Nevertheless, we observe the same trends for 25(OH)D3 serum levels according to CYP2R1, GC, and DHCR7 genotypes as described previously, and HCV-infected patients with HCC had slightly lower 25(OH)D3 serum levels selleck chem Regorafenib compared to those without HCC. More important, we believe that analyses of associations between punctual 25(OH)D3 serum levels and endpoints such as HCC can be misleading, since 25(OH)D3 serum levels strongly fluctuate during seasons, with age, and as a consequence of numerous other conditions (liver fibrosis, diabetes, obesity, supplementation, etc.) [6], [7]. A second limitation of our study is the relatively weak level of statistical significance in the analyses of the individual cohorts, with the consequence that associations for the three loci with HCC became only fully significant in the combined analyses of all included patients.
Most likely, this can be explained by the relatively low frequency of the risk genotypes (especially of GC and DHCR7), as well as by the numerous variables with influence 25(OH)D3 serum levels in a given individual [9]. In this regard, it is important to note that ORs for the risk genotype of all genes (CYP2R1, GC, and DHCR7) were similarly directed in the discovery cohort as well as in all replication cohorts, confirming a true association between these loci and HCV-induced HCC. In addition, the strengths of the associations between these loci and HCV-induced HCC were largely comparable to the strength of the associations between these loci and 25(OH)D3 serum levels (strong effect of GC and DHCR7, moderate effect of CYP2R1, according to Wang et al.
) [9]. Nevertheless, the observed subtle differences between the different cohorts included in our study may not only be explained by the relatively small sample size of individual cohorts, but also by specific cohort features such as different recruitment strategies (e.g. cohort study versus case-control studies). In this regard, our study also cannot fully rule out whether all investigated genes (CYP2R1, GC, DHCR7) have the same impact on progression to HCV-related HCC at different stages of liver disease or in populations of different ancestries. Furthermore, our study cannot clearly characterize whether the observed findings apply for patients who did or did not respond to antiviral therapy.
Though we show in the SCCS (in a minority of the whole study population), that SNPs in GSK-3 CYP2R1, GC, and DHCR7 were not associated with treatment outcome, it remains unclear whether these genetic variations are associated with HCC in both individuals with or without treatment-induced eradication of HCV. The heterogeneity of the four independent cohorts included in our analyses might be perceived as another limitation. This applies for important cohort features such as race, recording of treatment modalities, HCC frequency, or different allele frequencies for some SNPs (especially rs1278578).