Digoxin and Clinical Outcomes in Systolic Heart Failure Patients on Contemporary Background Heart Failure Therapy

Digoxin and Clinical Outcomes in Systolic Heart Failure Patients on Contemporary Background Heart Failure Therapy Amandeep Singh Dhaliwal, MDCMa, Audrius Bredikis, MDa, Gabriel Habib, MDa, Blase Anthony Carabello, MDa, Kumudha Ramasubbu, MDa, and Biykem Bozkurt, MDb,* Previous trials have shown that digoxin was beneficial in patients with heart failure (HF). However, these studies were conducted before the incorporation of ␤ blockers as standard therapy for patients with HF. The purpose of this study was to determine the effect of digoxin in patients with HF on a contemporary regimen of renin-angiotensin inhibition and ␤ blockade. In 347 almost exclusively men, data pertaining to the index hospitalization and occurrence of all-cause mortality or readmission for HF were collected. Cox proportional hazard modeling was used. Patients on digoxin therapy had a lower left ventricular (LV) ejection fraction (EF), higher prevalence of previous hospitalizations for HF and atrial fibrillation, and lower prevalence of hypertension. After adjustment for age, LVEF, history of HF hospitalizations, New York Heart Association class, presence of chronic renal insufficiency, presence of atrial fibrillation, and prescriptions for ␤ blockers and angiotensin converting enzyme inhibitors or angiotensin receptor blockers, HF hospitalizations (hazard ratio 1.08, 95% confidence interval [CI] 0.77 to 1.50, p ⴝ 0.66), total mortality (hazard ratio 1.03, 95% CI 0.78 to 1.35, p ⴝ 0.85), or the combined end point of HF hospitalization and total mortality (hazard ratio 1.11, 95% CI 0.81 to 1.53, p ⴝ 0.52) were not different in patients using digoxin compared with those not using digoxin. Clinical outcomes were not different in subgroups of patients with EF <25%, New York Heart Association class III or IV, atrial fibrillation, heart rate <60 beats/min, or patients on ␤-blocker therapy. In conclusion, digoxin use was not associated with a decrease in HF hospitalizations or overall mortality rates in a cohort of hospitalized patients with HF with LV systolic dysfunction on contemporary background HF treatment including angiotensinconverting enzyme inhibitors and ␤ blockers. © 2008 Elsevier Inc. All rights reserved. (Am J Cardiol 2008;102:1356 –1360) The purpose of this study was to determine the effect of digoxin on clinical outcomes in patients with heart failure (HF) on a contemporary regimen of renin-angiotensin axis inhibition and ␤ blockade. Methods Patients discharged from the Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas, with a diagnosisrelated group of congestive HF (diagnosis-related group 127) from April 2002 to June 2004 were identified. Patients were excluded for volume overload for other reasons (e.g., cirrhosis or renal failure); end-stage renal disease or hemodialysis therapy initiation within 12 months of the index hospitalization; other illnesses with life expectancy ⬍12 months; a competing admitting diagnosis that was as or more likely than decompensated HF (e.g., hypertensive

a

Section of Cardiology, Department of Medicine, Michael E. DeBakey Veterans Affairs Medical Center; and bWinters Center for Heart Failure Research, Baylor College of Medicine, Houston, Texas. Manuscript received April 12, 2008; revised manuscript received and accepted July 13, 2008. Dr. Bozkurt was supported by a Merit Entry Level Grant from Veterans Affairs Medical Research Service, Houston, Texas. *Corresponding author: Tel: 713-794-8019; Fax: 713-794-7551. E-mail address: [email protected] (B. Bozkurt). 0002-9149/08/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2008.07.014

emergency, cor pulmonale, pneumonia, or chronic obstructive pulmonary disease exacerbation); absence of follow-up at the Veterans Affairs Medical Center resulting in difficulty ascertaining end points, particularly HF hospitalizations; and absence of a complete data set or not fulfilling modified Framingham criteria for HF. Three hundred eighty-eight such patients were identified, of whom 347 had depressed left ventricular (LV) function, defined as LV ejection fraction (EF) ⱕ45%, similar to the definition used in the Digitalis Investigation Group (DIG) trial.1 Analysis was performed on only these 347 patients with HF with depressed LV function. The Baylor College of Medicine Institutional Review Board (Houston, Texas) and Michael E. DeBakey Veterans Affairs Medical Center Research and Development Committee approved this study. Data pertaining to the index hospitalization were collected and patients were followed up for the occurrence of the combined end point of all-cause mortality or first readmission for HF up to June 30, 2005. Vital status was confirmed using the Social Security Death Index. Glomerular filtration rate was calculated using the abbreviated Modification of Diet in Renal Disease Study equation of glomerular filtration rate (ml/min/1.73 m2) ⫽ 186 ⫻ (serum creatinine)⫺1.154 ⫻ (age)⫺0.203 ⫻ (0.742 if female) ⫻ (1.210 if African American). Chronic renal insufficiency was defined as glomerular filtration rate ⬍60 ml/min/1.73 m2. www.AJConline.org

Heart Failure/Digoxin in Contemporary Heart Failure Treatment

1357

Table 1 Baseline characteristics Variable

Age (yrs) Men White Ischemic cause LVEF (%) NYHA class II/III/IV (%)* Previous HF admission Hypertension Atrial fibrillation Diabetes mellitus Implantable cardioverter-defibrillator Pacemaker Nonsustained ventricular tachycardia Chronic renal insufficiency Body mass index (kg/m2) Medications at discharge Angiotensin-converting enzyme inhibitor or angiotensin receptor blocker ␤ Blocker Loop diuretics Hydralazine Long-acting nitrates Spironolactone Statin Vital signs and laboratory data at discharge Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Heart rate (beats/min) Sodium (mEq/L) Glomerular filtration rate (ml/min/1.73 m2) Brain natriuretic peptide (ng/L)† Hemoglobin (g/dL)

Overall Cohort (n ⫽ 347)

Digoxin

p Value

Yes (n ⫽ 155)

No (n ⫽ 192)

68 ⫾ 11 346 (100%) 192 (55%) 221 (64%) 23 ⫾ 10 19%/65%/16% 137 (39%) 275 (79%) 95 (27%) 164 (47%) 48 (14%) 42 (12%) 52 (15%) 196 (56%) 29 ⫾ 7

69 ⫾ 11 154 (99%) 81 (52%) 102 (66%) 20 ⫾ 9 18%/67%/14% 73 (47%) 110 (71%) 53 (34%) 65 (42%) 25 (16%) 23 (15%) 23 (15%) 89 (57%) 29 ⫾ 6

68 ⫾ 11 192 (100%) 111 (58%) 119 (62%) 25 ⫾ 10 20%/63%/18% 64 (33%) 165 (86%) 42 (22%) 99 (52%) 23 (12%) 19 (10%) 29 (15%) 107 (56%) 30 ⫾ 8

0.34 0.27 0.30 0.46 ⬍0.001 0.51 ⬍0.01 0.001 0.01 0.07 0.27 0.16 0.95 0.75 0.12

277 (80%) 219 (63%) 322 (93%) 16 (5%) 92 (27%) 51 (15%) 162 (47%)

130 (84%) 89 (57%) 149 (96%) 6 (4%) 35 (23%) 26 (17%) 63 (41%)

147 (77%) 130 (68%) 173 (90%) 10 (5%) 57 (30%) 25 (13%) 99 (52%)

0.09 ⬍0.05 ⬍0.05 0.56 0.14 0.33 ⬍0.05

120 ⫾ 24 69 ⫾ 13 78 ⫾ 16 137 ⫾ 4 57 ⫾ 22 614 (317–1,140) 13 ⫾ 2

113 ⫾ 22 67 ⫾ 12 78 ⫾ 15 137 ⫾ 4 60 ⫾ 22 686 (383–1,300) 13 ⫾ 2

125 ⫾ 24 71 ⫾ 13 79 ⫾ 15 137 ⫾ 4 56 ⫾ 22 562 (279–1,080) 12 ⫾ 2

⬍0.001 ⬍0.01 0.91 0.46 0.08 0.15 ⬍0.01

* At baseline. † Because of non-normal distribution, brain natriuretic peptide reported as median (interquartile range).

Data were presented as mean ⫾ SD unless otherwise specified. Baseline demographic characteristics were compared between groups using t test or analysis of variance for continuous variables and chi-square test for categorical variables. Nonparametric methods were used to compare non–normally distributed baseline characteristics. Cox proportional hazard modeling was used to determine the association of digoxin use and clinical outcomes. Included in the model were covariates significantly associated with outcomes in patients with HF, namely age, LVEF, history of HF hospitalizations, New York Heart Association (NYHA) class, presence of chronic renal insufficiency, presence of atrial fibrillation, and prescriptions of ␤ blockers and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. Prespecified analyses were performed subsequently according to the subgroups of LVEF ⱕ25%, patients with NYHA classes III and IV HF, use of ␤ blockers, presence of atrial fibrillation, and admission and discharge heart rates ⱕ60 beats/min. Analyses were performed using STATA 7.0 (Stata Corp., College Station, Texas) software for Windows (Microsoft Corp., Redmond, Washington), and p ⬍0.05 was considered statistically significant.

Results Baseline characteristics were listed in Table 1, reflecting predominantly men with depressed LVEF and a high prevalence of co-morbidities and advanced HF. At discharge, lisinopril was the most commonly used angiotensin-converting enzyme inhibitor (87%). Mean dose was 19 ⫾ 16 mg/day (73% had ⱖ10 mg/day, 48% had ⱖ20 mg/day). Metoprolol succinate was the most commonly used ␤ blocker (70%). Mean dose was 94 ⫾ 56 mg/day (52% had ⱖ100 mg/day, 13% had ⱖ200 mg/day). Carvedilol was used by 23%. Mean dose was 14 ⫾ 14 mg/day (44% had ⱖ12.5 mg/day, 18% had ⱖ25 mg/day). Median follow up was 303 days (interquartile range 69 to 564). A total of 179 admissions for decompensated HF occurred, and there were 184 cumulative deaths. Baseline characteristics according to digoxin use are listed in Table 1. In general, patients on digoxin therapy had lower LVEF, higher prevalence of previous HF hospitalizations, and higher prevalence of atrial fibrillation, but lower prevalence of hypertension. Overall, admission and discharge blood pressures were lower and admission heart rates were higher, but discharge heart rates were not different in patients on digoxin

1358

The American Journal of Cardiology (www.AJConline.org)

Table 2 Hazard ratios for total mortality and heart failure (HF) hospitalization by digoxin use Variable Combined HF hospitalization Mortality

Unadjusted Hazard Ratio (95% Confidence Interval, P value)

Adjusted Hazard Ratio (95% Confidence Interval, P value)

1.16 (0.90–1.49, 0.258) 1.30 (0.96–1.76, 0.096) 1.15 (0.85–1.55, 0.371)

1.03 (0.78–1.35, 0.852) 1.08 (0.77–1.50, 0.663) 1.11 (0.81–1.53, 0.521)

Hazard ratios are for digoxin:no digoxin use and are adjusted for age, LVEF, history of HF hospitalization, NYHA class, presence of chronic renal insufficiency, history of atrial fibrillation, and prescriptions of ␤ blockers and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers.

Figure 1. Adjusted hazard ratios with medians and 95% confidence intervals for the combined end point of HF hospitalizations or overall mortality for the overall cohort and within each subgroup. The referent group for each calculation was the group not on digoxin therapy within each subgroup. Hazard ratios were adjusted for age, LVEF, history of HF hospitalizations, NYHA class, presence of chronic renal insufficiency, presence of atrial fibrillation, and prescriptions of ␤ blockers and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers.

therapy. Other baseline characteristics were not significantly different between the 2 groups. There was a lower proportion of patients treated with ␤ blockers (57% vs 68%; p ⫽ 0.048) in the digoxin group. Daily doses of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or ␤ blockers did not differ between the 2 groups according to digoxin treatment. Proportions of patients on modest doses of lisinopril (ⱖ10 or ⱖ20 mg/day), metoprolol succinate (ⱖ100 mg/day), or carvedilol (ⱖ12.5 mg/day) were not different between the 2 groups. Admission and discharge laboratory markers, including blood urea nitrogen, glomerular filtration rate, sodium, and brain natriuretic peptide, did not differ according to digoxin use. There was no difference in the cumulative occurrence of the clinical end points of interest or duration of follow-up between the 2 groups. Of the original 155 subjects on digoxin therapy at discharge, 126 (81%) remained on digoxin therapy at 6 months and 112 (72%) remained on digoxin therapy at last follow-up. At last follow-up, 31 patients had been placed on

digoxin therapy. In adjusted and unadjusted analyses, digoxin use was not associated with the combined end point or individual end points of HF hospitalization or total mortality. Digoxin use was associated with a nonsignificant increase in HF hospitalization rates in unadjusted analyses (Table 2). The combined end point of HF hospitalization and total mortality (Figure 1) and individual end points were not different between patients on digoxin therapy and those not on digoxin therapy in any of the prespecified subgroup analyses according to subgroups of EF (ⱕ25% vs ⬎25%), NYHA class (III or IV vs I or II), use versus nonuse of ␤ blockers, presence or absence of atrial fibrillation, and admission or discharge heart rates of ⱕ60 or ⱖ60 beats/min. Discussion In a cohort of hospitalized patients with HF in a real-world setting, we first noted that digoxin was used in patients with

Heart Failure/Digoxin in Contemporary Heart Failure Treatment

more advanced HF, namely those with a lower LVEF, more previous HF hospitalizations, requiring more loop diuretics, and with lower blood pressure. Second, we found that digoxin use was not associated with a decrease in HF hospitalization rates or overall mortality in patients on contemporary HF treatment including concomitant use of angiotensin-converting enzyme inhibitors and ␤ blockers. Additionally, we noted no benefit of digoxin in reducing HF hospitalization or overall mortality rates in patients with different subgroups of LVEF, NYHA class, or ␤-blocker use, specifically even in patients with severe LV systolic dysfunction with LVEF ⱕ25% or advanced HF with NYHA class III or IV. American College of Cardiology/American Heart Association HF guidelines recommended consideration of adding digoxin therapy in patients with persistent symptoms of HF during therapy with diuretics, an angiotensin-converting enzyme inhibitor (or angiotensin receptor blocker), and a ␤ blocker.2 However, clinical data were limited about the beneficial role of digoxin with background therapy with ␤ blockers and angiotensin-converting enzyme inhibitors. The landmark DIG trial, a randomized controlled trial of digoxin in patients with chronic HF on background therapy with diuretics, showed that digoxin did not reduce overall mortality, but reduced the rate of hospitalizations for HF.1 In the DIG trial, the rate of ␤-blocker use was not reported, and at the time the trial was conducted, ␤ blockers were not accepted as standard therapy for patients with HF. Similar to the DIG trial, we did not show a survival benefit. However, contrary to the DIG trial, we failed to show a reduction in HF hospitalizations or any benefit in subgroups of patients with severe LV systolic dysfunction with LVEF ⱕ25% or NYHA class III or IV. It should be noted that similar to other cohorts of hospitalized patients with HF,3 our study population reflected a sicker HF population hospitalized for decompensated HF, with predominantly NYHA class III or IV HF, severely depressed LVEF, and a high prevalence of co-morbidities and previous HF hospitalizations. Notably, our study population was older with a higher proportion of patients with NYHA class III or IV HF1,4 –10 and more co-morbid medical conditions, particularly diabetes and hypertension, than in published trials.1,6 –10 Furthermore, use of diuretics and ␤ blockers in our study cohort was higher than in other former studies examining the role of digoxin.1,6 –10 The difference in our results in failure to show a decrease in HF hospitalizations or any benefit, even in patients with LVEF ⱕ25% or NYHA class III or IV HF, may relate to better background HF therapy with ␤ blockers and/or the higher prevalence of co-morbidities and other risk factors, which may not respond to digoxin. However, most importantly, our findings may be chance related given the multiple analyses performed, the retrospective nature of our cohort, and the limited power of the study, especially in subgroup analyses. Based on data from previous clinical trials,1,4,5 patients not on ␤-blocker therapy would be expected to derive some benefit from digoxin. Because of the smaller number of patients not on ␤-blocker therapy in this analysis, we may not have been able to show a beneficial effect of digoxin on clinical outcomes,1 a type II error. Additionally, although the overall prevalence of ␤-blocker

1359

use was 63% in our cohort, subjects on digoxin therapy had a lower proportion of ␤-blockade use compared with subjects not on digoxin therapy. Although not shown in our cohort, this may relate to the safety concerns of potential bradycardia with concomitant use of ␤ blockers and digoxin or the differences in baseline characteristics and severity of HF. We could not ascertain whether a combination of digoxin and ␤ blockade was detrimental on follow-up with regard to hospitalizations for digitalis toxicity, symptomatic bradycardia, and/or pacemaker requirement. Former studies with digoxin in patients with chronic HF reported improvements in LVEF,4-7 HF symptoms,5,8,9 and exercise performance.4,5 In our study, those were not measured, and thus we cannot exclude potential beneficial effects of digoxin on symptoms, exercise performance, and LV function. Our population consisted of almost exclusively men. Thus, we could not elaborate on the potential risk of digoxin in women with HF.11 Similarly, we did not have data for serum digoxin levels to elaborate on potential harmful effects with higher serum digoxin12 and/or the safety margin of digoxin in the elderly population.13,14 Although individual digoxin doses were not available in the data set, the maximal dose during data collection was 0.25 mg/day. Furthermore, although the age profile of our patients was older than in former trials and almost half our patients had chronic renal insufficiency; we found no evidence for increased rates of admissions for HF or mortality in the overall population or in the subgroup of subjects with glomerular filtration rate ⬍60 ml/min/1.73 m2. Thus, in this cohort of old patients with HF with a high prevalence of co-morbidities, digoxin appeared to have a nonbeneficial and nondetrimental profile regarding HF hospitalizations and overall mortality. Our study had several limitations, including the singlecenter observational retrospective nature and including almost exclusively men. There were significant differences in baseline characteristics, particularly those known to convey poorer outcome in patients on digoxin therapy. Although multivariable statistical models were used to adjust for heterogeneity between groups in this observational study, residual unmeasured confounding factors may be present. Our study population was an older sicker cohort captured in an inpatient setting. Thus, our findings may not be generalizable to a typical clinic cohort. Given the nature of the data set, we were unable to ascertain the length of therapy on digoxin before the index HF admission, and data for serum digoxin levels and optimization of drug doses during follow-up were not available. Crossovers were observed during follow-up, and we did not have data concerning reasons for discontinuation or starting therapy on follow-up. Additionally, we were not able to ascertain reasons for the low rates of defibrillator implantation and use of aldosterone antagonists, although this may reflect the gap between published clinical trials and the translation of results into clinical practice. Acknowledgments: The authors thank Dr. Douglas Mann for support and guidance.

1360

The American Journal of Cardiology (www.AJConline.org)

1. The Digitalis Investigation Group. The effect of digoxin on mortality and morbidity in patients with heart failure. N Engl J Med 1997;336: 525–533. 2. Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS, Ganiats TG, Jessup M, Konstam MA, Mancini DM, Michl K, et al. ACC/AHA 2005 guideline update for the diagnosis and management of chronic heart failure in the adult: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Update the 2001 Guidelines for the Evaluation and Management of Heart Failure): developed in collaboration with the American College of Chest Physicians and the International Society for Heart and Lung Transplantation: endorsed by the Heart Rhythm Society. Circulation 2005;112:e154 – e235. 3. Gheorghiade M, Filippatos G. Reassessing treatment of acute heart failure syndromes: the ADHERE Registry. Eur Heart J Suppl 2005; 7(Suppl B):B13–B19. 4. Uretsky BF, Young JB, Shahidi FE, Yellen LG, Harrison MC, Jolly MK. Randomized study assessing the effect of digoxin withdrawal in patients with mild to moderate chronic congestive heart failure: results of the PROVED trial. PROVED Investigative Group. J Am Coll Cardiol 1993;22:955–962. 5. Packer M, Gheorghiade M, Young JB, Costantini PJ, Adams KF, Cody RJ, Smith LK, Van Voorhees L, Gourley LA, Jolly MK. Withdrawal of digoxin from patients with chronic heart failure treated with angiotensin-converting-enzyme inhibitors. RADIANCE Study. N Engl J Med 1993;329:1–7.

6. The Captopril-Digoxin Multicenter Research Group. Comparative effects of therapy with captopril and digoxin in patients with mild to moderate heart failure. JAMA 1988;259:539 –544. 7. DiBianco R, Shabetai R, Kostuk W, Moran J, Schlant RC, Wright R. A comparison of oral milrinone, digoxin, and their combination in the treatment of patients with chronic heart failure. N Engl J Med 1989;320:677–683. 8. The German and Austrian Xamoterol Study Group. Double-blind placebo-controlled comparison of digoxin and xamoterol in chronic heart failure. Lancet 1988;1:489 – 493. 9. Guyatt GH, Sullivan MJ, Fallen EL, Tihal H, Rideout E, Halcrow S, Nogradi S, Townsend M, Taylor DW. A controlled trial of digoxin in congestive heart failure. Am J Cardiol 1988;61:371–375. 10. Lee DC, Johnson RA, Bingham JB, Leahy M, Dinsmore RE, Goroll AH, Newell JB, Strauss HW, Haber E. Heart failure in outpatients: a randomized trial of digoxin versus placebo. N Engl J Med 19825;306:699 –705. 11. Rathore SS, Wang Y, Krumholz HM. Sex-based differences in the effect of digoxin for the treatment of heart failure. N Engl J Med 2002;347:1403–1411. 12. Rathore SS, Curtis JP, Wang Y, Bristow MR, Krumholz HM. Association of serum digoxin concentration and outcomes in patients with heart failure. JAMA 2003;289:871– 878. 13. Ware JA, Snow E, Luchi JM, Luchi RJ. Effect of digoxin on ejection fraction in elderly patients with congestive heart failure. J Am Geriatr Soc 1984;32:631– 635. 14. Rich MW, McSherry F, Williford WO, Yusuf S. Effect of age on mortality, hospitalizations and response to digoxin in patients with heart failure: the DIG Study. J Am Coll Cardiol 2001;38:806 – 813.