Scribd Logo
Upload

Welcome to Scribd!

  • 63 views

    In Vitro Population Bioequivalence

    Uploaded by

    Metta Bodhiloka Susilo
    This document discusses parameters for assessing population bioequivalence (PBE) of particle size distributions (PSD) for injectable products. It provides examples of products that have been evaluated using D50, SPAN, Z-average, and PDI as primary PBE parameters. While Z-average and PDI appear to be increasingly used, the document argues they may be overly sensitive as PBE criteria compared to D50 and SPAN. It proposes modifying the acceptance criteria for PDI to address this issue. The document also questions the rationale for switching from D50/SPAN to Z-average/PDI and whether in vitro PBE testing is truly needed when in vivo BE is demonstrated.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    In Vitro Population Bioequivalence

    Uploaded by

    Metta Bodhiloka Susilo
    63 views14 pages
    This document discusses parameters for assessing population bioequivalence (PBE) of particle size distributions (PSD) for injectable products. It provides examples of products that have been evaluated using D50, SPAN, Z-average, and PDI as primary PBE parameters. While Z-average and PDI appear to be increasingly used, the document argues they may be overly sensitive as PBE criteria compared to D50 and SPAN. It proposes modifying the acceptance criteria for PDI to address this issue. The document also questions the rationale for switching from D50/SPAN to Z-average/PDI and whether in vitro PBE testing is truly needed when in vivo BE is demonstrated.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses parameters for assessing population bioequivalence (PBE) of particle size distributions (PSD) for injectable products. It provides examples of products that have been evaluated using D50, SPAN, Z-average, and PDI as primary PBE parameters. While Z-average and PDI appear to be increasingly used, the document argues they may be overly sensitive as PBE criteria compared to D50 and SPAN. It proposes modifying the acceptance criteria for PDI to address this issue. The document also questions the rationale for switching from D50/SPAN to Z-average/PDI and whether in vitro PBE testing is truly needed when in vivo BE is demonstrated.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    63 views14 pages

    In Vitro Population Bioequivalence

    Uploaded by

    Metta Bodhiloka Susilo
    This document discusses parameters for assessing population bioequivalence (PBE) of particle size distributions (PSD) for injectable products. It provides examples of products that have been evaluated using D50, SPAN, Z-average, and PDI as primary PBE parameters. While Z-average and PDI appear to be increasingly used, the document argues they may be overly sensitive as PBE criteria compared to D50 and SPAN. It proposes modifying the acceptance criteria for PDI to address this issue. The document also questions the rationale for switching from D50/SPAN to Z-average/PDI and whether in vitro PBE testing is truly needed when in vivo BE is demonstrated.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 14

    In-vitro Population Bioequivalence

    (PBE) Parameters for Particle Size


    Distribution (PSD)
    Mark Liu
    Mylan Global Pharmacokinetics, Morgantown, WV
    The views expressed in this discussion are based on broad
    industry commentary and should not be interpreted as unique or
    specific to Mylan N.V. or its subsidiaries unless specified.

    2
    Parameters required for Population BE (PBE)

    • The conventional population bioequivalence (PBE) analysis were based on


    D10, D50, D90 & Span.

    • D50 & Span as primary BE parameters, and D10 & D90 as supportive
    parameters.

    • Currently, Z-average & PDI appear as surrogates of D50 and Span as


    primary parameters for some products, either as option or requirement.

    • Additionally, the earth mover’s distance (EMD) method was introduced when
    the particle size distribution may not be monomodal (more than 1 peak).

    3
    Examples on PSG Iron Injectable products

    Product Date Guidance, in vitro criteria

    Iron Sucrose Nov,13 D50 and SPAN [i.e. (D90-D10)/D50] or PDI using PBE
    Sodium Ferric Jun,13 D50 and SPAN [i.e. (D90-D10)/D50] or PDI using PBE
    Gluconate
    Ferumoxytol Dec,12 D50 and SPAN [i.e. (D90-D10)/D50] or PDI using PBE
    Ferric Apr,16 Z-average size and polydispersity index using PBE
    Carboxymaltose
    Iron Dextran Oct,16 Harmonic intensity-weighted average particle diameter
    and PDI using PBE

    4
    Additional examples from PSG
    Product Dosage form Date Guidance, in vitro criteria
    Paclitaxel Suspension/Injectable Sept, 12 D50 and SPAN or PDI
    Azacitidine Powder; IV (Infusion), Nov,19 D50 and SPAN
    Subcutaneous
    Fluticasone propionate nasal Feb,19 D50 and SPAN
    Ciprofloxacin injectable Jul, 18 D50 and SPAN
    Triamcinolone acetonide injectable Jul, 19 D50 and SPAN
    Dantrolene sodium IV Oct,17 D50 and SPAN
    Budesonide Nasal May,19 D50 and SPAN
    Barium sulfate Oral Feb,18 EMD
    Cyclosporine Emulsion; ophthalmic Oct,16 EMD

    5
    Definition of D10, D50, D90, and SPAN

    • D90=Dv0.9
    • D10=Dv0.1
    𝐷𝐷𝐷𝐷−𝐷𝐷𝐷𝐷
    • SPAN=𝑥𝑥 =
    𝐷𝐷𝐷𝐷

    The span is a common


    calculation to quantify
    distribution width.

    6
    Definition of Z-average & PDI

    • Z-average = MEAN
    the intensity-weighted
    average
    𝑆𝑆𝑆𝑆𝑆𝑆2
    • 𝑃𝑃𝑃𝑃𝑃𝑃 = MEAN2

    the polydispersity index,


    which quantifies distribution
    width

    7
    Comparing SPAN & PDI

    𝜎𝜎2
    𝑃𝑃𝐷𝐷𝐼𝐼 = ---------------- Eq.1
    µ2
    𝐷𝐷𝐷𝐷−𝐷𝐷𝐷𝐷
    𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 = ---------------- Eq. 2
    𝐷𝐷𝐷𝐷

    The outcome of Equation 1 and Equation 2, both provide the distribution width
    relative to mean of a particle size distribution curve, based on dynamic light
    scattering techniques.

    8
    Relation between SPAN & PDI
    • Assume X ~ Normal(µ, σ), here X is the diameter of a Particle.
    • Here, µ is population Mean and σ is population standard deviation

    • D50 = µ

    • D90 = µ + 1.282 σ and D10 = µ - 1.282 σ


    𝐷𝐷𝐷𝐷−𝐷𝐷𝐷𝐷
    • SPAN = 𝐷𝐷𝐷𝐷
    = 2* 1.282 σ/ µ and PDI = (σ/ µ)^2, by definitions

    Rearranging these equations,

    • SPAN = 2.564 * sqrt(PDI) and thus PDI = 0.152*SPAN^2

    9
    T/R Ratio between SPAN & PDI

    PDI_T/PDI_R=(SPAN_T/SPAN_R)^2
    SPAN Ratio PDI Ratio
    0.80 0.64 PDI magnifies the T/R
    0.85 0.72 difference compared to SPAN.
    0.90 0.81
    0.95 0.90
    Passing PBE for PDI can be
    1.00 1.00
    more difficult compared to
    1.05 1.10
    1.10 1.21
    SPAN.
    1.15 1.32
    1.20 1.44
    1.25 1.56

    10
    T/R Margin for PBE
    1
    0.9 Because the values of
    0.8
    PDI can be very small
    0.7
    (~0.1), the margin for
    Test

    0.6
    passing the PBE is very
    0.5
    0.4
    small (10% of 0.1),
    0.3 making it much harder to
    0.2 Line of Identity pass when even the
    0.1 FDA Limit
    difference is minuscule.
    0
    0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

    Reference

    11
    Conclusions

    • The PDI is not the right parameter to replace SPAN.

    • At a minimum, it is not the appropriate parameter when


    employing the current PBE method.

    12
    New Proposal Based on T-R for PBE
    1
    0.9 • |T-R| < 0.1 instead of
    0.8 1.11 > T/R > 0.9 ratio
    0.7 when chosen to use
    Test

    0.6 PDI.
    0.5
    0.4 Line of Identity
    0.3
    FDA Limit
    0.2
    0.1 New Limit?
    0
    0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

    Reference

    13
    Questions

    • Is employing Z-average and PDI (in place of D50 & SPAN) overly
    sensitive?
    • Why the switch from D50/span to Zavg/PDI? What evidence is there
    that D50/ SPAN was problematic?
    • If PDI or Span represent distribution width (variability), could it be a one-
    sided test? Note: less variability for disparity should be beneficial.
    • Furthermore, do we need in-vitro PBE when in-vivo BE is required? Can
    it be supportive only?

    14

    You might also like