Abstract PO4-14-05: Germline and acquired genetic variants and long-term cancer-related fatigue among survivors of early-stage breast cancer (BC)

Abstract BACKGROUND: Previous studies suggested that bio-behavioral models explain part of the variability of cancer-related fatigue, and pro-inflammatory and age-related processes emerged as contributors to persistent fatigue. However, the biological underpinnings of this complex symptom, including its relevant genomic correlates, are still poorly understood. We aimed to comprehensively explore the association of genetic variants with long-term cancer-related fatigue. METHODS: We used the prospective, multicenter, longitudinal CANcer TOxicity cohort, enrolling patients at diagnosis of stage I-III BC (NCT01993498). Outcomes included severe (≥40/100) global (EORTC QLQ-C30) as well as physical, emotional, and cognitive (FA12) fatigue at year (Y)1, Y2, and Y4 after BC diagnosis. Applying a hypothesis-driven approach, we investigated (i) germline genetic variants in a pre-specified Genetic Risk Index (GRI) assaying expression-regulating single nucleotide polymorphisms (SNPs) in the promoter regions of 3 pro-inflammatory cytokine genes (IL1B [rs16944], IL6 [rs1800795], and TNF [rs1800629]; Bower J, JCO 2013; N=9035), and (ii) age-related expansions of hematopoietic clones carrying recurrent acquired mutations, commonly defined as clonal hematopoiesis of indeterminate potential (CHIP; NGS with unique molecular identifiers for 17 genes [including DNMT3A, TET2, ASXL1, PPM1D, ATM, JAK2, and TP53]; N=1219). The contribution of GRI and CHIP to fatigue was assessed by multivariable logistic regression models. Taking an agnostic approach, we then performed a genome-wide association study (GWAS) of fatigue and 1,894,511 germline SNPs (Illumina InfiniumExome24/Illumina GSA24; N=9056). All genetic variants were assessed at BC diagnosis. Tested associations were adjusted by validated clinical predictors (Di Meglio A, JCO 2022). RESULTS: In the overall cohort (N=9056) mean age was 56.5 years (SD 11.2), 51.0% had stage II or III BC, 52.6% received chemotherapy, and 82.1% endocrine therapy. The prevalence of severe fatigue was 26.1% at diagnosis and increased to 36.5%, 34.3%, and 32.6% at Y1, Y2, and Y4 after diagnosis, respectively. Most patients (77.4%) had 3-6 high-frequency alleles across the 3 SNPs in the pro-inflammatory GRI. Among patients with available CHIP data, 15.2% and 3.7% had a Variant Allele Frequency (VAF) >2% and >10%, respectively. The pro-inflammatory GRI was not associated with fatigue outcomes. We found significant associations between CHIP and severe long-term global fatigue at Y4 (adjusted OR [95%CI]: VAF ≥2% vs < 2%, 1.64 [1.08-2.51], p=0.021; VAF ≥10% vs < 10%, 2.22 [1.00-4.94], p=0.050). The GWAS identified several regions (each including ≥2 SNPs) as associated with severe fatigue outcomes (suggestive p-value cut-off < 5 × 10−5; see Table for number of identified regions by outcome and respective nearby genes). Overlap in biological pathways was observed in some of the gene regions associated with global and physical fatigue, but not for emotional or cognitive fatigue. Common biological processes that were associated (FDR < 0.05) with such genes included synaptic transmission, hemoglobin-oxygen binding, and response to stress. CONCLUSIONS: This study provides suggestive association data between gene variants and long-term cancer-related fatigue. While previously reported associations with pro-inflammatory GRI were not confirmed, some biological processes that may inform the mechanistic understanding of this symptom emerged, including associations with aging and response to stress that warrant further exploration. ///: marks different regions. All association analyses were adjusted by age, body mass index, smoke behavior, pre-treatment fatigue, pain, insomnia, and anxiety, and by the first ten axes of principal component analysis of the genetic data to control for population stratification. There was no overlap in biological pathways for the gene regions associated with emotional or cognitive fatigue (gene regions not shown). Table. Genes in proximity of regions associated with severe fatigue outcomes Citation Format: Antonio Di Meglio, Emilie Thomas, Youenn Drouet, Jean-Baptiste Micol, Davide Soldato, Maria Alice Franzoi, Martina Pagliuca, Julie Havas, Anne-Laure Martin, Sibille Everhard, Celine Besse, Anne Boland, Christophe Marzac, Nathalie Droin, Sandrine Boyault, Marina Rousseau, Olivier Trédan, Paul Cottu, Christelle Jouannaud, Marion Fournier, Laurence Vanlemmens, Charles Coutant, Asma Dhaini Merimeche, Baptiste Sauterey, Florence Joly, Mario Campone, Florence Lerebours, Marie-Ange Mouret-Reynier, Olivier Rigal, Thierry Petit, Sophie Guillermet, Antoine Arnaud, Mahmoud Ibrahim, Sylvie Giacchetti, Florence Dalenc, Johanna Wassermann, Olivier Arsène, Ariane Darut-Jouve, Aurélie Bertaut, Fabrice André, Jean-Francois Deleuze, Alain Viari, Ines Vaz Luis. Germline and acquired genetic variants and long-term cancer-related fatigue among survivors of early-stage breast cancer (BC) [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-14-05.