Session Item

Extraskeletal myxoid chondrosarcoma (EMC) is an ultra-rare sarcoma type characterized by a specific gene translocation. Surgery represents the mainstay of treatment, while data on the efficacy of radiation therapy in this disease are limited and conflicting. We report a retrospective multi-institutional series aimed at investigating the role of perioperative radiotherapy in a population of primary localized NR4A3-fused EMC patients treated within seven international sarcoma reference centers.

All adult patients with a diagnosis of primary or relapsed EMC treated from August 1991 to June 2021 with perioperative RT and surgery (RT/S) or surgery (S) alone at the participating institutions were retrospectively included in this analysis. In all patients, diagnosis was confirmed by the presence of NR4A3 translocation. Cases were extracted from prospectively maintained databases. Extensive pathological response was defined as ≤ 50 % of vital tumor cells and / or ≥ 50 % pathological necrosis and / or > 50 % hyalinization. Time to event analyses were conducted. The primary endpoint was local recurrence-free survival, defined as time to first local recurrence event or death from any cause. 

Seventy-eight patients (median age 54 years) were identified and included in the study. Median follow-up was 6.0 (IQ range 2.18-8.81) years. The most common site of origin was lower limb in 61 cases (78%), and 71 (91%) had localized stage disease. Treatments were RT/S in 45 (58%;) and S in 33 (42 %). Median dose was 50 Gy in the neoadjuvant (IQ Range: 50-50) and 60 Gy in the adjuvant setting (IQ Range: 60-66). Among patients receiving RT/S, 18 (54.5%) had an extensive pathological response.

In the unadjusted univariate analysis, patients in the RT/S group had improved local relapse free survival (LRFS) compared to the S group: median time to event 12 years versus 9 years (p=0.02). 5-y LRFS was 86% (95% CI 0.75-0.99) and 68% (95% CI 0.53-0.88) in the RT/S and S groups, respectively. In contrast, DFS as OS were not statistically significant between the two groups (p = 0.76 and p = 0.52, respectively). A significant multivariate model was also found for LRFS. This was significant both overall (Likelihood ratio test: χ²(6) = 46.64, p<0.001) and specifically regarding the prediction of a lower likelihood of events in the RT/S group (HR 0.07, 95%CI 0.02-0.37, p<0.01). Concerning DFS analogous results were found with Likelihood ratio test: χ²(7) = 48.39, p<0.001 and (HR 0.37, 95%CI 0.15-0.93, p<0.01).

In this large retrospective cohort of EMC patients with confirmed NR4A3 mutation, perioperative radiotherapy was associated with lower recurrence free survival and should be considered as part of the treatment armamentarium in a multidisciplinary discussion.