Positron emission tomography for predicting pathologic response after neoadjuvant chemoradiotherapy for locally advanced rectal cancer.

PURPOSE To investigate whether before and after chemoradiotherapy (CRT) positron emission tomography (PET) predict for pathologic response after preoperative CRT in patients with locally advanced rectal adenocarcinoma. METHODS Thirty-five patients who underwent pre-CRT and post-CRT PET scans before surgery were included. All patients were staged with endoscopic ultrasound or high resolution CT. CRT was given with 50.4 Gy at 1.8 Gy per fraction and concurrent 5-fluorouracil-based chemotherapy. Surgery occurred at a median of 46 days (range, 27 to 112 d) after completing CRT. The maximum standardized uptake value (SUV(max)) and the metabolic tumor volume (MTV) using various minimum SUV thresholds (2, 2.5, 3) on the PET scans (MTV(2.0), MTV(2.5), MTV(3.0)) were determined. Post-CRT PET scans were done 3 to 5 weeks after completion of CRT. Pathologic response was assessed using the tumor regression grade (TRG) scale. Patients with complete or near-complete response (TRG=0 to 1) were considered pathologic responders. The pre-CRT and post-CRT PET scan SUV(max) and MTV values were correlated with TRG. The ΔSUV(max) and ΔMTV were correlated with TRG. RESULTS No correlation was seen with SUV(max) (P=0.99), MTV(2.0) (P=0.73), MTV(2.5) (P=0.73), or MTV(3.0) (P=0.31) on the pre-CRT PET between pathologic responders versus nonresponders. No correlation was noted between SUV(max) (P=0.49), MTV(2.0) (P=0.73), MTV(2.5) (P=0.49), or MTV(3.0) (P=0.31) on the post-CRT PET scan and pathologic response. Finally, the ΔSUV(max) (P=0.32), ΔMTV(2.0) (P=0.99), ΔMTV(2.5) (P=0.31), ΔMTV(3.0) (P=0.31) did not correlate with pathologic response. CONCLUSIONS Changes seen on PET have limited value in predicting for pathologic response of rectal cancer after preoperative neoadjuvant therapy.