Active surveillance in patients with extra-abdominal desmoid-type fibromatosis: A combined analysis of three prospective observational studies.

11573 Background: Three observational studies on active surveillance (AS) in patients with extra-abdominal desmoid-type fibromatosis (DTF) provided prospective evidence supporting AS as frontline approach. Identification of prognostic factors for failure of AS will help determine the appropriate treatment strategy. The aim of this study was to confirm previous findings and investigate the prognostic impact of clinical and molecular variables on a larger series. Methods: Data available on January 31st 2024 from 3 prospective observational studies in Italy, the Netherlands and France in which DTF patients were followed for ≥3 years during an initial AS approach were merged. Patients ≥18 years, with primary extra-abdominal sporadic DTF (abdominal wall, extremity, head & neck, trunk) and with CTNNB1 mutation (T41A, S45F, S45P, WT, “other”) available were eligible for the analyses. The primary study endpoint was treatment-free survival (TFS). Secondary endpoints included incidence of RECIST progression, spontaneous RECIST regression and regression post-RECIST progression. Results: Two hundred and eighty-two patients (n= 282) with median follow up of 53 months (IQ, 39-63) were included. Three-year and five-year TFS and crude cumulative incidences (CCI) of RECIST progression, RECIST regression and regression post-RECIST progression were 69% (64-75%) and 66% (61-73), 34% (95% CI 28-40) and 35% (30-41), 26% (95% CI 21-32) and 34% (29-41), and 35% (95% CI 26-47) and 40% (30-52), respectively. Larger initial tumour size (P=0.031), CTNNB1 mutation (S45F and “other” mutation P=0.049), and head/neck and extremity tumour locations (P=0.032) were associated with worse TFS at multivariable analysis (Table 1). Consistently smaller size and abdominal wall were associated with higher probability of initial spontaneous RECIST regression and a specular trend for mutation was also seen (Table). Conclusions: This study confirms that spontaneous regression occurs in a significant proportion of patients and two third are treatment free at 5 years. Initial size, CTNNB1 mutation and location should be factored in the initial decision-making process. [Table: see text]