A new online assessment tool developed at Dana–Farber Cancer Institute can help rapidly identify people who should undergo genetic testing for Lynch syndrome.

Mutations in these so–called “mismatch repair” genes impair cells’ ability to heal breaks in their DNA, making genetically damaged cells prone to abnormal growth and the development of cancer. The five genes are MLH1, MSH2, MSH6, PMS2, and EPCAM.

The new online model, developed by Dana–Farber researchers and doctors, is called PREMM5, and enables people to quickly assess their likelihood of carrying one of these genes associated with Lynch syndrome. PREMM5 is described in a report in the Journal of Clinical Oncology.

The report’s senior author is Sapna Syngal, MD, MPH, Director of Research at the Center for Cancer Genetics and Prevention in the Population Sciences Division at Dana–Farber. “Most Lynch syndrome mutation carriers aren’t aware of it because genetic testing rates and knowledge about the condition are much lower than for other cancer–predisposing genes like BRCA1 and BRCA2, which are associated with breast and ovarian cancer,” said Syngal.

The 10 questions in PREMM5 are about the individual’s sex, age, personal history of cancer, and history of certain cancers among relatives. From the answers to these questions, which take less than 2 minutes to complete, the model calculates the probability that the person carries a mutation in any of the five Lynch syndrome genes in their germline – reproductive cells inherited from parents.

The researchers recommend that an individual whose risk is 2.5 percent or greater as identified by PREMM5 model be referred for genetic counseling and testing to determine if he or she has a Lynch syndrome mutation.

Because the PREMM5 tool is free, quick, and simple to use, it could help expand the reach of Lynch syndrome testing, says Syngal.

PREMM5 was developed by analyzing data from 18,734 patients who underwent germline genetic testing for the Lynch mutations, including numerous healthy patients who did not have cancer but had relatives with cancers. The test order forms provided information on the individuals’ age at genetic testing, sex, personal and family cancer histories – including age at diagnosis – and Syngal’s team concurrently analyzed the germline testing results. By correlating these two sets of data, the researchers produced an algorithm to predict the likelihood a person had inherited a mutation in one of the five genes. The model was then validated in a cohort of 1,058 patients with colorectal cancer who were recruited to an institutional sample registry at Dana–Farber. PREMM5 demonstrated it could effectively discriminate between carriers and non–carriers of the Lynch syndrome genes.

With the previous version of the model, called PREM1,2, 6, it was recommended that an individual whose score was 5 percent or greater of carrying Lynch mutations should undergo genetic testing. With PREMM5, a threshold of 2.5 percent or greater was determined to be the most useful. The PREMM5 model can be accessed on the Dana–Farber web site at http://premm.dfci.harvard.edu/.

The first author is Fay Kastrinos, MD, MPH, of Columbia University Medical Center.