Integration of gene mutations in risk prognostication for watch-and-wait follicular lymphoma patients initiating first-line treatment

Abstract

Follicular lymphoma (FL) patients with low tumor burden at diagnosis frequently undergo the watch-and-wait (W&W) strategy. The study aimed to facilitate risk assessment in predicting the time to lymphoma treatment (TLT) for W&W patients through an integrated analysis of clinical factors and genetic mutations. A retrospective study was conducted on 214 FL patients managed with W&W between 2016 and 2023. Among them, 184 patients underwent targeted sequencing. The median follow-up was 30.4 months (IQR 21.4-41.9, range 6.4-95.8). A clinico-genetic model m3-PRIMA-PI was developed using the multivariate Cox proportional hazards method, incorporating two clinical parameters (bone marrow involvement and elevated β2-MG) and three gene mutations (KMT2D, EP300, and TP53). Patients were categorized into low (69.0%), intermediate (21.7%), and high (9.2%) risk groups. Probabilities of treatment initiation at one year were 11.0% (95% CI, 5.2%-16.5%), 26.0% (95% CI, 10.7%-38.7%), and 54.3% (95% CI, 22.3%-73.1%); and at 2 years were 29.4% (95% CI, 20.2%-37.5%), 49.8% (95% CI, 31.1%-63.4%), and 93.5% (95% CI, 56.7%-99.0%), respectively. The predictive performance for TLT was superior with m3-PRIMA-PI, achieving a C-index of 0.66 (95% CI, 0.63-0.69), compared to established indexes like FLIPI (C-index 0.59, 95% CI, 0.56-0.62) and FLIPI2 (C-index 0.59, 95% CI, 0.55-0.61). The above results were further validated in an independent external cohort. The m3-PRIMA-PI may provide a promising tool for risk stratification in W&W FL patients.

Figure 1

Patient flow chart and genomic landscape of the watch‐and‐wait cohort. (A) Flow chart for FL patients of the watch‐and‐wait cohort. (B) The overall landscape of genetic mutation of 184 patients with targeted sequencing data. Multi‐hit indicating cases with more than one type of alteration. The right panel indicates the mutation rates and number of corresponding genes.

Figure 2

Patterns of treatment initiation and survival. (A) TLT curves in the watch‐and‐wait FL patients. (B) OS curves in the watch‐and‐wait FL patients. No. at risk, number at risk; OS, overall survival; TLT, time to lymphoma treatment.

Figure 3

Clinic and genetic factors related to inferior TLT. (A–C) TLT curves in watch‐and‐wait FL patients according to lymph node sites (A), bone marrow involvement (B), and β2‐MG (C). (D–F) TLT curves in watch‐and‐wait FL patients according to mutation status of KMT2D (D), EP300 (E), and TP53 (F). No. at risk, number at risk; β2‐MG, β2‐microglobulin.

Figure 4

The clinico‐genetic m3‐PRIMA‐PI model for TLT prediction in the training cohort. (A and B) TLT of the watch‐and‐wait cohort stratified by the original score (A) and the m3‐PIMA‐PI model (B) in the training cohort. (C) The distribution of predictors, scores, and risk groups for 184 patients in the training cohort. The left panel shows predictors incorporated in the model. The right panel indicates the rates of corresponding predictors. The bottom is the score calculated by the sum weight of each predictor and the risk group assigned to each patient. BM, bone marrow; No. at risk, number at risk; β2‐MG, β2‐microglobulin. (D) The reclassification of risk groups across FLIPI, FLIPI2, and m3‐PRIMA‐PI in the training cohort. HR, high risk; IR, intermediate risk; LR, low risk; TFT, test for trend.

Figure 5

The clinico‐genetic m3‐PRIMA‐PI model for TLT prediction in the validation cohort. (A and B) TLT of the watch‐and‐wait cohort stratified by the original score (A) and the m3‐PIMA‐PI model (B) in the validation cohort. (C) The distribution of predictors, scores, and risk groups for 184 patients in the validation cohort. The left panel shows predictors incorporated in the model. The right panel indicates the rates of corresponding predictors. The bottom is the score calculated by the sum weight of each predictor and the risk group assigned to each patient. BM, bone marrow; No. at risk, number at risk; β2‐MG, β2‐microglobulin. (D) The reclassification of risk groups across FLIPI, FLIPI2, and m3‐PRIMA‐PI in the validation cohort. HR, high risk; IR, intermediate risk; LR, low risk; TFT, test for trend.