Summary Basis of Decision for Azilect ®

3.2.1  Pharmacodynamics

The primary pharmacological activity of rasagiline was investigated in various in vitro and in vivo systems. Rasagiline was found to be a highly potent, selective, irreversible monoamine oxidase type B (MAO-B) inhibitor. Inhibition of the MAO-B enzyme causes an increase in extracellular levels of dopamine in the striatum, which in turn increases dopaminergic activity. It is likely that this is the main mechanism by which rasagiline improves models of dopaminergic motor dysfunction. Rasagiline does not inhibit other regulatory enzymes or bind to other receptors.

The ability of rasagiline to restore normal motor activities was studied in the rat for drug- and hypoxia-induced dopaminergic dysfunctions. Effects on learning and memory were also tested using models of passive avoidance, active avoidance, and the Morris water maze. Rasagiline treatment was found to improve passive avoidance, learning, and memory in both postnatal and adult rats.

In addition, results from non-clinical studies have indicated that rasagiline improved cognition in animal models and was neuroprotective in both in vitro and in vivo models. The neuroprotective activity of rasagiline is independent of its MAO inhibitory activity, and may be mediated by other mechanisms. One such mechanism may be the influence of aminoindan, a major metabolite of rasagiline. Aminoindan is not a MAO inhibitor, but in several animal models of impaired motor and cognitive functions it was demonstrated to ameliorate dopaminergic motor dysfunction and hypoxia-induced cognitive deficits, thus potentially contributing to rasagiline's beneficial effect.

The neuroprotective effects of rasagiline were compared to the effect of selegiline, another MAO-B inhibitor, using in vivo and in vitro models. Rasagiline exhibited more potent neuroprotective and neurotrophic effects compared to selegiline on dopaminergic neurons in primary mesencephalic cultures. Rasagiline also exhibited some protective effects against neurotoxin MPTP, which can induce parkinsonism in human and non-human primates. The required potency of a drug to prevent MPTP-induced neurotoxicity is indicative of its MAO-B inhibitory activity. In cell cultures, the drug effects were interpreted as inhibition of apoptotic cell death which was not due to the MAO inhibitory activity of these compounds. In vivo, the neuroprotective action of rasagiline was clearly evident and increased animal survival rates were noted. Increased neuronal survival was confirmed by histological, immunocytochemical, and MRI analyses. These results are currently considered preliminary however, and further investigation is required to draw definitive conclusions and establish clinical relevance.

The effects of rasagiline on the cardiovascular and central nervous system in animal models were investigated, however the safety pharmacological studies are considered to be incomplete. According to ICH guidelines, the core testing battery should also include analysis of the respiratory system and if there is a base of concern, other physiological systems such as the renal and gastrointestinal systems, should also be investigated.

No overt changes in arterial blood pressure or heart rates were seen in any of the four animals tested. There were no large or sustained changes in ECG waveform intervals or morphology. Rasagiline administered as a single dose up to 2 mg/kg produced no effect on motor activity, behaviour, or response to stimuli. At doses at least 10-fold higher than the proposed human therapeutic dose, no observable adverse effects in the cardiovascular or central nervous system were found.

ECG evaluation in free-moving dogs by telemetry, using a single dose (3 mg/kg), gave negative findings with respect to QT interval prolongation. Some increases in QT intervals were observed in toxicity studies with dogs treated with levodopa/carbidopa with or without rasagiline; however, there was no prolongation in the group treated with rasagiline alone. A study to test the potential of rasagiline to interact with potassium ion channels (hERG test) was performed and it did not indicate any potential to induce QT prolongation.

3.2.2  Pharmacokinetics

The pharmacokinetic studies indicated that absorption of rasagiline was rapid and substantial in all animal species studied. Despite these findings, oral bioavailability was relatively low, mainly due to intestinal and liver first-pass metabolism. The systemic exposure to rasagiline was evaluated in toxicokinetic studies, as part of each toxicity study in three animal species. Over the range of doses used in toxicity studies, exposure to rasagiline increased linearly with dose at low doses (up to 5 mg/kg/day) and non-linearly at high doses (above 5 mg/kg). Plasma protein binding was relatively high in all species tested but is not considered to affect pharmacological action. Volume of distribution was also high, indicating that the drug is distributed to all tissues.

The metabolism of rasagiline was investigated in the mouse, rat, and dog after single oral and intravenous doses. Rasagiline is primarily biotransformed by CYP450 enzymes to various metabolites. Similar metabolites were identified in all species studied.

In all species tested, rasagiline was excreted in the urine, mainly as polar metabolites. Excretion was rapid, and only small residues were noted 24-hours post-dosing. This may be related to the irreversible binding of rasagiline at MAO sites.

It should be emphasized that although rasagiline did not accumulate in the tissues in rats, the drug was retained at higher concentrations in pigmented cells, mainly in the eyes and skin, than in non-pigmented cells. It is possible that substantial accumulation of rasagiline in melanocytes may cause some interaction with endogenous and/or exogenous compounds that have adverse effects, such as induction of cell proliferation, genotoxic effects, or inhibition of the apoptotic process.

3.2.3  Toxicology

All toxicity studies used acceptable, standard protocols designed according to ICH Guidelines. Most of the studies were carried out in accordance with Good Laboratory Practices.

Short-Term

General toxicity studies did not reveal any remarkable or unexpected toxicity findings in any of the animal species tested. Rasagiline demonstrated low acute toxicity; acute toxicity is not anticipated at the human therapeutic dose or even at much higher doses. The lack of toxicity in many of the subchronic studies suggests that the dosages used were too low, although the toxicokinetic data indicate adequate exposure. The reduction of body weight in treated animals as observed in all species tested, clinical observations, and clinical pathological and histopathological findings, could be associated with the dopaminergic activity of rasagiline.

Signs of immunotoxic effects in both rat and dog studies were noted, primarily atrophy of lymphoid tissues at the highest tested dose levels. These were not followed up by functional testing. The sponsor considered that the changes in the lymphoid tissues were related to reduced body weight gains.

Genotoxicity

The genotoxic potential of rasagiline was evaluated in a number of in vitro and in vivo tests that examined bacterial and mammalian gene mutations, chromosomal aberrations in mammalian cell cultures, and micronucleus induction in the bone marrow of mice.

Rasagiline produced negative results in the bacterial mutation assays and in the mouse micronucleus assay. However, there were positive responses in the chromosomal aberration assay in lymphocytes and the mouse lymphoma cell assay for gene mutation, which together suggest that rasagiline has genotoxic potential.

The results obtained from experiments where changes were made to modify the culture medium were difficult to interpret.

Carcinogenicity

Carcinogenicity bioassays were conducted in rats at oral (gavage) doses of 0.3, 1, and 3 mg/kg (males) and 0.5, 2, 5, and 17 mg/kg (females) and mice at oral (gavage) doses of 1. 15, and 45 mg/kg. In rats, there was no increase in tumours at any dose tested. The sponsor's overall conclusion was that rasagiline has equivocal carcinogenic activity, not related to drug effects, despite the fact that positive results were obtained in the mouse study (the exposure ratio of mice at 15 mg/kg/day relative to humans treated with 1 mg/day is greater than 100-fold higher).

The sponsor argued that the tumours observed in mice at the two anatomical sites (lung and Harderian gland) were common tumours and thus the trend test for tumour rates should only be considered statistically significant at a p-value of equal or less than 0.005. This is based on a decision criterion developed by the FDA in evaluating common tumour incidences (Draft Guidance for Industry Statistical Aspects of the Design, Analysis, and Interpretation of Chronic Rodent Carcinogenicity Studies of Pharmaceutical, FDA, 2001). Based on the FDA's criterion for significance in incidence rates, the sponsor claimed that only the combined alveolar/bronchial adenoma/carcinoma of the lung in male mice (with a p = 0.002) should be considered a significant positive trend while the trend tests for all other sites, including tumour incidences in female mice of the lung (p = 0.0113) and Harderian tumours in male mice (p = 0.0194), should all be considered non-significant.

It should be noted that although the FDA's draft document regarding the statistical analysis of tumour data has been in existence for more than three years, it has not yet been finalized. Furthermore, in other carcinogenicity studies, notably in the studies conducted by the National Toxicology Program (NTP), different p-values for accepting significance in trend tests are being used (e.g., p-value of 0.05 and 0.01 for rare and common tumour incidences, respectively) than those suggested in the FDA's draft document.

Another important point to consider in demonstrating significant tumour rates is that, in addition to the trend test, pairwise comparison of tumour incidences between the control and a treated group should also be used. This is generally recognised as a test for significance and is also advocated in the FDA's document. The tumour rates in pairwise comparison are considered to be significant at p-value of 0.01 and 0.05 for common and rare tumours, respectively. By comparing the high dose and control groups in the mouse study a significant increase in tumour rates for adenomas/carcinomas in the lung of males can be demonstrated (p = 0.0046 for 45 mg/kg/d dose and p = 0.036 for 15 mg/kg/d dose).

As both statistical analyses (trend test and pairwise comparison), resulted in significant positive tumour incidences for lung tumours in male mice, it is concluded that there is clear potential for carcinogenic activity of rasagiline at high multiples of the clinical dose.

The increased incidences of Harderian gland adenomas/carcinoma with a p-value of 0.0131 when the high dose vs. control groups were compared, provide additional support to consider rasagiline carcinogenic in male mice.

In female mice, a significant increase in the adenoma/carcinoma incidences of the lung (p = 0.0237) was shown by pairwise comparison of the high dose vs. control groups. This result, together with a positive trend test for this type of tumour, provides some evidence that rasagiline also has carcinogenic activity in female mice.

Further information regarding carcinogenicity risks in humans can be found in section 3.3.5 Assessment of Melanoma Risk.

Reproduction and Development

Rasagiline administered to rats and rabbits of both sexes did not produce adverse effects on fertility. The no observed adverse effect level (NOAEL) was 3 mg/kg/day and 7 mg/kg/day for rats and rabbits, respectively. These doses represent a more than 50-fold higher dose than the human therapeutic dose, based on AUC ratios.

In combination treatment with levodopa (LD) and carbidopa (CD), rasagiline induced maternal toxicity in rabbits at doses above 0.1 mg/kg/day. Exposure to a 0.1/80/20 mg/kg/day combination (LD/CD/rasagiline) dose was similar to the exposure received by humans at the 1 mg/day clinical dose. No teratogenic effects were seen in rabbits. In rats however, exposure to combination treatment at a dose of 1/80/20 mg/kg/day, produced a high incidence of skeletal malformations in the form of wavy ribs. The sponsor argued that although the wavy ribs appeared to be treatment-related, these are reversible effects found in rodents only. Nevertheless, wavy ribs are considered to be a minor skeletal malformation. From these studies it is uncertain whether rasagiline alone is able to induce this malformation, or if their formation is due to the presence of LD or CD.

Based on these findings, it is recommended that rasagiline not be given to women during pregnancy. A warning has been included in the Product Monograph.

3.2.4 Summary and Conclusion

Based on studies of the pharmacological activity of rasagiline, the data demonstrate conclusively that it is an effective inhibitor of MAO-B, and at therapeutic doses it produces relatively minimal inhibitory activity of MAO-A. The interpretation of data

related to the neuroprotective effects of rasagiline is considered too speculative, and further investigations are needed to draw definitive conclusions. More importantly, further investigations are required to establish clinical relevance.

The most critical finding in the non-clinical assessment of this submission is the carcinogenic activity of rasagiline in mice. Due to the positive responses in some of the genotoxicity tests, including chromosomal aberrations and gene mutation in cultured mammalian cells, the carcinogenic activity of rasagiline is thought most likely to occur through a genotoxic mechanism. Together with the positive results of the mouse carcinogenicity study, the non-clinical data suggest that rasagiline could pose a potential carcinogenic risk to humans. Further analysis of the carcinogenic potential of rasagiline can be found in section 3.3.5 Assessment of Melanoma Risk.