Outcomes for Double Hit Lymphoma Patients Identified Via Routine Vs Selective Testing for MYC Rearrangement

Introduction

Patients (pts) diagnosed with double hit lymphoma (DHL) are reported to experience poor survival outcomes following first-line treatment with R-CHOP, although survival may be improved with receipt of intensive front-line immunochemotherapy. These results may be altered by selection bias, as prior retrospective series of DHL pts did not report performing routine cytogenetic testing to identify those with DHL. We hypothesized that outcomes might differ for DHL pts identified after the implementation of routine testing at our institution.

Methods

In January of 2015, our institution began routine testing for MYC-rearrangement (R) by fluorescence in situ hybridization (FISH) for all newly-diagnosed cases of DLBCL and HGBL/BCLU, with reflex testing for BCL2-R and BCL6-R by FISH if positive for MYC-R. Prior to this, FISH for MYC-R, BCL2-R and/or BCL6-R was performed at the discretion of the treating physician or interpreting hematopathologist. Pts included in our analysis were diagnosed with DHL from May 2009 through July 2018. Inclusion criteria were adults treated with first-line curative-intent immunochemotherapy. Exclusion criteria were HIV-associated lymphoma, primary CNS or testicular lymphoma, post-transplant lymphoproliferative disorder, and incomplete clinicopathologic and outcomes data. Primary outcome measures were progression free survival (PFS) and overall survival (OS) in the pre-January 2015 selective testing (ST) and post-January 2015 routine testing (RT) cohorts. Data were censored on 5/31/19.

Results

Sixty-five pts were included in the analysis. The ST and RT cohorts did not differ significantly on the baseline characteristics of age, sex, stage, performance status, B symptoms, elevated LDH, extranodal disease, stage III/IV, blood or bone marrow involvement, International Prognostic Index score, Ki67, cell of origin, expression of MYC or BCL2 protein or the combination, time from diagnosis to treatment >14 days, disease bulk > 7.5 cm, transformation from an indolent lymphoma, or front-line treatment with R-CHOP vs intensive immunochemotherapy. Eighty percent (n=52) of pts received intensive front-line immunochemotherapy, most with R-EPOCH (n=42).

Median length of follow-up was 32 months (61 months for ST and 24 months for RT). The RT group had a longer 2-year PFS 70% vs 43%, p=0.02; univariate hazard ratio (HR) for progression 2.4, 95% confidence interval (CI) 1.1-5.4, p=0.03, (Figure 1A). The only other variable associated with progression was presence of extranodal disease (HR 2.35, 95% CI 1.0-5.4, p=0.047). In a multivariate model, both ST and the presence of extra-nodal disease were independently associated with progression at 2 years (HR 2.84, 95% CI 1.3-6.4, p=0.012 and HR 2.79, 95% CI 1.2-6.5, p=0.018, respectively). Longer OS was observed with RT (median OS NR vs 16 months, 2-year OS 72% vs 41%, p=0.008; HR 2.97 95% CI 1.3-6.9, p=0.012, Figure 1B). No other factors were associated with longer OS. For pts who relapsed after first-line therapy, survival was poor with a median OS after relapse of 6 and 5 months in the RT and ST groups (p=0.11). When limiting the analysis to only those pts treated with intensive immunochemotherapy, median PFS and OS remained significantly longer for the RT group (2-year PFS 70% vs 38%, p=0.01; median OS NR vs 16 months and 2-year OS 74% vs 38%, p=0.007). Post-relapse survival was similarly poor for those treated with intensive induction regardless of RT vs ST (median post-relapse OS 6 vs 5 months, p=0.1).

Conclusions

Long-term survival may be improved for DHL pts identified through routine as opposed to selective FISH testing. These data could serve as a new baseline for outcomes of DHL pts identified in this manner. Survival outcomes for our cohort of pts treated with intensive immunochemotherapy are similar to those for DHL pts reported in a recent prospective multicenter study of front-line R-EPOCH (Lancet Haematol. 2018 Dec;5(12):e609-e617). While we suspected that the poorer outcomes experienced by ST pts may have been due to high-risk clinicopathologic factors that contributed to selection bias in testing, this does not seem to be the case for any measured factors. It may be that variation in additional molecular features within this high-risk cytogenetic population could explain differences in survival, and next generation sequencing analysis of DHL patient tissue specimens is planned.

View largeDownload slide

View largeDownload slide

Close modal