SUMMARY  

• CREDENCE (Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation), was a randomized, double-blind, placebo (PBO)-controlled, parallel group multicenter clinical trial assessing whether INVOKANA 100 mg has a renal and/or cardiovascular (CV) protective effect compared to PBO in patients with type 2 diabetes mellitus (T2DM) and albuminuric chronic kidney disease (CKD), who were also receiving standard of care.^{1, 2}
• Announcement of Early Termination Due to Efficacy: On July 16, 2018, the Janssen Pharmaceutical Companies of Johnson & Johnson announced that the CREDENCE trial was stopping early based on achievement of pre-specified efficacy criteria. The study’s Independent Data Monitoring Committee (IDMC) recommended this decision during a planned interim analysis based on demonstration of efficacy, as the trial had achieved pre-specified criteria for the primary composite endpoint of end-stage kidney disease (ESKD; defined as time to dialysis or kidney transplantation), doubling of serum creatinine (dSCr), and renal or CV death, when used in addition to standard of care.³
• Key Efficacy Outcomes: In CREDENCE, INVOKANA significantly reduced rates of the composite outcome of ESKD, dSCr, or renal or CV death (43.2 and 61.2 per 1000 patient-years [PY] in the INVOKANA and PBO arms, respectively), resulting in a 30% relative risk reduction (RRR) (hazard ratio [HR], 0.70; 95% confidence interval (CI), 0.59–0.82; P=0.00001;). Effects were consistent across prespecified subgroups.¹
  • The event rate of ESKD was 20.4 and 29.4 per 1000 PY (HR 0.68; 95% CI: 0.54-0.86; P=0.002) in the INVOKANA and PBO groups, respectively.
  • The event rate of dSCr was 20.7 and 33.8 per 1000 PY (HR 0.60; 95% CI: 0.48-0.76; P<0.001) in the INVOKANA and PBO groups, respectively.
  • The event rate of renal death was 0.3 and 0.9 per 1000 PY in the INVOKANA and PBO groups, respectively.
  • The event rate of CV death was 19.0 and 24.4 per 1000 PY (HR 0.78; 95% CI: 0.61-1.00) in the INVOKANA and PBO groups, respectively.
• Key Safety Results: Adverse events (AEs) reported in CREDENCE were similar in both groups and consistent with the known safety profile of INVOKANA, apart from lower limb amputation and fractures. Serious AEs related to study drug were reported more with INVOKANA vs PBO (12.2 vs 8.6 per 1000 PY; HR 1.45; 95% CI: 0.98-2.14).^{1, 4}
  • No imbalance in the risk of lower limb amputation with INVOKANA vs PBO was observed (12.3 vs 11.2 per 1000 PY; HR 1.11; 95% CI: 0.79-1.56).
  • Rates of fracture were similar between the INVOKANA and PBO treatment arms (11.8 vs 12.1 per 1000 PY; HR 0.98; 95% CI: 0.70-1.37).¹
• On April 15, 2019, Janssen announced the full results of the CREDENCE study in a late breaking clinical trial session at the International Society of Nephrology ISN) World Congress of Nephrology (WCN) in Melbourne, Australia and a simultaneous publication in the New England Journal of Medicine (Presentation slides available at: https://www.georgeinstitute.org/sites/default/files/credence-trial-results.pptx; Webcast available at: https://www.youtube.com/watch?v=gZC6PSN7Jt8).
• In CREDENCE, INVOKANA significantly reduced CV outcomes in both the primary and the secondary prevention cohorts.  Patients were classified as secondary prevention if they had a history of coronary, cerebrovascular, or peripheral vascular disease at baseline. All other participants belonged to the primary prevention cohort.⁵
• Additional analyses identified in the literature are referenced for your review.^{6, 7}

MECHANISM OF ACTION

Canagliflozin increases the delivery of sodium to the distal tubule by blocking SGLT2-dependent glucose and sodium reabsorption. This is believed to increase tubuloglomerular feedback. The feedback signal causes afferent arteriolar vasoconstriction, reducing intraglomerular perfusion and pressure, as well as a diminished extracellular plasma volume and BP. Additionally, these effects reduce atrial natriuretic peptide secretion, which may also be important in reducing intraglomerular pressure. These effects are clinically manifested as reductions in albuminuria and eGFR, followed by stabilization in eGFR.^8-10

CLINICAL DATA

CREDENCE (Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation), was a randomized, double-blind, PBO-controlled, parallel group multicenter, event driven clinical trial to assess the effects of INVOKANA (100 mg) compared to PBO on clinically important renal outcomes in people with T2DM and established CKD (estimated glomerular filtration rate [eGFR] 30 to <90 mL/min/1.73m²) and albuminuria (ratio of albumin to creatinine >300 to 5000 mg/g), who were receiving a stable, maximum tolerated labelled dose (for >4 weeks prior to randomization) of an angiotensin converting enzyme inhibitor (ACEi) or angiotensin II receptor blocker (ARB).^{1, 2, 4, 11-13}

Inclusion Criteria^{1, 11}

• ≥30 years of age with CKD in the setting of T2DM (HbA1c ≥6.5% to ≤12.0%; participants in Germany required a HbA1c range of ≥6.5% to <10.5%).
• Estimated glomerular filtration rate (eGFR) ≥30 to <90 mL/min/1.73m²
  • Prespecified to enroll ~60% with stage 3 CKD (eGFR 30 to <60 mL/min/1.73 m²)
  • eGFR was calculated using the CKD-EPI formula
• Albuminuria (urine albumin:creatinine ratio [UACR] >300 mg/g and ≤5000 mg/g)
• On a stable, maximum tolerated labelled daily dose of an ACEi or ARB for ≥4 weeks prior to randomization.

Exclusion Criteria (Select)^{1, 11}

• Type 1 diabetes
• Nondiabetic kidney disease
• History of kidney disease treated with immunosuppression
• History of treatment with chronic dialysis or kidney transplantation
• History of CV events within the previous 12 weeks or a history of New York Heart Association class IV heart failure at any time²
• Dual-agent treatment with an ACEi or an ARB
• Use of a direct renin inhibitor or a mineralocorticoid receptor antagonist
• Concomitant use of an SGLT2 inhibitor in the 12 weeks prior to randomization, current or past participation in another INVOKANA study, or known allergy, hypersensitivity, or intolerance to canagliflozin or excipients²
• History of atraumatic amputation within the 12 months prior to screening, or an active skin ulcer, osteomyelitis, gangrene, or critical ischemia of the lower extremity within 6 months of screening²

Methodology^{1, 11}

• The study consisted of a screening and run-in period, a double-blind treatment period (INVOKANA 100 mg vs. PBO), and a 30-day post treatment (after study completion or permanent discontinuation of study drug) follow-up. The total duration of the study was planned to be approximately 5.5 years.
• Patients meeting eligibility criteria at screening underwent a 2-week, single-blind, PBO run-in period, and were deemed eligible for randomization if they took >80% of the run-in study medication.
• Patients were randomly assigned in a 1:1 ratio to either INVOKANA 100 mg or matching PBO using randomly permuted blocks with stratification by baseline eGFR categories (30 to <45, 45 to <60, and 60 to <90 mL/min/1.73 m²).
  • Patients were instructed to take their study treatment once daily before the first meal of the day, with a log maintained of study capsule counts.²
  • Administration was to continue (even if eGFR drops below 30 mL/min/1.73 m²) until the completion of the study, the commencement of dialysis, kidney transplantation, incidence of diabetic ketoacidosis, pregnancy, or receipt of disallowed therapy.
  • INVOKANA 100 mg was selected due to the favorable benefit to risk profile compared to INVOKANA 300 mg in patients with CKD in the INVOKANA Phase 3 clinical development program.²
• Patients were followed up at weeks 3, 13, and 26, and then with telephone calls and in-clinic visits in 13-week intervals.
• Use of other therapies for glycemic management and CV risk factor control was recommended in accordance with clinical practice guidelines.
• Due to an increased risk of lower limb amputations in another INVOKANA study,¹⁴ an amendment was made to the protocol in May 2016, requiring investigators to examine the feet of participants at each visit, and temporarily interrupt treatment in participants who develop any active condition associated with amputation (i.e., lower extremity infection, skin ulcer, osteomyelitis, gangrene, or critical limb ischemia) until the condition has resolved. Restarting medication was encouraged after careful consideration of benefits and risks.^{1, 2, 11}

Outcomes^{1, 11}

• The primary outcome was the composite of:
  • ESKD (defined as chronic dialysis for >30 days, renal transplantation, or eGFR <15 mL/min/1.73m² sustained for >30 days)
  • Doubling of serum creatinine from baseline average sustained for >30 days
  • Death due to renal or CV disease
• Secondary outcomes were planned for sequential hierarchical testing. If INVOKANA was superior over PBO for reducing the risk of the primary efficacy endpoint, the treatment effects in the secondary endpoints would be tested subsequently. Statistical significance was required before testing the next hypothesis in the hierarchical test procedure in the following order:²
  • Composite of CV death or hospitalized heart failure (HHF)
  • Composite of Major Adverse Cardiovascular Events (MACE), comprised of CV death, nonfatal myocardial infarction (MI), or nonfatal stroke
  • Hospitalized heart failure (HHF)
  • Renal-specific composite of ESKD, dSCr, or renal death
  • CV death
  • All-cause mortality
  • Broader CV composite of CV death, nonfatal MI, nonfatal stroke, HHF, or hospitalized unstable angina
• Prespecified exploratory endpoints included: composite endpoint of ESKD, death from renal or CV causes; ESKD; dSCr; renal death; fatal and nonfatal MI; fatal and nonfatal stroke; hospitalized unstable angina; change in eGFR over time; change in albuminuria over time.² A post hoc exploratory outcome was reported for the composite of dialysis, kidney transplant, or renal death.
• All AEs were collected from randomization through 30 days after the last dose of study drug. Some AEs of interest to be collected include: all malignancies, fatal pancreatitis, hemorrhagic/necrotizing pancreatitis, nephrotoxicity/acute kidney injury, venous thromboembolic events, fractures, diabetic ketoacidosis, and amputation. ²
• All renal and CV events that are components in the primary and secondary composite endpoints of the study and all key safety outcomes (bone fractures, pancreatitis, ketoacidosis, renal cell carcinoma [RCC]) were adjudicated by independent blinded adjudication committees.²

Statistical Analysis^{1, 11}

• The CREDENCE study was designed to be an event-driven study with a projected duration of ~5.5 years to accrue an estimated occurrence of 844 events required to provide >90% power for a detection of a 20% relative risk reduction for the primary composite outcome for an alpha level of 0.045.²
• The primary and secondary outcomes were analyzed using a stratified Cox proportional hazard model with stratification of the baseline eGFR based on the intention-to-treat analysis set. Secondary outcomes were tested following a hierarchical sequence.
• A single interim analysis was planned to be conducted by an independent data monitoring committee once 405 confirmed primary efficacy endpoints were accrued. Early termination of the study would be considered if clear evidence of benefit was observed with the primary composite endpoint meeting a P<0.01 and the additional composite of ESKD, renal death, and CV death meeting a P<0.025.^{2, 3}
  • If the study was to stop at the interim analysis, the primary outcome would be tested at a two-sided level of 0.022 and the secondary outcomes at a two-sided level of 0.038, to account for type 1 error inflation.
  • Safety analysis used the on-treatment analysis dataset (all treated participants through 30 days after last study drug dose), and selected AEs of interest were also analysed using the on-study analysis dataset (all treated patients through the end of the trial, i.e., global trial end date).
  • The numbers needed to treat (NNT) to prevent one event over 2.5 years were calculated as the reciprocal of the between-treatment difference in cumulative incidence at 2.5 years based on Kaplan-Meier curve.

Baseline Characteristics^{1, 4, 11}

• A total of 4401 participants were randomized from 690 sites across 34 countries between March 2014 and May 2017 in the intention-to-treat (ITT) analysis set. There were 4 participants that were not dosed, leading to 4397 participants in the on-treatment and on-study analysis sets.
• Baseline characteristics were similar between the INVOKANA and PBO groups. These included a mean age of 63 years, 33.9% female participants, mean duration of T2DM of 15.8 years; mean HbA1c of 8.3%, mean eGFR of 56.2 mL/min/1.73m², median UACR of 927 mg/g.²
  • In the total population, 50.4% had prior CV disease, 14.8% had a history of heart failure, and 5.3% had a history of amputation.
• Across both treatment groups, the mean exposure to study drug was 115 weeks.
• For more information related to the CREDENCE baseline characteristics, please refer to Table: CREDENCE: Baseline Characteristics

Results

By July 2018, the number of confirmed primary endpoints to trigger the planned interim analysis had been accrued and the prespecified efficacy criteria for early cessation had been achieved. The independent data monitoring committee advised the Steering Committee to end the CREDENCE study early after a median follow-up duration of 2.62 years (range 0.02-4.53 years).¹

• A total of 4361 (99.1%) of participants were followed until the study completion for clinical and safety endpoints. Final vital status was collected in 99.9% of participants.¹
• The most frequent reason for study discontinuation was an AE (12% and 13% of INVOKANA and PBO groups, respectively).¹

Primary Composite Outcome

• In the CREDENCE study, INVOKANA significantly reduced the rates of the composite outcome of ESKD, dSCr, or renal or CV death (43.2 and 61.2 per 1000 PY in the INVOKANA and PBO arms, respectively), resulting in a 30% relative risk reduction (RRR) (hazard ratio [HR], 0.70; 95% confidence interval (CI), 0.59–0.82; P=0.00001; NNT=22 over 2.5 years for the primary composite endpoint). See Figure: Time to First Occurrence: Primary Composite Outcome and Table: Number Needed to Treat.¹
  • The effects were consistent across regions and all prespecified subgroups.¹ See tables: Primary Outcome based on Screening eGFR and Albuminuria and Primary Outcome: Demographic and Risk Factor Subgroups

Individual Components of the Primary Composite Outcome

• • The event rate of ESKD was 20.4 and 29.4 per 1000 patient years (HR 0.68; 95% CI: 0.54-0.86) in the INVOKANA and PBO groups, respectively; P=0.002).¹ See Figure: Time to First Occurrence: End Stage Kidney Disease and Table: Number Needed to Treat

• • The event rate of dSCR was 20.7 and 33.8 per 1000 patient years (HR 0.60; 95% CI: 0.48-0.76; P<0.001) in the INVOKANA and PBO groups, respectively.¹
  • The event rate of renal death was 0.3 and 0.9 per 1000 patient years in the INVOKANA (n=2) and PBO (n=5) groups, respectively. Since there were <10 events of renal death, a HR and 95% CI were not reported.¹
  • The event rate of CV death was 19.0 and 24.4 per 1000 patient years (HR 0.78; 95% CI: 0.61-1.00) in the INVOKANA and PBO groups, respectively.¹ See Figure: Time to First Occurrence: Cardiovascular Death.

• See Table: Summary of Efficacy Results.

Secondary and Exploratory Outcomes

• In the CREDENCE study, INVOKANA significantly reduced the risk of:¹
  • The composite of CV death and hospitalized heart failure (RRR 31%; HR 0.69; 95% CI 0.57-0.83; p<0.001)
  • MACE (RRR 20%; HR 0.80; 95% CI 0.67-0.95; p=0.01)
  • HHF (RRR 39%; HR 0.61; 95% CI 0.47-0.80; p<0.001)
  • Renal-specific composite of ESKD, dSCr, or renal death (RRR 34%; HR 0.66; 95% CI 0.53-0.81; p<0.001)
  • See figure: Rates of CV death or HHF / MACE / HHF / Renal composite in CREDENCE and Table: Number Needed to Treat.
• The risk of CV death was not found to be statistically significant (p=0.0502), therefore, due to the hierarchical testing sequence, the remaining two secondary endpoints were not formally tested.¹
• See Table: Summary of Efficacy Results.

Subgroup Analyses

Subgroup analyses were conducted to evaluate the primary composite outcome (ESKD, dSCr, or renal or CV death) according to screening eGFR and UACR at baseline.¹ See figure: Primary Outcome based on Screening eGFR and Albuminuria

Subgroup analyses were conducted to evaluate the primary composite outcome (ESKD, dSCr, or renal or CV death) according to demographic and risk factors at baseline.¹ See figure: Primary Outcome: Demographic and Risk Factor Subgroups

A secondary analysis of primary, prespecified secondary composites, and safety outcomes was conducted using Cox proportional hazards regression within each screening eGFR stratum. At screening, 1313 (30%), 1279 (29%), and 1809 (41%) of participants had an eGFR 30-<45, 45-<60, and 60-<90mL/min/1.73m². INVOKANA reduced the primary outcome (composite of ESKD, sustained doubling serum creatinine (SCr) or renal death), CV outcomes and serious adverse events with no impact on fractures or amputations. The effect of INVOKANA did not differ between eGFR subgroups (all P-interaction >0.11) for the primary and secondary outcomes. The efficacy outcomes of INVOKANA were individually significant in people with a screening eGFR 30-<45ml/min/1.73m² for the primary composite, renal composite and composite of CV death or hospitalization for heart failure (95% CI upper limit <1.00). There is evidence that there was a difference in eGFR subgroups for the adverse events, volume depletion and osmotic diuresis, which were less common in INVOKANA (P-interaction = 0.01 and 0.03, respectively). See Figure: Secondary Analysis: Renal, Cardiovascular and Select Safety Outcomes by Screening eGFR

Neuen et al (2020)¹⁶ conducted a secondary analysis of cardiovascular, renal and safety outcomes by baseline loop diuretic use in the CREDENCE trial population. At baseline, 479 patients in the INVOKANA group were on a loop diuretic and 476 patients in the placebo group. At baseline, both INVOKANA and placebo groups had 1723 patients not on a loop diuretic. The effect of INVOKANA on the primary composite outcome of end-stage kidney disease, doubling of serum creatinine, and renal or CV death was consistent across the overall, non-loop, and loop diuretic population (interaction, P=0.18). The effect of INVOKANA on safety outcomes, including all serious adverse events, volume depletion, renal-related adverse events, acute kidney injury, hyperkalemia, and amputation, was consistent among all subgroups.

Arnott et al (2020)¹⁷ conducted a secondary analysis of the pre-specified, hierarchical, secondary outcome hospitalization for heart failure or CV death by patient baseline characteristics including age (>65 or <65), sex (M/F), history of cardiovascular disease (Y/N), history of heart failure or corresponding New York Heat Association (NYHA) functional classification (no HF or NYHA I-III), estimated glomerular filtration rate (30 to <45, 45 to <60, and 60 to <90mL/min/1.73m²), UACR (>1000 or <1000), and baseline diuretic use (Y/N). In the CREDENCE trial, 432 patients experienced an HHF/CV event over a median follow-up of 2.6 years. The effect of INVOKANA on CV death or HHF did not show significant differences by these baseline subgroups (all P interactions >0.2 for HRs).

Li et al (2020)¹⁸ conducted a secondary analysis of the pre-specified, hierarchical, secondary outcome of hospitalization for heart failure by evaluating the effect of INVOKANA on the first, subsequent, and total HHF events. In the CREDENCE trial, 326 HHF events occurred, of those, 230 (71%) were first events, and an additional 96 (29%) were recurrent. Of the first events, 89 events were in the INVOKANA arm and 141 were in the PBO arm (p<0.001). Of the additional HHF events, 38 were in the INVOKANA arm and 58 were in the PBO arm (p<0.001). INVOKANA decreased the relative risk of total HHF events by 36% (RR 0.64; 95% [CI], 0.56-0.73) and the relative risk of a first HHF event by 39% (RR 0.61; 95% [CI], 0.47-0.80). The NNT to prevent a first HHF event was 46. When considering total HF events, the NNT to prevent an HHF event was 32.

Safety Outcomes

• The overall rates of AEs and serious AEs, were similar between the treatment groups.¹ See Table: Summary of Safety Results.
• There was no imbalance observed in the risk of lower limb amputation between the INVOKANA vs PBO arms (12.3 vs 11.2 per 1000 PY; HR 1.11; 95% CI: 0.79-1.56).¹
• Rates of fracture were similar between the INVOKANA and PBO treatment arms (11.8 vs 12.1 per 1000 PY; HR 0.98; 95% CI: 0.70-1.37).¹
• All renal-related AEs taken together occurred at a significantly higher rate with PBO compared to INVOKANA (79.1 vs 57.1 per 1000 PY; HR 0.71; 95% CI 0.61-0.82). Hyperkalemia events and acute kidney injury events were reported more in the PBO arm compared to the INVOKANA arm (hyperkalemia: 36.9 vs 29.7 per 1000 PY; HR 0.80; 95% CI: 0.65-1.00; acute kidney injury: 20.0 vs 16.9; HR 0.85; 95% CI: 0.64-1.13).¹
  • Renal-related preferred terms included: AKI, acute phosphate nephropathy, acute prerenal failure, anuria, azotaemia, blood creatinine increased, blood urea increased, continuous haemodiafiltration, dialysis, glomerular filtration rate decreased, haemodialysis, haemofiltration, hypercreatininaemia, neonatal anuria, nephritis, nephropathy toxic, oliguria, peritoneal dialysis, prerenal failure, renal failure, renal failure acute, renal failure neonatal, renal impairment, renal impairment neonatal. ²⁷
• Diabetic ketoacidosis rates were low overall, but higher in participants treated with INVOKANA (2.2 vs 0.2 events per 1000 PY; HR 10.8; 95% CI 1.39-83.65).¹
• Rates of cancer, including renal cell carcinoma, were similar in both treatment arms.¹

Intermediate outcomes

• The INVOKANA arm had a greater least-squares [LS] mean reduction in HbA1c at 13 weeks (-0.31%; 95% CI -0.26%, -0.37%) and a greater overall mean reduction throughout the study (-0.25%; 95% CI -0.20%, -0.31%) vs PBO.¹
  • Investigators were able to adjust background antihyperglycemic agent therapy during the course of the study.¹¹
• The INVOKANA arm had a greater mean change in systolic blood pressure (-3.30 mmHg; 95% CI -2.73, -3.87 mmHg), diastolic blood pressure (-0.95 mmHg; 95% CI -0.61, -1.28 mmHg), and body weight (-0.80 kg; 95% CI -0.69, -0.92 kg).¹
• The geometric mean of UACR was 31% lower on average during follow-up in the INVOKANA arm (95% CI -26%, -35%).¹
• The overall decline in eGFR slope over time was attenuated with INVOKANA vs PBO, with a difference of 1.52 mL/min/1.73m²/year (95% CI -1.11, -1.93 mL/min/1.73m²/year).¹
  • Patients treated with INVOKANA had a greater acute reduction in eGFR (difference of -3.17 mL/min/1.73m²/year), but a slower decline in eGFR thereafter (difference of 2.74 mL/min/1.73m²/year) compared to PBO. See figure: Effects on eGFR

Oshima et al (2020)¹⁹ conducted a secondary analysis on potential predictors of the acute decline in eGFR following initiation of INVOKANA and to assess the association between the initial decline with long-term eGFR trajectory and safety outcomes. An initial drop in eGFR was categorized as an acute eGFR decrease (>10%), acute modest eGFR decrease (>0 to 10%), or an acute eGFR increase (>0%), at week 3 of the study. An acute drop in eGFR occurred in 45% of patients, followed by an acute modest drop (28%) and an acute increase (27%).

• Compared to placebo, INVOKANA increased the likelihood of an acute eGFR drop >10% in all subgroups examined.
• Among the categories of eGFR groups, treatment discontinuation was consistent among INVOKANA and PBO groups.
• For each category of eGFR decline (or increase), the overall trajectory of eGFR decline was slower in the INVOKANA arm compared to PBO.
• Adverse events were consistent regardless of initial drop in eGFR among INVOKANA and PBO arms.
• There was an increased risk of a renal-related adverse events in patients experiencing an acute >30% decrease in eGFR.

A secondary analysis of potential mediators of INVOKANA’s effect on the outcome of heart failure and CV death was conducted by Li et al (2020).²⁰ Three criteria were required to perform a mediation analysis: INVOKANA must significantly reduce hospitalization for heart failure or CV death, potential mediators must be significantly changed by INVOKANA treatment compared to placebo, and levels of post-randomization potential mediators must be associated with the risk of HHF/CV death.

• In the mediation analysis, the post-randomization factors associated with the impact of INVOKANA on HHF/CV death were improved volume status/haematopoiesis and reduced albuminuria
  • Volume status & haematopoiesis % mediation (95% [CI]) in univariate models:
    • Hematocrit (%), 23.98% mediation (95% [CI], 8.16, 81.86)
    • Hemoglobin (g/dL), 32.08% mediation (95% [CI], 13.09, 100.35)
    • Erythrocytes (10¹² cells/L), 27.04% mediation (95% [CI], 10.36, 98.75)
• Additional post-randomization associations were reduced systolic BP and uric acid levels:
  • Systolic BP (mmHg), 10.31% mediation (95% [CI], 4.17, 28.06)
  • Serum urate (umol/L), 7.02% mediation (95% [CI], 2.78, 19.84)

Primary and Secondary Prevention

Analyses for the primary outcome in primary and secondary prevention cohorts were prespecified and originally presented at the American Diabetes Association 79^th Annual Scientific Sessions in June 2019.²¹ Patients were classified as secondary prevention if they had a history of coronary, cerebrovascular, or peripheral vascular disease at baseline. All other participants belonged to the primary prevention cohort. Analyses for the primary outcome in primary and secondary prevention cohorts were prespecified, but additional analyses were post hoc.⁵

A total of 2181 (49.6%) participants had no history of documented cardiovascular disease at entry and were in the primary prevention group, and 2220 (50.4%) participants were in the secondary prevention group. Primary prevention participants were younger (61.4 vs 64.6 years), more often female (36.6% vs 31.3%), and Asian (24.4% vs 15.5%), with a shorter duration of diabetes (15.2 vs 16.4 years) compared to secondary prevention participants. Primary and secondary prevention participants had similar mean eGFR (56.8 vs 55.5 mL/min/1.73 m²) and median UACR (943 vs 903 mg/g).⁵

INVOKANA significantly reduced efficacy outcomes, with no evidence of heterogeneity in the primary and secondary prevention groups (all interaction P values not significant).⁵ See Table: CREDENCE Efficacy Outcomes in the Primary, Secondary, and Overall Populations.

Summary of Safety in Primary and Secondary Prevention Cohorts

• The overall rates of AEs and serious AEs, were similar between subjects in the primary and secondary prevention cohorts⁵
  • The overall rate of AEs was significantly less for INVOKANA in both the primary (HR, 0.88; 95% CI, 0.80–0.96) and secondary (HR, 0.87; 95% CI, 0.80–0.96) prevention groups; with no evidence of heterogeneity between the sub-groups (P-interaction 0.97).
  • The overall rate of SAEs was less for INVOKANA in both the primary (HR, 0.84; 95% CI, 0.72–0.99) and secondary (HR, 0.89; 95% CI, 0.78–1.02) prevention groups; with no evidence of heterogeneity between the sub-groups (P-interaction 0.62).
• There was no imbalance in the risk of lower limb amputation between INVOKANA vs PBO in the primary HR, 1.09; 95% CI, 0.48–2.47) and secondary (HR, 1.11; 95% CI, 0.76–1.61) prevention groups; with no evidence of heterogeneity between the sub-groups (P-interaction 0.98).⁵
• The was no imbalance in the risk of fracture between INVOKANA vs PBO in the primary (HR, 1.05; 95% CI, 0.61–1.79) and secondary (HR, 0.93; 95% CI, 0.60–1.44) prevention groups; with no evidence of heterogeneity between the sub-groups (P-interaction 0.74).⁵

Literature Search

A literature search of MEDLINE^®, EMBASE^®, BIOSIS Previews^®, DERWENT^® (and/or other resources, including internal/external databases) was conducted on 10 March 2023.