Angiotensin receptor blocker therapy and risk of cancer

In their study published in this issue of the Journal of Hypertension, Rao et al.[1] examined the association between exposure to treatment with angiotensin receptor blockers (ARBs) and subsequent risk of lung cancer in individuals who received routine treatment in the US Department of Veteran Affairs Health System, which has maintained a robust database on which outcome questions such as this can be addressed. In this data record-linkage retrospective study using the Veteran Affairs electronic medical record system and other registries, the authors selected 1 299 902 patients as their analytic cohort, with 78 075 having received prescriptions for ARBs and 1 151 826 who did not. Following a rigorous statistical procedure in the selection of appropriate controls from those who did not receive ARB prescriptions, they found that the incidence of lung cancer was 0.44% in the treated and 0.57% in the nontreated patients, and concluded that there was no evidence to indicate an increased risk of lung cancer among new users of ARBs compared to nonusers [1]. Given the limitations of a study based on data linkage of the available administrative databases and the fact that a large adequately powered long-term clinical trial addressing this question will not be carried out, the authors have performed a remarkable analysis for the following reasons: by their choice of patients in the Veteran Affairs system, which provides long-term treatment for the large number of middle-aged and elderly patients who are at high risk of cancer because of their smoking habits and other social economic factors and thus were appropriate candidates for this study; the Veteran Affairs Healthcare System has a large robust medical records system that can be exploited readily; the authors had a sound strategy in using a large sample size of over a million patients and in using appropriate statistical methods to compare the users versus nonusers as controls, given that the data were nonrandomized and appropriate adjustments were essential in attempting to compare the two groups while reducing the effects of confounders and of chance. The study does raise a couple of concerns. The incidence of lung cancer in this cohort of largely male, high-risk, middle-aged patients appears surprisingly low when one compares this with US National Statistics. There might still be a possibility that in choosing the cohort of ARB users and nonusers, the authors have somehow chosen a lower-risk group. This question may not seem important in a study evaluating the causal relationship for ARB, but the results should be seen in the context of appropriateness of representative data. Also, since the time lag between exposure to a risk factor and manifest lung cancer usually is longer than the usual follow-up of 3–5 years, the relatively short duration in this study may not have adequately addressed the causal relationship, and may perhaps explain the low cancer incidence rates. These findings are, however, important since angiotensin-converting enzyme (ACE) inhibitors and ARBs reduce blood pressure and cardiovascular events, and are widely used in individuals with and without heart failure [2–6], and are also used by millions of people for hypertension. Increased risk of cancer would prevent the use of this class of effective and well tolerated therapeutic agents and deprive millions of people of their therapy. There had been conflicting data over their effects on cancer risk [7–11]. Angiotensin II stimulates neovascularization, which is required for tumor growth, and thus might act as a growth factor [12–14]. By blocking the effects of angiotensin II, ACE inhibitors and ARBs might thereby reduce the cancer risk [15–17]. However, whereas most of the information from studies of ACE inhibitors and ARB was neutral on cancer risk, some signals had been reported since the early days of renin–angiotensin system (RAS) modulation therapy, suggesting that although the data were inconclusive, the risk of associated cancer with ACE inhibitors, and ARB, could not be ruled out. A 2010 limited meta-analysis of selected ARB trials suggested a small increase in the risk of cancer, particularly of the lung, prostate and possibly breast [18]. Since then, several meta-analyses have reported that contrary to this earlier work, there was no evidence of an excess risk of cancers associated with long-term ARB therapy. The finding by Rao et al. in the current study that there is no excess risk of lung cancer is consistent with the results from later meta-analyses [19,20]. One might wonder at the reasons for the differences between the two sets of meta-analyses since by and large, they included many of the same large ARB and ACE inhibitor trials. There are several possible reasons. The first obvious reason is the limited number of studies included in the 2010 meta-analysis [18]. For example, whereas a later study included 15 trials on nearly 140 000 participants [20], this meta-analysis involved over 90 000 individuals in eight trials [18]. The analysis showing excess risk of lung cancer included data from only five trials. An important principle of meta-analyses is the inclusion of the totality of the data that can address the question, and not on a small number of selected trials. With such limited data in which information from other similar trials are not included, the findings cannot be regarded with any certainty. It is possible that the trials that published the results on cancer may have tended to report them if it indicated a worrisome trend. With increasing pressures to disclose information on adverse effects, it is possible that conducting an analysis based on published data may be misleading as trials with neutral results or a reduction in cancers are not disclosed. Thus, the cancer risk may unduly be over-emphasized, and conducting a meta-analysis solely based on published data on an adverse effect which is not uniformly reported may be biased. Indeed, it has been observed that in another comprehensive review and meta-analysis of the major ARB trials, a significantly excess cancer risk was seen in one trial, but significant reductions in risk were seen in other trials [20]. Since the totality of the data did not indicate an excess risk or benefit, one cannot reasonably put undue emphasis on the findings of some specific trials and claiming that these are unique and their results must be regarded as distinct from the others. It is also possible that with the relatively small number of events in each of the trials which are not specifically powered to show an a priori treatment effect, the findings could just be due to the play of chance. There is no known mechanism which might explain a possible excess of cancer risk from RAS blockade. The evidence from laboratory experiments and animal models suggests RAS blockade may have protective effects on cancer risk by suppressing tumor growth and in attenuating the carcinogenesis process in which angiogenesis is involved [12,15–17]. Angiotensin II stimulates neo-vascularization, which is a requirement for tumor growth, and it is itself also a growth factor, stimulating cell replication and increased expression of genes that control cell growth in tissue cultures in the absence of blood vessels [21,22]. ACE inhibitors decrease production of angiotensin II and ARBs block the effect of angiotensin II on the tissue receptors, and mechanisms of action of both classes of agents which inhibit the RAS have been shown to inhibit angiogensis and growth of induced cancer in rats [23,24]. In some human breast cancers in which the angiotensin II type 1 receptor was markedly increased due to gene over-expression, an ARB has been reported to reduce tumor size by 30% [25]. Alternative mechanisms of action may also influence the cancer risk and could perhaps explain the inconclusive findings on cancer risk in the many trials of RAS blockers. None of the above suggests a potential for an excess in cancer risks, but it is possible that other as yet unknown mechanisms may act to promote cancer development. The risk of developing cancer in individuals receiving long-term ARB therapy is of considerable clinical and public health importance. If the study by Rao et al. was to be judged on its own, its findings may be challenged due to the retrospective nonrandomized nature. However, the findings are consistent with those from meta-analyses that included data from the vast majority of major randomized clinical trials. Their findings have consistently found no evidence of an excess cancer risk. Ideally, these findings should be confirmed by a properly powered long-term randomized controlled clinical trial, but such a trial is unlikely for many reasons. In the meantime, individuals receiving ARBs can be reassured that, based on current evidence, the risk of developing cancer from long-term ARB use is minimal, if at all. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.