In their systematic review, Alexandre et al. seem to have missed or misinterpreted several important pieces of information.
Two of the three cell line studies they reviewed have significant methodological concerns and uncertainties in the results that make them far from convincing as supportive of anticancer effects.[2, 3] Both examined statin effects in serum-free cells: a poor model for oesophageal cancer in vivo. The reported reductions in other read-outs such a COX-2 or ICAM-1 expression without any data showing these changes have any functional significance considerably weaken any conclusions as to useful anticancer actions.
The study by Sadaria et al. also reported that simvastatin reduced proliferation by 59%, increased apoptosis by 19.8%, but only reduced viable cell numbers by 16.8%, and it is very difficult to explain those results biologically. Alexandre et al. suggest that anti-cancer effects of statins may be mediated via reduced ICAM-1 expression secondary to NF-kB inhibition, yet Sadaria et al. reported that statins reduced ICAM-1 expression without affecting NF-kB activity.
The authors also failed to reference a further study that is not only the one study which shows that statins inhibit proliferation and induce apoptosis in non-malignant Barrett's (as opposed to cancer) cells but also that statin effects (in this case rosuvastatin 0.1 mm) were produced at much lower concentrations.
Additional clinical studies, published before the submission date of Alexandre's paper, have not been included in the systematic review. One Barrett's cohort study showed a non-significant negative association between adenocarcinoma incidence and statin use, and a case–control study initially published in preliminary form, reporting significant negative associations when examining adenocarcinoma incidence compared with both Barrett's patients (relative risk 0.45) and gastroenterology outpatients (relative risk 0.52).