Reply to “Genetic Determinants of Mycophenolate-Related Anemia and Leukopenia After Transplantation”

Le résumé fourni par la source

We thank you for inviting our reply to a letter written by Bouamar et al. related to our recent publication. Their analysis of the genetic substudy of the clinical trial comparing fixed-dose mycophenolate mofetil treatment with concentration-controlled treatment (FDCC) supported our findings of an association between the CYP2C8 single-nucleotide polymorphism (SNP) (rs11572076) and mycophenolate-related anemia. However, they did not detect the associations that we found with polymorphisms within the IL12A and HUS1 genes. This study points directly to several important limitations and challenges in validating genetic association studies. There are many examples in the literature in which genetic associations have not been replicated in an independent validation cohort, and in most cases, efforts to replicate lead to nonsignificant findings. These discrepancies occur for many reasons including lack of sufficient events (resulting in poor statistical power in the discovery or the validation cohort) and differences in phenotype, clinical populations, treatment regimens, minor allele frequencies, and statistical methods. We have published our experience in validating SNPs of kidney transplant recipients that were previously found by others to be associated with acute rejection (1). There are several substantial differences between our study and the FDCC genetic substudy that greatly decrease the chance of similar findings. Table 1 shows the differences between the two studies that may account for the lack of validation reported by Bouamar et al. The most obvious difference is the sample size, and as Bouamar et al. acknowledge, their smaller sample could lead to discrepancies in findings. Another important difference is the definition of the anemia event. In our study, we required that anemia (hemoglobin [Hgb]<10 g/dL or hematocrit<0.30) be attributed to mycophenolate therapy by the treating physician and that it be coupled with a mycophenolate dose change, discontinuation of therapy or initiation of erythropoietin therapy after the onset of low Hgb. The FDCC study defined anemia as any Hgb level of less than 11.3 g/dL beyond posttransplantation day 28. Therefore, the associations between SNPs and mycophenolate-related anemia may be obscured, due to these differences in definition. This difference most likely accounts for the higher proportion of anemia events in the FDCC study compared with our study.TABLE 1: Major differences between this study and the FDCC genetic substudyIn our analysis, we used Cox proportional hazards regression statistical models with time to the first mycophenolate-related anemia event. We also performed analyses unadjusted and adjusted for clinical factors. It is not clear in the FDCC substudy what type of statistical analysis was conducted or if the analysis was adjusted for other potential causes of anemia. Not conducting the same type of analysis or controlling for such factors may lead to failure of replication. Finally, it is possible that our findings are false-positives, and we concur that our phenotype requires validation in an independent population but with a similar definition of anemia and the same statistical analyses. Because our study was a discovery analysis, we used a false discovery of 20% to determine the significance of the SNPs, and by definition, 20% may be false-positives. Therefore, we have completed enrollment of a cohort from our consortium in that we will analyze for validation of our top SNPs.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.