Dr Jonathan Sackner-Bernstein

The conflict began more than two years ago when Dr Jonathan Sackner-Bernstein (North Shore University Hospital, Manhasset, NY) presented preliminary analyses at conferences that he followed with publications in major journals [1,2] suggesting that nesiritide may promote or worsen renal insufficiency and perhaps increase 30-day mortality in patients with acute decompensated HF. As extensively covered by heartwire, his findings—which were based on a retrospective look at pooled data from several studies—alarmed some and failed to impress others.

Because his group's meta-analyses captured all available randomized trials that explored nesiritide therapy in patients with acute decompensated HF, Sackner-Bernstein said at the debate, in the absence of mortality trials, "It is the analysis best suited to investigate whether nesiritide is safe.

"That nesiritide is safe and effective would appear to be an obvious truth," he said, noting that many clinicians have seen how it can rapidly resolve dyspnea in many patients. "But we must acknowledge that we are at risk of being misled by our desire to help patients who have a terrible syndrome, who are at high risk and are suffering, by seeing what we want to see."

We must acknowledge that we are at risk of being misled by our desire to help patients who have a terrible syndrome.

He stated his thesis in a nutshell: "As nesiritide has not been shown to improve any meaningful clinical end point . . . and appears more likely to increase the risk of death rather than reduce the risk, it should not be used in clinical practice."

Sackner-Bernstein concluded by arguing that federal regulations require the FDA to mandate the drug's withdrawal: "The law is clear:  it does not require proof that nesiritide is dangerous. . . . The lack of proven safety is sufficient grounds for withdrawal."

30-day mean survival by treatment and nesiritide mortality HR in meta-analysis

Dr Clyde Yancy

When Sackner Bernstein made this case at the HFSA meeting, he was answered by Dr Clyde Yancy (University of Texas Southwestern Medical Center, Dallas). One of nesiritide's highest-profile investigators, Yancy presented a defense of nesiritide's value based on the existing prospective data and picked apart what he saw as flaws in the meta-analyses suggesting possible harm. The report that suggested a potential mortality risk, for example, "was a highly selected retrospective meta-analysis of trials that were not intended to show mortality."

Cardiologists, probably more than other physicians, Yancy said, "understand that it's extremely important to respect the primacy of prespecified end points and to be concerned and potentially worried about the corruption of data that go beyond those end points and get into small subgroup analyses." Accepting Sackner-Bernstein's conclusions, he said, "would violate our respect for the clinical-trial process.

"Let's be clear, exceedingly clear," Yancy said. "This debate exists because of the sensational nature of negative findings in today's risk-sensitive regulatory environment."

Is a naturally occurring peptide that promotes vasodilation, facilitates natriuresis, produces neurohormonal activation, and retards fibrosis in heart failure unsafe and ineffective?

Some of Yancy's other points:

• Compared with the control group, patients who received nesiritide showed significant improvements in the primary hemodynamic and symptom-related end points in the VMAC trial, the randomized study on which the drug's Food and Drug Administration approval was based. In that trial, nesiritide given at currently recommended dosage did not significantly elevate overall serum creatinine levels—and those whose levels increased did not have worse outcomes.
• In the Scios-sponsored ADHERE registry of patients who received nesiritide for acute decompensated HF, "we have shown a striking improvement in quality-of-care indicators, and we risk-adjusted the population so there is no apparent difference in mortality. Over the same time, there has been a decrease in in-hospital mortality," Yancy said.
• High-dose diuretics have been shown to increase the risk of all-cause mortality and sudden death in patients with acute decompensated HF and renal insufficiency, he noted. "Perhaps we need to worry about the negative influences of high-dose diuretic therapy, perhaps that's the culprit modality of risk."
• "Nesiritide is identical to human BNP, a naturally occurring peptide," he said. "Is a naturally occurring peptide that promotes vasodilation, facilitates natriuresis, produces neurohormonal activation, and retards fibrosis in heart failure unsafe and ineffective?"
• "We need more information from prospective, randomized mortality trials. This is the way we resolve questions, not by torturing data, but by going forward with new data acquisitions and understanding how best to use therapy."

Nesiritide was simply not better than nitroglycerin, which was used at quite a low dose.

Sackner-Bernstein responded both directly and indirectly to Yancy's arguments and continued to explore his rationale for taking nesiritide to task.

• "Safety refers to more than merely a favorable side-effect profile or whether a compound is naturally occurring. . . . For a drug to be used, data must show either that it's safe or that the impact on quality of life is substantial enough to warrant the risk." Nesiritide, he said, hasn't met that test.
• "When the FDA reviewed nesiritide in 2001, they concluded the data could not rule out a 50% increased risk of death with nesiritide when used in accordance with the current dosing recommendations," he observed.
• As for the ADHERE registry data, "We've all learned that there are problems with observational studies—including the valuable insight, reported in the literature, that their results are as likely to mislead or misinform as they are to tell us truth because of the inherent biases in them. . . . I do not prescribe hormone replacement therapy to my postmenopausal patients to reduce cardiovascular risk," he said. "Do you? Of course not. Because observational data can mislead us and in that case misled us for many years."
• "Whether an analysis of risk includes all the [nesiritide] studies in acute heart failure or is restricted to those studies using currently recommended dosing, it still shows an estimate of at least an 80% excess risk associated with the drug in either case."

Studies	HR (95% CI)	p
VMAC, PROACTION, NSGET, starting dose 0.01-0.02 g/kg/min	1.86 (1.02-3.41)	0.04
VMAC, PROACTION, starting dose 0.01 /kg/min	1.89 (0.98-3.63)	0.052

• In VMAC, fixed low-dose nesiritide was not shown to give consistent hemodynamic benefits compared with low-dose nitroglycerin, nor did it improve wedge pressure after the three-hour mark, Sackner-Bernstein said. The lack of difference in wedge pressure when it was measured between three and 24 hours is "particularly surprising given the lack of much up-titration of nitroglycerin over 24 hours. In fact, these data suggest the possibility exists that pharmacologic tolerance to nesiritide may develop."
• Also in that trial, dyspnea was significantly reduced at three hours in the nesiritide group, but there was no significant difference compared with nitroglycerin. In fact, results with nesiritide didn't differ significantly from those with nitroglycerin at any time during the trial's 24-hour blinded treatment phase, according to Sackner-Bernstein. "Nesiritide was simply not better than nitroglycerin, which was used at quite a low dose."
• If VMAC patients with worsening renal function still had good outcomes, it may be because "the nesiritide investigators were good doctors." Medical intervention for worsening renal function in the trial "was 229% more likely to be given in nesiritide-treated patients than in control patients. So the doctors did something about it, disrupted the natural history, and that's why the patients didn't have a worse outcome."