Abstract Background: Antibody drug conjugates (ADCs) have revolutionized the treatment of many solid tumours and hematological diseases. Little is known about the influence of body composition on efficacy and safety of ADCs. Methods: All patients treated with ADCs in early phase clinical trials between 03/2015 and 03/2023 at the Drug Development Department at Gustave Roussy were retrospectively included in the analysis. A deep learning software (Anthropometer3DNet) automatically measured anthropometric parameters in 3D on pretreatment scans, allowing multi-slice measurements of muscle body mass (MBM), fat body mass (FBM), subcutaneous fat mass (SFM) and visceral fat mass (VFM). Other clinical and biological parameters were also retrieved and evaluated. Primary endpoints were progression-free survival (PFS) and overall survival (OS). Survival analysis was performed using Cox regression models and Kaplan-Meier (KM) estimator. Cut-offs were determined using the maxstat method. Anthropometric parameters’ effect on toxicity was also evaluated using a ROC analysis. Results: A total of 136 patients were treated with ADCs. The most frequent tumor types were non-small cell lung cancer (56 patients, 41%) and colorectal cancer (31 patients, 23%). Median age, ECOG PS, albumin and number of previous lines of treatment were 60.8 years (30 to 85), 1, 42 g/L and 3 respectively. In the overall cohort, median PFS and OS were 2.8 and 8.6 months respectively, 90 (66%) patients had experienced toxicity, of which 46 were grade 3-4. The Royal Marsden Hospital (RMH) prognostic score of 2 or more was significantly associated with worse PFS (HR=2.13, p=0.007) and OS (HR=2.12, p=0.008). The number of treatment lines was significantly associated with worse OS (HR=1.21, p<0.001). For the anthropometric parameters, SFM (HR=0.89, p=0.013) and FBM (HR=0.89, p=0.02) were significantly associated to PFS. Higher SFM (cutoff: 3.3 kg/m2) and FBM (cutoff: 3.6 kg/m2) were significantly associated with longer PFS (mPFS=3.8 vs 2.3 months and 3.7 vs 1.9 respectively). Patients with higher SFM and FBM had numerically but non-significantly longer (mOS=8.5 vs 7.7 months and 8.6 vs 6.4 months respectively). All anthropometric parameters were significantly associated with all-grade toxicity in the univariate but not in the multivariate analysis. None of the parameters were associated with grade 3-4 toxicity. Conclusions: In this large monocenter cohort, 3D measured high SFM and FBM were significantly associated with PFS. Automatic extraction of body composition parameters using AI may help in anticipating the benefits of ADCs in patients included in early-phase clinical trials. Citation Format: Matthieu Delaye, Littisha Lawrance, Younes Belkouchi, Felix Wirth, Pierre Decazes, Alexandre Bône, Kaïssa Ouali, Antoine Hollebecque, Cristina Smolenschi, Rastislav Bahleda, Anas Gazzah, Stéphane Champiat, François-Xavier Danlos, Clémence Hénon, Kristi Beshiri, Madona Sakkal, Aurélien Marabelle, Jean-Marie Michot, Vincent Ribrag, Claudia Parisi, Christophe Massard, Vincent Goldschmidt, Pierre Vera, Santiago Ponce-Aix, Nathalie Lassau, Samy Ammari, Capucine Baldini. Subcutaneous fat mass predicts progression-free survival in patients treated with antibody-drug-conjugates in early-phase clinical trials [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6438.