We need to be more aggressive in identifying hypermutable cancers, because these patients might be great candidates for PD-1 inhibitor therapy.
—Michael Hall, MD
The genetic basis for inherited colorectal cancer is proving to be a much bigger and more complicated “pie” than was appreciated just a few years ago, according to Michael Hall, MD, Director of Gastrointestinal Risk Assessment at Fox Chase Cancer Center, Philadelphia.
“With next-generation sequencing, we can now test panels of tens to hundreds of genes at one time, and we are identifying more pieces of this pie,” Dr. Hall said in an interview following the 19th Annual Meeting of the Collaborative Group of the Americas on Inherited Colorectal Cancer (CGA) in Baltimore.
The ASCO Post asked Dr. Hall, a gastrointestinal medical oncologist and member of the organization, to describe the CGA, its mission, and some of the most provocative presentations from its recent conference.
What Is the CGA?
The CGA was established in 1995 to improve understanding of the basic science of inherited colorectal cancer and the clinical management of affected families. The group’s focus is on the growing group of inherited conditions that extend beyond the more recognized entities such as Lynch syndrome and familial adenomatous polyposis. The mission is to educate clinicians regarding genetics and management, improve access to trials, develop resources and support for registries, maintain a forum for the exchange of ideas, and serve as the authoritative voice of hereditary colorectal cancer health care in the Americas.
“CGA basically started out as a working group, and now 20 years later, we have grown into an organization that holds a yearly conference attended by a mix of gastroenterologists, oncologists, surgeons, pathologists, genetic counselors, and other researchers from around the world,” said Dr. Hall.
Oncologists can benefit from involvement in the CGA, he said. “We know that oncologists are aware of Lynch syndrome and have at least some experience with [familial adenomatous polyposis], but they may not be aware of the many other inherited gastrointestinal conditions,” Dr. Hall noted.
About one in six colorectal cancer patients under the age of 50 has an inherited oncogenic mutation, many of them in genes that would not typically be suspected, he pointed out. “We think it’s important for oncologists to understand that when they see a patient under 50 with [colorectal cancer], this could be an inherited condition,” he added.
While the primary focus of the CGA is colorectal cancer research, its reach also includes health services and behavioral research and heritable forms of other gastrointestinal malignancies. The group’s 20th Annual Meeting will be held October 2–3, 2016, in Seattle.
The following summaries describe abstracts of interest from this year’s CGA meeting.
Genetic Testing: Look Beyond Lynch Syndrome
Rachel Pearlman, MS, LGC, and colleagues from The Ohio State University Comprehensive Cancer Center, Columbus, evaluated the germline mutation spectrum in 250 colorectal cancer patients diagnosed under the age of 50 in the Ohio Colorectal Cancer Prevention Initiative.1 Genetic testing was performed on cohorts with mismatch repair–deficient tumors and mismatch repair–proficient tumors.
Of the 250 patients, 40 (16.8%) were found to have at least one pathogenic germline mutation (9.6% Lynch syndrome, 7.2% other hereditary cancer syndromes). Testing with a comprehensive panel of genes detected mutations that would have otherwise been missed.
“Had targeted testing for polyposis or a panel with only highly penetrant [colorectal cancer] genes been ordered, 9 of the detected germline mutations (22%) would have been missed,” the authors indicated. These pathogenic mutations in clinically actionable hereditary cancer genes provide valuable information for [colorectal cancer] patients and their families, they suggested.
“Ms. Pearlman’s group used the full force of the genetic testing that is available to screen tumors from high-risk individuals. This was a huge effort to recruit every colorectal cancer case in Ohio and get these patients in for genetic testing,” Dr. Hall observed.
“They found a high frequency of mutations in this under-50 group, and while quite a few were associated with Lynch syndrome genes, there were others that we would not necessarily expect but yet are clinically actionable, such as BRCA1 or BRCA2, PALB2, and P16 mutations,” he noted. “This argues that these under-50 colorectal cancer patients need a broader type of genetic investigation to make sure we are not missing other genes contributing to hereditary cancer risk and health risk.”
Modulating Genes and Reducing Polyp Burden
Deboarah Neklason, PhD, and colleagues from the Huntsman Cancer Institute, Salt Lake City, found that targeted inhibition of the cyclooxygenase-2 (COX2) and epidermal growth factor receptor (EGFR), in patients with familial adenomatous polyposis, downregulates oncogenes and results in regression of duodenal neoplasia.2 They conducted a phase II double-blind randomized trial evaluating the effect of inhibiting COX2 with sulindac and EGFR with erlotinib. They evaluated the drugs’ effect on polyp burden and on gene expression by comparing normal and polyp duodenal tissue before and after the 6-month treatment.
This comparison revealed more than 600 differentially expressed genes among the placebo group, compared to only 58 differentially expressed genes from patients on treatment, who also had fewer polyps.
“This suggests that pathways activated in polyp formation are blocked by this drug combination,” the authors wrote.
The study identified molecular drivers of familial andenomatous polyposis, resulting in information that may someday have practical implications, according to Dr. Hall. While colectomy in young adulthood can help prevent colorectal cancer in these patients, upper gastrointestinal polyposis remains problematic as a precursor to malignancy and ultimately could benefit from a pharmacologic intervention.
“At baseline and in the placebo group, we saw an overexpression of many genes we would not expect. In the treatment group, however, the number of differentially expressed genes fell tremendously, showing that the combination of these agents was suppressing many important driver pathways in polyps—ERK, ERBB2, SMAD4, beta-catenin and others,” Dr. Hall said.
“This provides proof of principle that there are molecular pathways that might be modulated therapeutically,” he continued. “Patients with dense polyposis, which can be very difficult to manage endoscopically, are a subset of patients with familial adenomatous polyposis who are at risk for upper gastrointestinal cancers. This drug combination could be a great option for them.”
Biallelic Mismatch Repair Syndrome Described
Biallelic mismatch repair deficiency syndrome is an aggressive inherited cancer predisposition syndrome. Individuals harboring biallelic germline mutations in the mismatch repair genes develop brain tumors, hematologic malignancies, and gastrointestinal cancers. In stark contrast to patients with Lynch syndrome, where the mean age of gastrointestinal cancer diagnosis is 47 years, almost all patients with biallelic mismatch repair deficiency syndrome develop gastrointestinal malignancies before age 18. There is evidence that these tumors may be amenable to treatment with immune checkpoint inhibitors.
An international team of investigators used a multiassay approach to detect biallelic mismatch repair deficiency syndrome in suspected patients and to assess mismatch repair capabilities in known patients.3 Exome and genome sequencing determined mutation frequency in all biallelic mismatch repair deficiency syndrome cancers. They found that 74% of all tumors and 37% of gastrointestinal tumors were microsatellite-stable. In contrast, there was a complete lack of staining of protein in the affected gene in 95% of cases. Exome sequencing of 29 biallelic mismatch repair deficiency syndrome cancers revealed extremely high mutation frequencies, very distinct from other childhood cancers.
Again, in contrast to tumors arising from Lynch syndrome, microsatellite-instability analysis was not informative in detecting biallelic mismatch repair deficiency syndrome, but both immunohistochemistry and the in vitro G/T mismatch repair assay were highly specific and sensitive for detection of the syndrome.
The findings point to a “hypermutant” phenotype that is unique to biallelic mismatch repair deficiency syndrome and diagnostic in these pediatric cancers. The extremely high mutation rate in gastrointestinal cancers predicts excellent response to immunotherapy, the authors maintained.
“This genetic syndrome is probably more common than we think—because Lynch syndrome itself is common—but it’s not on people’s radar,” Dr. Hall said. “What is very impressive about this work is that the researchers collected enough subjects to conduct these studies.”
Data presented in the past year suggested that programmed cell death protein 1 (PD-1) inhibitors may be effective in mismatch repair–deficient tumors such as these, he added. Indeed, the researchers of this study reported that these brain tumors had striking responses to PD-1 inhibitors.
“This suggests that we need to be more aggressive in identifying hypermutable cancers, because these patients might be great candidates for PD-1 inhibitor therapy,” he commented.
Other Meeting Highlights
Luis A. Diaz, MD, of Johns Hopkins Sidney Kimmel Comprehensive Cancer Center, Baltimore, reviewed data on the impact of the PD-1 inhibitor pembrolizumab (Keytruda) in mismatch repair–deficient colorectal cancer. Dr. Diaz was senior author of the phase II study recently published in The New England Journal of Medicine involving 41 patients with metastatic carcinoma, with or without mismatch repair deficiency.4
Patients with mismatch repair–deficient colorectal cancers in that study had a 40% response rate and 78% progression-free survival rate; median progression-free and overall survival was not reached. In contrast, those with mismatch repair–proficient colorectal cancers had poor outcomes. Pembrolizumab reduced the risk of progression by 90% (P < .001) and mortality risk by 78% (P = .05) in the mismatch repair–deficient group.
Dr. Diaz also indicated that tumors caused by mutations in the relatively newly described DNA polymerase genes, POLE and POLD1, can also be modulated by immunotherapy, suggesting that individuals with mutations in these genes might also derive benefit from checkpoint inhibitors.
A lecture by Judy Garber, MD, MPH, of Dana-Farber Cancer Institute, Boston, focused on the clinical challenge of genetic testing for large numbers of genes, especially the problem of variants of unknown significance. “Management of [variants of unknown significance] is a contentious issue,” Dr. Hall noted. “Dr. Garber gave a nice overview that explained why these incidental findings are here to stay, why they are a difficult aspect of the genetic testing arena, and how she thinks they should be managed.”
The CGA also bestowed a Lifetime Achievement Award on C. Richard Boland, MD, of Baylor Scott & White Health, Dallas. Dr. Boland’s seminal research helped explain microsatellite instability and its implications for cancer therapy. ■
Disclosure: Dr. Hall, and Dr. Neklason reported no potential conflicts of interest. Ms. Pearlman reported that The Ohio Colorectal Cancer Prevention Initiative is receiving free genetic testing from Myriad and discounted genetic testing from the University of Washington.
1. Pearlman R, Hampel H, Allen B, et al: Ohio Colorectal Cancer Prevention Initiative: Germline mutation spectrum in 250 colorectal cancer patients diagnosed under age 50. Collaborative Group of the Americas for Hereditary Colorectal Cancer Annual Meeting. Abstract 14. Presented October 11, 2015.
2. Neklason DW, Delker DA, Samadder NJ, et al: Molecular pathways impacted in FAP patients treated with COX and EGFR combined inhibition. Collaborative Group of the Americas for Hereditary Colorectal Cancer Annual Meeting. Abstract 6. Presented October 11, 2015.
3. Campbell B, Aronson M, Shlien A, et al: Functional characterization of biallelic mismatch repair deficiency and its implications for novel therapies. Collaborative Group of the Americas for Hereditary Colorectal Cancer Annual Meeting. Abstract 2. Presented October 11, 2015.
4. Le DT, Uram JN, Wang H, et al: PD-1 blockade in tumors with mismatch-repair deficiency. N Engl J Med 372:2509-2520, 2015.