Comparison of Methylated Septin 9 with Carcinoembryonic Antigen in monitoring of Colorectal Cancer Patients after Curative Surgery

Abstract

Background: Detection of circulating methylated septin 9 (mSEPT9) DNA in blood has been recently approved by the FDA as a non-invasive screening test for colorectal cancer (CRC) and is commercially available. As yet, whether mSEPT9 can be used in the post-operative monitoring of CRC patients remains unknown. Aim: To determine the performance of the second generation mSEPT9 assay in the post-operative monitoring of CRC patients and compare its performance with carcinoembryonic antigen (CEA). Methods: We enrolled consecutive CRC patients who were scheduled for surgical resection of CRC. Blood samples were prospectively collected from these patients immediately before surgery and post-operatively at 3-monthly intervals for determination of mSEPT9 and CEA. mSEPT9 was determined by commercially available assay (Epi proColon 2.0; Epigenomics AG, Germany) in an independent laboratory, which was blinded to the patient's diagnosis and timing of blood samples. All samples were analyzed by real time PCR in triplicate and the mSEPT9 assay was considered as positive when more than one PCR reactions was positive. Abnormal CEA was defined as value above 3 ng/ml as recommended by our hospital central laboratory. Results: A total of 90 CRC patients were enrolled in this study (60 men; mean age 67.6 years). The overall sensitivity of plasma mSEPT9 for the detection of CRC before surgery was 75.6% (95% CI, 64.0%-85.2%) and increased from stage I to IV cancers (I: 52.6%, II: 87.5%, III: 84.6%, IV: 100%). There was however no association between positive mSEPT9 with patient’s gender, age and location of CRC at baseline. The positive rate of mSEPT9 was significantly higher than elevated CEA for the detection of CRC in the pre-operative blood samples (75.6% vs 47.7%; P<0.001). For patients with a positive mSEPT9 or CEA at baseline, 57.9% and 40.7% turned negative at 3-month (P = 0.37), respectively. Among the 7 patients with tumor recurrence, 5 (71.4%) of them had persistently positive mSEPT9 or elevated CEA on follow-up. For patients with no clinical evidence of recurrence, mSEPT9 or CEA was negative in the following proportions of patients at 3 months (56.5% vs 67.9%; P =0.44), 6 months (55.8% vs 67.4%; P = 0.30) and 12 months (65.4% vs 73.0%; P =0.58), respectively. Conclusion: The second generation plasma mSEPT9 was more sensitive than CEA for the diagnosis of CRC patients, but may not be as specific as CEA for monitoring after curative resection.