Restaging after neoadjuvant chemoradiotherapy for rectal adenocarcinoma: role of F18-FDG PET

Abstract

Multimodality treatment of loco-regional advanced rectal cancer has demonstrated to improve local control and overall survival. Proctoscopy, digital rectal examination (DRE), computer tomography (CT), endorectal ultrasound (ERUS), and magnetic resonance imaging (MRI) cannot correctly detect downstaging in rectal tumors after chemo radiation therapy (CRT). New imaging techniques, like 18F-FDG PET, may play some role in predicting the pathologic response to CRT before surgical resection. Aim of the present study was to further investigate the accuracy and predictive value of 18F-FDG PET in a large series of patients with rectal cancer treated with preoperative intensified CRT. Between January 2000 and December 2003, 81 patients with histologically proven adenocarcinoma in clinical stage II–III disease, according to criteria of TNM classification, were included in this study. All patients were submitted to diagnostic staging workup with DRE, proctoscopy with biopsy, ERUS, CT scan of the abdomen and pelvis or pelvic MRI plus liver ultrasonography, coloscopy or barium colonic enema. One month later the end of CRT all patients were submitted to diagnostic restaging work-up (DRW) and 18F-FDG PET. Surgery was performed 8–9 weeks after the end of CRT and pathologic stage was defined. Moreover a pathologic assessment of tumor regression was made with tumor regression grade score (TRG). PET correctly identified 22/28 (79% specificity) patients with complete pathologic response (pCR). However, sensitivity was 45% (24/53) while PPV, and NPV were equal to 77 and 43%, respectively. Total PET accuracy rate was 56%. PET sensitivity increased from 45 to 56% if the end-point was pCR, or TRG score, respectively. The best correlation was found between PET findings and pathologic stage (P <0.01) or TRG score (P <0.01). The accurate identification of rectal cancer patients with major pathological response after preoperative CRT further supports the necessity of designing prospective studies with new and more accurate was imaging technologies with the main object of offering conservative treatment in responder patients.