Despite the well-established benefits of mineralocorticoid receptor agonists (MRAs) in heart failure with reduced ejection fraction, safety concerns remain in patients with concomitant diabetes mellitus (DM) because of common renal and electrolyte abnormalities in this population. We analyzed all-cause mortality and composite cardiovascular mortality and HF hospitalization over a median 9.9 months among 1,998 patients in the placebo arm of the Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study With Tolvaptan (EVEREST) trial by DM status and discharge MRA use. Of the 750 patients with DM, 59.2% were receiving MRAs compared with 62.5% in the non-DM patients. DM patients not receiving MRAs were older, more likely to be men, with an ischemic heart failure etiology and slightly worse renal function compared with those receiving MRAs. After adjustment for baseline risk factors, among DM patients, MRA use was not associated with either mortality (hazard ratio [HR] 0.93; 95% confidence interval [CI] 0.75 to 1.15) or the composite end point (HR 0.94; 95% CI 0.80 to 1.10). Similar findings were seen in non-DM patients (mortality [HR 1.01; 95% CI 0.84 to 1.22] or the composite end point [HR 0.98; 95% CI 0.85 to 1.13] [p >0.43 for DM interaction]). In conclusion, in-hospital initiation of MRA therapy was low (15% to 20%), and overall discharge MRA use was only 60% (with regional variation), regardless of DM status. There does not appear to be clear, clinically significant in-hospital hemodynamic or even renal differences between those on and off MRA. Discharge MRA use was not associated with postdischarge end points in patients hospitalized for worsening heart failure with reduced ejection fraction and co-morbid DM. DM does not appear to influence the effectiveness of MRA therapy.
Approximately 40% to 45% of patients hospitalized for worsening heart failure with reduced ejection fraction (HFrEF) have coexistent diabetes mellitus (DM). DM is an independent predictor of adverse postdischarge outcomes in hospitalized HFrEF patients and may modulate the risk-benefit ratio of certain pharmacotherapies. Mineralocorticoid receptor antagonist (MRA) have been shown to improve clinical outcomes in chronic HFrEF patients with mild-to-severe symptoms and patients with left ventricular dysfunction after myocardial infarction (MI). Accruing evidence suggests that the benefits of mineralocorticoid receptor (MR) blockade may be safely extended to the subset of HFrEF patients with DM. The widespread use of MRAs has been limited by ongoing clinician concern regarding worsening renal function and hyperkalemia, especially with concomitant use of angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers. In addition, type 2 DM was among the major risk factors for life-threatening hyperkalemia in a small case series of HFrEF patients. The immediate postdischarge period after hospitalization for HF is a vulnerable period marked by acute perturbations in electrolyte, neurohormonal, and renal function profiles, perhaps further augmenting MRA-associated side effects. Data are limited regarding the overall utilization and safety profile of MRA use in patients hospitalized for HFrEF with co-morbid DM. The Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study With Tolvaptan (EVEREST) trial included patients who largely met criteria for prescription of MRA (e.g., HFrEF, mild-to-severe symptomatology, without major baseline renal or electrolyte abnormalities). This trial experience offers an ideal setting to evaluate an in-depth, longitudinal characterization of the clinical profiles and MRA prescription patterns of patients hospitalized for worsening chronic HFrEF with comorbid DM.
Methods
The study design and primary results of the EVEREST trial have been previously described. In brief, EVEREST was a prospective, international, randomized, double-blind, placebo-controlled trial designed to explore the short- and long-term impact of tolvaptan, a vasopressin-2 receptor antagonist, when added to standard therapy, in patients hospitalized for worsening HF with an EF ≤40% and presenting with an evidence of fluid overload. Participants were randomized within 48 hours of hospitalization to receive either oral tolvaptan or matching placebo, in addition to standard therapy. Background HF therapy was left to the discretion of the treating physician, but guideline-based recommendations for optimal medical therapy were included in the study protocol. Significant exclusion criteria included refractory end-stage HF, hemofiltration or dialysis, supine systolic blood pressure (SBP) <90 mm Hg, serum creatinine concentration >3.5 mg/dl, and serum potassium >5.5 mEq/L.
Because tolvaptan interacts with the renin-angiotensin aldosterone system, we performed a post hoc analysis examining only patients in the placebo arm with available discharge MRA data. All patients who died during hospitalization were, thus, excluded. HFrEF patients were divided by MRA use at the time of discharge in the EVEREST trial and by the presence of DM. MRAs in the EVEREST database included canrenoic acid, canrenone, potassium canreonate, eplerenone, soludactone, and spironolactone. DM status was ascertained by baseline questionnaires obtained by study site coordinators from patient interviews and medical records in accordance to the American Diabetes Association criteria. Patients receiving insulin or oral hypoglycemic agents were also categorized as having DM. Chronic kidney disease was defined as estimated glomerular filtration rate ≤60 ml/min/1.73 m 2 on the day of enrollment, calculated using the Modification of Diet in Renal Disease Study equation.
The study was approved by the Institutional Review Boards and Ethics Committees of each participating site and was conducted in accordance with the Declaration of Helsinki. Clinical characteristics documented at baseline, with the exception of concomitant therapies that were obtained from discharge records, were used for the present analysis. The first outpatient visit occurred 7 days after discharge for those subjects discharged from the hospital on or before the tenth day or the seventeenth day after randomization for those still in the hospital on day 10. Outpatient assessments were performed after 1, 4, and 8 weeks and every 8 weeks thereafter up to 128 weeks.
An independent, blinded adjudication committee determined the specific causes of death and reasons for rehospitalization. This post hoc analysis used the 2 EVEREST co-primary end points: (1) all-cause mortality (ACM) and (2) the composite of cardiovascular (CV) mortality and HF hospitalization. Median follow-up in the EVEREST trial was 9.9 months (interquartile range 5.3 to 16.1 months).
For descriptive purposes, patients were stratified by discharge MRA use as MRA + and MRA − . Similarly, DM status was defined as DM + and DM − . Differences between MRA + versus MRA − were summarized separately for DM + and DM − patients. Baseline characteristics were compared by discharge MRA use in patients with and without DM using chi-square testing, Fisher’s exact test, and Kruskal-Wallis tests where appropriate. All continuous variables were reported as mean ± SD if normally distributed or median (interquartile range) if non-normally distributed.
The primary predictor for this analysis was MRA use at the time of discharge. Time-to-event data were analyzed with log-rank test, and hazard ratios (HRs) with corresponding 95% confidence intervals (CIs) were obtained from Cox proportional hazard models. The proportional hazards assumption (by Kolmogorov-type supremum tests for nonproportionality) was upheld for all end points, except for the composite end point in the non-DM cohort. For this group, the follow-up period was divided into 2 phases at 50 days after randomization (cutoff established by visual inspection of standardized score process plots). All multivariable Cox regression models were adjusted for known baseline predictors of mortality and morbidity: age, sex, region, EF, SBP, sodium, blood urea nitrogen, N-terminal pro-brain natriuretic peptide, QRS duration, discharge medication use (ACE inhibitors, β blockers, digoxin), in-hospital inotrope requirement, New York Heart Association (NYHA) class IV, atrial fibrillation/flutter, history of hypertension, coronary artery disease (CAD), chronic obstructive pulmonary disease, ischemic HF etiology, previous HF hospitalization, and chronic kidney disease. No evidence of significant collinearity between MRA utilization and the covariate set was detected. Testing for interaction between MRA use and outcomes by underlying DM status was performed.
Results
Of the 2,061 patients assigned to the placebo arm in the EVEREST trial, 3.3% (n = 63) of patients died during hospitalization or had missing discharge MRA data. Of those discharged alive with known MRA status, 62.3% (n = 1,245) received an MRA at discharge. Baseline DM was present in 37.5% (n = 750) of patients, and among these, 59.2% (n = 444) were discharged on MRA therapy, compared with 64.2% (n = 801) in patients without DM ( Figure 1 ). Most of these patients had been prescribed an MRA at the time of enrollment, which was continued through hospitalization (76.7% in the DM group and 80.3% in the non-DM group). Spironolactone was the predominant MRA used in the overwhelming majority of patients, regardless of DM status.
Among patients with DM, those not discharged on MRA therapy were generally older (p <0.001) and were more likely to be male (p <0.03), with a higher EF (p <0.003) and better NYHA functional class (p = 0.024). Discharge MRA use was less frequent in patients recruited from North America and Western Europe compared with South America and Eastern Europe (p <0.001). Lack of MRA prescription was associated with higher rates of co-morbidities such as hypertension (p = 0.017), peripheral vascular disease (p <0.001), and hyperlipidemia (p = 0.004) but lower rates of previous HF hospitalization and atrial fibrillation/flutter (p = 0.009). The prevalence of CAD and revascularization procedures was also significantly higher (both p <0.05) in DM patients who were not prescribed MRAs at discharge. SBPs were slightly, but nonsignificantly, higher in the non-MRA group (122.1 ± 19.9 vs 119.8 ± 19.3; p = 0.1). Serum creatinine was also slightly higher in patients not prescribed MRAs at discharge (1.5 ± 0.6 vs 1.4 ± 0.5; p = 0.003). Patients off MRAs at discharge were less likely to receive diuretics, digoxin, β blockers, and inotropic agents and more likely to receive antiplatelet drugs. In general, similar patterns were observed in baseline clinical profiles between patients discharged with and without MRA in both DM and non-DM cohorts ( Table 1 ).
| Variable | Diabetic | Non-Diabetic | ||||
|---|---|---|---|---|---|---|
| No MRA (n = 306) | MRA (n = 444) | p | No MRA (n = 447) | MRA (n = 801) | p | |
| Age (years) | 67.9 ± 10.7 | 65.1 ± 10.1 | <0.001 | 67.6 ± 12.6 | 63.6 ± 12.8 | <0.001 |
| Men | 242 (79.1%) | 320 (72.1%) | 0.029 | 347 (77.6%) | 605 (75.5%) | 0.403 |
| Non-Hispanic white | 259 (84.6%) | 373 (84%) | 0.3615 | 387 (86.6%) | 691 (86.3%) | 0.6121 |
| Black | 22 (7.2%) | 45 (10.1%) | 32 (7.2%) | 47 (5.9%) | ||
| Hispanic | 18 (5.9%) | 19 (4.3%) | 18 (4%) | 42 (5.2%) | ||
| Other Race/Ethnicity | 7 (2.3%) | 7 (1.6%) | 10 (2.2%) | 21 (2.6%) | ||
| Eastern Europe | 71 (23.2%) | 162 (36.5%) | <0.001 | 166 (37.1%) | 390 (48.7%) | 0.337 |
| North America | 163 (53.3%) | 145 (32.7%) | 158 (35.3%) | 139 (17.4%) | ||
| South America | 23 (7.5%) | 73 (16.4%) | 61 (13.6%) | 176 (22%) | ||
| Western Europe | 49 (16%) | 64 (14.4%) | 62 (13.9%) | 96 (12%) | ||
| Ejection fraction (%) | 28.6 ± 8.8 | 26.7 ± 7.8 | 0.003 | 28.3 ± 8.2 | 27.3 ± 8.1 | 0.039 |
| QRS duration (msec) | 122 (95–145) | 123 (97–150) | 0.677 | 123 (97–148) | 124 (98–152) | 0.514 |
| Body mass index (kg/m 2 ) | 29.3 ± 7.1 | 71.2 ± 856.3 | 0.078 | 48.5 ± 448.4 | 47 ± 623.2 | 0.962 |
| Atrial fibrillation on electrocardiogram | 70 (22.9%) | 117 (26.4%) | 0.009 | 130 (29.1%) | 253 (31.6%) | 0.364 |
| Ischemic heart failure etiology | 234 (77%) | 301 (68.3%) | 0.556 | 290 (65.9%) | 471 (59.7%) | 0.032 |
| Prior heart failure hospitalization | 237 (78.2%) | 375 (84.8%) | 0.001 | 313 (70.5%) | 635 (79.4%) | <0.001 |
| Prior myocardial infarction | 199 (65.2%) | 237 (53.4%) | 0.058 | 226 (50.7%) | 357 (44.6%) | 0.038 |
| Coronary artery disease ∗ | 257 (84.3%) | 326 (73.6%) | 0.001 | 325 (72.9%) | 497 (62%) | <0.001 |
| Hypertension † | 255 (83.3%) | 345 (77.7%) | 0.017 | 313 (70%) | 508 (63.4%) | 0.018 |
| Hypercholesterolemia ‡ | 200 (66%) | 254 (57.3%) | 0.004 | 204 (45.8%) | 293 (36.7%) | 0.002 |
| Peripheral vascular disease | 97 (31.7%) | 98 (22.2%) | <0.001 | 84 (18.8%) | 155 (19.4%) | 0.816 |
| Prior coronary artery bypass graft surgery | 112 (36.6%) | 108 (24.3%) | 0.006 | 102 (22.8%) | 107 (13.4%) | <0.001 |
| Prior percutaneous coronary intervention | 90 (29.4%) | 92 (20.7%) | 0.001 | 91 (20.4%) | 82 (10.2%) | <0.001 |
| Implantable cardioverter-defibrillator | 71 (23.2%) | 62 (14%) | 0.648 | 73 (16.3%) | 76 (9.5%) | <0.001 |
| Pacemaker | 62 (20.3%) | 84 (18.9%) | <0.001 | 86 (19.2%) | 103 (12.9%) | 0.003 |
| Chronic kidney disease | 142 (46.4%) | 144 (32.5%) | 0.797 | 123 (27.5%) | 127 (15.9%) | <0.001 |
| Chronic obstructive pulmonary disease | 36 (11.8%) | 55 (12.4%) | 0.556 | 44 (9.8%) | 65 (8.1%) | 0.3 |
| Dyspnea | 275 (91.4%) | 398 (92.6%) | 0.629 | 387 (88.2%) | 726 (91.9%) | 0.031 |
| Jugular venous distension | 76 (25.7%) | 116 (27.3%) | 0.982 | 113 (25.9%) | 213 (27.1%) | 0.638 |
| Rales | 241 (80.1%) | 344 (80%) | 0.061 | 363 (82.7%) | 656 (82.8%) | 0.95 |
| Edema § | 238 (79.1%) | 364 (84.5%) | 0.122 | 325 (74%) | 641 (80.9%) | 0.005 |
| Systolic blood pressure (mm Hg) | 122.1 ± 19.9 | 119.8 ± 19.3 | 0.115 | 122.3 ± 20.4 | 118.7 ± 18.4 | 0.002 |
| Heart rate (bpm) | 77.5 ± 15 | 79.4 ± 15.5 | 0.042 | 79.7 ± 15.5 | 80.6 ± 16 | 0.308 |
| New York Heart Association class IV | 121 (39.7%) | 185 (41.8%) | 0.024 | 141 (31.6%) | 332 (41.4%) | 0.001 |
| Albumin (g/dL) | 3.6 ± 0.6 | 3.7 ± 0.5 | 0.027 | 3.8 ± 0.5 | 3.8 ± 0.5 | 0.745 |
| NT-BNP (pg/mL) | 4839 (2079–8567) | 3581.5 (1865–7667) | 0.001 | 4820 (2287–10464) | 5028 (2462–9483) | 0.767 |
| Blood urea nitrogen (mg/dL) | 36.5 ± 22 | 31.6 ± 16.7 | 0.005 | 28.5 ± 13.7 | 27.5 ± 13.9 | 0.245 |
| Creatinine (mg/dL) | 1.5 ± 0.6 | 1.4 ± 0.5 | 0.003 | 1.4 ± 0.5 | 1.3 ± 0.4 | <0.001 |
| Estimated glomerular filtration rate (mL/min) ¶ | 50.3 ± 22.3 | 54.7 ± 20 | 0.002 | 56.1 ± 21.5 | 59.2 ± 20.1 | 0.016 |
| Sodium (mEq/L) | 139.2 ± 4 | 139.1 ± 4.8 | 0.462 | 140.5 ± 4.5 | 139.8 ± 4.8 | 0.008 |
| MRA at enrollment | 48 (15.8%) | 340 (76.7%) | <0.001 | 88 (19.9%) | 642 (80.3%) | <0.001 |
| Medications at discharge | ||||||
| Aspirin | 176 (57.5%) | 251 (56.5%) | 0.001 | 243 (54.4%) | 406 (50.7%) | 0.213 |
| ACE inhibitor or ARB | 237 (77.5%) | 385 (86.7%) | 0.07 | 371 (83%) | 703 (87.8%) | 0.02 |
| Beta-blocker | 226 (73.9%) | 353 (79.5%) | 0.009 | 325 (72.7%) | 582 (72.7%) | 0.986 |
| Calcium channel blocker | 43 (14.1%) | 36 (8.1%) | 0.727 | 53 (11.9%) | 39 (4.9%) | <0.001 |
| Warfarin | 114 (37.3%) | 171 (38.5%) | 0.042 | 173 (38.7%) | 317 (39.6%) | 0.762 |
| Digoxin | 132 (43.1%) | 225 (50.7%) | <0.001 | 177 (39.6%) | 421 (52.6%) | <0.001 |
| Diuretics | 272 (88.9%) | 444 (100%) | 0.012 | 381 (85.2%) | 801 (100%) | <0.001 |
| Nitrates | 117 (38.2%) | 131 (29.5%) | 0.374 | 158 (35.3%) | 222 (27.7%) | 0.005 |
| Amiodarone | 55 (18%) | 67 (15.1%) | 0.941 | 101 (22.6%) | 139 (17.4%) | 0.024 |
| Inotropes | 8 (2.6%) | 12 (2.7%) | <0.001 | 7 (1.6%) | 9 (1.1%) | 0.505 |
| Clopidogrel | 54 (17.6%) | 40 (9%) | 0.113 | 37 (8.3%) | 32 (4%) | 0.002 |
| Statins | 144 (47.1%) | 183 (41.2%) | 0.657 | 155 (34.7%) | 190 (23.7%) | <0.001 |
∗ Patient reported history of coronary artery disease.
† Patient reported history of hypertension.
‡ Patient reported history of hypercholesterolemia.
§ Peripheral edema was defined as slight to moderate to marked pedal or sacral edema.
Over a median follow-up period of 9.9 months, 26.5% (n = 199) of patients with DM and 22.8% (n = 285) of patients without DM experienced ACM, whereas 43.9% (n = 329) patients with DM and 35.8% (n = 447) patients without DM experienced the composite end point ( Table 2 ). In the DM subgroup, patients prescribed MRAs experienced lower rates of composite CV mortality and HF hospitalization (p = 0.002) and non-CV death (p = 0.025) but higher rates of sudden cardiac death (5.9% vs 4.6%; p = 0.030). No other differences in cause-specific outcomes were observed between MRA users and nonusers in patients with and without DM. In unadjusted analyses among DM patients, MRA use was associated with a 31% reduction in ACM (HR 0.69; 95% CI 0.52 to 0.91) and a 19% reduction in the composite end point (HR 0.81; 95% CI 0.65 to 1.01) ( Figure 2 ). MRA use was not associated with either ACM (HR 1.16; 95% CI 0.90 to 1.48) or the composite end point (HR 1.01; 95% CI 0.83 to 1.23) among patients without DM ( Figure 3 ). The unadjusted relation between MRA use and ACM differed significantly by DM status (p = 0.006). There was no significant interaction between MRA use and DM status for the composite end point (p = 0.39).
| Diabetes | Non-Diabetes | |||||
|---|---|---|---|---|---|---|
| MRA (−) (n = 306) | MRA (+) (n = 444) | p | MRA (−) (n = 447) | MRA (+) (n = 801) | p | |
| Primary endpoints | ||||||
| All-cause mortality | 98 (32) | 101 (22.7) | 0.172 | 95 (21.3) | 190 (23.7) | 0.319 |
| CV mortality + HF hospitalization | 147 (48) | 182 (41) | 0.002 | 157 (35.1) | 290 (36.2) | 0.702 |
| Cause of death | ||||||
| Cardiovascular | 66 (21.6) | 78 (17.6) | 0.443 | 69 (15.4) | 143 (17.9) | 0.276 |
| Sudden cardiac death | 14 (4.6) | 26 (5.9) | 0.030 | 27 (6) | 62 (7.7) | 0.263 |
| Heart failure | 43 (14.1) | 40 (9) | 0.310 | 30 (6.7) | 67 (8.4) | 0.296 |
| Myocardial infarction | 3 (1) | 1 (0.2) | 0.653 | 6 (1.3) | 4 (0.5) | 0.181 |
| Stroke | 1 (0.3) | 4 (0.9) | 1.000 | 0 (0) | 2 (0.2) | 0.540 |
| Other CV death | 5 (1.6) | 7 (1.6) | 0.154 | 6 (1.3) | 8 (1) | 0.581 |
| Non-CV death | 16 (5.2) | 14 (3.2) | 0.025 | 17 (3.8) | 25 (3.1) | 0.522 |
| Reason for hospitalization | ||||||
| Cardiovascular | 148 (48.4) | 178 (40.1) | 0.212 | 171 (38.3) | 282 (35.2) | 0.283 |
| Heart failure | 107 (35) | 136 (30.6) | 0.060 | 115 (25.7) | 202 (25.2) | 0.843 |
| Myocardial infarction | 10 (3.3) | 5 (1.1) | 0.483 | 9 (2) | 6 (0.7) | 0.049 |
| Stroke | 2 (0.7) | 6 (1.4) | 0.486 | 1 (0.2) | 10 (1.2) | 0.109 |
| Arrhythmia | 11 (3.6) | 12 (2.7) | 0.319 | 15 (3.4) | 19 (2.4) | 0.306 |
| Other CV hospitalization | 18 (5.9) | 19 (4.3) | 0.021 | 31 (6.9) | 45 (5.6) | 0.351 |
| Worsening heart failure | 141 (46.1) | 167 (37.6) | 0.056 | 143 (32) | 246 (30.7) | 0.640 |
| CV mortality + CV hospitalization | 175 (57.2) | 203 (45.7) | 0.002 | 191 (42.7) | 336 (41.9) | 0.789 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
