The outcome of endometrial carcinoma surveillance by ultrasound scan in women at risk of hereditary nonpolyposis colorectal carcinoma and familial colorectal carcinoma.

Abstract

We read with interest the recent article by Dove-Edwin et al. concerning endometrial carcinoma (EC) screening in women from hereditary nonpolyposis colorectal cancer (HNPCC) and HNPCC-like families.1 There are a number of issues we would like to raise in relation to their conclusions. Women with HNPCC pedigrees are at risk for two gynecologic malignancies: EC and epithelial ovarian carcinoma (EOC). The lifetime risk of EC in women from HNPCC families has been quoted as high as 60%; this equates to their risk of colorectal carcinoma (CRC), which is lower for women than for men in these families.2, 3 For EOC, the cumulative lifetime risk in the general population is 1%; in HNPCC families, this risk is increased to 9%. To date, there is no proven satisfactory screening modality for either of these gynecologic malignancies in the general population or in high-risk groups. The most studied form of screening has been for EOC in the general population and in breast/ovarian cancer families with or without identified BRCA1/BRCA2 mutations.4, 5 A combination of transvaginal ultrasound with Doppler studies (TV USS) and the tumor marker CA125 appears to be the best strategy, but this approach lacks sensitivity and specificity in the premenopausal population and, to date, has not been shown to lead to a reduction in mortality in the postmenopausal population. There are currently two large randomized controlled trials to assess its role, and until then, the consensus opinion is for women in defined high-risk groups to receive annual screening with TV USS and CA125 with an understanding of its limitations. This has significant resource implications in the United Kingdom, where radiology services already are overstretched and are unable to meet their commitment for diagnostic services, let alone screening services. The attention to finding a satisfactory method of screening has been greater for EOC than for EC because of the dramatically different survival rates for the two diseases. EOC typically presents as Stage III disease, and the overall 5-year survival rate is < 30%. In contrast, EC typically presents as Stage I disease with the symptom of postmenopausal bleeding. However, in the general population, one in four women diagnosed with EC will die from their disease. This differs from the study, quoted by Dove-Edwin et al., of 125 women in HNPCC families; that study yielded a mortality rate of 12% due to EC.6 EC also differs from EOC in that there is a loosely defined, premalignant phase of simple/complex hyperplasia with or without atypia. There is also a move to define more carefully a morphologic description of endometrial intraepithelial neoplasia. This is a more quantitative, reproducible definition supported with molecular evidence of monoclonality.7 Evidence of premalignant endometrial changes allows the choice of hormonal manipulation or less radical surgery than required for EC. The major symptom of EC in the general population is postmenopausal bleeding. A combination of ultrasound and outpatient endometrial sampling is a useful tool in excluding and diagnosing the disease. The gold standard investigation remains hysteroscopy and targeted biopsy. This can be done in an outpatient setting. Outpatient hysteroscopy is tolerated by > 90% of women, even in the postmenopausal age range.8 However, in women from HNPCC families, approximately 25% of ECs will be found in premenopausal women.2 The value of ultrasonography alone in the detection of endometrial lesions in premenopausal women with or without symptoms is limited. In symptomatic women, its sensitivity is improved by intrauterine infusion of saline solution and combining it with outpatient endometrial biopsy.9 The median age at first scan in the study by Dove-Edwin et al. was 42 years, suggesting that the majority of women in their study were in the premenopausal age range. We dispute the conclusion that outpatient endometrial biopsy is an inappropriate screening tool because of discomfort, especially compared with methods of screening for CRC, such as colonoscopy or barium investigations. We also argue that the most appropriate method of screening for EC may be a combination of outpatient hysteroscopy and endometrial biopsy. We agree that a randomized trial of USS screening versus no screening in women from HNPCC families is unethical. The most appropriate trial should have three arms: 1) annual TV USS and CA125; 2) annual TV USS, CA125, and outpatient endometrial biopsy; and 3) annual TV USS, CA125, outpatient hysteroscopy, and endometrial biopsy. Due to the small numbers involved, such a trial would need to be multicentered. We currently are using the third strategy for a one-stop screening clinic. We are assessing both the acceptability and patient satisfaction with the process. We look forward to collaborating with other centers in working toward a satisfactory screening modality for gynecologic malignancies in women in HNPCC families.