Stroke Risk Spikes After Cancer Dx

Cancer may be an under-appreciated, but common risk factor for stroke, which is most common the first month after cancer diagnosis.

A new diagnosis of cancer could signal elevated stroke risk, a nationally representative observational study affirmed.

The cumulative incidence of stroke was 25% to 7.43-fold higher within the first month after diagnosis across common cancer types compared with matched controls, Babak B. Navi, MD, of Weill Cornell Medical College and Memorial Sloan Kettering Cancer Center, both in New York City, and colleagues found.

The risk of ischemic or hemorrhagic stroke remained significantly elevated for all but prostate cancer through month three, the group reported online in the Annals of Neurology.

“Our findings suggest that cancer may be an underappreciated yet common risk factor for stroke,” they wrote in the analysis of the Surveillance Epidemiology and End Results (SEER)-Medicare linked database.

That conclusion agreed with an earlier population-based study from Sweden, which found 2-fold elevated stroke risk in the first 6 months after any cancer diagnosis. The risk dropped off rapidly after that point, as in the SEER data.

However, that study showed a persistent albeit small 20% elevated risk compared with the rest of the general population out to at least 10 years, whereas in Navi’s study the risks “were generally no longer present beyond 1 year” across cancer types.

Aside from the negative consequences of the event itself, “stroke is particularly relevant to cancer patients because strokes often preclude or delay cancer treatments, resulting in reduced survival,” Navi noted in an statement.

He recommended that both physicians and patients be vigilant for stroke symptoms in the early period after diagnosis.

These observational studies couldn’t determine exactly how cancer may be associated with stroke, although the hypercoagulability brought on by malignancy and its treatment is a prime suspect.

How to prevent stroke in cancer patients and whether cancer should be included in cardiovascular risk prediction instruments are questions that would require further research, Navi’s group noted.

However, two aspects of their SEER findings did support that the association is real.

“The heightened risk of stroke was greatest in the first 3 months after cancer diagnosis and then slowly decreased over time,” Navi’s group noted, “supporting the biological plausibility of an association between cancer and stroke given that cancer-mediated hypercoagulability and the effects of cancer therapy are generally greatest during this time period.”

Also, stroke risk was highest with cancer types that are typically more aggressive and more often presented with advanced stage disease. Three-month cumulative incidence rates versus the respective matched controls were (all P < .001):

  • For lung cancer, 5.1% versus 1.2%
  • For pancreatic cancer, 3.4% versus 1.3%
  • For colorectal cancer, 3.3% versus 1.3%

Stroke risk was lower but likewise significant among patients with breast cancer, which was often localized (1.5% versus 1.1%, P < .001).

The association wasn’t significant overall for prostate cancer patients (1.2% versus 1.1% in controls, P = .085), who also commonly had localized tumors.

“This finding further supports the biological plausibility of an association between cancer and stroke because cancer-mediated hypercoagulability and thrombotic risk is closely linked to cancer activity,” Navi and colleagues wrote.

The study included 327,389 pairs of patients with a new primary diagnosis of breast, colorectal, lung, pancreatic, or prostate cancer in the SEER database from 2001 through 2007 and controls individually matched by age, sex, race, registry, and medical comorbidities from among Medicare enrollees without cancer. Each pair was followed through 2009.

Ischemic strokes accounted for 89% to 93% of the strokes, as expected from the general population.

Limitations included reliance on administrative claims data, potential for residual confounding, and exclusion of people who were younger than 66, not enrolled in Medicare, or enrolled in a Medicare managed care plan.

Detection bias from greater monitoring among cancer patients was unlikely “because stroke risk was not uniform among cancer types and appeared to instead correlate with the general aggressiveness of the underlying cancer,” the researchers noted.

From the American Heart Association:

The study was supported by the NIH and the Florence Gould Endowment for Discovery in Stroke.

The researchers disclosed no relevant relationships with industry.


This article was first published on MedPage Today and reprinted with permission. (Free registration is required.)