Quantitative analysis of MYD88 L265P mutations by digital PCR is an independent prognostic factor for CNS relapse as well as systemic relapse and poor outcome

Abstract

Introduction: Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease. While 30% to 55% patients achieve a durable cure, the remaining patients eventually die of the DLBCL, especially central nervous system (CNS) relapse is potentially lethal with no standard salvage treatments. We previously reported that absolute peripheral monocyte count (AMC) at diagnosis predicts CNS relapse. Recently gain-of-function mutations in MYD88 L265P have been detected in part of DLBCL, however, an influence of the mutation on survival is unclear. Digital PCR method was introduced to analyze molecular mutations quantitatively, however, analysis of MYD88 by digital PCR has not been reported yet. The purpose of this study was to clarify the value of quantitative analysis of MYD88 L265P mutations by using digital PCR on biological features and clinical outcomes. Methods: We retrospectively analyzed the data from a total of 134 patients (74 men/60 women; median age, 67 y), consecutively diagnosed with DLBCL according to the WHO classification who were treated at our single institution between 2005 August and 2014 April. We exclude the data if IgH rearrangement by Southern blot analysis was absent expecting at least 5% of lymphoma cells in the sample. MYD88 L265P mutation was examined by using digital PCR assay and also investigated the associated clinicopathologic factors. Results: MYD88 L265P mutations at least 5% (range from 7.5% to 89%) were found in 22 (16.4%) patients. MYD88 L265P mutations were seen more frequently in patients with a non-GC subtype, CD5 positive, high AMC (>0.5 x 109/L), extra-nodal involvement, B symptom, significantly (P < .05). All 21 patients with inconclusive positive digital PCR (range from 0.01% to 1.01%) underwent allele-specific nested PCR; only 9 (42.8%) of 21 patients had a positivity. The remaining 103 patients were negative. We, in turn, analyzed survival data per three groups: positive (>5%), weakly positive, and negative. With a median follow-up periods of 64 months for the survival patients, only positive (>5%) patients showed inferior CNS relapse-free survival at 5 years (53% versus 100%, 96%; P = .0001), progression-free survival at 5 years (39.4% versus 88.9%, 83.1%, P = .0002), and overall survival at 5 years (52.3% versus 88.9%, P = .001). MYD88 L265P positive (>5%) was an independent variable predicting CNS relapse in the multivariate analysis (hazard ratio 5.1; 95% confidence interval, 1.2-22.9, P = .02). Conclusions: This is the first report that MYD88 L265P mutation detected by digital PCR is an independent prognostic factor for CNS relapse. This suggests that MYD88 L265P mutation can be a second hit mutation of DLBCL and play a crucial role in disease progression, however, R-CHOP may overcome as far as mutated cells are within a small amount. Allele-specific PCR results can be over-diagnosis for the marker. The additional drug such as Bruton's tyrosine kinase inhibitor might be helpful.