New Evidence Against Heart Failure and Cancer Link

Patrice Wendling

April 09, 2018

A new analysis of long-term follow-up data from two physician cohorts found no association between heart failure (HF) and incident cancer, despite previous findings suggesting HF is associated with higher cancer risk.

"We wondered whether the association was possibly confounded, for instance, by clinical comorbidities or surveillance bias. Heart failure patients tend to 'bump up' against the healthcare system more, and some cancers may be detected in that fashion," lead author, Senthil Selvaraj, MD, MA, Hospital of the University of Pennsylvania, Philadelphia, told | Medscape Cardiology in an email.

"Therefore, we searched for a database that was well equipped to answer the question and had more pertinent cancer data: risk factors, adjudicated outcomes, longer length of follow-up."

The investigators analyzed data from 28,341 male physicians free of either disease at enrollment in the Physicians' Health Studies I and II, two randomized trials of aspirin and vitamin supplements conducted from 1982 to 1995 and 1997 to 2011, respectively.

Over a median follow-up of 19.9 years, 1420 physicians developed HF and 7363 cancers were diagnosed. A median of 3.4 years passed between HF diagnosis and cancer diagnosis in the 177 physicians with both diseases, according to the study, published online ahead of the April 10 issue of the Journal of the American College of Cardiology.

Modeling HF as a time-varying exposure, the results showed no difference in the incidence of cancer among physicians with or without HF in unadjusted analysis (hazard ratio [HR], 0.92; 95% CI, 0.80 - 1.08).

Moreover, the relationship remained nonsignificant after adjustment for enrollment group, race, cigarette smoking, alcohol or aspirin use, family history of cancer, cirrhosis, proton-pump inhibitor or histamine-2 blocker use, and sun exposure (HR, 1.02; 95% CI, 0.84 - 1.25) and after further adjustment for any colonoscopy or sigmoidoscopy, physical exam, rectal exam, and prostate-specific antigen (PSA) testing (HR, 1.05; 95% CI, 0.86 - 1.29).

In an exploratory analysis of site-specific cancers, HF was significantly associated only with prostate cancer (HR, 0.65; 95% CI, 0.50 - 0.84). The relationship, however, weakened after adjustment for enrollment group, race, body mass index, and family history of prostate cancer (HR, 0.74; 95% CI, 0.56 - 0.99) and disappeared after further adjustment for any rectal exam or PSA level tested (HR, 0.78; 95% CI, 0.59 - 1.05).

A secondary analysis of physicians at significant landmark ages also revealed no association between HF and cancer risk.

"I think our results support current practice standards for cancer screening, insofar as HF patients should be screened similarly to the general population," Selvaraj said. "Of course, cancer screening should take into account life expectancy, which may be an important consideration in some advanced HF patients."

"Therefore, cancer screening should be made according to professional society guidelines and involves discussion and evaluation of the individual patient," he added. "What we do not show, conversely, is any increased need for cancer screening among HF patients."

Enough Already?

The results, however, conflict with prior work, including a recent prospective study in 1081 persons followed for nearly 5 years, in which the adjusted risk for cancer, excluding nonmelanoma skin cancer, was nearly twice as high in patients who developed HF after a myocardial infarction as in those who did not.

Senior author of that prospective study, Veronique Roger, MD, Mayo Clinic, Rochester, Minnesota, told | Medscape Cardiology the new report does not put the question of HF and cancer to rest.

"I don't think we have, in part, because larger is not always better and because what you gain in sample size, so to speak, you lose in precision," she said. Also, "it's only applicable to males, by definition."

"While we have to acknowledge that the case is not closed, there are directions for the scientific community and for the lay public," Roger said.

That means ensuring that studies are set up to look at broader sets of outcomes and that shared risk factors between HF and cancer are addressed.

"That's where the action item is for us as people close to healthcare and for the lay public," Roger said. "While we're waiting for more definitive evidence, what we really want to do is prevent both diseases. We've known for years that physical activity, a reasonable diet, controlling blood pressure are exceedingly important for health in general, cardiovascular health certainly, and some forms of cancers."

In a related editorial, Paola Boffetta, MD, MPH, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York City, and Jyoti Malhotra, MD, MPH, Rutgers Cancer Institute of New Jersey, New Brunswick, who also previously commented on the study by Roger et al, write that this is a "new, more definitive study, which provides strong evidence against the hypothesis of a link between HF and cancer."

Strengths include longer follow-up than previous investigations, consistency in a number of validation analyses, and high-quality data on the underlying comorbidities and potential confounders.

"It would be premature, however, to conclude definitively that the hypothesis of an association between HF and cancer should be abandoned," they write, noting the all-male population and that HF was determined primarily through self-report.

Boffetta and Malhotra call for additional, larger studies to look at the question, and for cancer to be included as one of the outcomes in HF trials.

Asked whether the evidence to date, including their results, support abandoning the cancer hypothesis in HF, Selvaraj replied, "I don't think we can be absolutely definitive. I think our results are reassuring but, as with any study, it has its own limitations."

"While the lack of women in the study is a limitation, I don't think there is clear biological rationale for difference in effect by gender, at least with current data," he said. "In addition, previous studies did not demonstrate a significant difference for cancer risk by gender."

Selvaraj noted that physicians are a unique population and tend to be healthier than those in epidemiologic or population-based studies; however, this may be a strength of the study because it reduces residual confounding, for example, by alcohol or smoking habits.

Going forward, Selvaraj said he's eager for the release of the HF endpoint of the landmark CANTOS trial to see whether an anti-inflammatory drug — already shown to lower cancer risk, particularly lung cancer risk — would also decrease HF risk.

"In addition, while inflammation may be important in both the development of HF and cancer, I believe research investigating specific immune mechanisms would be interesting," he said. "Our results suggest, given the lack of relationship between HF and cancer, that immune mechanisms may be distinct."

The study was partially supported by a grant from the National Heart, Lung, and Blood Institute. Selvaraj and Roger report no relevant financial relationships. Boffetta reports no relevant financial relationships. Malhotra has served on the advisory boards of Pfizer and AstraZeneca.

J Am Coll Cardiol. 2018;71:1501-1510, 1511-1512. Abstract, Editorial

Follow Patrice Wendling on Twitter: @pwendl. For more from | Medscape Cardiology, follow us on Twitter and Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.