ESTRO 37 Abstract book

S92

ESTRO 37

Purpose or Objective The role of neoadjuvant chemoradiotherapy (CRT) in the management of resectable esophageal cancer is growing. The advantage of nCRT followed by a 6-8-week period before surgery is that this approach can induce significant tumor regression and downstaging, which leads to an increased rate of microscopically radical resections, decrease in local recurrences, a pathologic complete response in 28-34% of the patients and an overall survival benefit. Accurate identification of non-responders early during nCRT would allow individualized decision making in continuation or discontinuation of nCRT. Furthermore, a tool is desirable to accurately assess the treatment response after nCRT to identify patients with a complete response. DW-MRI imaging during the first 2-3 weeks of nCRT has shown promising results in the prediction of pathological response. However, the optimal timing of scanning for pathological response prediction is unclear. The purpose of this study was to assess and evaluate the optimal time of DW-MRI scanning during nCRT to predict pathological response to nCRT with pathology of the resected specimen as reference standard. Material and Methods Patients with esophageal cancer who underwent nCRT according to the CROSS regimen (weekly administration of carboplatin/paclitaxel combined with 41.4Gy in 23 fractions), followed by an esophagectomy were eligible for inclusion. Patients underwent 6 sequential MRI scans. One MRI scan was performed in the week prior to nCRT in addition to the conventional diagnostic work-up. Subsequently, weekly MRI scans were performed during nCRT (Figure 1). The median tumor apparent diffusion coefficient (ADC) was determined at these six time points. Response to nCRT was reported using the Mandard system based on histopathologic evaluation of the resection specimen. Results To date, a total of 48 MRI scans of 8 patients with esophageal cancer were analyzed in this study. Most patients (5, 62.5%) had an adenocarcinoma, the remaining 3 patients (37.5%) had a squamous cell carcinoma. Mandard 1 was found at histopathological evaluation of the resection specimen in 3 patients (37.5%), Mandard 2 in 1 patient (12.5%), Mandard 3 and 4 both in 2 patients (25%). The percentage of change in tumor ADC values during nCRT with respect to the first MRI scan prior to treatment with nCRT is depicted in Figure 2. As highlighted in this figure, the difference in the change in tumor ADC values between good responders (Mandard 1 and 2) and to poor responders (Mandard 3 and 4) was most prominent in week 3 of nCRT (after 12-14 fractions). However, due to the limited number of patients enrolled, no statistical analyses could be performed to confirm these descriptive results. Figure 1: Study design

stratification was performed by dichotomizing patients into high- and low-risk groups based on mid-treatment response as well as p-16-status and smoking history, and compared in terms of loco-regional control, progression- free survival and overall survival. Results Out of the 96 patients, 74% had stage IVa or IVb (AJCC 7th ed) and 57% had p16 positive disease. With a median follow-up of 34 months, 14 patients experienced LRR. The median (inter-quartile range) reduction in total tumor volume was 18.7% (8.4% - 30.9%). Reduction in total tumor volume > median is an independent predictor of LRR (HR 0.22, 95% CI: 0.05 – 0.89, p=0.020), and the reduction in primary tumor volume is an even stronger predictor (HR 0.11, 95% CI: 0.02 – 0.57, p=0.002). For patients with p-16 negative disease, the three-year locoregional control rate for patients with total tumor reduction ≤ median was 55.3%, compared to 90.9% for >median reduction (p=0.035). For those with p-16 positive disease, the three-year locoregional control rate for patients with tumor reduction ≤ median (21.5%) for that group was 85.6%, compared to 95.0 % for p-16 positive > median reduction (p=0.30. Stratifying patients into a high-risk group with reduction in total tumor volume at mid-treatment ≤median, p-16 negative status, and smoking status >10 pack years, compared to a low- risk group without these factors showed a clear separation in Kaplan-Meier curves with actuarial 3-year loco-regional control, progression-free survival and overall survival rates for the high-risk patients of 45.7%, 38.2%, 71.8% compared to 90.7%, 70.6%, 89.8% for the low-risk patients, respectively (p≤0.021 for all).

Conclusion Our study shows that early response assessment based on mid-treatment CT is an independent predictor of LRR, especially for those with HPV negative disease and can be used to effectively distinguish high-risk and low-risk patients, allowing for risk-adaptive treatment stratification at the midway point.

Figure 2:

OC-0178 Optimal timing for tumor response assessment to nCRT with MRI in patients with esophageal cancer A.S. Borggreve 1 , S.E. Heethuis 2 , L. Goense 1 , P.S.N. Van Rossum 2 , A.L.H.M.W. Van Lier 2 , R. Van Hillegersberg 3 , J.P. Ruurda 3 , J.J.W. Lagendijk 2 , S. Mook 2 , G.J. Meijer 2 1 UMC Utrecht, Radiation Oncology and Surgical Oncology, Utrecht, The Netherlands 2 UMC Utrecht, Radiation Oncology, Utrecht, The Netherlands 3 UMC Utrecht, Surgical Oncology, Utrecht, The Netherlands