Abstract
Aim:
The authors assessed the comparative effectiveness of torsemide versus furosemide in the PROTECT trial.
Methods:
The authors assessed the relationship between loop diuretic at discharge and death or cardiovascular/renal hospitalization within 30 days, and death through 150 days postdischarge using inverse probability weighting.
Results:
Out of 1004 patients, 83.5% received furosemide and 16.5% torsemide. Torsemide patients had higher blood urea nitrogen, and more in-hospital worsening heart failure. Following adjustment, torsemide was associated with similar 30-day outcomes compared with furosemide (p = 0.93), but remained associated with increased 150-day death (hazard ratio: 2.26; 95% CI: 1.40–3.66; p < 0.001).
Conclusion:
Patients treated with torsemide had features of greater disease severity, similar 30-day outcomes but increased 150-day mortality. Prospective randomized trials are needed to investigate the effect of torsemide versus furosemide.
KEYWORDS : furosemide, heart failure, loop diuretics, outcomes, torsemide
Furosemide has historically been the primary loop diuretic used for the management of volume status in heart failure (HF) patients. However, other loop diuretics such as torsemide have increased half-life and more consistent bioavailability compared with furosemide [1]. Moreover, torsemide may have beneficial effects on myocardial fibrosis, ventricular structure and the neurohormonal milieu [2–7]. Several small observational studies from more than a decade ago [8–10] and a recent meta-analysis [11] suggest a reduction in morbidity and potentially mortality with torsemide compared with furosemide. Despite these data, furosemide remains the most commonly used loop diuretic [12] likely related in part to greater clinician experience with furosemide, historically lower medication cost and a lack of contemporary data supporting a benefit with torsemide. Therefore, the authors assessed the association between loop diuretics and postdischarge outcomes in a large, international acute HF trial. The authors hypothesized that use of torsemide at discharge would be associated with improved outcomes compared with furosemide use.
Methods
• Data source
The international PROTECT trial enrolled 2033 patients admitted to the hospital with acute HF and mild or moderate renal impairment. The design and results of PROTECT have been published [13,14]. Briefly, the inclusion criteria were: prior history of HF treated for at least 14 days with diuretic therapy; hospitalization for worsening breathlessness due to HF requiring intravenous diuretic therapy; admission creatinine clearance 20–80 ml/min by the Cockcroft-Gault equation; a brain natriuretic peptide (BNP) level ≥500 pg/ml or N-terminal pro-BNP level ≥2000 pg/ml; and systolic blood pressure (SBP) ≥95 mmHg. Key exclusion criteria were: use of inotropic agents or mechanical support; severe pulmonary disease; significant stenotic valvular disease; recent acute coronary syndrome or significant arrhythmias; and history of seizure/stroke within 2 years. Patients were enrolled from 2007–2009 and randomly assigned to the intravenous administration of the A1-receptor antagonist rolofylline or placebo. Loop diuretic was documented at baseline and hospital discharge/day 14. Worsening HF (WHF) was defined as a physician-determined assessment of worsening symptoms or signs of HF occurring greater than 24 h after the start of study drug to day 7 or discharge, whichever occurred first, that required institution or an increase in dose of intravenous or mechanical therapy for HF. The primary endpoint for the trial was treatment success (moderate or marked improvement in dyspnea at both 24 and 48 h after rolofylline administration) in the absence of any criterion for treatment failure (death or readmission for HF, WHF, persistent worsening renal function or initiation of dialysis up to and including day 7) or no change in the patient's clinical condition up to and including day 7. Death from any cause or rehospitalization for cardiovascular or renal causes through day 60 was a prespecified secondary endpoint. Vital status was also assessed at 180 days. An independent clinical events committee adjudicated the primary reason for rehospitalization and cause of death through day 60. The investigation conforms with the principles outlined in the Declaration of Helsinki. The PROTECT study was approved by the appropriate regulatory authorities and ethics committees prior to patient enrollment, and written informed consent was obtained from each patient before entry [15,16].
• Analysis cohort
For the present analysis, the authors limited the cohort to patients discharged alive on either furosemide or torsemide with follow-up after day 14. Given significant regional variation in the use of torsemide, the authors restricted the analyses to those countries where patients were treated with both furosemide and torsemide.
• Outcomes of interest
The outcomes for the present analysis were death or cardiovascular/renal hospitalization through 30-day postdischarge and death through 150-day postdischarge. The authors used different timepoints compared with the main PROTECT analysis (through 60- and 180-day postrandomization, respectively), since the authors were assessing outcomes associated with discharge loop diuretic rather than use at randomization. Similar methods have been used in previous PROTECT analyses to reduce bias related to shorter follow-up in those patients with longer index hospital length of stay (LOS) [17]. As additional endpoints, index hospital LOS, treatment success/no change/worsening (primary endpoint for the PROTECT trial) and event rates for WHF to 7 days were investigated. The authors were also interested in identifying clinical factors associated with patients being discharged on torsemide as compared with furosemide.
• Statistical methods
Demographics, physical and laboratory findings, medical history and therapies were summarized as frequencies and percentages for categorical variables and by the medians and 25th and 75th percentiles for continuous variables in patients discharged on either torsemide or furosemide. Diuretic dose was assessed with the conversion of 2 mg furosemide to 1 mg torsemide to present furosemide equivalents as described previously [1]. Patient characteristics were compared between groups using the Wilcoxon rank sum test for continuous variables, and χ2 or Fisher's mid-P tests for categorical variables as appropriate. The authors assessed the number of events for the outcomes of interest based on discharge diuretic. Cumulative event rates at the end of the study period, and cumulative event rate curves over time postdischarge were estimated using the Kaplan–Meier method.
Because the choice of diuretic at discharge was not randomized, the authors developed a propensity score logistic regression model to predict the use of torsemide versus furosemide. Covariates associated with postdischarge mortality were identified in previous analysis of PROTECT [17,18] and included age, country, hospitalization for HF in the past year, history of diabetes mellitus, history of atrial fibrillation/flutter, baseline measures (BMI, mean blood pressure [BP], blood urea nitrogen [BUN]) and day 14 measures (EQ5D score, congestion score, albumin, ALT, BUN and total cholesterol). Cox proportional hazards models were generated to assess the association between loop diuretic use at discharge and subsequent clinical outcomes. Hazard ratios (HRs) for torsemide versus furosemide were calculated with corresponding 95% CIs. The estimated propensity scores were used for inverse propensity weighting (IPW) of observations in these models [19]. IPW methods reweight the data to create a pseudo-population where, under the assumption that important confounders have been accounted for, patient characteristics are independent of treatment received [20]. Multivariate regression analysis adjusting for covariates above was also performed as a secondary analysis. Statistical significance was assessed using two-sided p-values (p < 0.05 was considered statistically significant). All statistical computations were generated by Duke Clinical Research Institute (Durham, NC, USA) using SAS version 9.2 or higher (SAS Institute Inc., Cary, NC, USA). No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the paper and its final contents.
Results
Figure 1 presents the patients included in this analysis. Out of 2033 patients in the PROTECT trial, 1876 patients were discharged on a loop diuretic. Furosemide was used in 1612 patients (86.0%), torsemide in 166 patients (8.8%) and bumetanide in only 98 patients (5.2%). The 11 enrolling countries with no use of torsemide at discharge were Belgium (total n = 6), France (n = 72), The Netherlands (n = 13), Sweden (n = 6), UK (n = 22), Czech Republic (n = 157), Hungary (n = 24), Romania (n = 86), Argentina (n = 52), Israel (n = 303) and Canada (n = 33). In the outcomes analysis, patients were included from six countries that discharged patients on both furosemide and torsemide: the USA, Poland, Russia, Germany, Italy and the Ukraine (Table 1). Out of the 1004 patients discharged alive on either furosemide or torsemide, from countries in which both drugs were used, 83.5% (n = 838) received furosemide and 16.5% (n = 166) received torsemide. Patients enrolled in Germany and the USA constituted the majority of the overall torsemide cohort (n = 87) where there was 61.8 and 13.1% use of torsemide, respectively. The median daily dose in the patients discharged on torsemide was 15 mg (IQR: 10–40), which is 30 mg (IQR: 20–80) when converted to furosemide equivalents, compared with 80 mg (IQR: 40–80) in the patients discharged on furosemide. Patient characteristics by discharge diuretic are presented in Table 1. Patients discharged on torsemide were more likely to have an HF hospitalization in the previous year and had more atrial fibrillation than patients discharged on furosemide. Torsemide patients had worse congestion as evidenced by more severe dyspnea on exertion, greater elevation in jugular venous pressure (JVP) and worse orthopnea compared with furosemide-treated patients. Patients receiving torsemide had higher BMI and BUN as well as lower mean blood pressure compared with patients discharged on furosemide. In general, HF medication use was similar in both patient groups, while implantable cardioverter defibrillator (ICD) use was greater in the torsemide patients. The baseline characteristics of patients excluded due to the lack of torsemide use in the enrolling country are presented in Supplementary Table 1. Compared with the analysis cohort, patients excluded from the analysis were older, had more diabetes and ICD use, higher blood pressure and worse renal function.
Figure 1. . Study population.
D/C: Indicates discharge.
Table 1. . Characteristics of the study population by discharge diuretic.
| Characteristic | Torsemide (n = 166) | Furosemide (n = 838) | p-value |
|---|---|---|---|
| Age (years) |
70 (60; 77) |
69 (60; 77) |
0.745 |
| Male sex |
118 (71.1) |
574 (68.5) |
0.510 |
| Country: | <0.001 | ||
| – Germany (n = 89, 8.9%) | 55 (61.8) | 34 (38.2) | |
| – Italy (n = 51, 5.1%) | 4 (7.8) | 47 (92.2) | |
| – Poland (n = 323, 32.2%) | 23 (7.1) | 300 (92.9) | |
| – Russia (n = 267, 26.6%) | 31 (11.6) | 236 (88.4) | |
| – Ukraine (n = 30, 2.9%) | 21 (70.0) | 9 (30.0) | |
| – USA (n = 244, 24.3%) |
32 (13.1) |
212 (86.9) |
|
| HF history: | |||
| – HF hospitalization within prior year | 102 (61.4) | 440 (52.5) | 0.035 |
| – Ischemic etiology | 112 (67.5) | 593 (70.8) | 0.397 |
| – Ejection fraction (%) | 30 (23, 41) | 28 (20, 38) | 0.236 |
| – Ejection fraction <40% | 64 (69.6) | 302 (76.6) | 0.156 |
| – NYHA class 1 month prior: | 0.003 | ||
| • None/I/II | 40 (24.1) | 115 (13.7) | |
| • III | 62 (37.3) | 364 (43.5) | |
| • IV | 64 (38.6) | 358 (42.8) | |
| – EQ-5D utility score (day 14) |
0.78 (0.57; 0.84) |
0.79 (0.63; 0.85) |
0.011 |
| Past medical history | |||
| – Hypertension | 137 (82.5) | 680 (81.1) | 0.675 |
| – Diabetes mellitus | 81 (48.8) | 353 (42.1) | 0.113 |
| – Atrial fibrillation or flutter | 107 (64.5) | 443 (53.2) | 0.008 |
| – Stroke or PVD |
31 (18.8) |
164 (19.6) |
0.811 |
| Signs and symptoms of HF on latest of days 5–14: | |||
| – Dyspnea on exertion: | 0.008 | ||
| • None | 4 (2.4) | 88 (10.6) | |
| • Mild | 73 (44.0) | 362 (43.5) | |
| • Moderate | 81 (48.8) | 348 (41.8) | |
| • Severe | 8 (4.8) | 34 (4.1) | |
| – Edema: | 0.054 | ||
| • 0 | 102 (61.4) | 587 (70.5) | |
| • 1+ | 40 (24.1) | 176 (21.1) | |
| • 2+ | 19 (11.4) | 54 (6.5) | |
| • 3+ | 5 (3.0) | 16 (1.9) | |
| –Rales: | 0.129 | ||
| • No rales | 132 (79.5) | 711 (85.4) | |
| • Rales <1/3 | 31 (18.7) | 115 (13.8) | |
| • Rales 1/3–2/3 | 3 (1.8) | 7 (0.8) | |
| – JVP: | <0.001 | ||
| • <6 cm | 72 (55.0) | 571 (75.3) | |
| • 6–10 cm | 43 (32.8) | 157 (20.7) | |
| • >10 cm | 16 (12.2) | 30 (4.0) | |
| – Orthopnea: | 0.002 | ||
| • None | 45 (27.1) | 353 (42.4) | |
| • One pillow (10 cm) | 69 (41.6) | 298 (35.8) | |
| • Two pillows (20 cm) | 40 (24.1) | 144 (17.3) | |
| • >30 degrees | 12 (7.2) | 38 (4.6) | |
| – Congestion score† |
3.5 (2.0; 6.5) |
3.0 (2.0; 4.0) |
<0.001 |
| Physical examination: | |||
| – BMI (kg/m2) | 29.9 (25.7; 33.9) | 27.9 (24.4; 32.7) | 0.006 |
| – Systolic BP (mmHg) | 120 (110; 140) | 124 (110; 140) | 0.274 |
| – Diastolic BP (mmHg) | 72 (63; 80) | 76 (70; 80) | 0.024 |
| – Mean BP (mmHg) | 89.5 (81.3; 97.7) | 91.7 (83.3; 100.0) | 0.045 |
| – Heart rate (bpm) |
80 (70; 91) |
80 (70; 94) |
0.151 |
| Laboratories on day 1: | |||
| – Sodium (mmol/l) | 140 (138; 142) | 140 (137; 143) | 0.512 |
| – Creatinine (mg/dl) | 1.4 (1.2; 1.8) | 1.3 (1.1; 1.7) | 0.094 |
| – eGFR‡ (ml/min/1.73 m2) | 47.6 (34.8; 58.0) | 49.6 (37.1; 62.1) | 0.061 |
| – BUN (mg/dl) | 32 (25; 41) | 27 (21; 38) | <0.001 |
| – Hemoglobin (g/dl) | 12.5 (11.2; 13.9) | 12.9 (11.5; 14.3) | 0.088 |
| – Albumin (g/dl) | 3.8 (3.6; 4.1) | 3.8 (3.6; 4.1) | 0.320 |
| – Total cholesterol (mg/dl) | 144 (117; 175) | 142 (118; 175) | 0.723 |
| – ALT (U/l) |
19 (15; 33) |
23 (15; 34) |
0.158 |
| Latest laboratories of days 5–14: | |||
| – Sodium (mmol/l) | 139 (136; 142) | 139 (136; 142) | 0.693 |
| – Creatinine (mg/dl) | 1.4 (1.2; 1.8) | 1.3 (1.1; 1.8) | 0.151 |
| – eGFR‡ (ml/min/1.73 m2) | 47.0 (33.8; 59.0) | 48.9 (35.5; 64.8) | 0.080 |
| – BUN (mg/dl) | 35 (25; 45) | 30 (22; 43) | 0.024 |
| – Hemoglobin (g/dl) | 12.7 (11.1; 14.4) | 13.2 (11.7; 14.9) | 0.016 |
| – Albumin (g/dl) | 4.0 (3.7; 4.3) | 4.0 (3.7; 4.3) | 0.633 |
| –Total cholesterol (mg/dl) | 150.5 (119.0; 191.0) | 156.0 (126.0; 189.0) | 0.392 |
| – ALT (U/l) |
22 (16; 30) |
23 (16; 32) |
0.767 |
| Baseline medications (two weeks prior to admission) and devices: | |||
| – ACE-inhibitor or ARB | 131 (78.9) | 610 (72.8) | 0.101 |
| – β-blocker | 133 (80.1) | 626 (74.7) | 0.138 |
| – Aldosterone antagonists | 69 (41.6) | 377 (45.0) | 0.418 |
| – Nitrates | 41 (24.7) | 227 (27.1) | 0.525 |
| – Digoxin | 60 (36.1) | 261 (31.1) | 0.207 |
| – Implantable cardioverter defibrillator | 44 (26.5) | 141 (16.8) | 0.003 |
| – Biventricular pacemaker |
19 (11.4) |
77 (9.2) |
0.369 |
| Medications at discharge (or day 7 if earlier): | |||
| – ACE-inhibitor or ARB | 138 (83.1) | 723 (86.3) | 0.290 |
| – β-blocker | 146 (88.0) | 737 (87.9) | 0.999 |
| – Aldosterone antagonists | 104 (62.7) | 541 (64.6) | 0.639 |
| – Nitrates | 28 (16.9) | 189 (22.6) | 0.102 |
| – Digoxin | 67 (40.4) | 313 (37.4) | 0.472 |
Values presented as median (IQR) or n (percentage of group), except where indicated otherwise.
For country, the data are presented as the number of patients (%) on each medication such that the row adds up to 100% for each country.
†The congestion score is an equally weighted aggregate of the scores for edema, rales, JVP, orthopnea and dyspnea on exertion.
‡eGFR calculated from central lab using MDRD formula.
ARB: Angiotensin receptor blocker; BP: Blood pressure; BUN: Blood urea nitrogen; EF: Ejection fraction; eGFR: Estimated glomerular filtration rate; EQ-5D: EuroQol questionnaire measuring health outcome; HF: Heart failure; JVP: Jugular venous pressure; NYHA: New York Heart Association; PVD: Peripheral vascular disease.
Factors associated with torsemide use at discharge are presented in Table 2. Country was the primary factor associated with torsemide use. The odds for torsemide use were highest in the Ukraine and Germany compared with the USA. Clinical factors associated with torsemide use included HF hospitalization in the previous year, mean blood pressure at baseline and congestion score at day 14.
Table 2. . Variables associated with torsemide use at discharge.
| Covariate | Multivariable odds ratio | 95% CI | χ2 | p-value |
|---|---|---|---|---|
| Age (5-year increase) |
0.96 |
0.88–1.06 |
0.597 |
0.440 |
| Country (vs USA): | 140.91 | <.001 | ||
| – Germany | 15.6 | 8.18–29.6 | ||
| – Italy | 0.49 | 0.15–1.55 | ||
| – Poland | 0.67 | 0.36–1.25 | ||
| – Russia | 1.39 | 0.74–2.60 | ||
| – Ukraine |
30.8 |
11.2–84.2 |
|
|
| Hospitalization for HF in past year (yes vs no) |
1.77 |
1.17–2.68 |
7.196 |
0.007 |
| History of diabetes mellitus (yes vs no) |
0.83 |
0.54–1.29 |
0.679 |
0.410 |
| History of atrial fib/flutter (yes vs no) |
1.50 |
0.98–2.30 |
3.504 |
0.061 |
| BMI at baseline (5 kg/m2 increase) |
1.03 |
0.88–1.20 |
0.126 |
0.723 |
| Mean BP at baseline (20 mmHg increase) |
0.65 |
0.45–0.94 |
5.377 |
0.020 |
| BUN at baseline (10 mg/dl increase) |
1.15 |
0.99–1.34 |
3.292 |
0.070 |
| BUN at day 14 (10 mg/dl increase) |
0.97 |
0.86–1.10 |
0.171 |
0.679 |
| Albumin at day 14 (0.5 g/dl increase) |
1.14 |
0.89–1.47 |
1.081 |
0.298 |
| ALT at day 14, U/l below 200 (10 U/l increase) |
1.05 |
0.96–1.15 |
1.266 |
0.260 |
| Total cholesterol at day 14 (10 mg/dl increase) |
0.99 |
0.94–1.04 |
0.236 |
0.627 |
| EQ5D utility score at day 14 (0.1 unit increase) |
0.99 |
0.89–1.09 |
0.067 |
0.795 |
| Congestion score at day 14 (1 unit increase) | 1.10 | 1.00–1.20 | 4.151 | 0.042 |
BP: Blood pressure; BUN: Blood urea nitrogen; EQ-5D: EuroQol EQ-5D questionnaire measuring health outcome; HF: Heart failure.
In-hospital outcomes by loop diuretic use at discharge are presented in Table 3. Index hospitalization length of stay was similar between the two groups. The event rate for in-hospital WHF was significantly higher in the patients that were later discharged on torsemide compared with the furosemide-treated group (14.5 vs 9.1%, p = 0.034). During index hospitalization, patients who were eventually discharged on torsemide had a higher percentage of treatment failure or unchanged status and less treatment success for the primary outcome of the original clinical trial of rolofylline compared with those patients discharged on furosemide.
Table 3. . Summary of in-hospital outcomes by loop diuretic use at discharge.
| Outcomes | Overall (n = 1004) | Torsemide (n = 166) | Furosemide (n = 838) | p-value† |
|---|---|---|---|---|
| Index hospitalization length of stay (days; IQR) |
10 (7, 15) |
9 (7, 15) |
10 (6, 15) |
0.43 |
| WHF through 7 days; n (%) |
100 (10.0) |
24 (14.5) |
76 (9.1) |
0.034 |
| Treatment success outcome; n (%) | 0.002 | |||
| – Failure | 166 (16.5) | 33 (19.9) | 133 (15.9) | |
| – Unchanged | 431 (42.9) | 85 (51.2) | 346 (41.3) | |
| – Success | 407 (40.5) | 48 (28.9) | 359 (42.8) |
†p-value for length of stay from Wilcoxon test. p-value for WHF from χ2 test. p-value for treatment success outcome from nonparametric ANOVA test.
IQR: median; WHF: Worsening heart failure.
Table 4 presents postdischarge outcomes in patients treated with furosemide or torsemide. Death or cardiovascular/renal hospitalization through 30-day and 150-day death were increased in patients discharged on torsemide compared with furosemide (19.1 vs 15.1% [p = 0.25] and 18.7 vs 9.1% [p < 0.001], respectively). Figure 2 displays the unadjusted event rate curves. Following adjustment for treatment propensity, torsemide use was associated with similar 30-day outcomes compared with furosemide (HR: 1.02; 95% CI: 0.64–1.63; p = 0.93), and increased 150-day death (HR: 2.26; 95% CI: 1.40–3.66; p<.001). Multivariate regression analysis yielded similar results to those with IPW methodology. Additional adjustment for index hospitalization length of stay and in-hospital WHF yielded similar results.
Table 4. . Outcomes associated with torsemide use at discharge (reference = furosemide).
| Endpoint |
Number of events/sample size (KM%) |
HR (95% Cl) | p-value | |
|---|---|---|---|---|
| Torsemide | Furosemide | |||
| Death or CV/renal hospitalization through 30-day postdischarge: | 31/165 (19.1%) | 126/836 (15.1%) | ||
| – Unadjusted | 1.261 (0.851–1.868) | 0.248 | ||
| – Adjusted IPW† | 1.020 (0.640–1.626) | 0.932 | ||
| – Covariate adjusted‡ |
|
|
0.860 (0.540–1.368) |
0.523 |
| Death through 150-day postdischarge: | 31/166 (18.7%) | 76/838 (9.1%) | ||
| – Unadjusted | 2.184 (1.438–3.316) | <0.001 | ||
| – Adjusted IPW† | 2.259 (1.396–3.657) | <0.001 | ||
| – Covariate adjusted‡ | 2.141 (1.307–3.507) | 0.002 | ||
Figure 2. . Unadjusted event rate curves by discharge loop diuretic.
Discussion
In a large international acute HF trial, the authors found that furosemide was the primary loop diuretic used in most countries (>85% use), except for Germany and the Ukraine which used greater than 60% torsemide. The observation that no patients in 11 of the countries that enrolled in PROTECT had any patients discharged on torsemide and only 5% of patients were discharged on bumetanide, highlights the substantial regional variation in the availability, cost and use of alternative loop diuretics. The furosemide equivalent dose was lower in the patients discharged on torsemide. Patients discharged on torsemide tended to have features of more severe disease, particularly prior HF hospitalization, worse congestion, higher BUN and more in-hospital WHF and overall in-hospital treatment failure compared with furosemide-treated patients. After adjustment for treatment propensity, torsemide was associated with similar 30-day outcomes but increased 150-day mortality compared with furosemide. Due to the potential for residual confounding, these data, which conflict with earlier studies [21], provide the equipoise for a prospective randomized trial that is adequately powered to investigate the effect of torsemide versus furosemide on clinical outcomes.
A primary finding of our analysis was that patients discharged on torsemide had features of more severe disease compared with furosemide patients. In particular, patients who received torsemide tended to have more significant congestion as well as higher BMI and BUN, and lower mean blood pressure. BUN captures renal dysfunction as well as metabolic status and neurohormonal activation suggesting worse systemic perturbations in patients treated with torsemide. Patients discharged on torsemide also were more likely to have been hospitalized for HF in the preceding year and had an increased comorbidity burden including atrial fibrillation. Prior HF hospitalization is a strong predictor of increased risk for future adverse events [22]. In addition, in-hospital WHF, which has been associated with worse long-term outcomes [23], was seen more often in the patients that were discharged on torsemide compared with those treated with furosemide. Interestingly, other HF medication use was similar between the two groups, while ICD therapy was increased in the torsemide group. This finding may have been related to between-group differences in the New York Heart Association (NYHA) class, prior HF hospitalizations and/or regional differences in the implantation rates of devices. Thus, despite similar background HF therapy in the torsemide and furosemide treated patients, patients discharged on torsemide had a high-risk profile including prior HF hospitalizations and in-hospital WHF.
Clinical factors strongly associated with torsemide use included prior HF hospitalization, congestion score and BUN. These findings suggest that clinicians use torsemide in the setting of perceived failure to respond to furosemide with refractory volume overload. Interestingly, the furosemide equivalent dose was lower in those patients discharged on torsemide. Torsemide has more consistent bioavailability compared with furosemide including in patients with intestinal edema [1]. Despite these pharmacologic advantages with torsemide, furosemide remains the most commonly used loop diuretic in most world regions given greater clinical experience and a historical patient-level cost advantage with furosemide. These two drugs now have comparable costs given the availability of generic torsemide. In fact, studies prior to the widespread availability of generic torsemide [24] suggested that torsemide may be associated with a reduction in total cost per patient due to reduced rehospitalization. However, the authors did not observe a rehospitalization benefit associated with torsemide use in their analysis.
Despite earlier studies demonstrating potential benefits with torsemide over furosemide [2–7], the authors did not observe improved outcomes associated with discharge torsemide prescription. As recently reviewed [21], studies have suggested beneficial effects of torsemide related to inhibition of aldosterone secretion and/or receptor binding and reduced myocardial collagen production. The authors did not observe a beneficial association with torsemide use in the present analysis, the specific reasons for which are unclear. The authors hypothesize that measured and unmeasured variables may have influenced these results and that the statistical methods were not able to adequately adjust for the increased severity of disease in torsemide patients. It is also possible that the antifibrotic effects that were observed in earlier studies with torsemide [6,25–26] may not result in improved clinical outcomes until there is a longer duration of exposure, particularly in the setting of contemporary guideline-directed HF medications including mineralocorticoid receptor antagonists. Additionally, the implications of a differential furosemide equivalent dose between the two study groups may have influenced the findings. A higher dose of torsemide than that used in this patient population (median dose of 15 mg torsemide; or 30 mg furosemide equivalent) may be necessary to mediate the beneficial effects. Previous studies have not demonstrated improved outcomes in association with diuretic therapies for HF patients [27]. Furthermore, data are not available for this cohort regarding medication crossover following discharge. Moreover, given the strong association between country and diuretic choice, these data may be confounded by international differences as has been described previously in the PROTECT trial [28] and additional recent heart failure trials including TOPCAT [29].
• Limitations
This was a retrospective analysis from a clinical trial and patients met specific entry criteria. These data may not apply to patients with different clinical characteristics to those included in the PROTECT trial. In particular, the regional variation in the use of torsemide is a major potential confounder [28]. Despite IPW adjustment, other measured and unmeasured variables may have influenced these results. The analysis population receiving torsemide was relatively modest in size and the analysis may have been underpowered. It was not a new-user design given the common use of diuretics to treat symptoms of volume overload. Furthermore, data were not available regarding postdischarge diuretic changes including medication type and dose. Thus, these data should be viewed as exploratory.
Conclusion
In a large international HF trial, patients discharged on torsemide had features of greater disease severity. After adjustment for treatment propensity, torsemide was associated with similar 30-day outcomes but increased 150-day mortality compared with furosemide.
Future perspective
Despite prior observational data suggesting improved outcomes with torsemide compared with furosemide, a relationship between torsemide use and clinical benefits was not observed in the present analysis. Randomized trials would be needed to support the preferential use of torsemide over furosemide.
EXECUTIVE SUMMARY.
In a large international heart failure trial, patients discharged on torsemide had features of greater disease severity than those discharged on furosemide.
After adjustment for treatment propensity, torsemide was associated with similar 30-day outcomes but increased 150-day mortality compared with furosemide.
These data provide the rationale for an appropriately powered randomized clinical trial of the comparative–effectiveness of furosemide versus torsemide in heart failure patients.
Supplementary Material
Footnotes
Financial & competing interests disclosure
RJ Mentz receives research support from the NIH, Amgen, AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Gilead, Novartis, Otsuka, and ResMed; honoraria from Novartis, Thoratec and HeartWare; and has served on an advisory board for Luitpold Pharmaceuticals, Inc. EJ Velazquez has received research grants and/or consultancy fees from the NHLBI, Medtronic, Pfizer, Alnylam, Abbott Vascular and Novartis. AA Voors received consultancy fees and/or research grants from: Alere, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Cardio3Biosciences, Celladon, Merck/MSD, Novartis, Servier, Sphingotec, Singulex, Torrent, Trevena, Vifor. M Metra received consulting incomes from Bayer, Novartis, Servier and honoraria for speeches from Abbott Vascular. JR Teerlink has received research funds and consulting fees from Merck and has also received research funds and consulting fees from Abbott, Amgen, Biogen Idec, Corthera, Cytokinetics, Johnson & Johnson/Scios, Novartis, Relypsa and Solvay for research in related areas. MM Givertz has received institutional research support and served on a scientific Advisory Board for Merck. CM O'Connor is a consultant to Merck and Amgen. C McKendry was supported by NIH grant T32 HL079896. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
Ethical conduct of research
The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.
References
Papers of special note have been highlighted as: • of interest; •• of considerable interest
- 1.Felker GM, Mentz RJ. Diuretics and ultrafiltration in acute decompensated heart failure. J. Am. Coll. Cardiol. 2012;59(24):2145–2153. doi: 10.1016/j.jacc.2011.10.910. [DOI] [PubMed] [Google Scholar]; • State-of-the-art review of diuretics in heart failure patients.
- 2.Clasen W, Khartabil T, Imm S, Kindler J. Torasemide for diuretic treatment of advanced chronic renal failure. Arzneimittel-Forschung. 1988;38(1A):209–211. [PubMed] [Google Scholar]
- 3.Uchida T, Yamanaga K, Nishikawa M, Ohtaki Y, Kido H, Watanabe M. Antialdosteronergic effect of torasemide. Eur. J. Pharmacol. 1991;205(2):145–150. doi: 10.1016/0014-2999(91)90812-5. [DOI] [PubMed] [Google Scholar]
- 4.Goodfriend TL, Ball DL, Oelkers W, Bahr V. Torsemide inhibits aldosterone secretion in vitro . Life Sci. 1998;63(3):PL45–PL50. doi: 10.1016/s0024-3205(98)00265-3. [DOI] [PubMed] [Google Scholar]
- 5.Harada K, Izawa H, Nishizawa T, et al. Beneficial effects of torasemide on systolic wall stress and sympathetic nervous activity in asymptomatic or mildly symptomatic patients with heart failure: comparison with azosemide. J. Cardiovasc. Pharmacol. 2009;53(6):468–473. doi: 10.1097/FJC.0b013e3181a717f7. [DOI] [PubMed] [Google Scholar]
- 6.Lopez B, Querejeta R, Gonzalez A, Sanchez E, Larman M, Diez J. Effects of loop diuretics on myocardial fibrosis and collagen type I turnover in chronic heart failure. J. Am. Coll. Cardiol. 2004;43(11):2028–2035. doi: 10.1016/j.jacc.2003.12.052. [DOI] [PubMed] [Google Scholar]; • Key mechanistic data supporting a beneficial role of torsemide in heart failure.
- 7.Kasama S, Toyama T, Hatori T, et al. Effects of torasemide on cardiac sympathetic nerve activity and left ventricular remodelling in patients with congestive heart failure. Heart. 2006;92(10):1434–1440. doi: 10.1136/hrt.2005.079764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Murray MD, Deer MM, Ferguson JA, et al. Open-label randomized trial of torsemide compared with furosemide therapy for patients with heart failure. Am. J. Med. 2001;111(7):513–520. doi: 10.1016/s0002-9343(01)00903-2. [DOI] [PubMed] [Google Scholar]
- 9.Cosin J, Diez J. Torasemide in chronic heart failure: results of the TORIC study. Eur. J. Heart Fail. 2002;4(4):507–513. doi: 10.1016/s1388-9842(02)00122-8. [DOI] [PubMed] [Google Scholar]; •• Large observational study highlighting potential clinical benefits with torsemide.
- 10.Muller K, Gamba G, Jaquet F, Hess B. Torasemide vs. Furosemide in primary care patients with chronic heart failure NYHA II to IV–efficacy and quality of life. Eur. J. Heart Fail. 2003;5(6):793–801. doi: 10.1016/s1388-9842(03)00150-8. [DOI] [PubMed] [Google Scholar]
- 11.Dinicolantonio JJ. Should torsemide be the loop diuretic of choice in systolic heart failure? Future Cardiol. 2012;8(5):707–728. doi: 10.2217/fca.12.54. [DOI] [PubMed] [Google Scholar]; • Summary of prior outcomes data with torsemide.
- 12.Bikdeli B, Strait KM, Dharmarajan K, et al. Dominance of furosemide for loop diuretic therapy in heart failure: time to revisit the alternatives? J. Am. Coll. Cardiol. 2013;61(14):1549–1550. doi: 10.1016/j.jacc.2012.12.043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Weatherley BD, Cotter G, Dittrich HC, et al. Design and rationale of the PROTECT study: a placebo-controlled randomized study of the selective α1 adenosine receptor antagonist rolofylline for patients hospitalized with acute decompensated heart failure and volume overload to assess treatment effect on congestion and renal function. J. Card. Fail. 2010;16(1):25–35. doi: 10.1016/j.cardfail.2009.10.025. [DOI] [PubMed] [Google Scholar]
- 14.Massie BM, O'Connor CM, Metra M, et al. Rolofylline, an adenosine α1-receptor antagonist, in acute heart failure. N. Engl. J. Med. 2010;363(15):1419–1428. doi: 10.1056/NEJMoa0912613. [DOI] [PubMed] [Google Scholar]
- 15.https://clinicaltrials.gov/ct2/show/NCT00328692 ClinicalTrials Database: NCT00328692.
- 16.https://clinicaltrials.gov/ct2/show/NCT00354458?term=NCT00354458&rank=1 ClinicalTrials Database: NCT00354458.
- 17.Cleland JG, Chiswell K, Teerlink JR, et al. Predictors of postdischarge outcomes from information acquired shortly after admission for acute heart failure: a report from the placebo-controlled randomized study of the selective α1 adenosine receptor antagonist rolofylline for patients hospitalized with acute decompensated heart failure and volume overload to assess treatment effect on congestion and renal function (PROTECT) study. Circ. Heart Fail. 2014;7(1):76–87. doi: 10.1161/CIRCHEARTFAILURE.113.000284. [DOI] [PubMed] [Google Scholar]
- 18.O'Connor CM, Mentz RJ, Cotter G, et al. The PROTECT in-hospital risk model: 7 day outcome in patients hospitalized with acute heart failure and renal dysfunction. Eur. J. Heart Fail. 2012;14(6):605–612. doi: 10.1093/eurjhf/hfs029. [DOI] [PubMed] [Google Scholar]
- 19.Lunceford JK, Davidian M. Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study. Stat. Med. 2004;23(19):2937–2960. doi: 10.1002/sim.1903. [DOI] [PubMed] [Google Scholar]
- 20.Curtis LH, Hammill BG, Eisenstein EL, Kramer JM, Anstrom KJ. Using inverse probability-weighted estimators in comparative–effectiveness analyses with observational databases. Med. Care. 2007;45(10):S103–S107. doi: 10.1097/MLR.0b013e31806518ac. [DOI] [PubMed] [Google Scholar]
- 21.Buggey J, Mentz RJ, Pitt B, et al. A reappraisal of loop diuretic choice in heart failure patients. Am. Heart J. 2015;169(3):323–333. doi: 10.1016/j.ahj.2014.12.009. [DOI] [PMC free article] [PubMed] [Google Scholar]; •• Recent review of the mechanistic and clinical data to support torsemide use in heart failure.
- 22.Solomon SD, Dobson J, Pocock S, et al. Influence of nonfatal hospitalization for heart failure on subsequent mortality in patients with chronic heart failure. Circulation. 2007;116(13):1482–1487. doi: 10.1161/CIRCULATIONAHA.107.696906. [DOI] [PubMed] [Google Scholar]
- 23.Devore AD, Hammill BG, Sharma PP, et al. In-hospital worsening heart failure and associations with mortality, readmission, and healthcare utilization. J. Am. Heart Assoc. 2014;3(4) doi: 10.1161/JAHA.114.001088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Stroupe KT. Healthcare costs of patients with heart failure treated with torasemide or furosemide. PharmacoEconomics. 2000;17(5):429–440. doi: 10.2165/00019053-200017050-00002. [DOI] [PubMed] [Google Scholar]
- 25.Lopez B, Querejeta R, Gonzalez A, Beaumont J, Larman M, Diez J. Impact of treatment on myocardial lysyl oxidase expression and collagen cross-linking in patients with heart failure. Hypertension. 2009;53(2):236–242. doi: 10.1161/HYPERTENSIONAHA.108.125278. [DOI] [PubMed] [Google Scholar]
- 26.Lopez B, Gonzalez A, Beaumont J, Querejeta R, Larman M, Diez J. Identification of a potential cardiac antifibrotic mechanism of torasemide in patients with chronic heart failure. J. Am. Coll. Cardiol. 2007;50(9):859–867. doi: 10.1016/j.jacc.2007.04.080. [DOI] [PubMed] [Google Scholar]
- 27.Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association task force on practice guidelines. J. Am. Coll. Cardiol. 2013;62(16):e147–e239. doi: 10.1016/j.jacc.2013.05.019. [DOI] [PubMed] [Google Scholar]
- 28.Mentz RJ, Cotter G, Cleland JG, et al. International differences in clinical characteristics, management, and outcomes in acute heart failure patients: better short-term outcomes in patients enrolled in eastern Europe and Russia in the PROTECT trial. Eur. J. Heart Fail. 2014;16(6):614–624. doi: 10.1002/ejhf.92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Pfeffer MA, Claggett B, Assmann SF, et al. Regional variation in patients and outcomes in the treatment of preserved cardiac function heart failure with an aldosterone antagonist (TOPCAT) trial. Circulation. 2015;131(1):34–42. doi: 10.1161/CIRCULATIONAHA.114.013255. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.