821 Comparison of the Performance of Flexible Sigmoidoscopy Alone and Combined Flexible Sigmoidoscopy and Fecal Immunochemical Test As Colorectal Cancer Screening Tool in 5406 Asymptomatic Chinese Adults
raymond s tang,martin c wong,jessica ching,jeffrey p shum,thomas y lam,s c ng,phyllis p lam,francis k l chan,joseph j y sung,justin c wu
DOI: https://doi.org/10.1016/S0016-5085(15)30538-2
IF: 29.4
2015-01-01
Gastroenterology
Abstract:Introduction While many aspects of colonoscopic surveillance program for colorectal cancer (CRC) in patients with ulcerative colitis (UC) have evolved, it is unknown how the risk of CRC changed over time.We report data from the largest and longest-running UC surveillance programs in the world, revealing a recent trend in CRC risk.Methods A retrospective chartreview of histologically confirmed extensive UC patients enrolled in long-term surveillance was performed.The primary end point was defined as death, colectomy, withdrawal from surveillance, or the census date (January 1, 2013).We compared per-decade CRC incidence rate (IR) over the last forty years -i.e.first (1973)(1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982), second (1983-1992), third (1993-2002) and fourth decade (2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012).Cancer detected in symptom-driven investigations or surgery prior to the next scheduled surveillance was defined as interval CRC.Results A total of 1,375 patients underwent 8,650 colonoscopies (median, 5 per patient; interquartile range (IQR), 3 -8) during 15,234 patient-years (PY) of follow-up (median, 11 years; IQR, 7 -17).Cancer was detected in 72 patients (5.2% of study population), of which 15 were interval CRC (20.8%).While the overall CRC IR had decreased over the first three decades, there was a non-significant increase in the fourth decade (4.9 per 1,000 PY) compared with the third decade (3.7 per 1,000 PY; Chi-squared, P=0.30), which was attributable to a significant increase in IR of Dukes' A or B cancer in the fourth decade (3.2 per 1,000 PY) compared with the third decade (1.3 per 1,000 PY; P=0.045; figure 1).Likewise, there was an increase in IR of dysplasia (17.7 versus 11.7 per 1,000 PY; P=0.01).The chromoendoscopy, which was introduced in our center since 2002, was superior in detecting dysplasia compared with white-light endoscopy (lesion detection rate, 8.4% (92/1,098) versus 4.0% (175/4,373); Chi-squared, P<0.001).The IR of Dukes' C or disseminated cancer showed a linear decrease over the last four decade (2.5 to 1.4 per 1,000 PY in the first decade to last decade; Pearson's correlation, R= -0.99; P=0.01; figure 1).Furthermore, despite the decrease in colectomy rate for dysplasia (linear regression, R= -0.43; P=0.007), IR of interval CRC showed significant reduction over time (2.5 to 0.4 per 1,000 PY in the first decade to last decade; Pearson's correlation, R= -0.99; P=0.007; figure 2).Conclusion The colonoscopic surveillance is increasingly becoming effective in reducing advanced cancer and interval cancer risks, which maybe attributable to the improving lesion detection rate and subsequent decrease in the rate of false negative examinations.Although decreasing colectomy rate allowed more patients to retain their colon, it may have resulted in increased early cancer risk, highlighting the need for close monitoring of patients with dysplasia managed non-surgically.