Genetic Testing in Gastrointestinal Cancers: A Case-Based Approach

Family History of Gastric Cancer

FIGURE 3

Pedigree of Miss Y's Family

Case: Miss Y is a healthy 20-year-old woman referred to you for gastroscopy screening because of a family history of diffuse gastric cancer (DGC) in her father and paternal uncle. You schedule her for an endoscopy and refer her to genetics. In the interim, results of Miss Y’s gastroscopy are negative. Given her family history of a lobular breast cancer diagnosis in her paternal aunt at age 55 years and the two diagnoses of DGC in her father at age 35 years and in her paternal uncle at age 50 years (Figure 3), you refer her for genetic counseling.

Genetic evaluation: A total of 21,320 new diagnoses of gastric cancer in US men and women is projected for 2012.[5] Approximately 3198 of these newly diagnosed patients (15%) will prove to have familial clustering of the disease, and in 852 (4%), the diagnosis will be due to high-risk Mendelian susceptibility genes (Figure 4).

FIGURE 4

Familial Gastric Cancer

This particular case is highly suspicious for hereditary DGC (HDGC), associated with germline mutations in CDH1, which encodes a cell-cell adhesion molecule (E-cadherin) and is the only gene associated with the syndrome.[27] Because Miss Y’s father and paternal uncle are deceased, testing of CDH1 is initiated in her paternal aunt who had lobular breast cancer. Even though this aunt’s cancer could potentially be a phenocopy (sporadic breast cancer unrelated to disease), because lobular breast cancer is associated with HDGC, she is the appropriate individual to test. Point mutations in CDH1 are found in 30% to 50% of HDGC families, with deletions in the CDH1 gene accounting for an additional 4% of such families.[28] In this case, sequencing reveals a germline frameshift mutation in CDH1 that has previously been reported in an unrelated HDGC family.

In view of the known mutation in the family, Miss Y can now be tested to determine her carrier status. Prior to genetic testing, she is counseled with regard to HDGC, its current management, and the benefits and limitations of genetic testing. It is discussed that a genetic result would help clarify her risk and therefore might aid her decision making around surveillance and preventive interventions. While it is known that virtually all carriers of CDH1 mutations have or will develop minute cancers within the lining of their stomachs, 20% of mutation carriers will not develop clinically significant DGC. What triggers the progression to clinically significant DGC is not known. Unfortunately, there is no effective means by which to find or monitor these lesions; therefore, the only way to reduce risk of DGC is by completely removing the stomach, which has significant associated morbidity and a 1% risk of mortality. Although there is a federal law prohibiting health care and employment discrimination, the potential for insurance discrimination, as well as the possible psychological implications for both the patient and the family, is discussed—as is done in every pretest counseling session. Based on her family history, Miss Y has a 50% chance of inheriting the mutation, since it is assumed that her father with DGC is an obligate carrier who shares the familial CDH1 mutation.

Following the pretest counseling, Miss Y undertakes predictive genetic testing for the familial mutation. The results indicate that Miss Y is positive for the mutation; she therefore is counseled regarding her 80% lifetime risk of DGC and 60% lifetime risk of lobular breast cancer,[29] as well as with regard to the recommended screening. Currently, the International Gastric Cancer Linkage Consortium (IGCLC) recommends annual upper endoscopies with multiple random biopsies for gastric cancer surveillance until prophylactic total gastrectomy is undertaken.[29] Multiple (minimum of 30) gastric biopsies must be taken randomly in order to increase the chance of detection.[29,30] Miss Y is referred to a multidisciplinary team for further discussion regarding total prophylactic gastrectomy. She ends up deciding to delay the procedure and continue with endoscopic surveillance until she has completed child bearing. It should be noted, however, that there are case reports of successful pregnancies post–prophylactic gastrectomy.[31] It is important to bear in mind that the timing of prophylactic gastrectomy is a highly individual decision that the patient should make with the help of a multidisciplinary team that includes a gastroenterologist, a geneticist, a genetic counselor, a surgeon, a nutritionist, and an oncologist, as well as a psychologist/counselor.

TABLE 3

A Selection of Other Genes That Can Have Gastrointestinal Manifestations, WIth Their Respective Syndromes

Miss Y discusses her mutation results with her sister. Her sister is keen to learn more information; however, her insurance provider is unwilling to cover the costs of genetic counseling and testing. In general, insurance providers differ in their specific policies regarding genetic counseling and testing; thus, it is worthwhile for a patient to investigate the policies of several companies. Alternatively, these services can be privately paid for—and for privacy concerns, some individuals elect this option. In the case of Miss Y’s sister, testing should be considered medically necessary. She has a 50% chance of having inherited a previously identified familial mutation associated with a highly penetrant cancer susceptibility syndrome for which surveillance has been shown to miss underlying cancer. It thus is reasonable for the ordering physician to appeal to the insurance provider on behalf of the patient to provide coverage for genetic counseling and testing. It should be emphasized that without being able to redefine the patient’s genetic risk, the insurance provider would by default need to cover the medically necessary and recommended surveillance for an at-risk individual from an HDGC family. In general, enrolling patients into disease-specific research studies is important to advance ongoing research and understanding of the disease. In some cases it is possible for study participants to learn preliminary genetic testing research results.

Following an appeal to the insurer for targeted testing for the family mutation, Miss Y’s sister is found not to be a carrier. Based on these results, Miss Y’s sister’s risk for gastric and breast cancer returns to that of the general population, and she does not require high-risk gastric or breast cancer surveillance (Table 3).

Early-Onset Colon Cancer

FIGURE 5

Pedigree of Mrs. V's Family

Case: Mrs. V is a 44-year-old woman referred by her gastroenterologist because of a new diagnosis of right-sided colorectal cancer (CRC) diagnosed on colonoscopy performed secondary to rectal bleeding. Review of the family history reveals that her maternal grandmother had uterine cancer at age 50 years and her maternal uncle had CRC at age 50 years (Figure 5). In view of the patient’s younger age at onset[5] and the family history of Lynch syndrome–related cancers, an underlying genetic susceptibility needs to be investigated.

Genetic evaluation: A total of 143,460 new diagnoses of CRC in US men and women is projected for 2012.[5] Approximately 21,519 of these newly diagnosed patients (15%) will prove to have familial clustering of the disease, and in 7173 (5%), the diagnosis will be the result of a mutation in a high-risk Mendelian susceptibility gene. Specifically, Lynch syndrome, caused by mutations in the DNA mismatch repair (MMR) genes, accounts for 2% to 5% of all CRC (Figure 6).

FIGURE 6

Familial Colon Cancer

Even without the family history of Lynch syndrome–related cancers, Mrs. V meets the revised Bethesda guidelines criteria for consideration of a diagnosis of Lynch syndrome (on the basis of a diagnosis of CRC under the age of 50 [Table 2]). The NCCN, in the revised Bethesda guidelines, recommends further examination of the tumor with either microsatellite testing and/or immunohistochemistry.[32] Lynch syndrome is an autosomal dominant cancer susceptibility syndrome caused by mutations in one of several DNA MMR genes (or a gene located nearby) that ultimately lead to loss or abnormal function of the MMR proteins (Table 2). Alterations in the MMR system cause errors in DNA replication to accumulate, particularly in sequences of DNA known as microsatellites. These repetitive regions of DNA easily mispair during normal DNA replication, causing gains or losses of the repeat sequences. If the MMR proteins are impaired, the cell cannot properly repair its DNA, resulting in the accumulation of microsatellite instability (MSI). In patients with a germline MMR gene mutation, a second copy of the affected MMR gene is somatically mutated, with progressive accumulation of altered microsatellites in the coding regions of genes involved in tumor initiation and progression.

Individuals with Lynch syndrome develop cancers in certain tissues that are more susceptible to losing the function of particular proteins; thus, there is a distinct pattern of cancers that is associated with the syndrome (Table 2). The optimal way to screen a cancer patient for Lynch syndrome is via tumor tissue testing. Analysis for the presence of MSI and/or immunohistochemical staining (IHC) for the four MMR proteins can be performed on colon and endometrial tumors. In patients with tumors exhibiting defective MMR either by MSI testing or by IHC, further germline genetic testing for the appropriate MMR genes is indicated. Complicating the diagnosis of Lynch syndrome is the fact that 15% to 20% of sporadic colon and endometrial cancers exhibit the same MMR repair defects as their hereditary counterparts.[33] Thus, the results of tumor testing need to be interpreted in conjunction with other clinical information. For example, when MLH1 protein is deficient in the tumor, as determined by IHC, it is possible either that there is an MLH1 germline mutation or that somatic epigenetic silencing has occurred. In such cases, testing for the presence of MLH1 promoter hypermethylation or the presence of a V600E mutation in BRAF, either of which would identify the tumor as sporadic, can help rule out Lynch syndrome.[34] Keep in mind that although MSI and IHC testing are useful screens, they are not 100% sensitive. The sensitivity of MSI testing is related to the panel of microsatellite markers used. While there have been efforts to standardize them, the panels can differ between institutions.[35] When using 3 or more single-nucleotide repeating microsatellites, the sensitivity for detecting germline MLH1 and MSH2 mutations is 89%, and 77% for MSH6; the sensitivity of IHC for MLH1, MSH2, and MSH6 is 83%.[36] Because either approach can theoretically lead to false-negative assessments for Lynch syndrome and thus potentially result in missed opportunities for surveillance and risk-reducing strategies, if there is sufficient clinical suspicion of Lynch syndrome based on personal or family history or other clinical parameters, the case for further genetic testing should be made. For example, because some tumors in patients with Lynch syndrome are missed by MSI testing but detected by IHC and vice versa, a gain in sensitivity may be achieved by the use of both tests and is advisable in high-risk clinical scenarios.

Patients with a diagnosis of Lynch syndrome have a 50% to 80% lifetime risk of CRC, with an associated accelerated progression of the adenoma-to-carcinoma sequence and a mean age of 45 years at CRC diagnosis in the proband (although diagnosis is generally later in the mutation-carrier family members).[37,38] Women with Lynch syndrome are also at a substantially increased risk for endometrial cancer, with an estimated lifetime risk of 40% to 60%.[39,40] Other Lynch syndrome–associated cancers include cancers of the stomach, small intestine, pancreas, ovaries, and biliary tract, and urothelial carcinoma of the renal pelvis and ureter. The presence of sebaceous neoplasms of the skin is seen in Muir-Torre syndrome, a variant of Lynch syndrome, while brain tumors, including glioblastomas and astrocytomas, are seen in Lynch families with Turcot syndrome.

Given her early age at onset and her family history of Lynch-associated cancers, Mrs. V’s colon tumor biopsy tissue was assessed by IHC for expression of the four DNA MMR proteins. An intact expression of MLH1 and PMS2 proteins was noted; however, staining was absent for the expression of both MSH2 and MSH6 proteins. IHC may be helpful in the diagnosis of Lynch syndrome, as results may direct gene-specific clinical genetic testing. In this case, the absence of MSH2/MSH6 protein expression suggested that the most likely culprit gene was MSH2. Thus, Mrs. V underwent germline genetic testing with sequencing, which revealed a deleterious mutation in the MSH2 gene.

In patients with a new diagnosis of colorectal cancer, the upfront identification of Lynch syndrome may have a significant impact on surgical management. First, given the high risk of metachronous colorectal cancer[41] (16% at 10 years, 40% at 20 years after initial CRC diagnosis[42]), subtotal colectomy as opposed to a segmental colon resection may be considered. Subtotal colectomy reduces the metachronous risk by 31% for every 10 cms of bowel removed.[42] Given the associated increased lifetime risk of endometrial and ovarian cancer, if a female patient has completed childbearing, a prophylactic total abdominal hysterectomy and bilateral salpingo-oophorectomy (TAH-BSO) may also be considered. After the presurgical clinical genetics evaluation and the identification of Lynch syndrome, Mrs. V underwent subtotal colectomy as well as risk-reducing TAH-BSO. Had Mrs. V elected not to undergo subtotal colectomy, following treatment and follow-up surveillance for her primary CRC, she would have been advised to adhere to recommendations for surveillance by colonoscopy every 1 to 2 years.[43]

It is notable that colorectal cancers exhibiting defective MMR, whether due to a somatic event or a germline MMR mutation, exhibit certain phenotypic features. Consistent with a Lynch syndrome–associated CRC, Mrs. V’s pathology revealed a right-sided, poorly-differentiated, mucinous invasive adenocarcinoma with medullary features and the presence of tumor-infiltrating lymphocytes.[44,45] More recent evidence also suggests that defective MMR (ie, the MSI-high tumor phenotype) may also be a prognostic and a predictive marker in CRC. Numerous retrospective studies, including large meta-analyses, have shown that patients with CRC with defective MMR have improved stage-independent survival compared with patients with proficient-MMR CRC.[45,46] In addition, by using data from randomized clinical trials of fluorouracil(Drug information on fluorouracil) (5-FU)-based therapy vs surgery-only controls, a predictive role for MMR status has also been demonstrated, with CRC patients who have defective MMR not appearing to benefit from treatment with 5-FU–based chemotherapy.[47,48]

Last but not least, with the identification of a germline MSH2 mutation in this patient, targeted genetic testing can now be undertaken in her family members, and at-risk family members can partake in appropriate cancer screening and prevention measures (Table 2).[32]