Selective Versus Universal Screening for Lynch Syndrome: A Six-Year Clinical Experience.

BACKGROUND: Lynch syndrome is the most common cause of hereditary colorectal cancer (CRC) and confers increased risk of other cancers. Identification of patients improves morbidity and mortality. Screening tumors for absent mismatch repair (MMR) protein expression by immunohistochemistry (IHC) is a recommended approach. Despite guidelines advocating universal screening, significant variation in clinical practice exists. AIMS/METHODS: A retrospective study of two different IHC-based Lynch syndrome screening protocols at an urban, university hospital was performed. Outcomes from a "selective" screening strategy utilized from August 2007-July 2010 on CRC tumors from patients with high-risk features were compared with a "universal" strategy of screening all CRC tumors from July 2010-August 2013. Positively screened patients were referred for genetic counseling and offered germline testing. RESULTS: A total of 392 patients with CRC were screened: 107 selectively and 285 universally. The prevalence of Lynch syndrome was 3.1 %, with no difference by strategy. There was a trend (p = 0.06) toward fewer universally screened patients agreeing to genetic counseling compared with those selectively screened. Selective criteria failed to identify one of eight cases of Lynch syndrome from the universal group, though the universal strategy screened 166 additional tumors to find this additional patient. CONCLUSIONS: Selective screening for Lynch syndrome has similar outcomes as universal screening in terms of identifying Lynch syndrome, despite screening far fewer patients. In addition, fewer eligible patients in our study agreed to undergo genetic counseling and germline testing than in prior studies. These lower rates may better reflect uptake of these services in clinical practice.

Characteristics of individuals with breast cancer rearrangements in BRCA1 and BRCA2.

BACKGROUND: Large rearrangements in BRCA1 and BRCA2 occur in a small percentage (< 1%) of patients tested for hereditary breast (BC) and ovarian cancer. It is unclear what factors predict BRACAnalysis Large Rearrangement Test (BART) positivity. METHODS: Data from 6 centers were included in this analysis. Individuals with negative Comprehensive BRACAnalysis tested for BART were included. RESULTS: From 1300 individuals, 42 (3.2%) were BART positivity. Factors positively associated with BART positivity were Myriad score, first-degree relatives with BC, infiltrating BC with ductal carcinoma in situ, younger age at BC diagnosis, estrogen receptor-negative BC for both the first and second BC, and Latin American/Caribbean ethnicity. Presence of unilateral BC was inversely associated with BART positivity. Several analyses were performed on the variables available to find the model that best predicts for BART positivity. The BART predictive model, including first BC, ovarian cancer, primary maternal ancestry being Latin America/Caribbean, number of first-degree relatives with BC of 1 or more versus 0, and family history of prostate and pancreatic cancer, had good predictive ability with an area under the curve of 0.77. CONCLUSIONS: Several factors are significantly associated with BART positivity. Among them we have found that Latin American/Caribbean ancestry, Myriad score, first degree relatives with BC, younger age at BC diagnosis, estrogen receptor-negative status of BC, and infiltrating ductal carcinoma with ductal carcinoma in situ features are significantly associated with BART positivity. A BART predictive model may help in a clinical setting.