Over the last 25 years of managing the care of patients and families with hereditary colorectal cancer syndromes (HCCS), I have witnessed many changes affecting both patients and practitioners. In particular, the discovery of new HCCS genes affords more patients and families a definitive cause for their disease.
New technology for mutational analysis with next generation sequencing (NGS) has decreased the cost, increased the yield of genetic testing, and broadened the use of commercially available multi-gene cancer panels. Scientists believe we now have discovered the genetic cause of most highly penetrant HCCS. In 1991, at the start of my fellowship, the landmark discovery of the first genetic cause of a HCCS, familial adenomatous polyposis (FAP) on chromosome 5q21 was identified and subsequently mutations in the APC gene were confirmed as the cause of FAP. A few years thereafter, the genetic cause of Lynch syndrome was traced to mutations in the DNA mismatch repair (MMR) genes—MLH1, MSH2 (and EPCAM), MSH6 and PMS2—solving the most common HCCS. Traditionally, clinicians used phenotypic criteria derived by experts to assign a presumptive diagnosis and test for one specific gene. The clinical criteria served another important role, to homogenize the most informative families which enabled discovery of the molecular and genetic bases of these syndromes. Hereditary nonpolyposis colorectal cancer (a clinical diagnosis) is a cogent example of the evolution from clinical criteria with low sensitivity (i.e., Amsterdam Criteria I, Amsterdam Criteria II)— where families appeared phenotypically identical—to identifying the etiology of Lynch syndrome due to germline defects in one of the mismatch repair genes. This discovery has allowed for the important prognostic distinction between Lynch syndrome (a germline genetic diagnosis), Lynch-like syndrome (due to double somatic mutations in the tumor) and familial colorectal type X (no MMR defects). Today, clinical criteria still have relevance for identifying patients and families with HCCS and are currently required to ascertain insurance coverage of germline genetic testing; however, their importance is waning. Two reasons for the decline are the use of universal tumor testing of colorectal cancer (CRC) for evidence of defective DNA MMR and the increasing use of commercial multi-gene cancer panels made possible by NGS.
Historically, the approach to genetic testing in HCCS included a paradigm of establishing the phenotypic features of polyps and tumors in the pedigree most suggestive of one specific syndrome. This led to germline testing of a single or a limited number of genes through high cost, low throughput, Sanger sequencing. Sanger sequencing is considered a “first generation” DNA sequencing method. The discovery of the highly penetrant genes causing the most common HCCS, including Lynch syndrome, the adenomatous polyposis syndromes [FAP and MutYH-associated polyposis (MAP)], and the rarer, hamartomatous polyposis syndromes (juvenile polyposis syndrome, Peutz-Jeghers syndrome and the PTEN hamartoma tumor syndromes), provided at least one answer for each syndrome. Whereas Lynch syndrome is caused by mutations in DNA MMR genes, MAP is caused by biallelic mutations in a base excision repair gene, a separate form of DNA repair.
Figure 1. Results of Next Generation Sequencing for Hereditary Cancer Syndromes
The “test for a single gene for a specific syndrome” algorithm was expensive and time-inefficient, and became impractical as the number of genes relevant to HCCS expanded. The development of NGS and the use of multi-gene panels has revolutionized genomic research and clinical care, replacing first generation DNA sequencing and expanded our understanding of additional genetic causes of the HCCS. NGS technology allows massive sequencing by a high throughput, rapid and affordable platform.
In the era of NGS it is likely that the last few highly penetrant genes associated with hereditary CRC have been discovered. Three recently identified are rare causes of attenuated adenomatous polyposis and CRC. These include POLE and POLD1 (also known as polymerase-proofreading associated polyposis), which have an extra-colonic tumor phenotypes with duodenal adenomas, brain tumors, and in POLD1 families, endometrial cancer. These are autosomal dominant disorders. NTHL1-associated polyposis is another disorder of base excision repair (making it similar to MAP), and results in an autosomal recessive oligopolyposis and a tumor spectrum which is not yet fully defined. Duplications in the GREM1 promoter cause hereditary mixed polyposis syndrome, which is seen in Ashkenazi Jewish families and is notable for multiple types of polyps including serrated, adenomatous and inflammatory components. After years of searching for the cause of serrated polyposis syndrome (SPS), mutations in RNF43 have been determined to be a cause of SPS in some families and should soon be available on commercial gene panels.
Emerging over the last five years and becoming commonplace in the practice of cancer genetics is the use of NGS-based, affordable, multi-gene cancer panels. Some commercial panels include as many as 25 genes associated with the most common cancers and interrogate high, moderate and low penetrance genes. Multi-gene panels can increase the yield of germline genetic testing for syndromes with genetic heterogeneity or variable and overlapping phenotypes. Panels may also present clinical challenges to clinicians, including the detection of unanticipated pathogenic variants and variants of uncertain significance (VUS). Recent studies of multi-gene panel testing in individuals with CRC or presumed hereditary cancer syndromes have demonstrated a pathogenic mutation yield of up to 15 percent. In some studies, more than 50 percent of the mutations are found in unanticipated, non-Lynch syndrome genes and 25-50 percent of patients lack a phenotype suggestive of their underlying mutation (e.g. MSH2 mutation in a breast cancer patient without a personal or family history of CRC) (Table). The detection of patients with “unanticipated” high-penetrance mutations highlights the genetic heterogeneity of hereditary cancer, the benefit of testing more than one or two genes, and raises the question of whether HCCS should be defined based on clinical criteria, genotype (as it is in LS) or both. Most large studies of cancer panel testing show a VUS detection rate of up to 38 percent, although this proportion of uncertainty will fall over time.
The use of commercially available, multi-gene mutation cancer panels afforded through NGS presents emerging opportunities and challenges for gastroenterologists who must be adept at understanding the cancer risks associated with the well-known and newly discovered HCCS. Our expertise will increasingly be called upon to counsel patients on the clinical implications and management strategies of those with pathogenic variants. The greatest immediate challenges are interpreting unanticipated sequence variants in the absence of an informative family history, interpreting moderate penetrance mutations where guidelines are silent on management and the proper interpretation of VUS.
Dr. Burke has no conflicts to disclose. Dr. Burke is the Immediate Past President of the American College of Gastroenterology.
1. Yurgelun M.B., Allen B., Kaldate R.R. et al, Identification of a variety of mutations in cancer predisposition genes in patients with suspected Lynch Syndrome. Gastroenterology. 2015; 149:604–613.
2. Yurgelun M.B., Kulke M.H., Fuchs C.S. et al, Cancer susceptibility gene mutations in individuals with colorectal cancer. J Clin Oncol. 2017; 35:1086-1095.
3. Susswein L.R., Marshall M.L., Nusbaum R. et al, Pathogenic and likely pathogenic variant prevalence among the first 10,000 patients referred for next-generation cancer panel testing. Genetics in Medicine. 2016; 18:823-832.