BACKGROUND: Although gastroenterologists are asked to perform colonoscopy in patients with metastatic cancer of unknown primary (MCUP), studies evaluating this practice are lacking. AIM: To determine the yield and cost of colonoscopy in patients referred for colonoscopy with an indication of MCUP. METHODS: We prospectively and retrospectively assessed colonoscopies performed from 2000 to 2011 at a county, a university, and a Veterans Administration medical centre to identify patients referred for colonoscopy for the indication of MCUP. Exclusion criteria included overt or occult bleeding, iron-deficiency anaemia, familial-colon-cancer syndrome, prior colon cancer, imaging suggesting colorectal lesion, and palpable rectal mass. Outcomes were the number of primary colon cancers and costs based on 2012 Medicare reimbursements. RESULTS: Two (1%) of the 160 patients meeting enrollment criteria had a primary colon cancer identified, and both died within 1 month after diagnosis without receiving therapy targeted at colon cancer. One patient without colon cancer had a perforation because of colonoscopy, which required surgery and colostomy. The cost of a strategy of routinely performing colonoscopy in patients referred with MCUP was $84 736 per colon primary identified. CONCLUSIONS: Primary colon cancer was rarely identified at colonoscopy in patients with MCUP and no standard indications for diagnostic colonoscopy. Furthermore, the cost to diagnose one additional colon primary was very high. Those with colon cancer had advanced disease and were unable to benefit from targeted therapy. Routine colonoscopy for MCUP cannot be recommended at present.
BACKGROUND: Although gastroenterologists are asked to perform colonoscopy in patients with metastatic cancer of unknown primary (MCUP), studies evaluating this practice are lacking. AIM: To determine the yield and cost of colonoscopy in patients referred for colonoscopy with an indication of MCUP. METHODS: We prospectively and retrospectively assessed colonoscopies performed from 2000 to 2011 at a county, a university, and a Veterans Administration medical centre to identify patients referred for colonoscopy for the indication of MCUP. Exclusion criteria included overt or occult bleeding, iron-deficiency anaemia, familial-colon-cancer syndrome, prior colon cancer, imaging suggesting colorectal lesion, and palpable rectal mass. Outcomes were the number of primary colon cancers and costs based on 2012 Medicare reimbursements. RESULTS: Two (1%) of the 160 patients meeting enrollment criteria had a primary colon cancer identified, and both died within 1 month after diagnosis without receiving therapy targeted at colon cancer. One patient without colon cancer had a perforation because of colonoscopy, which required surgery and colostomy. The cost of a strategy of routinely performing colonoscopy in patients referred with MCUP was $84 736 per colon primary identified. CONCLUSIONS:Primary colon cancer was rarely identified at colonoscopy in patients with MCUP and no standard indications for diagnostic colonoscopy. Furthermore, the cost to diagnose one additional colon primary was very high. Those with colon cancer had advanced disease and were unable to benefit from targeted therapy. Routine colonoscopy for MCUP cannot be recommended at present.