Antihypertensive therapy and cancer risks?

The debate on the risk of cancer in hypertensive patients (1–9) as well as any relationship to antihypertensive drugs (9–17) has been ongoing for decades.

Firstly, the association between hypertension and increased cancer risk has been debated since Dyer and colleagues first published their paper in 1975. Several cohort studies have confirmed this finding. The most common cancers reported in hypertensive patients occur in the kidney and the endometrium, but associations with hypertension have also been reported for other cancers (9).

Secondly, the relationship between cancer and antihypertensive drug use has been extensively debated, and essentially all major antihypertensive drug classes have been associated with increased cancer risk (9,14). If an association with antihypertensive drug usage and a risk of cancer were to exist, it would most likely be observed after prolonged treatment. Over the years, the debate has been fueled by conflicting data from observational studies as well as randomized controlled trials of antihypertensive treatments with cardiovascular disease outcomes. Recently, a meta-analysis of randomized controlled trials suggested that angiotensin II type 1 receptor blockers (ARBs) were associated with a modestly increased risk of new cancer diagnosis (14), although the authors were not able to draw any conclusions about the exact risk of cancer associated with each of the ARBs studied. In contrast, a larger, more recent network meta-analysis concluded that taking antihypertensive drugs is not associated with an increase in medium-term cancer risk. In this analysis (12), the authors identified 70 trials with a total of 324,168 participants in 148 comparator groups, with a mean follow-up of 3.5 years (range 1–9 years). The authors concluded that the data exclude even a 5–10% increase in relative risk of cancer or cancer-related death with the most commonly used antihypertensive drug classes, including ARBs, angiotensin-converting enzyme inhibitors (ACEI) and calcium antagonists.

Selection of individuals for cancer cohort studies based on their prior cardiovascular disease status is likely to lead to a situation where cancer risk factors may be negatively confounded with other residual risk factors. If this risk confounding combines with effect modification between the primary and residual risk factors, as exists in cardiovascular disease and the risk of cancer, then the aggregate effect is a non-linear distortion of the risk factor relation. Importantly, such non-linear risk factor associations can arise spuriously from selection mechanisms that are common in prospective cohort studies (18). Because of these circumstances, cancer risk associations observed in cardiovascular prevention cohorts should be interpreted with caution. Only a randomized study of the impact of antihypertensive drugs on cancer as a primary outcome could provide definitive information on this question. For obvious reasons, such a study will never be performed.

Thus, the story of antihypertensive therapy and risk of cancer has been lengthy and convoluted. After years of debate, a consensus seemed to have emerged that such a risk did not exist in the treated hypertensive population. However, observations that cancer risk is increased in patients with treated hypertension have re-surfaced in the literature time after time. Given the distorting effects that may appear in antihypertensive trials that enroll patients with additional risk factors, as discussed above, these observations should be interpreted with caution. When analyzing the relation between a primary risk factor and the occurrence of a cancer disease event, one should implement careful adjustment for residual risk factors and allow for effect modification between the primary and residual risk factors. If such an adjustment is done, it may eliminate an apparent association of cancer risk. The confounding between primary and residual cancer risk factors in a trial may arise from selection mechanisms common in such studies, where patient inclusion is restricted based on prior occurrence of disease. Because of such confounding, cardiovascular outcome trials may produce illusive cancer associations.