Growth Rate and Outcomes in Locally Recurrent Extremity and Truncal Soft Tissue Sarcoma

Importance Improved prognostic tools are needed for patients with locally recurrent extremity or truncal soft tissue sarcoma (STS). Objective To examine the association between average local recurrence (LR) growth rate and outcomes following resection of locally recurrent extremity or truncal STS. Design, Setting, and Participants This retrospective cohort study used a prospectively maintained database from a single high-volume tertiary sarcoma referral center in the US to identify patients 16 years of age or older who underwent repeat resection of a locally recurrent extremity or truncal STS between July 1, 1982, and December 31, 2021. Patients with atypical lipomatous tumors, desmoid tumors, dermatofibrosarcoma protuberans, angiosarcomas, and prior or synchronous distant recurrence were excluded. Data were analyzed from November 1, 2022, to June 17, 2024. Exposure Average LR growth rate, defined as the sum of recurrent tumor maximal diameters divided by the disease-free interval after index operation. Main Outcomes and Measures The primary outcomes were cumulative incidences of disease-specific death (DSD), with death from other causes as a competing risk, and second LR, with death from any cause as a competing risk. Results The study cohort included 253 patients (median IQR age, 64 51-73 years; 140 55.3% male). The 5-year cumulative incidence of DSD after repeat resection was 29%. Multivariable analysis indicated that LR growth rate (hazard ratio HR, 1.12 95% CI, 1.08-1.18; P lt; .001), younger age (HR, 0.98 95% CI, 0.97-0.99; P = .002), R1 or R2 margins (HR, 1.71 95% CI, 1.03-2.84; P = .04), high LR grade (HR, 2.90 95% CI, 1.17-7.20; P = .02), and multifocality (HR, 2.92 95% CI, 1.70-5.00; P lt; .001) were independently associated with higher incidence of DSD. Using the minimum P value method, the optimal cutoff for growth rate was found to be 0.68 cm/mo. Patients with values above this cutoff had higher 5-year incidences of DSD following repeat resection (63% vs 19%; permutation test P lt; .001) and higher amputation rates (19% vs 7%; P = .008). Only R1 margins were independently associated with higher incidence of second LR (HR, 1.81 95% CI, 1.19-2.78; P = .006). Conclusions and Relevance In this cohort study of patients undergoing resection of a locally recurrent extremity or truncal STS, LR growth rate was independently associated with DSD. These findings suggest that patients with growth rates higher than 0.68 cm/mo who undergo LR resection may have high disease-specific mortality and amputation rates and should be considered for perioperative systemic therapy.