Abstract
BACKGROUND: A major challenge in identifying candidates for nonoperative management of locally advanced rectal cancer is predicting pathologic complete response (pCR) following chemoradiation. We evaluated pre- and post-CRT PET-CT imaging to predict pCR and prognosis in this set of patients undergoing resection after neoadjuvant therapy. METHODS: We retrospectively identified patients from 2002 to 2015 with locally advanced rectal cancer who underwent CRT, pre- and post-CRT PET-CT imaging, and resection. Univariate and multivariate analysis was performed and receiver operating characteristic (ROC) curves were generated to evaluate the association of PET-CT characteristics with pCR and survival. ROC curves were generated to define optimal cutoff points for predictive PET-CT characteristics. RESULTS: 125 patients were included. pCR rate was 28%, and follow-up was 48 mo. On multivariable analysis, patients who had a pCR had lower median post-CRT maximal standardized uptake value (SUV(max)) (3.2 versus 5.2, P = 0.009) and higher median %SUV decrease (72 versus 58%, P = 0.009). ROC curves were generated for %SUV(max) decrease (AUC = 0.70) and post-CRT SUV (AUC = 0.69). Post-CRT SUV(max) <4.3 and %SUV(max) decrease of >66% were equally predictive of pCR with a sensitivity of 65%, specificity of 72%, PPV of 44%, and NPV of 86%. Median 5-y overall and relapse-free survival were improved for patients with post-CRT SUV <4.3 (OS: 86 versus 66%, P = 0.01; RFS: 75 versus 52%, P = 0.01) or %SUV decrease of >66% (OS, 82 versus 66%, P = 0.05; RFS, 75 versus 54%, P = 0.01). CONCLUSIONS: PET/CT may be useful in identifying patients who did not achieve pCR, as well as overall survival in patients undergoing CRT for rectal cancer. Patients with a post-CRT SUV of >4.3 should be considered for operative management, as an estimated 86% of these patients will not have a pCR.