[Anticancer drug adherence].

A large number of anticancer drugs have been introduced during the two last decades with significant impact for survival, making cancer a chronic disease in a growing number of indications. However, these drugs are costly, induce adverse effects and their efficacy frequently depends on the dose. For all these reasons, adherence in cancer therapy is critical for an optimal benefit-risk ratio. Patient adherence remains virtually unexplored in many cancers, such as malignant blood diseases. When measured, adherence is poor, especially when the drug is administered as oral and prolonged therapy (hormonotherapy in breast cancer, imatinib). Physician nonadherence represents another form of drug misadministration; poorly documented, its mechanism remains obscure. Adherence may be measured by a panel of methods, each of them displaying limits and pitfalls, suggesting that several complementary methods should be used in the context of prospective studies. Risk factors are age, socio-educative profile, disease stage and physician profile. This review emphasizes some methods to prevent nonadherence. Finally, this review argues for prospective studies, which should integrate a social pharmacology approach, including medicine, psycho-sociology and economics.