Abstract
Ziconotide, a potent, selective, reversible blocker of neuronal N-type voltage-sensitive calcium channels, is approved in the United States for the management of severe chronic pain in patients for whom intrathecal therapy is warranted, and who are intolerant or refractory to other treatment, such as systemic analgesics, adjunctive therapies, or intrathecal morphine. In the European Union, Ziconotide is indicated for the treatment of severe chronic pain in patients who require intrathecal analgesia. Nonclinical investigations of ziconotide included a comprehensive characterization of its toxicology, incorporating acute and subchronic toxicity studies in rats, dogs, and monkeys; reproductive toxicity assessments in rats and rabbits; and mutagenic, carcinogenic evaluations performed in vivo and in vitro. Additional investigations assessed the potential for cardiotoxicity (rats) and immunogenicity (mice, rats, and guinea pigs), and the presence or absence of intraspinal granuloma formation and local cell proliferation and apoptosis (dogs). The resulting nonclinical toxicology profile was predictive of human adverse events reported in clinical trials and consistent with ziconotide’s pharmacological activity. Frequently observed nonclinical behavioral effects included tremoring, shaking, ataxia, and hyperreactivity. Occurrences were generally transient and reversible upon cessation of treatment, and intolerable effects occurred at doses more than 45 times the maximum recommended clinical dose. Ziconotide was not associated with target organ toxicity, teratogenicity, or treatment-related gross or histopathological changes; it displayed no mutagenic or carcinogenic potential and no propensity to induce local cell proliferation or apoptosis. Although guinea pigs developed systemic anaphylaxis, antibodies to ziconotide were not detected in mice, rats, or guinea pigs, indicating low immunogenic potential. No evidence of granuloma formation was observed with intrathecal ziconotide treatment. In summary, the results from these nonclinical safety assessments revealed no significant toxicological risk to humans treated with ziconotide as recommended.
Ziconotide, also referred to in the literature as PRIALT (Elan Pharmaceuticals, San Diego, CA) and SNX-111, is the synthetic equivalent of ω-conopeptide MVIIA, a component of the paralytic venom of the piscivorous marine snail Conus magus (Miljanich 1997; Olivera et al. 1987). It is composed of 25 amino acids with three disulfide bonds, it has a molecular weight of 2639 Da, and its molecular formula is C102H172N36O32S7 (Figure 1). The chemical, biological, pharmacological, and therapeutic properties of ziconotide have been extensively studied and reviewed (Miljanich and Ramachandran 1995; Terlau and Olivera 2004; Newcomb and Miljanich 2002; Bowersox and Luther 1998; Miljanich 2004). Ziconotide selectively and reversibly blocks N-type voltage-sensitive calcium channels (VSCCs), which leads to the inhibition of neurotransmitter release (Olivera et al. 1987; Miljanich and Ramachandran 1995). In vitro binding studies with radiolabeled ziconotide in rat tissue have identified binding sites for ziconotide in the dorsal horn of the spinal cord (Gohil et al. 1994), where primary nociceptive afferent synapses are located. Ziconotide has also been shown to bind and block cloned human N-type calcium channels (NCCs); thus, intrathecal (IT) delivery of ziconotide in humans likely facilitates binding of NCCs in the dorsal horn to reduce pain signaling and produce potent analgesia. Ziconotide does not bind to μ- or κ-opioid receptors, and its affinity for the δ-opioid receptor is five orders of magnitude lower than its affinity for N-type VSCCs (Bowersox and Luther 1998).
Ziconotide, the first N-type VSCC analgesic to enter clinical trials, was approved in the United States in 2004 for the management of severe chronic pain in patients for whom IT therapy is warranted, and who are intolerant or refractory to other treatment, such as systemic analgesics, adjunctive therapies, or IT morphine (PRIALT [package insert] 2005). In 2005, ziconotide was approved in the European Union for the treatment of severe chronic pain in patients who require IT analgesia (PRIALT [summary of product characteristics] 2005). Nonclinical findings are described here to support and supplement the safety profile determined by clinical trials (Rauck et al. 2006; Staats et al. 2004; Wallace et al. 2006). In addition, because ziconotide is the first N-type VSCC-blocker to receive U.S. Food and Drug Administration approval, these findings may serve as a guide for future research on other drugs in this class.
MATERIALS AND METHODS
Test Materials
Ziconotide was manufactured under Good Manufacturing Practice regulations, with documented impurity profiles for each lot tested. With few exceptions, depending on study requirements, ziconotide was formulated in 0.9% sodium chloride for injection, with or without methionine (0.05 mg/ml), an antioxidant added to the final drug product to enhance stability. The stability of ziconotide in the drug product vehicle (± methionine) was determined by periodic analyses of the formulation before study initiation, during the study, and/or after study completion, as required by the conditions of the study.
Nonclinical Safety Studies
Animal studies were conducted in species commonly used in the development of pharmaceutical products and were selected according to internationally recognized and accepted regulatory guidelines for nonclinical animal studies. Procedures involving the care and use of animals were reviewed and approved by the appropriate Institutional Animal Care and Use Committee (IACUC) prior to conduct. During studies, animal care and use complied with all applicable sections of the Final Rules of the U.S. Animal Welfare Act Regulations (9 CFR) (U.S. Department of Agriculture 1989) and/or other appropriate regional animal welfare guidelines at the time of conduct.
The majority of studies forming the basis of the ziconotide safety pharmacology assessment were designed to identify unknown effects on organ system function and to broaden pharmacological characterization. Safety pharmacology end points were evaluated either early during ziconotide development as part of general pharmacology assessments or later as part of Good Laboratory Practice (GLP) acute and up to 28-day sub-chronic toxicology studies, which incorporated cardiovascular, respiratory, and target organ assessments that supported the findings from pharmacodynamic studies. Further, an in vitro electrophysiology study was performed to evaluate the influence of ziconotide on cardiac action potentials in mammalian cells stably transfected with cloned human ether-a-go-go-related (hERG), hKvLQT1/hminK, or rKv4.3 potassium ion-channel cDNA, thereby assessing potential effects on QT-interval prolongation in vivo.
Pharmacokinetic parameters (i.e., absorption, distribution, metabolism, and excretion) were evaluated using a radioimmunoassay (RIA) to quantify ziconotide plasma and/or cerebrospinal fluid (CSF) levels in dogs following bolus and continuous IT infusions, and in rats and monkeys following intravenous (IV) infusions. Species-specific protein binding was assessed by ultrafiltration and RIA. Studies of radiolabeled ziconotide were conducted in several species to characterize whole blood partitioning, protein metabolism, and brain distribution.
A toxicology program was conducted with ziconotide according to GLP regulations and was consistent with the International Conference on Harmonization Tripartite Safety Guidelines (Table 1). Acute and subchronic toxicity studies were performed in rats, dogs, and monkeys; reproductive toxicity, including fertility, embryo-fetal development, and prenatal/postnatal development, was studied in rats and/or rabbits. Mutagenic and/or carcinogenic assessments included a bacterial mutation (Ames) test in Salmonella typhimurium and Escherichia coli, a forward mutation assay in the L5178Y mouse lymphoma cell line at the thymidine kinase locus, an in vivo micronucleus assay in CD-1 mice, and a Syrian hamster embryo (SHE) cell transformation assay.
In addition to standard histopathology in rat and dog IT toxicology studies, detailed neurohistopathological examinations were performed on hematoxylin and eosin–stained neurological tissues including brain sections; spinal cord (cervical, thoracic, and lumbar); dorsal spinal nerve roots and ganglia; the sciatic, sural, and tibial nerves; and optic nerves. In a 28-day dog IT toxicology study, special stains and cell markers were used to assess the consequences of N-channel blockade on neuronal cell morphology and the local central nervous system (CNS) effects on non-neuronal cells. Additional stains included Luxol fast blue/periodic acid-Schiff for myelin, anti–glial fibrillary protein to specifically detect astrocytes and assess the extent of injury to the CNS, and Fluoro-Jade B to specifically identify degenerating and/or necrotic neurons and neuronal cell processes. In lieu of chronic carcinogenicity studies in rodents, cell proliferation and apoptosis were also evaluated to give assurance that cell cycle control is not disturbed. For this purpose, serially prepared slides of dog spinal cord and brain tissues from all vehicle control and high-dose ziconotide (0.240 μg/kg per hour) animals were stained for proliferating cell nuclear antigen and apoptosis using the terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling (TUNEL) assay. The possible formation of intraspinal granulomas at the IT catheter tip was also evaluated in the dog toxicology studies. Additional toxicological assessments (Table 2) included cardiotoxicity in rats at high parenteral sympatholytic doses of ziconotide and immunogenicity in mice, rats, and guinea pigs using conventional immunization protocols.
To mimic clinical use, ziconotide was delivered by continuous IT infusion to the subarachnoid space through a catheter introduced via the atlanto-occipital membrane and advanced caudally (rats, dogs) or inserted directly at the lumbar-sacral vertebra (dogs), so that the catheter tip lay between the L2 and L4 vertebrae. Plasma levels of ziconotide following IT infusion are extremely low, so ziconotide was administered IV in a number of studies to maximize the exposure of major organ systems outside the CNS and potentially elicit nonspecific toxic effects. Ziconotide was also administered epidurally (EPI) to dogs in subchronic toxicity evaluations and by other routes of administration in immunogenic investigations that included various induction/challenge combinations of IV, intradermal (ID), intraperitoneal (IP), intramuscular (IM), and subcutaneous (SC) delivery.
The duration of studies ranged from acute bolus and short continuous IT infusions (rats, dogs) to subchronic continuous IV infusions lasting up to 14 days (rats, cynomolgus monkeys) and continuous IT infusions (rats, dogs) or continuous EPI infusions (dogs) lasting 1 month. Although ziconotide is approved for long-term IT administration in humans, chronic toxicology studies, including carcinogenicity, were not performed due to several factors including (1) the technical limitations of chronic IT administration in animals; (2) the negligible systemic exposure or drug accumulation with IT administration of ziconotide; and (3) the chemical composition of ziconotide, a peptide containing amino acids that are naturally occurring in humans, poses no carcinogenic potential due to the classical mechanisms of chemical interaction with DNA, as demonstrated in a battery of mutagenic and cell transformation assays.
Exposure Multiples and Safety Ratios
The relative exposure of study animals to a drug, compared to humans administered the maximum recommended clinical dose, is often reported by calculating exposure multiples. Typically, this calculation is done by dividing the measured drug concentration in animal plasma by that measured in human patients; however, human plasma levels of ziconotide are often below the lowest quantifiable limit (0.04 ng/ml) over the recommended range of ziconotide IT infusion rates. For the exposure multiples reported here, the maximum human plasma exposure (0.09 ng/ml) was estimated by dividing the maximum recommended clinical infusion rate (0.8 μcg/h) (PRIALT [package insert] 2005) by the low-range plasma clearance value (9200 ml/h). Because this approximation assumes no ziconotide loss due to metabolism during CSF clearance into plasma, the exposure multiples reported are conservative estimates, and dose multiples are included to aid interpretation. The maximum human CSF exposure is reported as the mean maximum concentration after a 1-h IT infusion at the maximum recommended ziconotide infusion rate. The safety ratio was determined from plasma concentrations in animals at doses that did not cause adverse toxic effects (no observed adverse effect level [NOAEL]) and the maximum predicted human plasma exposure from IT administration. Determination of the NOAEL did not include adverse effects associated with the expected reversible pharmacological effects of the drug because these were not considered to be adverse toxic effects.
RESULTS
Summary of Relevant Nonclinical Safety Pharmacology
Ziconotide evokes a characteristic shaking behavior in laboratory animals when administered at or above analgesic doses by either the IT or IV route. This behavior is mediated by CNS mechanisms, may abate with continued treatment, and is reversible upon the cessation of ziconotide treatment. Cardiovascular effects are limited primarily to hypotension in rats when ziconotide is administered parenterally at several orders of magnitude above IT analgesic doses (Bowersox et al. 1992). This effect is most likely due, at least in part, to blockade of neurotransmitter release from sympathetic autonomic efferents innervating peripheral vasculature (Wang et al. 1998). Hypotension can also result from degranulation of mast cells when ziconotide is administered IV at very high doses, which is consistent with histamine release induced by polybasic peptides and, more specifically, appears to be dependent on the arginine residue at position 10 of the ziconotide amino acid sequence (Bowersox et al. 1992). Ziconotide showed no potential to prolong QT interval in an in vitro electrophysiology study of cardiac action potential or in numerous in vivo cardiovascular assessments of IT and IV ziconotide in dogs and monkeys at doses that were orders of magnitude above the clinical IT dose. Analgesic doses of IT ziconotide do not depress respiratory minute volume responses to carbon dioxide inhalation in rats (Dean, Bowersox, and Luther 1995). In vitro and in vivo studies show that ziconotide acts locally on the gut to decrease gastrointestinal (GI) motility, although the magnitude of this effect in vivo varies considerably depending on the route of administration. At doses 10 to 100 times greater than necessary for analgesia, bolus IT ziconotide decreases gastric motility by 50% in both mice and rats, suggesting a central mechanistic role. Ziconotide does not attenuate the characteristic effects of morphine withdrawal and does not maintain IV morphine self-administration in monkeys. Thus, ziconotide is unlikely to be associated with tolerance or withdrawal symptoms. Nonclinical investigations suggest that ziconotide has little propensity to produce adverse cardiovascular, respiratory, or GI effects at human IT therapeutic doses, and those effects observed at higher doses are generally transient and reversible.
Summary of Nonclinical Pharmacokinetics
Nonclinical pharmacokinetic (PK) studies in dogs have shown that ziconotide, a large hydrophilic molecule, distributes rapidly in CSF after lumbar IT administration, then redistributes relatively rapidly into plasma (Yaksh et al. 1999). The CSF clearance rate in dogs, as determined by continuous 48-h IT infusion of 0.35 μg/kg per hour, was 0.13 ml/min, with terminal CSF half-life (t1/2) of 1.34 h, whereas plasma t1/2 was 1.12 h. The terminal CSF t1/2 (1.79 h) observed after IT bolus injection of 0.69 μg/kg ziconotide is close to the CSF turnover rate, suggesting that ziconotide exits CSF through bulk flow. This was supported by a similar redistribution and clearance profile for coadministered inulin, a large hydrophilic nondiffusible compound used as a marker of rostral bulk-flow CSF clearance. The lumbar CSF:plasma ratios during continuous infusion reflect continuing ziconotide transport out of CSF, decreasing from 2400:1 after 8 h to 1200:1 after 24 h and 600:1 after 48 h.
Plasma PK parameters were determined after 24-h IV ziconotide infusions of 0.4 or 1.7 mg/kg per hour in rats and 0.5 or 1.0 mg/kg per hour in monkeys (Bowersox et al. 1997). In rats, a short initial mean t1/2 of 0.4 h and a longer terminal mean t1/2 of 4.6 h were observed. Results were similar in monkeys: a short initial mean t1/2 of 0.7 h and a longer terminal mean t1/2 of 6.5 h. In both species, more than 99% of the drug was cleared from the plasma by 18 h after cessation of the infusion; mean clearance ranged from 300 to 450 ml/h per kilogram, remained fairly constant between sexes, and was not affected by dose.
Plasma protein binding of ziconotide in rats, dogs, and humans is relatively low (54%, 62%, and 53%, respectively) and is not saturable over the concentration range likely to be encountered clinically. A relatively small proportion of ziconotide partitions into the cellular components of blood (partitioning coefficients range from 0.24 to 0.43 in rat, dog. monkey, and human) and is largely independent of drug concentration. The metabolism of ziconotide has been studied in rat brain and plasma after IV administration of radiolabeled ziconotide (Newcomb et al. 2000). The maximal amount of intact ziconotide in plasma (1.0%/g) and brain (0.006%/g) occurred within 20 min of IV injection and declined to undetectable levels (≤0.00025 %/g) in brain tissue 2 to 24 h after injection. Significant amounts of intact ziconotide and free radiolabeled tyrosine, as well as smaller amounts of unlabeled material, were found in both plasma and brain, but no intermediates were detected, indicating that ziconotide is metabolized at multiple sites by endogenous peptidases to yield constituent amino acids and no active metabolites.
Acute Toxicology Studies
Dogs received ascending, discontinuous, IT ziconotide doses by either bolus injection of 0.03 to 15 μg/kg over a 15-day period or continuous 24- or 48-h infusions of 0.005 to 0.48 μg/kg per hour over 11 days. Clinical signs consisted primarily of abnormal vocalization, lethargy, mild to moderate trembling with impaired movement, twitching, labored breathing, ataxia, and prostrate position. Ziconotide administration did not decrease blood pressure or cause any electrocardiogram (ECG) changes, and most clinical signs resolved during off-dose periods. Due to the severity of clinical signs at higher doses, the maximum tolerated dose (MTD) by acute IT administration was considered to be 0.048 μg/kg per hour.
Rats received continuous IV infusions of ziconotide from 0.08 to 67 mg/kg per hour for up to 24 h. Doses up to 1.7 mg/kg per hour were generally well tolerated but led to tremors, ptosis, hypoactivity, piloerection, and hunched posture, which interfered with the animals’ ability to feed, resulting in slight decreases in food intake and body weight. At doses greater than 8 mg/kg per hour for up to 6 h, all rats displayed severe tremors and died or were sacrificed moribund before the end of the dosing period. Monkeys received continuous 6-h IV infusions of ziconotide up to 29 mg/kg per hour. At doses greater than 0.58 mg/kg per hour, treatment-related weakness, tremor, incoordination, ataxia, and hypoactivity were observed. Heart rate and blood pressure were decreased in all treated animals, and with the exception of a slight QT prolongation in one high-dose male, ECGs were normal. Severe tremors and moribundity occurred at the highest dose.
Subchronic Toxicology Studies
Intrathecal ziconotide was administered for 28 days at doses up to 1.5 μg/kg per hour in rats and 1.2 μg/kg per hour in dogs (Shopp, Skov, and Yaksh 2004). These doses are approximately 100-fold greater than the maximum recommended human clinical IT dose and resulted in plasma exposure multiples of approximately 7 and 22, respectively (Table 3). As observed in the acute IT toxicity studies, findings were consistent with the expected pharmacological activity of ziconotide as a blocker of neuronal N-type VSCCs in the CNS (Table 4). Rats exhibited frequent tremoring, alopecia, chromodacryorrhea, and stiff tail at doses ≥0.15 μg/kg per hour, with no treatment-related effects on clinical pathology or neurohistopathology noted. A functional observation battery revealed gait abnormalities and impaired mobility in all groups, including controls, likely due to injury caused by spinal pressure from the implanted catheter. The NOAEL for general toxicity in rats, excluding expected CNS pharmacological effects, was determined to be > 1.5 μg/kg per hour. In dogs, CNS effects were noted in all treated groups, and severity increased with dose. At doses ≥0.10 μg/kg per hour, dogs exhibited dilated pupils, ataxia, and infrequent tremoring. At doses ≥0.30 μg/kg per hour, decreased activity, protruding nictitating membranes, ptosis, hypersensitivity, salivation, and frequent tremoring were observed. Four of 16 dogs were euthanized moribund due to the severity of CNS effects at doses ≥0.60 μg/kg per hour. There were no adverse toxic effects attributable to ziconotide on organ weights, clinical pathology, or histopathology at any dose, and CNS effects were reversible upon cessation of dosing. The MTD in dogs, due to the severity of expected CNS effects, was 0.3 μg/kg per hour IT ziconotide, approximately 23 times the recommended human IT dose; however, the NOAEL related to general toxicity and excluding adverse CNS pharmacological effects was > 1.2 μg/kg per hour. Neurohistopathological examinations in both rats and dogs revealed spinal cord compression with associated chronic inflammation in both control and ziconotide-treated animals, consisting of varying degrees of fibrosis, nerve fiber degeneration, and a variable admixture of lymphocytes, macrophages, and neutrophils, which was attributed to pressure exerted by the IT catheter.
In dogs receiving EPI administration of up to 3.0 μg/kg per hour ziconotide for 14 days or up to 1.2 μg/kg per hour ziconotide for 28 days, animals exhibited CNS effects similar to those seen with IT administration of ziconotide, including dose-related shaking, dilated pupils, partially closed eyes, and protruding nictitating membranes at doses ≥0.6 μg/kg per hour. Histologic preparations revealed catheter-related cord compression with chronic inflammation but no drug-related effects.
Ziconotide was administered IV for 14 days at doses up to 1.67 mg/kg per hour in rats and up to 1.04 mg/kg per hour in monkeys, yielding plasma concentrations about 1000-fold greater than levels from the highest IT dose in animals and 100,000 times greater than levels predicted from the maximum recommended human IT dose. Rats exhibited dose-related clinical signs of tremor, ptosis, cold-to-touch, dehydration, and hunched posture, primarily due to the pharmacological effects of ziconotide. Additional effects considered secondary to clinical signs included early reduction in food consumption, reduced body weight in males, reduced white blood cell counts (primarily lymphocytes and eosinophils) in all treated males, and a slightly increased incidence of thymic necrosis in the high-dose males and females. In monkeys, treatment-related clinical signs of tremor, ptosis, hypoactivity, rales, and diarrhea were observed at doses ≥0.21 mg/kg per hour, with slight transient reductions in heart rate and systolic blood pressure due to sympatholysis, possibly in combination with mast cell degranulation at the higher IV doses. There was no evidence of nonspecific target organ toxicity with continuous IV infusion of ziconotide for up to 14 days in rats and monkeys at systemic exposures greater than 85,000 times the predicted human exposure from IT administration.
Reproductive Toxicology Studies
A standard reproductive toxicology assessment of ziconotide, including fertility (segment I), teratology (segment II), and perinatal/postnatal studies (segment III), was completed. Ziconotide was administered by the IV infusion route, yielding plasma concentrations nearly 100,000-fold greater than those predicted from the maximum recommended human IT dose on a mg/kg basis, to ensure that peripheral reproductive organs were exposed to the compound.
In segment I fertility studies, continuous infusions of 42, 125, and 417 μg/kg per hour ziconotide were administered to male rats from 28 days prior to mating until necropsy (49 to 58 days total) and to female rats from 14 days prior to mating through gestation day 7 (22 to 29 days total). Parental toxicity was evident at all dose levels as decreased body weight and food consumption during the first week of dosing. High-dose males showed clinical signs of tremor and incoordination. There were no effects on male mating ability and fertility or on sperm motility, count, and morphology at any dose. Reductions in corpora lutea, implantation sites, and number of live fetuses were observed in females at the 417 μg/kg per hour dose. The NOAEL for male and female fertility was 417 and 125 μg/kg per hour, respectively.
Ziconotide was not teratogenic (segment II) in rats administered continuous IV infusions from gestation day 6 through day 15 at doses up to 1250 μg/kg per hour. Maternal toxicity, as indicated by decreased body weight gain and food consumption, was evident at all dose levels from 21 to 1250 μg/kg per hour. Clinical signs including ptosis, tremor, hunched posture, and decreased activity were generally observed at the higher doses. Fetal toxicity, as evidenced by reduced fetal weights and transient delayed ossification of the pubic bones, was observed at doses ≥625 μg/kg per hour. Postimplantation loss was significantly increased in rats at doses ≥63 μg/kg per hour. The NOAEL for embryo-fetal toxicity in rats was 21 μg/kg per hour. Ziconotide was not teratogenic in rabbits administered continuous IV infusions from gestation day 6 to day 18 at doses up to 208 μg/kg per hour. Maternal toxicity (as evidenced by decreased activity, body weight, and food consumption) and a slight reduction in fetal weight were evident at all ziconotide dose levels from 8 to 208 μg/kg per hour. At 208 μg/kg per hour, postimplantation loss was increased and the number of live fetuses reduced. Apart from a slight decrease in fetal weight, the NOAEL for embryo-fetal toxicity in rabbits was considered to be 42 μg/kg per hour.
In a segment III perinatal/postnatal study, continuous infusions of 42, 125, and 417 μg/kg per hour ziconotide were administered to female rats from gestation day 6 through weaning (postpartum days 20 to 23). Maternal (generation F0) toxicity, as indicated by clinical findings (e.g., ptosis, hunched posture, piloerection) and reduced body weight gain or food consumption, was evident at all dose levels. There were no effects on reproductive performance, parturition, or lactation. In F1 generation rats, there were no effects on survival, measures of physical development (other than slight decreases in postnatal body weight at the highest dose), behavior, or reproductive performance. Survival and early development of F2 generation pups were normal. The NOAEL for perinatal/postnatal effects in rats was considered to be 125 μg/kg per hour due to slight decreases in F1 generation postnatal body weights at the highest dose.
Mutagenicity and Carcinogenicity
A standard battery of genotoxicity assays was performed to assess the mutagenic potential of ziconotide. In vitro mutagenicity assays in the presence and absence of exogenous metabolic activation showed that ziconotide did not induce bacterial gene mutation at up to 5000 μg/plate in the Ames test, or mammalian gene mutation at up to 5000 μg/mL in the mouse lymphoma assay. In the in vivo mouse micronucleus assay, ziconotide did not increase micronuclei in bone marrow polychromatic erythrocytes after single IV injections of up to 32 mg/kg ziconotide. These data indicate that ziconotide is neither genotoxic nor clastogenic.
Chronic animal studies have not been performed to assess the carcinogenic potential of ziconotide. A nonclinical study was conducted to assess the potential transforming activity of ziconotide in cryopreserved SHE cells exposed to ziconotide for 7 days at concentrations ranging from 0.05 to 1.5 mg/ml. This system has been demonstrated to provide both the high sensitivity and high specificity needed to identify carcinogens and noncarcinogens in rodents. The resulting data showed that treatment with ziconotide at evaluable dose levels did not result in any statistically significant increase in transformation frequency. This negative result in the SHE cell transformation assay, when tested without exogenous metabolic activation, indicates a lack of carcinogenic potential for ziconotide.
Additional Toxicity Assessments
Additional toxicity studies and evaluations were conducted to assess (1) ziconotide-induced effects on cell proliferation and apoptosis in the CNS; (2) the local effects of IT ziconotide infusion on spinal cord histology in the rat and dog; (3) the effect of ziconotide-induced sympatholysis on cardiotoxicity in the rat; and (4) the immunogenic potential of ziconotide.
There was no evidence of drug-induced apoptosis or cell proliferation in the spinal cord or brain of ziconotide-treated dogs at IT-infused doses up to 0.240 μg/kg per hour for 28 days. Both apoptosis and cell proliferation labeling indices were very low, as expected in CNS tissue. In many animals, the labeling indices for both apoptosis and cell proliferation were 0. These data give assurance that cell cycle control is not disturbed by IT ziconotide treatment and provide additional evidence for the lack of carcinogenic potential associated with ziconotide.
The effects of IT ziconotide on spinal cord histology were evaluated in rats administered once-daily IT bolus injections of 1 to 10 μg ziconotide over a 4-day period. Histopathological changes in spinal cord tissue were present in rats from all groups, including controls. These changes (e.g., spinal cord compression, localized inflammation, dilatation of the periaxonal space) appeared to be induced by the presence of the IT catheter. Ziconotide appeared to exert a mild potentiation of the inflammatory reaction, as slightly more severe inflammatory reactions were noted in the high-dose drug-treated rats; however, the inflammatory reaction was qualitatively similar in all groups. Further, this finding was not corroborated in the subchronic toxicity studies of continuous ziconotide exposure in rats and dogs administered IT infusions at similar daily (24-h) doses for up to 28 days, suggesting that the rate of ziconotide introduction (fast bolus versus slow infusion) into the intrathecal space may play a minor role in potentiation of the inflammatory reaction to catheterization. Neurohistopathological examinations revealed no evidence of granuloma formation at the catheter tip in dogs treated by continuous IT infusion for up to 28 days at doses up to 1.2 μg/kg per hour (approximately 90-fold greater than the highest recommended human IT dose on a mg/kg basis).
At high parenteral doses of ziconotide in rats (e.g., ≥ 10 mg/kg), systemic hypotension occurs by a combination of blockade of sympathetic neurotransmission and mast cell degranulation, the latter of which is partially inhibited by pretreatment with histamine receptor antagonists (Bowersox et al. 1992). There were no differences in heart catecholamine levels in rats 72 h after treatment with a 90-min IV infusion of 15 mg/kg ziconotide when compared to untreated controls, indicating that ziconotide-induced sympatholysis does not deplete cardiac catecholamines. In addition, there were no cardiac-related histopathological findings in any nonclinical safety study of ziconotide.
Special toxicity studies evaluated the immunogenic potential of ziconotide in mice, rats, and guinea pigs. In mice, induction with IV ziconotide followed by ID challenge did not produce any detectable serum antiziconotide antibodies or any immediate or delayed response at the ID site. In rats induced IP and challenged IM or SC with ziconotide over 63 days, no serum antiziconotide antibodies were detected. Guinea pigs were susceptible to death by systemic anaphylaxis (35% of animals in two combined studies) with IP induction followed by IV challenge, but failed to exhibit a passive cutaneous anaphylactic response when induced ID with serum diluted from sensitized guinea pig or human and then challenged with IV ziconotide, suggesting a lack of serum antibodies in these samples.
DISCUSSION
Ziconotide was evaluated in a comprehensive program of nonclinical studies, and the resulting safety profile was consistent with the drug’s pharmacological action as a blocker of N-type VSCCs. Ziconotide-induced clinical and behavioral CNS effects in rats, dogs, and monkeys consisted primarily of frequent tremoring, shaking behavior, ataxia, and hyperreactivity and are predictive of human adverse events (e.g., abnormal gait, ataxia, hypertonia, tremor) reported in clinical evaluations. Behavioral effects were generally transient and reversible upon cessation of treatment and were not associated with any target organ toxicity. Other than a slightly increased incidence of thymic necrosis that was considered secondary to the stress of CNS pharmacology at the highest IV dose in rats, there were no treatment-related gross or histopathological changes at doses and systemic exposures that were orders of magnitude greater than human exposures during recommended clinical use. Detailed neurohistopathological examinations following continuous IT infusions lasting up to 28 days revealed only spinal cord compression with associated chronic inflammation in control and ziconotide-treated rats and dogs, which was attributed to pressure exerted by the IT catheter.
In reproductive toxicity studies, parental toxicity, as indicated by decreased body weight gain and food consumption, was evident at all dose levels, with associated clinical signs generally observed at the higher doses. There were no ziconotide-related adverse effects on male rat fertility at exposures > 10,000 times the plasma levels produced by the maximum recommended human IT dose, whereas effects on female reproductive parameters (i.e., reductions in corpora lutea, implant sites, and/or number of live fetuses) were noted at >750 times human plasma levels in rats and/or rabbits. Ziconotide was nonteratogenic in rats and rabbits. Fetal toxicity, as evidenced by reduced body weight and transient delayed ossification of the pubic bones, was observed in gestation day 20 rat fetuses at exposures > 10,000-fold human plasma levels. Delayed fetal skeletal ossification is commonly observed in teratology studies, may occur secondary to maternal toxicity (i.e., decreased body weight and/or food consumption), and is often transitory (Breslin, Kirk, and Zimmer 1989; Hanley et al. 1987; Kirk et al. 1995; Rodwell et al. 1989). Rat postnatal evaluations (Skov et al. 2004) further support the transient nature of this finding, because no effects were observed on the survival, skeletal development, or behavior of F1 generation pups. For a drug that is administered IT and produces exceptionally low observed plasma concentrations, the high systemic exposures needed to elicit these effects in animals indicate no significant reproductive risk to humans.
Ziconotide displayed no mutagenic or carcinogenic potential and no propensity to induce cell proliferation or apoptosis in the surrounding CNS tissues. Local toxicity consisted of spinal cord compression with associated chronic inflammation attributed to pressure exerted by the IT catheter in control and ziconotide-treated rats and dogs. Dogs treated with IT ziconotide displayed no evidence of granuloma formation at the catheter tip (Shopp, Skov, and Yaksh 2004), as has been observed clinically with high-dose IT opioid therapy (Hassenbusch et al. 2004) and in a canine IT infusion model at clinically relevant morphine doses (Yaksh et al. 2003). Ziconotide demonstrated no propensity to interfere with cardiac action potentials in vitro, induce QT prolongation in vivo, or deplete cardiac catecholamine levels at sympatholytic doses in rats. Although systemic anaphylaxis was observed in guinea pigs, the absence of antibodies to ziconotide in animals and human patients indicates that the drug has low immunogenic potential.
In patients, ziconotide is administered IT as a continuous infusion using programmable electronic infusion pumps. From clinical experience, a range of set doses was established for which the narrow therapeutic window defining efficacy versus adverse pharmacology was determined by reportable patient response to CNS effects. Infusion is recommended to begin at 0.1 μg/h (0.0017 μg/kg per hour in a 60-kg patient) or less, and the dose is then titrated to analgesia or intolerable adverse events, with a maximum recommended dose of 0.8 μg/h (0.013 μg/kg per hour in a 60-kg patient). The majority of clinical patients receiving ziconotide by IT infusion at doses up to 0.8 μg/h for as long as 9 months have no measurable systemic ziconotide exposure, and there is no evidence of ziconotide accumulation in those patients with quantifiable levels. Intolerable (moribund) adverse events in animals due to exaggerated CNS pharmacological activity of ziconotide occurred at IT doses at least 46 times the maximum recommended human dose, were predictive of human adverse events, and were reversible with cessation of treatment. Ziconotide was not associated with any CNS or peripheral target organ toxicity at IT doses greater than 92 times the maximum human dose, which resulted in plasma levels at least sevenfold greater than those predicted in humans. With IV administration of ziconotide in animals, systemic exposure is dramatically increased, thus ensuring exposure of peripheral organs and providing a more definitive evaluation of nonspecific drug toxicity. Safety ratios for general and reproductive toxicity in animals following subchronic IV infusions represent multiples that are orders of magnitude greater than the human exposure from IT administration (Table 5). In summary, the nonclinical safety assessment of ziconotide suggests no significant toxicological risk associated with the recommended IT treatment in humans.