Comparative Evaluation of the Sum of Longest Diameter Measurements as per RECIST 1.1 vs. CECT Based Volumetric Estimation for Response Assessment in Locally Advanced Head and Neck Cancer

Abstract

<h3>Purpose/Objective(s)</h3> Response assessment in solid tumors as per RECIST 1.1 is recommended to be based on the SLD (sum of longest diameter) of primary and/or lymph nodes (RECIST). Alternately, one could use tumor volumes computed on CECT to assess the tumor response (vRECIST). The study proposes to conduct a comparative evaluation of the response categorization as estimated by RECIST vs vRECIST in patients of locally advanced head neck cancers (LAHNC) treated with chemoradiotherapy (CTRT). <h3>Materials/Methods</h3> 38 patients of LAHNC being treated with CTRT alone were included. Baseline tumor dimensions were recorded using SLD criteria of RECIST 1.1 and tumor volumes as estimated on CECT (vRECIST). These were repeated at the mid-or end of CTRT. Two radiation oncologists evaluated these lesions independently for each patient using radiation therapy treatment planning system. The responses by both methods were categorized as per RECIST1.1 criteria into complete response (CR), partial response (PR), stable disease (SD) or progressive disease (PD). Chi-square test was used to compare the response categorization between the two methods. <h3>Results</h3> The 38 patients of primary inoperable tumors of LAHNC included in this prospective study had stages III (n=22), IVA (n=7) and IVB (n=9). They belonged to oral cavity (n=20), oropharynx (n=6), laryngopharynx (n=8), anaplastic cancer of the thyroid (n=1) and unknown primary with neck nodes (n=3). A total of 304 estimates of SLD and volumetric assessments on CECT of primary and/or nodes were carried out and analyzed in these patients by the two observers. Mean duration between two assessments was 5.9 weeks (4–11 weeks). Response categorization as per vRECIST was similar with both observers (CR: 0; PR: 25; SD: 10; PD: 3). However, there was a significant change in the response group categories (RECIST vs vRECIST) in 32/38 (84.2%) patients (p<0.001 for each of the observers) (Table). <h3>Conclusion</h3> Treatment responses in LAHNC during CTRT assessed using SLD for RECIST1.1 did not correlate with the response categories when compared with volumetric estimates as assessed with CECT (vRECIST) and was observer independent. Thus, SLD estimates could result in errors in interpretation of responses during treatment. For accurate evaluation of the treatment response, volumetric estimates should be preferred over uni-dimensional SLD measurements as proposed in RECIST 1.1. Table: Response categories between RECIST vs vRECIST for two observers.