B03 | Clonal hematopoiesis in follicular lymphoma: analysis of the phase III FIL FOLL12 trial

Abstract

Background: Clonal Hematopoiesis (CH) is an age-related phenomenon, characterized by somatic mutations in hematopoietic stem cells involved in myeloid malignancies in individuals without myeloid neoplasms. The clinical interplay between CH and follicular lymphoma (FL), has not been fully elucidated yet.
Methods: The study analyzed 242 FL patients treated with RCHOP or BR in the phase III FIL FOLL12 trial. Genomic DNA from baseline (N=242) and sequential samples (N=211) were available from whole peripheral blood and assessed for CH using CAPP-Seq-based NGS. The gene panel included 28 CH-related genes and a 1% VAF was set for variant calling. Clonal fitness (s, per year) was estimated from weighted regression on logit-transformed VAF. Clones were classified as increasing (s > 0.25/yr), decreasing (s < –0.25/yr), or stable when the 95% CI included zero.
Results: At baseline, CH mutations were identified in 88 cases (36.4%) with DNMT3A (16.9%) and TET2 (12.0%) being the most common affected genes. CH⁺ patients had a significantly higher median age of 66 years compared to 57 for CH^- patients (P<0.001). CH⁺ patients showed superimposable PFS and OS compared to CH^- patients with a 72-month PFS of 65.4% compared to 62.6% (P=0.579) and a 72-month OS of 91.3% compared to 93.9% in CH^- patients (P=0.532), respectively. CH was associated with increased hematologic toxicities, with TET2 mutations predicting higher rates of anemia (25.0% vs. 8.7%; P=0.002), thrombocytopenia (29.2% vs. 10.0%; P=0.002), and grade ≥3 neutropenia (72.2% vs. 35.7%; P=0.042).
The dynamics of CH were assessed in 211 paired samples collected after chemo-immunotherapy (CIT; median follow-up 30 months, IQR 12–36). CH prevalence increased from 36.4% to 48.3% (P=0.01), consistent with therapy-driven clonal selection. Canonical CH mutations (DNMT3A/TET2) remained largely stable over time (median s~+0.23/yr), whereas DNA damage response (DDR) gene mutations (PPM1D, CHEK2, TP53) exhibited markedly elevated clonal fitness (PPM1D~+1.23/yr; CHEK2~+0.90/yr; TP53~+0.79/yr), supporting strong positive selection under CIT. Both regimens promoted PPM1D/CHEK2 expansion, while TP53 growth was exclusive to BR (4.0% vs. 15.2%, P=0.007; RCHOP: 2.1% vs. 1.7%, P=0.794). Importantly, CIT-driven enrichment of high-fitness DDR clones was associated with a three-fold increase in secondary malignancy risk (OR=3.056; P=0.008), indicating a major contribution to long-term toxicity.
Conclusions: CH does not associate with poor outcomes in FL patients when treated in 1^st line with standard of care RCHOP or BR regimens. However, TET2 mutations associate with a higher frequency of hematological toxicity in FL treated with CIT. CIT promotes the expansion of DDR-mutated CH clones, with TP53 growth occurring exclusively after BR. High-fitness DDR clones predict a three-fold increase in secondary malignancy risk, supporting CH profiling as a biomarker to guide chemo-free approaches and limit treatment-related toxicity.