**3. Lynch syndrome: Mechanisms of carcinogenesis**

Six variants of the mismatch repair gene (MMR) have been cloned: *MSH2* (MutS homolog 2, chromosome 2p16), *MLH1* (MutL homolog 1, chromosome 3p21), *MSH3* (MutS homolog 3, interacts with *MLH1*), *MSH6* (MutS homolog 6, chromosome 2p16), *PMS1* (postmeiotic segregation 1, chromosome 2q31) and *PMS2* (postmeiotic segregation 2, chromosome 7p22) (Koessler et al., 2008). However, germline mutation analysis in four of these DNA-MMR genes (*MLH1, MSH2, MSH6*, and *PMS2*) is confirmatory diagnosis for LS (Hampel et al., 2005). A fifth and most recently identified gene, *EPCAM* (previously TACSTD1), is not a mismatch repair gene; however, large deletions in the 3' end in the upstream *EPCAM* gene affect *MSH2*. This occurs by transcriptional read-through into and subsequent epigenetic silencing of its downstream neighbor, *MSH2*, resulting in the LS phenotype (Ligtenberg et al., 2009).

The role of the MMR machinery is to maintain genomic integrity by correcting base-pair and small insertion-deletion mismatches that are generated during DNA replication. Two heterodimeric protein complexes, MutS-α and MutS-β, recognize the mismatch. MutS-α is a heterodimer of *MSH2* and *MSH6* proteins and MutS-β is an *MSH2/MSH3* heterodimer, see Figure 1 (Masuda et al., 2011). Either MutS-α and MutS-β heterodimers can recognize insertion/deletion loops with more than two bases, but MutS-α preferentially recognizes single base-pair mismatches or, one or two base pair insertion-deletion loops (Koessler et al., 2008). The repair components of the MMR machinery involve three other heterodimer pairs: MutL-α (*MLH1/PMS2*), MutL-β (*MLH1/PMS1*), and MutL-γ (*MLH1/MLH3*).

(Reproduced from Masuda et al., 2011).

Fig. 1. The DNA Mismatch Repair (MMR) machinery in humans.

In general, affected patients with LS carry a germline mutation in one allele of a MMR gene and acquire a second mutation within the tumor. Common mechanisms of the "second hit" include allele inactivation by mutation, loss of heterozygosity, or promoter hypermethylation leading to epigenetic silencing. Biallelic inactivation of MMR genes results in genomic instability due to failure in the repair of base pair mismatches that occur commonly during DNA replication (approximately 1 in 106 base pairs). DNA mismatches commonly occur in regions of tandem repeats of short DNA sequences called microsatellites that make up about 3% of human DNA (Baudhuin et al., 2005). Normally, the MMR machinery corrects errors in microsatellites, but mutations in the MMR genes in tumors cause expansion or contraction of these regions compared to normal tissue. These genetic alterations in microsatellite length are termed "microsatellite instability" (MSI) and are the molecular signature of LS-associated cancers (Lynch et al., 2009). Further, the increased mutation rate that results from MMR loss leads to alterations in nucleotide repeats in many other pathways; those that control cell growth, regulate cell death, and in the MMR genes themselves. Together, this accumulation of mutations drives the carcinogenetic process in LS.

#### **3.1 MMR genes and the risk of endometrial cancer**

The range of cancer risks in LS varies depending on the MMR gene involved. Approximately 70-80% of the clinical features of LS are accounted for by *MLH1* and *MSH2* mutations. Families with *MSH6* and *PMS2* mutations appear to have an attenuated cancer phenotype, presenting with a later age of diagnosis and a lower penetrance than *MLH1* and *MSH2*. *MSH6* may account for up to 15% and *PMS2* for up to 3-15% of all identified LS mutations (Hampel et al., 2005; Niessen et al., 2009). Recently, *EPCAM* has been thought to account for approximately 1-3% of LS mutations (Kuiper et al., 2011).

Endometrial cancer risk per MMR gene is as follows:

*MLH1 and MSH2:* Endometrial cancer in patients with *MLH1* and *MSH2* mutations often occur before the age of 50. The risk in *MLH1* and *MSH2* carriers is up to 20% by age 50 and up to 60% by age 70 according to some studies (Aarnio et al., 1999; Lynch et al., 2009). However, the diagnosis of EC after the age of 50 should still raise concern for LS if there is a positive family history.

*MSH6***:** *MSH6* mutation carriers appear to have the highest risk for endometrial cancer (up to 71%) of the MMR genes, higher than that of colorectal cancer (CRC) (Hendricks et al., 2004). The average age of onset of EC in *MSH6* mutation-positive individuals is 54 years. One study identified a somewhat lower risk for endometrial cancer in *MSH6* mutation carriers, however the risk was significantly increased above the general population, and appeared to be higher than the risk for CRC in women with LS. They reported a risk for endometrial cancer to age 70 of 26% (95% CI: 18-36%) and risk to age 80 of 44% (95% CI: 30- 58%) (Baglietto et al., 2009). These reports suggest that if a woman carries an *MLH1*, *MSH2*  or *MSH6* mutation, her risk of EC may be even higher than her risk of CRC.

*PMS2***:** One large series of *PMS2* carriers found the incidence of EC to be 7.5-fold higher than expected in the general population. This translates to a 15% risk to age 70 (Senter et al., 2008).

*EPCAM***:** The clinical features of *EPCAM*/*TACSTD1* mutations as a cause of Lynch syndrome is still being defined. Recent studies evaluating *EPCAM* 3'-end mutation carriers for their clinical phenotype found the risk for EC was dependent upon the type and size of

In general, affected patients with LS carry a germline mutation in one allele of a MMR gene and acquire a second mutation within the tumor. Common mechanisms of the "second hit" include allele inactivation by mutation, loss of heterozygosity, or promoter hypermethylation leading to epigenetic silencing. Biallelic inactivation of MMR genes results in genomic instability due to failure in the repair of base pair mismatches that occur commonly during DNA replication (approximately 1 in 106 base pairs). DNA mismatches commonly occur in regions of tandem repeats of short DNA sequences called microsatellites that make up about 3% of human DNA (Baudhuin et al., 2005). Normally, the MMR machinery corrects errors in microsatellites, but mutations in the MMR genes in tumors cause expansion or contraction of these regions compared to normal tissue. These genetic alterations in microsatellite length are termed "microsatellite instability" (MSI) and are the molecular signature of LS-associated cancers (Lynch et al., 2009). Further, the increased mutation rate that results from MMR loss leads to alterations in nucleotide repeats in many other pathways; those that control cell growth, regulate cell death, and in the MMR genes themselves. Together, this accumulation of mutations drives the

The range of cancer risks in LS varies depending on the MMR gene involved. Approximately 70-80% of the clinical features of LS are accounted for by *MLH1* and *MSH2* mutations. Families with *MSH6* and *PMS2* mutations appear to have an attenuated cancer phenotype, presenting with a later age of diagnosis and a lower penetrance than *MLH1* and *MSH2*. *MSH6* may account for up to 15% and *PMS2* for up to 3-15% of all identified LS mutations (Hampel et al., 2005; Niessen et al., 2009). Recently, *EPCAM* has been thought to

*MLH1 and MSH2:* Endometrial cancer in patients with *MLH1* and *MSH2* mutations often occur before the age of 50. The risk in *MLH1* and *MSH2* carriers is up to 20% by age 50 and up to 60% by age 70 according to some studies (Aarnio et al., 1999; Lynch et al., 2009). However, the diagnosis of EC after the age of 50 should still raise concern for LS if there is a

*MSH6***:** *MSH6* mutation carriers appear to have the highest risk for endometrial cancer (up to 71%) of the MMR genes, higher than that of colorectal cancer (CRC) (Hendricks et al., 2004). The average age of onset of EC in *MSH6* mutation-positive individuals is 54 years. One study identified a somewhat lower risk for endometrial cancer in *MSH6* mutation carriers, however the risk was significantly increased above the general population, and appeared to be higher than the risk for CRC in women with LS. They reported a risk for endometrial cancer to age 70 of 26% (95% CI: 18-36%) and risk to age 80 of 44% (95% CI: 30- 58%) (Baglietto et al., 2009). These reports suggest that if a woman carries an *MLH1*, *MSH2* 

*PMS2***:** One large series of *PMS2* carriers found the incidence of EC to be 7.5-fold higher than expected in the general population. This translates to a 15% risk to age 70 (Senter et al.,

*EPCAM***:** The clinical features of *EPCAM*/*TACSTD1* mutations as a cause of Lynch syndrome is still being defined. Recent studies evaluating *EPCAM* 3'-end mutation carriers for their clinical phenotype found the risk for EC was dependent upon the type and size of

carcinogenetic process in LS.

positive family history.

2008).

**3.1 MMR genes and the risk of endometrial cancer** 

Endometrial cancer risk per MMR gene is as follows:

account for approximately 1-3% of LS mutations (Kuiper et al., 2011).

or *MSH6* mutation, her risk of EC may be even higher than her risk of CRC.

*EPCAM* mutation (Kempers et al., 2011; Kupier et al., 2011). However, since deletions in *EPCAM* lead to disruption of the *MSH2* gene, following management guidelines for LS appears prudent at this point. Further research is needed to clarify the EC risks associated with *EPCAM* mutations and their association with LS.

#### **3.2 Microsatellite instability**

As discussed above, microsatellite instability (MSI) results from defects in the MMR machinery that correct the replication errors found in these regions of the human genome. MSI may occur via two mechanisms. MSI in the majority of EC is sporadic in nature, resulting from hypermethylation of the *MLH1* promoter leading to epigenetic silencing of the gene (Esteller et al., 1998). The second, and the one associated with LS, is a consequence of germline mutations in the DNA-MMR genes as discussed above. Thus, MSI is not pathognomonic of LS, and in fact, LS accounts for only a minority of MSI-high EC cases (Meyer et al., 2009).

MSI analysis may be performed on paraffin-embedded tissue sections. Amplification by PCR using five primers recommended by the National Cancer Institute-two mononucleotide (BAT25, BAT26) and three dinucleotide repeats (D2S123, D5S346, D173250)-are used to detect changes in the number of microsatellite repeats in tumor compared with normal tissue (Boland et al., 1998). Tumors are classified using the five marker panel as follows: MSI-high (MSI-H, highly unstable) if two or more of the five markers are positive, MSI-low (MSI-L, low instability) if one of the markers is positive, and MS-stable (MS-S, no instability) if none of the markers show MSI. MSI analysis has some limitations when used to detect LSassociated endometrial cancers. Many, but not all the ECs that are diagnosed in LS are MSI-H, while most, but not all MSI-H endometrial cancers are sporadic (Garg & Soslow, 2009). Thus, MSI analysis may fail to detect some LS-associated ECs, while it may turn out positive in a large percentage of sporadic ECs.

#### **3.3 Immunohistochemistry**

Mutations in the MMR genes typically result in truncated or absent protein products. Immunohistochemistry (IHC) staining using antibodies to the C-terminus of the MMR proteins can be used to identify LS-associated tumors for the absence of these gene products (Weissman et al., 2011). Like CRC, IHC in endometrial cancer has shown efficacy for identification of LS. However, results must be interpreted with caution since both absent MMR gene product and *MLH1* promoter hypermethylation are found in up to one-third of endometrioid adenocarcinomas (Modica et al., 2007).

Further, more than one gene product may be absent. This may be due to the heterodimerization of the MMR proteins. Thus, a loss in *MLH1* staining is almost always coupled with concurrent loss of *PMS2*, and loss of *MSH2* staining is accompanied by loss of *MSH6*. A deleterious mutation in either primary proteins *MLH1* and *MSH2* will most likely result in loss of the entire heterodimer (Wei et al., 2002). As an example, a lack of tumor staining for *MLH1* and *PMS2* is most likely the result of *MLH1* protein absence. In contrast, *PMS2* and *MSH6* are secondary proteins, and a deleterious mutation in either gene will result in loss of that isolated protein. In addition, large deletions in the upstream *EPCAM* gene can cause inactivation and absence of *MSH2* expression by IHC. As many as 20-25% of cases suspected of having a mutation in *MSH2*, are actually caused by germline deletions in *EPCAM* (Rumilla et al., 2011)

Assays to detect methylation of the *MLH1* promoter that can recognize epigenetic mechanisms that lead to MSI-H, should be considered along with IHC for MMR gene testing (Whelan et al., 2002). For example, studies have shown that methylation of the small proximal region in the *MLH1* promoter located -248 to -178 relative to the gene transcription start site invariably correlates with loss of *MLH1* expression (Kang et al., 2002). If methylation is present, the patient most likely has sporadic tumor rather than LS-associated carcinoma.

IHC has been shown to be a convenient and readily performed test for the detection of germline MMR gene mutations. There are, however, studies of mutations in the MMR genes that are not detected by IHC (Vasen et al., 2004). In fact, by most reports, there is an approximate 5-10 % false negative rate with both IHC and MSI. That is, up to 90-95% of CRCs and ECs seen in LS patients are MSI-H or lack at least one MMR protein product on IHC testing (Ferreira et al., 2009). Therefore, most experts recommend that IHC and MSI testing in combination, along with family and personal history, be used to maximize identification of patients at risk for LS so that germline genetic testing may confirm the diagnosis.
