Molecular screening of all colorectal tumors diagnosed before age 50 years followed by genetic testing efficiently identifies Lynch syndrome cases

In this issue, Schofield et al.1 provide population-based evidence essential for the design of cost-effective protocols to translate current genetic research into clinical practice. Their paper studies Lynch syndrome, a strong predisposition to colorectal and other cancers due to germline mutations in DNA mismatch repair (MMR) genes. People with this genetic predisposition can benefit from specialized clinical management. Screening the rectum and large bowel of unaffected carriers followed by prophylactic polypectomy has been estimated to reduce risk of colorectal cancer by 56%.2 Also, knowledge of non-carrier status within a carrier family can release some relatives from unnecessary screening. Screening and early identification of mutation carriers has the potential to be a highly cost-effective way to reduce the burden of colorectal cancer. To achieve this, it is important to have cheap and acceptable ways to identify mutation carriers. Mutation testing is complex and currently expensive because there is a considerable amount of genetic material to test in the encoding and regulating regions of the 4 genes, and many different types of genetic mutations can lead to MMR dysfunction.3 Therefore, an inexpensive but highly sensitive and specific preliminary test to identify likely carriers is needed. Molecular screening of newly diagnosed colorectal cancers followed by genetic testing is one way of doing this, and Schofield et al.1 have shown just how much the efficiency of this approach depends on the age at diagnosis of the tested cases. It appears that almost all Lynch syndrome colorectal tumors display a high degree of microsatellite instability (MSI) and fail to express the protein product of at least 1 MMR gene—the latter detected by immunohistochemical (IHC) testing.4 Therefore, either or both of MSI or IHC can be used as preliminary tests to identify likely Lynch syndrome cases. These molecular tumor tests are more sensitive (100% for IHC and 72% for MSI) than standard family history criteria for identifying germline mutation carriers in people with colorectal cancer diagnosed before age 45 years.5 In the context of population-based screening for Lynch syndrome cases, this set of molecular characteristics has been termed the Melbourne criteria.6 It is not known how well these tests perform for cases with older ages at diagnosis. They are likely to be less sensitive and specific because the later the onset of disease the more likely it is that MMR function will be lost by somatic mutation or age-related methylation of a MMR gene. Schofield et al.1 have now presented data which make it possible for policy-makers to decide, based on their local conditions and funding constraints, the most appropriate age at diagnosis threshold for their jurisdiction, below which all colorectal cancers should be tested for MSI as a preliminary screen for germline MMR mutation testing. Schofield et al.1 studied a population-based sample of colorectal cancer cases diagnosed before the age of 60 years in Western Australia from 2000 to 2006. They tested as many tumors as possible for MSI at the BAT-26 marker and called tumors MSI+ if they exhibited MSI at this single marker. This identified 105 MSI+ tumors, 7 of which were subsequently found to have the V600E BRAF mutation. This left 98 "red flag" cases who were considered to be likely MMR gene mutation carriers and for whom they tried to determine mutation status. Using the data presented in Supplementary File 2 and Table 2 of Schofield et al.,1 and assuming that the untested cases were in essence at random within each age at onset category, we generated figures showing the relationship between MSI and germline mutation by age category. Figure 1 shows how critical age at onset is in this whole story. The proportion of cases with MSI+ tumors decline from 18% in the under 35 years age-at-diagnosis group to 5% in the 50–59 years group; see Figure 1a. Also, the positive predictive value of this test drops dramatically after age 50 years: the proportion of MSI+ cases who are carriers is relatively stable in the under 50 years cases, averaging 78%, but is then 40% in the next 5 years and 9% for 55–59 years; see Figure 1b. Proportion of: (a) tumours which are MSI+, (b) MSI+ cases who are carriers, (c) cases who are carriers, assuming all carriers are MSI+, and (d)carriers diagnosed at various ages less than 60 years. These data also allow the calculation of a population-based estimate of the age-specific contribution of Lynch syndrome to colorectal cancer; see Figure 1c. The proportion of Lynch cases among all colorectal cases changes dramatically from about 1 in 7 prior to age 40 years, halves in the next decade to about 1 in 14, then halves again in the next 5 years to 1 in 35 for the 50–54 year age group and is many times less in the next 5 years. These estimates are likely to be slight under-estimates because: (i) mutations had not been sought among cases not MSI+; (ii) only 1 marker of MSI was used and (iii) existing mutations may not have been identified by the testing. Nevertheless, they accord with what little information there is on this issue in the literature: mutations in MLH1 and MSH2 have been estimated to account for 3.6% of colorectal cancers diagnosed before age 55 years,7 12% of colorectal cancers diagnosed before age 45 years5 and 28% of colorectal cancers diagnosed before age 30 years.8 How can the data from Schofield et al.1 and others be used to decide which age-group of colorectal cancers should be screened? Decisions will depend on local issues, including testing facilities and financial resources, but the data summarized here will be critical in helping make that decision. Figure 1d shows that, of all Lynch syndrome cases diagnosed before age 60 years, 80% were diagnosed before age 50 years. Therefore, this population-based study has provided evidence that molecular screening and genetic testing all colorectal tumors diagnosed before age 50 years will identify most Lynch syndrome cases.