The first multigene panels for colorectal cancer became clinically available in 2012.1 Prior to that date, clinical testing for inherited colorectal cancer syndromes typically proceeded in a sequential fashion. A clinician would develop a differential diagnosis and test corresponding genes in order of those most likely to have a causative mutation. Thus, analyzing multiple genes for inherited colorectal cancer risk was both time intensive and costly. Additionally, the genes selected for analysis were based on established diagnostic criteria. As more individuals tested positive for a particular cancer syndrome, the phenotypic criteria evolved. However, a bias still existed as patients used to modify syndrome characteristics were typically detected because they met defined clinical criteria. The true phenotype associated with gene mutations is challenging to determine without population-based prospective studies, and such studies rarely exist. However, with the advent of multigene panels for inherited colorectal cancer, clinicians are able to cast the net wider and examine many genes associated with colorectal cancer risk. As a result, patients are being identified who do not closely resemble established clinical criteria, and the phenotypes of inherited colorectal cancer syndromes are evolving.
Cragun and colleagues1 described clinical laboratory data for results reported for an inherited colorectal cancer panel between March 2012 and March 2013. In this data set, 1 in 4 individuals did not meet established syndrome-based testing guidelines. Of these individuals, there were patients with clinical histories that were suggestive of the condition diagnosed via genetic testing and also several atypical cases. Among the atypical cases were 2 individuals with SMAD4 mutations and no juvenile polyps, 1 individual with a PTEN mutation and greater than 100 adenomatous colonic polyps, and 1 individual with a CDH1 mutation who did not have breast or gastric cancer and whose family history did not report gastric cancer. Without multigene panels, it is likely these families would not have been genetically diagnosed.
As the number of individuals undergoing gene panels for inherited colon cancer increases, so do the data surrounding atypical cases and expanding phenotypes. This was recently seen at the 2014 Collaborative Group of the Americas on Inherited Colorectal Cancer (CGA-ICC) meeting. Data from both academic institutions and commercial laboratories were reported. Several authors suggested that an association between colorectal cancer and Li-Fraumeni syndrome exists and that TP53 analysis should be included in young colorectal cancer patients.2-4 Of interest, of the 42 patients with clinically actionable mutations described by Cragun and colleagues,1 only 1 patient was identified with a TP53 mutation, and this patient met Chompret criteria. However, as this paper was focused on results obtained from an inherited colorectal cancer panel, it is possible other colorectal cancer survivors could have been identified with a TP53 mutation via a multigene panel that covered colorectal cancer and other cancers, such as breast or gynecologic malignancies, which was not reported in this study.
Data presented at the CGA-ICC meeting also suggested overlapping phenotypes between hereditary breast and ovarian cancer syndrome and Lynch syndrome, two of the most well-described hereditary cancer syndromes. In a laboratory cohort described by Myriad Genetics in which a multigene panel was performed on samples previously submitted for Lynch syndrome testing, approximately 11% of the non–Lynch syndrome highly penetrant mutations (defined as BRCA1/2, APC, biallelic MUTYH, or STK11/BMPR1A mutation carriers) identified were in BRCA1/2. The majority of these individuals (93%) met National Comprehensive Cancer Network (NCCN) criteria for Lynch syndrome testing while only 33% met NCCN criteria for BRCA1/2 testing.5 Similar findings were demonstrated in a cohort described by GeneDx. In this laboratory cohort, results for any multigene panel ordered on an individual with colorectal cancer and/or polyps were analyzed. As such, the exact genes analyzed varied depending on the panel performed. However, approximately 20% of individuals testing positive for a non–Lynch syndrome highly penetrant gene (defined as BRCA1/2, APC, biallelic MUTYH, or STK11/BMPR1A mutation carriers) were found to carry a mutation in BRCA1.6 An association of breast cancer with PMS2 and MSH6 mutations was also suggested.7,8
Similar to those in the Cragun and colleagues paper, cases were also presented at CGA-ICC that did not meet the defined clinical criteria and would have likely been missed if testing strategies relied on single-syndrome approaches.9 Ambry Genetics presented several cases in which individuals with PTEN mutations met criteria for attenuated or classic familial adenomatous polyposis,10 as well as individuals who carried CDH1 mutations but did not have a personal and/or family history of gastric cancer.11 Additionally, many authors reported individuals testing positive for mutations in moderately penetrant genes,5,6,12,13 further emphasizing the need for better risk estimates and management guidelines surrounding these mutations.
Take-Home Messages
To sign up for our newsletter or print publications, please enter your contact information below.
