Painting a New CANVAS for SGLT-2 Inhibitors?


Written By

Reviewed By


Neal B, Perkovic V, Mahaffey KW, et al. Canagliflozin and Cardiovascular and Renal Events in Type 2 Diabetes. N Engl J Med. 2017;377(7):644-57.

While good glycemic control has been shown to prevent microvascular complications (e.g. retinopathy, nephropathy, neuropathy), only a few anti-diabetic agents have been shown to reduce macrovascular complications (e.g. cardiovascular events.1-5  Empagliflozin, a sodium glucose transporter-2 (SGLT2) inhibitor, not only reduced the risk of CV events but also all-cause mortality in the EMPA-REG OUTCOME study.4  (See iForumRx Commentary – The Heart of the Mattter: Is EMPA-REG a Game Changer for Diabetes Management?)  Based on this data, the SGLT2 inhibitors were given favorable second-line treatment status in the most recent AACE/ACE clinical practice guidelines.6   But do all SGLT2 inhibitors confer the same benefits… and risks?  The recently published Canagliflozin and Cardiovascular and Renal Events in Type 2 Diabetes (CANVAS) trial assessed the cardiovascular and renal benefits from long-term canagliflozin use.7  Dapagliflozin is also being studied but the results of the long-term outcome trial are not yet available.8


The CANVAS program was comprised of two trials: CANVAS, the cardiovascular trial, and CANVAS-R, an assessment of canagliflozin on renal outcomes.  Interim data from CANVAS in 2013 led to the creation of CANVAS-R and the results of the two studies were concurrently evaluated and reported. Table 1 provides inclusion and exclusion criteria.  Patients in CANVAS were randomized 1:1:1 to receive canagliflozin 300 mg daily, canagliflozin 100 mg daily, or placebo while patients in the CANVAS-R study were randomized 1:1 to receive canagliflozin 100 mg (with the option of increasing to 300 mg at week 13) or placebo.  After randomization, patients were evaluated in face-to-face visits three times during year one and every 6 months thereafter. 


The primary outcome was a composite of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke.  Secondary outcomes included: death from any cause, death from cardiovascular causes, progression of albuminuria, and the composite of death from cardiovascular causes and hospitalization for heart failure.  Progression of albuminuria was defined as an increase in albuminuria by 30% along with progression between stages of albuminuria.  The urinary albumin-to-creatinine ratio was measured every 26 weeks in CANVAS-R.  In CANVAS it was measured at week 12 and then annually.  Serum creatinine was collected every 26 weeks in both studies.  Exploratory analysis included nonfatal myocardial infarction, nonfatal stroke, hospitalization for heart failure as well as a renal composite endpoint.  The renal composite outcome was either a 40% reduction in eGFR sustained over two consecutive measurements, need for renal-replacement therapy, or death from renal causes.  An intention-to-treat approach was used.  The trial was powered for non-inferiority between treatment and placebo, and if statistically significant, superiority was analyzed.


Table 1. Notable Inclusion and Exclusion Criteria

Inclusion Criteria

  • Men or women diagnosed with type 2 diabetes

  • Patients 30 years and older with documented symptomatic atherosclerotic cardiovascular disease defined as stroke, myocardial infarction, unstable angina, coronary artery bypass graft, percutaneous coronary intervention, peripheral revascularization, symptomatic hemodynamically-significant carotid or peripheral vascular disease, or amputation secondary to vascular disease


  • Patients 50 years or older with two or more of the following risks: duration of type 2 diabetes of 10 years or more, systolic blood pressure >140 mm Hg while using antihypertensive treatment, current daily cigarette smoker, documented micro/macroalbuminuria or document HDL <39 mg/dL

Exclusion Criteria

  • History of diabetic ketoacidosis, type 1 diabetes, pancreas or beta-cell transplantation, or diabetes secondary to pancreatitis or pancreatectomy

  • Fasting fingerstick glucose >270 mg/dL at baseline

  • Fasting fingerstick glucose <110 mg/dL for patient on sulfonylurea or insulin at baseline

  • History of one or more severe hypoglycemic episodes within 6 months before screening

  • Renal disease requiring treatment with immunosuppressive therapy or a history of dialysis or renal transplant

  • Myocardial infarction, unstable angina, revascularization procedure, or cerebrovascular accident within 3 months before screening

  • Estimated glomerular filtration rate <30 mL/min/1.73 m2 at baseline


A total of 10,142 participants enrolled in the CANVAS program. The mean age of participants was 63.3 years old at baseline and 35.8% were female.  Most participants had long-standing diabetes (mean= 13.5 years) with a mean hemoglobin A1c of 8.2% and eGFR of 76.5 mL/min/1.73m2 at baseline.  Most participants had a history of cardiovascular disease (65.6%) and hypertension (90.0%).  Patients were followed for up to 6.5 years (mean follow-up = 3.6 years, median follow-up = 5.2 years).


The results (Table 2) demonstrated that the pooled canagliflozin group was not only non-inferior to placebo but superior for the primary outcome (death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke).  However, none of the individual components were non-inferior (or superior) to placebo.    Progression of albuminuria occurred less frequently in patients receiving canagliflozin however the effects were greater in the CANVAS-R study compared to CANVAS (HR 0.64, 95% CI 0.57-0.73 compared to HR 0.80, 95% CI 0.72-0.90 respectively).   A subgroup analysis showed that patients who were Caucasian, had a BMI ≥30, hemoglobin A1c ≥8%, eGFR of 30 to <60 mL/min/1.73m2,  or concurrently used beta-blockers and diuretics were more likely to benefit from canagliflozin. Canagliflozin reduced the A1c by 0.58% (95% CI -0.61 to -0.56), body weight by  -1.60 kg (95% CI -1.70 to -1.51), systolic blood pressure by 0.393 mm Hg (95% CI -4.30 to -3.56) and diastolic blood pressure by -1.39 mm Hg (95% CI -1.61 to -1.17) when compared to placebo (P<0.001 for all analysis).



Table 2. Study Outcomes


Canagliflozin (pooled)


Hazard Ratio


Number of patients per 1000 patient-year

Death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke




P<0.001 for non-inferiority

P=0.02 for superiority

Death from cardiovascular causes





Nonfatal myocardial infarction





Nonfatal stroke





Hospitalization for any cause





Hospitalization for heart failure





Death from cardiovascular causes or hospitalization for heart failure





Death from any cause





Progression of albuminuria





40% reduction in eGFR, renal-replacement therapy, or renal death





NS = Not significant; NR = Not reported


In terms of safety, adverse effects leading to treatment discontinuation was not more frequent in the canagliflozin groups when compared to placebo (35.5% vs 32.8%, p = 0.07) but, not surprisingly, genital infections, volume depletion, and diuresis were more frequent with canagliflozin use. Rates of hyperkalemia, acute kidney injuries, pancreatitis, malignancies, and thromboembolism were similar to placebo.  However, unexceptedly, the risk of amputation and fracture was significantly higher in the canagliflozin group. Patients who received canagliflozin experienced 6.3 amputations per 1000 patient years compared to 3.4 amputations in the placebo-treated group (P<0.001).  Most amputations (71%) occurred at the level of the toe or metatarsal.  Moreover, the canagliflozin group experienced 15.4 fractures per 1000 patient years compared to 11.9 fractures in the placebo group (P<0.001).  Most fractures occurred in the first 12 weeks and most frequently involved the arms or wrists.  It is conceivable these fractures are related to hypovolemia and falls.


The CANVAS program demonstrates that canagliflozin had a positive impact on the composite of death from cardiovascular causes, nonfatal myocardial infarctions, and nonfatal stroke.  While none of the individual outcomes were statistically different from placebo, the results were largely driven by reductions in death from cardiovascular causes and nonfatal myocardial infarctions.  When compared to EMPA-REG OUTCOME study, the findings in CANVAS are not as robust.  EMPA-REG OUTCOMES demonstrated statistically significant reductions in the composite outcome [P=0.04 for superiority] as well as all-cause mortality and cardiovascular mortality [P<0.001].  Both medications prevented the progression of albuminuria (5% absolute risk reduction with empagliflozin vs. approximately 4% with canagliflozin). The increased risk of amputation observed in the canagliflozin group was an unexpected finding.  However, the rate observed in the canagliflozin group is comparable to the national rate of 5.0 amputations per 1000 patient-years based on 2014 statistics and thus the rate of amputations observed in the placebo group is lower than expected.9 None-the-less, this finding can’t be dismissed and more research regarding its incidence and potential cause(s) is needed. Regarding fractures, the rate of fractures was increased in the CANVAS but not in the CANVAS-R.  EMPA-REG OUTCOME also evaluated the incidence of fractures, and no association was noted. Of note, there was a slight increase in the risk of stroke (fatal and non-fatal) with empagliflozin when compared to placebo in the EMPA-REG OUTCOME study (12.3 vs 10.5 per 1000 patients) but the incidence of stroke with canagliflozin was slightly reduced (7.9 vs 9.6 per 1000 patients) in the CANVAS program. Neither of these differences were not statistically significant, but require additional study.


The strengths of the trial include its large sample size and duration of follow-up as well as the inclusion of patients with and without cardiovascular disease.  However, the authors did not evaluate the effect of canagliflozin dose on the primary and secondary outcomes – thus we don’t know if one dose is better than another.  Limitations include the small subset of patients that were co-diagnosed with chronic kidney disease and the relative high treatment discontinuation rates (29.2% of patients in the canagliflozin group).


These data are compelling – however it is difficult to recommend canagliflozin over empagliflozin, despite improved renal and cardiovascular outcomes observed with both drugs. Both medications slowed the progression to albuminuria, but it is unknown why the benefit was greater in CANVAS-R compared to CANVAS.  The increased rate of fractures and amputations is worrisome, but perhaps these are random findings that won’t bear out over time.  What do you think?  Do the results of CANVAS show that SGLT2 inhibitors, as a class, provide cardiovascular and renal protection in patients with type 2 diabetes?  Are the SGLT2 inhibitors the most appropriate second line agents?