Abstract Background: The risk of late distant recurrence (≥ 5 years) in post-menopausal patients with ER+ BC remains significant even after standard adjuvant endocrine therapy (ET). Obesity, which has reached high proportions in Western Countries, is one of the modifiable factors that can contribute to BC recurrences. While body mass index (BMI) is an easily available surrogate measure of adiposity, it might not reflect accurately patient adiposity, especially in post-menopausal patients. We therefore hypothesize that mammary adipocyte measurements might better capture the adiposity of the patient. In this study, we retrospectively evaluated the prognostic role of both BMI and adipocyte measurements in the SOLE trial. Patients & Methods: The SOLE trial (NCT00553410) consisted of 4,884 postmenopausal women with hormone receptor positive, lymph node-positive BC treated with 5 years of letrozole after having completed 4-6 years of adjuvant ET. BMI was available for 3,606 patients with ER+/HER2- BC and categorized according to the WHO classification (underweight: < 18.5, normal weight 18.5 – 24.9, overweight 25 – 29.9, obesity ≥ 30 kg/m2). Considering the small sample size, patients with underweight (n=35) were grouped together with patients with normal weight for the subsequent analyses. We evaluated the association between clinico-pathological characteristics and BMI using Fisher exact test and Kruskal-Wallis test. Given the previously reported association of BMI with distant recurrences (PMID: 21115856), we considered distant-relapse free interval (DRFI) as the main study endpoint. Cox regression analyses were adjusted for histology, chemotherapy (yes/no), type of local therapy, in addition to previously reported clinico-pathological variables (PMID: 29158011). Then, we conducted a stratified case-cohort study consisting of 311 patients (97 with distant relapse-cases) with successful digital assessment of adipocytes on normal mammary adipose tissue sample collected at surgery, as previously described (PMID: 33120083). Stratification factors were treatment arm and type of prior ET. Weighted Cox regression models (PMID: 17554753, 18712477) were used to evaluate prognostic association with adipocyte size (area in µm²) categorized in quartile of the 75th percentile (Q4 vs Q1-3), as done previously (PMID: 33753731). Results: 1392 (39%) patients were normal weight, 1315 (36%) were overweight and 899 (25%) were obese. Older age at diagnosis, lymph node involvement (≥4), large tumor size and menopausal status were associated with overweight and obesity. After a median follow-up of 84 months, no associations were found between DRFI and BMI in the univariable and multivariable analysis (adjusted HR overweight vs normal=0.99, 95%CI: 0.76 – 1.29, p=0.977 and HR obese vs normal=1.11, 95%CI: 0.84 - 1.47, p=0.425). In the case-cohort study, patients with larger adipocytes had an increased hazard of distant recurrence in the univariable and multivariable analysis (see table 1). Conclusion: In this study, we did not detect any association between BMI and survival outcomes. We however found that larger mammary adipocytes were independently associated with an increased hazard of distant recurrence. These results suggest that mammary adipocytes, which can be easily evaluated on H&E slides using digital pathology, should be considered as a novel marker to evaluate the risk of late recurrence in patients with breast cancer. Citation Format: Edoardo Isnaldi, François Richard, Giuseppe Marano, Patrizia Boracchi, Marion Maetens, Giuseppe Floris, Guy Jerusalem, Andrea Gombos, Alastair M. Thompson, Erika Hitre, Stefan Aebi, Marco Colleoni, Patrizia Dell’Orto, Roswitha Kammler, Patrick Neven, Giuseppe Viale, Meredith Regan, Elia Biganzoli, Christine Desmedt. Impact of increased adiposity on extended aromatase inhibitor treatment in postmenopausal patients with estrogen receptor positive (ER+) breast cancer: a retrospective analysis of the SOLE trial [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P3-05-28.