ESTRO 2025 - Abstract Book

S803

Clinical - Gynaecology

ESTRO 2025

1131

Proffered Paper Normal Tissue Complication Probability models for gastrointestinal toxicity after adjuvant (chemo)radiotherapy for cervical cancer in the PARCER trial Anouk Corbeau 1,2 , Sofia Spampinato 2 , Mayuri Charnalia 3 , Nilesh Ranjan 4 , Sudeep Gupta 5 , Sadhana Kannan 6 , Tapas Dora 7 , Reena Engineer 4 , Amita Maheshwari 8 , Prachi Mittal 4 , Jeevanshu Jain 3 , Ankita Gupta 3 , Subhajit Panda 3 , Jaya Gosh 9 , Prachi Sawant 4 , Stephanie de Boer 1 , Mischa Hoogeman 2,10 , Carien Creutzberg 1 , Remi Nout 2 , Supriya Chopra 3 1 Department of Radiation Oncology, Leiden University Medical Center, Leiden, Netherlands. 2 Department of Radiotherapy, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, Netherlands. 3 Department of Radiation Oncology, Advanced Centre for Treatment, Research and Education in Cancer (ACTREC), Tata Memorial Centre, Homi Bhabha National Institute, Navi Mumbai, India. 4 Department of Radiation Oncology, Tata Memorial Hospital, Tata Memorial Centre, Homi Bhabha National Institute, Mumbai, India. 5 Department of Medical Oncology, Tata Memorial Hospital, Tata Memorial Centre, Homi Bhabha National Institute, Mumbai, India. 6 Epidemiology and Clinical Trials Unit, Advanced Centre for Treatment, Research and Education in Cancer (ACTREC), Tata Memorial Centre, Homi Bhabha National Institute, Navi Mumbai, India. 7 Department of Radiation Oncology, Homi Bhabha Cancer Hospital, Sangrur, Punjab, India. 8 Department of Gynecologic Oncology, Tata Memorial Hospital, Tata Memorial Centre, Homi Bhabha National Institute, Navi Mumbai, India. 9 Department of Medical Oncology, Advanced Centre for Treatment, Research and Education in Cancer (ACTREC), Tata Memorial Centre, Homi Bhabha National Institute, Navi Mumbai, India. 10 Department of Medical Physics & Informatics, HollandPTC, Delft, Netherlands Purpose/Objective: Radiotherapy for gynaecological cancers can cause late gastrointestinal (GI) adverse events. GI normal tissue complication probability (NTCP) models that provide individual risk estimates for gynaecological cancers are not yet available. The aim of this study was to develop NTCP models for late GI toxicity after postoperative (chemo)radiotherapy for cervical cancer. Material/Methods: Individual patient data from the PARCER trial (NCT01279135), randomizing women with cervical cancer between adjuvant intensity-modulated (IMRT) or three-dimensional conformal (3DCRT) radiotherapy, was used. GI toxicity (CTCAEv3.0) was assessed at baseline and follow-up until 5 years. Late toxicity was defined when occurring >90 days after randomization date(1). Two endpoints were investigated: 1) incidence of late moderate-to-severe (≥G2) GI toxicity; 2) persistent late GI toxicity defined according the “Cumulative-Month and Severity Score (C-MOSES)” (>0.70)(2). Predictors included patient- and treatment-related characteristics. Absolute volumes receiving 15Gy (V15Gy), 30Gy (V30Gy), and 40Gy (V40Gy) were considered for individual loops of the Small-Bowel (SB), Large-Bowel (LB), and Small+Large Bowel (Total-Bowel (TB)) and for Bowel-Bag (BB) as whole peritoneal space(3). Ten-fold cross-validation Least Absolute Shrinkage and Selection Operator (LASSO) was performed. Variables selected in ≥8 of the models were included in the final multivariable model. This procedure was performed including either SB, LB, and TB, or BB among the investigated variables. Cox-proportional hazard model was used for ≥G2 GI toxicity and logistic regression for persistent GI toxicity. Model performance was evaluated with C-index or area-under-the-curve (AUC). Results: The analysis included 283 patients (median follow-up: 43 months). The incidence of ≥G2 and persistent late GI toxicity was respectively 29.3% (n=83) and 31.4% (n=89). Table 1 shows the predictor selection by LASSO and the hazard/odds ratios for the final multivariable models. EBRT technique was most predictive for ≥G2 GI toxicity. When including BB, BB V30Gy and V40Gy and type of hysterectomy were selected for persistent GI toxicity. When including SB, LB, and TB, TB V40Gy was most predictive. No other patient- or treatment-related factors were selected by the model. Figure 1 displays the NTCP for persistent GI toxicity with increasing TB V40Gy.