González-Pérez A, Fernández-Vidaurre C, Rueda A, Rivero E, García Rodríguez LA

Cancer incidence in a general population of Asthma patients.

Pharmacoepidemiology and Drug Safety 2006;15:131-8.

ABSTRACT


Purpose: Several studies have assessed the association between asthma and cancer but none of them revealed a clear pattern of association of asthma with incidence of cancer. We aimed to examine the association between asthma, chronic obstructive pulmonary disease (COPD) and cancer.

Methods: We performed a cohort study with a nested case-control analysis using the General Practitioner Research Database in the UK. We defined three cohorts: patients with asthma, patients with COPD, and general population. During the follow up we identified a total of 5,263 incident cases of cancer. We conducted a nested case-control analysis that included all cancer cases as well as 20,000 frequency matched controls free of cancer frequency matched on age, sex, and calendar year.

Results: Patients with asthma did not have greater overall risk of cancer compared with the general population (Odds ratio=0.93, 95% Confidence interval:0.86-1.00). However, they presented an elevated risk of experiencing lung cancer (Odds ratio=1.84, 95%Confidence interval: 1.58-2.15). after controlling Controlling for smoking and other potential confounding factors yielded a much lower estimate (Odds ratio =1.35, 95% Confidence interval:1.15-1.59). This estimate contrasted with that observed for non smoking related cancer (0.87, 95%Confidence interval: 0.80-0.94)) Overall, respiratory drugs did not seem to be strongly associated with cancer among asthmatic patients. Patients with COPD had an Odds ratio of cancer of 1.26 (95% Confidence interval:1.12-1.43) compared with the general population.

Conclusion: Asthma does not seem to be associated with an increased risk of cancer. In fact the risk of non smoking related cancer appears to be somewhat reduced. However, we observed an elevated risk of lung cancer among asthmatic patients. Whether this result is a due to residual confounding and/or protopathic bias remains unclear. Further investigation is warranted to confirm or discard these associations.


Back to Publications List