ESTRO 2025 - Abstract Book

S1682

Clinical - Sarcoma & skin cancer & malignant melanoma

ESTRO 2025

2049

Digital Poster CORRELATION BETWEEN RADIOLOGIC AND PATHOLOGIC TUMOR RESPONSE FOLLOWING PREOPERATIVE CHEMORADIATION IN SOFT TISSUE SARCOMA Mauro Loi 1 , Daniela Greto 1 , Luisa Caprara 1 , Niccolò Bertini 1 , Carolina Orsatti 1 , Andrea Romei 1 , Giulio Frosini 1 , Ilaria Bonaparte 1 , Beatrice Bettazzi 1 , Luca Visani 1 , Marco Banini 1 , Gabriele Simontacchi 1 , Emanuela Olmetto 1 , Vieri Scotti 1 , Pietro Garlatti 1 , Icro Meattini 1 , Isacco Desideri 1 , Monica Mangoni 1 , Domenico A Campanacci 2 , Annarita Palomba 3 , Giuliana Roselli 4 , Lorenzo Livi 1 1 Radiation Oncology, Azienda Ospedaliero Universitaria Careggi, Università di Firenze, Florence, Italy. 2 Orthopaedic Oncology and Reconstructive Surgery, Careggi University Hospital, Florence, Italy. 3 Unit of Histopathology and Molecular Diagnostic, Azienda Ospedaliera Universitaria Careggi, Florence, Italy. 4 Department of Radiology, Azienda Ospedaliera Universitaria Careggi, Florence, Italy Purpose/Objective: In patients affected by Soft Tissue Sarcoma (STS) receiving preoperative radiotherapy with (CRT) or without concurrent chemotherapy (RT) followed by surgery, achievement of pathologic tumor response (pR) is correlated with improved survival. The aim of this study is to assess the correlation between radiologic response (rR) and pR. Material/Methods: Data from consecutive STS patients treated from 2014 to 2023 with preoperative CRT/RT (50 Gy in 25 fractions ± anthracycline-based chemotherapy was administered in selected cases) were retrospectively reviewed. Baseline and restaging MRI were performed respectively within 4 weeks before CRT/RT initiation and 3 weeks after CRT/RT end: rR was evaluated as tumor reduction in its major axis and Objective Response Rate (ORR) according to RECIST criteria. In resected samples, pR was evaluated in terms of necrosis rate, using a ≥90% threshold for near-complete response (n-pCR). Spearman correlation for continuous variables and Kruskall-Wallis test for median comparison was performed. ROC curves were calculated to assess Sensitivity (Se) and Specificity (Sp) of rR at different thresholds for detection of n-pCR. Results: Data from 118 patients (median age: 55 years, range 17-91) were collected. Concurrent chemotherapy was administered in 59 (50%) patients. All patients completed the planned RT treatment and underwent surgery after a median delay of 5 (range 3-12) weeks. Median tumor reduction was 10%, with ORR of 12%. Median necrosis was 80% (range 5-100%): n-pCR occurred in 56 patients (47%) and was statistically correlated with longer delay to surgery (median 6 versus 5 weeks, p=0.04). A moderate correlation between rR and pR was observed (r=-0.23, p=0.013). At ROC analysis, AUC was 0.658: a >15% tumor reduction was identified (Sp: 80%; Se: 52%) for n-pCR. In the subset receiving RT alone (n=59), a stronger correlation between rR and pR was found (r=-0.38, p=0.0041). At ROC analysis, AUC was 0.734: at >15% volume reduction threshold, Sp and Se were respectively 80% and 57%. In one patient experiencing n-pCR, volumetric increase due to intralesional bleeding rather than tumor progression was identified at radiologic review. Conclusion: Assessment of rR is correlated with pR. Due to time-dependent trend toward increased n-pCR following CRT/RT, MRI assessment at different time-points may guide timing of surgery. However RECIST criteria may not be appropriate for this purpose. Careful radiological review is recommended to identify tumor progression.

Keywords: PREOPERATIVE CHEMORADIATION, SOFT TISSUE SARCOMA