Two observational studies published Nov. 28 in the Journal of the American Medical Association produced heart failure (HF) treatment efficacy results that differ markedly from the results found in clinical trials of the treatments. But James C. Fang, MD, of University Hospitals Case Medical Center in Cleveland, Ohio, the author of an accompanying editorial, told Cardiovascular Business that the seemingly contradictory results “are just different ways of looking at the same information.”
In one of the studies, Adrian F. Hernandez, MD, of Duke Clinical Research Institute in Durham, N.C., and colleagues studied whether newly initiated aldosterone antagonist therapy for heart failure was effective in older patients with HF and reduced ejection fraction. They noted that the EMPHASIS-HF Study Group (N Engl J Med 2011; 364:11-21) and the Randomized Aldactone Evaluation Study Investigators (N Engl J Med 1999; 341; 709-717) clinical trials found that aldosterone antagonist therapy, “reduced mortality by 24 percent to 30 percent and readmission for heart failure by nearly 40 percent.” The researchers sought to discover whether the treatment yielded similar results in a clinical setting.
Using data from the American Heart Association’s Get With the Guidelines-Heart Failure Registry and Medicare claims data, the researchers identified 5,887 patients 65 years or older who were discharged home with a documented history of HF from 246 hospitals in the U.S. between Jan. 1, 2005, and Dec. 31, 2009. The patients were eligible for aldosterone therapy if they met these criteria: left ventricular ejection fraction of 35 percent of less or qualitative description consistent with moderate to severe left ventricular systolic dysfunction; serum creatinine at admission of 2.5mg/dL or less for men and 2mg/dL or less for women; no documented contraindications; and no previous exposure to aldosterone. The primary outcomes were all-cause mortality, cardiovascular readmission and heart failure readmission at three years, and readmission for hyperkalemia at 30 days and one year.
Of the 5,887 patients eligible for aldosterone therapy, 1,070 patients received it. All-cause mortality was 49.9 percent in the group who received therapy vs. 51.2 percent in the patients who did not; cardiovascular readmission rates also were similar between the two groups (63.8 percent vs. 63.9 percent); however, the rate of readmission for heart failure was lower in the treated group (38.7 percent vs. 44.9 percent). The treated group demonstrated higher rates for readmission for hyperkalemia at 30 days (2.9 percent vs. 1.2 percent) and one year (8.9 percent vs. 6.3 percent).
Explaining their results, the authors pointed out that clinical trials enroll carefully chosen patients who do not reflect the population seen in practice, and employ protocols that may be difficult or impossible to implement in a clinical setting. In order to see the best results from aldosterone therapy, “We need to simulate the systems in place in the clinical trials with our patients,” Hernandez said in an interview with Cardiovascular Business. “These results show a need for pragmatic trials that enroll patients like those we see in clinical practice, and with systems and processes in place that we can replicate in practice.”
The second observational study followed patients from the Swedish Heart Failure Registry with HF with preserved ejection fraction (HFPEF) between May 11, 2000, and October 10, 2011, to determine whether treatment with renin-angiotensin system (RAS) antagonists was associated with lower all-cause mortality. Lars Lund, MD, of Karolinska Institutet in Stockholm, Sweden, and colleagues, noted that several randomized controlled trials (CHARM, Lancet 2003; 362; 777-781) (PEP-CHF, Eur Heart J 2006; 27:2338-2345) (I-PRESERVE, N Engl J Med 2008;359;2456-2467) had found RAS treatment ineffective in the HFPEF population.
The study enrolled 16,216 patients with ejection fractions of 40 percent or greater for the main analysis. Of these, 12,543 received RAS antagonists: Most (73 percent) received an angiotensin converting enzyme (ACE) inhibitor; 25 percent received an angiotensin receptor blocker; and 2 percent received both. One-year survival was 86 percent in the treated group and 69 percent in the population who did not receive RAS antagonists. At five years, there was still a significant difference—55 percent vs. 32 percent.
The researchers calculated a propensity score for each participant in order to adjust for possible selection bias, confounding and differences between the groups, then matched each treated patient with an untreated patient. The matched patients were within five years of age of each other and had propensity scores that differed by 0.1 or less. In this matched group, one-year survival for treated patients was 77 percent and for untreated patients, 72 percent. Five-year survival was 36 percent for treated patients and 34 percent for untreated patients.
The researchers analyzed propensity for dose-matched HFPEF and found one-year survival was 82 percent for high dose, 80 percent for low dose and 76 percent for no dose. At five years, survival was 43 percent for high dose, 39 percent for low dose and 40 percent for no dose.
Lund et al concluded that the use of RAS antagonists was associated with reduced all-cause mortality in an unselected population of patients with HFPEF. In explaining their results in light of the limited efficacy findings in randomized clinical trials, the authors postulated that the disparity may be due to selection bias and underpowering in the randomized clinical trials and/or the more representative patient population in the observational study. The findings suggested that “heart failure may be on a continuous ejection fraction spectrum, with greater benefit of RAS antagonists the lower the ejection fraction,” they wrote.
Both observational studies produced conclusions that are at odds with the results of multiple randomized clinical trials, and both sets of researchers pointed out the disparity and suggested that a study population that more closely reflects the population encountered in clinical practice may be one reason.
Editorialist Fang agreed: “Clinical trials involve ‘perfect’ doctors and ‘perfect’ nurses treating ‘perfect’ patients under ‘perfect’ conditions,” he said. “But the gold standard clinical trials are still the gold standard. …The art of medicine is trying to determine the role of these evidence-based therapies on the patient in front of you.”