Summary Basis of Decision for Xeljanz

Clinical Pharmacology

Xeljanz contains tofacitinib, a potent, selective inhibitor of the Janus Kinase (JAK) family of kinases with a high degree of selectivity against other kinases in the human genome. Inhibition of JAK1 and JAK3 blocks specific cytokines that are integral to lymphocyte activation, proliferation, and function, resulting in modulation of multiple aspects of the immune response.

The clinical pharmacology included reports on the human pharmacodynamic and pharmacokinetic studies. The clinical pharmacological data support the use of Xeljanz for the recommended indication.

The pharmacokinetics of tofacitinib were altered in subjects that had moderate hepatic impairment, as well as in subjects with moderate and severe renal impairment; hence dosage adjustment is required in these patients.

Co-administration with potent immunosuppressants such as cyclosporine [P‑glycoprotein (P-gp) inhibitor] and tacrolimus is not recommended based on the risk of added immunosuppression. Co‑administration of cytochrome P450 (CYP) 3A4 and CYP2C19 inhibitors (ketoconazole and fluconazole) increased the exposure of tofacitinib and requires dosage adjustment, whereas a CYP3A4 and P-gp inducer (rifampin) decreased the exposure of tofacitinib leading to reduced efficacy/clinical benefit.

For further details, please refer to the Xeljanz Product Monograph, approved by Health Canada and available through the Drug Product Database.

Clinical Efficacy

The clinical efficacy of Xeljanz was assessed in five randomized, double-blind, multicentre, parallel‑group, placebo‑controlled, Phase III studies in patients ≥18 years of age with active rheumatoid arthritis (RA) diagnosed according to American College of Rheumatology (ACR) criteria. Patients had ≥6 tender and ≥6 swollen joints at randomization (≥4 swollen and ≥4 tender joints for Study II). The number of patients (n) randomized to receive Xeljanz or placebo was 610, 792, 717, 797, and 399 for Studies I, II, III, IV, and V, respectively. Baseline demographics were generally similar among the treatment groups in each study and comparable between the studies. The mean age ranged from 50 to 56 years. Most (80 to 87%) of the patients were female. With the exception of Study IV (46%), the majority (55% to 86%) of the patients in each study were white.

Xeljanz, 5 or 10 mg twice a day (BID), was administered as monotherapy in Study I, and in combination with non‑biologic disease-modifying anti-rheumatic drugs (DMARDs) in Study II in patients who had an inadequate response to DMARDs (non‑biologic or biologic). In Study III and Study IV, Xeljanz, 5 or 10 mg BID, was administered in combination with methotrexate (MTX) in patients who had an inadequate response to MTX. Adalimumab was used as an active control in Study III. In Study V, Xeljanz, 5 or 10 mg BID, was administered in combination with methotrexate (MTX) in patients who had inadequate efficacy or lack of tolerance to at least one approved tumour necrosis factor (TNF) inhibiting biologic agent.

The primary endpoints for Studies I and V were the proportion of patients who achieved American College of Rheumatology response criteria at the 20% improvement threshold (ACR20), the mean change from baseline in Health Assessment Questionnaire – Disability Index (HAQ-DI) and the proportion of patients who achieved a score <2.6 in the Disease Activity Score [DAS28-4(ESR)], at Month 3. The primary endpoints for Studies II, III, and IV were the proportion of patients who achieved an ACR20 response at Month 6, the mean change from baseline in HAQ-DI at Month 3, and the proportion of patients who achieved DAS28-4(ESR) <2.6 at Month 6. In Study IV, there was one additional co-primary endpoint of joint preservation as measured by the change in the van der Heijde modified Total Sharp score (mTSS).

Clinical Response

The ACR20 response criteria were assessed in all Phase III studies at either Month 3 (Studies I and V) or Month 6 (Studies II, III, IV). For the ACR20 analysis, the studies were powered to detect a clinically meaningful difference in response rates of at least 20% (with the placebo response at 30%).

Both of the Xeljanz doses, 5 mg and 10 mg BID, resulted in statistically significant and clinically meaningful differences from placebo in all of the Phase III studies on the ACR20 response. In comparison to placebo, the results were similar across most of the studies, with the exception of Study V which was conducted in patients who had previously failed a tumour necrosis factor inhibitor. In Study V, patients treated with Xeljanz 5 mg BID or 10 mg BID displayed slightly lower response rates, than the other studies. In addition, the 5 mg BID dose missed the 20% difference from placebo but was still statistically significant. The pooled Phase II/III data revealed that although there was a small dose response in the ACR20 with monotherapy (61% and 69% in the 5 mg and 10 mg groups, respectively), the difference between the 5 mg and 10 mg BID dose was minimal and clinically insignificant in the add-on to methotrexate studies (57% and 61% in the 5 mg and 10 mg BID dose groups, respectively). The results of Study III demonstrated that both Xeljanz dose groups had slightly higher ACR20 responses than the active comparator, adalimumab, but again with minimal and clinically insignificant difference between the 5 mg and 10 mg BID doses (51%, 52%, and 47%, in the Xeljanz 5 mg, 10 mg, and adalimumab dose groups, respectively).

The Health Assessment Questionnaire - Disability Index

The HAQ-DI assesses the degree of difficulty a patient has had in physical functioning using 8 domains of daily living activities; dressing and grooming, arising, eating, walking, hygiene, reach, grip, and other activities. Both doses of Xeljanz, 5 mg and 10 mg BID, led to statistically significant and clinically important differences in the HAQ-DI score in all of the Phase III studies. The placebo response was also very close to a clinically meaningful difference in some of the studies. Regardless, there were statistically significant differences between placebo and both doses of Xeljanz in all of the Phase III studies, supporting the efficacy of Xeljanz on this efficacy endpoint. Sensitivity analyses supported the re-analyses.

Disease Activity Score

The Disease Activity Score 28 based on erythrocyte sedimentation rate [DAS28-4(ESR)], is a composite endpoint with differential weighting given to each component. The components of the DAS28 arthritis assessment included tender joint count, swollen joint count, an acute phase reactant (ESR), and the patient’s global assessment of arthritis. In contrast to the ACR which describes relative improvement in disease activity, the DAS describes disease activity at a given point in time.

There were several methodological issues with the data collection and analysis of the DAS28-4ESR endpoint which limited the interpretation of the results. These issues were addressed but were not adequately resolved. In the sponsor’s primary analysis, statistical significance was reached for both doses only in Studies II, III, and V, but not in Study I. Also, this endpoint could not be tested statistically in Study IV as the mTSS endpoint failed to show statistical significance. When the data were re-analyzed (taking note of the uncertainty regarding proper data inclusion), this endpoint was no longer significant for the 5 mg BID dose in Studies III or V. Therefore, using the re-analysis, only one Phase III study (Study II) was statistically significant from placebo for the 5 mg BID dose.

Joint Structure Preservation

Joint structure preservation as measured by the change from baseline in the modified Total Sharp Score (mTSS) at Month 6 was only assessed in Study IV.

In the primary analysis, joint preservation was statistically significant for the Xeljanz 10 mg BID dose, but not for the 5 mg BID dose. The Per Protocol (PP) analysis showed no statistically significant differences from placebo for either of the Xeljanz doses. The results of the sensitivity analysis (non-parametric) using the Analysis of Covariance model on the rank conducted specifically for mTSS were not significant between placebo and Xeljanz 10 mg, while they were significant between Xeljanz 5 mg and placebo. Considering that radiological progression over time shows a highly skewed distribution pattern, the non-parametric analysis is considered most relevant for this endpoint. As the non-parametric testing failed to reject the hypothesis that Xeljanz 10 mg is not different from placebo in mTSS, the gatekeeping strategy does not allow testing of the 5 mg dose.

Joint preservation at Month 12 was assessed as a secondary endpoint. The results of the primary analysis were not significant for the 5 mg BID group, but were significant for the 10 mg BID group. The PP analysis at Month 12 was also significant for the 10 mg but not the 5 mg dose. Contrary to Month 6 non-parametric results, the non-parametric analyses at Month 12 were not significant for either of the doses. Given the lack of robustness across the different analyses, as well as the absence of corroborating data from another study, no definitive conclusions regarding Xeljanz treatment on the prevention of structural damage could be drawn. Study IV was not adequate to support a prevention of radiographic progression claim.

Xeljanz treatment did have beneficial effects on the majority of the symptomatic endpoints (with the exception of DAS28), however; joint preservation could not be claimed, and therefore put the efficacy of Xeljanz behind that of most other traditional DMARDs and biologics. Due to the absence of a decrease in disease activity and joint progression, together with the severity of the safety profile, Health Canada issued a Notice of Non‑Compliance (NON) to the sponsor as the benefits of treatment did not outweigh the risks.

In the Response to NON, the sponsor provided additional efficacy data from another Phase III study; a new Phase III study comparing Xeljanz to MTX in MTX-naïve RA patients. The data demonstrated statistically significantly superior effects of Xeljanz against MTX on both the symptomatic endpoints as well as on joint preservation. A major deficiency of this study was that the population studied was not MTX-resistant as proposed in the Xeljanz indication. Therefore, the sponsor provided published literature supporting the common mechanism of action of Xeljanz in both populations, linking the data from the clinical study to the population in the proposed indication.

Overall Analysis of Efficacy

The efficacy data from the five pivotal studies together with the data received from the Response to NON provided substantial evidence of symptomatic relief of moderate to severe RA, and modest evidence of some joint preservation. Xeljanz was shown to be superior to placebo in the majority of the co-primary endpoints in the Phase III clinical studies. Xeljanz demonstrated similar clinical efficacy to adalimumab, an approved DMARD used in Study III. Literature suggests that longer placebo-controlled studies would be required to demonstrate a statistically significant effect on joint preservation in this population, which is not possible due to the ethical constraints of giving placebo to patients for prolonged periods of time. Further, the United States Food and Drug Administration (FDA) released a new draft guidance (May 2013) on the development of treatments for RA which diminishes the need for evidence of joint preservation in RA therapies.

In response to the negative benefit/risk assessment in the NON, the sponsor restricted the indication and eliminated the 10 mg BID dose as many of the risks were dose‑dependent. The sponsor changed the indication to state that Xeljanz is to be used in combination with MTX, and removed its use with other DMARDs as DMARDs other than MTX were not sufficiently assessed by a single pivotal study. Also the claim for structural preservation was removed from the indication to better reflect the efficacy data from Study IV. While some uncertainty still exists for the effect of Xeljanz on joint preservation in MTX-resistant patients, the strength of the data provided together with the published literature support a positive benefit/risk assessment in that population.

For more information, refer to the Xeljanz Product Monograph, approved by Health Canada and available through the Drug Product Database.

Clinical Safety

The clinical safety review for the original drug submission evaluated the safety data from the Phase II and Phase III studies, as well as data from submitted long-term extension (LTE) studies. In total, 2,951 patients received Xeljanz 5 mg or 10 mg for at least 1 year (1,107 received 5 mg BID, 939 received 10 mg BID and 905 received a combination of doses between the index study and LTE study for at least 1 year) and 693 patients received either dose for at least 2 years.

The non-clinical and clinical data suggested an increased risk of serious infections, opportunistic infections, tuberculosis, herpes zoster, malignancy and lymphoma, hypertension, gastrointestinal (GI) perforations affecting primarily the lower GI tract, interstitial lung disease (ILD), decreases in neutrophils and neutropenia, erythropenia, decreases in lymphocyte counts and natural killer (NK) cells, decreases in platelets, lipid elevations, increases in creatine kinase (CK), and increases in body weight and body mass index when compared to placebo and in many cases when compared to adalimumab. As well, an increased incidence of alanine aminotransferase (ALT) elevations >3 times the upper limit of normal was seen in the Xeljanz groups in the Phase III background DMARD studies and one case of possible drug-induced liver injury in a patient treated with Xeljanz 10 mg was reported suggesting a possible major concern with respect to hepatic safety. In many cases these risks were dose-dependent with an increased risk demonstrated in the 10 mg BID group versus the 5 mg BID group. As well, some of these risks, such as infections and malignancies, were reported more commonly in patients 65 years and older, and others, such as herpes zoster, were reported more commonly in Asian patients. Some data suggested a clustering of cases of ILD in the Asian population.

Cardiovascular safety was assessed in the Phase III clinical studies of 0-12 months in duration. Adjudicated events of congestive heart failure were reported in 5/1,214 patients (incidence 0.4%, 0.55 events/100 patient years) that received Xeljanz 10 mg. There were no adjudicated events of congestive heart failure in the corresponding placebo group (n = 681) or the Xeljanz 5 mg group (n = 1,216). In the placebo-controlled Phase II clinical studies in patients with rheumatoid arthritis, steady-state treatment with 5 mg BID Xeljanz was associated with statistically significant 4-7 bpm decreases in heart rate and 4-8 ms increases in the PR interval compared with placebo.

A number of study design issues in the LTE studies made the interpretation of the long‑term safety data particularly problematic and no definite conclusions could be drawn. The results were considered inadequate to support the chronic use of the product in the proposed patient population. Given the severity of the safety profile, including the lack of adequate long-term safety data, together with the absence of a decrease in disease activity and joint progression, the benefits of treatment with Xeljanz did not outweigh the risks. As a result, a Notice of Non‑Compliance (NON) was issued due to a negative benefit-risk assessment.

In the Response to NON, updated safety information from on-going clinical studies was submitted. The most frequently reported adverse events in the Xeljanz treatment groups were upper respiratory tract infection, headache, nasopharyngitis, and diarrhea. In the LTE studies, the most frequent adverse events also included hypertension, herpes zoster, urinary tract infection, back pain and influenza. Safety issues of special interest included serious infections; opportunistic infections; malignancies; hypertension; hepatotoxicity; increased lipids; GI perforations; decreases in neutrophils, lymphocytes, platelets, NK cells, serum IgG, IgM, and IgA levels; and increases in creatinine, creatine kinase, and body weight. No new risk factors were identified, and there were no increases in the frequency or severity of the known risk factors. The sponsor provided evidence via the updated safety reports as well as through published literature to demonstrate that the safety profile of Xeljanz is similar to those of marketed biologic DMARDs, which are familiar to treating physicians.

Risk mitigation included withdrawal of the higher dose of 10 mg BID, as many of the risk factors identified in the drug submission were dose-dependent. The Xeljanz Product Monograph was updated to reflect the risks. The indication was restricted to use in combination with MTX, in patients who have an inadequate response to MTX, with the exception of patients intolerant to MTX where it can be used as monotherapy. A Serious Warnings and Precautions box was added for serious infections and malignancies. Warnings and precautions were also added for all of the identified safety events of interest, including hypertension, infections, GI perforations, hepatoxicity, ILD, and immunizations. The Risk Management Plan was also updated to address all of the significant safety concerns.

In summary, the pivotal studies from the NDS together with the additional data submitted in the Response to NON provided sufficient evidence to support a positive benefit‑risk assessment for the use of Xeljanz (5 mg BID) to be used in combination with MTX in MTX-resistant patients. Xeljanz may only be used as monotherapy in patients who are intolerant to MTX.

For more information, refer to the Xeljanz Product Monograph, approved by Health Canada and available through the Drug Product Database.

The non-clinical pharmacology, safety pharmacology, pharmacokinetic, and toxicology studies have characterized the non-clinical profile of tofacitinib in sufficient detail to support the intended clinical indication. The findings in the toxicology studies were, for the most part, consistent with effects that would be anticipated consequent to the pharmacologic activity of tofacitinib. These included decreases in circulating white blood cells and lymphocyte depletion in lymphoid tissues. Reductions of red blood cell parameters were also noted but generally occurred at higher doses than for decreases in lymphocytes. Target organs were mainly lymphoid tissues/organs and bone marrow, with gastrointestinal effects in some studies, and changes indicative of liver or lung effects manifested in rats. The severity and extent of adverse effects in repeat-dose studies were dose-dependent.

The results of the non-clinical studies as well as the potential risks to humans have been included in the Xeljanz Product Monograph. In view of the intended use of Xeljanz, there are no pharmacological/toxicological issues within this submission which preclude authorization of the product.

For more information, refer to the Xeljanz Product Monograph, approved by Health Canada and available through the Drug Product Database.

The Chemistry and Manufacturing information submitted for Xeljanz has demonstrated that the drug substance and drug product can be consistently manufactured to meet the approved specifications. Proper development and validation studies were conducted, and adequate controls are in place for the commercial processes. Changes to the manufacturing process and formulation made throughout the pharmaceutical development are considered acceptable upon review. Based on the stability data submitted, the proposed shelf‑life of Xeljanz is considered acceptable.

Proposed limits of drug-related impurities are considered adequately qualified; that is, within International Conference on Harmonisation limits and/or qualified from toxicological studies.

All sites involved in production are compliant with Good Manufacturing Practices.

All non-medicinal ingredients (described earlier) found in the drug product are acceptable for use in drugs according to the Food and Drug Regulations.

Letters of attestation confirming that the materials are not from a bovine spongiform encephalopathy and transmissible spongiform encephalopathy (BSE/TSE) affected country/area have been provided for this product indicating that it is considered to be safe for human use.