Genomic Characteristics Clinical Phenotype and Disease Course of TP53 Mutated Diffuse Large B Cell Lymphoma at Diagnosis and Relapse

INTRODUCTION

Next-generation sequencing of diffuse large B cell lymphoma (DLBCL) has revealed a vast landscape of genetic alterations. Unfortunately, very few of these carry a significant prognostic power or confer a different response to therapy. Unique, in this respect, are mutations in TP53, which have been implicated in over a third of the refractory cases to frontline regimes. This study aimed to evaluate the association between clinical phenotype, disease course and specific genomic alterations of TP53mut DLBCL, comparing de novo to relapsed or refractory disease (R/R), and chemo-sensitive to chemo-resistant cases.

METHODS

We performed a retrospective analysis of patients with DLBCL, treated at Memorial Sloan Kettering Cancer Center, harboring a TP53mut. Formalin fixed, paraffin embedded samples were sequenced using a custom DNA capture panel. Germline variants were filtered against the 1000 Genomes Project. Significant non-synonymous variants were identified by mapping to the COSMIC database, as well as clear inactivating mutations in established tumor suppressor genes. Analysis for TP53 mutations was performed per their affected p53 protein domains, including DNA binding motifs (Loop-L2, Loop-L3, Loop-Sheet-Helix), evolutionary conserved regions, non-conserved and truncating non-missense mutations.¹Sequencing was performed at various time-point in the course of the disease. We evaluated TP53mut allele frequency (AF), overall number of TP53mut /patient, and concurrent alterations in other genes. Disease characteristics and clinical course were extracted for first line treatment of DLBCL, as well as from R/R cases. Cell of origin (COO) determination and MYC, BCL2 and BCL6 rearrangements at the time of sequencing were recorded. We evaluated response to treatment and survival in the subset of patients sequenced at diagnosis. Transformed indolent lymphoma (t-DLBCL) cases were evaluated in respects to the time of transformation, but considered R/R if previously treated with chemotherapy.

RESULTS

Of 360 sequenced DLBCL cases, 84 were found to carry a TP53 mutation (35 de novo ; 49 R/R). Of these, 61% had a stage IV disease at presentation, involving extra-nodal sites (mostly musculoskeletal, lung, gastrointestinal and hepatic). Central nervous system involvement was rare throughout the course of the disease (2%). t-DLBCL represented a considerable part of the TP53mut population, more so in R/R disease (59% vs. 29%, p=0.01), where it originated more commonly from follicular lymphoma (76% vs. 50%, p=0.2). R/R was also enriched for Double-Hit disease (41% vs. 20%, p=0.06) and for MYC copy number alterations (59% vs. 20%, p=0.02), though the number of evaluable cases was small.

We identified 105 TP53 mutations, with most patients (86%) harboring only one. Overall, 49% of patients had a mutation involving a DNA binding motif, 26% a non-missense truncating mutations, and 35% a mutation affecting non- DNA binding motif regions (figure 1). R/R disease was associated with a higher AF (0.52 vs. 0.35, p=0.01), but not with a higher overall number of TP53mut or the affected p53 domain. Patients had a median of 4 concurrent mutations in other genes without a clear association with R/R .

Among the R/R patients, 65% were refractory to their first-line chemotherapy, with a median of 4 therapy lines overall. Further, of the 23 patients receiving autologous stem cell transplantation, 52% were refractory. Taken together, these observations suggest that TP53mut may be an early event conferring a high rate of multi-drug resistance. Refractoriness to frontline therapy was also observed in 49% of the de novo case. However, with a median follow-up of 30 months, 2y PFS was nearly 50%, demonstrating a subset of TP53mut patients who fare well (figure 2). There was no association between PFS and the site of TP53mut, total number of TP53 mutations or number and type of concurrent mutations.

CONCLUSIONS

TP53mut is associated with multi-drug resistance and a worse prognosis. However, nearly 50% of newly diagnosed patients can be expected to respond to frontline regimens, many of whom achieving long term remissions. We could not find a clear genomic feature to differentiate Responsive from Refractory patients. Future research should focus on prospectively examining the role of TP53mut in predicting treatment outcomes.

View largeDownload slide

View largeDownload slide

Close modal