Colonoscopic Screening of Average-Risk Women for Colorectal Neoplasia

ABSTRACT This investigation was performed as a tandem to the Veterans Affairs (VA) Cooperative Study 380, which compared the results of colonoscopy screening with the estimated results of screening with flexible sigmoidoscopy in a mostly male population. The study population for this series consisted of asymptomatic women referred for routine colorectal cancer screening whose colonoscopy was complete to the cecum. Estimates of diagnoses with flexible sigmoidoscopy were made by assuming that advanced lesions in the distal colon (defined as the rectum and the sigmoid colon) would be detected and that small adenomas in the distal colon, which were concurrent with advanced neoplasia in the proximal colon, would have triggered colonoscopy and therefore a correct diagnosis. All asymptomatic women 50 to 79 years of age at average risk for developing colorectal cancer who were screened at one of 4 military centers between July 1999 and December 2002 were eligible for the study. In addition, women with an elevated risk for colorectal cancer (first-degree relative with colon cancer) who were 40 years of age or older were included. Colonoscopy was complete to the cecum in 1463 participants, including 230 (15.7%) who had an elevated risk for colorectal cancer. Neoplastic lesions were identified in 299 women. Seventy-two (4.9%) patients were diagnosed with advanced neoplasia, including 46 women with tubular adenoma 1 cm or more in size, 26 with villous adenoma, 9 who had an adenoma with high-grade dysplasia, and one woman with invasive colorectal cancer. Advanced neoplasia was present in 16 (7.0%) of the 230 high-risk participants. Small or nonadvanced adenomas were seen in 227 women, including 60 high-risk patients. Older women were the most likely to have advanced disease. Nearly 12% (19 of 162) of those 70 to 79 years of age had advanced neoplasia compared with 3.3% (26 of 786) of women in the 50- to 59-year-old group (P = .002). In the 60- to 69-year age group, 5.5% (23 of 420) participants had advanced lesions. Of the 72 women with advanced lesions, 25 (34.7%) had neoplasia in the distal colon, which would have been identified by flexible sigmoidoscopy screening. Forty-seven women positive for advanced neoplasia in the proximal colon had no lesions in the distal colon (65.3%) and would have been missed with sigmoidoscopy. In the entire screening population, 1.7% (25 of 1463) women would have been diagnosed correctly with advanced neoplasias, and 3.2% (47 of 1463) would have been undetected if screening had been performed with flexible sigmoidoscopy. Stratification according to age or family history did not affect the estimated yield of sigmoidoscopy screening. When the distal colon was defined as the rectum and sigmoid colon, a total of 6.3% of participants had neoplastic lesions in the distal colon. Among the 1367 women with no lesions in the distal colon, 47 (3.4%) had advanced disease in the proximal colon and would not have been detected with flexible sigmoidoscopy done to the junction of the sigmoid and descending colon. The incidence of advanced neoplasia in the proximal colon with no lesions in the distal colon was similar for women with and without a family member with colorectal cancer (3.1% and 3.9%, respectively). Among high-risk patients, more advanced lesions were seen in the proximal colon in women with no neoplasia in the distal colon than in women with disease in the distal colon (5.2% vs. 0%, P = .32). When the definition of the distal colon was further expanded to include the descending colon as well as the rectum and sigmoid colon, no neoplasia was found in the distal colon of 1324 (90.6%) women, and small adenomas or advanced neoplasia were seen in 138 (9.4%) patients. Under this definition, 36 (2.7%) women had advanced lesions in the proximal colon but no lesions in the distal colon. Only 3 women (2.2%) had disease in both the proximal and distal colon. The results of this study were compared with the data from the men in the VA Cooperative Study 380. Men under 70 years of age with a negative occult blood test and negative family history who underwent colonoscopy screening were diagnosed with advanced neoplasia nearly twice as often as their female counterparts (15.5% vs. 7.9%). However, in the older age range, 70 to 79 years, the rate of advanced lesions was greater in women than in men (11.8% vs. 10.6%). Overall, the rate of advanced neoplasia was 8.6% in men and 4.5% in women. Screening with flexible sigmoidoscopy would have correctly identified two thirds (126 of 190, 66.3%) of all men with advanced neoplasia. In contrast, sigmoidoscopy screening would not have detected advanced disease in nearly two thirds of all women (64.8%, 10 of 54) who had advanced neoplasia in the proximal colon.