SAN DIEGO—Patients with acute renal failure (ARF) often receive diuretics despite a lack of proof that they benefit from such therapy. Now, however, an observational study has shown that diuretics may actually cause harm in this setting.

The cohort study by Ravindra L. Mehta, MD, and colleagues linked diuretic use to an increased risk of death and long-term renal dysfunction in critically ill patients with ARF.[1] “In the absence of compelling contradictory data from a randomized blinded clinical trial, the widespread use of diuretics in [these patients] should be discouraged,” advised Dr. Mehta, Professor of Clinical Medicine in the Division of Nephrology at the University of California, San Diego.

In additional research challenging assumptions about treating patients with renal impairment, investigators found no justification for using a creatinine clearance of 30 mL/min or less to exclude such patients from low-molecular-weight heparin (LMWH) therapy.[2] That cutoff has traditionally been used to distinguish patients who could be at increased risk of accumulating heparin and, thus, of hemorrhaging.

After a thorough literature review, Canadian researchers found evidence suggesting that this cutoff may be too low to identify at-risk patients when some LMWHs are administered. Interestingly, though, there was also evidence that the cutoff may be overly cautious when at least one other LMWH is given.

EFFECTS OF DIURETICS ON ARF

Dr. Mehta and colleagues reported outcomes for 552 patients with ARF who required intensive care between October 1989 and September 1995. For patients with no known kidney disease, ARF was defined as a blood urea nitrogen (BUN) of 40 mg/dL or higher or as a serum creatinine of 2.0 mg/dL or higher. For the remaining patients, a sustained rise in serum creatinine of 1 mg/dL or more from baseline was used to designate ARF.

All patients were prospectively followed from the time of initial nephrology consultation through hospital discharge. They were categorized as having recovered their renal function if, at discharge, they no longer required dialysis and their serum creatinine was 2.0 mg/dL or less or no more than 20% greater than it had been at baseline.

Fifty-nine percent (326) of the patients were receiving diuretics at the start of the study. These patients were older and more likely to have acute respiratory failure, a lower BUN, a history of congestive heart failure, and a nephrotoxic rather than ischemic or multifactorial origin of ARF. However, initial APACHE II or III scores were not different between the patients receiving and those not given diuretics.

By the end of the study, 53% (294) of the patients had died, and 7% (17) of the 258 survivors were dependent on dialysis. After controlling the analysis for baseline differences between diuretic users and nonusers, the investigators found that diuretic therapy markedly increased—by 77%—the risk of death or nonrecovery of renal function. Diuretic use was also associated with a longer length of hospital stay and longer interval between the initial nephrology consultation and initiation of dialysis.

The type of diuretic administered did not appear to have any impact on risk, but the dose did. Patients given low diuretic doses experienced no significant change in their risk status; those given high doses faced a marked increase, which rose further with time. Dr. Mehta and colleagues explained, “The increase in risk was borne largely by patients who were relatively unresponsive to diuretics.” They offered several possible explanations for this finding, including a direct toxic effect of diuretics or indirect effects not necessarily related to renal function. Diuretic therapy may also have led to a delay in the recognition of ARF or underestimation of its severity.

The study is clinically important because it is still relatively common for critically ill oliguric patients to receive diuretics, said Lameire et al in an editorial.[3] “A trial of high-dose loop diuretics in an oliguric patient should only be attempted after careful correction of the volume status, should be limited in time, and, more important, should not postpone consultation with a nephrologist experienced in ARF,” they stressed.

LMWH IN RENAL IMPAIRMENT

For their literature review, Jeff Nagge, BScPhm, and colleagues from Toronto searched MEDLINE, EMBASE, and the International Pharmaceutical Abstracts for articles about pharmacokinetic differences of LMWH in patients with varying degrees of renal impairment. They also checked the reference lists of the articles they identified for additional citations. A study was included if it was a prospective trial of LMWH in nondialyzed patients with renal impairment. Of the 304 articles they located, only nine met the inclusion criteria. Four of these articles were later excluded after being found unsuitable for the review; thus, three single-dose pharmacokinetic trials and two multiple-dose trials involving patients with deep venous thrombosis were available for the final analysis. Two of these trials evaluated nadroparin; the other three looked at enoxaparin, tinzaparin, or Cy 222 (an investigational 2.5 kDa LMWH).

Three trials contained sufficient creatinine clearance data for assessment of the traditional 30-mL/min cutoff for LMWH therapy. One trial of nadroparin showed that accumulation of anti-Xa heparin activity occurred at a creatinine clearance below 50 mL/min. In this case, use of the 30-mL/min cutoff would have missed at-risk patients, and was therefore potentially dangerous, the authors stated. The Cy 222 trial, on the other hand, showed that only a cutoff as low as 10 mL/min was needed to identify at-risk patients, which suggests that the 30-mL/min cutoff is too cautious.

Interestingly, no correlation between declining creatinine clearance and accumulation of anti-Xa heparin activity was seen in the tinzaparin trial. Possible explanations for this result include potential differences in LMWH clearance in patients with acute glomerular disease and patients with age-related declines in the glomerular filtration rate (the patients in this study were elderly) and tinzaparin’s relatively high molecular weight, which may reduce dependence on renal clearance. “This observation with tinzaparin raises the possibility that this [LMWH] might be the favored preparation in patients with impaired renal function; however, additional studies … would have to be performed before any evidence-based recommendations could be made,” Nagge and colleagues suggested.

While accumulation of anti-Xa heparin activity was observed in four of the five studies, the findings of the fifth (tinzaparin) trial argue against a class effect of LMWHs in patients with renal dysfunction. The authors emphasize that while the 30-mL/min cutoff appears unjustified and that conclusive recommendations on LMWH therapy cannot be made on the basis of their review, there is evidence that the anticoagulant effect of LMWH can accumulate in patients with renal impairment. They suggest that the type of renal injury be considered in further investigations and that additional studies are urgently needed.

—Gale Jurasek

References
1. Mehta RL, Pascual MT, Soroko S, Chertow GM, for the PICKARD Study Group. Diuretics, mortality, and nonrecovery of renal function in acute renal failure. JAMA. 2002;288:2547-2553.
2. Nagge J, Crowther M, Hirsh J. Is impaired renal function a contraindication to the use of low-molecular-weight heparin? Arch Intern Med. 2002;162:2605-2609.
3. Lameire N, Vanholder R, Van Biesen W. Loop diuretics for patients with acute renal failure. Helpful or harmful? JAMA. 2002;288:2599-2601.

Return to table of contents