A Prospective Observational Study of Significance of MRI Assessed Tumor Response for Locally Advanced Rectal Cancer Treated with Longcourse Neoadjuvant Chemoradiation

Abstract

<h3>Purpose/Objective(s)</h3> Use of MRI in comparing the morphological features of rectal cancer pre and post chemo radiation and to correlate the post treatment MRI appearances with the histological findings in resected tumors. <h3>Materials/Methods</h3> After staging MRI, immobilization and planning CT, target volumes for primary, nodal stations and OARs were drawn as per the RTOG atlas. Majority of the plans were 3DCRT such that 95% of PTV volume covered by at least 90% isodose. Patients were prescribed 50.4Gy in 28 fractions over 5 ½ weeks to the PTV along with oral capecitabine 825mg/m2 twice daily during treatment. 6-8 weeks after the treatment completion patients MRI pelvis was done and the radiologist was asked to grade the tumor response according to the Mandard system. Post-surgery the pathologist who was blinded for the MRI report was asked to grade the tumor regression according to Dworak et al. Both the tumor regression grades are compared and the correlation between two was recorded using statistical software. <h3>Results</h3> Based on the mrTRG, we have 3/37(8%) patients showed complete regression, 13/37(35.1%) showed good regression, 17/37(45.9%) showed moderate regression, 4/37(10.8%) showed slight regression. pTRG, we have 3/37(8.1%) patients showed total regression, majority of the patients 17/37(45.9%) showed pTRG 3, 13/37(35.1%) showed pTRG 2 and 4/37(10.8%) people showed slight regression. Nearly 43% (16/37) of the patients showed good response to the therapy, while remaining 57% (21/37) of the patients showed poor response. Nodal staging has prognostic importance, tumors with less nodal burden has good response to NACTRT. 15 out of 16 patients (93.75%) belonging to the favorable response group, (mrTRG 1–2) were also found to have a good pathological tumor regression (pTRG 3–4), the remaining patients being classified as having a pTRG of 2 (35%) or 0–1 (10.8%). Among 21 patients with predicted poor response due to suboptimal tumor regression on MRI scan (mrTRG 3–5), 4 (19.04%) were confirmed to have minimal pathological tumor regression (pTRG 0–1), while 12 (57.14%) and 5 (23.80%) showed a pTRG of 2 and 3–4, respectively. When a dichotomous classification (i.e., pTRG 0–3 vs pTRG 4 and mrTRG 3–5 vs mrTRG 1–2) was used to assess the ability of MRI to predict complete regression, 3 out of 3 patients (100%) with pathological complete response were correctly identified using mrTRG. Sensitivity, specificity of mrTRG were 94.4% (95% CI: 68.8– 96.5), 82.8% (95% CI: 74.5–90.6) respectively. Using Pearson's correlation, we found a correlation between mrTRG and pTRG significant at the level of 0.01(p value = 0.01). Negative correlation of, r² = -0.830 indicates that increasing trend of TRG in MRI correlates to decreasing trend of TRG on postsurgical pathology specimen. <h3>Conclusion</h3> Correlation and comparisons depicted in our study concluded that there is a significant association between mrTRG and pTRG. mrTRG can be used as surrogate marker for the prediction of pTRG.