How can a drug that is associated with higher rates of both renal dysfunction and death than placebo and that costs 50 times as much as standard therapies and for which there are no meaningful data on relevant clinical end points be given to more than 600,000 patients and be promoted throughout the United States for serial outpatient use, an indication not listed on the label?1,2,3,4,5 The answer to this question can be discerned, at least in part, from a review of the clinical development and marketing of nesiritide: recombinant human brain natriuretic peptide.
On May, 17, 2005, the New York Times reported that tens of thousands of patients around the country are receiving nesiritide treatment once or more per week, over a period of several months, for what is described as an outpatient “tune-up.” This application of nesiritide — which costs approximately $500 per dose, as compared with less than $10 for nitroglycerin or nitroprusside — is the subject of an aggressive marketing campaign by the manufacturer, Scios, which is encouraging physicians to start their own “infusion centers” for whose services they can bill Medicare as if they were providing chemotherapy. But when nesiritide was approved by the Food and Drug Administration (FDA) in 2001, it was designated for the treatment of acute, decompensated congestive heart failure. Moreover, it had apparent safety problems and no proven clinical advantage over existing treatments in terms of the key end points of improved survival and prevention of subsequent hospitalizations.
In a 2000 report, Colucci and colleagues concluded that “nesiritide would be a valuable addition to the initial treatment of patients admitted to the hospital for decompensated congestive heart failure.”1 But their placebo-controlled, dose-ranging trial was focused on short-term monitoring of the pulmonary-capillary wedge pressure. Follow-up data on these subjects, which were not part of the study design as reported in that article, suggested that nesiritide may have had an adverse effect on 30-day mortality, which was 7.1 percent in the nesiritide group, as compared with 4.8 percent in the placebo group (P=0.62).2 Over this longer period of follow-up, the incidence of substantial deterioration of renal function was more than three times as high among patients treated with nesiritide as among those given placebo (P=0.04).3
The data from that trial were reviewed by an FDA advisory panel in January 1999. Despite the panel’s recommendation that the drug be approved for the reduction of pulmonary-capillary wedge pressure, the FDA decided that more data were required. A larger trial, Vasodilatation in the Management of Acute Congestive Heart Failure (VMAC), was conducted in 498 hospitalized patients who had dyspnea at rest.4 Nesiritide was compared with intravenous nitroglycerin or placebo. Since there were statistically significant improvements in both the reduction of pulmonary-capillary wedge pressure (a difference of 4 mm Hg) and the self-reported dyspnea rating with nesiritide as compared with placebo, nesiritide was considered to have met the efficacy criteria. The study did not demonstrate any benefit of nesiritide over nitroglycerin in terms of death or the need for repeated hospitalization within 30 days. At a subsequent meeting, in May 2001, the majority of the advisory panel recommended granting approval for nesiritide, and in August 2001, the FDA formally approved this drug for commercial use.
The VMAC trial raised a number of concerns about nesiritide. Only 30 percent of the patients received furosemide or intravenous diuretics before enrollment, although use of these agents is a standard approach to acute decompensated heart failure. The dose of nitroglycerin was not titrated aggressively, and because of the monitoring of the pulmonary-capillary wedge pressure, the “double-blind” assessment was compromised. The length of stay in the hospital was greater among patients who received nesiritide than among those given nitroglycerin (10.0 vs. 8.1 days, P=0.008).5 An elevation of more than 0.5 mg per deciliter (44.2 µmol per liter) in the serum creatinine level occurred in 27 percent of the patients in the nesiritide group, as compared with 21 percent of the controls (P=0.11).3,4 There was no increase in urine output with nesiritide, and subsequent studies have shown that this drug does not have a natriuretic or diuretic effect in patients with decompensated heart failure. Furthermore, the rate of death at 30 days was 8.6 percent in the nesiritide group, as compared with 5.5 percent among the controls (relative risk, 1.56; 95 percent confidence interval, 0.75 to 3.24; P=0.20).2 Indeed, the FDA approval went forward despite an internal reviewer’s critical point that VMAC did not rule out a 50 percent increase in the risk of death. A meta-analysis also suggests that the use of nesiritide is associated with an increased frequency of abnormal renal function.3
There have been two different analyses of the effects of nesiritide treatment on mortality, the most important end point in a randomized trial of an intervention for heart failure (see graphs). In one of them, Sackner-Bernstein et al. pooled data from the three trials involving patients whose baseline treatment regimen was not required to include inotropes and for which 30-day mortality data were available. According to this analysis, there was an 81 percent increase in the death rate with nesiritide as compared with placebo.2 In contrast, Scios analyzed seven trials that had 30-day mortality data, including trials involving open-label and outpatient use, but did not take into consideration the baseline treatment regimen. The company reported a 24 percent increase in mortality (P=0.33), and this figure was incorporated into a revised package insert in April 2005.
Mortality at 30 Days among Patients Treated with Nesiritide as Compared with Controls in Randomized Trials.
The relative risk of death with nesiritide as compared with placebo was 1.81 (95 percent confidence interval, 1.02 to 3.27; P=0.04) in the analysis by Sackner-Bernstein et al.3 (Panel A) and 1.26 (95 percent confidence interval, 0.80 to 2.00; P=0.33) in the analysis by Scios (Panel B).
Even in the face of such findings, however, the manufacturer has been actively promoting the use of nesiritide. It has set up a toll-free telephone hotline for “Natrecor Reimbursement Support” and has published a 46-page “Natrecor Reimbursement and Billing Guide.” The guide provides physicians with specific Medicare code numbers to be used in billing for a professional fee for nesiritide infusion ($172 for the first hour, $39 for each additional hour, and $408 for eight hours of observation), as one would for chemotherapy. The company justifies this billing practice by noting that the codes for chemotherapy administration include “substances such as monoclonal antibody agents and other biologic response modifiers.”
Notwithstanding the fact that only one small, open-label feasibility study has been conducted, outpatient nesiritide use has become widespread, fulfilling sales objectives for the manufacturer and bringing in revenue for physicians. The overall sales figure for nesiritide is projected to be $700 million for 2005, nearly double last year’s tally; it represents payment for more than 1.4 million treatments. Given that nearly 10 times as much drug is used for serial administration in outpatients as for the one-time use in hospitalized patients, much of this growth clearly stems from the off-label “tune-up” application.
The nesiritide story reflects some recurring themes: in other recent cases, too, major safety problems have been uncovered after a drug has been approved. Nesiritide was approved on the basis of a single trial in which surrogate end points were assessed three hours after administration. In cardiovascular medicine, we learned long ago that therapies directed at surrogate end points such as the suppression of premature ventricular contractions or, for inotropic agents, an improved ejection fraction can be associated with excess deaths. With the low threshold set for regulatory approval, the FDA did not demand appropriate warnings on the label regarding an increased risk of death or worsened renal function and did not require the performance of trials that would have provided definitive verification of the safety and efficacy of nesiritide.
We practice medicine in an era in which there is one pharmaceutical-company representative for every five physicians and in which companies will stretch the limits in their marketing of drugs. The boundary lines that previously separated industry from the FDA and academia have unfortunately become blurred. Interestingly, the European Agency for the Evaluation of Medicinal Products, the counterpart of the FDA, has still not approved nesiritide and awaits the results of a trial involving 1900 patients before it will even consider doing so.5
In my view, nesiritide has not yet met the minimal criteria for safety and efficacy. Until a trial definitively proves that this drug reduces the risk of death or repeated hospitalization for heart failure, there will be questions about the appropriateness of the drug’s use or even commercial availability. We need a tune-up of our procedures to eliminate indiscriminate use of drugs, such as nesiritide, when there is not proper evidence of their safety.
Dr. Topol is provost of Cleveland Clinic Lerner College of Medicine of Case Western Reserve University and chair of the Department of Cardiovascular Medicine at the Cleveland Clinic, Cleveland.
An interview with Dr. Topol can be heard at www.nejm.org.
Colucci WS, Elkayam U, Horton DP, et al. Intravenous nesiritide, a natriuretic peptide, in the treatment of decompensated congestive heart failure, N Engl J Med 2000;343:246-53.
Sackner-Bernstein JD, Kowalski M, Fox M, Aaronson K. Short-term risk of death after treatment with nesiritide for decompensated heart failure. JAMA 2005;293:1900-1905.
Sackner-Bernstein JD, Skopicki HA, Aaronson KD. Risk of worsening renal function with nesiritide in patients with acutely decompensated heart failure. Circulation 2005;111:1487-1491.
Publication Committee for VMAC Investigators.
Intravenous nesiritide vs nitroglycerin for treatment of decompensated congestive heart failure: a randomized controlled trial. JAMA 2002;287:1531-1540. [Erratum, JAMA 2002;288:577.]
Teerlink JR, Massie BM. Nesiritide and worsening of renal function: the emperor’s new clothes? Circulation 2005;111:1459-1461.