Regulatory Decision Summary for Invokana

The effect of Invokana on cardiovascular risk was evaluated in the CANVAS Program (CANVAS and CANVAS-R studies). These two studies were multicenter, multi-national, randomized, double-blind, placebo-controlled parallel group, time- and event-driven, with similar inclusion and exclusion criteria and involving a total of 10,142 patients with type 2 diabetes who were either ≥30 years with a history of documented cardiovascular disease or ≥50 years with two or more risk factors for cardiovascular disease. In CANVAS, subjects were randomly assigned 1:1:1 to canagliflozin 100 mg, canagliflozin 300 mg, or matching placebo. In CANVAS-R, subjects were randomly assigned 1:1 to canagliflozin 100 mg or matching placebo, and titration to 300 mg was permitted at the investigators discretion. At baseline, 99% of patients were treated with one or more antidiabetic medications including metformin (77%), insulin (50%), and sulfonylurea (43%).

CANVAS successfully met its primary objective of demonstrating that the canagliflozin treatment group was non-inferior to placebo with regard to the MACE composite endpoint using a non-inferiority margin of 1.3, as well as demonstrating superiority for this endpoint (HR: 0.86; 95% CI: 0.75, 0.97; p = 0.0158 for superiority: 2-sided test). The reduction in MACE was accounted for by the subgroup of patients with established cardiovascular disease (HR" 0.82; 95% CI: 0.72, 0.95), whilst the subgroup of patients with only risk factors for cardiovascular disease at baseline had a hazard ratio that included zero (HR: 0.98; 95% CI: 0.74, 1.30).

Nominally statistically significant risk reductions were identified for several other adjudicated endpoints: hospitalisation for heart failure, heart failure resulting in hospitalisation or death, reduction in progression of albuminuria, and increased regression of albuminuria.

MACE was the primary composite endpoint (cardiovascular death, nonfatal myocardial infarction and nonfatal stroke) and is, therefore, considered eligible for an indication provided that this is limited to the secondary prevention subgroup of patients with established cardiovascular disease. The endpoints relating to heart failure and albuminuria are not considered eligible for indications because of failure of the second step of the hierarchical testing strategy and eligibility criteria that were not selected to enrich for patients at risk for these conditions.

The CANVAS/CANVAS-R program has limited utility as a source of adverse event (AE) data because only a circumscribed set of AEs of interest was collected regardless of the seriousness or study drug discontinuation. The overall incidence rates in the CANVAS integrated dataset can be determined only for these pre-specified events (i.e., male mycotic genital infection, malignancies, photosensitivity, venous thromboembolic events, atraumatic amputation, fracture, and diabetic ketoacidosis). Imbalances not in favour of canagliflozin were noted for the following of these AEs: male genital infection, renal cell carcinoma, photosensitivity, atraumatic lower limb amputation, fracture, and diabetic ketoacidosis. With the exception of renal cell carcinoma and photosensitivity, these AEs and associated risk mitigation measures are already dealt with in the PM. To address the risk of photosensitivity, appropriate text has been added to the Adverse Reactions section of the PM. The imbalance in events of renal cell carcinoma is described in the Adverse Reactions section of the PM. It cannot be determined from a clinical trial program of this size and duration whether this imbalance represents a treatment-related effect or a chance finding.

Interim safety results from the ongoing CANVAS program showed an approximately 2-fold increased risk of lower limb amputations associated with canagliflozin use. In CANVAS, canagliflozin -treated patients and placebo-treated patients had 5.9 and 2.8 amputations per 1,000 patients per year, respectively. In CANVAS-R, canagliflozin -treated patients and placebo-treated patients had 7.5 and 4.2 amputations per 1,000 patients per year, respectively. The risk of lower limb amputations was observed at both the 100 mg and 300 mg once daily dosage regimens. In September 2017, public and healthcare professionals were advised about the amputation risk, and the Canadian PM was updated to reflect this safety information. Overall, the number needed to treat (NNT) to prevent an event of MACE or heart failure resulting in death or hospitalization over 5 years is 28.54, which is smaller than the number needed to harm (NNH) to result in an event of lower limb amputation 68.23 over a 5-year time period.

The benefit-harm assessment of canagliflozin remains positive when it is prescribed under the conditions of use as outlined in the Canadian PM.