“…Consequently, various studies aimed at replicating this pharmacogenetic association with mixed results. A considerable body of research corroborated the finding by Uhr et al [] [Gex‐Fabry et al, ; Kato et al, ; Nikisch et al, ; Sarginson et al, ; Lin et al, ; Singh et al, ; Huang et al, ], but there were also a number of studies that failed to find a predictive value of ABCB1 polymorphisms on therapy outcome [Peters et al, ; Mihaljevic Peles et al, ; Menu et al, ; Perlis et al, ; Uher et al, ; GENDEP Investigators, MARS Investigators, and SI, ]. The possible reasons for these contradictory results are diverse and coincide with a generally observed difficulty in the replication of pharmacogenetic associations caused by, for example, heterogeneity of study designs and patient samples, variability in phenotype definitions (MD being a particularly heterogeneous phenotype), inadequate statistical power, environmental factors, and polymorphic heterogeneity, making it difficult to compare results across studies.…”