Chemical Stability of Admixtures Containing Ziconotide 25 Mcg-Ml

PEER REVIEWEDFEATURE IJPC

Chemical Stability of Admixtures Containing Ziconotide 25 mcg/mL and Morphine Sulfate 10 mg/mL or 20 mg/mL During Simulated Intrathecal Administration

David E. Shields, PhD ABSTRACT Jennifer Aclan, BSc The chemical stability of an intrathecally administered analgesic combination Aaron Szatkowski, BSc may infl uence the frequency of pump refi lls necessary to maintain safe and Elan Pharmaceuticals, Inc. effective analgesia. Previous work has shown that the stability of ziconotide South San Francisco, California at body temperature is reduced substantially by the presence of morphine sulfate 35 mg/mL. The current study was performed to evaluate the chemical INTRODUCTION stability of admixtures combining ziconotide with lower concentrations of Ziconotide is the synthetic equivalent of morphine sulfate during simulated intrathecal infusion under laboratory a naturally occurring conopeptide isolated conditions at 37°C. Admixtures containing ziconotide 25 mcg/mL and from the venom of the marine snail Conus morphine sulfate 10 mg/mL or 20 mg/mL were stored in implantable magus.1 In 2004, the U.S. Food and Drug intrathecal pumps at 37°C and in control vials at 37°C or 5°C. Samples were Administration approved ziconotide for obtained over 60 days (admixture containing morphine sulfate 10 mg/mL) the management of severe chronic pain or 28 days (admixture containing morphine sulfate 20 mg/mL) and drug in patients for whom intrathecal therapy concentrations were assessed by high-performance liquid chromatography. is warranted and who are intolerant of or Estimates of the time intervals that each admixture retained ≥90% and ≥80% whose pain is refractory to other treat- ments such as systemic analgesics, adjunc- of the initial concentrations of both drugs (i.e., the 90% and 80% stabilities) tive therapies, or intrathecal morphine.1 were based on 95% confi dence bounds obtained via linear regression. Ziconotide monotherapy is recommended Morphine was stable in both admixtures. In the admixture containing as a fi rst-line intrathecal therapy option by morphine sulfate 10 mg/mL, the mean ziconotide concentration declined to the 2007 Polyanalgesic Consensus Confer- 81.4% of the initial concentration in 60 days, and 90% and 80% stabilities ence (PACC) panel members.2 were maintained for 34 days and 65 days, respectively. In the admixture For patients who do not receive satisfac- containing morphine sulfate 20 mg/mL, the mean ziconotide concentration tory pain control from a single intrathe- declined to 85.3% of the initial concentration in 28 days, and 90% and 80% cal agent, intrathecal administration of a stabilities were maintained for 19 days and 37 days, respectively. Decreasing combination of analgesics is an accepted treatment approach.2,3 In theory, patients the concentration of morphine in an admixture containing ziconotide are most likely to benefi t from combina- improves the stability of ziconotide. tion therapy if the combined drugs have different mechanisms of action. Research in animals suggests that ziconotide inhibits that revealed that combination therapy with morphine sulfate,8 but that study used neurotransmission from primary nocicep- intrathecal morphine and ziconotide pro- a higher morphine concentration tive afferents in the dorsal horn of the spi- duced greater pain relief than therapy with (35 mg/mL) than is recommended in guide- nal cord by binding to and directly blocking either analgesic alone.6,7 Combination of lines developed by the 2007 PACC panel (20 N-type voltage-sensitive calcium channels. ziconotide with morphine is recommended mg/mL).3 The current study was performed Morphine administration also inhibits as a second-line intrathecal treatment op- to evaluate the chemical stability of admix- N-type voltage-sensitive calcium channels, tion by the 2007 PACC panel members.2 tures containing ziconotide 25 mcg/mL and but the mechanism is modulated indirectly The chemical stability of an analgesic morphine sulfate 10 mg/mL or 20 mg/mL through opioid receptors, and morphine combination is an important consideration under simulated clinical conditions. exerts additional effects on the activity when choosing medications because less of postsynaptic potassium channels.4,5 stable combinations could result in more Thus, the different analgesic mechanisms frequent pump refi lls and could compro- MATERIALS AND employed by ziconotide and morphine may mise the accuracy of the dose delivered. METHODS yield additive effects when delivered as an Previous work has shown that the stabil- Admixture Preparation intrathecal combination. This theory is ity of ziconotide at body temperature is This study consisted of two experiments. supported by two recent open-label studies reduced substantially by the presence of In the fi rst experiment, ziconotide (Lot

www.IJPC.com International Journal of Pharmaceutical Compounding Vol. 12 No. 6 | November/December 2008 553 PEER REVIEWED IJPC

CM2-345; PRIALT, ziconotide intrathecal and 23-gauge needle. Air was removed from and a waste vial was sealed to the catheter infusion, Elan Pharmaceuticals, Inc., South the syringe, and the fi lter and needle were by pushing the catheter connector pin San Francisco, California) was obtained in fl ushed with 1 mL of admixture solution. through the vial septum. A vent, composed 20-mL vials as a sterile, pyrogen-free solu- After the pump septum was wiped with of a 0.22-mcm sterile syringe fi lter and a tion containing ziconotide 25 mcg/mL, an alcohol swab, the needle was inserted 27-gauge needle, was provided to release methionine 50 mcg/mL (to reduce zi- through the pump septum and the reservoir headspace backpressure as the vial fi lled. conotide oxidation), and sodium chloride was fi lled aseptically with approximately The pump was programmed to deliver at 9 mg/mL. The pH was adjusted to between 5 mL of the admixture solution. The pump the maximum infusion rate and returned 4.0 and 5.0. Aseptic technique was used to was shaken gently to distribute the solution to the 37°C incubator until a suffi cient pool 200 mL of ziconotide solution into a throughout the reservoir, then the ad- amount of solution had been delivered sterile plastic bottle, and 2 g of lyophilized mixture was removed aseptically from the into the vial. Delivery was stopped, and morphine sulfate (Lot 6116-PMS50002; reservoir through the pump septum with the waste vial was replaced with a vented B&B Pharmaceuticals, Inc., Aurora, Colo- another needle attached to a clean syringe. sample collection vial. The pump was again rado) was added to the pooled solution for The pump was then fi lled with the remain- programmed to deliver at the maximum fi nal drug concentrations of ziconotide ing admixture in the fi lter-equipped syringe rate and placed in the incubator at 37°C un- 25 mcg/mL and morphine sulfate 10 mg/mL. and programmed to deliver the admixture til approximately 0.5 mL of the admixture In the second experiment, a similar tech- at the maximum possible rate in order to had been collected in the sample vial. The nique was used; 4 g of lyophilized morphine precondition the pump’s fl uid path. The catheter was detached, and the remaining sulfate was added to the pooled ziconotide admixture was delivered to waste for 30 catheter solution was pushed into the col- solution (Lot CM2-343; Elan Pharmaceu- minutes, then delivery was stopped, and all lection vial using a nitrogen-fi lled syringe. ticals, Inc.) to achieve fi nal drug concen- remaining admixture solution was removed When sampling was complete, the pump trations of ziconotide 25 mcg/mL and from the reservoir by using a clean syringe was programmed to deliver at the minimum morphine sulfate 20 mg/mL. and needle. rate and stored in the 37°C incubator. The Because the primary ziconotide deg- Pumps were then fi lled aseptically with sample collection vials, along with one radation pathway is oxidative, preparative 20 mL of the admixture solution by using a control vial each from the 37°C and 5°C techniques that minimize the presence of new fi lter-equipped syringe. Controls were storage conditions, were transferred to the dissolved oxygen in the admixtures were established at the beginning of the study HPLC apparatus for analysis. used. Dissolution of morphine sulfate was by aseptically fi lling polymethylpentene high-performance liquid chromatography facilitated with gentle shaking to minimize Analytical Equipment and the introduction of atmospheric oxygen (HPLC) vials to capacity with 750 mcL of into the admixtures. After the morphine admixture by using a needle and syringe. Conditions sulfate dissolved fully, the admixtures were Because intrathecal pumps are implanted in Ziconotide and morphine sulfate sparged by bubbling nitrogen through the clinical use, all pumps and half the control concentrations were measured with an solution to remove dissolved oxygen. vials were stored in the 37°C incubator Agilent Series 1100 HPLC system (Agilent during the study, except for the brief period Technologies, Inc., Santa Clara, California), when the pumps were removed from the or equivalent, equipped with a Phenom- Pump Preparation and Fill incubator to attach the catheter before enex Jupiter, 300-Å pore size, C18, 5-mcm, In both experiments, the same pump sampling. The other half of the control vials 250 × 4.6-mm column (Phenomenex, Inc., preparation, fi lling, sampling, and analytical were stored in a refrigerator at 5°C. Torrance, California) maintained at 40°C procedures were used, except as noted. as the stationary phase. The mobile phase After nitrogen sparging, the admixture Sampling consisted of gradient elution starting with and SynchroMed II pumps (Model 8637- Because of the limited number of 100% mobile phase A for 2 minutes, transi- 20; Medtronic Inc., Minneapolis, Min- samples obtainable from each pump, the tioning to 70% mobile phase A:30% mobile nesota) that had been exposed previously sampling schedules were determined while phase B over 12 minutes, then to 30% mo- to ziconotide were warmed for 1 hour in the study was ongoing. Pumps and control bile phase A:70% mobile phase B over the a calibrated 37°C incubator. Each pump vials were sampled within minutes of fi lling next 10 minutes, before returning to 100% septum was wiped with an alcohol swab to (day 0) and on days 5, 12, 26, 40, and 60 mobile phase A for approximately 6 minutes prevent bacterial contamination, and then (admixture containing morphine sulfate residual fl uid was removed from the pump 10 mg/mL) or on days 0, 6, 13, 21, and 28 to re-equilibrate the column. The fl ow rate reservoirs through the pump septa with a (admixture containing morphine sulfate used throughout the 30-minute procedure 23-gauge needle attached to a sterile 20-mL 20 mg/mL). Before sample collection, the was 2.0 mL/min. Mobile phase A consisted syringe. The fl uid was drawn aseptically pump’s fl uid path was equilibrated with the of 98% 25 mmol/L sodium perchlorate, into the syringe and discarded. reservoir contents by delivering approxi- 2% methanol, and 0.05% trifl uoroacetic Pump reservoirs were rinsed twice mately 1.5 mL of admixture solution into acid; mobile phase B consisted of 49.95% with approximately 5 mL of the admixture a waste vial. Each pump’s catheter port cap acetonitrile, 49.95% water, and 0.1% solution to ensure that drug concentra- was removed, and the catheter port was trifl uoroacetic acid. The ultraviolet detector tions were not affected by any residual fl uid swabbed with sterile 70% isopropyl alcohol. was set at 212 nm. Because of the nearly left in the pump reservoir. A sterile 20-mL A 40-cm length of catheter, with a catheter 1000-fold difference between the concen- syringe was fi lled with 10 mL of admixture connector pin pushed into the distal end, trations of morphine sulfate and ziconotide, solution and attached to a 0.22-mcm fi lter was attached to the pump’s catheter port, each sample solution was injected twice.

International Journal of Pharmaceutical Compounding www.IJPC.com 554 Vol. 12 No. 6 | November/December 2008 PEER REVIEWEDFEATURE IJPC

For ziconotide analysis, an injection volume FIGURE 1. The stability of an admixture containing ziconotide 25 mcg/mL of 150 mcL was used; 1 mcL was used for and morphine sulfate 10 mg/mL. morphine sulfate analysis. A qualifi cation was performed on this method, which included stability, range/linearity, accuracy/ FIGURE 1A recovery, specifi city, and precision; all met predefi ned acceptance criteria. The solution stability data indicated that 100 the admixture of ziconotide and morphine sulfate must be assayed within 10 days of sampling to prevent additional degrada- 90 tion. Linearity and range were established for ziconotide between 80% and 400% of standard concentration (25 mcg/mL). 80 Linearity for morphine sulfate was estab- Pumps, 37°C lished between 80% and 110% of standard Concentration Pumps, lower bound of 95% CI concentration (35 mg/mL). The correla- 70 Vial, 37°C tion coeffi cients of the linear regressions

Percentage Initial Ziconotide for each of the linearity series were within Vial, 5°C 60 the acceptance criteria, with correlation coeffi cients of 1.000 and 0.997 for zi- 0 102030405060 conotide and morphine sulfate, respectively. Time, d All results used to calculate linearity were within 3% of the results expected from Ziconotide concentrations. The concentrations of ziconotide measured in the determined regression line. Accuracy replicate implantable intrathecal pumps (open circles) stored at 37°C, and in vials (± 5%) was demonstrated for admixtures stored at 37°C (diamonds) or 5°C (open squares), are expressed as percentages of of ziconotide and morphine sulfate tested the concentration measured at study initiation (day 0). A linear least-squares fi t is within 80% to 110% of specifi ed concen- displayed for the available pump data (solid line), and the lower bound of the 95% confi dence interval (CI; dotted line) is also provided. trations. Specifi city was demonstrated, with both major peaks having a resolution of greater than 1.5. Precision was determined across instruments, analysts, and days. The results indicated that there was no FIGURE 1B signifi cant difference in either ziconotide or morphine sulfate results (<2.5%) when instrument, analyst, or day was varied.

100 Statistics For each pump, the concentrations of ziconotide and morphine sulfate on each 90 sampling day were converted to a percent- age of that measured on day 0. Statistical software package SAS version 8.2 (SAS Pumps, 37°C Institute Inc., Cary, North Carolina) was 80 used to conduct a linear regression analysis Vial, 37°C to predict the time dependence of the Sulfate Concentration Sulfate Vial, 5°C percentage of ziconotide remaining in Percentage Initial Morphine the pumps and the corresponding 95% 70 confi dence interval (CI). Admixture stability 0 102030405060 is reported as the number of days that the lower bound of the 95% CI for concentra- Time, d tions of both drugs remained at or above 90% (90% stability) or 80% (80% stability) Morphine sulfate concentrations. The concentrations of morphine sulfate of the initial concentrations. Values that measured in replicate implantable intrathecal pumps (open circles) stored at 37°C, exceed the duration of the study (i.e., 60 and in vials stored at 37°C (diamonds) or 5°C (open squares), are expressed as percentages of the concentration measured at study initiation (day 0). A linear days for the fi rst experiment or 28 days for least-squares fi t (solid line) of the pump data is displayed. the second experiment) were extrapolated, assuming linear degradation.

www.IJPC.com International Journal of Pharmaceutical Compounding Vol. 12 No. 6 | November/December 2008 555 PEER REVIEWED IJPC

RESULTS FIGURE 2. The stability of an admixture containing ziconotide 25 mcg/mL Morphine Sulfate 10-mg/mL and and morphine sulfate 20 mg/mL. Ziconotide 25-mcg/mL Admixture The mean concentration of ziconotide FIGURE 2A was 97.1% of the initial concentration on day 12 and 81.4% of the initial concentra- tion on day 60 (Figure 1A); the standard deviation (SD) was no more than 1.7% 100 on any sampling day. The lower bound of the 95% CI for ziconotide concentration remained above 90% and 80% of the initial 90 ziconotide concentration for 34 days and 65 days, respectively. A similar rate of decay 80 was observed in the control vials stored at Pumps, 37°C 37°C; the concentration of ziconotide fell to 96.2% of initial concentration on day 12 Concentration Pumps, lower bound of 95% CI 70 and 79.8% of initial concentration on day Vial, 37°C 60. Control vials stored at 5°C maintained

Percentage Initial Ziconotide Vial, 5°C 97.1% of the initial ziconotide concentra- 60 tion on day 60. 0 5 10 15 20 25 30 35 Concentrations of morphine sulfate in pump samples averaged 97.0% of the initial Time, d concentration on day 12 and 97.6% of the initial concentration on day 60 (Figure Ziconotide concentrations. The concentrations of ziconotide measured in replicate 1B); variability in pump measurements was implantable intrathecal pumps (open circles) stored at 37°C, and in vials stored ≤1.4% SD at all time points. The lower at 37°C (diamonds) or 5°C (open squares), are expressed as percentages of the bound of the 95% CI for morphine sulfate concentration measured at study initiation (day 0). A linear least-squares fi t is concentration remained above 90% of the displayed for the available pump data (solid line), and the lower bound of the 95% confi dence interval (CI; dotted line) is also provided. initial concentration for the duration of the study. In control vials stored at either 37°C or 5°C, the concentrations of morphine sul- fate remained at or near 100% of the initial FIGURE 2B concentration throughout the study.

Morphine Sulfate 20-mg/mL and Ziconotide 25-mcg/mL Admixture 100 The concentration of ziconotide in pump samples averaged 93.5% of the initial concentration on day 13 and 85.3% of the 90 initial concentration on day 28 (Figure 2A); variability was no greater than 0.9% SD on any sampling day. The lower bound of Pumps, 37°C the 95% CI for ziconotide concentration 80 Vial, 37°C remained above 90% and 80% of the initial

Sulfate Concentration Sulfate ziconotide concentration for 19 days and Vial, 5°C 37 days, respectively. Ziconotide concen- Percentage Initial Morphine trations in the control vials stored at 37°C 70 displayed a similar rate of decay, falling to 0 5 10 15 20 25 30 94.5% of the initial concentration on day 13 and to 82.4% of the initial concentra- Time, d tion on day 28. Control vials stored at 5°C retained 96.7% of the initial ziconotide Morphine sulfate concentrations. The concentrations of morphine sulfate measured concentration on day 28. in replicate implantable intrathecal pumps (open circles) stored at 37°C, and in vials Morphine sulfate concentrations in pump stored at 37°C (diamonds) or 5°C (open squares), are expressed as percentages of the concentration measured at study initiation (day 0). A linear least-squares fi t (solid line) samples averaged 100.1% of the initial of the pump data is displayed. concentration on day 13 and 97.8% of the initial concentration on day 28 (Figure 2B);

International Journal of Pharmaceutical Compounding www.IJPC.com 556 Vol. 12 No. 6 | November/December 2008 PEER REVIEWEDFEATURE IJPC

variability in pump measurements was when ziconotide and morphine sulfate are ACKNOWLEDGMENT ≤ 2.7% SD at all time points. The lower compounded for intrathecal administration, The authors acknowledge MedLogix bound of the 95% CI for morphine sulfate the stability of ziconotide is strongly de- Communications, LLC, for its contribu- concentration remained well above 90% of pendent on the concentration of morphine tions to the development of this manu- the initial morphine sulfate concentration sulfate. The clinical implications of these script. for the duration of the study. The concen- fi ndings are limited by the in vitro nature of trations of morphine sulfate in control vials the studies. REFERENCES stored at either 37°C or 5°C remained at In both admixtures, ziconotide and mor- 1. PRIALT [package insert]. San Diego, CA: or near 100% of the initial concentration phine sulfate were stable when stored in Elan Pharmaceuticals, Inc.; 2007. throughout the study. vials at 5°C. These results suggest that little, 2. Deer T, Krames ES, Hassenbusch SJ et al. if any, decline in concentration will occur Polyanalgesic consensus conference 2007: during refrigerated storage and shipment. Recommendations for the management DISCUSSION of pain by intrathecal (intraspinal) drug Identifi cation of the degradation Morphine is the most commonly pre- delivery: Report of an interdisciplinary products in these admixtures was beyond scribed intrathecal analgesic for chronic expert panel. Neuromodulation 2007; 10(4): the scope of this investigation. When pain management, and most physicians 300–328. ziconotide alone is stored in implantable 3. Hassenbusch SJ, Portenoy RK, Cousins M who use implantable infusion pumps have pumps, the primary ziconotide degradation et al. Polyanalgesic consensus conference prescribed intrathecal therapy that com- product is produced when oxygen reacts 2003: An update on the management of pain bines morphine with another intrathecally with the peptide methionine (data on fi le, by intraspinal drug delivery—report of an administered drug.9 Evidence from animal expert panel. J Pain Symptom Manage 2004; models and two recent open-label stud- Elan Pharmaceuticals, Inc.); therefore, 27(6): 540–563. ies suggest that simultaneous intrathecal reducing the presence of dissolved oxygen 4. Lacy CF, Armstrong LL, Goldman MP et administration of morphine and ziconotide in admixtures should improve stability. For al. Drug Information Handbook. 11th ed. Hudson, OH: Lexi-Comp; 2002. produces additive analgesia,6,7,10 but high this reason, the admixtures in this study were prepared with lyophilized morphine 5. Pirec V, Laurito CE, Lu Y et al. The com- concentrations of morphine sulfate have bined effects of N-type calcium channel been shown to reduce ziconotide stability,8 sulfate, rather than with a commercially available formulation that may contain blockers and morphine on A delta versus C which may infl uence the clinical utility of fi ber mediated nociception. Anesth Analg this analgesic combination. dissolved oxygen. Oxygen is removed from 2001; 92(1): 239–243. A previous study investigated the commercial ziconotide formulations by 6. Webster LR, Fakata KL, Charapata S et al. stability of an admixture containing zi- nitrogen sparging during manufacture; Open-label, multicenter study of combined conotide 25 mcg/mL and morphine sulfate the additional sparging performed during intrathecal morphine and ziconotide: Ad- preparation of the admixtures for this inves- dition of morphine in patients receiving 35 mg/mL by using the same protocol and ziconotide for severe chronic pain. Pain Med 8 tigation served primarily to remove oxygen statistical analyses described here. Mor- 2008; 9(3): 282–290. that may have entered the solution during phine sulfate was stable under all conditions 7. Wallace MS, Kosek PS, Staats P et al. Phase for the duration of this earlier study (17 compounding. Omitting this step may II, open-label, multicenter study of com- days), whereas ziconotide concentrations speed ziconotide degradation and reduce bined intrathecal morphine and ziconotide: declined to 90% of the initial concentra- admixture stability. Addition of ziconotide in patients receiving tion in 8 days and to 80% of the initial intrathecal morphine for severe chronic pain. Pain Med 2008; 9(3): 271–281. concentration in 15 days. However, the CONCLUSION 8. Shields D, Montenegro R, Ragusa M. Chem- morphine sulfate concentration used in the This report characterized the long-term ical stability of admixtures combining zi- study exceeded the maximum concentration chemical stability of admixtures that paired conotide with morphine or hydromorphone recommended by the 2007 PACC panel ziconotide 25 mcg/mL with morphine during simulated intrathecal administration. members (20 mg/mL).2 In the current sulfate 10 mg/mL or 20 mg/mL under Neuromodulation 2005; 8(4): 257–263. study, the stability of admixtures contain- conditions that simulated clinical use. When 9. Hassenbusch SJ, Portenoy RK. Current ing ziconotide and lower concentrations of stored in implantable pumps at 37°C, the practices in intraspinal therapy—a survey of clinical trends and decision making. J Pain morphine sulfate was investigated. admixture containing morphine sulfate Symptom Manage 2000; 20(2): S4–S11. 10 mg/mL retained at least 90% of the ini- Results of the present study reveal that 10. Wang YX, Gao D, Pettus M et al. In- the stability of ziconotide in admixtures tial ziconotide concentration for 34 days and teractions of intrathecally administered improves substantially when lower concen- 80% of the initial ziconotide concentration ziconotide, a selective blocker of neuronal trations of morphine sulfate are used. In for 65 days, and the admixture containing N-type voltage-sensitive calcium channels, the admixture containing morphine sulfate morphine sulfate 20 mg/mL retained at least with morphine on nociception in rats. Pain 90% of the initial ziconotide concentration 10 mg/mL, 90% stability and 80% stability 2000; 84(2–3): 271–281. for 19 days and 80% of the initial ziconotide were maintained for 34 days and 65 days, concentration for 37 days. Minimal zi- respectively; in the admixture containing conotide loss was observed in refrigerated Address correspondence to David E. morphine sulfate 20 mg/mL, 90% stabil- vials, and morphine sulfate concentrations Shields, PhD, Director of Analytical De- ity and 80% stability were maintained for were stable under all conditions. Clinical tri- velopment, Elan Pharmaceuticals, Inc., 800 19 days and 37 days, respectively. Taken als are necessary to evaluate the effi cacy and Gateway Boulevard, South San Francisco, together, these results demonstrate that safety of this drug combination. CA 94080. E-mail: [email protected]

www.IJPC.com International Journal of Pharmaceutical Compounding Vol. 12 No. 6 | November/December 2008 557