Loop Diuretics in Patients With Heart Failure: Time to Change to Torsemide?

Retrospective analysis of the studies of left ventricular dysfunction, digitalis investigator group, and evaluation study of congestive heart failure and pulmonary artery catherization effectiveness trials along with experimental studies1 has suggested that the use of non-potassium (K+)-sparing loop diuretics in patients with heart failure (HF) and a reduced left ventricular ejection fraction (REF)2-4 is associated with an increase in cardiovascular mortality and morbidity. Non-K+-sparing loop diuretics remain, however, an important part of our current therapeutic approach to patients with acute and chronic HFREF and in those with HF and a normal left ventricular ejection fraction (NEF) because they are important in relieving pulmonary and peripheral congestion and thus the signs and symptoms of HF. Although non-K+-sparing loop diuretics tend to increase norepinephrine levels and activate the rennin-angiotensin-aldosterone-system, it has been assumed that their concomitant use with a beta adrenergic receptor blocker and an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker would, at least in part, negate the activation of the sympathetic nervous system, the rennin-angiotensin-aldosterone-system, and their adverse consequences. Of the available non-K+-sparing loop diuretics, furosemide is the most widely used diuretic in patients with HF in most parts of the world in part due to the long experience with its use and in part due to its relatively low cost. Although non-K+-sparing loop diuretics such as bumetamide and torsemide have been shown to have better bioavailability than furosemide and are often used in patients with acute HF and diuretic resistance, their use in patients with chronic HFREF is less frequent. It has been assumed that except for bioavailability, duration of action, toxicity, and cost that the available non-K+-sparing loop diuretics are similar. There have, however, been several reports pointing out that torsemide in contrast to other non-K+-sparing loop diuretics may be unique in that it inhibits the secretion of norepinephrine and aldosterone and prevents myocardial collagen formation.5-7 These reports have, however, had relatively little impact on the choice of non-K+-sparing loop diuretic use in patients with HF, likely due to the fact that the importance of aldosterone and myocardial fibrosis in the pathophysiology of HF and the relation of aldosterone levels to cardiovascular mortality in patients with HF were poorly understood until recently. Harada et al8 in this issue of the Journal have compared the effect of torsemide with azosemide, a long-acting non-K+-sparing diuretic shown in experimental studies to prolong survival in comparison to the shorter acting diuretic furosemide that increases plasma norepinephrine levels,9 in 15 patients with HF using a crossover design. They found that torsemide significantly reduced left ventricular wall stress associated with a reduction in plasma aldosterone levels, whereas not increasing norepinephrine levels in comparison to azosemide. Lopez et al10 in another recent, small, randomized parallel group study in 24 patients with HFREF comparing equipotent doses of torsemide with furosemide have shown that torsemide but not furosemide improves myocardial collagen cross-linking and left ventricular chamber stiffness associated with a reduction in lysl oxidase (Lox) expression, which was found to be directly related to the degree of myocardial collagen cross-linking. Previous studies have also shown a beneficial effect of torsemide in comparison to furosemide on myocardial collagen formation and ventricular remodeling.7,11 The findings of Harada et al8 and Lopez et al10 and the earlier studies showing a beneficial effect of torsemide on norepinephrine and aldosterone production, myocardial collagen formation, and ventricular remodeling in comparison to furosemide have important implications for the therapy of patients with HFREF and HFNEF. Recent evidence suggests that an increase in plasma aldosterone levels is associated with an increase in cardiovascular risk in patients with acute and chronic HFREF and in patients with HF post myocardial infarction (MI).12,13 Of particular interest is the recent study by Palmer et al14 in patients with MI in which patients with clinical evidence of HF were excluded. In this study, plasma levels of aldosterone in the upper tertile were associated with an increase in cardiovascular mortality, and the combination of an increase in plasma brain natriuretic peptide and aldosterone predicted an even greater cardiovascular risk than either alone. Of particular interest in the study by Palmer et al14 in regard to the study by Harada et al8 was the finding that the plasma levels of aldosterone in the upper tertile associated with an increase in cardiovascular mortality were within normal limits. Thus, one might speculate that the use of torsemide rather than furosemide or another non-K+-sparing loop diuretic in these patients would reduce both plasma brain natriuretic peptide and aldosterone levels sufficiently to negate their increased cardiovascular risk. This may be especially important because in contrast to torsemide, furosemide and other available non-K+-sparring loop diuretics increase plasma aldosterone levels. Over the past decade, the results of the randomized aldactone evaluation study15 and Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study16 studies of aldosterone blockade in patients with chronic severe HFREF and HF post-MI along with an increasing number of smaller clinical trials17 have pointed out the pathophysiologic importance of aldosterone and the benefits of aldosterone blockade in patients with HFREF. Although aldosterone blockade has been shown to be effective in reducing cardiovascular and total mortality in patients with severe chronic HFREF and HFREF post-MI, its use is limited by the retention of K+, especially in patients with chronic kidney disease and/or diabetes mellitus. The use of torsemide rather than furosemide or another non-K+-sparing loop diuretic could be expected, as outlined above, to reduce plasma aldosterone levels, in some instances, below a critical level, thereby reducing aldosterone's cardiovascular risk and possibly the need for an aldosterone-blocking agent such as spironolactone or eplerenone with their attendant risks. In some patients, however, the reduction in aldosterone levels associated with the use of torsemide may only be partially effective and mineralocorticoid receptor activation could result in progressive HF and/or sudden cardiac death. In these circumstances, the dose of aldosterone blocker might be considerably lower than it might have been if furosemide or another non-K+-sparing loop diuretic had been used, thereby reducing the risk of serious hyperkalemia. Similarly, the study by Lopez et al10 and previous studies demonstrating a beneficial effect of torsemide in comparison to furosemide on myocardial collagen formation and ventricular remodeling have important implications for patients with HFREF and HFNEF in whom an increase in myocardial collagen formation and stiffness is thought to be critical in the transition from hypertensive heart disease to HFNEF or diastolic HF.18 The cost of torsemide in comparison to furosemide has been an impediment to its use. However, torsemide is now generic and its cost could likely be considerably reduced if it had greater use. Although it would be inappropriate to suggest a change in clinical practice affecting millions of patients with HF based upon the small crossover study of 15 patients by Harada et al,8 it should be emphasized as pointed out above that previous small studies have also shown that torsemide is unique among the non-K+-sparing loop diuretics in inhibiting aldosterone, norepinephrine, myocardial collagen formation, and ventricular remodeling, and in 1 open label nonrandomized postmarketing surveillance study of 1377 patients (TORIC), it was shown to significantly reduce cardiovascular mortality in comparison to furosemide.19 Although the TORIC study was not randomized or prospectively powered to detect a difference in cardiovascular mortality between torsemide and furosemide, its results are nevertheless provocative and at very least warrant, the consideration of a larger scale adequately powered prospective randomized study in the light of the unique properties of torsemide outlined above to refute or confirm its findings.The development of new drugs and devices for patients with HF holds great promise for the future reduction of cardiovascular risk. It would however be tragic if we were to miss the potential opportunity of suppressing norepinephrine, aldosterone, myocardial collagen formation, and ventricular remodeling provided by the availability of torsemide today, thereby potentially improving the outcome of patients with HFREF and NEF with a resultant improvement of mortality, morbidity, quality of life, and last but not least health care costs.