SUMMARY

• INVOKANA may increase the risk of hypoglycemia when combined with insulin or an insulin secretagogue (eg, sulfonylurea).¹
• The risk of hypoglycemia may be lowered by a reduction in the dose of sulfonylurea (or other concomitantly administered insulin secretagogues) or insulin.¹
• Please refer to the following section of the Full Prescribing Information which is relevant to your inquiry: DRUG INTERACTIONS.¹
• The Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation (CREDENCE) trial assessed the impact of canagliflozin vs placebo on kidney and cardiovascular (CV) outcomes in patients with chronic kidney disease (CKD) and patients with type 2 diabetes mellitus (T2DM). Beal et al (2025)² conducted a post hoc analysis of the CREDENCE trial to examine the impact of canagliflozin vs placebo on insulin usage in patients with CKD and T2DM by assessing dose initiation, intensification, reduction, and discontinuation.
  • Overall, 2884 patients were treated with insulin at baseline (mean age, 62.5 years; 35.6% were female); these patients had diabetes for a mean period of 17.5 years and exhibited a mean baseline HbA1c level of 8.4%.
  • Over a median treatment period of 2 years, insulin initiation or dose intensification occurred in 407/2202 (18.5%) patients receiving INVOKANA and 476/2199 (21.6%) patients receiving placebo.
  • INVOKANA lowered the occurrence of insulin initiation or dose intensification by 19% compared with placebo (HR, 0.81; 95% CI, 0.71-0.93).
• The phase 3 CANagliflozin cardioVascular Assessment Study (CANVAS) included an efficacy and safety insulin substudy.^3,4
  • At 18 weeks, INVOKANA 100 mg and 300 mg added on to background insulin therapy (≥30 international units [IU]/day) significantly improved glycemic control and reduced body weight and systolic blood pressure (BP) relative to placebo.⁵
  • At 52 weeks, both doses of INVOKANA added on to background insulin therapy (≥30 IU/day) continued to improve glycemic control and reduced body weight as observed at 18 weeks. INVOKANA was generally well tolerated with overall incidence of safety rates higher in treatment groups relative to placebo.⁶
  • In a substudy of CANVAS in which patients received INVOKANA 100 mg or 300 mg in addition to insulin therapy (≥20 IU/day), both INVOKANA doses met the primary efficacy endpoint, change from baseline in glycated hemoglobin (HbA1c) relative to placebo at week 18 (P<0.001) and sustained reductions through week 52. There was a greater incidence of hypoglycemia, genital mycotic infections, and hypovolemia with both INVOKANA doses as compared with placebo.⁷
• An analysis was performed of a subset of patients enrolled in the CANVAS study who were on stable insulin treatment ≥30 IU/day and metformin ≥2000 mg/day (N=432) at week 18. Significant improvements in glycemic control, reductions in body weight and systolic BP, and increases in low-density lipoprotein cholesterol and high-density lipoprotein cholesterol were seen with both INVOKANA 100 and 300 mg doses compared with placebo.⁸
• A post hoc analysis of the CANVAS study was conducted to assess the effects of INVOKANA on major adverse cardiovascular events (MACE) in patients with T2DM categorized into subgroups defined by disease duration, treatment intensity, and baseline HbA1c levels.⁹
  • INVOKANA diminished MACE in patients receiving insulin, with no evidence of benefits differing from those observed in other patients in the study (hazard ratio [HR], 0.85; 95% confidence interval [CI], 0.721.00).⁹
• In a randomized, double-blind, phase 2 study, Henry et al (2015)¹⁰ evaluated the safety and efficacy of INVOKANA as add-on to insulin in patients with type 1 diabetes (T1DM; N=351). At week 18, a greater proportion of patients receiving INVOKANA 100 mg and 300 mg met the primary composite endpoint of a lowered A1c ≥0.4% and no increase in body weight relative to baseline compared with placebo (36.9%, 41.4%, and 14.5%, respectively; P<0.001), as well as provided reductions in HbA1c, body weight, and insulin dose compared with placebo. Overall, the percentage of patients who experienced any adverse event (AE) was 55.6%, 67.5%, and 54.7% for INVOKANA 100 mg, 300 mg, and placebo, respectively.

BACKGROUND

The insulin substudy was part of the phase 3 CANVAS^3,4 clinical trial, as one of 2 studies^3,11 comprising the CANVAS Program. The CANVAS Program^12,13 (N=10,142) evaluated CV safety, as well as the potential for CV protection of INVOKANA in patients with T2DM who had either a prior history of CV disease (CVD) or at least 2 CV risk factors.^12,13

CLINICAL DATA

Post Hoc Analysis of the CREDENCE Trial

Beal et al (2025)² conducted a post hoc analysis of the CREDENCE trial to investigate the impact of canagliflozin vs placebo on insulin usage in patients with CKD and T2DM by assessing dose initiation, intensification, reduction, and discontinuation.

Study Design/Methods

• The trial included patients aged ≥30 years with T2DM (HbA1c, 6.5%-12%) and CKD, with a baseline eGFR of 30 to <90 mL/min/1.73 m², and a urine albumin:creatinine ratio (UACR) of ≥300-5000 mg/g.
• Of the 4401 patients included in the primary analysis, 65.5% (n=2884) were receiving insulin at baseline and were randomized to receive INVOKANA 100 mg once daily (n=1452) or matching placebo (n=1432).
• The primary composite endpoint were the initiation of insulin therapy or dose intensification (defined as a >25% increase in the insulin dose), along with an additional assessment of dose reduction and discontinuation.
• Clinical outcomes were doubling of serum creatinine, kidney failure or death due to kidney failure, heart failure hospitalization or CV death, nonfatal myocardial infarction, nonfatal stroke or CV death were analyzed by baseline insulin use. Key safety outcomes were also assessed.

Results

• Overall, 2884 patients were treated with insulin at baseline (mean age, 62.5 years; 35.6% were female); these patients had diabetes for a mean period of 17.5 years and exhibited a mean baseline HbA1c level of 8.4%.
• Over a median treatment period of 2 years, insulin initiation or dose intensification occurred in 407/2202 (18.5%) patients receiving INVOKANA and 476/2199 (21.6%) patients receiving placebo.
• INVOKANA lowered the occurrence of insulin initiation or dose intensification by 19% compared with placebo (HR, 0.81; 95% CI, 0.71-0.93).
• INVOKANA reduced insulin initiation by 28% in insulin-naïve patients (HR, 0.72; 95% CI, 0.55-0.93) and lowered the need for dose intensification by 16% in patients treated with insulin at baseline (HR, 0.84; 95% CI, 0.72-0.98).
• The impact on the primary outcome remained consistent in sensitivity analyses that accounted for baseline differences between patients using insulin and those not using insulin at baseline (HR, 0.82; 95% CI, 0.72-0.93).
• INVOKANA consistently reduced the relative risk of CV and kidney events, regardless of insulin use at baseline. However, event rates for these outcomes were higher among patients using insulin than among those not using insulin.
• A total of 12 patients developed ketoacidosis, with 11 of them receiving INVOKANA. Of these patients, 11 (91.7%) had been using insulin at baseline.

DIA 3008 - CANVAS Substudy (≥30 IU/Day of Insulin)

Matthews et al (2012)⁵ evaluated the efficacy and safety of INVOKANA compared with placebo in adult patients with T2DM on background insulin therapy (N=1718). Neal et al (2013)⁶ conducted an evaluation extending to 52 weeks. This substudy was conducted in patients receiving ≥30 IU/day of insulin, alone or in combination with other antihyperglycemic agents (AHAs), at study entry.

Study Design/Methods

• Patients were randomized to INVOKANA 100 mg (n=566), 300 mg (n=587), or placebo (n=565) once daily.
• Glycemic rescue therapy was administered during the initial 18-week period only if fasting plasma glucose (FPG) rose above predefined levels. Thereafter, glycemic management was otherwise at the discretion of the investigator.¹² Through week 18, patients were to remain on a stable AHA regimen (both insulin and other AHAs) unless down-titration was required to manage or avoid hypoglycemia, or glycemic rescue criteria were met; after week 18, AHA therapy could be adjusted to achieve individual glycemic targets, as determined by the investigator.⁶
• The primary efficacy endpoint evaluated at weeks 18 and 52 was change from baseline in HbA1c.^5,6
• The secondary efficacy endpoints were change from baseline in FPG, systolic BP, percent change from baseline in body weight, fasting plasma lipids, and proportion of patients reaching A1C <7.0%. Efficacy endpoints were analyzed in the modified intent-to-treat (mITT) population consisting of patients who received ≥1 dose of study drug.^5,6

Results

• Baseline demographic and disease characteristics for the mITT population (N=1718) were similar across groups: mean daily insulin dose 83 IU (majority basal plus bolus), mean age 62.8 years, men 67.0%, mean duration of T2DM 16.6 years, mean body mass index (BMI) 33.8 kg/m². Approximately 93% (N=1594/1718) of patients completed 18 weeks.
• Of the 1718 patients at study entry, 1449 (84.2%) remained on treatment at week 52. Baseline demographic and disease characteristics of patients who completed 52 weeks of treatment were generally similar to the mITT population.^5,6

Efficacy at 18 Weeks

• INVOKANA 100 mg and 300 mg once daily added onto background insulin therapy significantly reduced A1C relative to placebo. A summary of prespecified primary and secondary endpoints are shown in Table: Selected Endpoints at Weeks 18 and 52 (mITT Population).
• Compared with the INVOKANA 100 mg group (4.1%) and the INVOKANA 300 mg group (3.1%), more patients in the placebo group (8.7%) needed glycemic rescue therapy.⁵

Efficacy at 52 Weeks

• Between week 18 and week 52, mean daily insulin dose was unchanged (83 IU/day) for 70%, 69%, and 71% of patients treated with INVOKANA 100 mg, INVOKANA 300 mg, and placebo, respectively. Among patients who had a change in insulin dose (ie, increase or decrease of >15% for ≥7 consecutive days), the proportion in the INVOKANA 100 mg, INVOKANA 300 mg, and placebo groups who had an increase was 10%, 7%, and 19%, respectively, and the proportion who had a decrease was 19%, 22%, and 8%, respectively.
• Efficacy of both INVOKANA doses on glycemic control at 52 weeks were comparable to those achieved at 18 weeks (see Table: Selected Endpoints at Weeks 18 and 52 [mITT Population]). Based on results for the percent change from baseline in body weight, the weight reduction with INVOKANA increased in magnitude between weeks 18 and 52, whereas placebo resulted in negligible percent change from baseline.⁶

Safety

• Safety endpoints are shown in Table: Summary of Overall Safety and Select AEs at Weeks 18 and 52.

DIA 3008 - CANVAS Substudy (≥20 IU/Day of Insulin)

Neal et al (2015)¹⁴ evaluated the efficacy and safety on a larger subset of patients (N=2072) in the CANVAS substudy on INVOKANA compared with placebo in adult patients with T2DM. In contrast to the aforementioned analysis evaluating patients receiving ≥30 IU/day of insulin,^5,6 this subset analysis¹⁴ evaluated patients receiving ≥20 IU/day of insulin alone, or in combination with other AHAs at study entry.

Study Design/Methods

• Of 2072 patients included in the primary analysis, patients receiving insulin at baseline were randomized to receive INVOKANA 100 mg (n=692), INVOKANA 300 mg (n=690), and placebo (n=690) once daily.
• Patients were required to receive stable background glucose-lowering therapy, including insulin, for 8 weeks prior to screening and for the first 18 weeks after randomization. After the initial 18-week period, glycemic management was at the discretion of the investigator and applicable local guidelines.
• The primary endpoint of this substudy was the change in HbA1c from baseline among patients using insulin.
• The secondary efficacy endpoints were change from baseline in FPG, systolic and diastolic BP, and percent change in body weight and fasting plasma lipids, and the proportion of participants reaching an HbA1c of <7.0%. Effects were examined at 18 and 52-weeks.

Results

• Baseline demographics were balanced across randomized groups.
  • Patients were 66% male and had a median age of 63 years, mean HbA1c of 8.3%, BMI of 33.1 kg/m², estimated glomerular filtration rate (eGFR) of 75 mL/min/1.73 m², FPG of 9.2 mmol/L, and a median daily insulin dose of 60 IU. Most individuals were using basal/bolus insulin. Insulin pumps were permitted.
• Although all efficacy endpoints and safety results are not presented below, this larger group of patients receiving insulin dose of ≥20 IU/day (N=2072) yielded similar conclusions when compared with the smaller group of patients receiving insulin dose of ≥30 IU/day (N=1718).

Efficacy

• INVOKANA 100 mg and 300 mg significantly reduced the primary outcome of HbA1c relative to placebo at week 18 (P<0.001) and were sustained through week 52.
• Reductions in body weight and FPG increased in the proportion of patients achieving HbA1c <7.0 were observed with both INVOKANA 100 mg and 300 mg vs placebo at week 18 (P<0.001) and were sustained through week 52.

Safety

• At 52 weeks, AEs were reported for 77%, 78%, and 81% of patients treated with INVOKANA 100 mg, INVOKANA 300 mg, and placebo, respectively.
• There was a greater incidence of hypoglycemia, genital mycotic infections, and hypovolemia with both INVOKANA 100 mg and 300 mg doses as compared with placebo.

Post Hoc Analysis of the CANVAS Study (Effects on MACE)

Young et al (2021)⁹ conducted a post hoc analysis of the CANVAS study to assess the effects of INVOKANA on MACE in patients with T2DM categorized into subgroups defined by disease duration, treatment intensity, and baseline HbA1c levels.

Study Design/Methods

• The patients were divided into 3 subgroups based on the following:
  • Treatment intensity defined by glucose-lowering treatments at baseline (0 or 1 oral glucose-lowering agent, 2 oral glucoselowering agents, ≥3 oral glucose-lowering agents, or any combination of glucoselowering medication that included insulin therapy)
  • Disease duration defined by tertiles of disease duration at baseline (duration of <10 years, 10-16 years, or >16 years)
  • Baseline HbA1c (≤53.0 mmol/mol [<7.0%], >53.0-58.5 mmol/mol [>7.0-7.5%], >58.5-63.9 mmol/mol [>7.5-8.0%], >63.9-69.4 mmol/mol [>8.0-8.5%], >69.474.9 mmol/mol [>8.5-9.0%], and >74.9 mmol/mol [>9.0%]).
• The CV outcomes assessed included death from CVD, myocardial infarction, and stroke and a combined outcome comprising either CV death or admission for heart failure.
• The kidney outcome was a composite of end-stage renal disease, kidney death, and 40% decrease in eGFR.

Results

• Overall, 5095 (50%) patients were treated with insulin; 1844 (36.2%) were female, and the median age was 63.5 years. Mean disease duration in these patients was 16.3±7.7 years and the mean HbA1c was 68 mmol/mol (8.4±0.9%).
• INVOKANA lowered the risk of MACE across all subgroups (HR, 0.86; 95% CI, 0.75-0.97), with no substantial evidence of benefits varying between subgroups. INVOKANA diminished MACE in patients receiving insulin, with no evidence of benefits differing from those observed in other patients in the study (HR, 0.85; 95% CI, 0.72-1.00).
  • In the treatment intensity subgroup, the HRs for patients taking multiple glucoselowering treatments were as follows: 0 or 1 oral glucose-lowering agent (HR, 0.71; 95% CI, 0.51-0.99), 2 oral glucose-lowering agents (HR, 0.95; 95% CI, 0.711.26), and ≥3 oral glucose-lowering agents (HR, 1.02; 95% CI, 0.59-1.76).
• Indistinguishable results were obtained for other CV outcomes and the combined kidney outcome (HR for combined kidney outcome, 0.60; 95% CI, 0.47-0.77).

Drug Interactions With Insulin

Devineni et al (2012)¹⁵ conducted a phase 1, randomized, double-blind, placebocontrolled, parallel-group study evaluating the pharmacokinetics (PK), pharmacodynamics (PD), and safety of INVOKANA in patients with T2DM suboptimally controlled on insulin and up to 1 oral AHA (n=29; mean age: 42.7-52.8 years; BMI: 32.434.5 kg/m²). Patients received INVOKANA 100 mg/day, 300 mg twice daily, or placebo, in addition to either basal, mealtime, or both basal and mealtime insulin concomitantly. Total insulin dose at baseline per treatment group was 114.6 units/day in the INVOKANA 100 mg/day group, 51.4 units/day in the INVOKANA300 mg twice daily group, and 38.4 units/day in the placebo group.

Canagliflozin was rapidly absorbed in both treatment groups with an elimination half-life of 11.8 to 14.7 hours post dose in the 300 mg twice daily and 100 mg once daily groups, respectively. Insulin dose was not reduced in any patients. Canagliflozin was generally well tolerated with a low incidence of hypoglycemia when used in patients with T2DM currently receiving insulin therapy.

Insulin Dose Adjustments

Insulin dose adjustments should be individualized for each patients and the decision is ultimately up to the healthcare professional’s clinical discretion.

Exploratory Analysis

Rosenstock et al (2013)¹⁶ conducted an exploratory analysis of the insulin substudy⁵ evaluating the efficacy and safety of INVOKANA compared with placebo in a subset of adult patients with T2DM on stable basal (not prandial) insulin therapy at 18 weeks.

• A total of 86 patients in the INVOKANA 100 mg group, 104 in the INVOKANA 300 mg group, and 88 in the placebo group were on stable basal insulin therapy at baseline in the insulin substudy.
• A total of 278 patients were receiving stable basal insulin therapy, of which 262 (94.2%) completed the 18-week treatment period. Baseline demographics and disease characteristics were similar across groups: mean daily insulin dose 59 IU, mean age 63.1 years, males 62-67%, mean duration of T2DM 15.1-15.3 years, and mean BMI 34.4 kg/m².
• Mean insulin dose was unchanged for 85%, 86%, and 93% of patients in the INVOKANA 100 mg, 300 mg, and placebo groups, respectively. Mean change from baseline in insulin dose was -3.2 IU for the INVOKANA 100 mg group, -2.4 IU for the INVOKANA 300 mg group, and 1.1 IU for the placebo group.
• INVOKANA 100 mg and 300 mg improved both HbA1c and FPG compared with placebo.

Literature Search

A literature search of MEDLINE^®, Embase^®, BIOSIS Previews^®, and Derwent^® (and/or other resources, including internal/external databases) was conducted on 22 May 2025, did not identify any relevant citations pertaining to this topic.