Series: Landmark Trials in Endocrinology & Metabolism | Study #2
Category: SGLT2 Inhibitors | Design: Integrated analysis of two multicentre, double-blind, placebo-controlled RCTs (CANVAS and CANVAS-R) | n: 10,142 | Follow-up: 188.2 weeks (mean)
📋 Summary
Authors: Neal B et al., for the CANVAS Programme Collaborative Group
Journal: N Engl J Med 2017;377:644–657 | DOI: 10.1056/NEJMoa1611925
A total of 10,142 participants with type 2 diabetes mellitus and high cardiovascular risk were randomly assigned to canagliflozin (100 mg or 300 mg daily) or matching placebo across two integrated trials. Participants had a mean age of 63.3 years, a mean diabetes duration of 13.5 years, and 65.6% carried a prior diagnosis of cardiovascular disease. The primary composite endpoint comprised death from cardiovascular causes, nonfatal myocardial infarction, and nonfatal stroke, assessed over a mean follow-up of 188.2 weeks. The primary endpoint was met, with event rates of 26.9 per 1,000 patient-years in the canagliflozin group versus 31.5 per 1,000 patient-years in the placebo group (HR 0.86; 95% CI 0.75 to 0.97; p<0.001 for noninferiority; p=0.02 for superiority). Hospitalisation for heart failure was significantly reduced (HR 0.67; 95% CI 0.52 to 0.87), and a renal composite outcome demonstrated probable benefit (HR 0.60; 95% CI 0.47 to 0.77) and reduction in progression of albuminuria (HR 0.73; 95% CI 0.67 to 0.79). Lower-extremity amputation occurred at approximately double the rate in the canagliflozin group compared with placebo (6.3 vs 3.4 per 1,000 patient-years; HR 1.97; 95% CI 1.41 to 2.75), with events concentrated at the level of the toe or metatarsal.
📊 Key Findings
| Outcome | Canagliflozin | Placebo | Effect Size |
|---|---|---|---|
| 3-point MACE (primary) | 26.9/1,000 PY | 31.5/1,000 PY | HR 0.86 (0.75–0.97) · p=0.02 |
| HF hospitalisation | 5.5/1,000 PY | 8.7/1,000 PY | HR 0.67 (0.52–0.87) |
| Renal composite | see note | see note | HR 0.60 (0.47–0.77) |
| Progression of albuminuria | see note | see note | HR 0.73 (0.67–0.79) |
| Cardiovascular death | see note | see note | HR 0.87 (0.72–1.06) · NS |
| Lower-extremity amputation | 6.3/1,000 PY | 3.4/1,000 PY | HR 1.97 (1.41–2.75) · Safety signal |
| Absolute risk reduction (MACE) | Approximately 1.8% over 3.6 years | NNT approximately 56 | |
💬 Expert Commentary
The CANVAS Programme provided independent confirmation that cardiovascular and cardiorenal protection attributed to empagliflozin in the EMPA-REG OUTCOME trial was not compound-specific, extending these benefits to canagliflozin and strengthening the hypothesis of a class effect for SGLT2 inhibition. The concordant reductions in hospitalisation for heart failure and attenuation of albuminuria progression reinforced the mechanistic framework of haemodynamic and tubuloglomerular effects as the operative pathways underlying organ protection within this drug class. The cardiovascular mortality component did not achieve individual statistical significance (HR 0.87; 95% CI 0.72 to 1.06), a finding that contrasted with the prominent mortality reduction observed in EMPA-REG OUTCOME and raised early questions regarding whether inter-compound differences existed within the class.
The emergence of a significant lower-extremity amputation signal was the defining and clinically consequential finding of the CANVAS Programme. The approximately twofold increase in amputation risk, concentrated at the toe and metatarsal level, was unanticipated and prompted regulatory safety communications from the European Medicines Agency and the US Food and Drug Administration. The mechanism underlying this signal was not definitively established from trial data. Proposed contributory factors included exacerbation of volume depletion in the presence of peripheral arterial disease, diuretic co-administration, and reduced microcirculatory perfusion secondary to haemodynamic effects of SGLT2 inhibition. Critically, this signal was not reproduced with equivalent magnitude in EMPA-REG OUTCOME or in subsequent SGLT2 inhibitor trials, introducing ongoing uncertainty regarding whether the risk was compound-specific, population-specific, or confounded by study design differences. In clinical practice, initiation of canagliflozin requires pre-treatment assessment of lower-extremity vascular status and neuropathy, with structured foot surveillance throughout the treatment course.
Limitations: Integration of two trials with differing designs, participant enrolment criteria, and follow-up durations introduced methodological heterogeneity into the pooled analysis. Renal composite outcomes did not meet the prespecified primary hierarchical testing threshold for statistical significance. The amputation finding was not a prespecified primary safety endpoint, and formal causal inference is constrained accordingly. The study was industry-sponsored by Janssen Research and Development.
🔑 BOTTOM LINE
The CANVAS Programme confirmed cardiovascular and renal protective effects of canagliflozin consistent with a class effect for SGLT2 inhibition, but simultaneously identified a clinically important lower-extremity amputation signal that introduced a compound-specific safety qualification, mandating structured vascular assessment prior to initiation and vigilant foot surveillance throughout treatment.
⭐ Clinical Impact Rating: ●●●●○ Practice-changing with safety qualification
Next in the series: Study #3 DECLARE-TIMI 58: Dapagliflozin and Cardiovascular Outcomes in a Broader Type 2 Diabetes Population
