Optimal timing for assessment of tumor response to neoadjuvant chemoradiation in patients with rectal cancer: do all patients benefit from waiting longer than 6 weeks?

To estimate the metabolic activity of rectal cancers at 6 and 12 weeks after completion of chemoradiation therapy (CRT) by 2-[fluorine-18] fluoro-2-deoxy-d-glucose-labeled positron emission tomography/computed tomography ([(18)FDG]PET/CT) imaging and correlate with response to CRT.Patients with cT2-4N0-2M0 distal rectal adenocarcinoma treated with long-course neoadjuvant CRT (54 Gy, 5-fluouracil-based) were prospectively studied (ClinicalTrials.org identifier NCT00254683). All patients underwent 3 PET/CT studies (at baseline and 6 and 12 weeks from CRT completion). Clinical assessment was at 12 weeks. Maximal standard uptake value (SUVmax) of the primary tumor was measured and recorded at each PET/CT study after 1 h (early) and 3 h (late) from (18)FDG injection. Patients with an increase in early SUVmax between 6 and 12 weeks were considered "bad" responders and the others as "good" responders.Ninety-one patients were included; 46 patients (51%) were "bad" responders, whereas 45 (49%) patients were "good" responders. "Bad" responders were less likely to develop complete clinical response (6.5% vs. 37.8%, respectively; P=.001), less likely to develop significant histological tumor regression (complete or near-complete pathological response; 16% vs. 45%, respectively; P=.008) and exhibited greater final tumor dimension (4.3 cm vs. 3.3 cm; P=.03). Decrease between early (1 h) and late (3 h) SUVmax at 6-week PET/CT was a significant predictor of "good" response (accuracy of 67%).Patients who developed an increase in SUVmax after 6 weeks were less likely to develop significant tumor downstaging. Early-late SUVmax variation at 6-week PET/CT may help identify these patients and allow tailored selection of CRT-surgery intervals for individual patients.