Comparison of treatment response assessed by 18F-FDG PET/CT with the histopathological response using tumor regression grading on surgically resected specimen following neoadjuvant chemotherapy in squamous cell carcinoma of esophagus

Abstract

Background For the response assessment after chemotherapy, gold standard is always the histopathological response. However, metabolic response can also guide further treatment. Herein, this study aimed to evaluate metabolic response assessment to neoadjuvant chemotherapy in squamous cell carcinoma esophagus using PET response criteria in solid tumors (PERCIST) criteria with taking histopathological response by tumor regression grading as the standard method. Methods Total fifty-seven patients with squamous cell carcinoma esophagus were enrolled between April 2017 to December 2018 for this prospective study. All patients were undergone for baseline PET scan before going for neoadjuvant chemotherapy. Repeat PET scan was done after neoadjuvant chemotherapy. Operable patients were taken for surgery. Final histological response was assessed by Mandard grading. Three metabolic tools [maximum standardized uptake value (SUVmax), tumor regression grading, PERCIST] were compared. Results The mean SULpeak of the primary lesion was 11.7 ± 5.5 (median, 10.2, range 5.5–31.8). The average percentage change (%Δ) in SUVmax was 42.9 ± 26.3. On histopathology, 5 (13.1%) patients showed complete pathological response, whereas grade II, III, IV and V in 8 (21.1%), 12 (31.6%), 10 (26.3%) and 3 (7.8%) respectively. On comparison of PERCIST with Mandard grading, agreement analysis showed that there was moderate agreement (k, 0.48). %ΔSUV peak change showed a strong correlation with %ΔSUVmax (P = 0.01) and percentage tumor to liver ratio change (P = 0.01). On comparison, these metabolic response tools showed a weak agreement (k, 0.28 with tumor to liver ratio, k, 0.38 with SUVmax). Conclusion After neoadjuvant chemotherapy, 18F-fluorodeoxyglucose PET/CT can help to assess the response and guide the treatment. However, a larger study is warranted to evaluate their correlation with pathological response.