Angiotensin receptor blockers and cancer – Relationship dismissed by VALUE data while waiting for EMA and FDA reports

As editors of Blood Pressure, a hypertension research journal, we feel strongly obligated to point out a fundamental shortcoming in a recently published meta-analysis of cancer rates in patients treated with angiotensin receptor blockers (ARBs).

The online publication of the June 14 issue of Lancet Oncology contains a paper by Sipahi et al. entitled “Angiotensin receptor blockade (ARB) and risk of cancer; meta-analysis of randomized controlled trials” (1). A major conclusion of the report is based on Figure 4B, labeled “Cancer in patients without a background of treatment with angiotensin converting enzyme inhibitors”, which contains these data on cases of incident cancer as a proportion of total patient numbers: ARB, n = 1360/16,497 and control, n = 1262/16,527, representing a difference between the two groups of 98 patients with cancer. Relying on this narrow difference, the authors made a claim, supported by an uncritical editorial (2) in the same issue of the Journal, of a possible link between ARB use and cancer.

The VALUE Trial (3), which compared outcomes with an ARB and a calcium-channel blocker in hypertensive patients at high cardiovascular risk, is quoted by the report of Sipahi et al. (1), but it is erroneously stated that VALUE did not collect cancer information and therefore was not included in the meta-analysis. In reality, cancer data were collected and included in VALUE's formal database, completed in March 2004. Every case of cancer was described with a specific diagnosis. In VALUE, there were a total of 510/7649 cancer cases reported with the ARB valsartan (6.7%) vs 591/7596 cases reported with the calcium-channel blocker amlodipine (7.8%). In that large clinical trial, with 15,245 patients followed for several years, this finding obviously argues against an excess cancer risk associated with ARB therapy.

Of critical importance, if the VALUE data had been included appropriately in the meta-analysis of Sipahi et al. (1), the cancer event numbers would have become 1870/24,146 (7.7%) for ARBs vs 1853/24,123 (7.7%) for others, and clearly would have shown no difference whatever in cancer rates between ARB and control therapies (4).

With respect to risk of cancer, current antihypertensive drug regimens appear to be safe, and there is no reason to alter current guidelines. There is no convincing data that any of the established antihypertensive drugs act as carcinogens. However, the prevalence of cancer in the hypertensive population is increased, not as a function of detection bias, but most probably related to as yet undefined pathophysiological processes. Additional experimental and clinical research to identify and characterize relevant mechanisms and causes of excess neoplastic growth in the hypertensive population is of considerable interest and clinical importance.

This was the conclusion of an editorial in Blood Pressure in 1997 (5) that remains unchanged in 2010. We take it for granted that the extensive further investigations by the European Medicines Agency (EMA) and the Food and Drug Administration (FDA) in the USA will soon come to the same conclusions when they finish their announced investigations that include all available and relevant data.