Prospective analysis and clinical implications of therapy-related clonal hematopoiesis in adolescent and young adult survivors of acute lymphoblastic leukemia

Abstract Adolescents and young adults (AYA) with acute lymphoblastic leukemia (ALL) are exposed to cytotoxic chemotherapy early in life, which increases their risk of developing long-term complications such as cardiovascular disease (CVD), bone loss, and subsequent malignancies. Early exposure to genotoxic therapies may induce or select for therapy-related clonal hematopoiesis (CH), characterized by mutations in hematopoietic cells in the absence of an overt hematologic malignancy. CH has been associated with increased risk for blood cancers, CVD, osteoporosis, and other inflammatory disorders, all of which are complications that are routinely monitored for in ALL survivors. Therefore, we investigated the prevalence of CH and its association with long-term treatment-related complications in ALL survivors. We prospectively collected a total of 106 multi-timepoint samples from 70 unique AYA ALL survivors. Participants were seen at the Childhood, Adolescent and Young Adult Survivor Clinic at the University of Chicago and underwent comprehensive clinical assessment for treatment-related toxicities based on NCCN guidelines. All participants were treated with intensive pediatric or pediatric-inspired young adult chemotherapy regimens (i.e., CALGB 10403). For CH assessment, DNA was extracted from whole peripheral blood and sequenced using a hybrid capture-based targeted gene panel, covering for 22 genes and 95% of CH mutations found in the general population, reaching a depth of 2000x coverage and allelic frequency (AF) sensitivity of 0.01%. We defined presence of CH (CH+) when a variant was identified at AF >0.1%. A total of 76 multi-time point samples from 51 unique non-transplanted survivors were analyzed for CH mutations at this time. The median age of survivors analyzed was 30 (range 15 – 48) and 48% were Females. Participants were 46% White, 29% Hispanic, 13% Black, 2% Asian, and 13% did not report race. The median time between ALL diagnosis and assessment of CH in survivors was 7.4 years (range, 0.15 – 30 years). None of the survivors tested had CH with AF >1%, comparable to historically reported CH frequency in age-matched healthy controls (Jaiswal et al., NEJM 2014) and in contrast to treatment-related CH in an aged-matched cohort of solid cancers survivors (Coombs et al., Cell Stem Cell). We looked at CH clones with AF 0.1–1% and found a total of 33 (64%) of survivors had a least one CH clone: 19 (37.3%) had a mutation in one gene, 5 (9.8%) in two genes, and 9 (17.6%) in three genes. The most frequently mutated genes were: TP53 (66.1%, median AF 0.0011), DNMT3A (10.7%, median AF 0.0021), KRAS (7.1%, median AF 0.0013), and SF3B1 (7.1%, median AF 0.0011). When CH+ vs CH- survivors were compared, we did not observe any differences in median age (31 and 29, respectively, p= 0.16) or female sex (42% and 61%, respectively, p= 0.33). We observed that survivors who were further away from ALL diagnosis had fewer CH clones detected (alpha: -0.023, p= 0.2). Survivors who were ≥15 years from diagnosis had a significantly lower number of CH clones detected (p=5.2 × 10⁻⁵) and were significantly less likely to harbor mutations in DNA damage response genes TP53 or PPM1D (OR: 0.08, p=0.031). We did not observe any significant differences among hematologic parameters at the time of sample collection between both CH groups. There was a trend toward increased frequency of osteopenia (12.1% vs 0%, p = 0.278) and avascular necrosis (24.2% vs 5.6%, p = 0.134) in CH+ survivors compared to those without CH. However, none of these comparisons reached statistical significance. In this first-of-its-kind prospective evaluation of ALL survivors, we identified a high prevalence of low AF CH with a distinct mutational landscape across the survivorship spectrum. The clinical impact of these clones on long-term therapy-related complications is currently under investigation.