Early regression index (ERI) on MR images as response predictor in esophageal cancer treated with neoadjuvant chemo-radiotherapy: interim analysis of the prospective ESCAPE trial

Purpose: The early regression index (ERI) predicts treatment response in rectal cancer patients. Aim of current study was to prospectively assess tumor response to neoadjuvant chemo-radiotherapy (nCRT) of locally advanced esophageal cancer using ERI, based on MRI. Material and methods: From January 2020 to May 2023, 30 patients with esophageal cancer were enrolled in a prospective study (ESCAPE). PET-MRI was performed: i) before nCRT (tpre); ii) at mid-radiotherapy, tmid; iii) after nCRT, 2-6 weeks before surgery (tpost); nCRT delivered 41.4Gy/23fr with concurrent carboplatin and paclitaxel. For patients that skipped surgery, complete clinical response (cCR) was assessed if patients showed no local relapse after 18 months; patients with pathological complete response (pCR) or with cCR were considered as complete responders (pCR+cCR). GTV volumes were delineated by two observers (Vpre, Vmid, Vpost) on T2w MRI: ERI and other volume regression parameters at tmid and tpost were tested as predictors of pCR+cCR. Results: Complete data of 25 patients were available at the time of the analysis: 3/25 with complete response at imaging refused surgery and 2/3 were cCR; in total, 10/25 patients showed pCR+cCR (pCR=8/22). Both ERImid and ERIpost classified pCR+cCR patients, with ERImid showing better performance (AUC:0.78, p=0.014): A two-variable logistic model combining ERImid and Vpre improved performances (AUC:0.93, p<0.0001). Inter-observer variability in contouring GTV did not affect the results. Conclusions: Despite the limited numbers, interim analysis of ESCAPE study suggests ERI as a potential predictor of complete response after nCRT for esophageal cancer. Further validation on larger populations is warranted.