INTRODUCTION Prevalence of CHIP in cancer patients (pts) is estimated at about 25%, its presence being associated with inferior outcomes and with increased risk of development of therapy-related myeloid neoplasms (TRMN). Despite the increased body of knowledge on cancer and CHIP, processes driving the selection of clones and their latter malignant transformation have not been fully elucidated. We hypothesized that CHIP in cancer pts might not only lead to TRMN but also affect the prognosis of the primary neoplasm and its treatment-related toxicity. Our study aims to describe the prevalence and dynamics of CHIP in treatment-naïve pts with cancer and to analyze its impact on clinical outcomes. METHODS This study included 103 pts with a first cancer diagnosis at age ≥ 60 years and eligible for anticancer treatment without prior exposure to cytostatic agents. Peripheral blood (PB) samples were collected at diagnosis and 6 months after treatment. A customized NGS panel covering common CHIP genes ( DNMT3A, TET2, ASXL1, JAK2, PPM1D, TP53, SF3B1, GNB1, SRSF2, CHEK2, CBL, GNAS, and NRAS) was used to identify CHIP-positive (VAF≥ 1%) and CHIP-negative cases. Clonal dynamics were assessed through NGS at 6 months after treatment, categorized as ‘growing’ when their VAF increased by more than 25% compared to the baseline, ‘shrinking’ when VAF decreased by more than 25% compared to the baseline and ‘stable’ if it remains unaltered. RESULTS Baseline characteristics are shown in table 1.The prevalence of CHIP in our cohort was 35%, with an average of 1.5 somatic variants per patient and 54 identified variants, and a median VAF of 4.7% (IQR 2.5-11.0%). Notably, mutations in DNMT3A (33%), TET2 (28%), and PPM1D (13%) were the most prevailing gene aberrations, accounting for nearly 75% of all variants, while other common CHIP genes such as ASXL1 (9%) or JAK2 (0%) were less frequent. TP53 variants represented 9% of all mutations, whereas SF3B1, GNB1, SRSF2, and CBL each accounted for 2%. Breast cancer pts displayed a significantly higher prevalence of CHIP compared to other primary neoplasms (66% vs. 36%, p=0.01) whereas no patient with bladder neoplasm presented CHIP at diagnosis (p=0.04). These observations were not warranted only by differences in the age or smoking habit of these subgroup of neoplasms. The mutational spectrum of CHIP across different cancer categories was comparable. Among 20 paired samples sequenced (baseline and post-genotoxic exposure), 41% of all variants exhibited a growing pattern, 31% a shrinking pattern, and 28% remained static (Figure 1). Platinum-based therapy exposition promoted clonal expansion in DNMT3A mutations (p=0.03), while this effect was not observed in PPM1D or other genes, likely due to the low sample size. Age, tobacco use, and type of primary neoplasm did not appear to influence clonal fitness. There were no significant differences in the incidence of infectious complications or chemotherapy-induced hematologic toxicity between CHIP-positive and CHIP-negative cohorts. With the current follow-up time (16.8 months), the overall response (ORR) and complete response rates (CR) appeared comparable (ORR: 94% in CHIP-positive vs. 85% in CHIP-negative; CR: 79% vs. 70%, respectively). A patient with diffuse large B-cell lymphoma and CHIP ( ASXL1, PPM1D, and TP53 variants) developed a TRMN (MDS-MD) 7 months after completing R-CHOP treatment. CONCLUSION The prevalence of CHIP in our cancer pts cohort is 35%, with breast cancer cases displaying a CHIP occurrence around 62% not previously reported. Our study highlights an enrichment of mutations in PPM1D in treatment-naïve cancer pts, surpassing the frequency of ASXL1 in contrast to prior literature. Genotoxic therapy promotes clonal expansion in 41% of variants in our cohort; although factors influencing CHIP fitness remain poorly understood, DNMT3A showed heightened susceptibility to platinum therapy. Finally, and in contrast contrast with our initial hypothesis, we found no evidence of impaired outcomes in the CHIP population. These results emphasize the need for further longitudinal follow-up. Acknowledgements: This work was supported by two grants from the Instituto de Salud Carlos III (PI20/00881 and PI 20/00531)(Co-funded by European Regional Development Fund. ERDF, a way to build Europe). 2021 SGR 00560 (GRC) Generalitat de Catalunya; economical support from CERCA Programme.
Read full abstract