TP53 Mutations Detected by NGS Are a Major Clinical Risk Factor for Stratifying Mantle Cell Lymphoma

Abstract

Mantle cell lymphoma (MCL) is a rare subtype of B-cell non-Hodgkin's lymphoma characterized by significant clinical and biological heterogeneity. Recently, Bruton's tyrosine kinase inhibitors (BTKi) combined with chemoimmunotherapy (CIT) with or without autologous stem cell transplantation (ASCT) have shown promising results in younger patients in first-line clinical trials [1]. Therefore, there is a need to identify high-risk patients whose disease is refractory to standard chemotherapy in order to evaluate new BTKi-based treatments.

The current prognostic factors include the Mantle Cell Lymphoma International Prognostic Index (MIPI), histological and cytological characteristics, increased Ki-67 levels, and other genetic aberrations. TP53 mutations are consistently associated with a negative prognosis, exhibiting a strong and independent association with early disease progression and death, particularly in patients treated with conventional intensive CIT as well as R-bendamustine, lenalidomide, or BTKi alone [2]. Consequently, revising treatment algorithms to account for the TP53 status in both frontline and relapsed MCL patients represents a substantial step forward in optimizing patient-specific management.

One major challenge is improving the detection of TP53-mutated patients enrolled in clinical trials to prevent misclassification of patients with mutations and inaccurate assessment of drug efficacy. Currently, TP53 sequencing is not performed routinely in all MCL patients, and clinical trials have employed various detection methods with differing sensitivities and accuracies to identify TP53 abnormalities, leading to inconsistent results [3]. TP53 assessment often relies on immunohistochemical analysis of biopsy tissues, which can predict missense mutations but fails to detect TP53 null mutations. Only next-generation sequencing (NGS) can accurately detect the full spectrum of mutations leading to either the expression of a mutant protein (e.g., missense mutations) or the loss of P53 expression (e.g., nonsense mutations, frameshifts, and splice site alterations) with high sensitivity in various sample types [4]. NGS also enables the detection of coexisting mutations that reflect clonal heterogeneity and provides a reliable estimate of the size of the mutated TP53 clone according to the variant allele frequency (VAF), particularly for samples with high tumor cell purity. However, there are currently no data on the impact of mutation type or clonal heterogeneity on patient survival.

To assess the frequency and types of mutations and their impact on posttreatment survival, we conducted a retrospective, single-center study of 140 MCL patients diagnosed with circulating tumor cells by flow cytometry between 1986 and 2023 in a real-world clinical setting. All patients were from Avicenne and St. Louis Hospital and provided informed consent; these patients were enrolled in a “B-cell lymphoproliferative disorders” cohort (DC 2009 936). TP53 sequencing, encompassing the entire coding sequence and exon–intron junctions, was performed with NGS, with a VAF cutoff of 1% at the time of diagnosis or retrospectively on the diagnostic sample. The minimum average base coverage depth was approximately 5000×, with a minimum coverage threshold of 1000× for variant calling, requiring at least 10 variant reads. The pathogenicity assessment of the variants was conducted in accordance with the ERIC guidelines for chronic lymphocytic leukemia (CLL) and the recommended guidelines for TP53 variant reporting [5]. Polymorphisms were identified from the TP53 single-nucleotide polymorphism (SNP) data from the latest version of the UMD_TP53 database (https://p53.fr/tp53-database) and excluded (Figure S1). Sequencing was performed in peripheral blood samples with circulating tumor cells from 124 patients, tissue biopsies from 11 patients and bone marrow or other liquid samples from five patients. Clinical and biological characteristics at diagnosis, treatment details, and outcomes were collected (Table S1).

A total of 53 mutations were detected in 34% of the patients (47/140), including six patients with two mutations; this frequency is greater than those typically reported because of the high sensitivity of the technique (including the entire coding sequence with a VAF cutoff of 1%) and the type of sample analyzed, peripheral blood mononuclear cells (PBMC) with high tumor cell infiltration in the majority of cases. VAFs were notably high, exceeding 40% in 64% of the patients. Ten mutations had a VAF of less than 10% and would not have been detected by Sanger sequencing. The majority of mutations were missense mutations (74%) and were distributed mostly in the DNA binding domain, with hotspot mutations at the R278 codon (9/53 mutations). Fluorescence in situ hybridization (FISH) analysis was performed for 68 patients, revealing a 17p deletion in 20 patients, half of which were associated with a TP53 mutation (Figure 1A–C; Tables S1 and S2).

FIGURE 1

Open in figure viewerPowerPoint

Clinical impact of TP53 mutations. (A) Distribution of TP53 mutations in the TP53 protein. (B) Proportion of TP53 mutation types, expressed as a percentage of the total number of mutations detected (n = 53). (C) Frequencies of patients with mutations only, 17p deletions only, both mutations and deletions, and no TP53 abnormalities. Frequencies are calculated relative to the total number of cases screened for mutations and deletions (n = 68). (D) Kaplan–Meier estimates of overall survival (OS) based on the presence or absence of a TP53 mutation and null or missense mutations. (E) Kaplan–Meier estimates of PFS based on the presence (mut) or absence (WT) of TP53 mutations and the use of HD cytarabine or other treatments (tt). (F) Kaplan–Meier estimates of OS based on the presence or absence of TP53 deletions and mutations compared with patients without abnormalities. p values by log-rank test.

Over a median follow-up of 6.4 years, 76 out of 140 patients died. TP53-mutated patients exhibited inferior overall survival (OS) to that of unmutated patients (median OS: 2.45 vs. 9 years) (p = 0.001) We next assessed how the type of mutation influences patient prognosis. Missense mutations result in the expression of aberrant p53 protein with a dominant negative effect, whereas null mutations (nonsense, splice site, and frameshift mutations) lead to the complete absence of p53 expression. TP53 null mutations and missense mutations had the same impact on survival (Figure 1D).

A total of 131 patients required treatment that included high-dose cytarabine (RCHOP/RDHAP or RDHAX) as the primary treatment modality, followed by ASCT in a subset of patients, which were evenly distributed between the TP53-mutated and unmutated groups. Among the remaining nine patients who did not receive any treatment, only two harbored a TP53 mutation. Progression-free survival (PFS) after first-line treatment was also shorter in patients with TP53 mutations than in those without mutations (median PFS: 0.58 vs. 4.58 years; p = 0.013) (Table S1 and Figure S2C) and 1-year PFS rates were 45% versus 80%. The median PFS was shorter in TP53-mutant patients treated with high-dose cytarabine regimens (p = 0.001) and other treatments (R-bendamustine, RCHOP, or ibrutinib) (p = 0.016) than in unmutated patients (Figure 1E). These results are consistent with those of clinical trials; however, the TP53 mutant group was underrepresented in these trials [2].

Consistent with previous reports, blastoid morphology, a Ki-67 index > 30%, and high risk according to the MIPI were significantly associated with worse outcomes (p < 0.001, 0.002, and < 0.001, respectively). Notably, both blastoid morphology and a Ki-67 index > 30% were significantly associated with TP53 mutations (p = 0.027 and 0.006, respectively, chi-square test) (Figure S2A). Patients with a high risk according to the MIPI were distributed equally in both groups due to the advanced age of the patients and a predominant leukemic presentation (lymphocytes count > 5 × 10⁹/L and high leucocytes count) which led to an overestimation of the risk of progression.

MCL has a heterogeneous cellular origin; most cases are pregerminal, characterized by few/no IGHV somatic mutations, while others are postgerminal center in origin and associated with a higher somatic IGHV mutational burden. Navarro et al. published that mutated IGHV and non-nodal leukemic clinical presentation may correspond to a disease subtype with more indolent behavior. Our study focuses on leukemic MCL. IGHV mutational status was first analyzed to confirm the detection of clonality and to document these leukemic patients with IGHV mutational status and restriction in IGHV gene usage. IGHV gene rearrangements were analyzed by NGS (LymphoTrack IGHV Leader somatic hypermutation for MiSeq Illumina), and clonotypes were identified via the Vidjil web application (https://app.vidjil.org/) to document these leukemic cases and identify a possible indolent profile. The mutated IGHV patients represented one-third of the patients and were equally divided between the mutated and unmutated TP53 groups. The IGHV rearrangement profile was skewed, with a strong restriction in IGHV gene usage. The most common rearrangements were IGHV4-34 (14%), VH3-21 (9%), VH5-51 (6%), and VH1-8 (7%), which were predominantly unmutated. In contrast, the VH3-33 and VH4-39 rearrangements were found to contain mutations. The IGHV mutation status, regardless of whether a 98% or 97% identity threshold was used and regardless of TP53 status, did not affect survival or progression (Figure S2A,D) This finding contrasts with the study by Yi et al. [6], which indicated that mutated IGHV genes are enriched in the C1 group and are associated with a more favorable outcome. Furthermore, IGHV gene usage had no impact on survival (Figure S2B).

Although the data concerning the presence of a 17p deletion were incomplete, they enabled the identification of two groups with differing survival rates. The presence of a mutation, whether alone or in combination with a 17p deletion (31% of all patients) (Figure 1C), had a similarly negative impact on survival (Figure 1F). In contrast, having a deletion alone (15% of all patients) did not significantly affect survival, as indicated by the OS curve, which was not significantly different from that of the group without TP53 abnormalities (54% of all patients). Only one patient with indolent behavior had a TP53 nonsense mutation and a 17p deletion. These findings are consistent with those of previous studies and underscore that NGS is mandatory for assessing TP53 abnormalities to identify high-risk patients.

In conclusion, TP53 sequencing using NGS should be conducted before treatment in all patients enrolled in clinical trials as recommended by NCCN Guidelines [7]. This method is crucial for identifying groups at risk of relapsed/refractory disease and accurately interpreting treatment efficacy results. In our real-world cohort, only half of the patients had been subjected to TP53 sequencing at the time of diagnosis, with 34% of the overall cohort exhibiting a mutation. Unlike in CLL, TP53 mutation status has not yet been integrated into a therapeutic algorithm for MCL. However, as these data become more widely accessible, they will enable the stratification of patients in clinical trials according to TP53 mutations to assess the value of adding BTKi to first-line treatment.