Detection of recurrence in patients with rectal cancer: PET/CT after abdominoperineal or anterior resection.

Abstract

To assess diagnostic accuracy of combined positron emission tomography (PET) and computed tomography (CT) in detection of pelvic recurrence in patients with rectal cancer who underwent abdominoperineal or anterior resection. Sixty-two patients were enrolled; 37 were men, and 25 were women. Seventeen patients underwent abdominoperineal resection and 45 underwent anterior resection with an anastomosis in the pelvic region before referral for PET/CT. Pelvic sites of fluorine 18 ((18)F) fluorodeoxyglucose (FDG) uptake were rated separately on PET and PET/CT images as benign or malignant on the basis of shape, location, and intensity of (18)F FDG uptake (1-2 = benign and/or physiologic, 3 = equivocal, 4-5 = malignant). Two readers interpreted images in consensus. Altered pelvic anatomy and presence of presacral abnormalities were assessed with CT. Pelvic recurrence was confirmed with histologic analysis or clinical and imaging follow-up. Sensitivity, specificity, positive and negative predictive values, and accuracy of PET and PET/CT in the detection of pelvic recurrence were compared with lesion- and patient-based analyses by using the chi(2) test. Clinical relevance of PET/CT assessment was determined. Of 81 pelvic sites with increased (18)F FDG uptake, 44 were malignant. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for differentiating malignant from benign (18)F FDG uptake in the pelvis were 98%, 96%, 90%, 97%, and 93% for PET/CT and 82%, 65%, 73%, 75%, and 74% for PET, respectively. The most common cause for false-positive interpretation of PET findings was physiologic (18)F FDG uptake in displaced pelvic organs. Presacral CT abnormalities were present in 30 (48%) of 62 patients, and seven (23%) abnormalities were malignant. PET/CT was used to distinguish benign and malignant presacral abnormalities with a sensitivity, specificity, positive predictive value, and negative predictive value of 100%, 96%, 88%, and 100%, respectively. PET/CT findings were clinically relevant in 29 (47%) of 62 patients. PET/CT is an accurate technique in the detection of pelvic recurrence after surgical removal of rectal cancer.