Apart from these main findings, the preliminary evidence from a single study did MLN8237 not show any significant difference in the prevalence of MTHFR C677T mutation between patients with BCS or PVT and those with MVT, RVT and DVT. Certainly, due to a small number of patients, these conclusions should be interpreted with caution. The objective of subgroup analyses according to the regions where the studies were conducted was to explore

whether or not there is a racial difference in the etiology of BCS, non-cirrhotic PVT and cirrhotic PVT. Several findings were summarized as follows. First, the association between homozygous MTHFR C677T mutation or hyperhomocysteinemia and BCS was found in Asian studies, but not in European studies. Second, regardless of continents (Africa, Asia, Europe or South America), the subgroup analyses did not demonstrate any significant association of MTHFR C677T mutation with Kinase Inhibitor Library supplier non-cirrhotic PVT. Third, regardless of continents (Africa, Asia

or Europe), the subgroup analyses indicated an association of homozygous MTHFR C677T mutation with PVT in liver cirrhosis. Certainly, the subgroup analysis of Asian studies did not achieve any statistical significance. Fourth, regardless of continents (Africa or Europe), the subgroup analyses demonstrated a significant difference in the prevalence of hyperhomocysteinemia between liver cirrhosis with and without PVT. Taken together, the role of MTHFR C677T mutation or hyperhomocysteinemia in the development of

BCS may be different between Asian and European patients, which was similar to the results of our recent studies;[67] but their role in the development of cirrhotic or non-cirrhotic PVT was similar among studies from different continents. However, these findings should be cautiously interpreted due to a small number of studies included. Strengths of our study were as follows: (i) we performed an extensive published work search via four major databases without any publication language restriction; (ii) we separately evaluated the relationship between BCS or PVT and the presence of MTHFR PTK6 C677T mutation in different traits (total, homozygous and heterozygous); (iii) a majority of meta-analyses were lack of any significant heterogeneity or publication bias; (iv) in cases where the heterogeneity among studies was significant, we employed a random-effect model to calculate the OR value and sensitivity analysis to explore the source of heterogeneity; and (v) except for articles published in the abstract form, most of the articles were of relatively high quality. Our study had several limitations. First, only a relatively small number of studies were included in every meta-analysis. Therefore, the reliability of these conclusions should be confirmed in well-designed studies with a larger sample size.