect on absorption


Overall, inter- and intra-study comparisons suggest that, at equivalent doses, exposure to
siponimod is generally similar between oral formulations (CSF liquid and capsule, MF and FMI
tablets) evaluated in the clinical program.
In [Study A2101] in which the CSF formulation of siponimod was administered orally as a
capsule or liquid (Table 5-1), median Tlag and Tmax appeared to be slightly delayed for the
2.5 mg capsule (0.75 and 6.00 hours, respectively) compared to the 2.5 mg liquid solution
(0.13 and 4.00 hours, respectively). Geometric mean Cmax and AUC, however, were
comparable between the 2 formulations: Cmax of 19.3 and 21.9 ng/mL, respectively; AUC0-t
of 721 and 743 h*ng/mL, respectively; AUCinf of 745 and 766 h*ng/mL, respectively.
Across studies, subjects in [Studies A2105 and A2118] received multiple doses of 10 mg oral
siponimod as the CSF capsule and MF (IR) formulations, respectively. For the CSF capsule
formulation, the geometric mean (%CV geo mean) values on Day 28 for Cmax,ss and AUCtau
were 147 (24) ng/mL and 2580 (24) h*ng/mL, respectively. For the MF tablet, the geometric
mean (%CV geo mean) values for Day 18 Cmax,ss and AUCtau,ss were 152 (28.3) ng/mL and
2680 (30.1) h*ng/mL, respectively, following 10 mg qd from Days 14 to 18, demonstrating
similar overall exposure (AUC) between the 2 studies. Comparable median Tmax were
observed after multiple administrations for the CSF capsule (3 hours) and the MF tablet (4
hours).
Bioequivalence between MF and FMI formulations of siponimod (compared after single
0.25- and 4-mg doses) was demonstrated for Cmax, AUClast and AUCinf for both doses in
[Study A2111], as indicated by the fact that all 90% CI were in the prespecified range of 0.80
to 1.25. The results of the study are discussed in detail in [SBP-Section 3.3], along with a
cross-study comparison of the PK of the FMI tablet and MF formulations in [Study A2125] and
[Study A2107], respectively.
The impact of 3 formulation types (solution, capsule and MF tablet) on absorption parameters
was also investigated in the PopPK analysis of Phase 1/Phase 2 data (Section 3.1.9.1).
Compared to the tablet formulation, the duration of the zero-order absorption (D1) was 12%
and 59% smaller for the capsule and the solution, respectively. These results are in line with the
observations made on Tlag and Tmax mentioned above.

Absolute bioavailability
The PK results of Part 1 of [Study A2126] are discussed in detail in [SBP-Section 3.2.1].
The absolute bioavailability of siponimod (FMI tablet) as a single 0.25-mg dose administered
orally was 84% compared with a single 0.25-mg siponimod iv dose: geometric mean AUClast
and AUCinf were 63.6 and 67.4 h*ng/mL, respectively after oral administration, compared to
76.0 and 80.1 h*ng/mL, respectively, following iv administration. The ratios of the geometric
means reveal the Cmax of oral siponimod was approximately 52% of iv siponimod: oral Cmax
of 1.71 ng/mL, compared to iv Cmax of 3.22 ng/mL. Median oral siponimod Tmax was
observed 8 hours after dosing, while median iv siponimod Tmax was observed at the end of the
3-hour infusion. Similar apparent terminal T1/2 results were observed for oral (26.7 hours) and
iv (27.4 hours) administration.

Anonymized Page 44
Novartis Confidential Page 45
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.1.2 Effect of food


Overall, despite a slight delay in Tmax, food intake had no effect on the systemic exposure
(Cmax, AUC) to siponimod in [Study A2111] (FMI tablet, 0.25- and 4-mg doses) and
[Study A2101] (CSF capsule, 5-mg dose). In (Study A2111) (PK results discussed in detail in
[SBP-Section 3.5]), the geometric mean Cmax was about 10% lower for the 4 mg FMI fed
versus 4 mg FMI fasted treatment and was not considered clinically relevant. The statistical
analysis demonstrated that the FMI fasted and FMI fed treatments were bioequivalent for both
doses. Similarly, in [Study A2101], the geometric mean Cmax after food administration was
decreased by approximately 9% compared to that under fasted conditions ([Study A2101-Table
11-6]). All AUC values were similar regardless of dietary status at the time of study drug
administration. The ratio of the geometric means (with 90% CIs) for Cmax, AUC0-24h, and
AUClast were 0.91 (0.79, 1.05), 0.98 (0.91, 1.05), and 1.01 (0.97, 1.04), respectively ([Study
A2101-Table 14.2-1.4]), indicating no effect from food.
In [Study A2111] the Tmax was slightly delayed with food intake (6 to 7 hours across both
0.25 mg and 4-mg doses, versus 4 hours). For fed and fasted conditions in (Study A2101), the
median Tlag was 0.38 and 0.25 hours, respectively, and median Tmax 8.00 and 3.00 hours,
respectively.

3.1.3 Dose proportionality


Dose proportionality was assessed using power model, Y = α × Doseβ, where Y, α, and β
correspond to the PK parameter, proportionality constant, and exponent, respectively. The slope
estimate value of beta ≈ 1 indicates apparent dose-proportionality.
Overall, with siponimod single dose (0.1 to 75 mg) and multiple doses (0.3 to 20 mg)
administrations, siponimod Cmax and AUC rose in an apparent dose-proportional manner. The
PK of siponimod was also linear with regards to dose following multiple dose administrations
in the range of 0.25 to 10 mg in RRMS patients.

3.1.3.1 Dose proportionality after single dose


Overall, with siponimod single dose administrations, siponimod Cmax and AUC rose in an
apparent dose-proportional manner, over the range of 0.1 to 75 mg doses. Dose proportionality
analysis was performed for Cmax and AUC siponimod PK parameters in [Studies A2101,
Studies A1101 and A2105] (Day 1) (single doses of 0.1 to 75 mg, collectively). The observed
slope estimate and 90% CI from the power model for Cmax and AUC is summarized
individually for the 3 studies in Table 3-1. Pooled analyses of the (Studies A2101 and A2105)
single dose data (Day 1) are presented in Table 3-2, Figure 3-3 and Figure 3-4.
In (Study A2101), healthy subjects received siponimod single doses of 0.1, 0.3, 1, 2.5, 5, 10,
17.5, 25 and 75 mg. The slope estimates (90% CI) for Cmax, AUCinf and AUC0-24h were 1.01
(0.98, 1.04), 1.00 (0.97, 1.02) and 1.00 (0.97, 1.03), respectively. In (Study A2105), healthy
subjects received siponimod single doses of 0.3, 1, 2.5, 10 and 20 mg. The slope estimates (90%
CI) for Cmax and AUC0-24h were 1.03 (0.98, 1.09) and 1.03 (0.97, 1.09), respectively. In
(Study A1101), Japanese healthy subjects received siponimod single doses of 0.5, 2.5, 5 and
10 mg. The slope estimate (90% CI) for Cmax, AUCinf and AUC0-24h were 1.03 (0.98, 1.08),
0.99 (0.92, 1.05), and 1.01 (0.97, 1.05), respectively. In all 3 studies, siponimod Cmax,

Anonymized Page 45
Novartis Confidential Page 46
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

AUC0-24h and AUCinf ([Studies A2101 and A1101]) rose in an apparent dose-proportional
manner over the investigated dose range.

Table 3-1 Individual study estimation of siponimod dose-pharmacokinetic


relationship after a single dose
PK parameter Intercept Slope 90% CI (slope)
Study A2101
Cmax (ng/mL) 1.97 1.01 (0.98, 1.04)
AUC0-24h (h*ng/mL) 4.79 1.00 (0.97, 1.03)
AUCinf (h*ng/mL) 5.65 1.00 (0.97, 1.02)
Study A1101
Cmax (ng/mL) 2.18 1.03 (0.98, 1.08)
AUC0-24h (h*ng/mL) 4.93 1.01 (0.97, 1.05)
AUCinf (h*ng/mL) 5.68 0.99 (0.92, 1.05)
Study A2105
Cmax (ng/mL) 2.01 1.03 (0.98, 1.09)
AUC0-24h (h*ng/mL) 4.83 1.03 (0.97, 1.09)
Power model: the slope parameter estimate and the CI are obtained from the linear model of the log
transformed PK parameter value adjusted for an intercept and the log-transformed dose (covariate)
as fixed effects.
Notes: For [Study A2105]:
Dose PK relationship for Cmax are analyzed with model: ln(PK) = ln(DOSE)*DAYDAY;
Dose PK relationship for AUC0-24h are analyzed with model: ln(PK) = ln(DOSE).
Source: [Study A2101-Table 11-4, Study A1101-Table 11-3 and Study A2105-Table 11-4]

Exploratory pooled statistical analyses using a power model for PK parameters Cmax and
AUC0-24h and linear regression for dose-normalized PK parameters were performed.
The statistical analysis indicates Cmax and AUC0-24h increased in a dose-proportional manner
considering the slope (≈ 1.00) and 90% CI (0.98 to 1.04). From the regression analysis of
dose-normalized PK parameters versus dose, the slope ≈ 0 indicates that dose-normalized PK
parameters are comparable at different dose levels.

Table 3-2 Pooled study estimation of siponimod dose-pharmacokinetic


relationship after a single dose
90% CI for slope
PK parameter (unit) Slope estimate Lower Upper
AUC0-24h (hr*ng/mL) 1.01 0.98 1.03
Cmax (ng/mL) 1.02 0.99 1.04
Power model: The slope parameter estimate and the confidence interval are obtained from the
linear model of the log transformed PK parameter value adjusted for an intercept and the log-
transformed dose (covariate) as fixed effects.
Study A2101 data considered: 0.1-, 0.3-, 1-, 2.5-, 5-, 10-, 17.5-, 25- and 75-mg doses
Study A2105 data considered (Day 1): 0.3-, 1-, 2.5-, 10- and 20-mg doses
Source: (SCP Appendix-Table 11-1.1)

Anonymized Page 46
Novartis Confidential Page 47
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-3 Pooled study estimation of dose normalized Cmax dose


proportionality after a siponimod single dose

[Study A2101] data considered: 0.1-, 0.3-, 1-, 2.5-, 5-, 10-, 17.5-, 25- and 75-mg doses
[Study A2105] data considered (Day 1): 0.3-, 1-, 2.5-, 10- and 20-mg doses
Cmax intercept = 7.849; slope = -0.003.
Source: (SCP Appendix-Figure 11-1.1)

Anonymized Page 47
Novartis Confidential Page 48
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-4 Pooled study estimation of dose normalized AUC0-24h dose


proportionality after a siponimod single dose

[Study A2101] data considered: 0.1-, 0.3-, 1-, 2.5-, 5-, 10-, 17.5-, 25- and 75-mg doses
[Study A2105] data considered (Day 1): 0.3-, 1-, 2.5-, 10- and 20-mg doses
AUC0-24h intercept = 129.644; slope = -0.073.
Source: (SCP Appendix-Figure 11-1.1)

3.1.3.2 Dose proportionality following multiple doses


Overall, the PK of siponimod was linear with regards to dose following multiple dose
administrations in both healthy subjects and MS patients. In (Study A2105) (0.3-, 1-, 2.5-, 10-
and 20-mg doses), the relationship between dose and observed Cmax and AUC on Day 28 in
healthy subjects is summarized in Table 3-3. The slope estimate and 90% CI data indicate that
on Day 28, Cmax, AUCtau and AUClast approximately increased in a dose-proportional
manner in the investigated dose range. In RRMS patients in [Study A2201] (0.25 to 10 mg), the
PK linearity with regards to dose was confirmed in [CBAF312A-Phase 1-2-PopPK] through a
linear regression of normalized concentrations (normalized to a 1-mg dose) versus dose
(Figure 3-5). The slope was estimated as 0.009 with a SE of 0.082 (i.e., not statistically different
from 0).

Anonymized Page 48
Novartis Confidential Page 49
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-3 Estimation of siponimod dose-pharmacokinetic relationship after


multiple doses in healthy subjects
PK parameter Day Intercept Slope 90% CI (slope)
Cmax,ss (ng/mL) 28 2.79 0.99 (0.94, 1.05)
AUCtau 28 5.69 0.98 (0.94, 1.03)
(h*ng/mL)
AUClast 28 6.51 0.99 (0.92, 1.06)
(h*ng/mL)
Notes: Dose PK relationship for AUClast and Cmax are analyzed with model: ln(PK) =
ln(DOSE)*DAY
DAY and dose PK relationship for AUCtau are analyzed with model: ln(PK) = ln(DOSE)
Source: [Study A2105-Table 11-4]

Figure 3-5 Plasma concentrations of siponimod normalized to dose as a function


of the dose in RRMS patients

Source: [CBAF312A-Phase 1-2-PopPK – Figure 4-1]

Anonymized Page 49
Novartis Confidential Page 50
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.1.4 Pharmacokinetics of single doses


Overall, siponimod PK after administration of a single oral dose is characterized with a fast
absorption, small oral clearance, moderate apparent volume of distribution and moderate
apparent terminal half-life. The main PK parameters under fasted conditions in the
first-in-human SAD study (0.1 to 75 mg) are presented in Table 3-4 and the geometric mean
plasma concentrations in [Study A2101-Figure 11-3]. Siponimod was measurable in the plasma
as early as 0.25 hours post dose. The plasma concentration peaked between 3 and 6 hours post
dose (median) with individual subject values ranging from 1.5 to 24 hours. Cmax and
AUC0-24h appeared to increase dose-proportionally (Section 3.1.3.1). The inter-subject
variability was low to moderate for Cmax and AUC (% CV geometric mean values of 6% to
51% and 10% to 62%, respectively). The decay of siponimod plasma concentration was
mono-exponential with a geometric mean (%CV geometric mean) apparent terminal T1/2
between 27.02 (22) and 56.69 (12) hours and CL/F between 2.86 (28) and 4.14 (25) L/h.
The geometric mean (%CV geometric mean) values of Vz/F ranged from 127 (34) to
330 (58) L.

Anonymized Page 50
Novartis Confidential Page 51
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-4 Main PK parameters after siponimod single dose administration (0.1 to 75 mg) to healthy subjects
Siponimod 0.1 mga 0.3 mga 1.0 mga 2.5 mga 2.5 mgb 5.0 mgb 10.0 mgb 17.5 mgb 25.0 mgb 75.0 mgb
Treatment (N = 11) (N = 8) (N = 8) (N = 6) (N = 7) (N = 8) (N = 8) (N = 8) (N = 8) (N = 8)

Tlag 0.25 0.00 0.00 0.13 0.75 (0.50- 0.25 (0.00- 0.25 0.25 0.25 0.25
(h)c (0.00-0.27) (0.00-0.25) (0.00-0.25) (0.00-0.25) 0.77) 0.50) (0.00-0.50) (0.00-0.52) (0.00-0.50) (0.25-0.50)

Tmax 4.00 5.00 5.00 4.00 6.00 3.00 5.00 4.00 3.50 6.00
(h)c (3.00-8.00) (4.00-8.00) (4.00-15.67) (3.00-8.00) (4.00-8.02) (2.00-8.02) (3.85-16.00) (2.00-8.00) (1.50-12.00) (2.00-24.00)

T1/2 33.06 [26] 33.21 [28] 45.68 [26] 27.02 [22] 29.31 [17] 31.27 [19] 32.03 [17] 42.32 [38] 47.68 [20] 56.69 [12]
(h)d

Cmax 0.618 [39] 2.26 [6] 7.43 [42] 21.9 [35] 19.3 [19] 38.5 [21] 77.3 [25] 111 [32] 217 [29] 491 [51]
(ng/mL)d

AUCinf 24.6 [39] 89.3 [20] 349 [28] 766 [36] 745 [25] 1260 [20] 2710 [14] 4230 [25] 8140 [24] 18600 [51]
(h*ng/mL)d
AUC0-t 22.9 [41] 87.2 [20] 345 [27] 743 [36] 721 [26] 1240 [21] 2680 [13] 4200 [25] 8110 [24] 18500 [51]
(h*ng/mL)d

CL/F 4.07 [39] 3.36 [20] 2.86 [28] 3.26 [36] 3.36 [25] 3.95 [20] 3.69 [14] 4.14 [25] 3.07 [24] 4.03 [51]
(L/h)d

Vz/F 194 [31] 161 [19] 189 [32] 127 [34] 142 [23] 178 [21] 170 [22] 253 [36] 211 [29] 330 [58]
(L)d
a CSF liquid formulation
b CSF capsule formulation
c Median (min-max)
d Geometric mean [% geometric mean coefficient of variation]
Source: [Study A2101-Table 11-3, Table 16.2.5-2.23, Table 16.2.5-2.24, Table 16.2.5-2.25, Table 16.2.5-2.26, Table 16.2.5-2.27, Table 16.2.5-2.28, Table 16.2.5-2.30,
Table 16.2.5-2.31, Table 16.2.5-2.32, Table 16.2.5-2.33]

Anonymized Page 51
Novartis Confidential Page 52
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.1.5 Pharmacokinetics of repeated administration


Overall, siponimod exhibits time-independent PK: siponimod PK after multiple dose
administration is consistent with those after single dose administration.

3.1.5.1 Multiple ascending dose


The main PK parameters on Day 28 under fasted conditions in the MAD [Study A2105] (0.3 to
20 mg) are presented in Table 3-5 and the arithmetic mean (SD) concentrations of siponimod
in plasma over 28 days in [Study A2105-Figure 11-3]. A representative concentration-time plot
for siponimod on Days 1, 7 and 28 is presented for the 1.0-mg dose group in [Study A2105-
Figure 11-2]. After the first dose, the PK parameters of siponimod were consistent with those
estimated at the same dose levels in [Study A2101]. Siponimod was measurable in plasma as
early as 0.25 hours post dose. The plasma concentration of siponimod peaked between 3 and
4.5 hours post dose (median). Cmax and AUClast appeared to increase dose-proportionally
(Section 3.1.3.2). The range of geometric mean (%CV geometric mean) values for Cmax and
AUClast was 2.13 (13) to 162 (20) ng/mL and 36.2 (10) to 2740 (14) h*ng/mL, respectively.
Siponimod trough concentrations indicated that steady state was reached in all 5 cohorts after
approximately 6 days. The Tmax,ss ranged between 3.00 and 4.00 hours, similar to the Tmax
range after the first dose. Comparable PK parameters were defined on Days 7 and 28.
Mean AUCtau after once daily dosing was comparable to the mean AUCinf after a single dose
at the same dose levels in the SAD study. Mean Racc, calculated as the ratio of AUCtau over
AUClast (Day 1), was between 1.88 to 2.72 on Day 28, with similar values determined for
Day 7 (between 1.82 and 2.42). The decay in siponimod plasma concentration over time after
28 days of dosing appeared to be bi-exponential. A second elimination phase could be observed
after approximately 168 hours post dose. The apparent terminal T1/2 associated with this second
phase (T1/2,β) was between 71.5 and 110 hours. However, the effective T1/2 (T1/2,α) based on
drug accumulation at steady state was 22 to 36 hours, which was comparable to the apparent
terminal T1/2 determined in the SAD study. Mean apparent CL/F at steady state was 3.06 to
3.89 L/h, comparable to the values observed in the SAD study, indicating that siponimod PK is
time-independent.
The ranges of geometric mean (%CV geometric mean) values for Cmax,ss and AUClast were
5.31 (16) to 359 (17) ng/mL and 240 (38) to 17000 (46) h*ng/mL, respectively. A capping
maximum clinical exposure was incorporated into the protocol, to avoid the exposure at which
CNS symptoms had been observed in the 4-week chronic toxicology study in monkeys
([Table 2.6.7.7E-Study 0670007]).

Anonymized Page 52
Novartis Confidential Page 53
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-5 Main pharmacokinetic parameters after siponimod once daily


administration (0.3 to 20 mg) over 28 days to healthy subjects –
Day 28
Parameters 0.3 mg Daily 1.0 mg Daily 2.5 mg Daily 10 mg Daily 20 mg Daily
(N = 6) (N = 6) (N = 5) (N = 9) (N = 8)
Tlag (h)a 0.00 (0.00-0.00) 0.00 (0.00-0.00) 0.00 (0.00-0.00) 0.00 (0.00-0.00) 0.00 (0.00-0.00)
Tmax,ss (h)a 3.50 (3.00- 3.00 (2.00-8.00) 4.00 (3.00-8.00) 3.00 (2.00-6.00) 3.00 (3.00-4.00)
12.00)
Cmax,ss 5.31 [16] 14.9 [10] 38.3 [37] 147 [24] 359 [17]
(ng/mL)b
AUClast 240 [38] 608 [28] 145 [55] 5320 [35] 17000 [46]
(h*ng/mL)b
AUCtau 97.9 [19] 282 [16] 692 [45] 2580 [24] 6370 [23]
(h*ng/mL)b
CLz/F (L/h)b 3.06 [19] 3.55 [16] 3.61 [45] 3.89 [24] 3.21 (0.705)c
Vz/F (L)b 321 [22] 463 [40] 358 [64] 529 [36] 515 (148)c
Cavg (ng/mL)b 4.08 [19] 11.7 [16] 28.8 [45] 107 [24] 265 [23]
Cmin,ss 2.83 [27] 7.74 [20] 15.2 [120] 70.6 [29] 166 (88.0)c
(ng/mL)b
T1/2 (h)b,d 71.5 [27] 90.4 [27] 68.8 [28] 93.6 [28] 110 [9]
Fluc (%)b 54.3 [29] 48.2 [30] 69.3 [39] 63.5 [14] 52.5 [41]
Raccb 2.71 [19.1] 2.07 [12.3] 2.72 [47.3] 1.88 [19.2] 2.28 [22.9]
a Median (min–max)
b Geometric mean [% geometric mean coefficient of variation]
c Arithmetic mean (SD)
d Corresponds to T1/2,β. The effective T1/2 (based on drug accumulation at steady-state) is between 22 and 36
hours (Boxenbaum and Battle 1995).
Source: [Study A2105-Table 11-3]

For the 2-mg therapeutic dose, in [Study A2118] the geometric mean Cmax,ss was 30.4 ng/mL,
Cmin,ss was 14.5 ng/mL and AUCtau,ss was 558 h*ng/mL on Day 10 following 5 days of 2
mg qd dosing preceded by a 5-day DT, consistent with steady-state PK in the MAD study and
other previous studies in healthy subjects.

3.1.5.2 Dose titration


The main PK parameters under fasted conditions in [Study A2107] for siponimod 10 mg qd and
the 2 different DT regimens of siponimod up to 10 mg (Table 2-1) are presented in
[Study A2107-Tables 14.2-1.2 and 14.2-1.3] and the daily geometric mean plasma
concentrations in Figure 3-6. The time to steady state and the Racc estimated at steady state for
the 10 mg qd group were comparable to those observed in [Study A2105]. Siponimod trough
concentrations indicated that steady state was reached for the 10 mg qd group after
approximately 6 days for most of the subjects. The mean Racc for the 10 mg qd group,
calculated as the ratio of AUC0-24h on Day 12 over AUC0-24h on Day 1, was 2.33.

Anonymized Page 53
Novartis Confidential Page 54
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

The inter-subject variability in Cmax and AUClast remained constant over the whole treatment
duration for the DT1 cohort, with a similar variability as previously observed in [Study A2105].
For the 10 mg qd and the DT2 cohorts, a slightly higher variability was noticed toward the end
of the treatment period compared to the variability observed on the first 6 days of treatment.
This was mainly due to 2 outliers. Consequently, the mean siponimod exposure observed on
Day 12 for the 10 mg qd and for the DT2 cohorts were slightly higher compared to the DT1
cohort.

Figure 3-6 Daily geometric mean plasma concentrations of siponimod after once
daily dosing administered over 12 days as a fixed dose or uptitration
to 10 mg

T02 = siponimod 10 mg
T03 = DT#1
T04 = DT#2
Source: [Study A2107-Figure 11-1]

3.1.6 Distribution and protein binding


Overall, siponimod was moderately distributed within the human body following oral
administration. Siponimod and its metabolites were mainly confined within the plasma
compartment and very highly bound to human plasma proteins. No obvious differences in
[14C]siponimod plasma protein binding were observed between healthy subjects and subjects
with renal or hepatic impairment.
In the absolute bioavailability [Study A2126], the Vz/F was 143 L after a single oral dose of
siponimod. Siponimod and its metabolites were mainly confined within the plasma
compartment (mean blood/plasma AUC-ratio of radioactivity of approximately 0.67).
Siponimod and/or its metabolites displayed no special affinity to erythrocytes. The in vitro
blood distribution of [14C]siponimod was assessed in human and other animal species in

Anonymized Page 54
Novartis Confidential Page 55
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

[Table 2.6.5.6A-DMPK R0400881-01 and DMPK R1300902-01] ([PK Written Summary-


Section 4.1]). In the concentration range of 10 to 10000 ng/mL [14C]siponimod, no
concentration dependency of blood distribution was found. The fraction found in plasma was
68% in humans, corresponding to a ratio of blood-to-plasma concentrations (Cb/Cp) of 0.77.
The plasma protein binding of siponimod and its metabolites in [DMPK R1300334],
[DMPK RCBAF312A2129-01] and [DMPK R1400021] (Table 5-2) are presented in Table 5-
4. In the [DMPK RCBAF312A2129-01] and [DMPK R1300334] assessments in subjects with
hepatic and renal impairment, respectively, [14C]siponimod was very highly bound to plasma
proteins in all subjects (> 99.9%). The fu values in these studies were in line with those across
species in [Table 2.6.5.6A-DMPK R0400881-01 and DMPK R1300902-01]: 0.03% in rat, 0.03%
in monkey, 0.02% in human, and 0.01% in dog ([PK Written Summary-Section 4.1]). M3 and
M5 were also highly bound to human plasma proteins (99.4% and 99.6%, respectively;
[DMPK R1400021; Table 2.6.5.6A-DMPK R0400881-01 and DMPK R1300902-01]), and
M17 plasma protein binding was > 99.9% in human, mouse and rat ([Table 2.6.5.6B-
DMPK R1500677-01; PK Written Summary-Section 4.1]).
In human, plasma lipoproteins contributed strongly to the overall binding ([Table 2.6.5.6A-
DMPK R0400881-01 and DMPK R1300902-01]). [14C]siponimod was very highly bound
(≥ 99.8%) to human plasma lipoproteins (high density lipoprotein (HDL), low density
lipoprotein (LDL) and very low density lipoprotein (VLDL)), human serum albumin and
α1-acid glycoprotein (AGP). The contribution of lipoproteins to the total binding in human
plasma was larger as compared to the contribution of albumin and AGP.

3.1.7 Metabolism
This section describes the biotransformation pathways identified in the metabolism of
siponimod, the resultant metabolites formed and their clinical relevance in the pharmacological
activity of siponimod. Overall, no active or toxic siponimod metabolites were identified in
humans (Section 3.1.7.2, Section 3.1.7.3, [Toxicology Written Summary]). Consistent with in
vitro findings (Section 3.1.7.1), the primary phase I metabolic pathway identified in the
dedicated human ADME [Study A2104] was hydroxylation of siponimod to form M5, M6 and
M7 (Section 3.1.7.2). Subsequent phase II glucuronidation of M5 yields M3. An additional
metabolite, M17, was first identified in a mouse ADME study ([Table 2.6.5.9A-DMPK
R1300411-02; PK Written Summary-Section 5.2.1]); its presence in human was later confirmed
in [Study A2126]. The above 2 clinical studies together with [Studies A2118 and A2304]
revealed M3 and M17 as the main metabolites of siponimod (Section 3.1.7.3). Unbound Cmax
and EC50 values suggest that M3 and M17 systemic exposure are unlikely to translate into a
significant increase in pharmacological activity on S1P1 or S1P5 receptors. Additional details
of all of the siponimod metabolites in human are provided in the [PK Written
Summary-Section 5.2.4] and [PK Tabulated Summary], and the contribution of siponimod
metabolites to pharmacological activity on S1P receptors is further discussed in the
[Pharmacology Written Summary].

Anonymized Page 55
Novartis Confidential Page 56
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.1.7.1 Metabolism of siponimod in vitro


In vitro studies identifying metabolites described in [Study A2104] are included in
[Table 2.6.5.10A-DMPK R0400863-01]. Briefly, hydroxylation (M5, M6 and/or M7) was a
major metabolic pathway in incubations with liver microsomes of mouse, rat, monkey and
human but showed some differences across the species. Glucuronidation (M3) of the
hydroxylated M5 was observed as another major metabolic pathway in incubations with liver
microsomes from mouse, rat, monkey and human and with hepatocytes from mouse, rat and
human. Sulfate conjugation (M4) of a hydroxylated metabolite was found in incubations with
hepatocytes of rat and human. Additionally, the 3 hydroxylated metabolites M5, M6 and M7
were detected in [DMPK R0500432] following metabolism of [14C]siponimod in pooled HLM.
When siponimod was also incubated with a panel of 21 rhCYPs, all 3 hydroxylated metabolites
formed in HLM were found in the CYP2C9 incubates. With the results in rhCYPs and HLM
and the relative abundance of the currently known CYP isoforms, it was estimated that CYP2C9
was contributing predominantly (79.3%) to the oxidative metabolism of siponimod, with a
contribution from CYP3A of 18.5% (Figure 3-10; Table 5-6). It was also noted in the study that
compared to CYP2C9*1*1, substantially lower metabolism rates of siponimod in HLM were
observed for CYP2C9*2*2 (2.9-fold reduction) and CYP2C9*3*3 (up to 10-fold reduction)
donors, confirming the influence of CYP2C9 genetic polymorphism on the oxidative
metabolism of siponimod.

3.1.7.2 Metabolites identified in the human ADME Study A2104


A general scheme of the metabolism of siponimod in human from (Study A2104), including the
chemical structures of metabolites identified, in plasma (pl), urine (ur) and feces (fe), is shown
in Figure 3-7. The phase I metabolic reactions in the biotransformation of siponimod involved
C-hydroxylations (M5, M6 and M7), cleavage/hydrolysis at the oxime ether bound (M1, M2)
and further reduction yielding metabolite M8. Phase II reactions of hydroxylated metabolites
involved sulfation to yield M4a, M4b and M4c (from M5, M6 and M7, respectively) and
glucuronidation to yield M3 (from M5) and M12 (formed by hydroxylation followed by
glucuronidation) ([Study A2104d]).
In plasma, siponimod represented the main proportion of radioactivity (57.1% ± 5.9% of
[14C]-AUC0-120h). The major metabolite M3 accounted for 18.4% ± 5% of the plasma
[14C]-AUC0-120h (M3 AUCinf equaling 27.6% of the exposure to siponimod). Minor
proportions of other metabolite peaks were detected and attributed to M5, M6, M7, P29.6
and P30.5, each accounting on average for 1.5% to 3.7% of the plasma [14C]-AUC0-120h ([PK
Written Summary-Section 5.2.4], [PK Written Summary-Table 10-5]).
All minor metabolites detected in the study attributed to less than 10.0% of the total
[14C]-AUCinf and are thus exempt from further toxicological evaluation
[PK Written Summary-Section 5.2.5]. For M5, M6 and M7, the AUCinf was estimated using
the apparent elimination T1/2 of 34.0, 31.6 and 35.2 hours, respectively. In the case of P29.6
and P30.5, the apparent terminal T1/2 of total radiolabeled components (radioactivity) in
plasma was taken (T1/2 of 171 hours).

Anonymized Page 56
Novartis Confidential Page 57
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-7 General scheme of siponimod biotransformation pathways in human

Source: [Table 2.6.5.11F-Metabolic pathways in vivo in human]

Anonymized Page 57
Novartis Confidential Page 58
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.1.7.3 PK of main human plasma metabolites M3 and M17


Overall, M17 PK is characterized by a late Tmax and a long apparent terminal T1/2. M3 Tmax
and T1/2 are comparable to those of siponimod. M3 exhibits dose and time-independent PK.
The contribution of these metabolites to the overall siponimod pharmacological activity is
expected to be negligible.
This section focuses on the PK of the main human plasma metabolites of siponimod M3 and
M17 from [Studies A2118 and Study A2126], respectively. The PK of M3 in subjects with
hepatic and renal impairment is discussed in Section 3.2.2 and Section 3.2.3, respectively. M3
PK for co-administration of siponimod and rifampin and M3 and M17 PK for co-administration
of siponimod and itraconazole are discussed in Section 3.3.1.2.2, and M3 PK poor and
extensive metabolizers is discussed in Section 3.2.5.1.2.
M3, formed by glucuronidation of the hydroxylated metabolite M5, was identified as the main
circulating metabolite in the human ADME [Study A2104], where M3 AUCinf amounted to
27.6% of siponimod exposure. Similar ratios of exposure between M3 and the parent drug were
observed in (Study A2118) after multiple administrations. M17, a cholesterol ester, was
identified as the most prominent systemic metabolite in human based on AUC in the absolute
bioavailability (Study A2126).

Metabolite M3
The PK of M3 following 2- and 10-mg doses of siponimod is summarized in the addendum to
(Study A2118). Following multiple oral administration of siponimod, M3 plasma
concentrations peaked at approximately 6 hours post dose (median). Based on molar ratios, M3
AUClast and Cmax,ss represented 35% to 39% and 33% to 35% of siponimod AUCtau,ss and
Cmax,ss, respectively. These results for M3, as well as those for M5 also summarized in the
addendum, are consistent with (Study A2104) observations. Dose proportional increases in M3
Cmax,ss and AUClast were observed from Day 10 (2 mg) to Day 18 (10 mg). Taking into
consideration the single dose data of (Study A2104), it can be concluded that, like siponimod,
M3 was at steady state levels on Days 10 and 18. This is consistent with a similar T1/2 of
approximately 30 hours for siponimod and M3 (Study A2104).
Unbound M3 AUCtau,ss values and EC50 are presented in Table 3-6. The EC50 value of M3
tested on human S1P1 receptors was > 10000 nmol/L, demonstrating very weak or no activity
compared to the parent drug (EC50 of 1.1 ± 0.41 nmol/L ([Pharmacology Tabulated Summary-
Table 2.6.3.2-RD-2012-00145]). Hence the observed M3 unbound systemic exposure is
unlikely to translate into any pharmacological activity on S1P1 receptors).

Anonymized Page 58
Novartis Confidential Page 59
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-6 M3 unbound AUCtau,ss and EC50 on human S1P1 receptor


Dose Geo-Mean AUCtau,ss,u EC50 ± SD
(mg) (ng/mL*h)a (nmol/L*h)b (nmol/L)
2 (MD) 1.45 2.05 >10000
MD = multiple dose.
a Calculated from M3 PK values using fu of 0.572 for M3
b Calculated using MW of M3 of 708.7 g/mol.
Source: [Study A2118 Addendum 1-Table 2-1], [Table 2.6.3.2-RD-2012-00145]

Metabolite M17
In Part 1 of the absolute bioavailability study of siponimod (0.25 mg single dose;
[Study A2126]), discussed in detail in [SBP-Section 3.2.1], M17 represented about 81% to 97%
of the parent exposure (based on AUCinf) and was identified as the most prominent systemic
metabolite in human. A summary of M17 PK parameters is presented in Table 3-7, and the
concentration-time profile is depicted in [Study A2126-Figure 11-2a]. The PK of M17 was
comparable following 0.25 mg siponimod oral and iv administrations; however, an earlier peak
in the plasma concentration-time profiles of M17 was observed 6 hours after oral administration
only. For the 0.25 mg oral dose, M17 plasma levels peaked at 96 hours. The geometric means
of the T1/2, Cmax and AUCinf were 155 hours, 0.356 ng/mL and 112 h*ng/mL, respectively.
The geometric means of the individual metabolite-to-parent molecular ratios for AUCinf and
Cmax were 0.974 and 0.122, respectively.
Based on (Study A2126) data and assuming dose linearity and time independent PK for M17,
it is estimated that, following multiple administration of siponimod 2 mg qd, M17 AUCtau at
steady state should be about 896 h*ng/mL and Cav,ss about 37 ng/mL. M17 was measured at
the EOS Visit in [Study A2304], 24 hours after the last siponimod dose. The geometric mean
concentration of M17 after a 2 mg qd dosing was 30.6 ng/mL. The M17 to parent molar
concentration ratio (based on geometric means) was 0.889. The M17 concentration of
30.6 ng/mL reported in the current study, 24 hours after the last dose, is close to this estimated
Cav,ss. Considering the long M17 T1/2 observed in healthy subjects (~155 hours), a flat
exposure profile is expected for M17 at steady state following daily siponimod administration,
(i.e., limited difference between Cmin,ss and Cav,ss should be observed).
Unbound M17 AUCtau values and EC50 are presented in Table 3-8. While the EC50 values of
M17 tested on human S1P1 and S1P5 receptors were in the nanomolar range (341 ± 86 nmol/L
and 159 ± 9 nmol/L), M17 is ~70- to 80-fold less potent on S1P receptors than the parent drug
(EC50 of 4.3 ± 0.5 nmol/L ([Table 2.6.3.2-RD-2015-00301]). Hence the observed unbound
M17 systemic exposure is unlikely to translate into a significant increase in pharmacological
activity on S1P1 and S1P5 receptors (contribution of about 3% to total pharmacological activity
on S1P1 and S1P5).

Anonymized Page 59
Novartis Confidential Page 60
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-7 Summary of M17 pharmacokinetic parameters in Study A2126 Part 1


Statistics Siponimod 0.25 mg single oral Siponimod 0.25 mg/3 h single iv
dose infusion
(N = 15) (N = 15)
AUCinf 112 [20.5]a 105 [29.2]b
(h*ng/mL)b
AUClast 68.3 [38.0] 73.3 [38.7]
(h*ng/mL)b
Cmax (ng/mL)b 0.356 [25.7] 0.352 [27.2]
Tmax (h)a 96.0 (6.00-144) 96.0 (72.0-144)
T1/2 (h)b 155 [22.0]a 150 [12.6]b
MPR AUCinfb 0.974 [19.0]a 0.811 [23.1]b
MPR AUClastb 0.627 [29.6] 0.564 [23.9]
MPR Cmaxb 0.122 [24.0] 0.0639 [24.1]
a Median (min–max)
b Geometric mean [% geometric mean coefficient of variation]
cN=9
dN=8
MPR = Metabolite-to-parent ratio, calculated as (M17 parameter * siponimod MW
(516.61 D))/(siponimod parameter * M17 MW (884.5 D)); n = number of subjects with non-missing
values; N = number of subjects who received each treatment.
Source: [Study A2126-Table 14.2-5.4a]

Table 3-8 Unbound M17 AUCtau and EC50 on human S1P1 and S1P5 receptors
Dose Geo-Mean AUCtau,ss,u S1P1 EC50 ± SD S1P5 EC50 ± SD
(mg) (ng/mL*h)a (nmol/L*h)b (nmol/L) (nmol/L)
2 (multiple dose) 0.408c 0.460c 341 ± 86 159 ± 9
a Calculated using fu of 0.0455 for M17
b Calculated using MW of M17 of 885.25 g/mol
c Calculated from values extrapolated from [Study A2126] data assuming dose and time linearity PK
for M17
Source: [Study A2126-Table 14.2-5.4a], [Table 2.6.3.2-RD-2015-00301]

3.1.8 Elimination/Excretion
Overall, siponimod is primarily metabolized through hydroxylation forming M5, M6 and M7
and subsequent glucuronidation or sulfation, with fecal/biliary excretion as the major route of
elimination. The apparent clearance of siponimod was low. The effective T1/2 ranged from
22 to 36 hours.

Anonymized Page 60
Novartis Confidential Page 61
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Elimination
In [Study A2104] (10-mg dose), the apparent clearance of siponimod was 3.97 L/h.
The apparent terminal T1/2 of total radiolabeled components (radioactivity) and siponimod in
plasma were 171 and 56.6 hours, respectively (noncompartmental PK analysis). For M3, M5,
M6 and M7, the apparent elimination T1/2 ranged between 29.3 and 35.2 hours. For M17, the
geometric mean T/12 was 155 hours in [Study A2126] (0.25-mg dose; Table 3-7). Hence, M17
likely contributed to the longer T1/2 of total radioactivity.
Similarly, in other single and multiple dose studies in healthy subjects ([Studiy A2101 and
Study A2105]) an apparent CL/F of about 3.5 L/h was estimated, and in the 2 PopPK analyses
the typical CL/F value was 3.11 to 3.15 L/h. In the single dose study siponimod disappeared
from the systemic circulation in an apparent monophasic manner, with a geometric mean
apparent terminal T1/2 of 27.02 to 56.69 hours. In the multiple dose study, the decay in
siponimod plasma concentration over time after 28 days of dosing appeared to be biexponential
(Section 3.1.5.1). The effective T1/2 based on drug accumulation at steady state was 22 to
36 hours, comparable to the apparent terminal T1/2 in the single dose study.

Excretion of radioactivity in urine and feces


In [Study A2104], radioactivity was excreted mainly with feces (mean: up to 86.7% of dose;
[Study A2104-Figure 11-7]). The bulk of the radioactive dose was recovered within 216 hours
in feces and urine (mean: > 87% of dose). By 312 hours post dose, recovery of the radioactive
dose from the excreta was close to complete (mean: 90.4% of dose; range: 88.2% to 93.3%).
The samples collected after 312 hours post dose contained < 1% of the radioactive dose.

Metabolites in excreta
In (Study A2104), siponimod was eliminated mainly by metabolism with fecal/biliary excretion
as the major route of elimination. The main metabolic pathways identified were hydroxylation
forming M5, M6 and M7 and subsequent glucuronidation or sulfation of hydroxylated
metabolites (Figure 3-7). Other metabolites formed by cleavage/hydrolysis of the oxime ether
(sum of M1, M2 and M8 in excreta) were minor in human (≤ 2.1% of the dose;
[DMPK RCBAF312A2104c-Table 10]).
In the feces, unchanged siponimod accounted for 9.2% of the radioactive dose in the feces
during 0 to 192 hours. The major hydroxylated metabolite M5 amounted to 45.1% ± 9.0% of
the radioactive dose. M4a, M4b, M4c, M6 and M7 represented between 1.6% and 6.4% of the
excretion in feces (Study A2104); [DMPK RCBAF312A2104d]). Overall, M4, M5, M6 and
M7 covered 67.3% of the dose ([DMPK RCBAF312A2104c-Table 10]).
In urine, only a minor amount of radioactivity was excreted. Unchanged siponimod was not
detected in urine. M3, the major metabolite in urine, amounted to 2.1% of the dose
([DMPK RCBAF312A2104c-Table 10]). M1, M2 and M8 accounted for 0.2% to 0.5% of the
radioactive dose; the hydroxylated metabolites were detected in traces (< 0.1% each).

Anonymized Page 61
Novartis Confidential Page 62
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Based on the results of a study with bile-duct cannulated rats dosed intravenously with
[14C]siponimod, indicating that M3 was one of the major metabolites detected in bile while low
amounts only were observed in feces, M3 (glucuronide of M5) was likely de-conjugated to M5
in the feces [PK Written Summary-Section 5.2.4].

3.1.9 Population pharmacokinetics

3.1.9.1 Population pharmacokinetics on Phase 1/Phase 2 data


Based on previous knowledge on siponimod PK and the analysis of the pooled data in the
[CBAF312A-Phase 1-2-PopPK] from the Phase 1 studies in healthy subjects and Phase 2 study
in RRMS patients (Table 5-3), siponimod PK is well described by a 2-compartment disposition
model with first-order elimination and mixed zero- and first-order absorption. The typical
values of CL/F, Vc/F and apparent volume of peripheral compartment (Vp/F) were 3.15 L/h,
115 L and 28 L, respectively. The typical values of the absorption parameters absorption rate
constant (ka) and D1 were 0.687 h-1 and 1.7 h, respectively. All these parameters were estimated
with good precision (relative SE: 1% to 11%). The inter-individual variability (IIV) for all
parameters was high, with CV ranging between 49% and 73%.
The impact of predefined covariates was explored and the following were significant predictors
of siponimod PK: CYP2C9 genotype on CL/F and body weight on CL/F and Vc/F. Food effect
and formulation type impact the duration of the zero-order absorption into the depot
compartment, but this is not considered relevant for a chronic treatment such as that of
siponimod. There was no evidence that other covariates (gender, age, BMI, race, health status
(healthy subject or RMS patient), creatinine CL, bilirubin, AST and ALT) influence the PK of
siponimod. The main outcomes for these covariates are summarized in the relevant paragraphs
in Section 3.2 (PK in subpopulations) and Section 3.3 (drug interactions).

3.1.9.2 Population pharmacokinetics on Phase 3 data


[CBAF312A-Phase 3-PopPKPD] was conducted on Phase 3 data of 3454 siponimod
concentrations (on average, 3 per subject) from 1045 SPMS patients (Table 5-3). As there was
a sparse PK sampling schedule in this study (~3 hour post dose and trough samples), the
structural form of the PK model from the first PopPK analysis (Section 3.1.9.1) was retained
and only the following parameters were re-estimated: typical and IIV values of CL/F and Vc/F,
body weight effect on CL/F and Vc/F, as well as residual variability. The estimates of Q/F,
Vp/F, ka and D1 were fixed to the estimates from the previous analysis.
The impact of predefined covariates was explored and only weight and CYP2C9 genotypes
were significant predictors of siponimod PK. There was no evidence that other covariates
(gender, age, comedications affecting CYP2C9 or 3A4 enzymes and ethnicity (tested only for
Japanese and Chinese)) influence the PK of siponimod. The main outcomes for these covariates
are summarized in the relevant paragraphs in Section 3.2 (PK in subpopulations) and
Section 3.3 (drug interactions).

Anonymized Page 62
Novartis Confidential Page 63
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

As in the Phase 1/Phase 2 PopPK analysis, siponimod PK in the Phase 3 study was well
described by a 2-compartment disposition model with first order elimination and combined
zero- and first-order absorption. The typical values of CL/F (3.11 L/h), Vc/F (126 L) and Vp/F
(28 L), the body weight effect on CL/F and Vc/F and the CYP2C9 genotype effect on CL/F
were also consistent with the first PopPK analysis. All the parameters were estimated with good
precision (relative SE: 1% to 7%). The IIV for CL/F (CV = 24.2%) and Vc/F (CV = 33.0%)
were moderate. The shrinkages of individual random effects were estimated as 34% for CL/F
and 83% for Vc/F.

3.2 Pharmacokinetics in subpopulations

3.2.1 Patients versus healthy subjects

3.2.1.1 Multiple sclerosis


The influence of disease state on siponimod exposure was investigated in 2 clinical studies in
MS patients ([Studies A2201 and A2304]) and 2 PopPK evaluations (Section 3.1.9).
Overall, the PK of MS patients and healthy subjects do not have significant differences.
RRMS patients in the core (Study A2201) received placebo or untitrated regimens of siponimod
(0.5, 2 or 10 mg) followed by titrated regimens (0.25 or 1.25 mg) (in Section 2.3.1). The mean
siponimod plasma concentrations are provided in Table 3-9. The results suggest the PK of
siponimod in MS patients is comparable to that in healthy subjects: Cmin,ss of 15.2 ng/mL for
28 days of 2.5-mg qd treatment in [Study A2105] and 14.5 ng/mL for 5 days of 2-mg qd
treatment (preceded by 5 days of DT) in [Study A2118] (Section 3.1.5.1). In the extension
[Study A2201E1], patients received either 0.25, 0.5, 1.25, 2 or 10 mg qd dosing in the dose-
blinded phase and 2 mg qd dosing in the open-label phase. At the time of the switch to 2 mg qd
dosing at Month 24, geometric mean trough concentrations were comparable to those at the
same dose in (Studies A2105 and A2118). For all dose groups the geometric trough
concentrations remained in the range of 14.2 to 24.9 ng/mL (comparable to Study A2118)
between the post-switch 2-mg visit (2 weeks after dose change to 2 mg) and Month 54.
Additionally, in the Phase 1/Phase 2 PopPK analysis (Section 3.1.9.1), there was no evidence
of difference in the PK between healthy subjects and RRMS patients.
Similarly, in SPMS patients receiving siponimod 2 mg qd in (Study A2304), mean siponimod
concentrations (Table 3-10) were comparable to previous observations in healthy subjects
receiving 2 mg qd dosing in (Study A2118), indicating that the patients were adequately
exposed to siponimod throughout the study. Furthermore, in the Phase 3 PopPK analysis in
SPMS patients (Section 3.1.9.2), the typical values of CL/F and Vc/F were very close to those
found in the first PopPK analysis in healthy subjects and RRMS patients. This suggests that
siponimod PK in SPMS patients is comparable to that in healthy subjects and RRMS patients
and is independent from the population type. The dose-normalized steady-state PK trough
concentrations between healthy subjects (Studies A2105 and A2118) and patients with MS
(Studies A2201 and A2304) in Figure 3-8 suggest the PK of siponimod is comparable between
the populations, while a larger inter-subject variability is observed in MS patients compared to
healthy volunteers. For the 2-mg dose, the median trough concentration was approximately
8 ng/mL/mg for healthy subjects and 11 ng/mL/mg for MS patients. The scatter of individual
patients above the median could be due to the high sample size in the MS patient population

Anonymized Page 63
Novartis Confidential Page 64
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

compared to healthy subjects (92 samples). In addition, the PK samples were collected within
a larger elapsed time range in the MS trials compared to the healthy subject study.

Table 3-9 Geometric mean (percent geometric mean coefficient of variation)


siponimod plasma trough concentrations by treatment and visit in MS
patients
Visit Siponimod Siponimod Siponimod Siponimod Siponimod
10 mg 2 mg 1.25 mg 0.5 mg 0.25 mg
N=50 N=49 N=42 N=43 N=51
Month 1 114 (52) 22.6 (54) 12.8 (46) 6.05 (52) 2.68 (40)
Month 3 113 (51) 21.7 (48) 11.4 (49) 5.28 (60) 2.43 (57)
Month 6 52.7 (530) 23.7 (58) N/A 5.17 (66) N/A
N/A = not applicable.
Source: [Study A2201-Table 14.2-6.1]

Table 3-10 Geometric mean (percent geometric mean coefficient of variation)


trough concentrations by visit in MS patients (2 mg qd)
Visit Day 28 Month 3 Month 12 Month 24 EOS
Scheduled time 24h after last
point Pre-dose Post-dose Pre-dose Post-dose dose
N=1099 N=1099 N=1099 N=1099 N=1099
Siponimod (n) 1011 967 775 271 488
Geo-mean ng/mL 23.2 (51.0) 32.8 (56.9) 23.6 (53.7) 28.9 (62.3) 20.1 (66.5)
(CV% geo-mean)
Source: [Study A2304-Table 14.2-22]

Anonymized Page 64
Novartis Confidential Page 65
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-8 Box plot by dose for comparison of steady state pharmacokinetic
trough concentrations between healthy subjects and patients with
multiple sclerosis

Healthy subject studies considered: [Studies A2105 and A2118]


MS patient studies considered: [Studies A2201 and A2304]
Median = horizontal line in box; mean = symbol in box.
The upper (lower) edge of the box represents the 75th (25th) percentile.
A whisker is drawn from the upper (lower) edge of the box to the largest (smallest) value within
1.5 × interquartile range above (below) the edge of the box.
Source: (SCP Appendix-Figure 11-3.2)

3.2.1.2 Polymyositis and dermatomyositis


As discussed in Section 2.3.2, the 3 clinical studies of DT regimens of siponimod doses up to
10 mg in PM and/or DM patients (N ≤ 18 each) include [Studies A2202, X2205 and X2206].
Across the studies, siponimod steady-state plasma concentrations were generally reached at the
first PK assessment time point (taken on Day 28) or earlier (no previous sample available); this
is consistent with healthy subjects. Overall, although the inter-subject variability was generally
high, arithmetic mean siponimod trough plasma levels were comparable to previous
observations in healthy subjects at the same dose levels.
In (Study A2202), comparable siponimod exposure was observed between PM and DM patients.
The drug exposure in PM/DM patients (150 to 160 ng/mL) was approximately 2-fold higher
than expected from healthy subject studies at the same dose level of 10 mg (70.8 ng/mL in
(Study A2105); however, this observation should be considered with caution, considering the
limited patient data and the high inter-subject variability in plasma levels. In (Study X2205),
siponimod plasma exposure observed in the 2-mg group (mean trough concentrations ranging
from 21.4 to 25.3 ng/mL) was similar to that in previously healthy subject studies; for the
2 patients on 10 mg siponimod with post treatment concentrations, the siponimod plasma

Anonymized Page 65
Novartis Confidential Page 66
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

exposure was higher than observed in previous healthy subject studies ([Study A2105]). In
[Study X2206], arithmetic mean siponimod trough plasma levels were comparable to the mean
Cmin,ss values observed in [Studies A2118 and A2105] at the same dose levels, confirming
adequate exposure. In Period 1, values were more comparable to previous observations
(approximately 3.2 to 6 ng/mL in the 0.5 mg/2 mg dose group, 26.8 to 31.9 ng/mL in the
2 mg/2-mg dose group, and 53.6 to 100 ng/mL in the 10 mg/2-mg dose group), whereas
Period 2 values were more variable.

3.2.2 Hepatic impairment


Overall, the results of a single oral dose of siponimod (0.25 mg) in subjects with mild, moderate
and severe hepatic impairment in [Study A2122] suggest that no dose adjustment of siponimod
in subjects with impaired hepatic function is warranted.
The PK of siponimod was generally comparable between hepatic impairment subjects and their
healthy matched subjects ([Study A2122-Table 14.2-2.1]). Mean Cmax increased by 16% for
the mild impairment group and decreased by approximately 13% and 16% for the moderate and
severe impairment groups, respectively. Mean AUCinf and AUClast increased by 5% for the
mild impairment group and 15% for the severe impairment group, and a decrease in the mean
AUC of about 13% were observed in the moderate impairment group. Comparable mean CL/F
values were observed for the hepatic impairment groups (3.56 to 4.68 L/h) and their matched
healthy subjects (4.07 to 4.11 L/h).
In addition, no significant correlation was observed between Child-Pugh score, total bilirubin
and exposure (Cmax and AUCinf) of siponimod.
Siponimod ([14C]siponimod) was very highly bound to plasma proteins in all subjects (> 99.9%).
There was a slight increase in the fu with a decrease in hepatic function. Mean fu at 4 hours post dose
range from 0.000127 to 0.000141 in healthy subjects and 0.000150 to 0.000186 in subjects with
impaired hepatic functions. The results were in line with the in vitro fu in plasma from healthy
subjects ([DMPK R1300334]; [Table 2.6.5.6A-DMPK R0400881-01] and [DMPK R1300902-
01]; Section 3.1.6).
The unbound siponimod PK parameters were comparable in subjects with mild hepatic
impairment. Compared to matched healthy subjects, the 15% to 17% increase for subjects with
moderate hepatic impairment was not considered to be clinically relevant and the 50% increase
in subjects with severe hepatic impairment was not statistically significant due to high
variability.
Hepatically impaired subjects showed higher M3 and M5 mean Cmax and AUC compared to
healthy matched subjects for each group. M3 Cmax increased by 67%, 81% and 194% in mild,
moderate and severe groups, respectively. M3 AUClast increased by 95%, 159% and 455% in
mild, moderate and severe groups, respectively. M5 Cmax increased by 25% in the mild and
moderate groups and by 5% in the severe group. AUClast increased with the severity of the
hepatic impairment or decreased hepatic function: by 37%, 70% and 87% in mild, moderate
and severe groups, respectively, compared to matched healthy subjects.

Anonymized Page 66
Novartis Confidential Page 67
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

The increased systemic exposure of M3 and M5 associated with no significant changes in the
parent drug exposure suggests that the elimination pathway of the metabolites may be impacted
in subjects with hepatic impairment. M3, a glucuronidated metabolite, is likely a substrate of
the hepatic multidrug resistance-associated protein 2 (MRP2) transporters mediating its biliary
excretion. MRP2 is known to be down-regulated in cholestasis (Denson et al 2002). The
reduced excretion of M3 into the bile may lead to increased systemic concentrations as well as
an increase of hepatocytes M3 concentrations. In hepatocytes, M3 can possibly be
de-glucuronidated back to M5, consequently leading to an elevated systemic exposure of this
metabolite. However, as discussed in Section 3.1.7.3 and [Pharmacology Tabulated Summary-
Table 2.6.3.2-RD-2012-00145], considering the EC50 values of M3, M5 and siponimod
(> 10000, 470 and 1.1 nmol/L, respectively), the observed systemic exposure increase for both
M3 and M5 in hepatic impairment is unlikely to translate into any increase in pharmacological
activity on S1P1 receptors.
There was no rationale in the current study or in [Study A2129] (in renally impaired subjects)
to investigate M3/M5 protein binding, since these 2 metabolites are neither pharmacologically
active nor toxic.

3.2.3 Renal impairment


Overall, the results of a single oral dose of siponimod (0.25 mg) in subjects with severe renal
impairment in (Study A2129) suggest that no dose adjustment of siponimod in subjects with
impaired renal function is warranted.
Siponimod Cmax and AUC were marginally affected in renally impaired subjects, with slightly
lower Cmax (8%) and a 23% to 24% increase in AUC in severe renal impaired subjects
compared to healthy matched subjects; however, these differences were not clinically
meaningful. The mean apparent terminal T1/2 of siponimod was 25.4 and 36.2 hours in healthy
and severe renal impairment subjects, respectively ([Study A2129-Table 14.2-1.2]).
Siponimod ([14C]siponimod) was very highly bound (> 99.9%) to human plasma proteins in all
subjects. The fu (%) ranged between 0.0172% and 0.0550% in individual subjects, with no
relevant differences between the 2 groups. The results were in line with the in vitro fu in plasma
from healthy subjects ([Table 2.6.5.6A-DMPK R0400881-01 and DMPK R1300902-01];
Section 3.1.6).
Similar mean Cmax(u) of unbound siponimod was observed for severe renal impairment
subjects and matched healthy subjects. While not clinically meaningful, mean AUC(u)
increased by 33% in severe renal impairment subjects compared to healthy matched subjects.
Comparable M3 exposure was observed between subjects with severe renal impairment and
matched healthy subjects. Mean Cmax decreased by only 9% and AUClast and AUCinf
increased by 11% in the severe renal impairment subject group as compared to the matched
healthy subjects group. A slightly lower M5 exposure was observed in subjects with renal
impairment. Mean Cmax of M5 decreased by 26% and mean AUClast decreased by 16% as
compared to the matched control group.

Anonymized Page 67
Novartis Confidential Page 68
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

The data indicate that in subjects with severe renal impairment the relevant non-renal clearance
pathways of siponimod are not affected and metabolism remains unchanged, as also confirmed
by M3 and M5 data. The same results are expected for the mild and moderate renal impairment
stages.

3.2.4 Ethnic origin


The evaluations of race/ethnicity on siponimod PK in [Study A1101] (Japanese subjects) and
2 PopPK analyses ([CBAF312A-Phase 1-2-PopPK] and [CBAF312A-Phase 3-PopPKPD])
suggest that race/ethnicity does not significantly affect siponimod PK.
The PK profile of siponimod (0.5 to 10 mg) in Japanese subjects in (Study A1101) was similar
to the previous SAD [Study A2101] in non-Japanese subjects. Siponimod was measurable in
plasma as early as 0.25 hours post dose. The median Tmax across doses was 3.5 to 4 hours,
ranging between 2 to 12 hours in individual subjects. Cmax and AUC appeared to increase
dose-proportionally over the range investigated (Section 3.1.3.1). The inter-subject variability
as %CV geometric mean values ranged between 13% and 21% for Cmax and between 8% and
31% for AUC. The decline of siponimod plasma concentration was mono-exponential with a
geometric mean T1/2 of between 28.5 and 39.7 hours.
In addition, in the Phase 1/Phase 2 PopPK analysis (Section 3.1.9.1), race was tested as a
possible covariate influencing siponimod CL/F or Vc/F. Based on the data from 298 Caucasians,
21 blacks, 32 Japanese and 55 subjects of another race, there was no evidence of race impacting
siponimod PK. Similarly, in the Phase 3 PopPK analysis of SPMS patients (Section 3.1.9.2),
based on a small number of subjects (N = 14 for each ethnicity), Japanese and Chinese
ethnicities had no significant impact on siponimod PK.

3.2.5 Genotype
The polymorphic enzyme CYP2C9 is the major metabolizing enzyme for siponimod. Exposure
to siponimod is dependent on the CYP2C9 genotype. More than 50 single nucleotide
polymorphisms (SNPs) have been described in the regulatory and coding regions of the
CYP2C9 gene, but only 2 coding SNPs, namely CYP2C9*2 and CYP2C9*3, are considered to
lead to a clinically relevant reduction in enzyme activity. These 2 SNPs lead to 6 different
CYP2C9 genotype groups conferring to 3 functionally different phenotypes, namely extensive
(*1*1), intermediate (*1*2, *1*3 and *2*2) and poor metabolizers (*2*3 and *3*3). Table 3-
11 presents the prevalence of the 6 genotypes in Caucasian, Asian and African subjects.
Section 3.2.5.1 discusses differences in siponimod exposure between extensive and poor
metabolizers (CYP2C9*1*1 genotype versus *2*3 or *3*3 genotypes, respectively) in
[Study A2128]. Section 3.2.5.2 describes differences in siponimod PK across additional
CYP2C9 genotypes in the 2 PopPK analyses and [Study A2124]. Section 3.2.5.3 recommends
maintenance dose adjustments from 2 mg to 1 mg in CYP2C9*1*3 and *2*3 subjects with
reduced apparent metabolic clearance and exclusion of *3*3 subjects (not evaluated in Phase 3
[Study A2304], due to expected significantly higher chronic exposure with potential safety
implications) from a marketing authorization.

Anonymized Page 68
Novartis Confidential Page 69
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-11 Genotype frequencies of CYP2C9 polymorphisms in Caucasian, Asian


and African subjects
Predicted Population frequency (%)
metabolizer/CYP2C9 Caucasian subjects Asian (Japanese) African subjects
genotype subjects
Extensive metabolizers
*1*1 62-65 85-97 (95-96) 75-87
Intermediate metabolizers
*1*2 20-24 0-5 (0) 5-9
*1*3 9-12 4-7 (4-5) 4
*2*2 1-2 0-0.1 (0) 0-0.1
Poor metabolizers
*2*3 1.4-1.7 0-0.2 (0) 0-0.1
*3*3 0.3-0.4 0-0.2 (0-0.1) 0
Source: Kirchheiner and Brockmoller 2005, Scott et al 2010, Lee et al 2002, Gaikwad et al 2014

3.2.5.1 CYP2C9 extensive and poor metabolizers


An exploratory PK/pharmacogenetic (PG) analysis in the first-in-human study indicated that
heterozygous CYP2C9*3 carriers tend to have a higher AUC of siponimod compared to
subjects not carrying the *3 allele. Consequently, the PK of siponimod was assessed in CYP2C9
extensive and poor metabolizers (CYP2C9*1*1 genotype and CYP2C9*2*3 or *3*3 genotypes,
respectively) in [Study A2128] following a single 0.25-mg dose of siponimod in Part 1 and
0.25-mg (Days 1 and 2) and 0.5-mg (Day 3) doses in Part 2 (Table 5-1). Overall, in Part 1 the
siponimod AUCinf and AUClast was higher in poor than in extensive metabolizers. A minor
increase of Cmax was observed in poor metabolizers. In Part 2, subjects with CYP2C9*2*3 and
*3*3 genotypes were comparably exposed to siponimod. For metabolites M3 and M5, in Part 1
the Cmax and AUC were markedly lower and Tmax delayed in poor metabolizers; the apparent
terminal T1/2 was also prolonged. In Part 2, the M3 and M5 Cmax and AUC0-24h were lower
in CYP2C9*3*3 than CYP2C9*2*3 subjects, while the median Tmax was similar.

3.2.5.1.1 Pharmacokinetics of siponimod


Siponimod was rapidly absorbed in Part 1, with a median Tmax of 4 to 5 hours for all
3 genotypes ([Study A2128-Figure 11-1], Table 3-12). Siponimod mean Cmax increased by 21%
and 16% in CYP2C9*2*3 and *3*3 subjects, respectively, compared to CYP2C9*1*1 subjects;
however, these differences were not deemed clinically relevant. Mean AUCinf and AUClast
increased by 2.05-fold in CYP2C9*2*3 genotype subjects and by 3.84- to 3.87-fold in
CYP2C9*3*3 subjects as compared to their matched control group. The inter-subject variability
(%CV) for Cmax was comparable between CYP2C9*1*1 and *2*3 subjects (approximately 17%
and 13%, respectively) and slightly higher for *3*3 subjects (29%). The inter-subject variability
measured for AUC was between 16% and 23% for all 3 genotypes. The mean apparent terminal
T1/2 was prolonged in poor metabolizers (50.9 and 126 hours in CYP2C9*2*3 and *3*3
subjects, respectively, compared to 28.1 hours in *1*1 subjects). The mean CL/F was reduced
in CYP2C9*2*3 (1.73 L/h) and *3*3 (0.923 L/h) subjects as compared to their matched control
group (3.55 L/h).

Anonymized Page 69
Novartis Confidential Page 70
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

In Part 2, Day 3 PK parameters cannot be considered as steady state PK parameters considering


the long elimination T1/2 subjects with CYP2C9*2*3 and *3*3 genotypes in Part 1. Subjects
with CYP2C9*2*3 and *3*3 genotypes were comparably exposed to siponimod with a
geometric mean Cmax between 6.87 and 7.54 ng/mL, respectively, and a mean AUC0-24h of
125 to 148 h*ng/mL, respectively. (For the same dosing regimen for the DT2 group in
[Study A2107], at Day 3 a similar Cmax was observed (geometric mean of 5.08 ng/mL) while
the AUClast geometric mean was 88.90 h*ng/mL.) As in Part 1, the inter-subject variability
(%CV) for Cmax and AUC0-24h was low to moderate and comparable between CYP2C9*2*3
and *3*3 carriers (approximately 17% to 26%). Median Tmax appeared to be slightly delayed
(6 hours) in Part 2 for both CYP2C9*2*3 and *3*3 genotypes compared to Part 1 (4 to 5 hours).

Table 3-12 Summary statistics of siponimod pharmacokinetic parameters per


CYP2C9 genotype after a single 0.25-mg dose in Study A2128 Part 1
(pharmacokinetic analysis set)
PK parameter (unit) CYP2C9 *1*1 CYP2C9 *2*3 CYP2C9 *3*3
(N = 12) (N = 6) (N = 6)
Cmax (ng/mL)b 2.03 (17.2) 2.45 (13.1) 2.35 (29.0)
Tmax (h)a 4.00 (2.00-6.00) 5.00 (4.00-8.00) 4.00 (4.00-16.00)
AUClast (h*ng/mL)b 68.6 (21.3) 140 (15.6) 266 (22.9)
AUCinf (h*ng/mL)b 70.5 (21.2) 144 (15.7) 271 (22.4)
AUC0-24h (h*ng/mL)b 31.6 (17.0) 44.7 (9.4) 42.7 (26.6)
T1/2 (h)b 28.1 (18.5) 50.9 (31.7) 126 (12.7)
Tlag (h)a 0.250 (0.000-0.500) 0.250 (0.000-0.500) 0.500 (0.000-0.500)
Vz/F (L)b 144 (18.6) 127 (20.9) 167 (24.2)
CL/F (L/h)b 3.55 (21.2) 1.73 (15.7) 0.923 (22.4)
a Median (min-max)
b Geometric mean [% geometric mean coefficient of variation]
Source: [Study A2128-Table 14.2-2.2a]

3.2.5.1.2 Pharmacokinetics of M3 and M5


In Part 1, the PK of M3 and M5 in subjects with CYP2C9*1*1 genotype was consistent with
historical study data in [Study A2129]. Cmax and AUC of M3 and M5 were markedly lower in
poor metabolizers. For M3, mean Cmax decreased by 49% and 88% in CYP2C9*2*3 and *3*3
subjects, respectively, compared to CYP2C9*1*1 subjects. M3 mean AUCinf and AUClast
decreased by 12% to 23%, respectively, in CYP2C9*2*3 carriers and by 66% to 76%,
respectively, in *3*3 subjects as compared to their matched control group. For M5, mean Cmax
decreased by 47% and 74% in CYP2C9*2*3 and *3*3 subjects, respectively, compared to *1*1
subjects. M5 mean AUCinf and AUClast decreased by 38% in CYP2C9*2*3 carriers and 95%
in *3*3 subjects as compared to their matched control group.

Anonymized Page 70
Novartis Confidential Page 71
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

For M3, the median Tmax was 12 and 24 hours in subjects with CYP2C9*2*3 and *3*3
genotypes compared to 6 hours in *1*1 carriers. A similar delay in median Tmax was observed
in M5 poor metabolizers: 7 and 48 hours in CYP2C9*2*3 and *3*3 genotypes, respectively,
compared to 4 hours in *1*1 carriers. Overall, metabolite mean apparent terminal T1/2 was
prolonged in poor metabolizers. For M3, T1/2 was 54.8 and 96 hours in CYP2C9*2*3 and *3*3
subjects compared to 32.9 hours in *1*1 subjects. For M5, T1/2 was 105 hours in CYP2C9*2*3
subjects compared to 37.9 hours in *1*1 subjects. The M5 T1/2 could not be determined in
CYP2C9*3*3 subjects due to the very low plasma level allowing for only very few observations
48 hours post dose.
In Part 2, while subjects with CYP2C9*2*3 and *3*3 genotypes were comparably exposed to
siponimod, the Cmax and AUC0-24h for M3 and M5 were lower (range: 4- to 5-fold) in
CYP2C9*3*3 subjects compared to CYP2C9*2*3 carriers. For M3, median Tmax was 8 hours
for both groups. For M5, median Tmax was 10 hours in CYP2C9*2*3 subjects and 7 hours in
the *3*3 carriers.

3.2.5.2 Other CYP2C9 genotypes


The 2 PopPK analyses (Section 3.1.9) identified CYP2C9 genotype as a significant predictor
of siponimod CL/F. Consistent with the PopPK findings, differences in siponimod
PK parameters, including CL/F, were noted between CYP2C9*1*2 and *1*3 genotype subjects
in [Study A2124] (study design in Table 5-1).
The different CYP2C9 genotypes of the 268 subjects in the Phase 1/Phase 2 analysis and
1045 SPMS patients in the Phase 3 analysis are shown in Table 3-13. Patients with a
CYP2C9*3*3 genotype were not included in Phase 3 [Study A2304] due to expected
significantly higher chronic siponimod exposure with potential safety implications. The results
of the 2 analyses, conducted on 2 different populations, are very consistent. There were no
apparent differences in CL/F between CYP2C9*1*1 and *1*2 subjects (Table 3-14). Compared
to these subjects, individuals with the *2*2, *1*3 and *2*3 heterotypes had 20%, 35% to 38%
and 45% to 48% smaller CL/F values, respectively; individuals with the *3*3 genotype had a
74% smaller clearance. These results are consistent with the clearance values in Table 3-12 for
CYP2C9*1*1, *2*3 and *3*3 subjects in [Study A2128]. Based on this data, a phenotypic
classification is proposed in Table 3-14 to relate to CYP2C9 genotypes and siponimod
metabolism performance. Overall, CYP2C9*1*2 subjects behave like extensive, rather than
intermediate metabolizers regarding elimination of siponimod.

Table 3-13 Number of individuals (percent) with different CYP2C9 genotypes


across studies (population pharmacokinetic analysis sets)
CYP2C9 Genotype/PopPK *1*1 *1*2 *1*3 *2*3 *2*2 *3*3a
analysis
Phase 1/Phase 2 190 (71) 41 (15) 28 (10) 5 (2) 2 (0.7) 2 (0.7)
Phase 3 664 (64) 203 (19) 131 (13) 29 (3) 18 (2) -
a Two *3*3 were taken from the initial genotype analysis performed during the study inclusion
Source: [CBAF312A-Phase 1-2-PopPK-Table 5-4], [CBAF312A-Phase 3-PopPKPD-Table 5-1]

Anonymized Page 71
Novartis Confidential Page 72
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-14 Estimated siponimod CL/F per CYP2C9 genotype in the two
population pharmacokinetic analyses for a typical 69 to 70 kg subject

CYP2C9 Estimated CL/F (L/h) % of CYP2C9*1*1 CL/F


Genotype Phase 1/Phase 2 Phase 3 Phase 1/Phase 2 Phase 3
Extensive metabolizers
CYP2C9*1*1 3.3 3.1 100 100
CYP2C9*1*2 3.3a 3.1 99a 100
Intermediate metabolizers
CYP2C9*2*2 2.6a 2.5 80a 80
CYP2C9*1*3 2.1 1.9 65 62
Poor metabolizers
CYP2C9*2*3 1.8 1.6 55 52
CYP2C9*3*3 0.9 - 26 -
a Parameters estimated with a relative SE much higher than 50%
Source: [CBAF312A-Phase 1-2-PopPK-Table 5-12], [CBAF312A-Phase 3-PopPKPD-Table 5-9]

Differences in the PK of CYP2C9*1*2 and *1*3 subjects following a single 0.25-mg dose of
siponimod in [Study A2124] are consistent with the PopPK results. On Day 1, siponimod PK
in CYP2C9*1*2 subjects (Table 3-18) was consistent with historical data ([Study A2111]).
Siponimod was rapidly absorbed in CYP2C9*1*2 subjects, with a median Tmax of 4 hours. In
CYP2C9*1*3 subjects, median Tmax was slightly delayed (6 hours). While geometric mean
Cmax was comparable between CPY2C9*1*2 and *1*3 subjects, geometric mean AUCinf and
AUClast were approximately 29% higher in *1*3 compared to *1*2 subjects. Accordingly,
siponimod geometric mean T1/2 was longer and CL/F smaller in CYP2C9*1*3 versus *1*2
subjects (39.9 versus 27.2 hours and 2.58 L/h versus 3.33 L/h, respectively).

3.2.5.3 CYP2C9 genotype-based dosing approach


The dependence of the apparent metabolic clearance on CYP2C9 genotype affects subject
exposure to siponimod. Dose adjustments are to be considered for subjects with reduced
apparent metabolic clearance compared to that in CYP2C9*1*1 subjects, to avoid untoward
safety findings due to chronic high exposure.
As discussed above, reduced clearance is expected in subjects with the following CYP2C9
genotypes: *2*2, *1*3, *2*3 and *3*3. In CYP2C9*1*3 and *2*2 intermediate metabolizers,
estimations that systemic apparent clearance is reduced by approximately 35% to 38% and
approximately 20% (Section 3.2.5.2, Table 3-14), respectively, corresponds to an increase in
siponimod exposure by approximately 61% and 25%, respectively, compared to *1*1 extensive
metabolizers ([CBAF312A-Phase 1-2-PopPK], [CBAF312A-Phase 3-PopPKPD]).
In CYP2C9*2*3 and *3*3 subjects, systemic apparent clearance is estimated to be reduced by
approximately 45% to 48% and approximately 74% (Section 3.2.5.2, Table 3-14), respectively,
and exposure in [Study A2128] was found to be approximately 2-fold and 4-fold higher (Section
3.2.5.1, Table 3-12), respectively, compared to *1*1 subjects.

Anonymized Page 72
Novartis Confidential Page 73
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Based on this data, a maintenance dose of 1 mg instead of 2 mg is proposed in subjects with


CYP2C9*1*3 and CYP2C9*2*3 genotypes (Table 3-15). The 1 mg maintenance dose is
expected by Novartis to adjust for the respective reduced apparent metabolic clearance
compared to CYP2C9*1*1 subjects and achieve an exposure that is comparable to that of
CYP2C9*1*1 subjects receiving a 2 mg qd dose, and thus avoiding potential long-term safety
risks of chronic higher exposure. No dose adjustment is proposed for CYP2C9*1*2 and *2*2
subjects considering their comparable or only slightly reduced predicted systemic clearance
compared to *1*1 subjects. As CYP2C9*3*3 subjects have not been studied in the Phase 3
[Study A2304], due to expected significantly higher chronic exposure, Novartis proposes to
exclude such subjects from a marketing authorization.

Table 3-15 Recommended siponimod maintenance doses per CYP2C9 genotype


CYP2C9 Genotype Calculated maintenance dose Recommended maintenance
(mg)a dose (mg)
Extensive metabolizers
CYP2C9*1*1 2 2
CYP2C9*1*2 2 2
Intermediate metabolizers
CYP2C9*2*2 1.6 2
CYP2C9*1*3 1.24 1
Poor metabolizers
CYP2C9*2*3 1.04 1
a Calculated according to CL/F values
Source: [CBAF312A-Phase 1-2-PopPK-Table 5-12], [CBAF312A-Phase 3-PopPKPD-Table 5-9]

3.2.6 Gender
The CL/F and Vc/F data in 224 males and 182 females in the Phase 1/Phase 2 PopPK analysis
(Section 3.1.9.1) and the CL/F data in 413 males and 632 females in the Phase 3 PopPK analysis
of SPMS patients (Section 3.1.9.2) suggest that gender had no significant impact on these
siponimod PK parameters.

3.2.7 Body weight


Body weight was shown to influence siponimod CL/F and Vc/F in the 2 PopPK analyses.
However, considering the limited effect on siponimod exposure, no dose adjustment based on
body weight is warranted.
In the Phase 1/Phase 2 PopPK analysis (Section 3.1.9.1), the effect of weight on Vc/F and CL/F
was implemented linearly as shown below for CL/F:
CL i  θ CL  1  θ WT on CL  ( WT  WT med )   exp η CL, i ,
with CL as the typical CL/F, WT on CL as the weight effect on CL/F, WT as the individual
weight, WTmed as the median weight (70.5 kg) and CL,i as the individual random effect.

Anonymized Page 73
Novartis Confidential Page 74
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

The weight effects on CL/F and Vc/F were 0.0085 and 0.0069 kg-1, respectively. For the range
of weights in the analysis (42 to 126 kg), CL/F ranged from 18% lower to 28% higher and Vc/F
from 24% lower to 47% higher compared to a 70.5-kg individual.
In the Phase 3 PopPK analysis of SPMS patients (Section 3.1.9.2), both Vc/F and CL/F
increased with weight with the power coefficient of 1 and 0.75, respectively (i.e., as (BW/70)1
and (BW/70)0.75). For the range of weights in the population (40 to 142 kg), Vc/F ranged from
43% lower to 101% higher and CL/F ranged from 35% lower to 69% higher compared to a
70.5-kg individual. Subjects with a body weight of 40 and 142 kg have a 53% higher and 40%
lower exposure, respectively, compared to subjects with a body weight of 70.5 kg. Considering
the limited effect on siponimod exposure, no dose adjustment is warranted to compensate for
the body weight effect.

3.2.8 Elderly (greater than or equal to 65 years old)


There was no study conducted in elderly subjects. In the 2 PopPK analyses, age (range assessed:
18 to 61 years) was not identified as a covariate affecting siponimod CL/F. As hepatic and renal
impairment do not impact siponimod PK (Section 3.2.2 and Section 3.2.3, respectively),
differences in siponimod PK are not expected between elderly and younger subjects.

3.2.9 Children (less than 18 years old)


Not applicable.

3.3 Drug interactions


The primary focus of this section is the results of in vitro and clinical studies of potential DDI
effects on siponimod PK, while DDI results from clinical studies primarily evaluating PD
effects of comedications (T cell-dependent and T cell-independent vaccines, monophasic OC,
propranolol) are detailed in Section 2.2.7.2. Additionally, this section discusses predictions of
PBPK modeling with CYP2C9/3A4 inducers/inhibitors in subjects/patients with different
CYP2C9 genotypes, and recommendations for siponimod CYP2C9 genotype-based DDI
management.
Figure 3-9 provides an overview of the effect of concomitant medications on siponimod PK and
the effect of siponimod on comedication PK. Geometric mean ratios (fold change) and 90% CI
for siponimod and comedication Cmax and AUC are displayed. The clinical relevance of these
results is given by the recommendations in the figure. The results of the individual studies are
detailed in Section 2.2.7.2 and Section 3.3.1.2.2.

Anonymized Page 74
Novartis Confidential Page 75
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-9 Effect of concomitant medications on siponimod PK and effect of siponimod on comedication PK (geometric
mean ratios and 90 percent confidence intervals)

Source: [Study A2108-Table 11-4], [Study A2125-Table 11-9], [Study A2124-Table 11-13], [Study A2121-Table 11-3], [Study A2116-Table 11-5].

Anonymized Page 75
Novartis Confidential Page 76
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.3.1 Pharmacokinetic interactions

3.3.1.1 Siponimod as a causative agent of interaction

3.3.1.1.1 In vitro data


Overall, siponimod and its major systemic metabolites M3 and M17 do not show any clinically
relevant DDI potential at the therapeutic dose of 2 mg qd for all investigated CYP enzymes and
transporters, and do not necessitate clinical investigation. Table 5-5 provides a summary of all
in vitro inhibition and/or induction data on human metabolic enzymes and drug transporter
proteins. All enzymes and transporters for which IC50/Ki, KI and kinact values were reported
for siponimod and M3 are discussed, including, if pertinent, an overview about the possible
clinical consequences at the different organ levels assessed using models described in health
authority guidelines for the investigation of drug interactions (FDA 2012 and EMA 2012). All
concentrations used in calculations as well as the enzymes and transporters without relevant
inhibition or induction are listed in the footnotes below the table, including all tested enzymes
for M17.

3.3.1.1.2 Clinical studies


Not applicable. Siponimod does not significantly inhibit or induce CYP450 enzymes in vitro
and has a low potential to cause transporter-mediated DDIs (Section 3.3.1.1.1).

3.3.1.2 Siponimod as an object of interaction


The fractional elimination pathway contributions in human, as currently anticipated from
available animal and human ADME studies, are depicted in Figure 3-10. Hence, for all DDI
risk assessments the contribution of hepatic clearance to total in vivo clearance is anticipated at
100% and, in consequence, no DDI effects at the level of renal and intestinal secretion or
extrahepatic metabolism are expected.

Anonymized Page 76
Novartis Confidential Page 77
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-10 Schematic description of the drug disposition pathways for


siponimoda

a Oral absorption of siponimod was estimated based on metabolite patterns in excreta and
extrapolation to total administered dose, assuming that the drug and metabolites are stable against
intestinal bacterial enzymes.
b Based on SimCYP simulations ([DMPK R1600759-01])
c Fractional contribution of CYP enzymes based on phenotyping data ([DMPK R0500432])
Source: DMPK R1600759-01

3.3.1.2.1 In vitro data


Table 5-6 summarizes in vitro human phenotyping information for siponimod and provides a
summary about the resulting interaction potential at the level of absorption, distribution and
elimination evaluated according to health authority recommendations (FDA 2012, EMA 2012).
A detailed PBPK investigation on the DDI of siponimod as a victim of CYP inhibitors and
inducers is described in the SimCYP report (DMPK R1600759-01) (Section 3.3.1.2.3). In
human, no significantly active or toxic siponimod metabolites were identified (Section 3.1.7).
Consequently, metabolites were not further elucidated for their interaction potential as
enzyme/transporter substrates.
Overall, CYP2C9 is the major metabolizing enzyme for siponimod (79.3%), followed by
CYP3A4 (18.5%). CYP2C9 is polymorphic and the CYP2C9 genotype influences the fractional
contributions of the 2 oxidative metabolism pathways to overall elimination. The DDI effect in
the presence of CYP3A or CYP2C9 perpetrator drugs will be highly variable and mainly
dependent on the CYP2C9 genotype.

Anonymized Page 77
Novartis Confidential Page 78
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.3.1.2.2 Clinical studies


Based on the in vitro data summarized in Section 3.3.1.2.1, 3 clinical DDI studies were
conducted to assess the effect on siponimod PK of inhibitors and inducers of the main CYP450
enzymes involved in its metabolism. Co-administration of rifampin (a CYP2C9/3A4 inducer)
in CYP2C9*1*1 subjects decreased exposure of siponimod (Cmax, AUC) but not M3 or M5
([Study A2125]). In CYP2C9*1*1 subjects upon co-administration with fluconazole (a
CYP2C9 inhibitor) siponimod AUC was increased and the apparent terminal T1/2 was
prolonged ([Study A2108]). While an increased siponimod exposure was expected, co-
administration of itraconazole (CYP3A4 inhibitor) led to a decrease in siponimod exposure.
Varying degrees of decreases in AUC observed in CYP2C9*1*2 and *1*3 subjects suggest a
possible influence of CYP2C9 genotype on the effect of itraconazole (CYP3A4 inhibitor)
co-administration ([Study A2124]). Clinical DDI management, based on subject CYP2C9
genotype and the comedication effect on CYP enzymes, is discussed in Section 3.3.1.2.4.

Rifampin, moderate CYP2C9 and strong CYP3A4 inducer:


In the 2-period (Study A2125), subjects received siponimod uptitrated to 2 mg before
co-administration of rifampin in Period 2 (Table 5-1). As expected, rifampin significantly
decreased the steady-state exposure of siponimod (Table 3-16). The Cmax,ss of siponimod was
decreased by 45% while AUCtau,ss was decreased by 57% in the presence of rifampin.
Siponimod trough concentrations decreased over Period 2 when rifampin was co-administered
and appeared to reach a new steady-state level between Days 22 and 24 in most subjects.
This observation confirms that the maximum enzyme induction by rifampin was achieved on
Day 24 when PK assessments were performed. The results indicated that while the Cmax,ss of
M3 increased significantly in the presence of rifampin, the AUCtau,ss of M3 displayed little
change only. M5 Cmax,ss did not change under co-administration of rifampin, but the
AUCtau,ss of M5 decreased significantly. Therefore, while induction of CYP2C9 and CYP3A4
did significantly increase the clearance of siponimod and increase the peak exposure to M3,
rifampin did not increase the total exposure of M3 and M5.

Anonymized Page 78
Novartis Confidential Page 79
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-16 Summary statistics of pharmacokinetic parameters for


co-administration of siponimod and CYP2C9/3A4 inducer rifampin -
Days 12 and 24 (pharmacokinetic analysis set)
Siponimod Alone, Siponimod + Rifampin
Study Day 12a Study Day 24b
PK Parameter (Unit) N = 16 N = 15
AUClast,ss (h*ng/mL)c 546 (17.6%)e 235 (16.9%)
AUCtau,ss (h*ng/mL)c 546 (17.6%)e 235 (16.9%)
Cmax,ss (ng/mL)c 28.6 (17.1%)e 15.7 (17.9%)
Cmin,ss (ng/mL)c 15.6 (20.6%)e 4.32 (58.6%)
Cav,ss (ng/mL)c 22.8 (17.6%)e 9.80 (16.9%)
Tmax,ss (h)d 4.00 (1.50-8.00)e 4.00 (1.50-4.05)
Fluc (%)c 55.5 (23.4%)e 112 (15.9%)
N = number of subjects who received treatment in the study period.
a Period 1: Siponimod uptitrated from 0.25 to 2 mg qd (Days 1 through 12).
b Period 2: Siponimod 2 mg qd + Rifampin 600 mg qd. (Days 13 through 24).
c Geometric mean (% geometric mean coefficient of variation)
d Median (minimum-maximum)
e Subjects with non-missing values: n = 15
Source: [Study A2125-Table 11-4]

Fluconazole, moderate CYP2C9/CYP3A4 inhibitor:


In the 2-period [Study A2108], subjects received a 4 mg single dose of siponimod before
co-administration of fluconazole (dosing details in Section 2.2.7.1) in Period 2 (Table 5-1).
Plasma concentrations of siponimod in presence of fluconazole were overall higher than those
of siponimod when administered alone (Table 3-17). Co-administration of siponimod and
fluconazole led to approximately 2-fold increases of AUClast and AUCinf of siponimod in
plasma. Cmax increased by only 10%. CL/F was decreased by approximately 2-fold and
apparent terminal T1/2 during fluconazole treatment increased by 50% compared to siponimod
alone. Tmax or Tlag remained unchanged, thus suggesting that fluconazole did not affect the
absorption phase of siponimod.

Anonymized Page 79
Novartis Confidential Page 80
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-17 Main pharmacokinetic parameters for co-administration of siponimod


and CYP2C9 inhibitor fluconazole
PK Parameter (Unit) Period 1 (siponimod alone) Period 2 (siponimod + fluconazole)
(N = 14) (N = 11)
AUClast (h*ng/mL)b 1111 (23.74) 2164 (31.57)
AUCinf (h*ng/mL)b 1116 (23.75) 2190 (32.07)
Cmax (ng/mL)b 31.2 (20.03) 33.98 (19.77)
T1/2 (h)b 40.63 (16.47) 61.65 (12.25)
Tmax (h)b 3.908 (36.53) 4.941 (33.32)
Tlag (h)a 0.25 (0.0- 0.50) 0.0 (0.0-0.0)
CL/F (L/h)b 3.586 (23.75) 1.826 (32.07)
Vz/F (L)b 210.2 (30.66) 162.4 (23.22)
a Median (minimum–maximum)
b Geometric mean (% geometric mean coefficient of variation)
Source: [Study A2108-Tables 14.2-1.2 and 14.2-1.3]

Itraconazole, selective and strong CYP3A4 inhibitor:


In the 3-period [Study A2124], CYP2C9*1*2 and CYP2C9*1*3 genotype subjects (for whom
a larger contribution of the CYP3A4 pathway to the total metabolic clearance is predicted)
received a single 0.25-mg dose of siponimod in Period 1 (Day 1), itraconazole in Period 2
(Days 15 to 18), and a second single dose of siponimod (Day 19) with continued itraconazole
administration in Period 3. For siponimod-only treatment on Day 1, the PK of siponimod in
CYP2C9*1*2 subjects (Table 3-18) was consistent with historical data; differences observed
between CYP2C9*1*2 and *1*3 subjects are discussed in Section 3.2.5.2.
Upon co-administration with itraconazole, siponimod geometric mean AUClast and AUCinf
were decreased by 9% to 10% and 24% in CYP2C9*1*2 and *1*3 subjects, respectively.
Geometric mean CL/F only marginally increased in CYP2C9*1*2 as compared to Day 1
(3.78 versus 3.33 L/h); a slightly larger increase was observed in *1*3 subjects (3.42 versus
2.58 L/hr). Co-administration of siponimod with itraconazole did not alter the geometric mean
Cmax or median Tmax (4 hours) in either CYP2C9 genotype. Overall, the decreases in AUC
observed in CYP2C9*1*2 and *1*3 subjects in Period 3 indicate a possible 2C9 genotype
influence on the itraconazole effect.
Additionally, M3 Cmax decreased by 19% to 20%, AUCinf decreased by 24% and AUClast
decreased by 23% to 25% in both CYP2C9 genotypes in the presence of itraconazole. For M5,
Cmax decreased by 11% and 16% in CYP2C9*1*2 and *1*3 subjects, respectively, AUCinf
was reduced by 17% in *1*2 subjects, and AUClast decreased by 21% and 36% in *1*2 and
*1*3 subjects, respectively. For M17, Cmax decreased by 28% and 38% and AUClast decreased
by 58% and 77% in *1*2 and *1*3 subjects, respectively.

Anonymized Page 80
Novartis Confidential Page 81
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-18 Siponimod pharmacokinetic parameters on Day 1 (siponimod alone)


and Day 19 (siponimod co-administered with itraconazole)
Siponimod 0.25 mg sd +
PK Parameter (Unit) Siponimod 0.25 mg sd itraconazole 100 mg bid
Cohort 1 N=17 N=16
(CYP2C9*1*2)
AUCinf (h*ng/mL)b 75.0 (26.7) 66.2 (29.4)
AUClast (h*ng/mL)b 71.4 (27.4) 63.6 (30.3)
CL/F (L/h)b 3.33 (26.7) 3.78 (29.4)
Cmax (ng/mL)b 1.91 (24.6) 1.91 (23.6)
T1/2 (h)b 27.2 (23.1) 24.6 (20.0)
Tmax (h)a 4.00 (3.00-8.00) 4.00 (3.00-8.00)
Vz/F (L)b 131 (16.4) 134 (19.9)
Cohort 2 N=13 N=13
(CYP2C9*1*3)
AUCinf (h*ng/mL)b 96.8 (25.7) 73.2 (27.2)
AUClast (h*ng/mL)b 92.2 (26.4) 69.8 (27.4)
CL/F (L/h)b 2.58 (25.7) 3.42 (27.2)
Cmax (ng/mL)b 1.81 (15.3) 1.70 (15.6)
T1/2 (h)b 39.9 (23.9) 31.1 (23.5)
Tmax (h)a 6.00 (3.00-12.0) 6.00 (3.00-8.00)
Vz/F (L)b 148 (23.1) 153 (15.2)
sd = single dose; N = number of subjects who received treatment.
a Median (minimum-maximum)
b Geometric mean (% geometric mean coefficient of variation)
Source: [Study A2124-Table 14.2-1.2]

Population pharmacokinetic investigations CYP2C9 and CYP3A4 inducers and


inhibitors:
In the [CBAF312A-Phase 3-PopPKPD] analysis conducted on data from SPMS patients
(Section 3.1.9.2), a sub-analysis including the Phase 2 [Study A2201] in RRMS patients
investigated the effect of comedications affecting the activity of CYP2C9 or CYP3A4 on
siponimod PK. Out of 1257 subjects (total of 3933 siponimod concentrations, measured while
on concomitant medication or not more than 7 days after), the following numbers of subjects
were on CYP2C9 and CYP3A4 inducer(s) or inhibitor(s):
 13 on CYP2C9 inducer(s) and 38 on CYP3A4 inducer(s) (< 1% and 2% of siponimod
concentrations, respectively).
 192 on CYP3A4 weak inhibitor(s), 21 on moderate CYP3A4 inhibitor(s), 12 on strong
CYP3A4 inhibitor(s) and 1 on CYP2C9 inhibitor(s) (13%, < 1%, < 1% and < 1% of
siponimod concentrations, respectively).

Anonymized Page 81
Novartis Confidential Page 82
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Graphical displays did not reveal differences in siponimod concentrations if measured while on
CYP3A4 inducer(s) or not or on CYP3A4 inhibitor(s) or not. CYP2C9 inducers or inhibitors,
always administered along with comedications affecting CYP3A4, did not seem to influence
siponimod concentrations either.
Also siponimod plasma concentrations associated with CYP3A4 inducers or CYP3A4 weak,
moderate or strong inhibitors were predicted using the PopPK model (that did not include
concomitant medication as a covariate) and compared with the observed values. There appeared
to be no bias between the predicted and the observed values. This analysis could not be repeated
on CYP2C9 inducers or CYP2C9 inhibitors as there was not enough data.
Finally, the inclusion of weak CYP3A4 inhibitors as a covariate in the PopPK model was found
not to be significant.
Overall, these investigations do not confirm the clinical DDI results and suggest that CYP3A4
inducers or inhibitors do not impact siponimod PK. These results should however be interpreted
with caution as the pooled analysis of comedications was irrespective of their daily dose and
treatment duration, which may have led to heterogeneous induction or inhibition conditions. In
addition, the effect of comedications affecting CYP3A4 activity are most evident, especially
for inhibitors, in the CYP2C9 intermediate and poor metabolizer genotypes (Section 3.3.1.2.3),
which were either excluded (CYP2C9*3*3) or represented a small percentage of the PopPK
population (Table 3-13). No conclusion can be drawn from these analyses for drugs affecting
CYP2C9 due to the limited data available.

3.3.1.2.3 In silico data


This section describes PBPK modeling conducted as a complementary analysis to the clinical
PK DDI studies. Overall, PBPK modeling indicates a differential CYP2C9 genotype-dependent
inhibition and induction of CYP3A4 pathways. With decreased CYP2C9 metabolic activity, a
stronger effect of the CYP3A4 perpetrators on siponimod exposure is anticipated. No CYP2C9
genotype-dependent inhibition and induction of CYP2C9 pathway was detected due to the fact
that dual CYP2C9/3A4 inhibitors and inducers were selected for the SimCYP software (Version
16) predictions. Clinical DDI management, based on CYP2C9 genotype and comedication
effect on CYP enzymes, is discussed in Section 3.3.1.2.4.
The CYP2C9 genotype influences the fractional contributions of the 2 oxidative metabolism
pathways (CYP2C9 and CYP3A4) for overall elimination of siponimod: the hepatic CYP2C9
contribution to metabolism is anticipated to be 80.4% and 7.4% in CYP2C9*1*1 and *3*3
genotypes, respectively, according to the PBPK model. CYP3A4 plays a minor role in
siponimod metabolism for the CYP2C9*1*1 genotype, with a contribution of 17.5% to the
overall elimination of siponimod; its contribution to the overall systemic clearance is expected
to be greater in subjects with lower CYP2C9 activity associated with a reduced total clearance
(82.2% contribution in the *3*3 genotype). Considering the variable CYP3A4 and CYP2C9
contributions to the overall clearance of siponimod in the different CYP2C9 genotypes, it is
expected that the CYP2C9 genotype influences the effects of CYP3A4 and CYP2C9 inhibitors
and inducers.
SimCYP software was used to predict the DDI potential of siponimod as substrate in the
presence of typical CYP2C9/3A inhibitors and inducers for the 6 different CYP2C9 genotypes

Anonymized Page 82
Novartis Confidential Page 83
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

([DMPK R1600759-01]): *1*1 and *1*2 extensive metabolizers, *1*3 and *2*2 intermediate
metabolizers and *2*3 and *3*3 poor metabolizers. The PBPK model was established using a
mixed approach combining in vitro data and PK parameters derived from clinical studies as
detailed in [DMPK R1600759-01-Section 3.2].
The model was used to evaluate the effects of itraconazole (strong CYP3A inhibitor),
fluconazole (moderate CYP3A/CYP2C9 inhibitor), fluvoxamine (weak CYP3A/CYP2C9
inhibitor) (Table 3-19), rifampin (strong CYP3A inducer with moderate induction potential for
CYP2C9) and efavirenz (moderate CYP3A inducer) (Table 3-20) on siponimod systemic
exposure. As there are no marketed drugs showing strong CYP2C9 inhibition or induction
potential so far, only moderate and weak CYP2C9 perpetrator drugs were used clinically and/or
for the modeling. Although a PBPK model of a strong CYP2C9 inhibitor sulfaphenazole is
available in SimCYP, the performance to predict the DDI effects might not be fully evaluated.
In addition, as no selective CYP2C9 inhibitors or inducers are available on the market, dual
perpetrator drugs impacting also the CYP3A4 pathways were selected. However, the fractional
contribution of the CYP2C9 pathway was confirmed using poor metabolizer PK data
(CYP2C9*2*3 and *3*3). The DDI simulation design parameters used are summarized in
[DMPK 1600759-01-Table 3-3 and DMPK 1600759-01-Table 3-4] for siponimod as a victim
with CYP enzyme inhibitors and inducers, respectively.

Effect of inhibitors
Strong CYP3A4 inhibitors
In the presence of the strong CYP3A4 inhibitors itraconazole and ketoconazole, there was a
tendency of an increase in predicted siponimod AUCinf inhibition ratios with the decrease in
CYP2C9 metabolic activity. The ratios increased from 1.17 in CYP2C9*1*1 up to 1.37 in
CYP*2*3 in presence of itraconazole, and from 1.24 in CYP2C9*1*1 up to 1.51 in
CYP2C9*2*3 in presence of ketoconazole. The CYP2C9*3*3 genotype population shows a
significantly higher DDI risk, with predicted AUC ratios of 3.25 and 4.42 for itraconazole and
ketoconazole, respectively (Table 3-19). Ratios for Cmax (predicted and observed) were not
affected in the presence of any of the inhibitors and did not differ among the CYP2C9 genotypes;
the geometric means for the predicted Cmax ratios ranged from 1.00 to 1.07.
In the itraconazole DDI [Study A2124] siponimod AUCinf decreased by 10% and 24% in
subjects with the CYP2C9*1*2 and *1*3 genotypes, respectively, in the presence of
itraconazole, while SimCYP predictions indicated 18% and 30% increases, respectively, in
these 2 genotypes. A potential explanation for this discrepancy could be the induction of another
metabolizing pathway, such as CYP1A1, which was identified in vitro as a minor siponimod
metabolic pathway ([DMPK R0500432]). It was not implemented in the SimCYP model, as in
a non-induced state the abundance of this enzyme is very low (Kim et al 2004). Hepatocyte
investigations performed to better characterize the itraconazole inhibition/induction properties
in vitro indicated that while itraconazole weakly induces CYP1A1 (about 1% relative to the
induction by the positive control), the itraconazole net effect on CYP1A1 is an inhibition effect
([DMPK R1701078]). Itraconazole did not lead to a significant induction of CYP2B6, 2C9 or
3A4. The influence of smoking, a known potent CYP1A inducer, on siponimod PK was
investigated graphically with data from [Study A2304]. Siponimod concentrations were
stratified by CYP2C9 genotype and compared between smokers and non-smokers. The graphs

Anonymized Page 83
Novartis Confidential Page 84
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

did not reveal any apparent difference in steady-state siponimod concentrations as a function of
the smoking status ([BAF312_exposure_and_smoking_status]). Based on these results likely
indicating no clinically relevance of CYP1A pathway in siponimod metabolism, and
considering that the fraction metabolized via CYP3A4 was verified by the fluconazole and
rifampicin DDI study data, the SimCYP model was still considered as valid. While the
underlying mechanism leading to a reduced exposure of siponimod in presence of itraconazole
is not understood, these results should therefore not be extrapolated to other strong and specific
CYP3A4 inhibitors (such as clarithromycin).
Moderate CYP3A4 inhibitors
For erythromycin (moderate CYP3A4 inhibitor), the predicted AUC inhibition ratios were also
larger for the CYP2C9*3*3 genotype (2.46) compared to the other genotypes (1.14 to 1.32;
Table 3-19).
Moderate CYP3A4/2C9 inhibitors
In the presence of fluconazole (moderate CYP3A4/2C9 inhibitor) and fluvoxamine
(weak CYP3A4/2C9 inhibitor), little difference between the 6 different CYP2C9 genotypes
was predicted, with an AUC ratios ranging from 2.15 to 2.52 for fluconazole and 1.12 to 1.41
for fluvoxamine as both metabolic pathways are inhibited by this dual inhibitor (Table 3-19).
The predicted AUC ratio for CYP2C9*1*1 subjects (2.15) in the presence of fluconazole was
in line with the corresponding clinical observation (1.98) in [Study A2128].

Effect of inducers
Strong CYP3A4/moderate CYP2C9 inducers
In presence of rifampin (strong CYP3A4 and moderate 2C9 inducer), both siponimod metabolic
pathways are impacted and only a small difference in the induction effect was predicted between
CYP2C9 genotypes (AUC ratios ranging from 0.32 to 0.21, Cmax ratios ranging from 0.5 to
0.27; Table 3-20). Predicted siponimod ratios for CYP2C9*1*1 subjects (AUC ratio of 0.32,
Cmax ratio of 0.5) when co-administered with rifampin were comparable to the corresponding
clinical observations (AUC ratio of 0.43, Cmax ratio of 0.55) in [Study A2125].
Moderate CYP3A4 inducers
When co-administered with efavirenz (moderate CYP3A4 inducer), slightly higher siponimod
AUC and Cmax ratios were predicted for CYP2C9*3*3 (AUC ratio of 0.35, Cmax ratio of 0.41);
similar AUC and Cmax ratios were estimated for the other CYP2C9 genotypes (AUC ratios
between 0.58 and 0.71 and Cmax ratios between 0.65 and 0.79; Table3-20).

Table 3-19 Predicted siponimod single dose inhibition area under the curve
ratios in the presence of CYP3A4 and CYP3A4/CYP2C9 inhibitors in
various CYP2C9 genotypes
Moderate
Strong 3A4 Strong 3A4 Moderate 3A4 2C9/3A4 Weak 3A4/2C9
inhibitor: inhibitor: inhibitor: inhibitor: inhibitor:
CYP2C9 Itraconazole Ketoconazole Erythromycin Fluconazole Fluvoxamine
genotype AUCi/AUCa AUCi/AUCa AUCi/AUCa AUCi/AUCa AUCi/AUCa
*1*1 1.17 1.24 1.14 2.15 1.41

Anonymized Page 84
Novartis Confidential Page 85
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Moderate
Strong 3A4 Strong 3A4 Moderate 3A4 2C9/3A4 Weak 3A4/2C9
inhibitor: inhibitor: inhibitor: inhibitor: inhibitor:
CYP2C9 Itraconazole Ketoconazole Erythromycin Fluconazole Fluvoxamine
genotype AUCi/AUCa AUCi/AUCa AUCi/AUCa AUCi/AUCa AUCi/AUCa
(1.16 – 1.18) (1.23 – 1.26) (1.13 – 1.16) (2.11 – 2.18) (1.38 – 1.45)
1.18 1.26 1.16 2.15 1.41
*1*2
(1.17 – 1.19) (1.25 – 1.28) (1.15 – 1.17) (2.11 – 2.19) (1.37 – 1.44)
1.24 1.34 1.21 2.18 1.37
*2*2
(1.23 – 1.26) (1.32 – 1.37) (1.19 – 1.22) (2.14 – 2.22) (1.34 – 1.41)
1.30 1.42 1.26 2.20 1.35
*1*3
(1.28 – 1.32) (1.39 – 1.45) (1.24 – 1.28) (2.16 – 2.24) (1.32 – 1.38)
1.37 1.51 1.32 2.22 1.32
*2*3
(1.35 – 1.40) (1.48 – 1.55) (1.29 – 1.35) (2.18 – 2.26) (1.29 – 1.34)
3.25 4.42 2.46 2.52 1.12
*3*3
(2.99 - 3.51) (4.14 – 4.70) (2.29 – 2.63) (2.45 – 2.58) (1.11 – 1.13)
a Geometric mean (90% CI)
Source: [DMPK 1600759-01-Tables 6-11, 6-12, 16-13, 6-14 and 6-15]

Table 3-20 Predicted siponimod steady state induction area under the curve and
maximum concentration ratios in the presence of CYP3A4 and
CYP3A4/CYP2C9 inducers in various CYP2C9 genotypes
Strong 3A4/moderate 2C9 inducer:
CYP2C9 Rifampin Moderate 3A4 inducer: Efavirenz
genotype AUCi/AUCa Cmaxi/Cmaxa AUCi/AUCa Cmaxi/Cmaxa
*1*1 0.32 (0.31 – 0.33) 0.50 (0.49 – 0.51) 0.71 (0.69 – 0.73) 0.79 (0.77 – 0.80)
*1*2 0.32 (0.31 – 0.33) 0.49 (0.48 – 0.50) 0.70 (0.68 – 0.71) 0.77 (0.76 – 0.79)
*2*2 0.31 (0.30 – 0.32) 0.46 (0.45 – 0.47) 0.65 (0.63 – 0.67) 0.73 (0.71 – 0.74)
*1*3 0.30 (0.29 – 0.31) 0.43 (0.42 – 0.44) 0.61 (0.59 – 0.63) 0.69 (0.67 – 0.71)
*2*3 0.29 (0.28 – 0.30) 0.41 (0.40 – 0.42) 0.58 (0.56 – 0.60) 0.65 (0.63 – 0.67)
*3*3 0.21 (0.20 – 0.22) 0.27 (0.27 – 0.29) 0.35 (0.33 – 0.38) 0.41 (0.39 – 0.43)
a Geometric mean (90% CI)
Source: [DMPK 1600759-01-Tables 6-16 and 6-17]

3.3.1.2.4 Siponimod CYP2C9 genotype-based drug-drug interaction management


When considering the CYP2C9 genotype-based dosing approach for all but CYP2C9*3*3
patients, who are recommended to be excluded from a marketing authorization (Section 3.2.5.3),
the estimated net effect on siponimod exposure in the presence of CYP2C9/CYP3A4 inhibitors
or inducers are summarized in Table 3-21 and Table 3-22. The net effect is the simulated
inhibition or induction ratio ([DMPK 1600759-01-Table 3-3 and DMPK 1600759-01-Table 3-
4]) multiplied by the corresponding CYP2C9-genotype ratio effect on exposure remaining after
dose adjustment. This net effect corresponds to the expected inhibition or induction effect on
siponimod exposure in comparison to that in CYP2C9*1*1 subjects receiving a 2 mg qd
maintenance dose without co-administration of any CYP2C9/CYP3A4 perpetrator drug.

Anonymized Page 85
Novartis Confidential Page 86
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Inhibitors
Under the proposed genotype-based dosing recommendations and considering an acceptance of
a maximum 2-fold exposure increase (conservative approach), siponimod may be combined
with all types of CYP3A4/CYP2C9 inhibitors without relevant implications on safety or
efficacy for concomitant treatment in most patients. Particular caution should be applied in
patients with a CYP2C9*2*2 genotype for combination treatment with moderate
CYP2C9/CYP3A4 inhibitors (e.g. fluconazole) due to a higher exposure (AUC ratio 2.7; Table
3-21). Dosage adjustment to 1 mg daily may be considered in these patients.

Inducers
Strong CYP3A4/moderate CYP2C9 inducers are predicted to reduce siponimod exposure by
approximately 61% to 76% (Table 3-22). Moderate CYP3A4 inducers are predicted to reduce
siponimod exposure by approximately 19% to 51% (Table 3-22). Caution should therefore be
applied when siponimod is used in combination with:
 Strong CYP3A4/moderate CYP2C9 inducers for all genotypes (e.g., rifampin,
carbamazepine)
 Moderate inducers of CYP3A4 in patients with a CYP2C9*1*3 or *2*3 genotype
(e.g., efavirenz, modafinil)

Table 3-21 Net inhibition effect ratios (inhibition x CYP2C9 genotype exposure
ratios)
CYP2C9- Moderate
genotype Strong 3A4 Strong 3A4 Moderate 3A4 2C9/3A4 Weak 3A4/2C9
based inhibitor: inhibitor: inhibitor: inhibitor: inhibitor:
CYP2C9 dosing Itraconazole Ketoconazole Erythromycin Fluconazole Fluvoxamine
genotype (mg) AUC ratioa,b AUC ratioa,b AUCi/AUCa,b AUC ratioa,b AUC ratioa,b
*1*1 2 1.17 1.24 1.14 2.15 1.41
*1*2 2 1.18 1.26 1.16 2.15 1.41
*2*2 2 1.55 1.68 1.51 2.73 1.71
*1*3 1 1.05 1.15 1.02 1.78 1.09
*2*3 1 1.32 1.46 1.27 2.13 1.27
a Calculated using DDI ratios from Table 3-19 multiplied by the genotype effect on exposure
(CYP2C9-genotype based dosing adjustment factor)
b Geometric mean (90% CI)
Source: [DMPK 1600759-01-Tables 6-11, 6-12, 16-13, 6-14 and 6-15]

Table 3-22 Net induction effect ratios (induction x CYP2C9 genotype exposure
ratios)
Therapeutic Strong 3A4/moderate 2C9 Moderate 3A4 inducer:
CYP2C9 maintenance inducer: Rifampin Efavirenz
genotype dose (mg) AUC ratioa,b Cmax ratioa,b AUC ratioa,b Cmax ratioa,b
*1*1 2 0.32 0.50 0.71 0.79
*1*2 2 0.32 0.49 0.70 0.77
*2*2 2 0.39 0.58 0.81 0.91

Anonymized Page 86
Novartis Confidential Page 87
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Therapeutic Strong 3A4/moderate 2C9 Moderate 3A4 inducer:


CYP2C9 maintenance inducer: Rifampin Efavirenz
genotype dose (mg) AUC ratioa,b Cmax ratioa,b AUC ratioa,b Cmax ratioa,b
*1*3 1 0.24 0.35 0.49 0.56
*2*3 1 0.28 0.39 0.56 0.62
a Calculated using DDI ratios from Table 3-20 multiplied by the genotype effect on exposure
(CYP2C9-genotype based dosing adjustment factor)
b Geometric mean (90% CI)
Source: [DMPK 1600759-01-Tables 6-16 and 6-17]

3.4 Pharmacodynamics
The primary focus of this section is the across-study pooled analyses of primary and secondary
PD effects of siponimod in healthy subjects (single and multiple dose studies) and PM/DM
patients (multiple dose studies); a summary of the assessment methodology, including types of
assessments for each analysis and categories of interest, is presented in Table 5-3. A by-study
summary of clinical PD evaluations of siponimod is presented for healthy subjects in
Section 2.2, including hepatic and renal impairment subjects (not included in the pooled
analyses) in Section 2.2.10.1, and in MS and PM/DM patients in Section 2.3.1 and Section 2.3.2,
respectively.

3.4.1 Efficacy assessments


As a selective S1P receptor modulator, siponimod induces internalization and degradation of
the S1P1 receptor and thereby acts as a functional antagonist on S1P1. The resulting ALC
reduction in peripheral blood due to prevention of lymphocyte recirculation from lymphatic
tissue to target organs constitutes the primary efficacy-related PD effect of siponimod. The
primary focus of this section is the pooled categorical analyses of ALC levels across single and
multiple dose Clinical Pharmacology and PM/DM studies. Efficacy assessments across MS
studies are discussed in [SCE-Sections 3.2 and 3.3].

3.4.1.1 Lymphocyte effects


The pooled categorical analyses of ALC levels, described in Table 5-3, are available in
(SCP Appendix-Tables 12-5.1 to 12-5.3) for studies in healthy subjects (single and multiple
dose) and PM/DM patients. The dose-dependent reduction of ALC by siponimod observed in
individual single and multiple dose studies in healthy subjects is discussed in Section 2.2.
The effect of siponimod on ALC in the 3 PM/DM studies is described in Section 2.3.2.
The effect of siponimod on leukocyte subsets is discussed in Section 2.2.2 ([Study A2105]).
The exposure-lymphocyte relationship investigated in a PopPKPD analysis is described in
Section 3.4.3.
Overall, in the pooled analyses of multiple doses of siponimod, at the therapeutic (2 mg) dose
the majority of siponimod-treated subjects (57.6%) reached therapeutic ALC levels of
≥ 0.6 × 109/L and < 1.2 × 109/L (Category 2) followed by subjects (30.3%) with ALC values
≥ 0.2 × 109/L and < 0.6 × 109/L (Category 3). At daily doses < 2 mg, the majority of ALC levels
were in Categories 1 and 2 (range: 85.2% to 93.4%). This supports the effectiveness of the

Anonymized Page 87
Novartis Confidential Page 88
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

selected therapeutic maintenance dose of 2 mg in reducing ALC to a therapeutically relevant


level, while at the same time avoiding ALC reduction to levels < 0.2 × 109/L.
Across all investigated multiple dose levels, most siponimod-treated subjects showed ALC
levels in Categories 2 (45.0%) or 3 (43.0%) while 11.6% and 0.4% of subjects were showing
ALC levels within normal range or < 0.2 × 109/L, respectively (Table 3-23). There was a clear
increase in incidence of Category 3 events at siponimod doses > 2 mg. Among placebo-treated
subjects, most subjects had ALC levels in Category 1 (95.2%, 3.3%, 0.5%, 0.3% in Categories
1, 2, 3 and 4, respectively). At therapeutic multiple doses of siponimod and below (≤ 2 mg
(titrated and non-titrated)), the greatest proportion of subjects had ALC levels in Category 2,
with only a small proportion of subjects showing ALC levels in Category 3 (average: 13.0%;
range: 0% to 30.3%) and no subjects with ALC levels in Category 4.

Table 3-23 Lymphocytes categorical outlier analysis – multiple dose studies


Absolute lymphocyte levels (109/L) event categories
Category 1 2 3 4
9 9
Within normal range ≥ 0.6 × 10 /L ≥ 0.2 × 10 /L < 0.2 × 109/L
(1.2 × 109/L to and and
3.5 × 109/L) < 1.2 × 109/L < 0.6 × 109/L
Treatment n (%) n (%) n (%) n (%)
BAF312 0.3 mg (N=15) 7 (46.7) 7 (46.7) 1 (6.7) 0
BAF312 0.5 mg (N=94) 49 (52.1) 44 (46.8) 1 (1.1) 0
BAF312 1 mg (N=101) 23 (22.8) 63 (62.4) 15 (14.9) 0
BAF312 2 mg (N=66) 8 (12.1) 38 (57.6) 20 (30.3) 0
BAF312 2.5 mg (N=16) 0 5 (31.3) 11 (68.8) 0
BAF312 4 mg (N=64) 3 (4.7) 36 (56.3) 25 (39.1) 0
BAF312 10 mg (N=32) 0 7 (21.9) 25 (78.1) 0
BAF312 20 mg (N=18) 0 0 17 (94.4) 1 (5.6)
BAF312 0.5 mg preceded by 10 (83.3) 2 (16.7) 0 0
up-titration(N=12)
BAF312 2 mg preceded by 30 (12.7) 114 (48.3) 59 (25.0) 0
up-titration (N=236)
BAF312 10 mg preceded by 7 (5.8) 50 (41.7) 63 (52.5) 0
up-titration (N=120)
BAF312 in combinationa 68 (38.9) 64 (36.6) 42 (24.0) 1 (0.6)
(N=175)
Comparator alone (N=188) 163 (86.7) 2 (1.1) 0 0
Placebo (N=393) 374 (95.2) 13 (3.3) 2 (0.5) 1 (0.3)
All BAF (N=544) 63 (11.6) 245 (45.0) 234 (43.0) 2 (0.4)
Any treatment (N=872) 376 (43.1) 257 (29.5) 235 (26.9) 3 (0.3)
N = total number of subjects who got administered with at least one dose of respective treatment;
n = number of unique subjects meeting respective category criteria.
Included [Studies A2102, A2105, A2107, A2110, A2116, A2118, A2121, A2125, A2128 (Part 2) and
A2130].
Only events occurring at or after first drug intake were considered. A subject minimum result was
considered and was counted only once in the worst category.

Anonymized Page 88
Novartis Confidential Page 89
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Absolute lymphocyte levels (109/L) event categories


Category 1 2 3 4
Within normal range ≥ 0.6 × 109/L ≥ 0.2 × 109/L < 0.2 × 109/L
(1.2 × 109/L to and and
3.5 × 109/L) < 1.2 × 109/L < 0.6 × 109/L
Treatment n (%) n (%) n (%) n (%)
a Combination studies included: [Studies A2116, A2121 and A2125] (propranolol, monophasic OC and
rifampin, respectively)
Source: (SCP Appendix 1-Table 12-5.2)

3.4.2 Safety assessments


A broad evaluation of safety in Clinical Pharmacology and PM/DM studies as assessed through
pooled analyses of AEs is discussed in this section, followed by safety assessments of
cardiac/hemodynamic, pulmonary and other safety-related (secondary) PD effects of siponimod
in Clinical Pharmacology and PM/DM studies. Safety assessments of the 2 MS studies is
discussed in the [SCS-Sections 2, 3 and 4].

3.4.2.1 Adverse events


The pooled analyses of AEs by preferred term (PT), system organ class (SOC), and category
(e.g., intensity, severity, etc.), described in Table 5-3, are available in [SCS Appendix 1-Tables
12-1.1.1 to 12-1.1.3, 12-1.2.1 to 12-1.2.3 and 12-1.3.1 to 12-1.3.3, respectively] for studies in
healthy subjects (single and multiple dose) and PM/DM patients. Together the analyses
demonstrate that siponimod was well tolerated and showed a favorable safety profile in the
populations studied. The majority of AEs was mild in intensity for siponimod and placebo
treatments in healthy subjects in single and multiple dose studies and PM/DM patients. In
healthy subjects in single and multiple dose studies, only 1 (0.3%) and 4 (0.5%) subjects,
respectively, reported 6 SAEs, the majority of which were suspected as being related to study
drug ([SCS Appendix 1-Listings 16.2.7-1.1p and 16.2.7-1.2p]). Of the 8 (16.3%) PM/DM
patients who reported a total of 14 SAEs, 4 patients (6 SAEs) were receiving placebo treatment;
from the other 4 patients on active treatment (8 SAEs), 5 SAEs (2 patients) were considered
drug-related.
For the single and multiple dose studies in healthy subjects (0.1 to 75 mg and 0.3 to 20 mg,
respectively), Table 3-24 and Table 3-25, respectively, list (by SOC and PT) the siponimod and
placebo frequencies of the AEs that occurred in ≥ 2% of siponimod-treated subjects. For both
single and multiple doses of siponimod, the 3 most frequent AEs were headache (32.5% and
20.8%, respectively), dizziness (12.1% and 5.7%, respectively) and nausea (5.5% and 4.2%,
respectively), supporting the safety of siponimod in healthy subjects. Cardiac disorders with
frequencies ≥ 2% only occurred in subjects receiving single doses of siponimod and included
bradycardia (4.3%) and AV block second degree (Mobitz 1, 2:1 or 3:2 conduction) (3.4%)
(further discussed in Section 3.4.2.2.1 and Section 3.4.2.2.2, respectively).

Anonymized Page 89
Novartis Confidential Page 90
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-24 Incidence of AEs (at least 2 percent) by SOC and PT in single dose
studies in healthy subjects
Subjects with at least 1 AE Siponimod Placebo
N = 348 N = 41
n (%) n (%)
Cardiac disorders 37 (10.6) 0
AV block second degree 12 (3.4) 0
Bradycardia 15 (4.3) 0
Gastrointestinal disorders 30 (8.6) 1 (2.4)
Nausea 19 (5.5) 1 (2.4)
General disorders and administration site conditions 28 (8.0) 0
Fatigue 12 (3.4) 0
Infections and infestations 19 (5.5) 0
Upper respiratory tract infection 7 (2.0) 0
Nervous system disorders 125 (35.9) 4 (9.8)
Dizziness 42 (12.1) 1 (2.4)
Headache 113 (32.5) 2 (4.9)
Somnolence 8 (2.3) 1 (2.4)
n = number of subjects with at least 1 AE in the category.
Source: [SCS Appendix 1-Tables 12-1.1.1 and 12-1.2.1 and Listing 16.2.7-1.1p]

Anonymized Page 90
Novartis Confidential Page 91
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-25 Incidence of AEs (at least 2 percent) by SOC and PT in multiple dose
studies in healthy subjects
Subjects with at least 1 AE Siponimod Placebo
N = 544 N = 393
n (%) n (%)
Gastrointestinal disorders 72 (13.2) 30 (7.6)
Constipation 11 (2.0) 6 (1.5)
Diarrhoea 12 (2.2) 3 (0.8)
Nausea 23 (4.2) 6 (1.5)
General disorders and administration site conditions 41 (7.5) 11 (2.8)
Fatigue 15 (2.8) 2 (0.5)
Investigations 52 (9.6) 3 (0.8)
Alanine aminotransferase increased 19 (3.5) 1 (0.3)
Musculoskeletal and connective tissue disorders 45 (8.3) 14 (3.6)
Back pain 15 (2.8) 2 (0.5)
Nervous system disorders 137 (25.2) 31 (7.9)
Dizziness 31 (5.7) 3 (0.8)
Headache 113 (20.8) 25 (6.4)
Respiratory, thoracic and mediastinal disorders 25 (4.6) 13 (3.3)
Oropharyngeal pain 13 (2.4) 5 (1.3)
n = number of subjects with at least 1 AE in the category.
Source: [SCS Appendix 1-Tables 12-1.1.2 and 12-1.2.2 and Listing 16.2.7-1.2p]

3.4.2.2 Cardiac/hemodynamic effects


This section summarizes the results of pooled categorical analyses of cardiac/hemodynamic
effects of siponimod across single and multiple dose studies, specifically the effects on HR
(Section 3.4.2.2.1), cardiac rhythm (AV blocks, sinus pauses, cardiac repolarization (QT) in
Section 3.4.2.2.2) and hemodynamics (BP in Section 3.4.2.2.3 and orthostatic
hypotension/dysregulation in Section 3.4.2.2.4).

3.4.2.2.1 Effect on heart rate/negative chronotropic effects


As discussed for healthy subjects in Section 2.2, a transient, dose-dependent decrease in HR
was observed after siponimod administration in the SAD and MAD studies ([Studies A2101
and A2105]). DT regimens of siponimod over 8 or 9 days in [Study A2107] robustly attenuated
the bradyarrhythmic effects typically seen at treatment onset with S1P receptor modulators.
The pooled categorical analyses of HR effects by 12-lead ECG and Holter ECG assessments,
described in Table 5-3, are available in [SCS Appendix 1-Table 12-2.1.1 to SCS Appendix 1-
Table 12-2.1.3 and SCS Appendix 1-Table 12-2.2.1 to SCS Appendix 1-Table 12-2.2.2,
respectively] for studies in healthy subjects (single and multiple dose) and PM/DM patients.

Anonymized Page 91
Novartis Confidential Page 92
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Overall, under single dose conditions in healthy subjects, across all investigated dose levels HR
changes from Baseline in siponimod-treated subjects were predominantly Category 1 events
(HR change from Baseline > 25% and ≤ 35%) which occurred at a lower frequency than in
placebo-treated subjects. New HR values in Category 4 (HR value below 30 bpm) only occurred
at doses ≥ 10 mg. Under multiple dose conditions in healthy subjects across all investigated
doses and regimens, cases of HR change from Baseline in siponimod-treated subjects were
mainly events of Category 1 and Category 2 (HR change from Baseline > 35% and ≤ 50%). No
new HR values of Category 4 were observed. Importantly, in subjects under the 2-mg siponimod
DT regimen, the incidence of Category 1 and 2 events and cases in the combined Category 8
(HR decrease > 25% from Baseline to a HR < 50 bpm) were similar to placebo, underlining the
effectiveness of the DT regimen in attenuating the bradyarrhythmic effects of siponimod.
Under multiple dose conditions (non-titrated and titrated) in healthy subjects, the percentages
of siponimod-treated subjects with Category 1 and 2 events were 54.2% and 31.1%, respectively,
based on Holter ECG analyses (Table 3-26) and 11.2% and 5.5%, respectively, based on 12-lead
ECG analyses. The corresponding percentages of placebo-treated subjects were 38.7% and
21.8%, respectively, based on Holter ECG analyses and 8.7% and 4.3%, respectively, based on
12-lead ECG analyses. At doses ≤ 2 mg, this difference was mainly driven by the siponimod
treatment groups without a preceding DT and the majority of Category 1 and 2 events occurred
at doses > 2 mg. The Category 1 and 2 percentages for the 2 mg DT dose group were 18.6%
and 6.8%, respectively, based on Holter ECG analyses and 0.4% and 0%, respectively, based
on 12-lead ECG analyses, indicating a placebo-like occurrence of these events. None of the
subjects showed a HR value below 30 bpm in any group (Category 4) based on 12-lead ECG
or Holter analyses. Cases in the combined Category 8 were seen in 19.8% of siponimod-treated
subjects (mainly at doses > 2 mg) compared to 0.4% of placebo-treated subjects based on Holter
ECG analyses, while no cases were seen in 12-lead ECG analyses. In the 2 mg DT group only
1.7% of subjects showed Category 8 events based on Holter ECG analyses, indicating a
placebo-like occurrence of these events.

Anonymized Page 92
Novartis Confidential Page 93
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-26 Heart rate categorical outlier analysis (Holter data) – multiple dose studies
HR change from Baseline New HR values compared to Combined category
Baseline
Category 1 2 3 4 5 6 7 8 9
HR decrease HR increase
≥ 30 bpm ≥ 40 bpm ≥ 45 bpm > 25% from > 25% from
> 25% and > 35% and and and and Baseline to a Baseline to a
≤ 35% ≤ 50% > 50% < 30 bpm < 40 bpm < 45 bpm < 50 bpm HR < 50 bpm HR > 100 bpm
Treatment n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%)
BAF312 0.3 mg (N=15) 11 (73.3) 8 (53.3) 4 (26.7) 0 0 0 1 ( 6.7) 0 6 (40.0)
BAF312 1 mg (N=15) 9 (60.0) 4 (26.7) 2 (13.3) 0 0 2 (13.3) 2 (13.3) 3 (20.0) 2 (13.3)
BAF312 2.5 mg (N=16) 13 (81.3) 8 (50.0) 3 (18.8) 0 0 3 (18.8) 8 (50.0) 8 (50.0) 4 (25.0)
BAF312 10 mg (N=32) 18 (56.3) 7 (21.9) 3 ( 9.4) 0 3 ( 9.4) 6 (18.8) 16 (50.0) 14 (43.8) 5 (15.6)
BAF312 20 mg (N=18) 11 (61.1) 6 (33.3) 3 (16.7) 0 4 (22.2) 7 (38.9) 10 (55.6) 12 (66.7) 5 (27.8)
BAF312 0.5 mg 4 (33.3) 2 (16.7) 0 0 0 3 (25.0) 4 (33.3) 2 (16.7) 0
preceded by up-titration
(N=12)
BAF312 2 mg 22 (18.6) 8 ( 6.8) 0 0 1 ( 0.8) 1 ( 0.8) 16 (13.6) 2 ( 1.7) 0
preceded by up-titration
(N=118)
BAF312 10 mg 75 (62.5) 47 (39.2) 6 ( 5.0) 0 1 ( 0.8) 3 ( 2.5) 8 ( 6.7) 2 ( 1.7) 3 ( 2.5)
preceded by up-titration
(N=120)
BAF312 in 4 (10.5) 1 ( 2.6) 1 ( 2.6) 0 1 ( 2.6) 4 (10.5) 16 (42.1) 11 (28.9) 0
combinationa (N=38)
Comparator alone 80 (61.1) 54 (41.2) 13 ( 9.9) 0 0 3 ( 2.3) 14 (10.7) 2 ( 1.5) 6 ( 4.6)
(N=131)

Anonymized Page 93
Novartis Confidential Page 94
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

HR change from Baseline New HR values compared to Combined category


Baseline
Category 1 2 3 4 5 6 7 8 9
HR decrease HR increase
≥ 30 bpm ≥ 40 bpm ≥ 45 bpm > 25% from > 25% from
> 25% and > 35% and and and and Baseline to a Baseline to a
≤ 35% ≤ 50% > 50% < 30 bpm < 40 bpm < 45 bpm < 50 bpm HR < 50 bpm HR > 100 bpm
Treatment n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%)
Placebo (N=266) 103 (38.7) 58 (21.8) 19 ( 7.1) 0 0 1 ( 0.4) 8 ( 3.0) 1 ( 0.4) 27 (10.2)
All BAF (N=273) 148 (54.2) 85 (31.1) 22 ( 8.1) 0 10 ( 3.7) 29 (10.6) 75 (27.5) 54 (19.8) 25 ( 9.2)
Any treatment (N=558) 331 (59.3) 197 (35.3) 54 ( 9.7) 0 10 ( 1.8) 33 ( 5.9) 94 (16.8) 56 (10.0) 58 (10.4)
N = total number of subjects who got administered with at least one dose of respective treatment; n = number of unique subjects meeting respective
category criteria.
Included [Studies A2102, A2105, A2107, A2116, A2118 and A2128 (Part 2)].
For all studies, time matched Baseline is considered; for [Study A2118], latest results prior to the dosing is considered as Baseline. A subject with
multiple occurrences within each category was counted only once for that category and treatment.
a Combination studies included: Study A2116 (propranolol)
Source: [SCS Appendix 1-Table 12-2.2.2]

Anonymized Page 94
Novartis Confidential Page 95
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.4.2.2.2 Effect on cardiac rhythm


This section summarizes the results of the pooled categorical analyses of cardiac rhythm (AV
blocks, sinus pauses, cardiac repolarization (QTcF)).

Bradyarrhythmia (AV blocks and sinus pauses)


Across the Clinical Pharmacology studies and studies in PM/DM patients summarized
individually in Section 2.2 and Section 2.3.2, respectively, there were no second degree AV
Blocks Mobitz II or higher degree observed. Second degree AV blocks Mobitz I or second
degree AV blocks with 2:1 or 3:2 conduction identified in Holter ECGs or in continued online
ECG monitoring were uncommon, mostly asymptomatic and only in very few cases
accompanied by mild AEs, such as headache or dizziness, with uncertain causal relationship to
the AV block occurrences. Detected sinus pauses were mostly asymptomatic and were not
considered to be of clinical relevance, and all resolved spontaneously without requiring medical
intervention. In many cases the AV block and sinus pauses detected occurred during resting or
nocturnal hours, which represent episodes of increased vagal tone known to be associated with
negative chronotropic and dromotropic cardiac effects and higher incidence rates of AV blocks
and sinus pauses.
The pooled analysis of bradyarrhythmic events by Holter ECG, 12-lead ECG and telemetry
assessments, discussed below and described in Table 5-3, are available in [SCS Appendix 1-
Table 12-2.3.1 to SCS Appendix 1-Table 12-2.3.2, SCS Appendix 1-Table 12-3.3.1 to SCS
Appendix 1-Table 12-3.3.3 and SCS Appendix 1-Table 12-4.3.1 to SCS Appendix 1-Table 12-
4.3.3], respectively for studies in healthy subjects (single and multiple dose) and PM/DM
patients.
Overall, observations of AV blocks were mainly first degree AV blocks or second degree AV
blocks Mobitz I (Wenckebach) type, observed at similar or lower frequencies in the 12-lead
ECG and telemetry analyses compared to Holter ECG analyses. All observations of sinus pauses
were based on the Holter ECG analysis, except for 1 observation each for 12-lead ECG analysis
(single dose subject) and online cardiac monitoring (telemetry) analysis (PM/DM patient). The
following section primarily focuses on AV Block and sinus pause events in the pooled Holter
ECG based analyses. Importantly, most AV blocks and sinus pauses were detected at
doses > 2 mg and their incidence was notably higher under non-titrated conditions compared to
DT conditions, demonstrating the efficient mitigation of bradyarrhythmic effects of siponimod
under DT of siponimod.

AV blocks
Under single dose conditions in healthy subjects, asymptomatic first degree AV blocks were
observed in Holter ECGs in 2.8% of the siponimod-treated subjects and had a comparable
incidence of 0% in placebo-treated subjects. As some of the pooled single-dose studies were
open-label studies and did not include a placebo group, this slight imbalance in the incidence
of first degree AV blocks should be interpreted with caution. All detected second degree AV
blocks were Mobitz I type, which resolved spontaneously without requiring medical
intervention and were mostly asymptomatic; no second degree AV blocks with 2:1 or 3:2

Anonymized Page 95
Novartis Confidential Page 96
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

conduction or higher degree AV blocks were observed. The incidence of AV blocks Mobitz I
in Holter ECGs was 6.5% after a single dose of siponimod compared to 0% in placebo subjects
and none of the observed second degree AV blocks occurred at doses below 2.5 mg. The
corresponding figures for second degree AV blocks based on online cardiac monitoring
(telemetry) data were 2.7% in siponimod-treated subjects compared to 0% in placebo subjects,
with 80% of these observations occurring at doses of 4 mg.
In multiple dose studies, asymptomatic first degree AV blocks were observed in Holter ECGs
in 5.0% of the siponimod-treated subjects and in 1.2% of placebo-treated subjects, with the
incidences ranging between 0% and 6.3% in non-titrated groups compared to only 0% to 1.7%
in DT conditions for siponimod-only treatment (Table 3-27). As some of the pooled multiple-
dose studies were open-label studies and did not include a placebo group, this imbalance in the
incidence of first degree AV blocks should be interpreted with caution. In Holter ECG-based
analyses, the majority of detected second degree AV blocks was Mobitz I type and observed at
an incidence of 3.5% in siponimod-treated subjects compared to 1.8% in placebo-treated
subjects, while second degree AV blocks with 2:1 or 3:2 conduction did occur in 1% and 0.2%
of the siponimod subjects, respectively (0% in placebo subjects). The Holter-based incidence
of second degree AV blocks of Mobitz I type showed a clear imbalance between non-titrated
groups (0% to 18.8% (0% to 13.3% at doses ≤ 2 mg)) compared to 0% to 1.7% in DT groups,
indicating the efficient mitigation of bradyarrhythmic events under the DT regimen.

Table 3-27 Bradyarrhythmia events categorical analysis (Holter data) - multiple


dose studies
1st degree 2nd 2nd degree 2nd degree Sinus Sinus
AV Block degree AV AV Block, AV Block, pause pause
Treatment (PR > 200 Block, 2:1 3:2 (RR > 2 (RR > 3
msec) Mobitz I conduction conduction sec) sec)
n (%) n (%) n (%) n (%) n (%) n (%)
BAF312 0.3mg
0 2 (13.3) 0 0 1 (6.7) 0
(N=15)
BAF312 0.5mg
5 (5.3) 0 0 0 1 (1.1) 0
(N=94)
BAF312 1mg
5 (5.0) 2 (2.0) 1 (1.0) 0 3 (3.0) 0
(N=101)
BAF312 2mg
3 (4.5) 1 (1.5) 0 0 0 0
(N=66)
BAF312 2.5mg
1 (6.3) 3 (18.8) 1 (6.3) 0 5 (31.3) 0
(N=16)
BAF312 4mg
1 (1.6) 0 0 0 1 (1.6) 0
(N=64)
BAF312 10mg
2 (6.3) 3 (9.4) 1 (3.1) 1 (3.1) 9 (28.1) 0
(N=32)
BAF312 20mg
0 0 0 0 10 (55.6) 0
(N=18)
BAF312 0.5mg
preceded by
0 0 0 0 1 (8.3) 1 (8.3)
up-titration
(N=12)

Anonymized Page 96
Novartis Confidential Page 97
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

1st degree 2nd 2nd degree 2nd degree Sinus Sinus


AV Block degree AV AV Block, AV Block, pause pause
Treatment (PR > 200 Block, 2:1 3:2 (RR > 2 (RR > 3
msec) Mobitz I conduction conduction sec) sec)
n (%) n (%) n (%) n (%) n (%) n (%)
BAF312 2mg
preceded by
2 (1.7) 1 (0.8) 0 0 2 (1.7) 0
up-titration
(N=118)
BAF312 10mg
preceded by
1 (0.8) 2 (1.7) 1 (0.8) 0 5 (4.2) 0
up-titration
(N=120)
BAF312 in
combinationa 5 (13.2) 0 0 0 3 (7.9) 0
(N=38)
Comparator
alone 8 (6.1) 0 1 (0.8) 0 1 (0.8) 0
(N=131)
Placebo
4 (1.2) 6 (1.8) 0 0 6 (1.8) 0
(N=325)
All BAF
20 (5.0) 14 (3.5) 4 (1.0) 1 (0.2) 40 (9.9) 1 (0.2)
(N=403)
Any treatment
28 (4.0) 20 (2.9) 5 (0.7) 1 (0.1) 47 (6.8) 1 (0.1)
(N=696)
N = number of subjects dosed with respective treatment; n = number of unique subjects meeting the
criteria.
Included [Studies A2102, A2105, A2107, A2110, A2116, A2118 and A2128 (PART 2)].
Only events occurring at or after first drug intake were considered.
a Combination studies included: (Study A2116) (propranolol)
Source: [SCS Appendix 1-Table 12-2.3.2]

Sinus pauses
Under single dose conditions in healthy subjects, asymptomatic sinus pauses (defined as
RR > 2 sec) were observed in Holter ECGs in 19.6% of the siponimod-treated subjects
compared to 2.4% in placebo-treated subjects. The majority of detected sinus pauses had a RR
interval of < 3 sec and was observed mainly at oral doses above or equal to 2.5 mg. As some of
the pooled single dose studies were open-label studies and did not include a placebo group, this
imbalance in the incidence of sinus pauses should be interpreted with caution. All detected sinus
pauses resolved spontaneously without requiring medical intervention.
Under multiple dose conditions in healthy subjects, the overall incidence of sinus pauses (RR >
2 sec) based on Holter ECG data was 9.9% in siponimod-treated subjects compared to 1.8% in
placebo-treated subjects (Table 3-27). All identified sinus pauses were asymptomatic, and
resolved spontaneously without medical intervention. The detected sinus pauses occurred
mainly at oral doses > 2 mg; within individual dose groups, incidences ranged between 0 and
6.7% at oral non-titrated doses ≤ 2 mg, while incidences ranged between 1.6% and 55.6% at
non-titrated oral doses between 2.5 and 20 mg. As with second degree AV blocks, the incidence

Anonymized Page 97
Novartis Confidential Page 98
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

of sinus pauses was notably lower in DT groups, with 1.7% for DT to 2 mg and 4.2% for DT to
10 mg, demonstrating the efficient mitigation of bradyarrhythmic events by the DT regimen.

QT interval/cardiac repolarization
Treatment with siponimod did not reveal an arrhythmogenic potential related to QT
prolongation ([Study A2118]).
The pooled categorical analyses of predefined QTcF events by 12-lead ECG and Holter ECG
asssessments, described in Table 5-3, are available in [SCS Appendix 1-Table 12-3.1.1 to SCS
Appendix 1-Table 12-3.1.3 and SCS Appendix 1-Table 12-3.2.2], respectively for studies in
healthy subjects (single and multiple dose) and PM/DM patients. Fridericia-corrected QT
values were used as an appropriate HR correction method for a drug with known effects on HR
and to enhance the QTcF-based categorical analysis in the thorough QT study.
The QTcF profile in the pooled analyses across single and multiple dose studies was overall
consistent with the categorical QTcF analyses in the dedicated thorough QT (Study A2118).
Holter ECG analyses (data only available for multiple dose study healthy subjects) and 12-lead
ECG analyses revealed that the majority of events were QTcF changes from Baseline > 30 msec
and ≤ 60 msec (Category 1) or new QTcF values > 450 msec and ≤ 480 msec (Category 3)
which were observed in siponimod-treated subjects/patients with an incidence comparable to
placebo. QTcF changes from Baseline > 60 msec (Category 2) or new on-treatment QTcF
values > 480 msec (Categories 4 and 5) or the combination of both (Category 6) were seen in a
limited number of subjects (0% for most subjects, for both siponimod and placebo treatments).
As some of the pooled studies were open-label studies and did not include a placebo group, this
imbalance in the incidence of QTcF categories of interest should be interpreted with caution.
In single dose studies, 12-lead ECG data did not reveal any QTcF changes in Categories 2, 4, 5
or 6 in siponimod-treated subjects. QTcF changes in Category 1 were detected in 8.3% of
siponimod treated subjects compared to 9.8% in placebo subjects and did not show any clear
relation to the dose level. New QTcF values in Category 3 were seen in 2.8% of
siponimod-treated subjects compared to 0% in placebo subjects and occurred at single doses of
≥ 4 mg or at 1 mg iv over 24 hours.
Under multiple dose conditions in healthy subjects, 12-lead ECG data revealed occurrences of
predefined events only in siponimod-treated subjects for Categories 2 (3 subjects (0.6%)),
4 (1 subject) and 6 (1 subject). QTcF changes from Baseline in Category 1 were detected in 5.7%
of siponimod-treated subjects compared to 2.3% in placebo subjects and did not show any clear
relation to the dose level or difference between non-titrated and titrated dosing regimens. Holter
ECG analyses from 3 multiple dose studies did not reveal any events in Categories 2, 4, 5 or 6
in siponimod-treated subjects (Table 3-28). QTcF changes from Baseline in Category 1 were
detected in 18.4% of siponimod-treated subjects compared to 4.2% of placebo-subjects and did
not show any clear relation to the dose level. Category 3 events were seen in 3.4% of siponimod-
treated subjects compared to 0.4% in placebo-treated subjects and occurred at single doses of
10 mg.

Table 3-28 QTcF events categorical analysis (Holter data) - multiple dose studies

Anonymized Page 98
Novartis Confidential Page 99
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

QTcF change from Combined


New QTcF values
Baseline category
Category 1 2 3 4 5 6
QTc > 480
msec and
> 450 and > 480 and
> 30 > 60 > 500 change
≤ 480 ≤ 500
msec msec msec from
msec msec
Baseline >
60 msec
Treatment n (%) n (%) n (%) n (%) n (%) n (%)
BAF312 10mg
8 (57.1) 0 (0.0) 3 (21.4) 0 (0.0) 0 (0.0) 0 (0.0)
(N=14)
BAF312 2mg preceded
by up-titration 3 ( 2.5) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
(N=118)
BAF312 10mg
preceded by up-titration 17 (14.2) 0 (0.0) 3 ( 2.5) 0 (0.0) 0 (0.0) 0 (0.0)
(N=120)
BAF312 in combinationa
8 (21.1) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
(N=38)
Comparator alone
18 (13.7) 1 ( 0.8) 2 ( 1.5) 0 (0.0) 0 (0.0) 0 (0.0)
(N=131)
Placebo
10 ( 4.2) 0 (0.0) 1 ( 0.4) 0 (0.0) 0 (0.0) 0 (0.0)
(N=238)
All BAF
33 (18.4) 0 (0.0) 6 ( 3.4) 0 (0.0) 0 (0.0) 0 (0.0)
(N=179)
Any treatment
61 (14.0) 1 ( 0.2) 9 ( 2.1) 0 (0.0) 0 (0.0) 0 (0.0)
(N=436)
N = total number of subjects who got administered with at least one dose of respective treatment;
n = number of unique subjects meeting respective category criteria.
Included [Studies A2107, A2116 and A2118].
Time matched Baseline is considered. A subject with multiple occurrences within each category was
counted only once for that category and treatment.
a Combination studies included: Study A2116 (propranolol)
Source: [SCS Appendix 1-Table 12-3.2.2]

In PM/DM patients, 12-lead ECG analyses showed a placebo-like incidence of Category 1 and
3 QTcF events, while the incidence of events in Categories 2, 4, 5 and 6 was higher in
siponimod-treated patients compared to patients under placebo. Due to the overall small sample
sizes in the pooled PM/DM studies and potential confounding effects of the underlying disease
condition and concomitant use of standard of care drugs for treatment of PM/DM, the described
imbalance in the incidence of certain QTcF categories of interest should be interpreted with
caution.

Anonymized Page 99
Novartis Confidential Page 100
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.4.2.2.3 Effect on blood pressure


This section summarizes the results of pooled categorical analyses for SDP and DBP events
across multiple dose studies. The pooled categorical analyses of BP by SBP and DBP
assessments, described in Table 5-3, are available in [SCS Appendix 1-Table 12-7.1.1 to SCS
Appendix 1-Table 12-7.1.3 and SCS Appendix 1-Table 12-7.2.1 to SCS Appendix 1-
Table 12-7.2.3, respectively] for studies in healthy subjects (single and multiple dose) and
PM/DM patients.
Pooled categorical analyses under multiple dose conditions in both healthy subjects and
PM/DM patients revealed a higher incidence of categorical changes in SBP and DBP under
siponimod treatment than under placebo treatment, which were considered to be of no clinical
relevance. Occasional observations in higher BP event categories (i.e., more pronounced
changes) were seen at similar incidences in siponimod-treated and placebo-treated
subjects/patients and therefore did not indicate any clinically relevant effect of siponimod on
hemodynamic variables.
Overall, in single dose studies in healthy subjects there were no Category 6 observations for
SBP (new value > 160 mmHg) or DBP (new value > 110 mmHg), and incidences of
Categories 1 to 5 events in siponimod-treated subjects were similar to placebo. For both single
and multiple dose studies in healthy subjects, SBP and DBP observations for siponimod
treatment in the combined Category 7 (SBP increase > 15% from Baseline to a
SBP > 140 mmHg; DBP increase > 15% from Baseline to a DBP > 90 mmHg) were
placebo-like in incidence.
Under multiple dose conditions in healthy subjects, the incidences of SBP events were higher
in siponimod- compared to placebo-treated subjects (64.5% and 43.8%, respectively, in
Category 1; 48.5% and 32.1%, respectively, in Category 2; 40.8% and 26.2%, respectively, in
Category 3; [SCS Appendix 1-Table 12-7.1.2]). The frequencies of Category 5 and 6 events for
SBP were overall low and comparable for siponimod and placebo treatments (4.6% and 2.0%,
respectively, in Category 5; 1.3% and 1.0%, respectively, in Category 6).
Similarly, the incidences of DBP events were higher in siponimod- than placebo-treated subjects (65.3%
and 43.0%, respectively, in Category 1; 49.1% and 31.6%, respectively, in Category 2; 44.1% and
26.7%, respectively, in Category 3; [SCS Appendix 1-Table 12-7.2.2]). While there were no
observations in Category 6, the incidences of Category 5 DBP events were overall low and
comparable for siponimod and placebo treatments (2.0% and 1.3%, respectively).

3.4.2.2.4 Orthostatic hypotension/dysregulation


Orthostatic hypotension AEs were reported infrequently across the Clinical Pharmacology
studies and in studies with PM/DM patients. In [Study A2101], orthostatic hypotension was
reported only at the 75-mg dose. The relationship of the event of orthostatic hypotension to
treatment was considered suspected by the Investigator. One (1) AE of orthostatic
lightheadedness in a subject receiving propranolol on Days 1 to 20 was reported in
[Study A2116]; the event resolved spontaneously during treatment. Orthostatic hypotension
was among the 4 moderate AEs reported in subjects in the siponimod IR treatment group of
[Study A2119]. Overall, due to the low rate of orthostatic hypotension events, no clear dose
dependency or relation to siponimod treatment could be identified.

Anonymized Page 100


Novartis Confidential Page 101
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.4.2.3 Pulmonary effects


Due to the potential for bronchial smooth muscle constriction with S1P modulators such as
siponimod, PFTs were routinely performed in clinical trials. Overall, no clinically meaningfully
changes in pulmonary function were observed in healthy subjects or PM/DM patients.
Single doses of siponimod did not produce a substantial or significant effect upon FEV1 or
FEF25-75% ([Study A2101]). At doses of 10 mg and below, mean FEV1 values at 6 hours post
dose ranged between 2.881 and 3.668 L (observed for 2.5- and 5-mg doses, respectively), in
comparison to 3.41 L observed among subjects receiving placebo. A slight downward trend was
noted at doses above 10 mg, with FEV1 mean values of 3.000, 2.926 and 2.618 L for non-
therapeutic doses of 17.5, 25.0, and 75.0 mg, respectively.
No clinically significant changes in pulmonary function were noted in healthy subjects
receiving multiple doses of siponimod. In [Study A2105], changes in PFT variables (treatment
differences in all siponimod groups compared to time-matched placebo) ranged from -0.45 to -
0.09 L (between 12.2% decrease and 2.4% decrease, respectively) for FEV1 and from -0.70 to
+0.10 L/s (between 17.5% decrease and 2.5% increase, respectively) for FEF25-75%. These
changes were considered mild to moderate and their magnitude seemed to be unrelated to dose
and day in any treatment groups. Furthermore, across all siponimod dose groups, no significant
difference or trend was observed in the proportion of subjects with less than 10% FEV1
decrease. In [Study A2116], all subjects in the 4 dosing groups for either siponimod or
propranolol given alone or as combination treatment (Section 2.2.7.2.3) had less than 15%
decrease in FEV1 or slight increase (< 5%) at steady state (Day 19) in comparison to Baseline.
In the 3 multiple dose studies of siponimod in PM/DM patients, the only PFT findings were
mild to moderate AEs of lung function impairment or decreased diffusing capacity for carbon
monoxide (DLCO) reported in 1 subject in [Study X2205] and 2 subjects in [Study X2206].

3.4.2.4 Other pharmacodynamic effects


This section discusses the results of pooled analyses for the key secondary safety-related PD
effect of siponimod including on LFTs (Section 3.4.2.4.1), prospective assessment of suicidal
ideation/behavior (Section 3.4.2.4.2), infection rates (Section 3.4.2.4.3) and the results of the
comprehensive abuse/dependence potential assessment (Section 3.4.2.4.4).

3.4.2.4.1 Hepatoxicity/liver transaminase and bilirubin elevations


Elevated liver transaminases have been observed in siponimod clinical trials at a higher
incidence under active siponimod treatment compared to placebo. This section summarizes the
results of pooled analyses for selected routine clinical laboratory variables (ALT, AST, GGT,
AP and bilirubin) related to hepatic function, based on pre-defined categories of interest under
single and multiple dose conditions and in healthy subjects and patients with PM or DM. LFTs
were also evaluated in the individual studies in Section 2.2 and Section 2.3. Events of elevated
LFTs were mostly asymptomatic and not exceeding CTCAE Grade 3. There were no cases of
Hy’s law.
The pooled categorical analysis of LFTs (ALT, AST, GGT, AP and total bilirubin events),
described in Table 5-3, are available in [SCS Appendix 1-Tables 12-6.1.1 to 12-6.1.3] and [SCS

Anonymized Page 101


Novartis Confidential Page 102
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Appendix 1-Table 12-6.2.1 to SCS Appendix 1-Table 12-6.2.3] (combined categories) for
studies in healthy subjects (single and multiple dose) and PM/DM patients.
In both healthy subjects and PM/DM patients there were no events in the combined Categories
5 and 6 (ALT or AST > 3-fold and bilirubin > 2-fold above ULN), confirming that no Hy’s law
cases were identified. Under single and multiple dose conditions in healthy subjects, ALT, AST
and GGT increases ≥ 3-fold ULN (Categories 2 to 4) generally occurred at placebo-like
frequencies of approximately 1% or less; these instances occurred primarily above the 2 mg
therapeutic dose in siponimod-treated subjects.
Under multiple dose conditions in healthy subjects, no events of AP elevations falling within
the predefined categories were identified. For bilirubin only Category 1 events were identified
and the overall incidence was very low (1.3% in siponimod-treated and 0.3% in placebo
subjects). For ALT, AST and GGT, the majority of events were in Category 1 (12.4%, 4.4%
and 6.5%, respectively) and were higher in siponimod-treated subjects (17.5%, 6.3% and 8.1%,
respectively) than placebo-treated subjects, generally occurring more often at doses > 2 mg
(siponimod-only titrated and non-titrated treatment; average: 30.0% (15.8% to 46.9%), 12.7%
(3.1% to 38.9%) and 13.2% (3.1% to 33.3%), respectively). For bilirubin only Category 1
events were identified and the overall incidence was very low (1.3% of siponimod-treated
subjects and 0.3% of placebo subjects). For ALT, AST and GGT, the majority of events were
in Category 1. For AST, 6.3% of siponimod-treated subjects (mainly at doses ≤ 2 mg) and 0.5%
of placebo-treated subjects had values in Category 1, compared to < 1% each for Categories 2,
3 or 4 for either treatment. For ALT, the incidence of Category 1 events was 17.5% of
siponimod-treated subjects compared to 3.3% of placebo-treated subjects, with an increase in
incidence at doses > 2 mg. Category 2 events were identified in 4.2% of siponimod-treated
subjects compared to 0.5% of placebo-treated subjects, with a clear increase in incidence at
doses > 2 mg. In Categories 3 and 4, the overall incidences were very low (1.0% and 0.1% of
subjects, respectively). For GGT, the incidence of Category 1 events was 17.5% of siponimod-
treated subjects (with similar proportions of subjects at doses ≤ 2 mg or > 2 mg), compared to
3.3% of placebo-treated subjects. Category 2 events were identified in 2.2% of siponimod-
treated subjects compared to 0% of placebo-treated subjects, with an increase in incidence at
doses > 2 mg. In Categories 3 and 4, the overall incidences were very low (0.7% and 0.1% of
subjects, respectively).
Due to the underlying disease effects on liver enzymes (Nathwani et al 2005, Wada et al 2016)
and multiple concomitant and standard of care medications in the PM/DM studies, an influence
of the disease activity and of co-administered drugs on the analyzed laboratory variables cannot
be ruled out; thus, the pooled LFT analyses in PM/DM patients should be interpreted in this
context. There were no patients in the PM/DM multiple dose studies with values in Categories 3
or 4 for any of the LFT parameters. For AP and bilirubin, there were also no patients with values
in Categories 1 or 2. The incidence of values in Categories 1 and 2 for ALT, AST and GGT
were higher in siponimod-treated patients compared to siponimod-treated healthy subjects:
ALT 20.9% and 11.6%, respectively; AST 16.3% and 7.0%, respectively; and GGT 12.2% and
8.2%, respectively. Comparing to placebo-treated PM/DM patients, similar percentages were
observed for ALT; however, Category 1 and 2 percentages were 10.0% and 15.0%, respectively,
for AST and 5.0% and 0%, respectively, for GGT in placebo-treated patients.

Anonymized Page 102


Novartis Confidential Page 103
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

3.4.2.4.2 Suicidal ideation and behavior


Prospective assessments of suicidal ideation and behavior were included as part of the safety
assessments in multiple dose studies with siponimod. Suicidal ideation and behavior assessment
in MS patients are discussed in [SCS-Section 2.1.5.9].
Prospective suicidal ideation and behavior assessments from 9 multiple dose studies in healthy
subjects and PM/DM patients using the Columbia Suicide Severity Rating Scale (C-SSRS) did
not reveal any risk for suicidal ideation or behavior ([Studies A2116, A2118, A2121, A2122,
A2125, A2128, A2130, X2205 and X2206]). In (Study A2130), 1 subject reported a prior
episode of suicidal ideation at Screening as part of the medical history, while none of the
subjects reported any sign of suicidal ideation or behavior during the study.

3.4.2.4.3 Infections
Considering the primary PD effect of siponimod on lymphocytes, potential translation to an
increased rate of infections was explored and is discussed in detail in [SCS-Section 2.1.5.1 and
SCS-Sections 2.1.5.2]. A pooled analysis of AEs by PT (detailed in Table 5-3) assessed the
incidence of infection-related AEs as reported in the Clinical Pharmacology and PM/DM
studies.
The most commonly reported infection-related AE was upper respiratory tract infection, with
overall slightly higher incidence rates for siponimod treatment (2.0%, 1.8% and 7.0% in single
dose, multiple dose and PM/DM studies, respectively) compared to placebo treatment (0%, 1.0%
and 0% in single dose, multiple dose and PM/DM studies, respectively). However, the
incidence of such infections was similar to placebo at the therapeutic dose of 2 mg (1.7%
incidence for siponimod uptitration to 2 mg, compared to 1.0%). Urinary tract infections were
reported at low incidence rates across all types of studies, at incidence rates of 0.3%, 0.2% and
4.7% for siponimod treatment in single dose, multiple dose, and PM/DM studies, respectively,
compared to no events for placebo treatment in the single dose, multiple dose or PM/DM studies.
Other types of infection-related AEs were reported in single subjects or at incidence rates ≤ 0.3%
without a distinguishable difference between siponimod- and placebo-treated subjects/patients;
these included herpes zoster infection (mild, skin involvement only), viral upper respiratory
tract infections, respiratory tract infection, tooth infection, viral infection and vulvovaginal
mycotic infection.

3.4.2.4.4 Abuse and dependence potential (8-factor drug abuse liability assessment)
A comprehensive 8-factor abuse potential assessment was carried on the basis of relevant
chemistry, nonclinical and clinical data of siponimod and including post-marketing data on
fingolimod ([Abuse Potential Assessment]).
Overall, chemistry, nonclinical and clinical data with siponimod do not indicate any signals of
abuse, misuse or dependence potential in animals or humans, nor do the data demonstrate any
potential pharmacological similarities to existing drugs of abuse or psychoactive effects that
may be of interest for drug abuse, such as reinforcing, mood-elevating, sedative, stimulant,
hallucinogenic or acute cognitive effects. These data are consistent with post-market data for
the pharmacologically similar drug fingolimod, which has not shown any signs of abuse, misuse,
diversion or dependence in the community. Therefore, it can be concluded that siponimod has

Anonymized Page 103


Novartis Confidential Page 104
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

no abuse or dependence potential and is not expected to be subject to abuse, misuse or diversion
in the community, or result in harm to public health as a result of abuse, misuse or dependence.
Thus, siponimod does not meet the criteria for any of the schedules defined under the Controlled
Substances Act and therefore should not be scheduled.

3.4.3 Exposure – lymphocyte relationship


This section discusses the exposure-lymphocyte relationship investigated in
[CBAF312A-Phase 3-PopPKPD] (Table 5-3). Overall, the dose-dependent inhibition of ALC
was well captured by the indirect response model described below. At the therapeutic dose of 2
mg qd, it is expected that for the majority of subjects siponimod treatment will not result in a
reduction of ALC values below 0.2 × 109/L. Among covariates that were examined for potential
to affect ALC, the only predictors of lymphocyte response were population type (healthy
subjects different from RRMS patients different from SPMS patients), gender and Japanese
ethnicity. A discussion of the recommended therapeutic dose based on a MRI-based CUAL
analysis from the Phase 2 RRMS [Study A2201], which revealed a near-maximal efficacy at
2 mg, is discussed in [SCE-Section 4.1].
In the PopPKPD analysis, a total of 25003 ALC records from 2591 subjects in SPMS and
RRMS patient studies and healthy subject studies were included. Of these, 8837 were from
healthy subjects, 1486 from RRMS patients and 14680 from SPMS patients. There were 17529
observations collected under active siponimod treatment and 7474 observations from
placebo-treated subjects. The population included 63 Japanese and 21 Chinese subjects.
The final model for ALC was an indirect response model with inhibition of lymphocyte
circulation by siponimod concentrations. The percent decrease of circulation rate was a
saturable Imax function of drug concentrations. In the reference population of male healthy
subjects, the Imax was estimated to be 79% (95% CI: 78% to 80%). The siponimod IC50 was
estimated to be 4.94 ng/mL.
Among the covariates that were tested (age, body weight, gender, ethnicity (Chinese and
Japanese) and disease status/population type), only population type, gender and Japanese
ethnicity were predictors of lymphocyte response. Table 3-29 displays the effect of these
covariates on the PD model parameters. Both RRMS and SPMS patients had a 17% lower
Baseline lymphocyte count compared to healthy subjects. Japanese subjects had a 10% higher
Imax than non-Japanese. Females had a 37% lower IC50 than males and SPMS patients had a
39% lower IC50 than both RRMS patients and healthy subjects. The effects of gender and
population type on IC50 and of ethnicity on Imax should however be interpreted with caution
as they might be significantly biased. Due to sparseness of the data the individual estimates of
Imax and IC50 were not well estimated (high shrinkage of 69% for Imax and 83% for IC50),
making the estimation of covariate effects on these parameters rather uncertain.

Anonymized Page 104


Novartis Confidential Page 105
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-29 Effect of covariates on the lymphocyte response model parameters


Model parameter Effect of the covariate
Baseline lymphocyte count 17% (95% CI: 15% to 19%) lower in both RRMS and SPMS
patients than in healthy subjects
Imax 10% (95% CI: 7% to 13%) higher in Japanese than in
non-Japanese subjects
IC50 37% (95% CI: 31% to 43%) lower in females than in males
IC50 39% (95% CI: 34% to 45%) lower in SPMS patients than in both
RRMS patients and healthy subjects
Source: [CBAF312A-Phase 3-PopPKPD-Section 5-11]

The impact of population type, gender and Japanese ethnicity and also body weight and
CYP2C9 genotypes (significant predictors of siponimod plasma concentrations, as described in
Section 3.1.9.2) on lymphocyte response was evaluated using trough simulations
(500 lymphocyte count profiles were simulated for each subject). All were conducted over
130 days with 2 mg siponimod administered qd for 40 days (steady-state conditions) for a
typical subject (SPMS patient, male, 70.5 kg, non-Japanese with CYP2C9*1*1 or *1*2
genotype), or different subjects changing 1 covariate at a time and keeping the others at typical
values.
Figure 3-11 represents the obtained simulated median ALC for the typical subject and by gender,
population type (RRMS and healthy subjects) and Japanese ethnicity. This shows, for all
simulated populations, a sharp decrease in the lymphocyte count upon treatment initiation
(Day 0), then values remaining at a plateau during the chronic treatment followed by a return
to Baseline starting promptly after treatment discontinuation (Day 39). Similar graphs for
lymphocyte count profiles were produced for different body weights and different CYP2C9
genotypes, showing the same type of trajectories of lymphocyte count
([CBAF312A-Phase 3-PopPKPD]).

Anonymized Page 105


Novartis Confidential Page 106
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Figure 3-11 Simulated median of absolute lymphocyte count profiles following


administration of 2 mg siponimod once daily for 40 days in typical
SPMS patients, by gender, ethnicity and population type

HV = healthy volunteer.
500 patients per profile, residual variability taken into account
Source: [CBAF312A-Phase 3-PopPKPD-Figure 5-20]

For each covariate, the lymphocyte count at steady state and the lymphocyte recovery time to a
level of 1.0 × 109/L after treatment discontinuation are displayed in Table 3-30 and Table 3-31,
respectively. The lymphocyte recovery time to a level of 90% of the Baseline value after
treatment interruption was also computed ([CBAF312A-Phase 3-PopPKPD-Table 5-15]).
These values are approximately 2 to 2.5 times greater than the values reported for the recovery
time to a level of 1.0 × 109/L (see below).

Anonymized Page 106


Novartis Confidential Page 107
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Compared to the typical SPMS patients (70.5 kg), for a 2-mg dose, lighter patients (49 kg, 5th
percentile of weight in the Phase 3 study) had a 9% lower nadir median steady-state lymphocyte
count and identical lymphocyte recovery time and heavier patients (100 kg, 95th percentile of
weight) had a 4% higher nadir median steady-state lymphocyte count and slightly faster
lymphocyte recovery ([CBAF312A-Phase 3-PopPKPD-Figure 5-16]). Patients with
CYP2C9*2*3 and *1*3 genotypes had comparable nadir lymphocyte count. The CYP2C9*2*2,
*1*3, *2*3 and *3*3 genotypes had lower nadir (9%, 12%, 16% and 26%, respectively)
lymphocyte count than the reference CYP2C9*1*1 (wild type) and *1*2 genotypes.
Lymphocyte counts recovered in the following order, by CYP2C9 genotype: *1*1 and *1*2
first, then *2*2, next *1*3 then *2*3 and lastly *3*3. In both cases, the lower lymphocyte
counts in lighter patients and poor metabolizers (CYP2C9*2*2, *1*3, *2*3 and *3*3) were due
to higher plasma siponimod exposures in lighter patients compared to heavier patients and
higher plasma siponimod exposure in intermediate and poor metabolizers compared to patients
with CYP2C9*1*1 or *1*2 genotypes. (PD assessments in CYP2C9 extensive and poor
metabolizers are described in Section 2.2.10.3.)
Females had an 8% lower nadir median steady-state lymphocyte count than males, and
lymphocyte counts in females recovered to 1.0 × 109/L at the same speed as in males. The lower
nadir lymphocyte count in females was due to the 37% (95% CI: 31% to 43%) smaller IC50 in
females compared to males.
Japanese patients had a 23% lower nadir median steady-state lymphocyte count than
non-Japanese, and lymphocyte recovery time to 1.0 × 109/L in Japanese patients was slightly
slower than in non-Japanese patients (1 day slower). The lower nadir lymphocyte count in
Japanese was due to the 10% (95% CI: 7% to 13%) higher Imax in Japanese compared to non-
Japanese. (PD assessments in Japanese subjects are described in Section 2.2.10.2.)
RRMS patients had a 16% higher nadir median steady-state lymphocyte count than SPMS
patients, and RRMS and SPMS patients had a 17% and 29%, respectively, lower nadir median
lymphocyte count than healthy subjects. In addition, lymphocyte counts in healthy subjects
recovered slightly more quickly to 1.0 × 109/L than in RRMS patients, both of whom had
quicker lymphocyte recovery time than SPMS patients. This was a result of RRMS and SPMS
patients having lower (17% (95% CI: 15% to 19%)) Baseline lymphocyte count than healthy
subjects, and SPMS patients having lower IC50 (39% (95% CI: 34% to 45%)) compared to
healthy subjects and RRMS patients. The median simulated lymphocyte counts at Baseline were
1.72 and 1.73 × 109/L for RRMS and SPMS patients, respectively, which was 17% and 16%
lower, respectively, than the median simulated lymphocyte count at Baseline for healthy
subjects.

Anonymized Page 107


Novartis Confidential Page 108
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-30 Median lymphocyte count at the end of the last day of dosing in
patients with secondary progressive multiple sclerosis following
administration of 2 mg siponimod once daily for 40 days (steady-state
conditions)
Median (90% PI) lymphocyte count at
Category of patients end of last day of dosing (× 109/L)
Typical (70.5 kg male, *1*1 or *1*2, non-Japanese SPMS
0.560 (0.271-1.08)
patient)
Typical, but 49 kg only 0.510 (0.224-1.03)
Typical, but 100 kg only 0.583 (0.285-1.09)
Typical, but *2*2 CYP2C9 genotype only 0.509 (0.276-1.02)
Typical, but *1*3 CYP2C9 genotype only 0.491 (0.239-0.970)
Typical, but *2*3 CYP2C9 genotype only 0.473 (0.247-0.927)
Typical, but *3*3 CYP2C9 genotype only 0.416 (0.186-0.851)
Typical, but female only 0.515 (0.245-0.922)
Typical, but healthy subject only 0.784 (0.402-1.40)
Typical, but RRMS patient only 0.652 (0.309-1.22)
Typical, but Japanese only 0.434 (0.217-0.811)
Source: [CBAF312A-Phase 3-PopPKPD-Table 5-14]

Table 3-31 Time in days needed for absolute lymphocyte count recovery to
1.0 x 109/L for 2 mg siponimod once daily for 40 days (steady-state
conditions)
Median time for ALC recovery to 1.0 × 109/L
Categories of patients and doses
(90% PI) (days)
Typical (70.5 kg male, *1*1 or *1*2, non-
5 (1-11)
Japanese SPMS patient)
Typical, but 49 kg only 5 (1-13)
Typical, but 100 kg only 4 (1-10)
Typical, but *2*2 CYP2C9 genotype only 6 (1-15)
Typical, but *1*3 CYP2C9 genotype only 8 (2-18)
Typical, but *2*3 CYP2C9 genotype only 10 (2-23)
Typical, but *3*3 CYP2C9 genotype only 24 (4-48)
Typical, but female only 5 (2-12)
Typical, but healthy subject only 2 (1-6)
Typical, but RRMS patient only 3 (1-9)
Typical, but Japanese only 6 (2-12)
Source: [CBAF312A-Phase 3-PopPKPD-Table 5-16]

The PopPKPD model was also used to simulate the trough siponimod plasma concentration and
the absolute lymphocyte count at steady state as a function of siponimod dose in 70.5 kg healthy,
non-Japanese male subjects with CYP2C9*1*1 or *1*2 genotypes. The results, displayed in
Table 3-32 and Figure 3-12, show that while the trough siponimod concentrations increase in a
dose-proportional manner, the nadir lymphocyte count approaches a minimum value
asymptotically.

Anonymized Page 108


Novartis Confidential Page 109
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 3-32 Median (90 percent prediction interval) simulated trough siponimod
plasma concentration and trough lymphocyte count at steady state in
healthy subjects by dose
Median (90% PI)
Trough siponimod concentration (ng/mL) Trough lymphocyte count
Dose (mg) (× 109/L)
0.1 1.05 (0.54, 1.75) 1.73 (1.07, 2.60)
0.25 2.66 (1.35, 4.51) 1.43 (0.837, 2.22)
0.5 5.18 (2.62, 8.80) 1.21 (0.660, 1.90)
1 10.5 (5.57, 16.5) 0.978 (0.461, 1.59)
2 21.1 (11.2, 35.7) 0.782 (0.364, 1.28)
4 41.5 (20.9, 70.4) 0.636 (0.284, 1.10)
10 102 (53.3, 182) 0.547 (0.216, 0.935)
20 210 (108, 351) 0.517 (0.187, 0.914)
Source: [CBAF312A-Phase 3-PopPKPD-Table 5-17]

Figure 3-12 Median simulated steady-state trough siponimod concentration


versus trough lymphocyte count in healthy subjects by dose

The width and height of the boxes represent the 90% PI.
Source: [CBAF312A-Phase 3-PopPKPD-Figure 5-23]

4 Special studies
Not applicable.

Anonymized Page 109


Novartis Confidential Page 110
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

5 Appendix
Table 5-1 Clinical Pharmacology studies in healthy subjects and special populations and studies in patients with MS, PM
and DM
Study (status) Study objective Number of subjects per dose/route Study design Siponimod dose/regimen
(formulations)
Healthy subjects and special population studies (N = 20)
[Study A2101] SAD study to explore safety, N = 98 healthy subjectsa: Two-part, single center, 0.1, 0.3, 1 and 2.5 mg
(completed) tolerability, PK and PD N = 80 to siponimod randomized, double-blind, (liquid); 2.5, 5, 10, 17.5, 25
placebo-controlled, SAD and 75 mg (CSF capsules)
N = 18 to placebo Phase 1 study
[Study A2102] MAD study to evaluate N = 60 healthy subjectsa: Randomized, parallel, double- 0.3 and 1.0 mg (liquid); 2.5,
(completed) safety, tolerability, PK and N = 45 to siponimod blind, placebo-controlled, 10 and 20 mg (CSF
PD time-lagged, MAD Phase 1 capsules)
N = 15 to placebo study
[Study A2105] MAD study to evaluate N = 50 healthy subjectsa: Randomized, parallel, 0.3 and 1.0 mg (liquid); 2.5,
(completed) safety, tolerability, PK and N = 37 to siponimod double-blind, 10 and 20 mg (CSF
PD (repeat of Study A2102) placebo-controlled, capsules)
N = 13 to placebo time-lagged, MAD Phase 1
study
[Study A2107] DT and fixed multiple dose N = 56 healthy subjectsa in 4 parallel treatment Double-blind, Uptitration (8 or 9 days) to
(completed) study to investigate the groups: placebo-controlled, 10 mg (0.25, 1, 4 and 5 mg
negative chronotropic effect N = 42 to siponimod 10 mg, including parallel-group Phase 1 study MF tablets)
of siponimod 2 DT regimens and 1 non-titrated regimen
(N = 14 each)
N = 14 to placebo

Anonymized Page 110


Novartis Confidential Page 111
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study (status) Study objective Number of subjects per dose/route Study design Siponimod dose/regimen
(formulations)
[Study A2110] Multiple dose study to N = 138 healthy subjectsa to: Randomized, partially 0.5, 1, 2 and 4 mg (MF
(completed) investigate the effect of 7 treatment/discontinuation sequences (N = 19 double-blind, tablets)
siponimod treatment to 20 each), each consisting of 3 periods to placebo-controlled Phase 1
re-initiation on the initial evaluate a total of 4 dose levels of siponimod or study with 3 periods (10 days
negative chronotropic effect placebo and 5 discontinuation periods of single dose drug treatment,
(21 conditions total) drug discontinuation, 1 day of
single dose drug re-initiation)
[Study A2104] Single dose ADME study in N = 4 healthy male CYP2C9*1*1 subjects to Open-label, single oral dose 10 mg/2 mL [14C]siponimod
(completed) healthy subjects with the [14C]-siponimod Phase 1 study concentrate, diluted and
CYP2C9*1*1 genotype administered as 4 mL oral
solution
[Study A2111] Single dose study in healthy N = 62 healthy CYP2C9*1*1 subjects to 1 of 12 Randomized, open-label, 3- 0.25 mg MF tablet fasted
(completed) subjects with the sequences: period, 3-treatment, 0.25 mg FMI tablet fasted
CYP2C9*1*1 genotype to 6 sequences with 0.25 mg siponimod 6-sequence, single dose,
assess both the crossover Phase 1 study 0.25 mg FMI tablet fed
bioequivalence of the 6 sequences with 4 mg siponimod 4 mg MF tablet fasted
siponimod FMI tablet 4 mg FMI tablet fasted
formulation as compared to
4 mg FMI tablet fed
the siponimod MF and the
effect of food on the PK of
the FMI
[Study A2119] Single dose study to assess N = 60 healthy subjectsa: Randomized, double-blind, 4 mg for F16 and F10 (MR
(completed) the tolerability, PD and PK of N = 45 to siponimod F16, F10 and IR placebo-controlled, parallel formulations tablets) and IR
2 MR siponimod tablets tablet formulations (N = 15 each) group, single dose Phase 1 formulation (MF tablets)
(F16, F10) compared to the study
IR tablet (MF) and placebo N = 15 to placebo

Anonymized Page 111


Novartis Confidential Page 112
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study (status) Study objective Number of subjects per dose/route Study design Siponimod dose/regimen
(formulations)
[Study A2125] Multiple dose, 2-period, N = 16 healthy CYP2C9*1*1 subjects: Open-label, 2-period, drug Uptitration (5 days) to 2 mg
(completed) single-sequence study in Period 1 (Days 1 to 12): siponimod interaction Phase 1 study (0.25, 0.5 and 2 mg FMI
healthy subjects with the uptitration to 2 mg tablets)
CYP2C9*1*1 genotype to
evaluate the effect of the Period 2 (Days 13 to 24): siponimod 2 mg
CYP2C9/3A4 inducer and rifampin 600 mg
rifampin on siponimod PK
[Study A2108] Single dose study in healthy N = 14 healthy CYP2C9*1*1 subjects: Open-label, single dose, 4 mg (MF tablets)
(completed) subjects with the Period 1 (14 days): siponimod single dose 2- period, drug interaction
CYP2C9*1*1 genotype to (Day 1) Phase 2 study
evaluate the safety,
tolerability and PK of Period 2 (20 days): siponimod single dose
siponimod when given alone (Day 3) and 200 mg qd fluconazole (Days
and in combination with the 1 to 19)
CYP2C9/3A4 inhibitor
fluconazole
[Study A2124] Single dose study in healthy N = 30 healthy CYP2C9*1*2 (17) and *1*3 (13) Open-label, 3-period, single- 0.25 mg (FMI tablets)
(completed) subjects with CYP2C9*1*2 subjects: sequence, crossover, drug
and *1*3 genotypes to Period 1 (Days 1 to 14): 0.25 mg interaction Phase 1 study
evaluate the effect of the siponimod single dose (Day 1)
CYP3A4 inhibitor
itraconazole on siponimod Period 2 (Days 15 to 18): 100 mg
PK, safety, and tolerability itraconazole bid
Period 3 (Days 19 to 32/36): 0.25 mg
siponimod single dose (Day 19), and
100 mg itraconazole bid (Days 19 to 31
(*1*2) or Days 19 to 35 (*1*3))

Anonymized Page 112


Novartis Confidential Page 113
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study (status) Study objective Number of subjects per dose/route Study design Siponimod dose/regimen
(formulations)
[Study A2130] Multiple dose study to N = 136 healthy subjects (CYP2C9*3*3 Randomized, double-blind, Uptitration (5 days) to 2 mg
(completed) evaluate the modulation of genotype subjects excluded) to 4 treatment placebo-controlled, (0.25, 0.5, 1 and 2 mg FMI
immune response to T-cell groups: parallel-group, multiple dose tablets)
dependent and T-cell N = 34 to vaccination (Day 21) with Phase 1 study
independent antigen stimuli interrupted siponimod treatment (Days 1
by preceding, concomitant to 10 and 35 to 48)
and interrupted
administration of multiple N = 34 to vaccination (Day 21) with
therapeutic doses of preceding siponimod treatment (Days 1 to
siponimod 13)
N = 34 to vaccination (Day 21) under
concomitant siponimod treatment (Days
11 to 48)
N = 34 to vaccination (Day 21) with placebo
treatment
[Study A2121] Multiple dose study in N = 24 healthy female CYP2C9*1*1 subjects: Open-label, multiple dose, Uptitration (6 days) to 4 mg
(completed) female healthy subjects with Period 1 (Days 1 to 21): OC (30 µg EE 2-period Phase 1 study (0.25, 1 and 4 mg MF
the CYP2C9*1*1 genotype and 150 µg LVG) tablets)
to evaluate the effect of oral
siponimod on the PK and Period 2 (Days 29 to 49): 4 mg siponimod
PD of monophasic oral (uptitration beginning on Day 23, during a
contraceptive 7-day OC pill-free period) and OC (30 µg
EE and 150 µg LVG)

Anonymized Page 113


Novartis Confidential Page 114
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study (status) Study objective Number of subjects per dose/route Study design Siponimod dose/regimen
(formulations)
[Study A2116] Multiple dose study to N = 76 healthy subjects (CYP2C9*3*3 genotype Randomized, double-blind, Uptitration (5 days) to 2 mg
(completed) evaluate PD and/or PK subjects excluded) to 1 of 4 parallel treatment placebo-controlled, multiple (0.25 and 1 mg MF tablets)
interaction of siponimod and arms (N = 19 each): dose Phase 1 study
propranolol Group A: propranolol 80 mg (Days 11 to
co-administration 20) added on top of siponimod 2 mg
(Days 1 to 20)
Group B: siponimod 2 mg (Days 11 to 20)
added on top of propranolol 80 mg (Days
1 to 20)
Group C: placebo (Days 1 to 20)
Group D: propranolol 80 mg (Days 1 to
20)
[Study A2126] Single dose, 2-part study in N = 33 healthy CYP2C9*1*1 subjects: Open-label, single dose, Part 1: 0.25 mg 3-hour iv
(completed) healthy subjects with the Part 1: N = 16 to 1 of 2 sequences of 2-part Phase 1 study infusion, and 0.25 mg FMI
CYP2C9*1*1 genotype to 2 periods each (2 sequences with 0.25 mg tablet
measure the absolute siponimod (oral and iv)) Part 2: 1.0 mg 24-hour iv
bioavailability, safety, infusion (4 consecutive
tolerability, and PD of oral Part 2: N = 17 (1 mg siponimod (iv))
6-hour iv infusions)
and iv siponimod
[Study A2118] Multiple dose thorough QT N = 304 healthy subjects (CYP2C9*3*3 Randomized, double-blind, Uptitration (5 days) to 2 mg
(completed) study to assess the effects genotype subjects excluded): parallel-group, placebo- and and (3 days) to 10 mg (0.25,
on QT interval (cardiac N = 99 to siponimod uptitration to 2 mg moxifloxacin-controlled Phase 1, 4 and 5 mg MF tablets)
repolarization) at oral (Days 1 to 10) and siponimod uptitration 1 study
therapeutic and to10 mg (Days 11 to 18)
supratherapeutic doses of
siponimod N = 103 to 400 mg moxifloxacin (Days 10
and 18)
N = 102 to placebo

Anonymized Page 114


Novartis Confidential Page 115
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study (status) Study objective Number of subjects per dose/route Study design Siponimod dose/regimen
(formulations)
[Study A2122] Single dose study in N = 40 CYP2C9*1*1 subjects to siponimod: Single dose, open-label, 0.25 mg (MF tablets)
(completed) subjects with the N = 24 hepatic impairment (mild, parallel-group Phase 1 study
CYP2C9*1*1 to compare the moderate and severe; N = 8 each)
PK, safety and tolerability of
siponimod in subjects with N = 16 healthy control subjects
mild, moderate and severe
hepatic impairment and
healthy control subjects
[Study A2129] Single dose study in N = 16 CYP2C9*1*1 subjects to siponimod: Single dose, open-label, 0.25 mg (FMI tablets)
(completed) CYP2C9*1*1 subjects (wild N = 8 severe renal impairment parallel-group Phase 1 study
type genotype) to compare
the PK, safety and N = 8 healthy control subjects
tolerability of siponimod in
subjects with renal
impairment and normal renal
function
[Study A1101] SAD study to evaluate N = 40 healthy male Japanese subjectsa: Randomized, double-blind, 0.5, 2.5, 5 and 10 mg (MF
(completed) safety, tolerability, PK and N = 8 each to 0.5 to 10 mg siponimod placebo-controlled, ascending tablets)
PD in Japanese subjects single dose Phase 1 study
N = 8 to placebo
[Study A2128] Study to assess the safety, N = 24 CYP2C9 poor and extensive Open-label, 2-part, single and 0.25 and 0.5 mg (MF tablets)
(completed) tolerability, and PK of metabolizer subjects: multiple dose Phase 1 study
siponimod in subjects with N = 12 CYP2C9*1*1 extensive
CYP2C9 extensive metabolizers
metabolizers (CYP2C9*1*1
genotype) and poor N = 12 CYP2C9*2*3/*3*3 poor
metabolizers (CYP2C9*2*3 metabolizers
or *3*3 genotype) Part 1: single dose of 0.25 mg siponimod
Part 2 (poor metabolizers only): 0.25 mg
siponimod (Days 1 and 2) and 0.5 mg
siponimod (Day 3)

Anonymized Page 115


Novartis Confidential Page 116
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study (status) Study objective Number of subjects per dose/route Study design Siponimod dose/regimen
(formulations)
Studies in patients with MS (N = 2)
[Studies A2201 and Study, with long-term Core: N = 297 RRMS patients (CYP2C9*3*3 Double-blind, randomized, Core: 0.5, 2 and 10 mg in
A2201E] extension, to evaluate genotype patients excluded): multi-center, adaptive, Period 1; 0.25 mg or
(completed) safety, tolerability, efficacy, N = 235 to siponimod dose-ranging, uptitration (4 days) to
and dose response curve of placebo-controlled, 1.25 mg in Period 2 (MF
siponimod in patients with N = 62 to placebo parallel-group, 2-period tablets)
RRMS Extension: N = 184 (from core study) to Phase 2 study Extension: 0.25 mg or
siponimod uptitration (2 days) to
0.5 mg, (4 days) to 1.25 mg,
(5 days) to 2 mg or (9 days)
to 10 mg in dose-blinded
phase; 2 mg (5-day
uptitration if had received
0.25- or 0.5-mg dose prior)
in open-label phase (MF and
FMI tablets)
[Study A2304] Study to evaluate efficacy, Core: N = 1651 SPMS patients: Multi-center, randomized, Core: Uptitration (5 days) to
(ongoing extension safety, and tolerability data N = 1105 siponimod double-blind, parallel-group, 2 mg (dose reduction to
part) in patients with SPMS, and placebo-controlled variable 1 mg was permitted) (FMI
to provide long-term data in N = 546 to placebo treatment duration Phase 3 tablets)
patients who complete the study
core part
Studies in patients with PM and DM (N = 3)
[Study A2202] Study to evaluate the N = 18 patients with PM (N = 11) and DM (N = Multi-center, parallel, double- Uptitration (9 days) to 10 mg
(completed) efficacy and tolerability of 7) (CYP2C9*3*3 genotype patients excluded): blind, placebo controlled, (0.25, 1, 4 and 5 mg tablets)
siponimod in patients with N = 8 to siponimod/siponimod proof-of-concept Phase 2 (MF tablets)
PM and DM (Period 1/Period 2 extension) study with an open-label
extension phase (Period 2)
N = 10 to placebo/siponimod
(Period 1/Period 2 extension)

Anonymized Page 116


Novartis Confidential Page 117
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study (status) Study objective Number of subjects per dose/route Study design Siponimod dose/regimen
(formulations)
[Study X2205] Proof-of-concept study to N = 14 patients with PM (excluding Double-blind, randomized, Uptitration (5 days) to 2 mg
(completed) evaluate the efficacy and CYP2C9*3*3 genotype patients and potentially placebo-controlled, or (9 days) to 10 mg (0.25,
tolerability of siponimod in *1*3, 2*2*, and *2*3 genotype patients, parallel-group Phase 2 study 0.5, 1 and 2 mg tablets) (MF
patients with PM depending on concomitant medication use): with an extension phase and FMI tablets)
N = 7 to siponimod 2 mg/2 mg (Period 2)
(Period 1/Period 2 extension)
N = 2 to siponimod 10 mg/10 mg
(Period 1/Period 2 extension)
N = 4 to placebo/siponimod 2 mg
(Period 1/Period 2 extension)
N = 1 to placebo/siponimod 10 mg
(Period 1/Period 2 extension)
[Study X2206] Study to evaluate safety, N = 17 patients with DM (excluding Double-blind, randomized, Uptitration (2 days) to
(completed) tolerability, efficacy and CYP2C9*3*3 genotype patients and potentially placebo-controlled Phase 2 0.5 mg, (5 days) to 2 mg
preliminary dose-response *1*3, 2*2*, and *2*3 genotype patients, study with an extension and/or (9 days) to 10 mg
of siponimod in patients with depending on concomitant medication use) to: phase (Period 2) (0.25, 0.5, 1 and 2 mg
DM N = 5 to placebo/siponimod 2 mg tablets) (FMI tablets)
(Period 1/Period 2 extension)
N = 4 to siponimod 0.5 mg/2 mg
(Period 1/Period 2 extension)
N = 4 to siponimod 2 mg/2 mg
(Period 1/Period 2 extension)
N = 4 to siponimod 10 mg/2 mg
(Period 1/Period 2 extension)
a There were no specific inclusion/exclusion criteria related to subject CYP2C9 genotype.

Anonymized Page 117


Novartis Confidential Page 118
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 5-2 In vitro, ex vivo, and in silico studies using human biomaterials
Study Study objective Matrix Siponimod
concentration
In vitro and ex vivo plasma protein binding and stability
[DMPK R1300334] To determine the ex vivo binding of [14C]siponimod to Plasma samples from 24 subjects with single 0.25 mg oral
plasma proteins in selected human plasma samples impaired hepatic function (mild, moderate, dose of siponimod
from the clinical [Study A2122] severe) and from 16 healthy subjects
[DMPK To determine the free plasma circulating fraction of Human plasma samples from healthy and single 0.25 mg oral
RCBAF312A2129-01] siponimod (and M3 and M5) to assess whether protein renal impaired subjects dose of siponimod
binding was affected by renal insufficiency in selected
human plasma samples from the clinical [Study
A2129]
[DMPK R1400021] To assess the in vitro plasma protein binding and Human plasma 100 ng/mL
stability at 37°C of M3 and M5
Metabolism by CYP450 enzymes
[DMPK R0500432] To investigate the human CYP enzymes involved in Human liver microsomes, panel of 21 up to 300 μM
the oxidative metabolism of siponimod and assess the recombinant human CYPs [14C]siponimod
risk of potential drug-drug interactions by co
medications including in vitro metabolism of
[14C]siponimod and enzyme kinetics parameters (Km
and Vmax)
CYP450 induction and inhibition
[DMPK R0500496] To assess the potency of siponimod to induce Reporter gene cells (DPX2) up to 100 μM
CYP3A4 via the human pregnane X receptor (PXR)
[DMPK R0500497] To assess the potential of siponimod to inhibit CYP Human liver microsomes, 9 CYP-isoenzyme up to 200 μM
enzyme activity selective substrate probes
[DMPK R1200710] To examine siponimod for its potential to induce CYP Primary human hepatocytes up to 50 μM
enzyme 1A2, 2B6, 2C9, and 3A4 messenger
ribonucleic acid (mRNA) and activities

Anonymized Page 118


Novartis Confidential Page 119
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study Study objective Matrix Siponimod


concentration
[DMPK R1300188] To determine the potential of siponimod as reversible Human liver microsomes up to 100 μM
in vitro inhibitor of CYP2B6 and as time-dependent
inhibitor of CYP enzymes 1A2, 2C9, 2D6 and 3A4/5
as an addition to the CYP inhibition results of
(DMPK R0500497).
[DMPK R1300932] To determine the potential of M3 to function as an in Human liver microsomes up to 100 μM (M3)
vitro inhibitor of CYP enzymes 1A2, 2A6, 2B6, 2C8,
2C9, 2C19, 2D6, 2E1, and 3A4/5.
[DMPK R1300933] To examine the potential of M3 to induce CYP Primary human hepatocytes up to 100 μM (M3)
enzyme 1A2, 2B6, 2C9 and 3A4 mRNA and activities
[DMPK R1500795] To examine the potential for M17 to induce CYP Primary human hepatocytes up to 10 μM (M17)
enzyme 1A2, 2B6, 2C9, 3A4 mRNA and activities
[DMPK R1500796] To determine in vitro the potential of M17 to function Human liver microsomes 12.5 μM (M17)
as inhibitor of CYP-mediated reactions
Drug interaction studies
[DMPK 1600759-01] To predict the DDI potential of siponimod as a N/A (performed using simulator software N/A
substrate in the presence of typical CYP2C9/3A SimCYP Version 16, based on in vitro and in
inhibitors and inducers for 6 different CYP2C9 vivo siponimod data)
genotypes
[DMPK 1701078] To characterize the inhibition/induction properties of Primary human hepatocytes up to 10 μM
the strong CYP3A inhibitor itraconazole (itraconazole)
In vitro cellular uptake
[DMPK R0800531] To study the transport potential of siponimod into the Caco 2 cell monolayers up to 50 μM
intestinal barrier
[DMPK R1300921] To study the uptake mechanism of siponimod into the Caco 2 cell monolayers up to 50 μM
intestinal barrier

Anonymized Page 119


Novartis Confidential Page 120
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study Study objective Matrix Siponimod


concentration
Transporter inhibition: efflux transporters
[DMPK R1200722] To assess the potential of siponimod to inhibit human Recombinant MDCKII (Madin-Darby canine up to 24.7 μM
ATP-binding cassette (ABC) transporter activity kidney II) and LLC-PK1 (porcine kidney) cells
overexpressing human breast cancer resistant
protein (MXR, BCRP, ABCG2) and human
P-glycoprotein (MDR1, P-gp, ABCB1)
[DMPK R1300847] To assess the potential of siponimod to inhibit ABC Recombinant Sf9 (inside-out) vesicles up to 40 μM
transporter activity overexpressing the human bile salt export
pump (BSEP, ABCB11) protein
[DMPK R1300849] To assess the potential of siponimod to inhibit human Recombinant Sf9 (inside-out) vesicles up to 52.6 μM
ABC transporter activity overexpressing human multidrug-resistance
protein 2 (MRP2, ABCC2)
[DMPK R1300852] To assess the potential of M3 to inhibit human ABC Recombinant Sf9 (inside-out) vesicles up to 400 μM
transporter activity overexpressing the human bile salt export
pump (BSEP, ABCB11) protein
[DMPK R1300853] To assess the potential of M3 to inhibit human ABC Recombinant MDCKII and LLC-PK1 cells up to 100 μM
transporter activity overexpressing human breast cancer resistant
protein (MXR, BCRP, ABCG2) and human
P-glycoprotein (MDR1, P-gp, ABCB1)
[DMPK R1500825] To assess the potential of M17 to inhibit human ABC Mammalian cell based (inside-out) vesicles up to 1.35 μM
transporter activity overexpressing the human P-glycoprotein
(MDR1, P-gp, ABCB1) protein and
recombinant Sf9 (inside-out) vesicles
overexpressing human breast cancer resistant
protein (MXR, BCRP, ABCG2)
[DMPK R1500826] To assess the potential of M17 to inhibit human ABC Recombinant Sf9 (inside-out) vesicles up to 2.5 to 4 μM
transporter activity overexpressing the human bile salt export
pump (BSEP, ABCB11) protein and
recombinant Sf9 (inside-out) vesicles
overexpressing human multidrug-resistance
protein 2 (MRP2, ABCC2)

Anonymized Page 120


Novartis Confidential Page 121
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study Study objective Matrix Siponimod


concentration
Transporter inhibition: uptake transporters
[DMPK R1200723] To assess the potential of siponimod to inhibit human Recombinant HEK293 (human embryonic up to 24.3 μM
organic anion transporting polypeptide 1B1 kidney) cells overexpressing OATP1B1 and
(OATP1B1) and human organic anion transporting OATP1B3
polypeptide 1B3 (OATP1B3) transporter activity
[DMPK R1200724] To assess the potential of siponimod to inhibit human Recombinant HEK293 cells overexpressing up to 24.3 μM
organic anion transporter 1 (OAT1) and organic anion OAT1 and OAT3
transporter 3 (OAT3) activity
[DMPK R1200725] To assess the potential of siponimod to inhibit human Recombinant HEK293 cells overexpressing up to 24.3 μM
organic cation transporter 1 (OCT1) and human OCT1 and OCT2
organic cation transporter 2 (OCT2) transporter
activity
[DMPK R1300856] To assess the potential of M3 to inhibit OAT1 and Recombinant HEK293 cells overexpressing up to 100 μM
OAT3 activity OAT1 and OAT3
[DMPK R1300855] To assess the potential of M3 to inhibit human OCT1 Recombinant HEK293 cells overexpressing up to 100 μM
and human OCT2 transporter activity OCT1 and OCT2
[DMPK R1300857] To assess the potential of M3 to inhibit OATP1B1 and Recombinant HEK293 cells overexpressing up to 400 μM
OATP1B3 transporter activity OATP1B1 and OATP1B3
[DMPK R1500827] To assess the potential of M17 to inhibit OATP1B1 Recombinant HEK293 cells overexpressing up to 4 μM
and OATP1B3 transporter activity OATP1B1 or OATP1B3
[DMPK R1500828] To assess the potential of M17 to inhibit OCT1 and Recombinant HEK293 cells overexpressing up to 4 μM
OCT2 transporter activity OCT1 and OCT2
[DMPK R1500829] To assess the potential of M17 to inhibit OAT1 and Recombinant HEK293 cells overexpressing up to 4 μM
OAT3 activity OAT1 and OAT3

Anonymized Page 121


Novartis Confidential Page 122
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Study Study objective Matrix Siponimod


concentration
Transporter inhibition: SLC transporters
[DMPK R1300848] To assess the potential of siponimod to inhibit the Recombinant HEK293 cells overexpressing up to 40 μM
transport activity of the human multidrug and toxin MATE1 and MATE2K
extrusion proteins 1 (MATE1, SLC47A1) and 2K
(MATE2K, SLC47A2)
[DMPK R1300854] To assess the potential of M3 to inhibit the transport Recombinant HEK293 cells overexpressing up to 100 μM
activity of the human multidrug and toxin extrusion MATE1 and MATE2K
proteins 1 (MATE1, SLC47A1) and 2K (MATE2K,
SLC47A2)
[DMPK R1500830] To assess the potential of M17 to inhibit the transport Recombinant HEK293 cells overexpressing up to 4 μM nominal
activity of the human multidrug and toxin extrusion MATE1 and MATE2K (2.3 μM measured)
proteins 1 (MATE1, SLC47A1) and 2K (MATE2K,
SLC47A2)
N/A = not applicable.

Anonymized Page 122


Novartis Confidential Page 123
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 5-3 Summary of pooled analyses included in the Summary of Clinical Pharmacology
Analysis/Studies Pooled Sample Size Description
Dose proportionality after single dose
[Studies A2101 and A2105] N = 117 healthy subjects Power model
Phase 1/2 Population PK
[Studies A2101, A2105, A2107, A1101 and A2201] N = 190 healthy subjects Nonlinear mixed effects modeling
N = 216 RRMS patients Subjects on active treatment only
Phase 3 Population PK
[Study A2304] N = 1045 SPMS patients Nonlinear mixed effects modeling
Subjects on active treatment only
Steady-state PK in healthy subjects and MS patients
[Studies A2105, A2118, A2201 and A2304] N = 126 healthy subjects Subjects on active treatment and with steady state PK
N = 217 RRMS patients concentration data. considered.
N = 1043 SPMS patients A2105 (Day 28, 0 hour)
A2118 (Day10 and Day 18, 0 hour)
A2201 (Month 1, Month 3)
A2304 (Day 28, Month 12)
Exposure-lymphocyte relationship
[Studies A2101, A2105, A2107, A1101, A2110, A2118, N = 688 healthy subjects Nonlinear mixed effects modeling
A2201 and A2304] N = 287 RRMS patients Subjects on active treatment and on placebo
N = 1616 SPMS patients
Categorical analysis of lymphocyte count
Healthy subjects, SD: [Studies A1101, A2101, A2104, Healthy subjects, SD: N = 365 Categories of interest for ALC level:
A2108, A2111, A2119, A2124, A2126 and A2128] Healthy subjects, MD: N = 872 1) Within Normal Range (1.2 × 109/L to 3.5 × 109/L)
Healthy subjects, MD: [Studies A2102, A2105, A2107, PM/DM patients, MD: N = 49 2) ≥ 0.6 × 109/L and < 1.2 × 109/L
A2110, A2116, A2118, A2121, A2125, A2128 and A2130]
3) ≥ 0.2 × 109/L and < 0.6 × 109/L
PM/DM patients, MD: [Studies A2202, X2205 and X2206]
4) < 0.2 × 109/L
Categorical analysis of heart rate

Anonymized Page 123


Novartis Confidential Page 124
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Analysis/Studies Pooled Sample Size Description


12-lead ECG: Healthy subjects, SD: N = 365 Categories of interest for HR events:
Healthy subjects, SD: [Studies A1101, A2101, A2104, Healthy subjects, MD: N = 872 1) HR change from Baseline > 25% and ≤ 35%
A2108, A2111, A2119, A2124, A2126 and A2128] PM/DM patients, MD: N = 49 2) HR change from Baseline > 35% and ≤ 50%
Healthy subjects, MD: [Studies A2102, A2105, A2107, 3) HR change from Baseline > 50%
A2110, A2116, A2118, A2121, A2125, A2128 and A2130]
4) New HR value < 30 bpm
PM/DM patients, MD:[Studies X2205 and X2206]
5) New HR value ≥ 30 and < 40 bpm
Holter ECG: Healthy subjects, SD: N = 124
6) New HR value ≥ 40 and < 45 bpm
Healthy subjects, SD: (Studies A1101, A2119 and A2128) Healthy subjects, MD: N = 558
7) New HR value ≥ 45 and < 50 bpm
Healthy subjects, MD: (Studies A2102, A2105, A2107,
8) HR decrease > 25% from Baseline to a HR < 50 bpm
A2116, A2118 and A2128)
9) HR increase > 25% from Baseline to a HR > 100 bpm
PM/DM patients, MD: Not applicable.
Categorical analysis of bradyarrhythmia
12-lead ECG: Healthy subjects, SD: N = 278 Categories of interest for bradyarrhythmia events:
Healthy subjects, SD: (Studies A1101, A2101, A2104, Healthy subjects, MD: N = 652 1) First degree AV block (PR > 200 msec)
A2108, A2111, A2119, A2124, A2126 and A2128) PM/DM patients, MD: N = 31 2) Second degree AV block, Mobitz I (Wenckebach)
Healthy subjects, MD: (Studies A2102, A2105, A2107, 3) Second degree AV block, 2:1 conduction
A2110, A2116, A2118, A2121, A2125, A2128 and A2130)
4) Second degree AV block, 3:2 conduction (Holter
PM/DM patients, MD: (Studies A2202, X2205 and X2206) analysis only)
Holter ECG: Healthy subjects, SD: N = 255 5) Sinus pause (RR > 2 sec)
Healthy subjects, SD: (Studies A1101, A2101, A2119, Healthy subjects, MD: N = 696 6) Sinus pause (RR > 3 sec)
A2126 and A2128)
Healthy subjects, MD: (Studies A2102, A2105, A2107,
A2110, A2116, A2118, and A2128)
PM/DM patients, MD: Not applicable.

Anonymized Page 124


Novartis Confidential Page 125
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Analysis/Studies Pooled Sample Size Description


Telemetry: Healthy subjects, SD: N = 210
Healthy subjects, SD: (Studies A1101, A2104, A2108, Healthy subjects, MD: N = 324
A2111, A2119, and A2124) PM/DM patients, MD: N = 49

Healthy subjects, MD: [Studies A2102, A2105, A2110,


and A2116]
PM/DM patients, MD: [Studies A2202, X2205 and X2206]
Categorical analysis of QTcF
12-lead ECG: Healthy subjects, SD: N = 365 Categories of interest for QTcF events:
Healthy subjects, SD: [Studies A1101, A2101, A2104, Healthy subjects, MD: N = 872 1) QTcF change from Baseline > 30 msec and ≤ 60
A2108, A2111, A2119, A2124, A2126 and A2128] PM/DM patients, MD: N = 31 msec
Healthy subjects, MD: [Studies A2102, A2105, A2107, 2) QTcF change from Baseline > 60 msec
A2110, A2116, A2118, A2121, A2125, A2128 and A2130] 3) New QTcF value > 450 msec and ≤ 480 msec
PM/DM patients, MD: (Studies X2205 and X2206) 4) New QTcF value > 480 msec and ≤ 500 msec
Holter ECG (hourly data): Healthy subjects, MD: N = 436 5) New QTcF value > 500 msec
Healthy subjects, SD: Not applicable 6) New QTcF value > 480 msec combined with QTcF
Healthy subjects, MD: [Studies A2107, A2116 and change from Baseline > 60 msec
A2118]
PM/DM patients, MD: Not applicable
Categorical analysis of vital signs
Healthy subjects, SD: (Studies A1101, A2101, A2104, SBP: Categories of interest for SBP:
A2108, A2111, A2119, A2124, A2126 and A2128) Healthy subjects, SD: N = 365 1) SBP change from Baseline > 5% and ≤ 10%
Healthy subjects, MD: (Studies A2102, A2105, A2107, Healthy subjects, MD: N = 872 2) SBP change from Baseline > 10% and ≤ 15%
A2110, A2116, A2118, A2121, A2125, A2128 and A2130)
PM/DM patients, MD: N = 49 3) SBP change from Baseline > 15%
PM/DM patients, MD: (Studies A2202, X2205 and X2206)
DBP: 4) New SBP > 140 and ≤ 150 mmHg
Healthy subjects, SD: N = 365 5) New SBP > 150 and ≤ 160 mmHg
Healthy subjects, MD: N = 872 6) New SBP > 160 mmHg
PM/DM patients, MD: N = 49

Anonymized Page 125


Novartis Confidential Page 126
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Analysis/Studies Pooled Sample Size Description


7) SBP increase > 15% from Baseline to a SBP > 140
mmHg
8) SBP decrease > 15% from Baseline to a SBP < 90
mmHg
Categories of interest for DBP:
1) DBP change from Baseline > 5% and ≤ 10%
2) DBP change from Baseline > 10% and ≤ 15%
3) DBP change from Baseline > 15%
4) New DBP > 90 and ≤ 100 mmHg
5) New DBP > 100 and ≤ 110 mmHg
6) New DBP > 110 mmHg
7) DBP increase > 15% from Baseline to a DBP > 90
mmHg
8) DBP decrease > 15% from Baseline to a DBP < 50
mmHg
Categorical analysis of liver function test parameters
Healthy subjects, SD: [Studies A1101, A2101, A2104, Healthy subjects, SD: N = 365 Categories of interest for LFT parameters (ALT, AST,
A2108, A2111, A2119, A2124, A2126 and A2128] Healthy subjects, MD: N = 872 GGT, alkaline phosphatase and bilirubin):
Healthy subjects, MD: [Studies A2102, A2105, A2107, PM/DM patients, MD: N = 49 1) Values ≥ 1.5-fold and < 3-fold increase above upper
A2110, A2116, A2118, A2121, A2125, A2128 and A2130] limit of normal (ULN)
PM/DM patients, MD: [Studies A2202, X2205 and X2206] 2) Values ≥ 3-fold and < 5-fold increase above ULN
3) Values ≥ 5-fold and < 8-fold increase above ULN
4) Values ≥ 8-fold increase above ULN
5) Combined category of ALT > 3-fold and
bilirubin > 2-fold increase above ULN
6) Combined category of AST > 3-fold and
bilirubin > 2-fold increase above ULN
ULNs: ALT and AST 40 U/L, GGT 50 U/L, AP 200 U/L
and bilirubin 20 µmol/L.

Anonymized Page 126


Novartis Confidential Page 127
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Analysis/Studies Pooled Sample Size Description


Categorical analysis of adverse events
Healthy subjects, SD: [Studies A1101, A2101, A2104, AEs by PT and Treatment: Analyses of AEs:
A2108, A2111, A2119, A2122, A2124, A2126, A2128 and Healthy subjects, SD: N = 389 1) By PT and Treatment
A2129]
Healthy subjects, MD: N = 872 2) By SOC and Treatment
Healthy subjects, MD: [Studies A2102, A2105, A2107,
PM/DM patients, MD: N = 49 3) By Category
A2110, A2116, A2118, A2121, A2125, A2128 and A2130]
AEs by SOC and Treatment:
PM/DM patients, MD: [Studies A2202, X2205 and X2206]
Healthy subjects, SD: N = 389
Healthy subjects, MD: N = 872
PM/DM patients, MD: N = 49
AEs by Category:
Healthy subjects, SD: N = 389
Healthy subjects, MD: N = 872
PM/DM patients, MD: N = 49
MD = multiple dose; SD = single dose.

Anonymized Page 127


Novartis Confidential Page 128
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 5-4 Summary of in vitro and ex vivo plasma protein binding of siponimod
and metabolites
Compound Group fu ± SD (%)
(DMPK R1300334)
(Study A2122) Siponimod Healthy subjects 0.0134 ± 0.00264 (N = 14)a
Siponimod Mild impairment 0.0150 ± 0.00258 (N = 8)
Siponimod Moderate impairment 0.0155 ± 0.00183 (N = 8)
Siponimod Severe impairment 0.0186 ± 0.00606 (N = 8)
In vitro Siponimod In vitro human plasma pool 1 0.0136 ± 0.00135 (N = 1)b
Siponimod In vitro human plasma pool 2 0.0124 ± 0.000339 (N = 1)b
(DMPK RCBAF312A2129-01)
(Study A2129) (N = Siponimod Healthy subjects 0.0263 ± 0.00820
8)
Siponimod Severe impairment 0.0292 ± 0.0160
In vitrod Siponimod Blank human plasma pool 1 0.0206c
Siponimod Blank human plasma pool 2 0.0269c
d
[DMPK R1400021]
In vitro M3 In vitro human plasma 0.572 ± 0.00292
M5 In vitro human plasma 0.401 ± 0.0689
a 16 samples were received, but 2 samples were not analyzed (tube broke during centrifugation)
b One plasma pool from 2 different individuals, SD represents 3 runs
c One plasma pool from 2 different individuals
d fu results might be slightly underestimated as equilibrium was not completely reached after
6 hours incubation
n = number of subjects/samples
Source: [DMPK R1300334-Table 1-1], [DMPK RCBAF312A2129-01-Table 2-1] and
[DMPK R1400021-Table 1-1]

Anonymized Page 128


Novartis Confidential Page 129
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Table 5-5 Siponimod, M3 and M17c acting as a perpetrator drug


Affected Relevant perpetrator DDI assessment potential clinical implicationsd
protein constants
Siponimoda M3b
DDI risk concerning enzymes in the liver and in the intestine ([DMPK R1300188], [DMPK R1300932],
[DMPK R1300857])
The reversible and irreversible in vitro inhibition
CYP2B6 na IC50: 94 µM potential of siponimod or M3 towards CYP2B6,
CYP2C9 and CYP3A4/5 is unlikely to translate into
IC50: 230 µM clinical significance as the unbound steady-state
IC50: 80 µM concentrations (Ch,in,u and Cmax,ss,u) at
IC50,u: 1.3 µM therapeutic dose are much lower than the
experimentally determined unbound inhibition
CYP2C9 KI: 54 µM constants (> 200-fold). However, siponimod is
KI,u: 0.26 µM projected to slightly inhibit CYP3A4/5 activity in the
na intestine (Cgut,ent/IC50,u ≈ 0.35). The predicted net
Kinact: 0.006 exposure change following CYP3A4/5 inhibition
min-1 assuming fugut=1 is up to 14% for the specific probe
substrate midazolam which is unlikely to translate
into clinical significance though, given the high
IC50: 100 µM plasma protein binding of siponimod, indicating that
CYP3A4/5 na fugut should be less than 1e. In vitro, no CYP450
IC50,u: 0.6 µM
induction potential by siponimod, M3 or M17 was
observed in human hepatocytesa,b,c.
DDI risk concerning solute-carrier transporters at the liver inlet ([DMPK R1200723],
[DMPK R1300855], [DMPK R1300857])
OATP1B1 Ki: 1.3 µM Ki: 3.7 µM The in vitro inhibition towards OATP1B1, OATP1B3
OATP1B3 Ki: 2.4 µM Ki: 4.1 µM and OCT1 is unlikely to translate into clinical
significance at therapeutic doses as the
OCT1 Maximal Maximal experimentally determined Ki values are more than
inhibition: 21% inhibition: 32% 1000-fold higher than the siponimod Ch,in,u or M3
at 18.3 µM at 100 µM Cmax,u.
DDI risk concerning ubiquitous efflux transporters ([DMPK R1300852])
BSEP na Maximal The weak in vitro inhibition potential of M3 towards
inhibition: BSEP is unlikely to result in any clinical meaningful
38.5% at 400 effect as Cmax,ss,u is significantly lower than the
µM experimentally determined Ki value.
Cgut,ent = concentration in enterocytes; IC50,u = unbound IC50; Ki = inhibition constant; Kinact =
maximal rate of enzyme inactivation; KI,u = unbound KI; na = not applicable; Ch,in,u = hepatic
unbound inlet concentration; Cmax,ss,u = maximum unbound concentration at steady state.
a Additional processes tested for siponimod but not affected ([DMPK R0500496], [DMPK R0500497],
[DMPK R1200722], [DMPK R1200724], [DMPK R1200725], [DMPK R1300188], [DMPK R1200710],
[DMPK R1300847], [DMPK R1300848], [DMPK R1300849]): reversible inhibition of CYP1A2,
CYP2A6, CYP2B6, CYP2C8, CYP2C19, CYP2D6, CYP2E1 (tested up to 100 µM siponimod) and
P-gp, BCRP, BSEP, MRP2, OAT1, OAT3, OCT1, OCT2, MATE1 and MATE2K (tested up to 24 µM
siponimod); time-dependent inhibition of CYP1A2, CYP2D6 and CYP3A4/54 (tested up to 100 µM
siponimod); induction of CYP1A2, CYP2B6, CYP2C9 and CYP3A4/5 (tested up to 10 µM siponimod)
b Additional processes tested for M3 but not affected ([DMPK R1300932], [DMPK R1300933],
[DMPK R1300852], [DMPK R1300853], [DMPK R1300854], [DMPK R1300855], [DMPK R1300856],
[DMPK R1300857]): Reversible inhibition of BCRP, P-gp, BSEP, OAT1, OAT3, OCT1, OCT2 (tested
up to 100 µM), CYP1A2, CYP2A6, CYP2C8, CYP2C19, CYP2D6, CYP2E1 and CYP3A4/5 (tested up

Anonymized Page 129


Novartis Confidential Page 130
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Affected Relevant perpetrator DDI assessment potential clinical implicationsd


protein constants
Siponimoda M3b
to 100 µM); irreversible inhibition of CYP1A2, CYP2C9, CYP2D6 and CYP3A4/54 (tested up to
100 µM); induction of CYP1A2, CYP2B6, CYP2C9 and CYP3A4/5 (tested up to 100 µM in
hepatocytes).
c Additional processes tested for M17 but not affected ([DMPK R1500796], [DMPK R1500795],
[DMPK R1500825], [DMPK R1500826], [DMPK R1500827], [DMPK R1500828], [DMPK R1500829],
[DMPK R1500830]): Reversible inhibition of BCRP, P-gp, BSEP, MRP2, OATP1B1, OATP1B3, OAT1,
OAT3, OCT1, OCT2, MATE1, MATE2-K (tested up to 4 M, 2.5 M for MRP2), CYP1A2, CYP2A6,
CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP2E1 and CYP3A4/5 (tested up to 12.5 µM);
irreversible inhibition of CYP1A2, CYP2C9, CYP2D6 and CYP3A4/54 (tested up to 10 µM); induction
of CYP1A2, CYP2B6, CYP2C9 and CYP3A4/5 (tested up to 10 µM in hepatocytes).
d DDI risk calculations (according to HA guidelines (FDA 2012, EMA 2012) for siponimod after multiple
oral doses of 2 mg qd were done using the measured unbound therapeutic steady-state blood
concentration (Cmax,ss,u) of 0.0006 M, an estimated unbound hepatic inlet blood concentration
(Ch,in,u) of 0.001 M (assumed plasma fu: 0.01, measured plasma fu: 0.0002 ([Table 2.6.5.6A-
DMPK R0400881-01 and DMPK R1300902-01]), an anticipated maximal concentration in the gut
lumen (Cgut) of about 15.5 M and a projected maximal enterocyte concentration (Cgut,ent) of 0.21
M. Cmax,ss,u (≈ Ch,in,u) for the metabolite LNL925 was assessed with about 0.000196 M
(assumed plasma fu: 0.01, measured plasma fu: 0.00572 ([DMPK R1400021]). Cmax,ss,u (≈ Ch,in,u)
for the metabolite LYS815 was assessed with about 0.00145 M (assumed plasma fu: 0.01, measured
plasma fu: 0.000455 ([DMPK R1500677-01]).
e Net effect assessment according to Fahmi et al 2008 (note: depending on the final therapeutic dose,
the ultimate therapeutic concentrations and the probe substrate of interest the denoted DDI projection
can change.

Table 5-6 Siponimod acting as a victim of drugs


Process Involved DDI assessment potential clinical implications
proteina
DDI risk concerning transporters and enzymes involved in absorption ([DMPK R0500432],
[DMPK R0800531], [DMPK R1300921])
Siponimod was not identified as efflux transporter
substrate (investigated transporters: P-gp, BCRP or
Intestinal absorption na
MRP2). Intestinal efflux is therefore not expected to impact
rate and/or extent of siponimod absorption.

Siponimod was in vitro identified as CYP2C9 substrate and


to a lesser extend as a CYP3A4/5 substratea. Due to its low
Intestinal first-pass CYP2C9 clearance, siponimod is unlikely to be subject to major
effect CYP3A4/5 intestinal metabolic first-pass via CYP2C9 and CYP3A in
humans and consequently no significant DDI following
intestinal enzyme inhibition is expected.
DDI risk concerning transporters involved in sinusoidal uptake in the liver ([DMPK R0700963])
Siponimod uptake into hepatocytes is solely governed by
a high passive diffusion process (investigated
Hepatic uptake na
transporters: OATPs, OCTs, OATs). Interactions on
hepatic uptake transporters are therefore irrelevant.
DDI risk concerning enzymes and transporters involved in (pre-)systemic hepatic metabolism and
canalicular secretion (DMPK R0500432, [DMPK R0900164-01])

Anonymized Page 130


Novartis Confidential Page 131
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Process Involved DDI assessment potential clinical implications


proteina
Siponimod can be categorized as BDDCS/ECCCS Class 2
compound in line with hepatic metabolism being the major
elimination mechanism (Camenisch 2016). Based on rat
data and in line with its BDDCS/ECCCS classification
Biliary secretion na biliary secretion is considered to be a minor elimination
pathway for siponimod (Camenisch 2016). Hence, no
direct DDI effects at the level of renal uptake carriers and
efflux pumps as well as canalicular efflux transporters are
expected.
Based on in vitro data, siponimod oxidative metabolism is
predominantly driven by CYP2C9 (about 79%) and to a
lesser extent by CYP3A4 (about 19%). CYP2C9 is
polymorphic and the CYP2C9 genotype influences the
fractional contributions of the 2 oxidative metabolism
pathways to overall elimination (the hepatic fm,CYP2C9 is
anticipated with 80.4% (CYP2C9*1*1) or 7.4%
(CYP2C9*3*3)). For the CYP2C9*1/*1 genotype, CYP3A4
plays a minor role in siponimod metabolism with a
contribution of 17.5% to the overall elimination of
siponimod, but its contribution is expected to be greater in
subjects with lower CYP2C9 activity associated with a
reduced total clearance (82.2% contribution in
CYP3A4/5, CYP2C9*3/*3 genotype). Considering the variable
Hepatic metabolism CYP3A4 and CYP2C9 contributions to the overall
CYP2C9
clearance of siponimod in the different CYP2C9
genotypes, it is expected that the CYP2C9 genotype
influences the effects of CYP3A4 and CYP2C9 inhibitors
and inducers.
Dynamic DDI predictions for *1*1 and *3*3 subjects using
the population-based PBPK modeling software SimCYP
projected a respective exposure increase of 2.15 to 2.52-
fold for siponimod in the presence of the moderate
CYP2C9 and CYP3A4 inhibitor fluconazoleb
([DMPK R1600759-01]). The effect of itraconazole, a
strong CYP3A4 inhibitor, is predicted to lead to a 17%
AUC increase in CYP2C9*1*1 carriers and to a 3.25-fold
exposure increase in CYP2C9*3*3 subjectsb.
DDI risk concerning transporters involved in renal elimination ([Study A2104])
Renal excretion of siponimod represents a minor
Active tubular
na elimination pathway in human. Hence, no DDI effects at
secretion
the level of renal transporters are expected.
DDI risk concerning enzymes and transporters involved in extra-hepatic/renal elimination
([DMPK R0500432], [DMPK R1300921], [DMPK R0800531])
In vitro, siponimod was found to be metabolized by
CYP1A1. As such, it cannot be excluded that exposure to
siponimod will be impacted by co-administered inducers of
Extrahepatic
CYP1A1 CYP1A1. However, in the non-induced state CYP1A1
metabolism
abundance is considered to be low and a significant
metabolism in human is considered rather unlikely
(Kim et al 2004).

Anonymized Page 131


Novartis Confidential Page 132
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Process Involved DDI assessment potential clinical implications


proteina
Not being an efflux transporter substrate (see absorption),
intestinal secretion is unlikely to be a relevant systemic
Intestinal secretion na
elimination pathway for siponimod. Consequently, no DDI
effects are anticipated.
DDI risk concerning extra-hepatic/renal distributional transporters and plasma proteins
Transporter-mediated extra-hepatic/renal distribution (e.g.,
brain penetration) is not anticipated for siponimod (see
Tissue distribution na
absorption/hepatic uptake). Consequently, no related DDI
effects are expected.
Siponimod has a low hepatic extraction ratio, is given
orally and is to a large extent eliminated hepatically. A
displacement effect, if at all, is likely to be compensated by
Plasma protein
na changes in clearance or volume of distribution
binding
(Benet and Hoener 2002). Consequently, the potential for
alterations of unbound siponimod exposure due to
displacement interactions by comedications is very low.
na = not applicable.
a Additional processes tested but without significant in vitro turnover or low estimated clearance
contribution ([DMPK R0500432]): CYP1A1, CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C18,
CYP2C19, CYP2D6, CYP2E1, CYP2J2, CYP3A5, CYP3A7, CYP4A11, CYP4F2, CYP4F3A,
CYP4F3B, CYP4F12, CYP19.
b Dynamic risk predictions using population-based PBPK modeling software SimCYP®
([DMPK R1600759-01])

Anonymized Page 132


Novartis Confidential Page 133
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

5.1 Literature references


[Benet L and Hoener B (2002)] Changes in plasma protein binding have little clinical
relevance. Clin Pharmacol Ther; 71:115-21.
[Boxenbaum H and Battle M (1995)] Effective half-life in clinical pharmacology. J Clin
Pharmacol; 35(8):763-6.
[Brinkmann V, Cyster JG, Hla T (2004)] FTY720: Sphingosine 1-phosphate receptor-1 in the
control of lymphocyte egress and endothelial barrier function. Am J Transplant; 4:1019-25.
[Bryan L, Kordula T, Spiegel S, et al (2008)] Regulation and functions of sphingosine kinases
in the brain. Biochim Biophys Acta; 1781:459-66.
[Camenisch G (2016)] Drug Disposition Classification Systems in Discovery and
Development: A Comparative Review of the BDDCS, ECCS and ECCCS Concepts. Pharm
Res; 33(11):2583-93.
[Denson LA, Bohan A, Held MA, et al (2002)] Organ-Specific Alterations in RAR: RXR
Abundance Regulate Rat Mrp2 (Abcc2) Expression in Obstructive Cholestasis.
Gastroenterology; 123(2):599–607.
EMA (2012) Guideline on the investigation of drug interactions (Internet) Available from:
http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2012/07/WC5
00129606.pdf (Accessed 25-Jul-2016)
[Fahmi OA, Boldt S, Kish M, et al (2008)] Prediction of drug-drug interactions from in vitro
induction data: application of the relative induction score approach using cryopreserved
human hepatocytes. Drug Metab Dispos; 36:1971-4.
FDA (2012) Guidance for Industry, Drug Interaction Studies - Study Design, Data Analysis,
Implications for Dosing, and Labeling Recommendations (Internet) Available from:
http://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/uc
m292362.pdf (Accessed 25-Jul-2016)
[Gaikwad T, Ghosh K, Shetty S (2014)] VKORC1 and CYP2C9 genotype distribution in
Asian countries. Thromb Res; 134(3):537-44.
[Kim JH, Sherman ME, Curriero FC, et al (2004)] Expression of cytochromes P450 1A1 and
1B1 in human lung from smokers, non-smokers, and ex-smokers. Toxicol Appl Pharmacol;
199(3):210-9.
[Kirchheiner J, Brockmoller J (2005)] Clinical consequences of cytochrome P450 2C9
polymorphisms. Clin Pharmacol Ther; 77(1):1-16.
[Lee CR, Goldstein JA, Pieper JA (2002)] Cytochrom P450 2C9 polymorphisms: A
comprehensive review of the in-vitro and human data. Pharmacogenetics; 12(3):251-63.
[Legangneux E, Gardin A, Johns D (2013)] Dose titration of BAF312 attenuates the initial
heart rate reducing effect in healthy subjects. Br J Clin Pharmacol; 75(3):831-41.
[Matloubian M, Lo CG, Cinamon G, et al (2004)] Lymphocyte egress from thymus and
peripheral lymphoid organs is dependent on S1P receptor 1. Nature; 427:355-60.

Anonymized Page 133


Novartis Confidential Page 134
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

[Nathwani RA, Pais S, Reynolds TB et al (2005)] Serum alanine aminotransferase in skeletal


muscle diseases. Hepatology; 41(2):380-2.
[Pellegrino R, Viegi G, Brusasco V, et al (2005)] Interpretative strategies for lung function
tests. Eur Respir J; 26(5):948-68.
[Sallusto F and Lanzavecchia A (2009)] Heterogeneity of CD4+ memory T cells: functional
modules for tailored immunity. Eur J Immunol; 39(8):2076-82.
[Scott SA, Khasawneh R, Peter I, et al (2010)] Combined CYP2C9, VKORC1 and CYP4F2
frequencies among racial and ethnic groups. Pharmacogenomics; 11(6):781-91.
[Subei AM and Cohen JA (2015)] Sphingosine 1-phosphate receptor modulators in multiple
sclerosis. CNS Drugs; 29(7):565-75.
[Wada T, Abe G, Kudou T, et al (2016)] Liver damage in patients with polymyositis and
dermatomyositis. Kitasoto Med J; 46:40-6.

5.2 Post-text tables and figures

Anonymized Page 134


Novartis Confidential Page 135
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Cmax [ng/mL/mg]

15

10

0.1 5 10 17.5 25 75
Dose (mg)
Regression

For Cmax: Intercept = 7.849 , Slope = -0.003


For AUC0-24h: Intercept = 129.644 , Slope = -0.073
/tf_pk_doseprop_sd.sas@@/main/5 04MAR18:10:11

Anonymized Page 135


Novartis Confidential Page 136
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

A UC0-24h [hr* ng/mL/mg]

200

150

100

50

0.1 5 10 17.5 25 75
Dose (mg)
Regression

For Cmax: Intercept = 7.849 , Slope = -0.003


For AUC0-24h: Intercept = 129.644 , Slope = -0.073
/tf_pk_doseprop_sd.sas@@/main/5 04MAR18:10:11

Anonymized Page 136


Novartis Confidential Page 137
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod

Anonymized Page 137


Novartis Confidential Page 138
1

CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod


BAF312

Table 11-1.1
Dose proportionality of BAF312 PK parameters of primary interest for
siponimod single oral doses(0.1, 0.3, 1, 2.5, 5, 10, 17.5, 20, 25, 75 mg)
PK analysis set

90% CI for
slope
Slope
PK parameter (unit) estimate Lower Upper
AUC0-24h [hr*ng/mL] 1.01 0.98 1.03
Cmax [ng/mL] 1.02 0.99 1.04

Power model: The slope parameter estimate and the confidence interval are obtained from the
linear model of the
log transformed PK parameter value adjusted for an intercept and the log-transformed dose
(covariate), as fixed effects.
Data considered - CBAF312A2101 (Doses: 0.1 mg, 0.3 mg, 1 mg, 2.5 mg, 5 mg, 10 mg, 17.5 mg,
25 mg, 75 mg)
Data considered - CBAF312A2105 (Day 1) (Doses: 0.1 mg, 0.3 mg, 2.5 mg, 10 mg, 20 mg)

/tf_pk_doseprop_sd.sas@@/main/5 04MAR18:10:11

Anonymized Page 138


Novartis Confidential Page 139
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod
1

BAF312

Table 12_5.1
Lymphocytes categorical outlier analysis - Single dose studies
Safety analysis set

Absolute lymphocyte levels (10E9/L) event


Categories
Within LYM >= LYM >=
normal 0.6 and < 0.2 and
range LLN < 0.6 LYM < 0.2
Treatment n (%) n (%) n (%) n (%)
BAF312 0.1mg (N=11) 10 (90.9) 1 (9.1) 0 0
BAF312 0.25mg (N=100) 95 (95.0) 5 (5.0) 0 0
BAF312 0.3mg (N=8) 7 (87.5) 0 1 (12.5) 0
BAF312 0.5mg (N=8) 1 (12.5) 7 (87.5) 0 0
BAF312 1mg (N=8) 8 (100.0) 0 0 0
BAF312 2.5mg (N=21) 3 (14.3) 8 (38.1) 10 (47.6) 0
BAF312 4mg (N=90) 38 (42.2) 29 (32.2) 22 (24.4) 0
BAF312 5mg (N=16) 0 4 (25.0) 12 (75.0) 0
BAF312 10mg (N=20) 4 (20.0) 2 (10.0) 12 (60.0) 2 (10.0)
BAF312 17.5mg (N=8) 0 1 (12.5) 7 (87.5) 0
BAF312 25mg (N=8) 0 0 8 (100.0) 0
BAF312 75mg (N=8) 0 0 8 (100.0) 0
BAF312 in combination (N=40) 37 (92.5) 3 (7.5) 0 0
Comparator alone (N=42) 31 (73.8) 0 0 0

Anonymized Page 139


Novartis Confidential Page 140
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod
2

Absolute lymphocyte levels (10E9/L) event


Categories
Within LYM >= LYM >=
normal 0.6 and < 0.2 and
range LLN < 0.6 LYM < 0.2
Treatment n (%) n (%) n (%) n (%)
BAF312 0.25mg i.v. (N=15) 12 (80.0) 3 (20.0) 0 0
BAF312 1mg i.v. (N=17) 5 (29.4) 12 (70.6) 0 0
Placebo (N=41) 37 (90.2) 4 (9.8) 0 0
All BAF (N=324) 169 (52.2) 72 (22.2) 80 (24.7) 2 (0.6)
Any treatment (N=365) 206 (56.4) 76 (20.8) 80 (21.9) 2 (0.5)

Studies included: A1101, A2101, A2104, A2108, A2111, A2119, A2124,A2126,A2128 (Part1 only).
Only events occurring at or after first drug intake were considered.
A subject minimum result was considered and was counted only once in the worst category.
N- Total number of subjects who got administered with at least one dose of respective treatment.
n- number of unique subjects meeting respective category criteria.
Lab parameter ranges (lower limit-upper limit): Lymphocytes ranges (10E9/L)(1.2 - 3.5)

/report/pgm_saf/t_lrs.sas@@/main/8 27FEB18:20:13

Anonymized Page 140


Novartis Confidential Page 141
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod
1

BAF312

Table 12_5.2
Lymphocytes categorical outlier analysis - Multiple dose studies
Safety analysis set

Absolute lymphocyte levels (10E9/L) event


Categories
Within LYM >= LYM >=
normal 0.6 and < 0.2 and
range LLN < 0.6 LYM < 0.2
Treatment n (%) n (%) n (%) n (%)
BAF312 0.3mg (N=15) 7 (46.7) 7 (46.7) 1 (6.7) 0
BAF312 0.5mg (N=94) 49 (52.1) 44 (46.8) 1 (1.1) 0
BAF312 1mg (N=101) 23 (22.8) 63 (62.4) 15 (14.9) 0
BAF312 2mg (N=66) 8 (12.1) 38 (57.6) 20 (30.3) 0
BAF312 2.5mg (N=16) 0 5 (31.3) 11 (68.8) 0
BAF312 4mg (N=64) 3 (4.7) 36 (56.3) 25 (39.1) 0
BAF312 10mg (N=32) 0 7 (21.9) 25 (78.1) 0
BAF312 20mg (N=18) 0 0 17 (94.4) 1 (5.6)
BAF312 0.5mg preceded by up-titration(N=12) 10 (83.3) 2 (16.7) 0 0
BAF312 2mg preceded by up-titration (N=236) 30 (12.7) 114 (48.3) 59 (25.0) 0
BAF312 10mg preceded by up-titration (N=120) 7 (5.8) 50 (41.7) 63 (52.5) 0
BAF312 in combination (N=175) 68 (38.9) 64 (36.6) 42 (24.0) 1 (0.6)
Comparator alone (N=188) 163 (86.7) 2 (1.1) 0 0
Placebo (N=393) 374 (95.2) 13 (3.3) 2 (0.5) 1 (0.3)

Anonymized Page 141


Novartis Confidential Page 142
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod
2

Absolute lymphocyte levels (10E9/L) event


Categories
Within LYM >= LYM >=
normal 0.6 and < 0.2 and
range LLN < 0.6 LYM < 0.2
Treatment n (%) n (%) n (%) n (%)
All BAF (N=544) 63 (11.6) 245 (45.0) 234 (43.0) 2 (0.4)
Any treatment (N=872) 376 (43.1) 257 (29.5) 235 (26.9) 3 (0.3)

Studies included: A2102, A2105, A2107,A2110, A2116, A2118, A2121, A2125, A2128 (Part2), A2130.
Only events occurring at or after first drug intake were considered.
A subject minimum result was considered and was counted only once in the worst category.
N- Total number of subjects who got administered with at least one dose of respective treatment.
n- number of unique subjects meeting respective category criteria.
Lab parameter ranges (lower limit-upper limit): Lymphocytes ranges (10E9/L)(1.2 - 3.5)

/report/pgm_saf/t_lrs.sas@@/main/8 27FEB18:20:13

Anonymized Page 142


Novartis Confidential Page 143
CTD 2.7.2 Summary of Clinical Pharmacology BAF312/siponimod
1

BAF312

Table 12_5.3
Lymphocytes categorical outlier analysis – PM/DM studies
Safety analysis set

Absolute lymphocyte levels (10E9/L)


event Categories
Within LYM >= LYM >=
normal 0.6 and < 0.2 and
range LLN < 0.6 LYM < 0.2
Treatment n (%) n (%) n (%) n (%)
BAF312 10mg (N=16) 0 4 (25.0) 6 (37.5) 6 (37.5)
BAF312 0.5mg preceded by up-titration(N=4) 0 2 (50.0) 2 (50.0) 0
BAF312 2mg preceded by up-titration (N=23) 0 2 (8.7) 15 (65.2) 6 (26.1)
BAF312 10mg preceded by up-titration (N=6) 0 0 4 (66.7) 2 (33.3)
Placebo (N=20) 12 (60.0) 8 (40.0) 0 0
All BAF (N=43) 0 6 (14.0) 24 (55.8) 13 (30.2)
Any treatment (N=49) 4 (8.2) 8 (16.3) 24 (49.0) 13 (26.5)

Studies included: A2202,X2205,X2206.


Only events occurring at or after first drug intake were considered.
A subject minimum result was considered and was counted only once in the worst category.
N- Total number of subjects who got administered with at least one dose of respective treatment.
n- number of unique subjects meeting respective category criteria.
Lab parameter ranges (lower limit-upper limit): Lymphocytes ranges (10E9/L)(1.2 - 3.5)

/report/pgm_saf/t_lrs1.sas@@/main/8 27FEB18:20:14

Anonymized Page 143

You might also like