Summary Basis of Decision for Champix

A conventional non-clinical testing program in support of Champix followed ICH and FDA guidelines. Studies included pharmacodynamic (PD), pharmacokinetic (PK), and toxicologic studies in mice, rats, ferrets, rabbits, dogs, and monkeys. All pivotal studies were Good Laboratory Practice (GLP) compliant.

3.2.1 Pharmacodynamics

In vitro receptor binding studies demonstrated that of the four nicotinic receptor sub-types tested (α4β2, αβ4, α7, or α1β1γδ) varenicline specifically binds with high affinity to the human and rat α4β2 neuronal nicotinic acetylcholine receptor sub-type. In vitro functional studies showed that varenicline acts as a partial agonist, at the α4β2 subtype. It also acts as a partial agonist at the α3β4 nicotinic subtypes, but as a full agonist at the α7 nicotinic subtype (central pre-synaptic receptor), for which varenicline has very low binding affinity. The α4β2 subtype is believed to mediate the behaviorally reinforcing effects of nicotine. Nicotine is known to cause complex CNS behavioural, neuromuscular, endocrine, renal, metabolic and cardiovascular effects in humans, due to variety of peripheral and central receptor subtypes. It is not possible to know the extent to which varenicline is similar in effects, but varenicline did not have significant interactions with the peripheral α3β4 or α1β1γδ nicotinic subtypes.

Varenicline was shown to bind moderately to serotonergic 5HT3 receptor, known to be associated with nausea/vomiting and changes to bowel motility.

No significant interaction was shown with 56 other neurotransmitter receptors, ion channels, and neurotransmitter uptake sites.

Functional activity in vitro was demonstrated in oocytes and human embryonic kidney (HEK) cells expressing α4β2 neuronal nicotinic receptors in which maximal currents obtained were less than half those for nicotine, consistent with a partial agonist action. In both in vitro and in vivo models of mesolimbic dopamine function, activation by varenicline was about 50% of that induced by nicotine, consistent with a partial agonist mechanism of action.

As expected, when varenicline was substituted for nicotine in drug-discrimination studies, there was a dose-dependent generalization of the nicotine response, with the rats responding to the maximum dose of 1 mg/kg varenicline as if it were nicotine. In self-administration studies, varenicline reduced nicotine intake and the rate at which nicotine was self-administered by rats. Varenicline was also less reinforcing than nicotine, since rats worked harder for nicotine than for varenicline. Furthermore, varenicline (unlike nicotine, as determined from the literature) did not produce detectable effects in a rat model of withdrawal.

The behavioural and urine electrolyte effects seen in the safety pharmacology studies were mild and generally at high dose/exposures and therefore not suggestive of significant risk for humans. In vitro and in vivo non-clinical experiments demonstrated that varenicline performs as expected for a partial agonist of the α4β2 nicotinic receptor subtype, and the data are consistent with utility as a smoking-cessation treatment. 

3.2.2 Pharmacokinetics

Overall, varenicline is characterized by high absorption, negligible first-pass effect, moderate distribution across many tissues, moderate half-life and extensive renal excretion.

Absorption

Drug exposure levels determined by the PK parameters C_max and AUC in single and multiple dose studies were dose-dependent in all species (mice, rats, ferrets, rabbits, dogs, and monkeys) with no significant gender differences.

Distribution

Plasma protein binding was low in all species tested with percent-bound values of 18%, 45%, 19%, 41%, and 20% in mice, rats, dogs, monkeys, and humans, respectively. In rats, oral administration of radiolabelled varenicline revealed drug-related material distributed throughout all tissues and organs in the body, except for lens and vitreous humor of the eye, with maximal radioactivity concentrations achieved at the first sampling point (1 hour post-dosing) in the gastrointestinal tract, and in other eye tissues. Varenicline demonstrated an increased but reversible affinity for melanin-containing tissues with drug-related material present in the eye and skin at 168 hours post-dose. Radiolabelled concentrations in pigmented skin were much higher (up to 18 fold) than in non-pigmented skin and showed accumulation from days 1 to 3 of dosing. In the reproductive toxicology studies, varenicline was detected in fetal rat and rabbit serum and was also detected in the serum of nursing offspring of varenicline-dosed rats.

Metabolism

A total of 13 metabolites of varenicline were identified in the circulation and urine of all laboratory animals and healthy human subjects tested, and all were products of oxidation and conjugation pathways. In mice, rats, monkeys, and humans the drug-related material in circulation and excreta was primarily unchanged drug (75-93%), indicating that metabolism is not a primary route of varenicline clearance in these species. No metabolite was present in excreta at more than 4.6% of the dose in any species. All metabolites observed in humans were observed in one or more species tested. In rabbits, two metabolites were present in higher concentrations in circulation than varenicline.

Excretion

Varenicline excretion in mice, rats, monkeys, and humans was mainly via the urine. In vitro data suggest that the renal excretion of varenicline occurs by both passive filtration and an active transport process (renal transporter OCT-2). The percentages of total dose recovered in the urine were 83%, 68%, 75%, and 87%, respectively; while the amounts in feces were 11%, 22%, 6.5%, and 0.9%, respectively.

Drug Interactions

Drugs cleared by, or which affect, cytochrome P450 enzymes

In vitro studies demonstrated that varenicline does not inhibit cytochrome P450 enzymes. The P450 enzymes tested for inhibition were: 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and 3A4/5. Also, in human hepatocytes in vitro, varenicline did not induce the activity of cytochrome P450 enzymes 1A2 and 3A4. Therefore, varenicline is unlikely to alter the pharmacokinetics of compounds that are primarily metabolised by cytochrome P450 enzymes. Furthermore, since metabolism of varenicline represents less than 10% of its clearance, drugs known to affect the cytochrome P450 system are unlikely to alter the pharmacokinetics of varenicline and therefore a dose adjustment of Champix should not be required for these types of drugs.

Drugs cleared by, or which affect, renal secretion

In vitro studies demonstrated that varenicline does not inhibit human renal transport proteins at therapeutic concentrations. Therefore, drugs that are cleared by renal secretion (e.g., metformin) are unlikely to be affected by varenicline.

In vitro studies demonstrated the active renal secretion of varenicline is mediated by the human organic cation transporter, OCT2. Although varenicline is a substrate and weak inhibitor of the human organic cation transporter (hOCT-2), and cimetidine and probenacid (as hOCT-2 inhibitors) reduce the clearance of varenicline, the drug interaction effects between these drugs were modest and occurred at much higher drug concentrations than achieved in humans. In cases of severe renal impairment, it is possible that the increase in systemic exposure to varenicline with an inhibitor of OCT2 such as cimetidine, might be clinically meaningful. This is reflected in the Product Monograph.

3.2.3 Toxicology

Single-Dose Toxicity

Acute oral studies were conducted in rat (30, 100, 200, 300 mg/kg), and monkey (3 mg/kg) and intravenous studies in monkeys (0.08 - 0.3 mg/kg).

Rats that were administered oral (PO ) doses of varenicline at ≥ 200 mg/kg (approximately 235-300 fold above the expected human exposure) were uncoordinated and developed convulsions and tremors with decreased body weight and loose stools. There was one death at 300 mg/kg. Effects observed were generally rapid in onset (with 2.5 hours post- dosing) and transient (reversed by the end of the 14-day observation period. The single-dose No Observed Adverse Effect Level (NOAEL) was ≤ 100 mg/kg corresponding to an exposure approximately 100-120 fold above the expected human exposure.

In monkeys at a dose of 3 mg/kg (approximately 2-4 fold above the expected human exposure) the findings were also in the GI (emesis), and central nervous system (tremours), along with decreased activity, and decreased food consumption. This dose was also associated with decreased heart rate and QT interval, and increased PRQ interval and P wave width on the electrocardiogram (ECG). All of the findings were reversed by the next day. The NOAEL was 0.2 mg/kg corresponding to an exposure of approximately 1.3 fold above the expected human exposure. ECG alterations were not seen in any of the subsequent multiple repeat-dose monkey studies or the cardiac safety pharmacology study, therefore it would be reasonable to suggest based on this data that there is not significant cardiac risk in humans.

Multiple-Dose Toxicity

Eleven oral, repeat-dose studies were completed in rats and monkeys. Rat studies included 6-week, 3- and 6-month studies. Monkey studies included 6-week, 3- and 9-month studies.

The multiple-dose NOAEL in rats was 10 mg/kg/day in the 3- and 6-month studies, which corresponds to approximately 40-65 fold above the maximum recommended human exposure.

Decreased in food consumption were seen in rats at doses of 10 mg/kg, likely due to decreases in gastric motility. Rats dosed with varenicline at 30 mg/kg (approximately 75-140 fold above the expected human exposure) had clinical signs related to the nervous system and gastrointestinal tract: decreased food consumption, decreased body weight, loose stool, dehydration, microscopic and macroscopic evidence of intestinal dilatation (ileum, jejunum, cecum, colon). These changes likely reflect the pharmacologic effects of varenicline on gastric emptying and intestinal motility in the rats. Decreases in body temperature were also seen, with body temperature perturbation a well known effect of nicotine.

Slight increases in alanine aminotransferase (ALT), alkaline phosphatase (ALP), and total bilirubin, suggestive of hepatobilary alteration without microscopic hepatic findings, were reported in rats dosed with varenicline at ≥ 10 mg/kg. The significance of the reported increases in absolute liver weights in females at ≥ 10 mg/kg is uncertain as similar increases were not observed in previous studies. Hepatic changes occurred at the highest dose of the 10-day and 6-week studies (i.e., 100 mg/kg/day and 30 mg/kg/day, respectively), but were not seen in the 6-month study. These changes included: small increases in ALT, ALP, aspartate amino transferase (AST)), and/or total bilirubin, and in some cases, microscopic evidence of hepatocellular necrosis. The changes are noted in the Product Monograph.

The multiple-dose NOAEL in monkeys was 0.2 mg/kg/day in the 3- and 6-months studies, which corresponds to approximately 3 fold above the maximum recommended human exposure.

Monkeys dosed with varenicline at 1.2 mg/kg (approximately 10-12 fold above the maximum recommended human exposure) were all sacrificed prematurely due to the severity of decreased food consumption, decreased body weight, and severe/prolonged dehydration. One female monkey at this dose was found dead of megacolon, secondary to colonic torsion and ischemic necrosis. Megacolon is an uncommon spontaneous finding in monkeys. This finding was likely secondary to the gastrointestinal dysfunction (loose stools) that was evident prior to and during treatment, although a drug treatment effect can not be excluded.

Slight increases in fibrinogen in male and female monkeys were observed at 1.2 mg/kg. Increased fibrin was also previously observed in monkeys at 1.2 mg/kg (0.6 BID) with concurrent increases in neutrophils and monocytes. This is suggestive of inflammation, but is of uncertain significance. Decreases in body temperature were also seen at this dose, and at the lower dose of 0.6 mg/kg/day.

Genotoxicity

Varenicline was non-genotoxic in a battery of in vitro and in vivo genetic toxicity tests.

Carcinogenicity

Two pivotal 2-year studies investigated the carcinogenic potential of varenicline, one in mice and one in rats. The studies were GLP compliant.

The mice studies showed no evidence of carcinogenicity associated with the oral administration of varenicline tartrate. Mice were administered doses up to 20 mg/kg. Exposure at 20 mg/kg was approximately 120 fold above the expected maximum human exposure.

The incidence of neoplasms associated with varenicline was increased in the rat studies, however the results were not statistically significant. Rats were administered doses of 1, 5, and 15 mg/kg. One male at 5 mg/kg had a benign hibernoma (tumor of brown fat in the mediastinum) and 2 males at 15 mg/kg had malignant hibernoma which was the cause of death in both animals. Exposure at 1, 5, and 15 mg/kg represented an exposure of approximately 15, 30, and 50 fold, respectively, above the expected maximum human exposure.

Hibernoma is a rare neoplasm and a review of a few cases in control animals from published reports was discussed. Discussions included the physiologically relevant differences between rodent and human brown fat. It was proposed that because of the physiological differences between human and rodent brown fat, the human risk is theoretical and probably non-existent. Varenicline is not genotoxic. Also, there was no abnormal histopathological evidence of brown fat proliferation at doses of 30 mg/kg in the 3- and 6-month rat studies. Hibernomas were not observed in female rats or male and female mice. The weight of evidence suggests that the risk is extremely low but the relevance to humans is uncertain. The presence of the benign and malignant hibernomas is acknowledged as treatment-related in the Product Monograph.

Reproductive and Developmental Toxicity

Five studies designed to examine the reproductive toxicity of varenicline were conducted in rats (Segments I, II, and III) and rabbits (Segment II).

Rats in the Segment I studies were administered doses of 0.3, 3, and 15 mg/kg. There were no effects on fertility and the male and female reproductive parameters, and no effects on early embryonic development. Therefore, the fertility and early embryonic NOAEL was 15 mg/kg corresponding to an exposure approximately 37-43 fold above the expected human exposure. The NOAEL for the F₀ males and females was 3 mg/kg, approximately 16 fold above the expected human exposure.

Segment II studies were conducted in rats and rabbits. Rabbits that were administered doses of 30 mg/kg demonstrated significant reductions in fetal and placental weights. The rabbit F₀ maternal toxicity and F₁ fetotoxicity NOAEL were both 10 mg/kg, corresponding to an exposure approximately 15-28 fold above the expected human exposure. In rats and rabbits, the maternal and fetal serum varenicline concentrations increased as the doses increased. Fetal concentrations were higher than the maternal concentrations confirming that the fetus was exposed in utero following administration of varenicline to the dam. The Product Monograph acknowledges this fetotoxicity and recommends that pregnant women not use varenicline.

Segment III studies were completed in rats dosed at 0, 0.3, 3, and 15 mg/kg and there was no effect on the reproductive parameters. The NOAEL for F₀ maternal reproductive toxicity is 15 mg/kg/day, corresponding to an exposure approximately 39 fold above the expected human exposure. Body weights for the F₁ male and female offspring in the 15 mg/kg/day group were reduced and some alterations were noted in functional development. A treatment-related increase in the mean amplitude of the auditory startle response (ASR) was noted in the 15 mg/kg/day group males and considered to be treatment-related. The rat studies also noted a reduction in fertility of the F₁ offspring in the 15 mg/kg/day group. The Product Monograph acknowledges these post natal developmental findings and recommends that pregnant and nursing women not use varenicline.

Serum-drug concentrations of the F₀ females and F₁ pups indicate that pups are exposed to varenicline through milk after oral administration of varenicline to dams. The Product Monograph indicates nursing women should discontinue nursing or discontinue taking varenicline.

Local Toxicity

Varenicline was classified as a mild ocular irritant and a mild skin irritant in rabbits. However, no deaths or evidence of systemic toxicity following a single 24-hour semi-occluded dermal application to intact skin at a dose level of 2000 mg/kg in rats were observed. Varenicline was non-sensitizing following dermal challenge in guinea pigs. Oral administration of varenicline at 100 mg/kg/day for three days did not cause phototoxicity in rats.

3.2.4 Summary and Conclusion

Overall, the pharmacology and toxicology studies support the use of Champix for the proposed indication. Treatment-related clinical signs and alterations on food consumption and body weight in all animal species which could be anticipated based on the pharmacology of varenicline generally occurred at dose levels and corresponding systemic exposures above the anticipated exposure in humans. Treatment-related pathologic findings in the gastrointestinal tract of rats occurred at doses resulting in exposure at significant levels above the anticipated human exposure. The carcinogenicity studies in mice did not indicate carcinogenic potential. However, the occurrence of hibernomas in rats is discussed in the Product Monograph as the relevance to humans is uncertain. Adverse effects were observed in the animal reproduction studies. The Product Monograph indicates that Champix is not recommended in pregnant women and recommends discontinuation of the drug or nursing in nursing women.