Résumé
Background: More than half of patients with early BC on aromatase inhibitors (AI) experience AIMT. Inflammatory traits linked with treatment-related estrogen depletion have been associated with AIMT. A potential role for germline variants in genes regulating inflammation, bone turnover, estrogen metabolism and AI pharmacokinetics has also been suggested (Reinbolt 2018, Stearns 2024, Lintermans 2016, Wang 2016, Hertz 2017). Yet study were relatively small and results inconsistent or not validated. Therefore, the understanding of the biological and potential genetic underpinnings of AIMT is still limited. We aimed to comprehensively explore the association of genetic variants with AIMT in order to identify targets for impactful interventions. Methods: We included postmenopausal patients with stage I-III BC treated with adjuvant AI from the longitudinal CANTO cohort (NCT01993498). AIMT was evaluated at month 3-6, year (Y) 1, Y3 and Y5 of AI treatment and defined as presence of any grade [G] articular or muscular pain by CTCAE v4.0. All germline variants were assessed on blood samples obtained at BC diagnosis. We performed a GWAS between 1,894,475 single nucleotide polymorphisms (SNPs) (Illumina InfiniumExome24 / Illumina GSA24) and AIMT, using multivariable logistic regressions. Each model was adjusted on pre-specified clinical and behavioral factors, including age, socio-economic status, comorbidities, tumor and treatment-related variables (including type of AI), and patient-reported health outcomes, all assessed at BC diagnosis. Missing data was handled by multiple imputation with a set of 15 complete datasets. The study had approximately 80% power to detect associations with AIMT at Y3 at suggestive GWAS significance alpha level of 10-6 targeting variant with minor allele frequency (MAF) of 0.25 and odds ratio (OR) of 1.5. Therefore, our primary outcome of interest was AIMT at Y3. Given the exploratory nature of these analyses, we report relevant associations at an alpha threshold of 10-4. Genes harboring significant top SNPs were annotated to identify potential explicatory biological pathways. Results: Overall 4854 patients treated with AI were included. Of these, 3847 had available genetic data. Mean age was 63.9 years (SD 7.2), mean BMI was 27.0 (5.6). 63% patients reported history of articular disease, 41% received chemotherapy. 55% received letrozole, 40% anastrozole and 5% exemestane as first AI. 65% of patients reported AIMT at Y3 (of whom G3 9%), while 86% patients reported AIMT overall (at any time point; of whom G3 18%). Patients with AIMT at Y3 were generally younger, with higher BMI, had previous history of articular or muscular disease, and reported more frequently anxiety or depression at diagnosis. The GWAS identified 9 SNPs as associated with AIMT at Y3: rs4096589 (MAF 0.39; OR 1.36 [95% CI 1.21-1.53]), rs983495 (MAF 0.46; OR 0.76 [0.68-0.85]), rs12964962 (MAF 0.25; OR 0.73 [0.64-0.84]), rs348780 (MAF 0.42; OR 0.77 [0.69-0.86), rs72485589 (MAF 0.13; OR 0.67 [0.56-0.81]), rs2559854 (MAF 0.42; OR 0.78 [0.69-0.87]), rs895876 (MAF 0.16; OR 0.72 [0.61-0.83]), rs2926866 (MAF 0.17; OR 1.40 [1.20-1.66]), rs2240027 (MAF 0.08; OR 1.58 [1.28-1.96]). rs2926866 was also associated with AIMT at Y1 (1.22 [1.05-1.41]), Y5 (1.36 [1.12-1.66]), and overall (1.48 [1.24-1.76]). Relevant molecular functions that were associated with the identified SNPs included protein homodimerization activity (rs4096589), identical protein binding (rs895876), actin- and actin-filament binding (rs2240027). Conclusions: This was an exploratory study aimed to assess associations between genetic variants and AIMT in postmenopausal patients with early BC. Results of this relatively large prospective cohort identified several SNPs that might be involved in AIMT. Further analyses are warranted for model validation and to better elucidate the mechanistic role of the described genetic variants from a functional perspective.
Citation Format: Pietro Lapidari, Youenn Drouet, Emilie Thomas, Jérémie Jacquemin, Maria Alice Franzoi, Martina Pagliuca, Sophie Laurent, Chayma Bousrih, Maryam Lustberg, Olivier Tredan, Anne-laure Martin, Catherine Gaudin, Christelle Jouannaud, Marion Fournier, William Jacot, Marina Rousseau, Jean Francois Deleuze, Alain Viari, Ines Vaz-Luis, Antonio Di Meglio. Genome-wide association study (GWAS) of aromatase inhibitor musculoskeletal toxicity (AIMT) among early-stage breast cancer (BC) survivors [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr PS15-02.