Patients With Carcinoma Of Unknown Primary And "Colon Cancer Profile": Clinicopathologic Features And Survival Data

4130 Background: We have previously proposed a “colon cancer profile” (CCP-CUP) favorable subset. CCP-CUP is ~ 8% of all CUP with features resembling lower gastroinestional cancers. Distinguishing this entity from other CUP pts is of significance given the progress in the treatment of metastatic colorectal cancer (CRC). Additionally, emerging data suggests a high level of agreement between histology+IHC and tissue of origin profiling for CCP-CUP. Methods: The retrospective cohort includes 74 pts from MD Anderson (50) and Sarah Cannon (24) Cancer Centers from 2004 -2010. Pts with CDX2+ tumors and pathology suggesting a "GI profile" were chosen. All pts met the definition of CUP and most had a negative colonoscopy. Results: 2 cohorts were created - Cohort 1 (34 pts), labeled “consistent with lower GI profile” [CDX2+, CK20+, CK7-] and Cohort 2 (40 pts), labeled “probable lower GI profile” [CDX2+, irrespective of CK7/CK20 status]. Most pts had a good PS; 58% women, median age 59 yrs; 20 (27%) pts had ascites on presentation and the predominant sites of metastases included liver (30%), carcinomatosis (50%), and nodes (51%). 53 pts received first line CRC regimens (FOLFOX or FOLFIRI based), 15 pts received gemcitabine or taxane based and 3 ‘other’ regimens. OS was 37 mo (C.I 22- 52). 6 of these were "outliers" (Stage 4 NED or indolent pathology). Excluding these, cohorts 1 and 2 had 32 and 36 pts - their OS were 37 and 21 mo respectively. There was no difference in OS of pts with or without ascites on presentation. Kras data was available for 17; 12 were Kras mutant. On multivariate analysis, the factors found to negatively influence survival were age, PS (of 2) and 3 or more sites of disease. Conclusions: Survival of IHC defined CCP-CUP pts (which may include colorectal, appendiceal and small bowel profile) exceeds historical control and illustrates a new "favorable" subset. IHC is not without its limitations – nonetheless, pts with CDX2 + tumors and CUP should undergo evaluation for GI cancers and likely best served with an armamentarium of drugs used for CRC. Since carcinomatosis and liver mets are predominant sites, all patients with abdominal CUP should undergo optimal IHC (CDX-2, CK7, and CK20) testing to rule out a CCP-CUP.