Characterization and implications of the initial estimated glomerular filtration rate ‘dip’ upon sodium-glucose cotransporter-2 inhibition with empagliflozin in the EMPA-REG OUTCOME trial

Treatment with sodium-glucose co-transporter-2 inhibitors induces an initial 3–5 ml/min/1.73 m2 decline in estimated glomerular filtration rate (eGFR). Although considered to be of hemodynamic origin and largely reversible, this ‘eGFR dip’ may cause concern clinical practice, which highlights the need better understand its incidence implications. In post hoc analysis EMPA-REG OUTCOME trial, 6,668 participants randomized empagliflozin 10 mg, 25 mg or placebo eGFR available at baseline week four were categorized by change into three groups; over 10% (‘eGFR dipper’), 0 up intermediate’), no non-dipper’). Baseline characteristics intermediate’ non-dipper’ generally comparable. An was observed 28.3% versus 13.4% placebo-treated participants; odds ratio 2.7 [95% Confidence Interval 2.3–3.0]. multivariate logistic regression, diuretic use higher KDIGO risk category independently predictive placebo. Safety beneficial treatment effects on cardiovascular kidney outcomes consistent across subgroups based these factors. The did not have a major impact effect subsequent death, hospitalization for heart failure, incident worsening disease. Thus, patients type 2 diabetes more advanced disease and/or therapy likely experience empagliflozin, but reduction relevantly modified such dip.’ Sodium-glucose (SGLT2i) developed as glucose-lowering agents represent new option (CV) (T2D). Improvements CV been several SGLT2i trials.1Zinman B. Wanner C. Lachin J.M. et al.Empagliflozin, outcomes, mortality diabetes.N Engl J Med. 2015; 373: 2117-2128Crossref PubMed Scopus (6811) Google Scholar, 2Wanner Inzucchi S.E. al.Empagliflozin progression 2016; 375: 323-334Crossref (1935) 3Neal Perkovic V. Mahaffey K.W. al.Canagliflozin renal events 2017; 377: 644-657Crossref (2050) 4Wiviott S.D. Raz I. Bonaca M.P. al.Dapagliflozin 2019; 380: 347-357Crossref (2762) 5Mosenzon O. Wiviott Cahn A. al.Effects dapagliflozin development diabetes: from DECLARE-TIMI 58 randomised trial.Lancet Diabetes Endocrinol. 7: 606-617Abstract Full Text PDF (350) 6Perkovic Jardine M.J. Neal nephropathy.N 2295-2306Crossref (2374) Scholar Owing mechanism action, is associated transient decrease (eGFR), also termed dip,’ shortly after initiation.2Wanner Scholar,7Heerspink H.J. Johnsson E. Gause-Nilsson reduces albuminuria hypertension receiving renin-angiotensin blockers.Diabetes Obes Metab. 18: 590-597Crossref (135) Scholar,8Heerspink Desai M. slows function glycemic effects.J Am Soc Nephrol. 28: 368-375Crossref (238) has raised concerns it predispose acute injury (AKI). Adverse event (AE) post-marketing reporting AKI led U.S. Food Drug Administration issue warnings used caution AKI.9U.S. AdministrationFDA drug safety communication: FDA strengthens medicines canagliflozin (Invokana, Invokamet) (Farxiga, Xigduo XR). [06-14-2016].https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-strengthens-kidney-warnings-diabetes-medicines-canagliflozinGoogle However, data trials1Zinman Scholar,3Neal Scholar,4Wiviott Scholar,6Perkovic large observational cohorts10Nadkarni G.N. Ferrandino R. Chang al.Acute SGLT2 inhibitors: propensity-matched analysis.Diabetes Care. 40: 1479-1485Crossref (108) 11Ueda P. Svanström H. Melbye al.Sodium glucose cotransporter serious adverse events: nationwide register cohort study.BMJ. 2018; 363: k4365Crossref (201) 12Patorno Pawar Franklin failure routine care.Circulation. 139: 2822-2830Crossref (137) 13Iskander Cherney D.Z. Clemens K.K. al.Use cotransporter-2 older adults population-based study.CMAJ. 2020; 192: E351-E360Crossref (33) show reduced SGLT2i.An reported renin-angiotensin-aldosterone system (RAAS) inhibition.14Bakris G.L. Weir M.R. Angiotensin-converting enzyme inhibitor-associated elevations serum creatinine: concern?.Arch Intern 2000; 160: 685-693Crossref Scholar,15Brenner B.M. Cooper M.E. de Zeeuw D. losartan 2001; 345: 861-869Crossref (6092) Until recently, RAAS inhibition sole nephroprotective disease, remains widely used. Nevertheless, value controversial.16Holtkamp F.A. Thomas M.C. al.An fall during predicts slower long-term function.Kidney Int. 2011; 80: 282-287Abstract (246) 17Clase C.M. Barzilay J. Gao does predict outcomes.Kidney 91: 683-690Abstract (51) 18Ohkuma T. Jun Rodgers increases creatinine starting angiotensin-converting inhibitor-based continuation mellitus.Hypertension. 73: 84-91Crossref 19Fu E.L. Trevisan Clase al.Association plasma blockade outcomes.Clin 14: 1336-1345Crossref (14) ScholarThe top limit use, especially within lower range. Therefore, implications understood. We characterized various degrees investigated whether influenced had outcomes.MethodsThe design methods double-blind, placebo-controlled, multinational trial (ClinicalTrials.gov identifier: NCT01131676) described previously.1Zinman study population included 7020 treated T2D, established ≥30 ml/min per 1.73 (MDRD; modification diet disease). Participants assigned random receive (1:1:1) once daily, addition standard care. median duration 2.6 years, observation time 3.1 years. primary outcome prespecified secondary (defined nephropathy) Scholar,2Wanner ScholarFor analysis, 6668 who received ≥1 dose 4 values percent (equation Chronic Kidney Disease Epidemiology Collaboration [CKD-EPI]) 3 categories: >10% >0% ≤10% This pragmatic categorization, using clinically relevant, memorable, easy-to-apply cutoffs, resulted similar-sized groups arm. For each category, we discontinuation. Additional sensitivity performed >30% baseline. all analyses, compared pooled (10 mg) groups. Serum albumin, urinary albumin spot urine, measured central laboratories calculate urine-albumin-to-creatinine (UACR). Disease: Improving Global Outcomes (KDIGO) categorization conducted according heat map, 2-dimensional classification system, identifying low UACR levels, are elevated outcomes.20Kidney (KDIGO)KDIGO 2012 practice guideline evaluation management chronic disease.Kidney 2013; : 1-150Google time, CKD-EPI equation. Mixed-model, repeated-measures evaluate changes including glycated hemoglobin (HbA1c) level (CKD-EPI) linear covariates geographic region, body mass index (BMI), treatment, visit, visit-by-treatment interaction, interaction between HbA1c visit fixed effects. Changes year (i.e., slope) obtained intercept/random coefficient model, previously.21Wanner Heerspink Zinman slope trial.J 29: 2755-2769Crossref (103) model applied dipping’ separately, only discontinuation where assessed absolute last (LVOT) first (follow-up).Baseline further analyses categories focus one harmonized event, defined occurrence 4. evaluated potential regression factors, factors investigate interactions hence Following that approach, backward selection significance P < 0.05 factor retained model. 0.1 determined step Relevant calculated model.To association CV-related (HHF), independent combined Cox proportional hazards group, variables age, sex, BMI, HbA1c, eGFR, additional adjustment systolic blood pressure (SBP), diastolic (DBP), fasting glucose.While actual randomization results loss thus allow comparison dipper’ other categories, approaches. analyzed relevant event. These subgroup, treatment-by-subgroup interaction. addition, mediator outcomes. done accordance previously concept traditional mediation proposed Baron Kenny,22Baron R.M. Kenny D.A. moderator-mediator variable distinction social psychological research: conceptual, strategic, statistical considerations.J Pers Psychol. 1986; 51: 1173-1182Crossref (53345) similarly death.23Inzucchi Fitchett al.How reduce mortality? Insights trial.Diabetes 41: 356-363Crossref (427) onward model,1Zinman adjusted follows: % = 100 ∗ ([lnHR - lnHRC]/lnHR), HR (hazard ratio) denotes group alone HRC adjusting dip’. To mediator, should studied outcome, must Mediation indicated if closer unity than (primary model). Complete would 1.0 A positive indicates partially mediated negative diminished dip.’All nominal α without correction multiple hypothesis testing.Safety overall AEs descriptive narrow Standardised Medical Dictionary Regulatory Activities (MedDRA) Query (ARF) investigators, preferred term AKI. Statistical SAS version 9.4 (SAS Institute, Cary, NC).ResultsThere wide interindividual variability among participants. Whereas (IQR) –0.05 (IQR, –4.04 +4.27) (–21.4 +21.4 1st 99th percentiles, respectively), shifted toward –2.69 –7.87, +1.30) (–24.9 +17.7 respectively).Baseline categoriesCategorization dipper,’ 41.1% 39.5% intermediate,’ 30.5% 47.1% groups, respectively (Figure 1a). shown Table 1. Among empagliflozin-treated participants, most comparable except SBP slightly latter group. contrast, there some differences dippers’ non-dippers’; older, longer-standing history diabetes, rates impaired category. Hemoglobin, hematocrit, levels dippers.’ suboptimal control, even though they taking antihypertensive medications. insulin fewer metformin, while nonusers, metformin users nonusers.24Inzucchi Jurišić-Eržen al.Are benefits therapy?.Diabetes 22: 631-639Crossref (49) Characteristics participant subset experienced initiation (1.4%; n 64) summarized Supplementary S1. Overall, their those dippers,’ tended comorbidities factors.Table 1Baseline categoriesCharacteristics‘eGFR (>10% decline)‘eGFR (>0% (no decline)Number (%)1259 (28.3)1827 (41.1)1357 (30.5)Age, yr, mean ± SD64.6 8.3cP 0.0001 non-dipping group.62.8 8.462.2 8.8Male (%)875 (69.5)1316 (72.0)980 (72.2)BMI, kg/m2, SD30.9 5.4bP 0.001 group.30.6 5.2aP group.30.2 5.2HbA1c, %, SDdN value: 1258 group.8.11 0.88.04 0.88.05 0.9FPG, mg/dl, SDeN values: 1352, 1818, 1257 no, >0 ≤10%, respectively.150.4 44.1aP group.152.7 42.8153.8 43.4Time since diagnosis (%)cP group.aP ≤1 yr22 (1.7)51 (2.8)52 (3.8) >1 5 yr161 (12.8)286 (15.7)239 (17.6) >5 yr293 (23.3)457 (25.0)362 (26.7) >10 yr783 (62.2)1033 (56.5)704 (51.9)eGFR, m2, SD68.3 18.1cP group.79.5 22.9cP group.72.9 20.6eGFR group.cP ≥90 m2132 (10.5)599 (32.8)272 (20.0) 60 <90 m2716 (56.9)859 (47.0)731 (53.9) <60 m2411 (32.6)369 (20.2)354 (26.1)UACR (median, IQR)fN 1337, 1813, 1245 respectively.27.4 (7.1–121.1)cP group.16.8 (7.1–59.2)15.0 (6.2–55.7)UACR <30643 (51.1)1137 (62.2)870 (64.1) 300409 (32.5)513 (28.1)339 (25.0) >300193 (15.3)163 (8.9)128 (9.4)KDIGO Low CKD473 (37.6)961 (52.6)690 (50.8) Moderate CKD391 (31.1)510 (27.9)371 (27.3) High CKD246 (19.5)251 (13.7)165 (12.2) Very high CKD135 (10.7)91 (5.0)111 (8.2)Hemoglobin, g/dl, SDgN 1356, 1826, 1259 respectively.13.5 1.5cP group.13.8 1.413.9 1.5Hematocrit, SDhN 1353, 1821, respectively.40.6 4.5cP group.41.6 4.2aP group.41.9 4.4Albumin, SDiN group.4.39 0.32cP group.4.43 0.29aP group.4.46 0.31Cholesterol, SD HDLjN 1339, 1807, 1243 respectively.44.3 12.144.6 11.444.9 12.2 LDLkN 1805, 1242 respectively.85.0 33.7aP group.85.0 35.2aP group.88.2 38.5 TGjN respectively.176.8 147167.8 129167.4 116SBP, mmHg, SD137.3 17.4cP group.135.3 16.4bP group.133.3 16.8DBP, SD76.6 9.976.8 9.476.6 9.8BP categorical, <140 DBP mmHg728 (57.8)1130 (61.9)898 (66.2) ≥140 mmHg531 (42.2)697 (38.1)459 (33.8)Concomitant medication, (%) ACEi/ARB1094 (86.9)cP group.1462 (80.0)1054 (77.7) Beta-blocker853 (67.8)aP group.1186 (64.9)856 (63.1) Diuretic685 (54.4)cP group.695 (38.0)539 (39.7) Loop diuretic252 (20.0)cP group.234 (12.8)178 (13.1) CCB441 (35.0)aP group.581 (31.8)426 (31.4) Statin973 (77.3)1431 (78.3)1033 (76.1) ASA1068 (84.8)aP group.1500 (82.1)1112 (81.9) Metformin864 (68.6)cP group.1389 (76.0)1031 (76.0) Sulfonylurea513 (40.7)aP group.787 (43.1)615 (45.3) Insulin666 (52.9)cP group.858 (47.0)592 (43.6)CV factor, CHD history965 (76.6)1379 (75.5)1008 (74.3) Stroke history294 (23.4)427 (23.4)307 (22.6) PAD history301 (23.9)aP group.366 (20.0)270 (19.9) HF history138 (11.0)149 (8.2)135 (9.9)Race, (%)aP White920 (73.1)1328 (72.7)970 (71.5) Asian242 (19.2)412 (22.6)308 (22.7) Black/African-American90 (7.1)71 (3.9)64 (4.7)ACEi, inhibitor; ARB, angiotensin receptor blocker; ASA, acetylsalicylic acid; index; CCB, calcium channel CHD, coronary disease; CKD, CV, cardiovascular; DBP, pressure; rate; HDL, high-density lipoprotein; HF, failure; KDIGO, Outcomes; LDL, low-density PAD, peripheral artery SBP, diabetes; TG, triglycerides; UACR, urine albumin-to-creatinine ratio.a group.b group.c group.d N group.e respectively.f respectively.g respectively.h respectively.i group.j respectively.k respectively. Open table tab categoryThe 68.3 18.1, 79.5 22.9, 72.9 20.6 respective –12.6 5.7, –3.3 2.4, +5.4 5.7 m2. Few (1.4% [n 64] 0.9% 20] placebo). these, 1 patient discontinued following 4.In remained stable 12 1b), well empagliflozin-induced (Supplementary Figure S1). placeb