Predicted Functional Consequences of TP53 Alterations and Prognosis in TP53-Mutated Mantle Cell Lymphoma

Abstract

Background: Mantle cell lymphoma (MCL) cases with TP53 mutations ( TP53mut) are associated with poor responses to chemoimmunotherapy (Eskelund Blood 2017). Across myeloid and solid tumors, predicted functional consequences and pathogenicity of specific TP53 mutations have been linked to differential patient outcomes (Pollock Breast Cancer Res Treat 2022, Yabe Cancers 2021). We aimed to characterize the impact of these features and associated molecular alterations on treatment outcome in TP53mut MCL patients treated with standard frontline regimens and on the ongoing phase 2 BOVEN trial. The BOVEN trial aims to assess the efficacy of frontline treatment with zanubrutinib, obinutuzumab, and venetoclax in TP53mut MCL (Kumar Blood 2021). Methods: The BOVEN trial enrolled 25 patients with TP53mut MCL. Retrospective chart review was performed for 22 patients with TP53mut MCL treated at frontline with standard chemoimmunotherapy regimens (Joffe Blood 2019). Tumor histology and FISH were obtained for each case. Mutational profiling (SNV/indels, CNA) was obtained prior to therapy with various CLIA-certified targeted sequencing platforms (Cheng J Mol Diagn 2015, Dias-Santagata EMBO Mol Med 2010). Mutations in TP53 were annotated using predicted phenotypes and pathogenicity from the TP53 Database (de Andrade Cell Death Differ 2022) and from phenotypic annotation of TP53 mutations (Giacomelli Nat Genet 2018). The primary outcome measures were overall and progression-free survival (OS/PFS). Results: Each patient included in the BOVEN cohort featured variants in TP53 with 26 SNV/indels across all 25 patients. There were 25 SNV/indels across all 22 patients treated with standard regimens. Within the trial cohort 14/25 cases had evidence of biallelic inactivation of TP53 with co-occurring SNV/indels in the gene and 17p loss by FISH or TP53 copy losses by SNParray; the comparator cohort featured 11/22 cases with biallelic losses. Across cohorts, most mutations were missense (>80% in both), with the majority occurring in the DNA binding domain of the p53 protein (76% and 91%, respectively), and with >30% at known hotspot codons in both cohorts. Most mutations (>70% in both cohorts) were predicted to adversely impact WT p53 function, including nonfunctional transactivation activity and dominant negative effect on the WT allele. The majority were also predicted to be highly pathogenic by bioinformatic methods (AGVGD, SIFT, PolyPhen2, PHANTM, REVEL, BayesDel). However, none of these features correlated with outcome. In the BOVEN cohort the variant allele frequency (VAF) of the TP53 SNV/indels approached significant association with poor PFS (HR 33.2, p=0.05) but not OS (HR 28.6, p=0.09). In contrast, in the comparator cohort TP53 VAF appeared to have an opposite effect direction, though this was not significant for PFS (HR 0.98, p=0.2) or OS (HR 0.97, p=0.06). We next investigated the impact of co-occurring molecular alterations on outcome. Besides TP53, the most frequently mutated genes in the trial cohort were KMT2D (24%), ATM (16%), NOTCH2 (12%), and CCND1 (12%). Alterations in these genes did not correlate with outcome. Within the comparator cohort the recurrently mutated genes were SMARCA4 (23%), ARID1A (18%), ARID1B (14%), ATM (14%), and CCND1 (14%). As described previously in this cohort (Joffe Blood 2019), alterations in SMARCA4 were associated with worse PFS (HR 4.2, p=0.02) but not OS (HR 2.2, p=0.27). Mutational burden (SNV and CNA load) was not prognostic in either 12/25 BOVEN samples sequenced with the same panel or the comparator cohort. There was no significant difference in outcome between patients with single mutations in TP53 vs. those with biallelic losses in TP53 in either cohort. Conclusion: Across patients with TP53mut MCL treated either with standard frontline chemoimmunotherapy or on the BOVEN trial, predictions of intrinsic pathogenicity and functional consequences of the TP53 alterations were not prognostic. The impact of TP53 mutation VAF remains unclear, with seemingly opposite effect directions between the two cohorts, though neither finding was significant. As previously described in the comparator cohort, alterations in SMARCA4 are associated with worse PFS in TP53mut MCL, though this was not observed in the BOVEN cohort. There were no significant differences in outcome among patients with biallelic inactivations in TP53 or with higher burdens of somatic alterations.