Inconsistency associated with SOX11 immunohistochemistry in mantle cell lymphoma: a meta-analysis

SOX11 immunohistochemistry (IHC) in mantle cell lymphoma (MCL) is known to show varied results. Our aim was to evaluate the factor responsible for this variation among different studies. A meta-analysis was performed with the original data including the proportion and number of SOX11-positive MCL cases, host and clonality of SOX11 antibodies, clone or catalog number of antibodies, MCL subtypes, number of cases with indolent traits, number of aggressive variants, and cut-off for SOX11 IHC interpretation. A total of 21 published studies were analyzed. The combined proportion of SOX11-positive MCL cases was 0.80 (95% CI = [0.72, 0.87]), and substantial heterogeneity was observed (I2 = 83%). To explore sources of heterogeneity, subgroup analysis and meta-regression were done. Subgroup analysis with moderators of antibody clone or catalog number, antibody clonality, and monoclonal antibody clones showed substantial residual heterogeneity. Meta-regression with moderators of the proportion of cut-off value showed statistically significant result, although that with the aggressive cases did not. However, meta-regression with the cut-off value as a moderator showed substantial heterogeneity. The current meta-analysis of SOX11 immunohistochemistry in MCL showed the cut-off value to be important sources of overall heterogeneity.