In a recent issue of the Journal of Clinical Oncology,1 and as reviewed in this issue of The ASCO Post, an ASCO expert endorsement panel reviewed and endorsed, with minor qualifications, the European Society for Medical Oncology (ESMO) clinical practice guidelines for management of familial/genetic colorectal cancer predisposition.2 The guidelines address genetic testing strategies, clinical surveillance, and treatment logics for hereditary nonpolyposis colorectal cancer (HNPCC or “Lynch syndrome”) and for familial adenomatous polyposis.
With this commentary, which generally represents a positive endorsement of this endorsement, I will note a few emerging areas of interest that were not raised by either ESMO or ASCO and will critique some specifics that could bear a revisit in the future.
Major Recommendations
Key features of the ASCO/ESMO guidelines include the following:
(1) All cases of colorectal cancer or, alternatively, patients aged < 70 years and those over 70 years who meet criteria suggesting higher likelihood of underlying HNPCC/Lynch syndrome (so-called Bethesda Guidelines) should be referred for tumor testing with immunohistochemistry for evidence of defective mismatch repair. This process selects patients who would benefit from germline mutation testing. Recognition is made of the frequent nonfamilial/genetic or sporadic loss of MLH1 expression due to promoter hypermethylation and the consequent need for a secondary assay (BRAF mutation or methylation).
(2) The only firm surveillance recommendation for HNPCC/Lynch syndrome, and the only one with a solid evidence base, is colorectal screening by means of colonoscopy starting at age 20 to 25 years and repeated at 1- to 2-year intervals. While the lack of evidence base is acknowledged, there is nevertheless a recommendation for annual endometrial sampling and transvaginal ultrasound (ASCO, unlike ESMO, does not recommend CA-125). Other than looking for evidence of Helicobacter pylori and perhaps esophagogastroduodenoscopy in high-risk geographies, no recommendation for upper gastrointestinal screening is offered, and no other extracolonic at-risk organs carry a positive screening recommendation.
(3) For treatment in HNPCC/Lynch syndrome, aspirin is supported as a chemopreventive intervention, but uncertainty as to proper dose leaves this as a “soft” recommendation. Consideration of risk-reducing total abdominal hysterectomy with bilateral salpingo-oophorectomy after childbearing is offered and does have some evidence base behind it. No real recommendation as to extent of surgery for colorectal cancer is offered, supporting, in effect, the status quo of segmental surgery/postsurgical surveillance or subtotal colectomy, the choice depending on outcome of individualized surgeon-patient negotiation.
(4) In familial adenomatous polyposis, APC (adenomatous polyposis coli) mutation testing of affected individuals is recommended as a foundation for predictive testing in at-risk relatives. In young mutation carriers, sigmoidoscopy at 2-year intervals starting at age 12 to 14 is recommended, with colonoscopy once adenomas are encountered. ASCO recommends starting younger (10 to 11 years), at intervals of 1 to 2 years, and would consider full colonoscopy as an alternative to sigmoidoscopy. These minor distinctions do, in my experience, echo a real difference in practice patterns between the United States and Europe.
Surgical choice of colectomy/ileorectal anastomosis vs proctocolectomy/ileal J-pouch-anal anastomosis is based on well-established considerations. Upper gastrointestinal surveillance and endoscopic and surgical interventions are predicated on the high rate of duodenal adenomas, though little note is made of the tremendous difficulties that the multiple, large, flat polyps pose to both the endoscopist and surgeon. The challenges of managing desmoids and the controversial role for chemopreventive measures (typically nonsteroidal anti-inflammatory drugs) are noted and follow conventional wisdom. The ASCO guidelines suggest a greater willingness to consider routine imaging for desmoids in high-risk familial adenomatous polyposis patients.
(5) The approach to recessively inherited MYH-associated polyposis is clinically similar to that of the attenuated form of familial adenomatous polyposis.
Unaddressed Issues
As indicated above, the differences between the ESMO and ASCO guidelines are fairly minimal. My remaining comments are intended to call attention to several areas that are not addressed by either ESMO or ASCO. Although it is likely that these simply relate to lack of an established evidence base or expert consensus on controversial topics, the practitioner wishing to manage these high-risk patients must be aware of certain problem areas.
In the case of HNPCC/Lynch syndrome, risk assessment models have been offered as a basis for comprehensive mismatch-repair gene mutation testing (alone or as part of panels, as noted below) in a fashion similar to that used in familial breast/ovarian cancer to predict mutation likelihood.3-5 Such testing has been offered as an alternative to tumor-based testing (microsatellite instability/immunohistochemistry). This may be appropriate, so long as the clinician is aware of the limitations and is willing to return to the tumor if, as is often the case, germline DNA testing is not informative.
Neither ESMO nor ASCO devotes much attention to this issue. The extent of use of risk models as a basis for germline mutation testing is unknown. Testing based on such models is being aggressively marketed in the United States, likely much more so than in Europe, and may account for the lack of attention by ESMO.
Panel Testing
As the number of cancer predisposition genes increases, including those with lower penetrance or those that overlap other disorders, and as the cost of sequencing decreases, several commercial entities have recently begun to offer “panel” testing of 16 or more genes. Examples include APC, ATM, AXIN2, BMPR1A, CDH1, CHEK2, EPCAM, MLH1, MSH2, MSH6, MUTYH, PMS2, PTEN, SMAD4, STK11, and TP53 (http://www.genedx.com/oncology-genetics/).
Other labs offer a different number of genes that are either relatively specific for gastrointestinal cancer or are even broader in scope (Ambry, 14 genes: http://www.ambrygen.com/hereditary-cancer-panels; Myriad, 25 genes: http://www.myriad.com/products-services/hereditary-cancers/myrisk-hereditary-cancer/). Academic centers such as the University of Washington and the City of Hope Cancer Center offer similar panels. They consist of massively parallel or next-generation DNA sequencing with or without rearrangement studies.6,7
Many elitists have scoffed at such “shotgun” testing. Yet, we must admit that there are cases in which a patient’s personal and family histories are sufficiently limited, conflicting, or otherwise confusing on the one hand, while on the other hand, his or her zeal for testing is high, so as to warrant consideration of such testing. This appears to make the testing easier for the clinician to whom the panels are being aggressively marketed. However, the less clear the clinical picture, the lower the yield of such testing. Thus, the clinician needs to guard against providing the patient a false sense of security when the panel is a nondiagnostic negative. Some form of clinical surveillance may still be warranted.
Conversely, we are starting to see tests that are positive for mutations in genes that would not have been considered under a more-targeted testing strategy. This may be good in that it suggests a greater overall sensitivity for a shotgun approach. However, it may be problematic to the extent that a given positive test may be indicative of a very attenuated expressivity that will be hard to translate into an appropriate predictive testing and clinical surveillance strategy.
I would encourage expert panels that have taken up guideline development to place panel testing on their agenda, even if lack of consensus for or against prevents taking a position. I emphasize this not because there is a strong evidence base, but simply because such testing is in fact already commercially available, and clinicians deserve some expert guidance on the matter.
Variants of Uncertain Significance
The matter of “variants of uncertain significance”—that is, DNA germline sequence variants that may or may not be pathogenic—is not taken up by ESMO or ASCO. In patients with informative microsatellite instability/immunohistochemistry, suggestive of underlying HNPCC/Lynch syndrome, this can be especially vexing. In as many as 30% of cases, a variant of uncertain significance is identified.8 The International Society for Gastrointestinal Hereditary Tumors (InSiGHT) has put together an expert panel to more properly classify variants of uncertain significance,9 but the proportion that remains unclassified represents an ongoing challenge.
Reports from reference labs performing clinical mutation testing generally provide excellent commentary on the meaning and limitations of variants of uncertain significance. Although such variants are occasionally reclassified as pathogenic or harmless and supplemental clinical lab reports provided, these nuances of interpretation underscore the importance of having access to genetic counselors. Genetic counselors can provide vital assistance in helping clinicians to make sense of it all and to explain the limitations to patients.
Because the genetic counseling ranks are well developed in the United States, relative to Europe, it is not surprising that ESMO does not comment on the role of genetic counselors. Nevertheless, ASCO and other U.S. guideline experts should, in this commentator’s view, provide a positive endorsement of the routine use of genetic counselors in managing high-risk patients.
Other Diagnostic Considerations
Most of the ESMO guidelines assume that genetic counseling and testing have been conducted and a mutation found. In the case of familial adenomatous polyposis, this is not really a critical matter, as clinical surveillance eventually demonstrates the presence or absence of polyp phenotype. But in HNPCC/Lynch syndrome and in patients with oligopolyposis not otherwise specified, inability to identify a definite underlying germline mutation, in many cases, leaves considerable uncertainty regarding diagnosis and, consequently, the approach to clinical surveillance.
In general, when microsatellite instability is present and is not attributable to methylation and/or shows BRAF mutation, HNPCC/Lynch syndrome is considered present even when no mutation is detected. Since this failure to detect germline mutation occurs in as many as half of all cases with informative microsatellite instability/immunohistochemistry, it is not a trivial consideration. First-degree relatives of such microsatellite instability–high but mutation nondiagnostic cases are considered at risk and screened “as if” they are carriers, since they cannot be proven not to be carriers. This is a challenging concept for such patients and their providers.
This problem of the microsatellite instability–high but mutation nondiagnostic case has recently become more intriguing in light of evidence that colorectal cancers, in some cases—typically those with little or no family history of HNPCC/Lynch syndrome tumors—may actually be sporadic as evidenced by apparently somatic mutations in both alleles of, say, MSH2, in the absence of methylation or BRAF mutation.10,11 Note this discussion is different from the situation in which no microsatellite instability is present but family history is strong, the so-called “familial colon cancer syndrome X.”
Need for Clinical Trials
Guidelines for screening and management of HNPCC/Lynch syndrome and for familial adenomatous polyposis show minimal disagreement between ASCO and ESMO. A distressing thread that runs through the guidelines is the inability to offer positive recommendations for clinical screening and intervention in a host of settings. While this lack of evidence base supports the conservative positions taken by ESMO/ASCO, it should serve to prompt the clinical research community to design trials that would provide an evidence base for or against, for example, extracolonic surveillance in HNPCC/Lynch syndrome.
Up to now, this would have been challenging due to the relatively small cohort of patients with known mismatch-repair mutations. However, a constellation of factors should be changing this and leading to such trials: (1) more routine microsatellite instability/immunohistochemistry for colorectal cancer and endometrial cancer; (2) availability of mutation testing; (3) resultant increasing number of mutation-positive families tracked by various registries or otherwise accessible; (4) increasing patient advocacy; and (5) the existence of collaborative societies—ESMO, ASCO, American Society of Colon and Rectal Surgeons, GI societies such as the American Gastroenterological Association and American Society for Gastrointestinal Endoscopy, Collaborative Group of the Americas on Inherited Colorectal Cancer, InSiGHT, Cooperative Family Registries, and corresponding consortia in Europe—that individually or collectively ought to be taking up the challenge.
I am optimistic that the data on genetic testing and testing strategies will continue to improve and refine clinical practice guidelines in this area. Clinical surveillance guidelines, requiring carefully designed clinical trials, will take much longer to develop. I would challenge the professional societies to take up the challenge so that the era of “no recommendation for or against” will come to an end. ■
Disclosure: Dr. Lynch reported no potential conflicts of interest.
References
1. Stoffel EM, Mangu PB, Grube SB, et al: Hereditary colorectal cancer syndromes: American Society of Cinical Oncology clinical practice guideline endorsement of the familial risk-colorectal cancer: European Society for Medical Oncology clinical practice guidelines. J Clin Oncol 33:209-217, 2015.
2. Balmana J, Balaguer F, Cervantes A, et al: Familial risk-colorectal cancer: ESMO clinical practice guidelines. Ann Oncol 24(suppl 6):vi73-vi80, 2013.
3. Kastrinos F, Steyerberg EW, Mercado R, et al: The PREMM(1,2,6) model predicts risk of MLH1, MSH2, and MSH6 germline mutations based on cancer history. Gastroenterology 140:73-81, 2011.
4. James PA, Doherty R, Harris M, et al: Optimal selection of individuals for BRCA mutation testing: A comparison of available methods. J Clin Oncol 24:707-715, 2006.
5. Dinh TA, Rosner BI, Atwood JC, et al: Health benefits and cost-effectiveness of primary genetic screening for Lynch syndrome in the general population. Cancer Prev Res 4:9-22, 2011.
6. Walsh T, Lee MK, Casadei S, et al: Detection of inherited mutations for breast and ovarian cancer using genomic capture and massively parallel sequencing. Proc Natl Acad Sci USA 107:12629-12633, 2010.
7. Hiraki S, Rinella ES, Schnabel F, et al: Cancer risk assessment using genetic panel testing: Considerations for clinical application. J Genet Couns 23:604-617, 2014.
8. Maxwell KN, Wubbenhorst B, D’Andrea K, et al: Prevalence of mutations in a panel of breast cancer susceptibility genes in BRCA1/2-negative patients with early-onset breast cancer. Genet Med. December 11, 2014 (early release online).
9. Thompson BA, Spurdle AB, Plazzer JP, et al: Application of a 5-tiered scheme for standardized classification of 2,360 unique mismatch repair gene variants in the InSiGHT locus-specific database. Nat Genet 46:107-115, 2014.
10. Carethers JM: Differentiating Lynch-like from Lynch syndrome. Gastroenterology 146:602-604, 2014.
11. Mensenkamp AR, Vogelaar IP, van Zelst-Stams WA, et al: Somatic mutations in MLH1 and MSH2 are a frequent cause of mismatch-repair deficiency in Lynch syndrome-like tumors. Gastroenterology 146:643-646.e8, 2014.
Dr. Lynch is Professor of Gastroenterology, Hepatology and Nutrition, The University of Texas MD Anderson Cancer Center, Houston.
