Clinical Impact of TP53 Gene Mutations in Diffuse Large B-Cell Lymphoma (DLBCL): An International DLBCL Rituxan-CHOP Consortium Program Study

Abstract 967 Mutations of the TP53 tumor suppressor gene are associated with a poor clinical outcome in DLBCL patients treated with CHOP. The impact of TP53 mutations on clinical outcome of DLBCL patients treated with Rituxan-CHOP has not been comprehensively analyzed. The purpose of this study was to analyze the frequency and type of TP53 mutations in Rituxan-CHOP treated DLBCL patients from twenty-two medical centers, and to correlate these with clinical outcome. TP53 mutations were identified in 138/604 (22.7%) Rituxan-CHOP treated DLBCL cases and included missense (n=133), nonsense (n=16), splice site (n=9) and frameshift (n=1) mutations. The presence of any TP53 mutation correlated with poor overall survival (OS) with a median OS of 50 months in the TP53 mutation group versus 69 months in the wild-type group (wt-TP53, P=0.0042). Seventy-three of 138 cases (53%) had mutations in the DNA binding domains of the TP53 gene, which were found to be the most important predictor of poor OS (P=0.0044). In contrast, mutations in the non-DNA binding domains did not correlate with poor OS (P=0.157). Overexpression of p53 protein significantly correlated with only TP53 missense mutations (P=0.002), but not with other types of TP53 mutations, while TP53 deletion did not correlate with mutation or OS. In comparison to our previous series of patients treated only with CHOP, Rituxan-CHOP regimen improved OS in both wt-TP53 and TP53 mutated groups. The 5-year survival rate was 42% in patients with any TP53 mutation (median survival=50 months) and 41% in patients with the DNA-binding domain mutations (median survival=49 months) compared to 52% for those with wt-TP53 (median survival=69 months). The complete remission rate was 51% in patients with any TP53 mutation and 44% in patients with the DNA-binding domain mutations, compared to 77% for those with wt-TP53. However, the clinical outcome and treatment response to the Rituxan-CHOP varied in patients with mutations in different regions of the DNA-binding domains. Patients with mutations in the DNA minor binding groove motif (Loop L3, 17% of all mutations) had significantly decreased median OS (17 months) when compared to patients with Loop L2 (16% of all mutations) or loop-sheet-helix motifs (Loop L1-S10-H2, 20% of all mutations) with median OS of 49 and 50 months, respectively. In contrast to our previous CHOP series study, median survival was significantly improved for Rituxan-CHOP treated DLBCL patients with mutations in the loop-sheet-helix motifs (43 months). Multivariate analysis confirmed that TP53 mutations and activated B-cell-like (ABC)/germinal center B-cell-like (GCB) subtype classification were independent predictors of OS with a hazard ratio of 0.69 (GCB vs ABC, 95% CI 0.49-0.98) and 1.60 (TP53 vs wt-TP53, 95% CI 1.10-2.31), respectively. Similar to our previous CHOP study, the TP53 mutation profile, regardless of location, was found to stratify GCB-DLBCL, but not ABC-DLBCL, into molecularly distinct subsets with different clinical outcomes in Rituxan-CHOP treated DLBCL patients. This study demonstrates the importance of TP53 mutational profile for predicting clinical outcome. Elucidation of the roles of specific TP53 domain mutations, as documented in our study, will help in refining prognostic models for DLBCL patients treated with either the CHOP or Rituxan-CHOP regimen. These findings also provide the rationale and strategies for p53 targeted therapeutic intervention in DLBCL patients. Disclosures: Kahl: Milllennium: Consultancy, Research Funding; Cephalon: Consultancy, Research Funding.