Browse the corpus

Empagliflozin in Patients with Chronic Kidney Disease. BACKGROUND: The effects of empagliflozin in patients with chronic kidney disease who are at risk for disease progression are not well understood. The EMPA-KIDNEY trial was designed to assess the effects of treatment with empagliflozin in a broad range of such patients. METHODS: We enrolled patients with chronic kidney disease who had an estimated glomerular filtration rate (eGFR) of at least 20 but less than 45 ml per minute per 1.73 m2 of body-surface area, or who had an eGFR of at least 45 but less than 90 ml per minute per 1.73 m2 with a urinary albumin-to-creatinine ratio (with albumin measured in milligrams and creatinine measured in grams) of at least 200. Patients were randomly assigned to receive empagliflozin (10 mg once daily) or matching placebo. The primary outcome was a composite of progression of kidney disease (defined as end-stage kidney disease, a sustained decrease in eGFR to <10 ml per minute per 1.73 m2, a sustained decrease in eGFR of ≥40% from baseline, or death from renal causes) or death from cardiovascular causes. RESULTS: A total of 6609 patients underwent randomization. During a median of 2.0 years of follow-up, progression of kidney disease or death from cardiovascular causes occurred in 432 of 3304 patients (13.1%) in the empagliflozin group and in 558 of 3305 patients (16.9%) in the placebo group (hazard ratio, 0.72; 95% confidence interval [CI], 0.64 to 0.82; P<0.001). Results were consistent among patients with or without diabetes and across subgroups defined according to eGFR ranges. The rate of hospitalization from any cause was lower in the empagliflozin group than in the placebo group (hazard ratio, 0.86; 95% CI, 0.78 to 0.95; P = 0.003), but there were no significant between-group differences with respect to the composite outcome of hospitalization for heart failure or death from cardiovascular causes (which occurred in 4.0% in the empagliflozin group and 4.6% in the placebo group) or death from any cause (in 4.5% and 5.1%, respectively). The rates of serious adverse events were similar in the two groups. CONCLUSIONS: Among a wide range of patients with chronic kidney disease who were at risk for disease progression, empagliflozin therapy led to a lower risk of progression of kidney disease or death from cardiovascular causes than placebo. (Funded by Boehringer Ingelheim and others; EMPA-KIDNEY ClinicalTrials.gov number, NCT03594110; EudraCT number, 2017-002971-24.).

Details of rationale of the present study and trial design have been reported previously.14,15 Our study was designed and led by a Steering Committee that included representatives from the central coordinating office at the University of Oxford, each recruiting region, the sponsor (Boehringer Ingelheim), and other clinical and statistical experts. An independent Data Safety Monitoring Board (DSMB, known as the Data Monitoring Committee) was responsible for regular review of unblinded data to ensure participant safety, and for a Protocol-defined formal interim analysis for efficacy. The Protocol and the Data Analysis Plan (DAP) are available with the full text of this article at NEJM.org and at empakidney.org. The trial was conducted at 241 centers in eight countries. Regulatory authorities and ethics committees for each center approved the trial.

ocol-defined formal interim analysis for efficacy. The Protocol and the Data Analysis Plan (DAP) are available with the full text of this article at NEJM.org and at empakidney.org. The trial was conducted at 241 centers in eight countries. Regulatory authorities and ethics committees for each center approved the trial. Adults with a race-adjusted CKD-EPI16 eGFR of at least 20 but less than 45 ml per minute per 1.73 m2 of body-surface area (irrespective of level of albuminuria); or an eGFR of at least 45 but less than 90 ml per minute per 1.73 m2 with a urinary ACR of at least 200 mg/g at the screening visit were eligible provided they were prescribed a clinically appropriate dose of single-agent RAS-inhibitor. Patients could also be included if an investigator judged that such treatment was either not tolerated or not indicated. Patients with or without diabetes were eligible. Those with polycystic kidney disease or a kidney transplant were excluded. Full details of eligibility criteria are provided in the Protocol (see Supplementary material available at NEJM.org). All participants provided written informed consent. All eligible participants entered a pre-randomization run-in phase and were provided with a 15-week supply of once daily placebo tablets. During this time, local investigators reviewed screening data, assessed current RAS-inhibitor use, and approved potential participants for later randomization. Throughout the trial, clinical responsibility for participants remained with their local doctors.

provided with a 15-week supply of once daily placebo tablets. During this time, local investigators reviewed screening data, assessed current RAS-inhibitor use, and approved potential participants for later randomization. Throughout the trial, clinical responsibility for participants remained with their local doctors. After completing at least 6 weeks of run-in, willing participants had central samples of blood and urine collected for central analysis and storage, and were randomly allocated to receive empagliflozin (10 mg once daily) or matching placebo using minimized randomization with a 10% stochastic element.17 At follow-up visits, participants provided information on their kidney status (i.e., any dialysis treatment or receipt of a kidney transplant), adherence to study treatment (with reasons for stopping) and details of concomitant medication. They were also asked in a structured interview about any serious adverse events (and Protocol-specified non-serious adverse events), underwent clinical measurements of blood pressure and weight, and had blood collected for local safety assessments of creatinine, liver function and potassium. Blood samples and, at selected visits, urine samples were sent to the central laboratory for efficacy analyses and archiving. Adaptations due to coronavirus-19 and assay methods are provided in the Supplementary Appendix.

and had blood collected for local safety assessments of creatinine, liver function and potassium. Blood samples and, at selected visits, urine samples were sent to the central laboratory for efficacy analyses and archiving. Adaptations due to coronavirus-19 and assay methods are provided in the Supplementary Appendix. The prespecified primary outcome was the first occurrence of the composite outcome of kidney disease progression or cardiovascular death. Kidney disease progression included end-stage kidney disease (ESKD), defined as commencing maintenance dialysis or receipt of a kidney transplant; a sustained decline in eGFR to less than 10 ml per minute per 1.73 m2; a sustained decline in eGFR of at least 40% from baseline; or renal death. The term ‘sustained’ was defined as either as measured at two consecutive scheduled study follow-up visits at least 30 days apart, or as measured at the final follow-up visit or the last scheduled visit before death (or withdrawal of consent or loss to follow-up). Central laboratory serum creatinine measurements were used to estimate GFR, with local laboratory creatinine measurements used when central results were missing. The prespecified key secondary outcomes were hospitalization for heart failure or cardiovascular death; all-cause hospitalizations (first and subsequent, combined); and death from any cause. The other secondary outcomes were the components of the primary outcome: kidney disease progression; death from cardiovascular causes; and ESKD or death from cardiovascular causes. Details of the tertiary, safety and laboratory assessments and planned exploratory assessments are in the Data Analysis Plan in the Protocol, available at NEJM.org. Key subgroup analyses of the primary outcome were prespecified to be by diabetes status, eGFR, and urinary ACR at baseline. All deaths, potential hospitalizations for heart failure, myocardial infarction, stroke, liver injury, ketoacidosis, lower limb amputation, acute kidney injury and serious genital infections were subject to adjudication by blinded clinicians using prespecified definitions and source documents collected from sites. Clinical outcome definitions are provided in the Supplementary Appendix.

infarction, stroke, liver injury, ketoacidosis, lower limb amputation, acute kidney injury and serious genital infections were subject to adjudication by blinded clinicians using prespecified definitions and source documents collected from sites. Clinical outcome definitions are provided in the Supplementary Appendix. Follow-up was planned until at least 1070 participants had experienced a first primary outcome, in order to provide 90% power at two-sided P=0.05 to detect an 18% relative reduction in risk.14 The Protocol specified that a single formal interim analysis for efficacy should be conducted when 150 participants had experienced a first ESKD event. Based on the number of primary outcomes at the time (n=624), the two conditions for recommending an early stop for efficacy were prespecified as a hazard ratio for the primary outcome and the other secondary outcome of ESKD or death from cardiovascular causes of <0.778, with two-sided P values of <0.0017 and <0.05, respectively (see Protocol for details).

es at the time (n=624), the two conditions for recommending an early stop for efficacy were prespecified as a hazard ratio for the primary outcome and the other secondary outcome of ESKD or death from cardiovascular causes of <0.778, with two-sided P values of <0.0017 and <0.05, respectively (see Protocol for details). All analyses were performed according to the intention-to-treat principle and included data from all randomized participants including information collected between the formal interim analysis and final follow-up visits.18–20 A Cox proportional hazards regression model adjusted for baseline variables specified in the minimization algorithm (age, sex, prior diabetes, eGFR, urinary ACR, and region) was used to estimate the hazard ratio and 95% confidence intervals (CI) for empagliflozin versus placebo for time-to-event analyses.21 Key secondary outcomes were prespecified to be adjusted for multiple testing using the Hochberg “step-up” procedure with a family-wise error rate of 0.029. For the outcome of first and subsequent all-cause hospitalizations, a semi-parametric joint frailty model was used.22 Effects of empagliflozin on the tertiary and exploratory outcomes based on annual rate of change in eGFR were assessed with shared parameter models.23 Further statistical details are provided in supplementary statistical methods and the pre-specified DAP at NEJM.org. The original full database is held and analyses performed by the Nuffield Department of Population Health at the University of Oxford using SAS software, version 9.4 (SAS Institute). The Steering Committee was responsible for manuscript writing and the decision to publish.

Adults with a race-adjusted CKD-EPI16 eGFR of at least 20 but less than 45 ml per minute per 1.73 m2 of body-surface area (irrespective of level of albuminuria); or an eGFR of at least 45 but less than 90 ml per minute per 1.73 m2 with a urinary ACR of at least 200 mg/g at the screening visit were eligible provided they were prescribed a clinically appropriate dose of single-agent RAS-inhibitor. Patients could also be included if an investigator judged that such treatment was either not tolerated or not indicated. Patients with or without diabetes were eligible. Those with polycystic kidney disease or a kidney transplant were excluded. Full details of eligibility criteria are provided in the Protocol (see Supplementary material available at NEJM.org). All participants provided written informed consent.

All eligible participants entered a pre-randomization run-in phase and were provided with a 15-week supply of once daily placebo tablets. During this time, local investigators reviewed screening data, assessed current RAS-inhibitor use, and approved potential participants for later randomization. Throughout the trial, clinical responsibility for participants remained with their local doctors. After completing at least 6 weeks of run-in, willing participants had central samples of blood and urine collected for central analysis and storage, and were randomly allocated to receive empagliflozin (10 mg once daily) or matching placebo using minimized randomization with a 10% stochastic element.17 At follow-up visits, participants provided information on their kidney status (i.e., any dialysis treatment or receipt of a kidney transplant), adherence to study treatment (with reasons for stopping) and details of concomitant medication. They were also asked in a structured interview about any serious adverse events (and Protocol-specified non-serious adverse events), underwent clinical measurements of blood pressure and weight, and had blood collected for local safety assessments of creatinine, liver function and potassium. Blood samples and, at selected visits, urine samples were sent to the central laboratory for efficacy analyses and archiving. Adaptations due to coronavirus-19 and assay methods are provided in the Supplementary Appendix.

The prespecified primary outcome was the first occurrence of the composite outcome of kidney disease progression or cardiovascular death. Kidney disease progression included end-stage kidney disease (ESKD), defined as commencing maintenance dialysis or receipt of a kidney transplant; a sustained decline in eGFR to less than 10 ml per minute per 1.73 m2; a sustained decline in eGFR of at least 40% from baseline; or renal death. The term ‘sustained’ was defined as either as measured at two consecutive scheduled study follow-up visits at least 30 days apart, or as measured at the final follow-up visit or the last scheduled visit before death (or withdrawal of consent or loss to follow-up). Central laboratory serum creatinine measurements were used to estimate GFR, with local laboratory creatinine measurements used when central results were missing. The prespecified key secondary outcomes were hospitalization for heart failure or cardiovascular death; all-cause hospitalizations (first and subsequent, combined); and death from any cause. The other secondary outcomes were the components of the primary outcome: kidney disease progression; death from cardiovascular causes; and ESKD or death from cardiovascular causes. Details of the tertiary, safety and laboratory assessments and planned exploratory assessments are in the Data Analysis Plan in the Protocol, available at NEJM.org. Key subgroup analyses of the primary outcome were prespecified to be by diabetes status, eGFR, and urinary ACR at baseline. All deaths, potential hospitalizations for heart failure, myocardial infarction, stroke, liver injury, ketoacidosis, lower limb amputation, acute kidney injury and serious genital infections were subject to adjudication by blinded clinicians using prespecified definitions and source documents collected from sites. Clinical outcome definitions are provided in the Supplementary Appendix.

Follow-up was planned until at least 1070 participants had experienced a first primary outcome, in order to provide 90% power at two-sided P=0.05 to detect an 18% relative reduction in risk.14 The Protocol specified that a single formal interim analysis for efficacy should be conducted when 150 participants had experienced a first ESKD event. Based on the number of primary outcomes at the time (n=624), the two conditions for recommending an early stop for efficacy were prespecified as a hazard ratio for the primary outcome and the other secondary outcome of ESKD or death from cardiovascular causes of <0.778, with two-sided P values of <0.0017 and <0.05, respectively (see Protocol for details).

From February 2019 to April 2021, 8544 potential participants attended a screening visit from which 8184 (96%) entered the pre-randomization run-in and 6609 were randomized (Supplementary Appendix, Fig. S1). At randomization, mean age was 63.8 years, 33% of participants were women and 54% did not have diabetes (Table 1), and broadly representative of patients with CKD at risk of progression (Table S1). Mean±standard deviation eGFR was 37.3±14.5 ml per minute per 1.73 m2 and 35% had an eGFR less than 30 ml per minute per 1.73 m2. Median urinary ACR was 329 mg/g, and 48% had a urinary ACR below 300 mg/g (Tables 1 and S2). On March 07 2022, the independent DSMB reported that based on 624 first primary outcomes, both conditions for stopping early for efficacy were met at the formal interim analysis. Follow-up was completed on July 05 2022, at which time median follow-up was 2.0 years (interquartile range, 1.5 to 2.4 years). In all, 6552 participants (99.1%) were alive and completed final follow-up or had died during follow-up. Vital status was missing for 18 (0.3%) participants, and 39 participants (0.6%) withdrew consent (Fig. S1). All eligible events were adjudicated.

dian follow-up was 2.0 years (interquartile range, 1.5 to 2.4 years). In all, 6552 participants (99.1%) were alive and completed final follow-up or had died during follow-up. Vital status was missing for 18 (0.3%) participants, and 39 participants (0.6%) withdrew consent (Fig. S1). All eligible events were adjudicated. At 12 months of follow-up (the approximate midpoint), 2909 [89.6%] of the empagliflozin group and 2924 [90.3%] of the placebo group reported taking most (i.e. >80%) of their study treatment. By final follow-up, study treatment was discontinued by 557 (16.9%) surviving participants allocated empagliflozin, and by 640 (19.4%) allocated placebo. This included 18 (0.5%) participants in the empagliflozin group and 31 (0.9%) in the placebo group who started treatment with an open-label SGLT2 inhibitor. Table S3 provides details of the reasons for discontinuation. The primary outcome of kidney disease progression or death from cardiovascular causes occurred in 432 participants (13.1%) in the empagliflozin group and 558 participants (16.9%) in the placebo group (hazard ratio, 0.72; 95% CI 0.64 to 0.82; P<0.001) (Fig. 1).

At 12 months of follow-up (the approximate midpoint), 2909 [89.6%] of the empagliflozin group and 2924 [90.3%] of the placebo group reported taking most (i.e. >80%) of their study treatment. By final follow-up, study treatment was discontinued by 557 (16.9%) surviving participants allocated empagliflozin, and by 640 (19.4%) allocated placebo. This included 18 (0.5%) participants in the empagliflozin group and 31 (0.9%) in the placebo group who started treatment with an open-label SGLT2 inhibitor. Table S3 provides details of the reasons for discontinuation. The primary outcome of kidney disease progression or death from cardiovascular causes occurred in 432 participants (13.1%) in the empagliflozin group and 558 participants (16.9%) in the placebo group (hazard ratio, 0.72; 95% CI 0.64 to 0.82; P<0.001) (Fig. 1). After controlling the family-wise error rate for the three key secondary outcomes, there were significantly fewer first and subsequent hospitalizations from any cause in the empagliflozin group (24.8 versus 29.2 hospitalizations per 100 patient years: hazard ratio, 0.86; 95% CI 0.78 to 0.95, P=0.003) (Tables 2 and S4). There was no statistically significant effect on the composite of hospitalization for heart failure or death from cardiovascular causes (hazard ratio, 0.84; 95% CI 0.67 to 1.07; P=0.15), or on death from any cause (hazard ratio, 0.87; 95% CI 0.70 to 1.08; P=0.21) (Table 2 and Fig. S2).

0.95, P=0.003) (Tables 2 and S4). There was no statistically significant effect on the composite of hospitalization for heart failure or death from cardiovascular causes (hazard ratio, 0.84; 95% CI 0.67 to 1.07; P=0.15), or on death from any cause (hazard ratio, 0.87; 95% CI 0.70 to 1.08; P=0.21) (Table 2 and Fig. S2). The hazard ratios for the comparison of empagliflozin with placebo on kidney disease progression and for death from cardiovascular causes separately were 0.71 (95% CI 0.62 to 0.81) and 0.84 (95% CI 0.60 to 1.19), respectively (Fig. S3). For the composite of ESKD or death from cardiovascular causes, the hazard ratio was 0.73 (95% CI 0.59 to 0.89) (Tables 2 and S5). The effect of empagliflozin on the primary outcome was generally consistent across the prespecified subgroups. In particular, the benefits were consistent in patients with or without diabetes and regardless of eGFR at randomization. There was some evidence that the proportional risk reduction may be larger in those with higher urinary ACR (Figs. 2 and S4). Results were similar in prespecified exploratory subgroup analyses of the kidney disease progression outcome (Fig. S5).

patients with or without diabetes and regardless of eGFR at randomization. There was some evidence that the proportional risk reduction may be larger in those with higher urinary ACR (Figs. 2 and S4). Results were similar in prespecified exploratory subgroup analyses of the kidney disease progression outcome (Fig. S5). There was an acute drop in eGFR on commencing study treatment, followed by a slowing of the rate of annual decline. Overall, the between-group difference in total slope was 0.75 (95% CI 0.54 to 0.96) ml per minute per 1.73 m2 per year. For chronic slopes (i.e. from 2 months to final follow-up), there was a between-group difference of 1.37 (95% CI 1.16 to 1.59) ml per minute per 1.73 m2 per year (Figs. 3 and S6). Prespecified exploratory analyses by subgroups showed the rate of decline in the chronic slope was slower in the empgalfilozin group in all the key subgroups, including patients with low urinary ACR. Between group differences in rate of eGFR decline were larger in the subgroups of participant with faster rate of annual decline (i.e. patients with diabetes, higher eGFR, or higher baseline urinary ACR) (Fig. S7). There were no significant effects of empagliflozin on any specific cause of death (Figure S3), major cardiovascular events (hazard ratio, 0.93 95% CI 0.76-1.12), self-reported episodes of gout, or development of new-onset diabetes (Tables S5 and S6).

There was an acute drop in eGFR on commencing study treatment, followed by a slowing of the rate of annual decline. Overall, the between-group difference in total slope was 0.75 (95% CI 0.54 to 0.96) ml per minute per 1.73 m2 per year. For chronic slopes (i.e. from 2 months to final follow-up), there was a between-group difference of 1.37 (95% CI 1.16 to 1.59) ml per minute per 1.73 m2 per year (Figs. 3 and S6). Prespecified exploratory analyses by subgroups showed the rate of decline in the chronic slope was slower in the empgalfilozin group in all the key subgroups, including patients with low urinary ACR. Between group differences in rate of eGFR decline were larger in the subgroups of participant with faster rate of annual decline (i.e. patients with diabetes, higher eGFR, or higher baseline urinary ACR) (Fig. S7). There were no significant effects of empagliflozin on any specific cause of death (Figure S3), major cardiovascular events (hazard ratio, 0.93 95% CI 0.76-1.12), self-reported episodes of gout, or development of new-onset diabetes (Tables S5 and S6). Ketoacidosis occurred in 6 patients in the empagliflozin group versus 1 patient in the placebo group (0.09 versus 0.02 per 100 patient-years). Lower limb amputations occurred in 28 patients in the empagliflozin group and 19 in the placebo group (0.43 versus 0.29 per 100 patient-years). The incidence of serious urinary tract infections, hyperkalemia, acute kidney injuries, serious or symptomatic dehydrations, liver injuries, and bone fractures were broadly similar in each group (Tables 2 and S7). There was no apparent evidence that empagliflozin increased the incidence of serious adverse events overall, or in any particular MedDRA system organ class (Table S8).

ia, acute kidney injuries, serious or symptomatic dehydrations, liver injuries, and bone fractures were broadly similar in each group (Tables 2 and S7). There was no apparent evidence that empagliflozin increased the incidence of serious adverse events overall, or in any particular MedDRA system organ class (Table S8). There were reductions in weighted-average differences [standard error, SE] in mean body weight (-0.9 [0.1] kg) and blood pressure (systolic -2.6 [0.3] mmHg; diastolic -0.5 [0.2] mmHg), but no significant effect on glycated hemoglobin (Table S9). The geometric mean urinary ACR was reduced by 19% (95% CI 15% to 23%). Table S10 provides details of the observed increases in hematocrit and hemoglobin, and the absence of clinically relevant differences in blood calcium, phosphate, or sodium measured in a subset of participants at 18 months.

dian follow-up was 2.0 years (interquartile range, 1.5 to 2.4 years). In all, 6552 participants (99.1%) were alive and completed final follow-up or had died during follow-up. Vital status was missing for 18 (0.3%) participants, and 39 participants (0.6%) withdrew consent (Fig. S1). All eligible events were adjudicated. At 12 months of follow-up (the approximate midpoint), 2909 [89.6%] of the empagliflozin group and 2924 [90.3%] of the placebo group reported taking most (i.e. >80%) of their study treatment. By final follow-up, study treatment was discontinued by 557 (16.9%) surviving participants allocated empagliflozin, and by 640 (19.4%) allocated placebo. This included 18 (0.5%) participants in the empagliflozin group and 31 (0.9%) in the placebo group who started treatment with an open-label SGLT2 inhibitor. Table S3 provides details of the reasons for discontinuation.

The primary outcome of kidney disease progression or death from cardiovascular causes occurred in 432 participants (13.1%) in the empagliflozin group and 558 participants (16.9%) in the placebo group (hazard ratio, 0.72; 95% CI 0.64 to 0.82; P<0.001) (Fig. 1). After controlling the family-wise error rate for the three key secondary outcomes, there were significantly fewer first and subsequent hospitalizations from any cause in the empagliflozin group (24.8 versus 29.2 hospitalizations per 100 patient years: hazard ratio, 0.86; 95% CI 0.78 to 0.95, P=0.003) (Tables 2 and S4). There was no statistically significant effect on the composite of hospitalization for heart failure or death from cardiovascular causes (hazard ratio, 0.84; 95% CI 0.67 to 1.07; P=0.15), or on death from any cause (hazard ratio, 0.87; 95% CI 0.70 to 1.08; P=0.21) (Table 2 and Fig. S2). The hazard ratios for the comparison of empagliflozin with placebo on kidney disease progression and for death from cardiovascular causes separately were 0.71 (95% CI 0.62 to 0.81) and 0.84 (95% CI 0.60 to 1.19), respectively (Fig. S3). For the composite of ESKD or death from cardiovascular causes, the hazard ratio was 0.73 (95% CI 0.59 to 0.89) (Tables 2 and S5).

The effect of empagliflozin on the primary outcome was generally consistent across the prespecified subgroups. In particular, the benefits were consistent in patients with or without diabetes and regardless of eGFR at randomization. There was some evidence that the proportional risk reduction may be larger in those with higher urinary ACR (Figs. 2 and S4). Results were similar in prespecified exploratory subgroup analyses of the kidney disease progression outcome (Fig. S5). There was an acute drop in eGFR on commencing study treatment, followed by a slowing of the rate of annual decline. Overall, the between-group difference in total slope was 0.75 (95% CI 0.54 to 0.96) ml per minute per 1.73 m2 per year. For chronic slopes (i.e. from 2 months to final follow-up), there was a between-group difference of 1.37 (95% CI 1.16 to 1.59) ml per minute per 1.73 m2 per year (Figs. 3 and S6). Prespecified exploratory analyses by subgroups showed the rate of decline in the chronic slope was slower in the empgalfilozin group in all the key subgroups, including patients with low urinary ACR. Between group differences in rate of eGFR decline were larger in the subgroups of participant with faster rate of annual decline (i.e. patients with diabetes, higher eGFR, or higher baseline urinary ACR) (Fig. S7). There were no significant effects of empagliflozin on any specific cause of death (Figure S3), major cardiovascular events (hazard ratio, 0.93 95% CI 0.76-1.12), self-reported episodes of gout, or development of new-onset diabetes (Tables S5 and S6).

Ketoacidosis occurred in 6 patients in the empagliflozin group versus 1 patient in the placebo group (0.09 versus 0.02 per 100 patient-years). Lower limb amputations occurred in 28 patients in the empagliflozin group and 19 in the placebo group (0.43 versus 0.29 per 100 patient-years). The incidence of serious urinary tract infections, hyperkalemia, acute kidney injuries, serious or symptomatic dehydrations, liver injuries, and bone fractures were broadly similar in each group (Tables 2 and S7). There was no apparent evidence that empagliflozin increased the incidence of serious adverse events overall, or in any particular MedDRA system organ class (Table S8).

There were reductions in weighted-average differences [standard error, SE] in mean body weight (-0.9 [0.1] kg) and blood pressure (systolic -2.6 [0.3] mmHg; diastolic -0.5 [0.2] mmHg), but no significant effect on glycated hemoglobin (Table S9). The geometric mean urinary ACR was reduced by 19% (95% CI 15% to 23%). Table S10 provides details of the observed increases in hematocrit and hemoglobin, and the absence of clinically relevant differences in blood calcium, phosphate, or sodium measured in a subset of participants at 18 months.

In this population of patients with a wide range of causes of CKD, GFR and levels of albuminuria, empagliflozin safely reduced the risk of the primary outcome of kidney disease progression or death from cardiovascular causes by about 28%. Treatment was effective irrespective of whether patients had diabetes, and across a broad range of eGFR down to around 20 ml per minute per 1.73 m2. Risk of hospitalization for any cause was also reduced by 14%.

risk of the primary outcome of kidney disease progression or death from cardiovascular causes by about 28%. Treatment was effective irrespective of whether patients had diabetes, and across a broad range of eGFR down to around 20 ml per minute per 1.73 m2. Risk of hospitalization for any cause was also reduced by 14%. The effect of SGLT2 inhibition on kidney disease progression or cardiovascular death seen in the present trial is quantitatively similar to that seen in two other large placebo-controlled trials in CKD populations.6,7 The CREDENCE trial of canagliflozin required all participants to have type 2 diabetes and a urinary ACR of at least 300 mg/g, and excluded patients with an eGFR of less than 30 ml per minute per 1.73 m2.6 The DAPA-CKD trial of dapagliflozin required participants to have a urinary ACR of 200 mg/g and an eGFR of 25 to 75 ml per minute per 1.73 m2. It included 1398 participants without diabetes and 624 participants with an eGFR below 30 ml per minute per 1.73 m2.7 EMPA-KIDNEY adds substantially to the existing evidence by demonstrating consistent benefits among 3569 (54%) participants without diabetes and, separately, among 2282 (35%) participants with an eGFR below 30 ml per minute per 1.73 m2. Despite recruiting 3192 (48%) participants with a urinary ACR below 300 mg/g, there was a limited number of primary outcomes in these types of patient as their CKD was progressing at a slower rate than participants with a urinary ACR of at least 300 mg/g. Prespecified exploratory analyses of the annual rate of change in eGFR - an accepted surrogate for kidney disease progression24 – showed empagliflozin slowed the rate of chronic eGFR decline in patients with a urinary ACR below 300 mg/g at baseline (including those with urinary ACR <30 mg/g).

ACR of at least 300 mg/g. Prespecified exploratory analyses of the annual rate of change in eGFR - an accepted surrogate for kidney disease progression24 – showed empagliflozin slowed the rate of chronic eGFR decline in patients with a urinary ACR below 300 mg/g at baseline (including those with urinary ACR <30 mg/g). Key trial strengths are its large size and broad eligibility criteria, the high level of adherence to study treatment, and the almost complete follow-up of all participants. The trial has certain limitations, including the lower-than-expected cardiovascular event rate, which reduced statistical power to assess the secondary or tertiary cardiovascular outcomes. Nevertheless, the hazard ratios for cardiovascular outcomes are consistent with the totality of the evidence: a meta-analysis of the other CKD trials indicated that SGLT2 inhibitors lower risk of cardiovascular death by 16% (0.84; 95% CI 0.73 to 0.97) and the composite of hospitalization for heart failure or cardiovascular death by 27% (0.73; 95% CI 0.65 to 0.82).13 In summary, in a broad range of patients with CKD, including large numbers without diabetes, with an eGFR below 30 ml per minute per 1.73 m2, and with low urinary ACR, we found that empagliflozin reduced the risk of kidney disease progression or death from cardiovascular causes in a broad range of patients with CKD at risk of progression.