Antibody-drug conjugate–associated pneumonitis (AAP) across novel first-in-human phase I trials: A pooled single-center analysis.

Abstract

524 Background: Antibody-drug conjugates (ADCs) are increasingly used in cancer care; toxicities like Antibody-Drug Conjugate Associated Pneumonitis (AAP) have emerged as clinically significant & fatal adverse events. Data characterizing AAP across phase 1 trials is limited. Methods: We retrospectively collected data on patients (pts) enrolled across Phase 1 ADC trials (2015-2025) at our center who were diagnosed (Dx) with AAP. Descriptive statistics were used for analysis. Time to AAP was analyzed using Kaplan–Meier (KM) method & cumulative incidence calculated as 1 – KM. Results: Of 290 phase 1 trials, 25 trials used ADCs with 247 pts enrolled; 14.5% (n = 36/247) developed AAP. Median age at AAP Dx was 70 yrs (IQR = 64.7-74). Leading cancers were ovarian (55.6%, n = 20/36) & endometrial (30.6%, n = 11/36). Preexisting lung disease was present in 8.3% (n = 3/36; COPD: n = 2/3, asthma: n = 1/3); 16.7% (n = 6/36) had prior chest radiation. Median no. of prior therapies (Tx) was three. Prior to ADC start, 50% (n = 18/36) received checkpoint inhibitors. Baseline PFT was done in 36.1% (n = 13/36), with a median baseline DLCO of 58% (IQR 48–90). ADC monotherapy (75%, n = 27/36) was the commonly used Tx. Median duration of ADC Tx was 5.5 months (m). Top ADC targets were FOLR1 (33.3%, n = 12/36), TROP2 (30.6%, n = 11/36) & NaPi2b (19.4%, n = 7/36). Radiographic patterns of AAP are shown (Table). Median time from ADC start to AAP was 3.6 m. Estimated cumulative incidence of AAP was 41.7% (<3 m), 22.2% (3-6 m), & 22.2% (6-12 m). Grade (G) distribution was: G1 (63.9%; n = 23/36), G2 (22.2%; n = 8/36), G3/G4 (5.6%; n=1/36 each), G5 (8.3%; n=3/36). Steroids were given in 75% (n = 27/36) [oral<1 mg/kg (66.6%, n = 18/27); oral ≥1 mg/kg (25.9%, n = 7/27) & IV steroids (7.4%, n = 2/27)]. ADC Tx was permanently discontinued due to AAP in 55.6% (n = 20/36), with G1 AAP in 55% (n = 11/20). 13.9% (n = 5/36) were rechallenged (G1 AAP in 80%, n = 4/5), with 40% (n = 2/5) continuing on trial & 20% (n = 1/5) each discontinuing due to cancer progression, other toxicity, or AAP progression (G2 to G3). ADC therapy continued uninterrupted in 30.6% (n = 11/36), of which 72.7% (n = 8/11) had G1 AAP. Conclusions: To our knowledge, this is the first and largest pooled dataset characterizing AAP across novel first-in-human phase 1 ADC trials. High cumulative incidence of AAP within the first 3m of Tx underscores the need for early recognition with implications for safety monitoring in Phase 1 ADC drug development. CT pattern distribution(not mutually exclusive) % [Pts with CT pattern/ Total pts (36)] Patchy ground-glass opacities (GGO) 22.2 (8) Diffuse GGO 55.6 (20) Consolidation 25 (9) Nodules 25 (9) Fibronodular 16.7 (6) Reticular 2.8 (1) CT pattern distribution ADC Targets % (CT pattern/Total pts with target) FOLRα TROP2 NaPi2b Patchy GGO 8 (1/12) 55 (6/11) 14 (1/7) Diffuse GGO 50 (6/12) 36 (4/11) 57 (4/7) Consolidation 42 (5/12) 9 (1/11) 14 (1/7) Nodules 0 (0) 0 (0) 43 (3/7)