“Physicians have adopted nesiritide widely because they believe that it improves outcomes beyond the transient improvement in dyspnea that led to its approval,” said Barry Massie, M.D., chief of cardiology at the VA Medical Center, San Francisco.
The two articles by Dr. Sackner-Bernstein and associates “dispel the belief that nesiritide is beneficial with regard to renal function and survival. Given the potentially increased risk of mortality and worsening renal failure, it makes sense to use it only in patients who have not responded to usual therapies.” Dr. Massie receives consultant fees from GlaxoSmithKline and from Scios, the division of Johnson & Johnson responsible for nesiritide.
The Food and Drug Administration was also unconvinced by the metaanalyses done by Dr. Sackner-Bernstein.
“The data currently available are not materially different than when the cardio-renal advisory committee [of the FDA] discussed them prior to [nesiritide's] approval. In fact, the advisory committee was so unimpressed with the slightly adverse lean that they recommended not saying anything about mortality in the label,” an FDA spokeswoman said.
“The problem with metaanalyses is that the contribution each study makes is known when you are deciding what studies to combine. There is no way to compensate for the biases this introduced,” she said.
“Dr. Sackner-Bernstein and colleagues made a reasonable set of choices, but other reasonable alternatives are also possible. Most other combinations of studies will yield an estimate of risk lower than the one in their paper,” she noted.
That was exactly what Johnson & Johnson found when it ran its own metaanalysis of the seven controlled studies of nesiritide in which 30-day mortality was tallied. In this analysis, which was one of two that they added to the drug's labeling in April—the other analysis looked at 180-day mortality in four studies—the 30-day death rate was 5.3% in 1,059 patients treated with nesiritide and 4.3% among 658 patients treated with a comparator drug.
According to a rebuttal of the JAMA metaanalysis released by Johnson & Johnson in April, “the recently published metaanalysis is a selective reinterpretation of existing published data. If all available mortality data for all Natrecor [nesiritide] trials are taken together, Natrecor treatment does not significantly increase the risk of mortality, particularly after adjustment for baseline differences.”
But Dr. Sackner-Bernstein defended his group's more selective approach. “The metaanalysis we performed is more relevant. Our metaanalysis focused on the patients for whom the drug is intended, in studies that were randomized and double-blind and performed in a more rigorous fashion,” he told this newspaper.
Despite Johnson & Johnson's contention that the nesiritide database doesn't indicate a mortality problem, in April the company picked Eugene Braunwald, M.D., chairman of the Thrombosis in Myocardial Infarction (TIMI) study group at Brigham and Women's Hospital in Boston, to head an advisory panel that will review nesiritide's mortality data and advise the company on how to design additional mortality studies. Dr. Elkayam was named to head a similar panel that will review the effects of nesiritide on renal function.
Decompensation: A Changing Paradigm
The controversy about nesiritide's safety comes when some heart failure experts are crafting a new understanding of acute decompensated heart failure and how it is best managed.
“This is a complex syndrome, and patients can be very different,” Mihai Gheorghiade, M.D., said in an interview. “Treatment must be tailored to the clinical presentation.”
Recent observations from a pair of registries that together have more than 150,000 patients with acute decompensated heart failure have shown that these patients fall into three categories based on their systolic blood pressure. About 50% of patients are hypertensive (systolic pressure more than 140 mm Hg); about 48% are normotensive; and about 2% are hypotensive (systolic pressure less than 90 mm Hg).
Hypertensive patients should initially be treated with a vasodilator drug such as nesiritide or a nitrate, said Dr. Gheorghiade, associate chief of cardiology at Northwestern University in Chicago. Normotensive patients can initially receive either a diuretic or a vasodilator.
Another way to distinguish acute decompensated patients is by the clinical events that led to their acute state. One pathway is marked by chronic congestion that gradually builds over days or weeks to high pulmonary capillary wedge pressure and leg edema. The second pathway involves rapid-onset fluid overload triggered by high blood pressure (vascular failure). This may result from neurohormonal activation and increased sympathetic tone that raises blood pressure and redistributes blood from the systemic to the pulmonary circulation.
Patients with gradual-onset fluid overload tend to be normotensive or hypotensive, and it makes sense to initially treat them with a diuretic. Patients with rapid-onset fluid redistribution tend to be hypertensive, and the best initial therapy is with a vasodilator, Dr. Gheorghiade said. In patients with a rapid fluid shift, the goal is to quickly lower blood pressure without losing too much fluid.