Empagliflozin and incidence of events consistent with acute kidney injury: Pooled safety analysis in more than 15 000 individuals.

Abstract

Chronic kidney disease (CKD) is a common co-morbidity that will develop in about 40% of patients with type 2 diabetes (T2D).1 Sodium-glucose co-transporter-2 inhibitors (SGLT2is) are key treatment options for patients with T2D to reduce the risk of cardiovascular (CV) disease, heart failure (HF), and kidney disease. More recently, beneficial effects with SGLT2is in patients with HF and CKD have also been shown, independent of the prevalence of T2D.2-5 SGLT2is have been shown to acutely decrease estimated glomerular filtration rate (eGFR), often referred to as an ‘initial eGFR dip’.6-8 Although largely considered a haemodynamic and reversible effect, there has been concern that SGLT2is potentially predispose patients to acute kidney injury (AKI) as a result of a reduction in intraglomerular pressure, induction of kidney medullary hypoxic injury, or hypovolaemia.9 This may be of particular concern in patients who have impaired kidney function, are elderly, or receiving loop diuretic medication. In the EMPA-REG OUTCOME trial, treatment with empagliflozin was not associated with an increased risk of AKI, even in the patient subgroup with eGFR levels of less than 60 ml/min/1.73m2 at baseline.8 While patients were more probable to experience an ‘eGFR dip’ of more than 10% with empagliflozin, CV and kidney outcomes were consistent.10 We analysed pooled safety datasets in the global empagliflozin clinical trial programme to investigate the incidence of investigator-reported acute kidney events, focusing on patient groups at risk of acute kidney events overall and during the first 30 days after treatment initiation. We analysed pooled safety data from 20 trials (Table S1) all in patients with T2D who were randomized to receive empagliflozin or matching placebo. Where studies included other empagliflozin doses, only patients who received empagliflozin 10/25 mg were included; dose escalation from 10 to 25 mg empagliflozin was used in some cases (Table S1). Safety topics of interest were analysed using Medical Dictionary for Regulatory Activities (MedDRA) version 21.0 preferred terms. AKI was assessed using investigator-reported adverse events (AEs) and captured by the MedDRA preferred term ‘acute kidney injury’. Acute kidney disease (AKD) is a broader umbrella term that includes AKI events, but also includes events that were haemodynamically driven (Table S2). AKD was assessed using investigator-reported AEs (without adjudication) coded as ‘acute renal failure’ according to the narrow standardized MedDRA query (see Table S2). Descriptive analyses were performed based on all patients treated with at least one dose of trial drug (placebo or empagliflozin). Exposure-adjusted incidence rates were calculated per 100 patient-years as 100 × n/T, where n was the number of patients with the event and T was the total patient-years at risk of the event. Patient-years at risk were defined as the time from the first dose to the onset of the first event (for patients with an event) or to the last dose + 7 days (for patients without an event). Time to first occurrence of AKI or AKD was evaluated by Kaplan–Meier analysis. A total of 15 081 patients were included, with a combined total empagliflozin exposure in patient-years of 16 480 and 7857 for the empagliflozin 10/25 mg and placebo groups, respectively. The studied population was 64% male with a mean age of 60.5 years in the placebo group and 60.3 years in the empagliflozin 10/25 mg group (Table S3). Mean (standard deviation) eGFR at baseline in the placebo group was 80.9 (21.0) ml/min/1.73m2 compared with 82.3 (20.2) ml/min/1.73m2 in the empagliflozin group. At baseline, 25.6% of the placebo group and 24.8% of the empagliflozin group had microalbuminuria, while 8.9% of the placebo group and 7.7% of the empagliflozin group had macroalbuminuria. The incidence of AKI and AKD events were numerically lower in the the empagliflozin versus placebo group (Figure 1 and Table 1). Information on acute dialysis was only available for the EMPA-REG OUTCOME trial, and the majority of AKI events in this pooled group were from EMPA-REG OUTCOME (placebo, n = 37; empagliflozin 10 mg, n = 26; empagliflozin 25 mg, n = 19).8 Similarly, the majority of AKD events in the pooled group were from EMPA-REG OUTCOME (placebo, n = 155; empagliflozin 10 mg, n = 121; empagliflozin 25 mg, n = 125).8 Although no formal statistical analyses were used to compare the groups, the overall AKI risk with empagliflozin and placebo appeared similar. AKI events increased with decreasing kidney function, but the overall incidence rate with empagliflozin was numerically lower than with placebo across all subgroups of baseline eGFR (Table S4). Further, the incidence of AKI events remained numerically lower with empagliflozin across all baseline categories by urine albumin: creatinine ratio, history of HF and baseline use of angiotensin-converting enzyme (ACE) inhibitors/angiotensin receptor blockers (ARBs), non-steroidal anti-inflammatory drugs (NSAIDs), diuretics or loop diuretics for empagliflozin versus placebo. The incidence of AKD events was also seen to increase with decreasing kidney function (Table S5) but remained similar between treatments across various baseline subgroups (Figure S1). There were no differences between the treatment groups in the number of AKI outcomes that occurred within the first 30 days of treatment initiation (five events [0.1%], 1.25/100 patient-years [placebo] vs. six events [0.1%], 0.72/100 patient-years [empagliflozin 10/25 mg]), or in the number of AKD events occurring over this period (18 events [0.4%], 4.50/100 patient-years [placebo] vs. 49 events [0.5%], 5.91/100 patient-years [empagliflozin]; Table S5). In this post hoc, pooled analysis of more than 15 000 patients from the global empagliflozin clinical trial programme, risk for AKI and AKD with empagliflozin was comparable with placebo, including in patients with impaired kidney function or chronic HF. In patients receiving medications associated with potential AKI risk (e.g. ACE inhibitors/ARBs, NSAIDs, diuretics and loop diuretics), risk was generally higher compared with those not receiving any of these drugs but was comparable for empagliflozin versus placebo. The overall number of kidney-related events consistent with AKD or AKI within the first 30 days after treatment initiation was low and generally comparable between treatment groups. While there were more reports of kidney impairment in the first 30 days of treatment, no initiation of kidney replacement therapy was reported during this period, suggesting that this may reflect the ‘eGFR dip’ known with empagliflozin and other SGLT2is and is probably of low clinical relevance. Clinical interpretation of these findings may be limited by the clinical trial setting, and while these findings are consistent with real-world evidence to date,11 confirmation by further analysis of patients in real-world clinical settings may be required. Monitoring of kidney function in line with local prescribing information and with consideration of factors that may predispose patients to AKI may be advisable when initiating SGLT2is. Only a low number of AKI and AKD events were reported (101 and 67, respectively) and these were based on investigator reports without any formal adjudication. In clinical settings such as the emergency room, AKI and AKD diagnosis via Kidney Disease Improving Global Outcomes (KDIGO) guidelines is often difficult to obtain,12 and different definitions of AKD exist; for example, this analysis did not use the KDIGO definition. As AKI tends to be under-reported, the results presented here probably represent a conservative estimate, but this should be comparable between treatment arms. No adjudication process implemented for renal events in this analysis, although as all trials were placebo-controlled, no bias is expected to be introduced by the absence of an adjudication process. Although the analysis population included a preponderance of male patients, this phenomenon is common among clinical trials and not specific to the empagliflozin trial programme and was balanced between treatment arms. This comprehensive analysis indicates that empagliflozin was not associated with an increased risk of AKI or acute kidney failure compared with placebo treatment. These results are supported by observational studies in propensity score-matched cohorts and meta-analyses, which reported that SGLT2i use reduced the risk of AKI events.13, 14 RA, SJH, and CW reviewed and edited the initial research proposal, reviewed the data and analysis, and reviewed/edited the manuscript. SJH, DCW, IR, and KD reviewed/edited the initial research proposal and the manuscript. AE conducted the statistical analysis and reviewed/edited the manuscript. DS reviewed the data and analysis, and reviewed/edited the manuscript. The authors thank Audrey Koitka-Weber and Maximilian von Eynatten, who were employees of Boehringer Ingelheim at the time the research was carried out, for their contributions to the analysis. Medical writing assistance, supported financially by Boehringer Ingelheim, was provided by Andy Shepherd and Katarina Kolaric of Elevate Scientific Solutions. This analysis was funded by the Boehringer Ingelheim & Eli Lilly and Company Diabetes Alliance. RA is supported by research grants from NIH 5R01HL126903 and VA 5I01 CX001753 and has served as a consultant for Boehringer Ingelheim, Bayer, Akebia, AstraZeneca, Reata, Relypsa, Chinook, and Janssen. DCW has acted as a consultant and/or received honoraria from Astellas, AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, Janssen, Napp, Mitsubishi Tanabe, Mundipharma, and Vifor Fresenius. KD has no conflicts to declare. CW has received personal fees from Boehringer Ingelheim; and has received personal fees from Akebia, AstraZeneca, Bayer, Eli Lilly, GlaxoSmithKline, Gilead, Merck Sharpe Dohme, Mundipharma, Sanofi-Genzyme, and Vifor Fresenius outside the submitted work. SJH, AE, IR, and DS are employees of Boehringer Ingelheim. Appendix S1. Supporting Information Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.