ORIGINAL CONTRIBUTION

Intrathecal Ziconotide in the Treatment

of Refractory Pain in Patients
With Cancer or AIDS
A Randomized Controlled Trial
Peter S. Staats, MD Context Ziconotide (formerly SNX-111) selectively blocks N-type voltage-sensitive
Thomas Yearwood, MD, PhD calcium channels and may be effective in patients with pain that is refractory to opioid
therapy or those with intolerable opioid-related adverse effects.
Steven G. Charapata, MD
Objective To assess the safety and efficacy of intrathecal ziconotide in patients with
Robert W. Presley, MD pain that is refractory to conventional treatment.
Mark S. Wallace, MD Design, Setting, and Patients Double-blind, placebo-controlled, randomized trial
Michael Byas-Smith, MD conducted from March 12, 1996, to July 11, 1998, at 32 study centers in the United
States, Australia, and the Netherlands. Patients were 111 individuals ages 24 to 85
Robert Fisher, MD
years with cancer or AIDS and a mean Visual Analog Scale of Pain Intensity (VASPI)
David A. Bryce, MD score of 50 mm or greater. Patients were randomly assigned in a 2:1 ratio to receive
Eugene A. Mangieri, MD ziconotide or placebo treatment.
Interventions Intrathecal ziconotide was titrated over 5 to 6 days, followed by a
Robert R. Luther, MD
5-day maintenance phase for responders and crossover of nonresponders to the op-
Martha Mayo, PharmD posite treatment group.
Dawn McGuire, MD Main Outcome Measure Mean percentage change in VASPI score from baseline
to the end of the initial titration period.
David Ellis, MD, PhD
Results Of the evaluable population, 67 (98.5%) of 68 patients receiving ziconotide

Z
ICONOTIDE (FORMERLY SNX- and 38 (95%) of 40 patients receiving placebo were taking opioids at baseline (median
111, Neurex Pharmaceuti- morphine equivalent dosage of 300 mg/d for the ziconotide group and 600 mg/d for
cals, Menlo Park, Calif ) is the the placebo group; P=.63, based on mean values), and 36 had used intrathecal mor-
synthetic equivalent of ␻-cono- phine. Mean (SD) VASPI scores were 73.6 (1.8) mm in the ziconotide group and 77.9
peptide MVIIA, a 25–amino-acid poly- (2.3) mm in the placebo group (P=.18). Mean VASPI scores improved 53.1% (95% con-
fidence interval [CI], 44.0%-62.2%) in the ziconotide group and 18.1% (95% CI, 4.8%-
basic peptide present in the venom of
31.4%) in the placebo group (P⬍.001), with no loss of efficacy of ziconotide in the main-
Conus magus, a marine snail. 1 Zi- tenance phase. Pain relief was moderate to complete in 52.9% of patients in the ziconotide
conotide produces potent antinocicep- group compared with 17.5% in the placebo group (P⬍.001). Five patients receiving zi-
tive effects2 by selectively binding to conotide achieved complete pain relief, and 50.0% of patients receiving ziconotide re-
N-type voltage-sensitive calcium chan- sponded to therapy compared with 17.5% of those receiving placebo (P=.001).
nels3,4 on neuronal somata, dendrites, Conclusion Intrathecal ziconotide provided clinically and statistically significant an-
dendritic shafts, and axon terminals, algesia in patients with pain from cancer or AIDS.
thus blocking neurotransmission from JAMA. 2003;291:63-70 www.jama.com
primary nociceptive afferents.
Ziconotide is the first selective N- Author Affiliations: Division of Pain Medicine, Johns Rhino Pias Consulting LLC, Reno, Nev (Dr Luther); Eunoe
Hopkins University School of Medicine, Baltimore, Md Inc, Redwood City, Calif (Drs Mayo and McGuire); Elan
type voltage-sensitive calcium channel (Dr Staats); Integrated Spine Clinic, Daphne, Ala (Dr Pharmaceuticals Inc, San Diego, Calif (Dr Ellis).
blocking agent to be tested in clinical Yearwood); Pain Management Associates, Research Financial Disclosure: Dr Staats is a paid member of
trials. There is no evidence of tolerance Medical Center, Pain Institute, Kansas City, Mo (Dr the Analgesia Advisory Board of Elan Pharmaceuti-
Charapata); Pain Care Specialists, Colorado Springs, Colo cals Inc. The terms of this arrangement are being man-
to ziconotide5 or of addictive behavior (Dr Presley); Center for Pain and Palliative Medicine, aged by the Johns Hopkins University in accordance
in animals (Elan Pharmaceuticals Inc, University of California, San Diego, La Jolla (Dr Wal- with its conflict of interest policies. Dr Wallace has re-
lace); Department of Anesthesia, Emory University ceived an unrestricted research grant from Medtronic
unpublished data), and the drug must School of Medicine, Atlanta, Ga (Dr Byas-Smith); R. C. Neurological.
be administered intrathecally to maxi- Goodman Institute for Pain Management, Sparks Re- Corresponding Author and Reprints: Peter S. Staats,
gional Medical Center, Fort Smith, Ark (Dr Fisher); MD, Division of Pain Medicine, Johns Hopkins Uni-
mize antinociceptive effectiveness and Marshfield Clinic, Marshfield, Wis (Dr Bryce); Ano- versity School of Medicine, 550 N Broadway, Suite 301,
minimize sympatholysis.6 dyne Research Services, Northport, Ala (Dr Mangieri); Baltimore, MD 21205 (e-mail: [email protected]).

©2004 American Medical Association. All rights reserved. (Reprinted) JAMA, January 7, 2004—Vol 291, No. 1 63

Downloaded from jama.ama-assn.org at Capes Consortia on February 16, 2011

INTRATHECAL ZICONOTIDE AND REFRACTORY PAIN IN CANCER OR AIDS

conotide below the quantitation assay States, Australia, and the Netherlands
Figure 1. Study Flow
minimum of 0.039 ng/mL (Elan Phar- needed to have a mean VASPI score of
maceuticals Inc, unpublished data). The 50 mm or greater during the 3 days be-
111 Patients Randomized
rate of CSF clearance of intrathecal zi- fore enrollment, despite a regimen of
conotide approximated the CSF turn- systemic or intrathecal analgesics. The
71 Assigned to Receive 40 Assigned to Receive
Ziconotide Placebo over rate (mean elimination half-life of VASPI is a scale commonly used in as-
4.6 hours). sessments of pain therapy, in which pa-
27 Discontinued 9 Discontinued Placebo An initial, open-label feasibility study tients rate their pain on a scale of 0 mm
Ziconotide 4 Adverse Events
12 Adverse Events 5 Other
evaluated 31 male patients with chronic (no pain) to 100 mm (worst pain imag-
15 Other pain states of diverse indications, in- inable). We discontinued intrathecal
cluding cancer, AIDS, spinal cord in- medications in the patients with im-
68 Included in Efficacy
Analysis
40 Included in Efficacy
Analysis
jury, thalamic pain, and brachial plexus planted pumps at least 3 days prior to
3 Excluded avulsion.14 None of the patients had re- study enrollment. According to ac-
(Insufficient Data)
ceived adequate pain control with opi- cepted clinical practice, the investiga-
oid therapy, including, in many cases, tors weaned patients from baseline in-
intrathecal opioid therapy. Patients re- trathecal medications and maximized
Many patients with cancer or AIDS ceived a continuous infusion of zi- systemic and oral medications to con-
do not receive satisfactory pain re- conotide via an intrathecal catheter be- trol pain during this period.
lief7,8 from systemic administration of ginning at a dosage of 0.3 ng/kg per Exclusion criteria included preg-
opioids and become potential candi- hour, which was titrated upward to the nancy, sepsis or inadequately treated in-
dates for intraspinal analgesia.9-11 How- point of pain relief or intolerable ad- fection, investigational drug use, or pal-
ever, regardless of delivery route, con- verse effects, with a maximum dosage liative surgical procedure(s) within the
cerns about addiction and abuse, and of 300 ng/kg per hour. Of the 24 pa- preceding 30 days; dementia; un-
the potential for developing tolerance, tients who completed the study, 19 ex- treated affective disorders; nonpatent
adverse effects, and pain refractori- perienced an average reduction of 43% spinal canal; severe asthma, cardiac fail-
ness limit the effectiveness of opioids. in their Visual Analog Scale of Pain In- ure, or bradyarrhythmias; and neuro-
In experimental studies of absorp- tensity (VASPI) score. Fifteen patients cardiogenic syncope. Prior to enroll-
tion, distribution, metabolism, and were able to reduce their concomitant ing patients, each study center obtained
elimination, intrathecal ziconotide ap- use of opioids by at least 50%. institutional review board/institu-
peared and diminished rapidly in Investigators have reported the fol- tional ethics committee approval and
plasma and resulted in relatively little lowing central nervous system ad- obtained written informed consent from
plasma protein binding. Intravenous zi- verse effects associated with the use of all patients.
conotide degraded in rat brain tissue in intrathecal ziconotide: dizziness, nys-
2 to 24 hours, produced no detectable tagmus, confusion, abnormal gait, som- Study Design
intermediates, and cleared quickly from nolence, speech difficulties, ambly- For this double-blind, placebo-
both cerebrospinal fluid (CSF) and the opia, ataxia, amnesia, and abnormal controlled, randomized study, we used
circulatory system. This accelerated thought processes.15 These effects di- a central call-in system to randomly
clearance rate likely means that the dis- minished or resolved when the inves- assign patients in a 2:1 ratio to receive
tribution of ziconotide throughout the tigators reduced the infusion rate or ziconotide or placebo treatment, strati-
CSF and its metabolism within the CSF ceased infusion. The patients who re- fied within each center by cancer or AIDS
are also rapid.6 ported major adverse effects had been diagnosis and by history of intrathecal
Investigators demonstrated the an- receiving 0.2 to 5.3 µg/h of intrathecal morphine use. The study began with a
tinociceptive effects of ziconotide in vari- ziconotide. screening phase and preinfusion evalu-
ous animal models of acute and chronic Encouraged by this preliminary evi- ation (1-7 days). In patients without pre-
pain (eg, formalin, hot-plate, and dence of human analgesia, we con- viously implanted pumps, we implanted
ligation).12,13 Toxicology studies re- ducted a randomized, double-blind, pla- an intrathecal catheter and used an exter-
vealed no organ specificity, mutogenic- cebo-controlled trial to test the nal infusion system. Given the risk of
ity, or teratogenicity, even at higher doses effectiveness of intrathecal ziconotide infection with the use of external sys-
than would be used in humans (Elan in treating refractory pain in patients tems, we limited the total time frame for
Pharmaceuticals Inc, unpublished data). with cancer or AIDS. drug infusion to 2 weeks for all patients.
In clinical pharmacokinetic studies In the initial titration phase, partici-
using high intravenous doses and low METHODS pants received ziconotide or placebo for
intrathecal doses, ziconotide exhib- Study Population 5 days (for nonresponders, a discre-
ited a linear clearance from plasma and To be eligible, patients with cancer or tionary additional day at any given dos-
CSF, with plasma concentrations of zi- AIDS from 32 centers in the United age or increased dosage could be given).
64 JAMA, January 7, 2004—Vol 291, No. 1 (Reprinted) ©2004 American Medical Association. All rights reserved.

Downloaded from jama.ama-assn.org at Capes Consortia on February 16, 2011

 INTRATHECAL ZICONOTIDE AND REFRACTORY PAIN IN CANCER OR AIDS

Responders received an additional 5 tion period (mean of the last 2 VASPI (blood chemistry, hematology, urinaly-
days of maintenance therapy while non- scores or of the last 3 VASPI scores if sis, and electrocardiography). For the
responders crossed over to the oppo- the first 2 scores differed by more than timed cognitive tests, we asked pa-
site group for an additional 5 or 6 days. 15 mm). Responders had a 30% or tients to draw a line between num-
Ziconotide responders could enroll in greater decrease in VASPI scores, with bered circles or between numbered and
a long-term, open-label study. no concomitant increase in opioid use lettered circles, or to decode a series of
The protocol defined responders as or change in opioid class. 9 simple symbols.
patients with a 30% or greater decrease We also calculated the percentage
in mean VASPI score and no increase in change in the CPRS, the WBPI, and the Masking
concomitant opioid use or change in opi- KPSS; the change in opioid use; and the Study sponsors, principal investiga-
oid type. The study protocol was con- change in responder status from base- tors, and patients did not know group
sistent across all study centers. line to the end of the maintenance phase assignments during the trial. Only the
and of the crossover phase, as dictated pharmacists who prepared the study
Data Collection prospectively by the protocol. We re- drugs knew each patient’s status.
Starting from screening and preinfu- corded adverse events, vital signs, and
sion visits and in 24-hour intervals, we the results of cognitive assessments Statistical Analysis
calculated values for the VASPI, the (trail-making and digit symbol tests), The planned sample size was 70 evalu-
5-point Category Pain Relief Scale and clinical laboratory evaluations able patients receiving ziconotide (with
(CPRS), the Wisconsin Brief Pain In-
ventory (WBPI), the Karnofsky Perfor-
Table 1. Demographic and Baseline Characteristics of the Evaluable Sample
mance Status Scale (KPSS), as well as
Initial Treatment Assignment, No (%)
mean percentage change in opioid use
and response to treatment. Ziconotide Placebo
Characteristic (n = 68) (n = 40) P Value
Drug Administration Age, y
Mean (SE) 55.3 (1.72) 56.6 (2.07)
Dosing of ziconotide (an aqueous iso- .78
Median (range) 57.5 (24.0-81.0) 56.0 (33.0-85.0)
tonic solution of 100 µg/mL ziconotide Sex
free-base with L-methionine and so- Women 34 20
.97
dium chloride as excipients) and pla- Men 34 20
cebo (identical vehicle) during the early Race
White 57 (83.8) 38 (95.0)
stage of enrollment in the study was 5 Black 8 (11.8) 1 (2.5)
ng/kg per hour. To minimize variabil- .09*
Hispanic 3 (4.4) 0
ity, we changed this dosage to 0.4 µg/h Asian 0 1 (2.5)
with incremental increases in titration Diagnosis
every 12 hours to the maximum toler- Cancer 59 (86.8) 36 (90.0)
Breast 11 12
ated dosage of ziconotide. Based on
Lung 13 5
safety evaluations of the first 48 en-
Colorectal 10 5
rolled participants, we decreased the Prostate 5 2
.62
starting dosage for 60 subsequent par- Myelogenous/lymphatic 7 3
ticipants to 0.1 µg/h or less, with up- Skin 3 2
ward titrations once every 24 hours to Other 10 7
the point of analgesic effect or to a dis- AIDS 9 (13.2) 4 (10.0)
cretionary maximum dosage of 2.4 µg/h. Currently receiving systemic analgesics 64 (94.1) 38 (95.0) .85
We adjusted concentrations as re- Oral morphine equivalent, mg/d†
Mean (SE) 6200 (3310) 4200 (1200)
quired by final daily doses and appara- .63
Median (range) 300 (0-201 800) 600 (0-30 000)
tus used. Patients could receive other
Prior intrathecal morphine 22 (32.4) 14 (35.0) .78
systemic medications, including opi- Pain responsive to intrathecal morphine‡ 9 (40.9) 6 (42.9) .91
oids, as clinically indicated, but not ad- KPSS§
juvant intrathecal agents. Mean (SE) 64.4 (1.92) 60.5 (2.41)
.33
Median (range) 70.0 (30.0-90.0) 60.0 (30.0-90.0)
End Points and Measurements Abbreviation: KPSS, Karnofsky Performance Status Score.
*For white vs black, Hispanic, and Asian combined.
The primary efficacy variable was mean †Compared using analysis of variance; P value is based on mean values.
‡Percentages based on number of patients who received previous intrathecal morphine therapy (n = 22 for ziconotide;
percentage change in VASPI score from n = 14 for placebo).
baseline (score immediately prior to in- §Scores on the KPSS range from 0 (dead) to 100 (normal, with no evidence of disease). Values reported are based on
n = 62 for ziconotide and n = 38 for placebo.
fusion) to the end of the initial titra-
©2004 American Medical Association. All rights reserved. (Reprinted) JAMA, January 7, 2004—Vol 291, No. 1 65

Downloaded from jama.ama-assn.org at Capes Consortia on February 16, 2011

INTRATHECAL ZICONOTIDE AND REFRACTORY PAIN IN CANCER OR AIDS

completed the study on July 11, 1998.

Figure 2. Reduction in Pain Intensity by Dosage and Diagnosis
Patients ages 24 to 85 years were en-
80 rolled at 32 study centers. Seven cen-
Change in VASPI Score From Baseline, %

Ziconotide Placebo P = .001

70 ters enrolled 5 or more patients, 13 cen-
P <.001 P = .02∗ ters enrolled 2 to 4 patients, and 12
60 P = .02∗

50
P = .04 centers enrolled only 1 patient each. We
40
excluded 3 patients in the ziconotide
group because of insufficient VASPI
30
data, leaving an evaluable group of 68
20
patients receiving ziconotide and 40 re-
10
ceiving placebo, an ITT group of 71 pa-
0
tients receiving ziconotide and 40 re-
–10
ceiving placebo, and a total population
–20
All Evaluable Initial Dosage Initial Dosage Patients Patients With
of 112 patients (including 1 open-
Patients ≤0.1 µg/h >0.1 µg/h With AIDS Cancer label, compassionate-use patient)
(n = 108) (n = 60) (n = 48) (n = 13) (n = 95)
(FIGURE 1).
Subsets of the Evaluable Patient Sample (n = 108) More than 25 types of cancer were rep-
resented in the patient population, and
VASPI indicates Visual Analog Scale of Pain Intensity. Error bars indicate SEM. *After adjustment for diagnosis.
more than half of the patients had wide-
spread metastatic disease. Many of the
change in VASPI scores and KPSS nonmetastatic cancers directly infil-
Figure 3. Percentage of Patients With scores, with treatment, previous use of
Various Degrees of Improvement in Pain
trated neural structures, including the
Intensity and Those Meeting Criteria for intrathecal morphine (yes/no), and spinal cord. The most frequently men-
Response to Treatment treatment ⫻ previous use of intrathe- tioned cancer complications causing pain
cal morphine as interaction terms. We were neuropathy, postherpetic neural-
80 Ziconotide (n = 68) used the Cochran-Mantel-Haenszel test gia, pathologic fractures, and complica-
70 Placebo (n = 40) stratified by history of intrathecal opi- tions of cancer radiotherapy. Among the
60 oid use to analyze differences in the patients with AIDS, the most frequent
50
P = .001 CPRS and the WBPI scores between causes of pain were peripheral neuropa-
Patients, %

study baseline and the end of initial ti- thy, Kaposi sarcoma, and postherpetic
40
tration. To compare incidence of ad- neuralgia. The majority of patients had
30 verse events between study groups, we undergone 1 or more surgical proce-
20 used the Cochran-Mantel-Haenszel dures for cancer excision, including hys-
10
general association test stratified by his- terectomy, mastectomy, pneumec-
tory of intrathecal opioid use. For ad- tomy, and amputation (2 patients had
0
<30 30-49 50-89 ≥90 Treatment verse events with fewer than 5 re- undergone hip disarticulation). Three
Response ports, we used the Fisher exact test. All patients had undergone bone marrow
Decrease in VASPI Score, %
statistical tests were 2-sided, and we transplantation.
VASPI indicates Visual Analog Scale of Pain Intensity. considered results statistically signifi- At baseline (TABLE 1) the mean KPSS
Response to treatment defined as a ⱖ30% decrease cant if Pⱕ.05. score was near 60, and 36 patients had
in VASPI scores, without a concomitant increase in opi-
oid use or a change in opioid class.
The intent-to-treat (ITT) sample in- received intrathecal morphine. Medi-
cluded patients who received a study cation use included a median oral mor-
drug and had baseline and follow-up phine equivalent dosage of 300 mg/d
complete VASPI data at the end of the VASPI scores. The evaluable sample had in the ziconotide group and 600 mg/d
initial titration period and no substan- sufficient VASPI data at the end of the in the placebo group. Baseline medi-
tial violations of the protocol) and 35 initial titration period and did not sub- cations included opioids (92% of
evaluable patients receiving placebo. stantially violate the protocol. Effi- patients), antidepressants (44%), anx-
This sample size has a 96% statistical cacy and demographic results cover the iolytics (40%), anticonvulsants (27%),
power at the .05 significance level to de- evaluable sample, and safety results anti-inflammatory agents (28%), and
tect a greater than 30% change in VASPI cover the ITT sample plus 1 open- antipsychotics (16%). The ziconotide
scores in the 2 treatment groups. Sta- label, compassionate-use patient. and placebo groups had comparable
tistical analyses were performed using demographic and clinical characteris-
SAS version 6.12 (SAS Institute Inc, RESULTS tics, including mean VASPI scores (73.6
Cary, NC). We conducted 2-way analy- The study’s first patient was enrolled on [SD, 1.8] mm vs 77.9 [SD, 2.3] mm, re-
ses of variance of the mean percentage March 12, 1996, and the last patient spectively; P=.18).
66 JAMA, January 7, 2004—Vol 291, No. 1 (Reprinted) ©2004 American Medical Association. All rights reserved.

Downloaded from jama.ama-assn.org at Capes Consortia on February 16, 2011

 INTRATHECAL ZICONOTIDE AND REFRACTORY PAIN IN CANCER OR AIDS

From baseline to the end of the ini- The therapeutic effect (ie, improve- statistically significant in both starting-
tial titration phase, mean VASPI scores ment in VASPI score for the zi- dose subgroups and in each underly-
for the evaluable group improved 53.1% conotide group minus the improve- ing disease subgroup (Figure 2). Age,
(95% confidence interval [CI], 44.0%- ment for the placebo group) was sex, or prior treatment with intrathe-
62.2%) in the ziconotide group and
18.1% (95% CI, 4.8%-31.4%) in the pla-
cebo group (P⬍.001) (FIGURE 2). Ef- Table 2. Adverse Events Reported in 5% or More of Patients in the Safety Sample
(Initial Titration Phase)
ficacy results in the ITT sample were
No. (%)
not appreciably different from those in
the evaluable sample; the mean changes Ziconotide Placebo
in VASPI score for the ITT sample were Body System/Adverse Event (n = 72) (n = 40)
51.4% in the ziconotide group and Patients with any adverse event 70 (97.2) 29 (72.5)
Patients with any serious adverse event 22 (30.6) 4 (10.0)
18.1% (95% CI, 17.3%-49.4%) in the
Body as a whole 33 (45.8) 14 (35.0)
placebo group (P⬍.001).
Fever 18 (25.0) 3 (7.5)
Pain relief (based on CPRS scores)
Headache 11 (15.3) 6 (15.0)
was moderate to complete in 52.9% of
Asthenia 5 (6.9) 2 (5.0)
the ziconotide group and in the same
Pain 2 (2.8) 2 (5.0)
range but never reaching complete in
Injection site reaction 0 2 (5.0)
17.5% of the placebo group (P⬍.001).
Cardiovascular system 24 (33.3) 4 (10.0)
Five patients receiving ziconotide
Postural hypotension 17 (23.6) 2 (5.0)
achieved complete pain relief; 50.0% of
Hypotension 6 (8.3) 2 (5.0)
those receiving ziconotide responded
Digestive system 42 (58.3) 16 (40.0)
to therapy, compared with 17.5% of Nausea 21 (29.2) 7 (17.5)
those receiving placebo (P = .001) Vomiting 13 (18.1) 5 (12.5)
(FIGURE 3). Opioid use decreased in the Constipation 9 (12.5) 7 (17.5)
ziconotide group by 9.9% but in- Nausea and vomiting 9 (12.5) 0
creased in the placebo group by 5.1%. Anorexia 5 (6.9) 0
From the protocol-defined re- Diarrhea 5 (6.9) 2 (5.0)
sponder group of 34 patients receiv- Dyspepsia 1 (1.4) 2 (5.0)
ing ziconotide and 7 receiving pla- Metabolic and nutritional system 2 (2.8) 2 (5.0)
cebo, 2 patients (1 from each group) did Peripheral edema 1 (1.4) 2 (5.0)
not continue into the maintenance Musculoskeletal system 7 (9.7) 1 (2.5)
phase. An additional 15 patients re- Myasthenia 6 (8.3) 1 (2.5)
ceiving ziconotide and 5 receiving pla- Nervous system 60 (83.3) 14 (35.0)
cebo were identified as responders by Dizziness 36 (50.0) 4 (10.0)
the investigators (though they did not Nystagmus 33 (45.8) 4 (10.0)
meet strict protocol-defined re- Somnolence 17 (23.6) 3 (7.5)
sponder criteria) and proceeded to the Confusion 15 (20.8) 2 (5.0)
maintenance phase. In patients receiv- Abnormal gait 9 (12.5) 0
ing ziconotide who proceeded to the Nervousness 7 (9.7) 0
maintenance phase of the study (n=48), Ataxia 4 (5.6) 1 (2.5)
ziconotide maintained efficacy, result- Insomnia 4 (5.6) 0
ing in a 69.2% change in VASPI scores Abnormal thinking 4 (5.6) 0
at the end of the initial titration phase, Respiratory system 14 (19.4) 6 (15.0)
compared with 69.4% at the end of the Dyspnea 4 (5.6) 2 (5.0)
maintenance phase. The 26 patients re- Lung disorder 4 (5.6) 0
ceiving placebo who crossed over to the Hypoxia 1 (1.4) 2 (5.0)
ziconotide group during the second Pneumonia 1 (1.4) 2 (5.0)
phase experienced a 44.9% mean re- Skin and appendages 7 (9.7) 4 (10.0)
duction in VASPI score at the end of the Pruritus 3 (4.2) 2 (5.0)
crossover phase. The 12 patients re- Special senses 9 (12.5) 0
ceiving ziconotide who crossed over to Amblyopia 4 (5.6) 0
the placebo group experienced a 4.2% Urogenital system 23 (31.9) 0
mean reduction in VASPI score at the Urinary retention 13 (18.1) 0
end of the crossover phase. Urinary tract infection 7 (9.7) 0

©2004 American Medical Association. All rights reserved. (Reprinted) JAMA, January 7, 2004—Vol 291, No. 1 67

Downloaded from jama.ama-assn.org at Capes Consortia on February 16, 2011

INTRATHECAL ZICONOTIDE AND REFRACTORY PAIN IN CANCER OR AIDS

cal morphine did not affect the im- ziconotide during the initial titration TABLE 4 shows the mean dose and
provement in VASPI scores. The treat- phase were confusion, somnolence, timing of onset for the most frequently
ment groups did not register a and urinary retention (4.2% each); all reported adverse events in patients re-
significant difference in WBPI subsets others were reported in less than 3% ceiving ziconotide. Confusion oc-
or KPSS scores at the end of the initial of the study population. curred much more often in patients re-
titration phase. Nine types of adverse events oc- ceiving ziconotide who were older than
During the initial titration phase, 4 curred with significantly greater fre- 60 years than in those aged 60 years or
(10.0%) of the 40 patients receiving quency in the ziconotide group com- younger. Central nervous system ad-
placebo reported a total of 4 serious pared with the placebo group (TABLE 2), verse events (ie, cognitive dysfunction,
adverse events, and 22 (30.6%) of the but starting at the lower dosage, using vestibular symptoms, and somno-
72 receiving ziconotide in the safety smaller dose increments, and increas- lence) had a median time to resolution
group reported a total of 31 serious ing the interval between dose titrations of 4 days (range, 0-58 days). Adverse
adverse events. Of these 31 events, tended to reduce this frequency. TABLE 3 events led to early discontinuation in 12
we considered the 14 (45.2%) that shows the proportion of patients in each patients receiving ziconotide and in 4 re-
involved the nervous system (5 mod- starting dosage group that experienced ceiving placebo. An additional 15 pa-
erate, 9 severe) to be related to these adverse events. Confusion was the tients receiving ziconotide and 5 receiv-
ziconotide treatment. The remaining only of these 9 adverse events that oc- ing placebo discontinued the study for
serious adverse events were not deter- curred in a larger percentage of pa- such reasons as external catheter com-
mined by the reporting investigators tients with a starting ziconotide dosage plications, patient request, unrelated
to be related to ziconotide treatment. of 0.1 µg/h or less compared with pa- medical conditions, or lack of therapeu-
The most common serious adverse tients with a starting ziconotide dosage tic effect. There were 5 cases of menin-
events for patients who received of more than 0.1 µg/h. gitis in the ziconotide group and 2 in the
placebo group—all in patients with ex-
ternal infusion systems.
Table 3. Ziconotide-Specific Adverse Events in the Safety Sample, by Starting Dosage Thirteen patients died during the
Adverse Events, No. (%) study or the 30-day follow-up period,
Ziconotide Placebo
and we recorded 2 additional deaths af-
ter follow-up (TABLE 5). Death rates
⬎0.1 µg/h ⱕ0.1 µg/h ⬎0.1 µg/h ⱕ0.1 µg/h were not significantly different be-
Event (n = 31) (n = 41) (n = 19) (n = 21)
tween the ziconotide and placebo
Fever 10 (32.3) 8 (19.5) 1 (5.3) 2 (9.5)
groups. Twelve patients died from can-
Postural hypotension 10 (32.3) 7 (17.1) 1 (5.3) 1 (4.8)
Nausea 13 (41.9) 8 (19.5) 5 (26.3) 2 (9.5)
cer, 1 committed suicide, 1 suc-
Nausea and vomiting 5 (16.1) 4 (9.8) 0 0
cumbed to pneumonia, and 1 died from
Vomiting 9 (29.0) 4 (9.8) 2 (10.5) 3 (14.3)
an unknown cause. Of the 5 patients
Abnormal gait 5 (16.1) 4 (9.8) 0 0
who died while receiving placebo, 2 had
Confusion 6 (19.4) 9 (22.0) 1 (5.3) 1 (4.8)
crossed over from the ziconotide group.
Dizziness 18 (58.1) 18 (43.9) 2 (10.5) 2 (9.5) The patient who died due to un-
Somnolence 9 (29.0) 8 (19.5) 1 (5.3) 2 (9.5) known causes received only placebo.
Urinary retention 7 (22.6) 6 (14.6) 0 0 Other than these deaths, no clini-
cally significant changes in vital signs,
laboratory analyses, or cognitive func-
tion test scores occurred in either treat-
Table 4. Dose and Timing of Onset of Most Frequent Ziconotide Adverse Events
ment group from baseline to study ter-
Mean (SE) Mean (SE)
No. of Days Mean (SE) Dosage Cumulative Dose mination. There were no reports of
Adverse Event No. to Onset at Onset, µg/h at Onset, µg anaphylaxis or hypersensitivity to zi-
Urinary retention 13 2.23 (0.5) 1.44 (0.8) 37.4 (17.3) conotide.
Amblyopia 6 2.33 (0.5) 1.41 (1.0) 42.4 (21.0)
Nausea 34 2.44 (0.3) 0.89 (0.2) 34.2 (11.0) COMMENT
Somnolence 17 2.47 (0.3) 0.84 (0.2) 30.8 (9.3) We demonstrated the clinically and sta-
Dizziness 36 2.50 (0.2) 0.97 (0.2) 27.3 (5.0) tistically significant analgesic effect of
Hypotension 23 2.70 (0.4) 1.38 (0.6) 54.3 (19.0) intrathecal ziconotide in a heterog-
Nystagmus 33 2.76 (0.2) 1.45 (0.6) 51.3 (19.8) eneous, complex, and treatment-
Confusion 15 3.00 (0.4) 0.62 (0.1) 39.8 (14.0) refractory patient population. Our
Asthenia 11 3.18 (0.5) 1.01 (0.4) 44.9 (12.5) results validate reports of the antino-
Abnormal gait 13 3.31 (0.4) 1.14 (0.2) 45.0 (11.6) ciceptive effects of ziconotide, 12,13
68 JAMA, January 7, 2004—Vol 291, No. 1 (Reprinted) ©2004 American Medical Association. All rights reserved.

Downloaded from jama.ama-assn.org at Capes Consortia on February 16, 2011

 INTRATHECAL ZICONOTIDE AND REFRACTORY PAIN IN CANCER OR AIDS

Table 5. Deaths Reported During the Study and Within 30 Days Poststudy
No. of Days No. of Days No. of Days Cumulative
Receiving Receiving After Last Dose Ziconotide Relationship
Patient Sex, Age Ziconotide Placebo Until Death Dose, µg Cause of Death to Treatment
Initial Titration Phase
Ziconotide
Female, 59 y 4 ... 55 80.40 Carcinoma of lung Unrelated
Male, 51 y 3 ... 1 116.40 Carcinoma of lung Underlying condition
Placebo
Female, 53 y ... 5 40 Unknown Unknown Unknown
Maintenance Phase
Ziconotide
Female, 66 y 11 ... 15 166.57 Endometrial carcinoma Unrelated
Female, 57 y 9 ... 0 26.09 Carcinoma Unrelated
Female, 53 y 10 ... 2 84.40 Carcinoma of lung Unrelated
Female, 45 y 10 ... 4 40.06 Endometrial carcinoma Unrelated
Male, 63 y 8 ... 3 531.26 Suicide Possible
Placebo
Male, 40 y ... 11 13 0 Carcinoma Unrelated
Crossover Phase
Ziconotide
Male, 60 y 5 5 29 36.07 Prostatic carcinoma Unrelated
Male, 55 y 5 5 20 20.70 Carcinoma of lung Unrelated
Male, 67 y 5 1 7 33.21 Pneumonia Probable
Placebo
Female, 52 y 5 5 7 78.60 Breast carcinoma Underlying condition
Male, 70 y 5 5 3 24.13 Carcinoma of lung Underlying condition
Male, 70 y 5 6 8 58.39 Carcinoma of lung Unrelated

as well as the findings of an open-label cerebellum.16 These effects were eas- and with refractory pain. Compared
feasibility study in patients with intrac- ily recognizable and reversible, how- with placebo, ziconotide was associ-
table pain.2 Animal studies suggested ever, and their incidence decreased ated with a number of (typically dose-
that our patients would not develop tol- when we decreased the initial dose and related) adverse events: abnormal gait,
erance to the analgesic effects of zi- the frequency of titration. dizziness, nystagmus, confusion, som-
conotide within the time frame of this The incidence of meningitis in this nolence, fever, postural hypotension,
trial, perhaps because ziconotide binds trial can be attributed to 2 factors. First, urinary retention, nausea, and vomit-
directly to calcium channels, bypass- the study patients were very likely im- ing. Completed and ongoing trials us-
ing G-protein–dependent secondary munosuppressed. Second, the limited ing lower dosages of ziconotide and
messenger mechanisms.13 life expectancy of many of the pa- longer durations of treatment will bet-
The robust analgesic effect of zi- tients made it prudent to administer ter define the long-term risk-benefit
conotide observed in our study popu- ziconotide via implanted catheter or profile of this potent analgesic.
lation underscores the finding that in- external infusion systems or via preex-
trathecal ziconotide is 1000-fold more isting implanted pumps. Each of the 7 Author Contributions: Dr Staats, as principal inves-
tigator of this study, had full access to all of the data
potent than intrathecal morphine in cases of meningitis occurred in pa- in the study and takes responsibility for the integrity
blocking the phase-2 responses in the tients with external systems. Thus, the of the data and the accuracy of the data analyses.
Study concept and design: Staats, Wallace, Mang-
formalin test that are thought to de- high rate of infection appears to be due ieri, Luther, McGuire.
pend on central sensitization.12 to poor physiological status and pres- Acquisition of data: Staats, Yearwood, Charapata, Pres-
ley, Wallace, Byas-Smith, Fisher, Bryce, Mangieri,
Because this was a safety and effi- ence of an externalized catheter, not to Luther, Mayo, McGuire, Ellis.
cacy study, we deliberately maxi- an idiosyncratic effect of the drug. Analysis and interpretation of data: Staats, Mang-
mized the incidence of ziconotide ad- The overall frequency of deaths in ieri, Luther, Mayo, McGuire, Ellis.
Drafting of the manuscript: Staats, Mangieri, Mayo, Ellis.
verse events. Ziconotide produced our study was consistent with the un- Critical revision of the manuscript for important in-
significantly more vestibular effects (eg, derlying disease status of the patients. tellectual content: Staats, Yearwood, Charapata, Pres-
ley, Wallace, Byas-Smith, Fisher, Bryce, Luther,
somnolence, confusion, urinary incon- Our randomized, double-blind, pla- McGuire, Ellis.
tinence, and fever) than did placebo, cebo-controlled trial revealed the con- Statistical expertise: Ellis.
Obtained funding: Luther, Mayo, McGuire.
perhaps because ziconotide blocks N- siderable efficacy of ziconotide in pa- Administrative, technical, or material support: Staats,
channels in the granular cell layer of the tients with end-stage cancer or AIDS Luther, Mayo, Ellis.

©2004 American Medical Association. All rights reserved. (Reprinted) JAMA, January 7, 2004—Vol 291, No. 1 69

Downloaded from jama.ama-assn.org at Capes Consortia on February 16, 2011

INTRATHECAL ZICONOTIDE AND REFRACTORY PAIN IN CANCER OR AIDS

Study supervision: Staats, Presley, Fisher, Mangieri, conduct of the study and the collection, analysis, and listings from the trial were provided to the correspond-
Luther, Mayo, McGuire. interpretation of the data obtained. Contract re- ing author for his use in the preparation and review
Funding/Support: Neurex and Medtronic provided fi- search organizations (IBAH Inc and Clinmetrics) were of the manuscript. Medtronic was responsible for pro-
nancial support for the study; Medtronic also pro- responsible for the monitoring of the trial and the study viding SynchroMed infusion systems and for analyz-
vided the Synchromed infusion systems. After the sites, the resolution of data queries, and the transfer ing and reporting the performance of the Syn-
merger of Neurex with Elan Pharmaceuticals Inc, Elan of completed case report forms to the data manage- chroMed infusion system.
provided continuing financial support for the final re- ment group. Data management was performed by Co- Acknowledgment: We thank Pacific Research Asso-
porting of the trial results to the Food and Drug Ad- vance and included data entry, statistical analysis, and ciates, Palo Alto, Calif, and Covance, Princeton, NJ,
ministration. Neurex (now Elan) provided both the generation of data tables and listings. The prepara- for statistical analysis and data management; Kather-
study drug and the placebo. tion of the final trial report was performed jointly by ine K. Bass, MD, for assistance in manuscript prepa-
Role of the Sponsor: The study sponsor at the time Covance and the sponsor. The preparation and re- ration; Jere Fellmann, PhD, and Judith Lynn, MBA,
of the initiation of the trial was Neurex. The sponsor view of the manuscript were a joint effort among the for clinical trials and project management; Frederick
designed the study with the advice of physician in- authors, the sponsor, and a contract medical writer. W. Luthardt, MA, for checking all factual details of
vestigators who were expert in the use of intrathecal Elan pharmaceuticals provided financial support for the the study; and MedLogix Communications for edito-
therapy. The sponsor was responsible for the overall medical writer. All pertinent original data tables and rial support.

REFERENCES
1. Olivera B, Gray WR, Zeikus R, et al. Peptide neu- 6. McGuire D, Bowersox S, Fellmann JD, Luther RR. siol Scand Suppl. 1987;85:47-53.
rotoxins from fish-hunting cone snails. Science. 1985; Sympatholysis after neuron-specific, N-type, voltage- 12. Bowersox SS, Gadbois T, Singh T, Pettus M, Wang
230:1338-1343. sensitive calcium channel blockade: first demonstra- YX, Luther RR. Selective N-type neuronal voltage-
2. Brose W, Pheifer B, Hassenbusch S, et al. Analge- tion of N-channel function in humans. J Cardiovasc sensitive calcium channel blocker, SNX-111, pro-
sia produced by SNX-111 in patients with morphine Pharmacol. 1997;30:400-403. duces spinal antinociception in rat models of acute,
resistant pain. Presented at: 15th Annual Meeting of 7. Breitbart W, Rosenfeld BD, Passik SD, McDonald persistent and neuropathic pain. J Pharmacol Exp Ther.
the American Pain Society; November 14-17, 1996; MV, Thaler H, Portenoy RK. The undertreatment of 1996;279:1243-1249.
Washington, DC. pain in ambulatory AIDS patients. Pain. 1996;65:243- 13. Malmberg AB, Yaksh TL. Effect of continuous in-
3. Olivera BM, Cruz LJ, de Santos V, et al. Neuronal 249. trathecal infusion of omega-conopeptides, N-type cal-
calcium channel antagonists: discrimination between 8. Zech DF, Gron S, Lynch J, Hertel D, Lehmann KA. cium-channel blockers, on behavior and antinocicep-
calcium channel subtypes using omega-conotoxin from Validation of World Health Organization Guidelines tion in the formalin and hot-plate tests in rats. Pain.
Conus magus venom. Biochemistry. 1987;26:2086- for cancer pain relief: a 10-year prospective study. Pain. 1995;60:83-90.
2090. 1995;63:65-76. 14. Mathur VS. Ziconotide: a new pharmacological
4. Miljanich G, Ramachandran J. Antagonists of neu- 9. Staats PS. Cancer pain: beyond the ladder. J Back class of drugs for the management of pain. Semin
ronal calcium channels: structure, function, and thera- Musculoskeletal Rehabil. 1998;10:69-80. Anesth. 2000;19:67-75.
peutic implications. Annu Rev Pharmacol Toxicol. 10. Jacox AK, Carr DB, Payne R, et al. Clinical Prac- 15. Penn RD, Paice JA. Adverse effects associated with
1995;35:707-734. tice Guideline Number 9: Management of Cancer Pain. the intrathecal administration of ziconotide. Pain. 2000;
5. Mathur VS, McGuire D, Bowersox SS, Miljanich GP, Rockville, Md: Agency for Health Care Policy and Re- 85:291-296.
Luther RR. Neuronal N-type calcium channels: new search; 1994. AHCPR publication 294-0592. 16. Vandaele SF, Reader TA. Ca(2+)-sensitive and in-
prospect in pain therapy. Pharm News. 1998;5:25- 11. Ventafridda V, Spoldi E, Caraceni A, DeConno F. sensitive omega-conotoxin GVIA binding sites in rat
29. Intraspinal morphine for cancer pain. Acta Anaesthe- brain. Neuroreport. 1994;5:1121-1124.

70 JAMA, January 7, 2004—Vol 291, No. 1 (Reprinted) ©2004 American Medical Association. All rights reserved.

Downloaded from jama.ama-assn.org at Capes Consortia on February 16, 2011