1581 Background: Patients considered for liver resection (LR) for hepatic metastases from metastatic colorectal cancer (mCRC) have an 18FDG-PET CT scan (PET) to exclude extrahepatic disease (EHD). The prognostic significance of an equivocal PET on overall survival (OS) for patients who proceed to LR is not entirely clear. The aim of the study is to compare OS for patients with equivocal PET prior to LR to those with a PET negative for EHD. Methods: The South Australian Metastatic Colorectal Cancer Registry collects data for mCRC patients diagnosed after February 1, 2006. Patients were included if they had LR and a PET prior to LR. PETs were coded as no EHD and possible EHD. The Cox proportional hazard model was applied to analyse the outcome of patients with an equivocal PET for EHD on OS, adjusting for possible confounders. Results: Of the 2,480 patients on the registry, 273 had had LR. Of these, 183 (67.0%) had a PET prior to LR, with 137 having no EHD and 46 having possible EHD. The no EHD and possible EHD groups were well balanced for patient, tumour and treatment characteristics – mean age: 66.7 yrs-vs-68.4 yrs, male gender: 61.3%-vs-63.0%, KRAS wildtype: 11.0%-vs-16.3%, stage IV disease at initial diagnosis: 49.6%-vs-54.3%, colonic primary: 74.4%-vs-65.2%, one LR: 82.5%-vs-89.1%, one line of chemotherapy: 52.4%-vs-48.6% and well-moderate tumour differentiation: 85.7%-vs-86.4%. The median follow-up was 32.9 months for no EHD and 33.6 months for possible EHD (P-value = 0.84). The OS for no EHD compared with possible EHD at 1-year was 98.5%-vs-93.5%, at 2-years was 87.6%-vs-88.0%, and at 5-years was 61.5%-vs-59.4%. The unadjusted hazard ratio for OS was 1.22 (95% CI 0.64–2.34, P-value = 0.54) for possible EHD. On adjustment for age, gender, stage at diagnosis, primary site, number of LRs, lines of chemotherapy and tumour differentiation, the hazard ratio remained non-significant; however lower (HR=0.76 (95% CI 0.37–1.59, P-value = 0.47)), for possible EHD. Conclusions: A PET was only performed in 67.0% of patients who had LR for mCRC. There was no difference in OS between patients with no EHD and possible EHD on PET who proceed to LR.
Serial FDG–PET/CT for early outcome prediction in patients with metastatic colorectal cancer undergoing chemotherapy
