Histopathological Response After Neoadjuvant Chemotherapy for High-Risk Soft-Tissue Sarcomas: A Secondary Analysis of a Randomized Clinical Trial

Importance: Treatment of high-risk soft-tissue sarcoma (STS) of extremity or trunk wall involves neoadjuvant chemotherapy (NACT) followed by surgery. Histopathological response could estimate patient outcomes. Objective: To characterize morphological changes in surgical specimens of patients treated with NACT with or without radiotherapy (RT) to identify histopathological features that stratify risk of recurrence and ultimately estimate the benefit from neoadjuvant treatments. Design, Setting, and Participants: This was a preplanned prospective secondary analysis of the ISG-STS 1001 clinical trial, a study with both a randomized clinical trial (conducted between 2011 and 2016) and a nonrandomized patient cohort (included between 2016 and 2020) at 32 centers across Italy, Spain, France, and Poland. Participants were patients with STS randomly assigned to receive either anthracycline plus ifosfamide or histotype-tailored (also termed histology tailored) NACT. Data analyses were performed from January to June 2023. Intervention: Participants received 3 cycles of anthracycline plus ifosfamide or histotype-tailored NACT with or without RT followed by surgery. Main Outcomes and Measures: The primary outcome was disease-free survival (DFS). Histopathological features considered included the proportion of stainable tumor cells, tumor necrosis, hemorrhage, fibrohistiocytic reaction with hemosiderin, sclerosis or fibrosis, and sclerohyalinosis. The proportion of stainable tumor cells was classified according to the European Organization for Research and Treatment of Cancer-Soft Tissue and Bone Sarcoma Group categories or as absent or present. The continuous variable of sclerohyalinosis, expressed as a percentage, was categorized based on the second tertile of its distribution (20%). Tumor necrosis, hemorrhage, fibrohistiocytic reaction with hemosiderin, sclerosis or fibrosis, which were also expressed as a percentage, were classified as absent or present. Results: A total of 388 patients (201 in randomized cohort, 187 in nonrandomized cohort; median [IQR] age, 50 [41-60] years; 245 males [63.1%]) were evaluable for histopathological response. In the randomized cohort, after a median (IQR) follow-up of 86 (70-99) months, 115 of 201 patients (57.2%) developed a disease recurrence. The proportion of stainable tumor cells (>1%) was not associated with DFS (hazard ratio [HR], 1.47; 95% CI, 0.36-5.98; P =.59). Necrosis (>1%) was associated with shorter DFS (HR, 3.11; 95% CI, 1.36-7.14; P =.007), while sclerohyalinosis greater than 20% was associated with longer DFS (HR, 0.51; 95% CI, 0.28-0.94; P =.03). Exclusion of patients who received preoperative RT did not alter these associations. In patients randomly assigned to anthracycline plus ifosfamide (n = 98), sclerohyalinosis greater than 20% remained associated with longer DFS (HR, 0.24; 95% CI, 0.09-0.67; P =.007). These findings were confirmed when a broader cohort (n = 187) was included. Conclusions and Relevance: In this secondary analysis of a randomized clinical trial, the proportion of stainable tumor cells, currently considered as the most relevant posttreatment change, did not stratify patient risk. The findings support consideration of the presence of sclerohyalinosis (>20%) to identify patients with the best outcome after NACT.