Genetic Counseling in Hereditary Nonpolyposis Colorectal Cancer

Introduction

Two main forms of hereditary colorectal cancer can be distinguished: hereditary nonpolyposis colorectal cancer (HNPCC) and familial adenomatous polyposis (FAP). Both conditions have an autosomal dominant pattern of inheritance. Hereditary nonpolyposis colorectal cancer accounts for approximately 1%-5% of all colorectal cancer cases and FAP accounts for 1% of such cancers [1]. With some exceptions, FAP can be attributed to a germ-line defect in the adenomatous polyposis coli (APC) gene on chromosome 5q [2]. This gene was localized in 1987 and isolated in 1991. As a result, presymptomatic DNA-based diagnosis of FAP has become possible.

Patients with FAP exhibit a characteristic clinical picture of multiple adenomatous polyps in the large bowel. In contrast, patients with HNPCC do not present with pathognomonic signs. Therefore, clinical diagnosis of HNPCC has depended on family studies. Until recently the basic genetic defect of HNPCC was unknown. However, in 1993 a gene for HNPCC was localized on chromosome 2p through linkage analysis of two large HNPCC kindreds [3].

An important additional finding was provided by studies of tumors from HNPCC patients. Most of these tumors have a characteristic pattern of widespread genetic alterations, the so-called microsatellite instability or replication error positive phenotype [4]. This finding directed further studies on the nature of the HNPCC-associated gene on chromosome 2. This gene could then be identified as one normally involved in DNA mismatch repair [5,6]. Thus, HNPCC appears to be due basically to a disturbance of DNA mismatch repair, which leads to genetic instability in somatic cells. Within a year, three additional HNPCC-associated DNA mismatch repair genes were cloned [7,8]. Apparently, HNPCC is a genetically heterogeneous condition, with different genes being involved in different families.

Identification of the gene defects underlying HNPCC introduces the prospect of presymptomatic diagnosis for at-risk family members. Although DNA testing is potentially beneficial for these individuals, it also is associated with various psychosocial consequences.

In this article, genetic counseling of HNPCC families will be addressed, with emphasis on presymptomatic DNA-based diagnosis. For the purpose of illustration, clinical information on and mutation assays in two of our HNPCC pedigrees will be presented.

HNPCC: Definition and Diagnosis

As a group, HNPCC patients share certain features: early onset of disease, predominantly proximal tumor localization, a high incidence of multiple primary colorectal cancers, and possible manifestation of other tumor types, notably, endometrial cancer (Table 1). However, as mentioned above, pathognomonic clinical characteristics have not been identified. Therefore, HNPCC can be recognized only by its autosomal dominant inheritance pattern. Penetrance (the percentage of gene carriers who exhibit disease) is approximately 90%; skipped generations seldom occur [1].

In 1990, at the second meeting of the International Collaborative Group on HNPCC in Amsterdam, minimal criteria for the identification of HNPCC kindreds to be included in collaborative studies were proposed. These criteria are generally known as the Amsterdam criteria (Table 2) [9]. Since extracolonic tumors are not included, the diagnosis of HNPCC may be missed if one strictly adheres to these criteria for diagnostic purposes [10].

Clinically, HNPCC cannot be identified if the family history is negative. A patient with HNPCC due to a new mutation can be identified only by means of direct mutation studies.

In summary, the clinical diagnosis of HNPCC is based on observation of early-onset colorectal cancer and other tumors, notably, endometrial cancer, in successive generations. In large families with many affected individuals, the diagnosis may be straightforward. In small families with only a few affected individuals, the clinical diagnosis of HNPCC must often remain tentative.

DNA Mismatch Repair Genes and HNPCC

The DNA in every cell is continuously exposed to injury by various intrinsic and extrinsic factors. Usually, the consequences of damage to the DNA are not severe since the molecule is subject to various control systems that prevent the occurrence of mutations. These systems recognize insults, and subsequent pathways may lead to either repair of the errors or programmed cell death.

DNA repair systems can be separated into two main groups: nucleotide excision repair and mismatch repair. DNA mismatch repair is the repair of base-pair anomalies that occur during DNA replication [11]. The four DNA mismatch repair genes implicated in HNPCC are human homologs of bacterial and yeast DNA mismatch repair genes. The characteristics of these four genes are summarized in Table 3. In most HNPCC families, the condition appears to be due to a germ-line mutation in either the hMSH2 or hMLH1 gene [12].

Approaches to DNA-Based Diagnosis

There are two main approaches to presymptomatic DNA-based diagnosis: linkage analysis and direct mutation analysis. In these studies, the genomic DNA is usually isolated from blood samples.

In linkage analysis, haplotypes (chromosome regions) that harbor the gene of interest are examined. This method is based on a comparison of haplotypes from affected and unaffected family members. Since different genes are involved in HNPCC, linkage analysis, if feasible, is performed as the first step in identifying the gene locus involved in the family under investigation. Often, linkage studies are of limited value due to small family size or the limited availability of blood samples.

Direct mutation analysis focuses on the individual patient and is not dependent on family studies. Various DNA studies used to investigate HNPCC families are summarized in Table 4.

Some mutations, in particular, missense mutations that lead to an amino-acid substitution in the protein product, may not be causally related to the disease. Mutations that result in protein truncation (as a result of base-pair substitutions, deletions, or insertions leading to the generation of stop codons) are expected to be pathogenic. Therefore, new methods for the in vitro detection of protein truncation will probably become increasingly important for diagnostic purposes [13].