Abstract
The present approach at our institution for the treatment of patients with colorectal (CRC) cancer and with liver metastases planned for metastasectomy is the neoadjuvant administration of Bevacizumab with Irinotecan based therapy. Metabolic imaging of tumor viability with 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET), and simultaneous anatomic localization provided by low-dose non-enhanced computed tomography (CT), can be obtained in a combined modality FDG-PET/CT scan. The purpose of this study was to evaluate the possible contribution of FDG-PET/CT as a surrogate marker to evaluate treatment response of liver metastases in vivo. This is a retrospective evaluation of 18F-FDG PET and CT findings in the first seven consecutive patients. FDG-PET/CT scans were performed before the start of the neoadjuvant and after four cycles of therapy, just prior to surgery. Results were compared to concurrent contrast-enhanced CT, when required, and pathology. Response to treatment was determined according to RECIST size criteria obtained from data from thin (3–5mm) slice CT, and changes in uptake of 18F-FDG uptake on PET.
A total of 20 liver lesions were evaluated in seven patients. Overall, 6/7 patients had favorable response to treatment, and only one had progression of disease. One patient was found to be inoperable at surgery. Biopsy was obtained in 1/4 lesions in this patient, while pathology was unable for the remaining three lesions. As such, pathologic validation of findings was available for 17/20 lesions. Complete response (CR) was evident on FDG-PET in 10/17 (58%) lesions, whereas only 4/17(23%) were deemed CR by CT. Similarly, only 1/17 (6%) lesion appeared stable by FDG-PET criteria, whereas three (18%) were termed stable disease (SD) according to size on CT.
FDG-PET findings correlated better than CT with pathology, and were more indicative of pathology. Overall PET/CT correctly predicted necrosis at pathology in 70% vs. 35% by CT. Our results suggest that 18F-FDG PET may be instrumental for predicting the pathologic response to Bevacizumab based therapy.