Recurrence will develop in 30–50% of colorectal cancer (CRC) cases despite apparent clearance following treatment. Carcinoembryonic antigen (CEA) is the only guideline‐recommended blood test for monitoring cases for recurrence, but its sensitivity and specificity are suboptimal. This observational study compared a novel 2‐gene (methylated BCAT1 and IKZF1 DNA) blood test with CEA for detection of recurrent CRC. We conducted a paired comparison of the BCAT1/IKZF1 test with CEA (cut‐off 5 ng/mL) in blood from patients in remission after treatment for primary CRC and undergoing surveillance. Blood collected in the 12 months prior to or 3 months after complete investigational assessment of recurrence status were assayed and the results compared by McNemar's test. Of 397 patients enrolled, 220 underwent satisfactory assessment for recurrence and 122 had blood testing performed within the prescribed period. In 28 cases with recurrent CRC, CEA was positive in 9 (32%; 95% CI 16–52%) compared to 19 (68%; 95% CI 48–84%) positive for methylated BCAT1/IKZF1 (P = 0.002). All samples that were CEA positive were also BCAT1/IKZF1 positive. In 94 patients without clinically detectable recurrence, CEA was positive in 6 (6%, 95% CI 2–13%) and BCAT1/IKZF1 in 12 (13%, 95% CI 7–21%), P = 0.210. The odds ratio of a positive CEA test for recurrence was 6.9 (95% CI 2–22) compared to 14.4 (5–39) for BCAT1/IKZF1. The BCAT1/IKZF1 test was more sensitive for recurrence than CEA and the odds of recurrence given a positive test was twice that of CEA. The BCAT1/IKZF1 test should be further considered for monitoring cases for recurrence.

Reversal of loop ileostomy may be associated with significant morbidity and mortality. Increasing the delay from creation to closure may result in fewer complications.There is an increased risk in older patients with more comorbidity, particularly when the primary procedure is for diverticular disease with significant peritoneal contamination.

No abstract

No abstract

495 Background: Annual computed tomography (CT) scans and periodic carcinoembryonic antigen (CEA) testing are monitoring methods for colorectal cancer (CRC) recurrence. We have previously described a 2-gene blood test for CRC (methylated BCAT1 and IKZF1) that may indicate tumour DNA shedding into the blood. The aim was to compare the 2-gene and CEA blood tests at detecting recurrence in CRC surveillance patients. Methods: Methylated BCAT1/IKZF1and CEA (positive ≥5ng/mL) were measured in patients previously diagnosed with CRC (excluding those undergoing active treatment). McNemar’s test was used for statistical analyses, using clinical findings including CT to diagnose recurrence. Results: At study midpoint, 340 patients were enrolled (64% men, average 64yr at diagnosis), including 59 patients with blood testing done pre- and post-resection, and 105 with CT surveillance scans (median 18 months after primary diagnosis). Following resection (median 2.3 months), 91% of patients who were 2-gene positive prior to treatment showed either no detectable methylated BCAT1/IKZF1(26/35) or significantly reduced levels in blood (6/35). Residual disease was found in two patients (2/3) who remained gene positive post treatment. Recurrence was identified in 30/105 (29%) patients with surveillance CT scans. Of these, 67% and 27% were 2-gene and CEA positive, respectively, with 8 (27%) cases positive by both tests (Table 1, p<0.001). In 13 cases with local recurrence, 54% were 2-gene positive, with only 1 (8%) positive by both tests (p=0.03). In 17 cases with distant recurrence, respective 76% and 41% were 2-gene and CEA positive (p=0.03). In patients with no evident disease, 20% were positive for one test but not the other (2-gene test, 16%; CEA, 4%; p = 0.04). Conclusions: Following resection, most patients had either reduced or no methylated BCAT1/IKZF1 in blood, indicating a correlation with the presence of cancer. Two-gene positivity correlated with local (54%) and distant (76%) recurrence with 2.5-fold more recurrence cases detected than with CEA. The 2-gene test may be better than CEA for recurrence monitoring. [Table: see text]