1

© American College of Medical Genetics and Genomics Commentary

Genetic testing for inherited disease susceptibility is a power-ful tool that helps to reduce morbidity and mortality associated with familial cancer syndromes. The identification of individu-als with germ-line mutations in BRCA1 and BRCA2 (associated with hereditary breast and ovarian cancers); MLH1, MSH2, MSH6, and PMS2 (associated with Lynch syndrome); and RET (associated with familial medullary thyroid cancer), among many others, has led to targeted screening and interventions that can lead to reductions in morbidity, and in some cases, mortality.1,2 Although it is generally assumed that familial cancer syndromes are largely identifiable through a thorough assessment of the family history, there are overlaps in cancer susceptibility syndromes (e.g., ovarian cancer can be due to either a BRCA1/2 or a Lynch syndrome–associated mutation); moreover, family history can be limited by adoption, small family size, and poor communication within families. Genetic evaluation usually has proceeded in a serial fashion, starting with the most likely genes and then, if testing for these is nega-tive, consideration of additional testing. However, this process can lead to “testing fatigue” in both providers and patients, incomplete evaluation when patients do not return to complete the recommended testing, and considerable lengthening of the time frame required for testing. Against this background, advances in technology have allowed development of multi-plex gene panels in which many genes (from 6 to 110) can be assessed simultaneously by massively parallel sequencing. The potential advantage of this approach is an efficient and timely evaluation that may be only marginally more expensive than standard-of-care genetic testing, or in the case of serial test-ing, potentially less expensive. However, little is known to date about key aspects of such testing, including the likelihood of a positive result, the rates of variants of unknown significance, and the clinical utility of testing using multiplex panels.

In this issue of Genetics in Medicine, LaDuca et al.3 describe their initial results obtained from tests using four multiplex panels performed in a Clinical Laboratory Improvement Amendments–certified commercial laboratory (Ambry Genetics) on a large series of individuals. These panels, simi-lar to those offered by other commercial laboratories (Invitae,

GeneDx, Fulgent Diagnostics, University of Washington, and Myriad Genetics), contain a mix of high-penetrance and pre-sumed moderate-penetrance genes. In general, guidelines for clinical management are available for individuals with muta-tions in high-penetrance, but not in moderate-penetrance, genes. It is worth noting that BRCA1 and BRCA2 were not ini-tially included in these versions of the Ambry panels due to pat-ent concerns (although they are currently included). Questions and concerns regarding the use of these panels have been previously published4 and include (i) finding unanticipated high-penetrance mutations (e.g., detecting a TP53 mutation in a family without features consistent with Li–Fraumeni syn-drome); (ii) managing high rates of variants of uncertain signif-icance (VUSs); (iii) determining the clinical utility of mutations in moderate-penetrance susceptibility genes; and (iv) the ratio-nale for the selection of genes on the panels because mutations in some have not been definitively associated with susceptibility either for cancer at all (e.g., XRCC2—not included in the Ambry Genetics panel but part of other commercially available panels) or the cancer for which the panel is intended (e.g., MUTYH in a breast cancer panel).

The study by LaDuca et al.3 provides critical data to begin to answer these important questions. Of 2,079 patients, 8.3% had deleterious mutations in the genes tested (deter-mined by the proprietary Ambry Variant Analyzer accord-ing to the guidelines of the American College of Medical Genetics and Genomics), with frequencies ranging from 7.2% of those tested with the ovarian cancer panel to 9.6% with the “cancer” panel. Of the total, importantly, only 71 patients (3.4%) had mutations in genes for which clinical guidelines are available (TP53, PTEN, MLH1, MSH2, MSH6, PMS2, SMAD4, APC, biallelic MYH, CDH1, and STK11). The majority (47/74 (64%)) of these mutations were found in the colon cancer panel and had corresponding clinical guide-lines. Of the patients tested with the colon cancer panel, 8% (46/557) had a positive clearly clinically actionable result, as compared with <1% (7/874) with such a result detected using the breast cancer panel. This result is not surprising because of the number of genes in which mutations are known to

Submitted 10 April 2014; accepted 21 April 2014; advance online publication 3 July 2014. doi:10.1038/gim.2014.56

Genet Med

00

00

2014

Genetics in Medicine

10.1038/gim.2014.56

Commentary

00

00

10April2014

21April2014

© American College of Medical Genetics and Genomics

3July2014

1Division of Hematology/Oncology, Department of Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA; 2Basser Research Center for BRCA, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA; 3Abramson Cancer Center, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA; 4Division of Translational Medicine and Human Genetics, Department of Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA. Correspondence: Katherine L. Nathanson (knathans@exchange.upenn.edu)

Panel testing for inherited susceptibility to breast, ovarian, and colorectal cancer

Susan M. Domchek, MD1–3 and Katherine L. Nathanson, MD2–4

Genetics in medicine

2

DOMCHEK and NATHANSON | Panel testing for inherited susceptibility to cancerCommentary

confer a greatly increased risk of colorectal cancer and thus illustrates a much higher clinical utility for panel testing in this context. However, the utility of identifying mutations in moderate-penetrance genes even on the colon cancer panel remains unclear.

Unanticipated findings in high-penetrance genes can be viewed as both an advantage and a disadvantage of multiplex testing. As an example, the authors report that 30% of patients with mutations in the colon panel would not meet the criteria for clinical testing. However, this result must be interpreted with caution because neither all the family history nor all the patient information may have been reported; moreover, the particular mutations that were detected were not specified. The counterargument is that the detection of unanticipated high-penetrance mutations associated with specific syndromes—such as mutations in (i) CDH1 (associated with hereditary diffuse gastric cancer), for which the standard recommenda-tion would be prophylactic gastrectomy, and (ii) TP53 (linked to Li–Fraumeni syndrome), which is associated with a partic-ularly high morbidity and mortality—are challenging to man-age in the absence of a supportive family history. Interestingly, no STK11 (Peutz–Jehgers syndrome) mutations were iden-tified using any panel, suggesting that it is recognized only clinically. In this cohort, 10 TP53 and 2 CDH1 mutations were detected. Three of the TP53 mutation carriers were diagnosed with breast cancer before the age of 35 years, so genetic testing would have been warranted clinically. Pretest counseling for panel mutation testing must include a discussion of the low, but real, likelihood of an unanticipated finding that could have major implications for medical management. On the basis of previous work in presymptomatic testing for TP53 mutations in family members of those with known mutations, 45% of individuals decline testing.5 Counseling models for multiplex testing must incorporate this consideration.

Most of the deleterious mutations found in this series were in moderate-penetrance genes for which there are no clinical management guidelines and thus no clear clinical utility. Even for genes about which much is known (such as CHEK2), how to use mutational information for clinical care is uncertain. In addition, the discovery of a mutation in such a moderate-penetrance gene may not provide a compelling explanation for the family history that may have prompted testing in the first place. Cancer penetrance associated with CHEK2 may be influenced by specific mutations (e.g., CHEK2 1100delC versus I157T),6 as well as the associated family history.7 Risks of other cancers, such as colon cancer, appear to be modestly elevated, but whether this should alter screening recommendations (including age at which to begin or frequency of screening) is unknown. For many of the other genes (such as NBN, BRIP1, and MRE11), even less is known about their true association with cancer susceptibility. Clinical genetic testing should be considered in the context of its use as a tool to improve the medical management of patients. Thus, because the identifica-tion of deleterious mutations in moderate-penetrance genes currently does not influence medical management, their

inclusion in commercially available genetic testing panels should be carefully considered going forward.

As anticipated, VUSs were frequently observed, with rates of 15–26% (approximately double the deleterious mutation rate). Critical for ordering providers is the need to prepare patients for the possibility of this type of result. Clinical management remains based on the family history in this set-ting. Providers and testing laboratories also need to establish methods to reexamine these results periodically, particularly VUSs in high-penetrance genes. In addition, programs that offer tracking of VUSs in families with disease, without bur-dening the patient with additional cost, are of great value in providing additional information to determine pathogenicity. Although the high rate of VUSs is of concern, it is anticipated that with additional testing, rates will decline as they have with BRCA1/2 testing. However, it may be more difficult to determine the significance of VUSs in moderate-penetrance genes, adding to complexity and potential confusion for both patient and provider.

Multiplex panel testing for cancer susceptibility holds promise, and the study by LaDuca et al.3 has provided us with important initial information. The VUS rate is reasonable for this early stage of panel use. The yield of mutations in genes with associated clinical management guidelines, particularly in breast cancer patients, is low but is appreciable in those evaluated for colon cancer. Panel testing for inherited muta-tions may equip non–cancer geneticists with a mechanism with which they can identify individuals who carry a mutation associated with a cancer susceptibility syndrome so that they can be appropriately referred to cancer genetics centers, which may be particularly important for diseases with high inher-ited mutation rates, such as pheochromocytoma.8 However, the inclusion of moderate-penetrance genes in these panels remains a concern because currently the identification of del-eterious mutations in these genes does not (and should not, without proper evidence) lead to changes in medical manage-ment. The results of this study emphasize the need for interna-tional collaborative efforts to study rare moderate-penetrance mutations and to develop standard-of-care guidelines based on accurate estimates of cancer risk for the benefit of both patients and their providers.

ACKNOWLEDGMENTSThe authors acknowledge funding support from the Breast Cancer Research Foundation (K.L.N.), the Rooney Family Founda-tion (K.L.N. and S.M.D.), the Basser Center for BRCA Research at the University of Pennsylvania (K.L.N. and S.M.D.), the MacDonald Cancer Risk Evaluation Program (K.L.N. and S.M.D.), the Susan G Komen Foundation (S.M.D.), and the CURE (Commonwealth Uni-versal Research Enhancement) Program (K.L.N.). The Pennsylvania Department of Health specifically disclaims responsibility for any analyses, interpretations, or conclusions.

DISCLOSUREThe authors declare no conflict of interest.

Genetics in medicine

3

Panel testing for inherited susceptibility to cancer | DOMCHEK and NATHANSON Commentary

REfERENCES 1. Domchek SM, Friebel TM, Singer CF, et al. Association of risk-reducing surgery

in BRCA1 or BRCA2 mutation carriers with cancer risk and mortality. JAMA 2010;304:967–975.

2. Finch AP, Lubinski J, Møller P, et al. Impact of oophorectomy on cancer incidence and mortality in women with a BRCA1 or BRCA2 mutation. J Clin Oncol 2014; e-pub ahead of print 24 February 2014.

3. LaDuca H, Stuenkel AJ, Dolinsky JS, et al. Utilization of multigene panels in hereditary cancer predisposition testing: analysis of more than 2,000 patients. Genet Med 2014; e-pub ahead of print 24 April 2014.

4. Domchek SM, Bradbury A, Garber JE, Offit K, Robson ME. Multiplex genetic testing for cancer susceptibility: out on the high wire without a net? J Clin Oncol 2013;31:1267–1270.

5. Lammens CR, Aaronson NK, Wagner A, et al. Genetic testing in Li-Fraumeni syndrome: uptake and psychosocial consequences. J Clin Oncol 2010;28: 3008–3014.

6. Cybulski C, Wokolorczyk D, Kladny J, et al. Germline CHEK2 mutations and colorectal cancer risk: different effects of a missense and truncating mutations? Eur J Hum Genet 2007;15:237–241.

7. Cybulski C, Wokolorczyk D, Jakubowska A, et al. Risk of breast cancer in women with a CHEK2 mutation with and without a family history of breast cancer. J Clin Oncol 2011;29:3747–3752.

8. F i shbe in   L , Merr i l l   S , F raker   DL, Cohen  DL, Nathanson  KL. Inherited  mutations in pheochromocytoma and paraganglioma: why all patients should be offered genetic testing. Ann Surg Oncol 2013;20:1444–1450.

Genetics in medicine