The advent of next-generation sequencing, and its transition further into the clinic with the US Food and Drug Administration approval of a cystic fibrosis assay in 2013, have increased the speed and reduced the cost of DNA sequencing. Coupled with a historic ruling by the Supreme Court of the United States that human genes are not patentable, these events have caused a seismic shift in genetic testing in clinical medicine. More labs are offering genetic testing services; more multigene panels are available for gene testing; more genes and gene mutations are being identified; and more variants of uncertain significance, which may or may not be clinically actionable, have been found. All these factors, taken together, are increasing the complexity of clinical management. While these developments have led to a greater interest in genetic testing, risk assessment, and large-scale population screening, they also present unique challenges. The dilemma for clinicians is how best to understand and manage this rapidly growing body of information to improve patient care. With millions of genetic variants of potential clinical significance and thousands of genes associated with rare but well-established genetic conditions, the complexities of genetic data management clearly will require improved computerized clinical decision support tools, as opposed to continued reliance on traditional rote, memory-based medicine.
In 2013, the US Food and Drug Administration (FDA) approved the use of next-generation DNA sequencing in clinical practice—specifically, a cystic fibrosis assay run on Illumina’s MiSeq instrument. Next-generation sequencing (NGS) refers to massively parallel, high-throughput DNA sequencing, and it has deployed a suite of technologies that have drastically increased the speed and decreased the cost of sequencing compared with the traditional Sanger method. The Supreme Court of the United States, also in 2013, invalidated the BRCA patent held by Myriad Genetics, ruling that human genes are not patentable. These two seminal events have created a seismic shift in the genetic testing world, accentuating forces already at work to markedly increase the use of genetic testing for cancer susceptibility in clinical medicine, and companies are vying to make less expensive, more efficient genetic tests.
These profound changes in genetic testing have altered the practice of medicine (Figure 1). It took 13 years and cost $2.7 billion to sequence the first human reference genome. Today, using NGS, a genome can be sequenced for $1,245 in under a week, or in 26 hours for approximately $6,500. As of this writing, the upcoming launch of a $999 genome with clinical grade sequencing coverage and interpretation, and with short turnaround time, has been announced.
In this rapidly evolving field, rather than asking, “What’s new in genetic testing?” perhaps the better question is, “What isn’t new?” New realities include but are not limited to:
• More laboratories doing cancer genetic testing.
• More panels containing varying combinations of genes.
• More genes to understand and manage, with more incidental findings.
• More variants of uncertain significance (VUSs).
• More interest in gene testing.
• More to cover in counseling.
• More interest in population screening.
• More need for clinical decision support (CDS).
More Laboratories Doing Cancer Genetic Testing
Almost immediately after the Supreme Court’s 2013 decision, multiple laboratories began to market BRCA testing, as well as testing for other cancer genes (Table 1). The challenge for physicians, nurse practitioners, and genetic counselors was and remains how to select the appropriate laboratory for a given situation. In choosing between labs, one must consider test accuracy, types of panels offered, accuracy and thoroughness of the interpretation, support in determining management of mutation carriers, coverage by insurance companies, and cost both to the healthcare system and to the patient (in potential out-of-pocket costs). Each clinician and/or health system must weigh these costs and benefits to choose the best laboratory for the situation.
More Panels Containing Varying Combinations of Genes
In the past, a patient at risk for hereditary breast and ovarian cancer (HBOC) syndrome was initially tested solely for BRCA1 and BRCA2. If the results of these tests were negative, the patient’s risk of having other mutations, such as PTEN or TP53, would be assessed and additional tests ordered. This “diagnostic odyssey” (as it is known in pediatric genetics) was expensive, time consuming, and frustrating for both the patient and the clinician; in addition, it was often stopped prematurely, missing the actual gene that was involved.
The marked decrease in the cost of DNA sequencing, and the large number of genes now identified as increasing a person’s cancer susceptibility (Table 2), have led to the development of panel testing, in which large numbers of genes are tested simultaneously at a cost similar to or even less than the cost of sequencing just one or two genes. Various laboratories now offer a variety of panels (see Table 1), from disease-specific panels, such as a breast panel, to “pan-cancer” panels, which test multiple genes that may increase cancer susceptibility for a variety of cancers and inherited cancer syndromes. However, not all these gene mutations have corresponding clinical management guidelines, and this may pose a significant dilemma for physicians.
While the experts have urged caution in using these tests, clinicians have moved ahead in large numbers. Many practices now routinely order cancer panels as first-line genetic testing, and recent publications in the breast cancer literature are reporting identification of actionable mutations by sequencing more broadly across a larger number of genes.[7-9] Actionable mutations in cancer genetics have been defined as those that “have significant diagnostic, prognostic, or therapeutic implications in subsets of cancer patients and for specific therapies.” In breast cancer, patients with actionable mutations may consider additional breast imaging and/or elective risk-reducing surgery (Figure 2). However, it should be noted that in this new era, almost no randomized clinical trials have been performed, and “actionable” is determined by expert opinion. In addition to actionable genes, most panels include genes that are less well characterized or that lack management guidelines. This is undoubtedly confusing for both the patient and the physician. As the scale of testing increases, and as more clinical outcomes data accumulate and are published (or at least shared via a growing number of online variant databases), it is hoped that either the discovery of mutations in these genes will be found to have utility or that these genes will be removed from the panels. In addition, it is important to recognize that finding mutations does not always lead to changes in management recommendations, and caution should be exercised when attributing causality to genes with low or moderate penetrance.
Data are accumulating rapidly, as demonstrated by the cumulative results of panel testing as it relates to HBOC syndromes. Tung (two series), Walsh, and Couch have reported the results of panel tests in patients who were at risk for HBOC syndromes and had not yet undergone BRCA testing.[7,8,12,13] Desmond and Maxwell reported on patients at risk for HBOC syndromes who had previously undergone BRCA testing.[9,14] With a small amount of mathematical manipulation (the number of patients who were BRCA-positive in the Desmond and Maxwell series was extrapolated by multiplying the number tested by 1.1155 [based on the percentage of patients who were BRCA-positive on de novo testing, 10.37%—see below]), we have combined these series into a summary of the results from this type of panel testing (Table 3, Figure 3). In those patients who underwent de novo testing, 10.3% were BRCA-positive and 5.41% were positive for other genes. This highlights the fact that about one-third of mutation carriers will be missed if only BRCA1/2 is tested (or that 50% more carriers will be found if panel testing is performed). When BRCA results are excluded, the series can be combined as second-tier testing to show that 6% of patients who are BRCA-negative will test positive for other genes (Table 3). These results raise an important question with regard to the hundreds of thousands of patients who tested as BRCA-negative in the pre-panel era. Should all of these patients be retested now with panels? Or should we only retest the highest-risk patients? Will insurers pay for this additional testing? Obviously many questions remain.
These studies highlight the potential benefit of multigene panels in identifying patients and families with clinically significant mutations that may alter their management recommendations. It almost goes without saying that the current support systems must be extended to help patients who carry such mutations cope with their new diagnosis and the potential associated psychological, financial, and family strains.
Desmond et al took their analysis a step further by looking at the potential actionability of panel results. They re-evaluated the 1,046 HBOC candidates who had tested negative for BRCA1/2 mutations using multigene panel testing, and found that 40 of the subjects (3.8%) had deleterious mutations in non-BRCA genes. In terms of the importance of identifying mutations in non-BRCA genes and potential actionability, the study was expanded to include 23 more patients outside the original cohort in whom mutations were found using gene panels. When all these mutation carriers were taken together, 92% of the patients had mutations correlating with cancers found either in themselves or in their families (strengthening the suggested relationship between these genes and diseases); 33 patients (52%) required a change in management (28 increases in cancer screening, 1 prophylactic colectomy, and 4 prophylactic gastrectomies); and testing of relatives was indicated for 72%, owing to potential changes in management for those relatives. Compared with BRCA1/2 testing, multigene panel testing finds a larger number of patients who may require a change in their clinical management and that of their relatives. For some, however, this may lead to more questions than answers, and it may be unclear what the long-term risks and benefits of these findings will be for any given individual.
More Genes to Understand and Manage, With More Incidental Findings
Increasing numbers of genes related to a variety of cancers have been identified (see Table 2). Mutations in these genes have different incidence rates in the population, cause different spectrums of cancers in a family, and have different levels of penetrance. The same gene mutation can cause different cancers, and the same cancer can be caused by different gene mutations. The permutations are astronomical. The observation has been made that there is an inverse relationship between incidences of mutations in the population and cancer penetrance. When evaluating a patient with a strong family history of HBOC, we can no longer consider just BRCA1 and BRCA2. We must consider other genes as well (see Table 2).
Easton et al effectively summarized our current knowledge about the penetrance of major breast cancer genes. The lifetime risk of developing breast cancer ranges from 75% for BRCA1 to 23% for NBN; the lifetime risk of developing breast cancer for an average woman is 12.4%. This difference in penetrance clearly influences the management recommendations for unaffected women: whereas MRI screening might be suggested for both mutations, risk-reducing mastectomies should only be strongly considered for BRCA1/2, and not NBN (see Figure 2). Because we are now identifying genes of moderate-to-low penetrance, and with different spectra of associated disease, our strategies must be flexible. Medical management should be informed by the penetrance of the gene mutation and by the age, family history, gender, and health of the patient; whether he or she has cancer or not; and if so, what type and stage of cancer it is. A woman in her twenties with a BRCA1 mutation who does not have cancer should have breast MRI screening, and in her late thirties, she should strongly consider prophylactic bilateral oophorectomy and might consider prophylactic mastectomies. If she develops breast cancer, regardless of age, she should consider bilateral mastectomies.[18,19] In comparison, a male BRCA1 carrier will likely do little with regard to this mutation other than inform his relatives and increase prostate screening.[18,19] In yet another scenario, a woman with a CHEK2 mutation might consider breast MRI screening, while prophylactic mastectomies would be discouraged per established guidelines[18,19]; oophorectomy or ovarian screening would be unnecessary, as that is not part of the disease spectrum associated with this mutation. It should be noted that the guidelines for CHEK2 are based on expert opinion, as no outcomes data for MRI screening currently exist. In addition, it is important to recognize that this field is continuously evolving, and recommendations may change frequently as more outcomes are reported.
As we test more and more genes, we will encounter more incidental findings. That is, we will identify gene mutations that we did not anticipate based on the family history. Because many panels include genes for multiple cancers, we now routinely see patients who are being tested for a hereditary breast cancer syndrome who turn out to have a hereditary colon cancer syndrome instead, and vice versa.
Frey et al evaluated 127 patients who underwent noninformative primary genetic screening and subsequent multigene panel testing. Four patients were found to have pathogenic mutations that could be characterized as “incidental” in the context of their personal and family histories. Similarly, Yurgelun et al performed multigene panel testing in 1,260 patients with suspected Lynch syndrome. They reported that 71 patients (5.6%) had a clinically unsuspected mutation in a non–Lynch syndrome cancer susceptibility gene, most commonly BRCA1/2. Of note, the long-term implications of these findings have yet to be verified. Regardless, treating these mutation carriers by established guidelines would be prudent.
TO PUT THAT INTO CONTEXT
Phuong L. Mai, MD, MS
Magee-Womens Hospital, University of Pittsburgh Cancer Institute
How Has the Landscape of Cancer Genetic Testing Changed, and What Are Some of the Implications of These Changes?
Genetic testing for hereditary cancer susceptibility has evolved rapidly over the last decade. The introduction of next-generation sequencing (NGS) technologies has made possible the simultaneous testing of multiple genes that may contribute to inherited cancer risk (so-called “gene panel testing”), at an affordable cost. Many laboratories are now offering a variety of cancer gene panels. However, the inconsistency in the gene selection and laboratory interpretation of pathogenicity, coupled with the lack of data regarding phenotype and the absence of clinical management guidelines for many of the genes in these panels, has complicated the testing process. For most of the low- to moderate-penetrance genes included, the utility of the test results is uncertain. In the absence of established penetrance and an established spectrum of associated cancers, the ability to make management recommendations for individuals carrying a pathogenic variant is limited.
What Caveats Should Be Kept in Mind When Considering Use of Gene Panel Testing?
Although it is reasonable to consider cancer gene panel testing in many scenarios, caution must be exercised before embracing panel testing for all patients being evaluated for cancer risk. Gene panel testing may miss some mutations detectable by traditional single-gene analytic methods but not with NGS technologies. In addition, adequate pretest counseling must be provided to ensure that patients are making informed decisions regarding testing. Furthermore, it is unclear what the long-term outcomes associated with panel testing are, given the uncertainty in risk estimates and the lack of established effective management strategies for many of the moderate-penetrance genes.
More precise cancer risk estimates and better-defined spectra of associated cancers are needed in order to establish clinically/medically meaningful risk management strategies for individuals in whom gene panel testing identifies a mutation.
Although incidental findings can be distressing for the patient and the clinician, it may be in the best interest of the patient to identify these syndromes before a cancer occurs. While controversial, the American College of Medical Genetics and Genomics (ACMG) guidelines suggest that if an incidental finding is potentially actionable, it should be reported to the patient. The ACMG produced a minimum list of 56 genes they considered theoretically actionable, and this list is expected to grow rapidly. More specifically, they sought to include “conditions for which confirmatory approaches for medical diagnosis would be available…disorders for which preventive measures and/or treatments were available, and disorders in which individuals with pathogenic mutations might be asymptomatic for long periods.”
Genes have variations in their DNA sequences. Many of these variants are benign, such as those that result in a person having blue eyes vs brown eyes, while others are pathogenic and markedly increase the risk of disease. Mutations in or variants of cancer susceptibility genes can be classified into five categories: benign, likely benign, VUS, likely pathogenic, and pathogenic.
Classification of variants can sometimes be easy, such as when the mutation in a gene causes truncation of the protein whose pathogenicity is known to be mediated through loss of function; others are harder to classify, such as when the variant is a change in a single amino acid in the protein, and the effect of that change is not obvious. If it is not possible to determine whether a mutation is disease-causing, the variant is classified as a VUS. In this case, the clinician must rely on the family history to determine the appropriate management. VUSs can be challenging for both the clinician and the patient. One would expect, and it has been shown, that for genes that are rarely tested, the VUS rates will be markedly higher, because there is less clinical and family history data to link with the mutation. However, as more mutations are identified within families that have known clinical diseases/outcomes, the phenotypes can be more clearly defined. Eggington et al showed that the VUS rate for BRCA1 and BRCA2 declined by 84% between 2002, when the rate was 12.8%, and 2013, by which time the rate had dropped to 2.1%, mostly due to the markedly increased rate of testing.
Based on the studies mentioned previously, Frey et al found a 42% VUS rate when panel testing was undertaken. In both cohorts studied by Tung et al, around 40% of individuals had at least one VUS. Yurgelun et al detected one or more VUSs in 38% of their cohort. The most common VUSs were found in ATM, APC, NBN, and BRIP1. The PROMPT Registry was initiated to help classify the VUSs being found in the plethora of genes that are now being tested, as well as to improve our understanding of the penetrance of these genes. This registry allows patients who have a variant in one of the less-studied genes to enter their data online into a central registry.
As more and more testing is performed for more and more genes, we anticipate that many VUSs will be reclassified, and the number of VUSs for each gene tested will eventually drop significantly. In the interim, it is important to emphasize that VUS findings should not be used to determine the management of patients. In these patients, management recommendations must be informed by the family history and other patient factors.
1. Veritas Genetics. Our latest news. https://www.veritasgenetics.com/#sec-4. Accessed April 2, 2016.
2. Ellinger MS. Supreme Court decision in Association for Molecular Pathology v. Myriad Genetics, Inc. June 13, 2013. http://www.fr.com/news/myriad-decision. Accessed February 18, 2016.
3. National Institutes of Health; National Human Genome Research Institute. The Human Genome Project completion: frequently asked questions. https://www.genome.gov/11006943. Accessed February 18, 2016.
4. National Institutes of Health; National Human Genome Research Institute. The cost of sequencing a human genome. https://www.genome.gov/sequencingcosts. Accessed February 18, 2016.
5. Miller NA, Farrow EG, Gibson M, et al. A 26-hour system of highly sensitive whole genome sequencing for emergency management of genetic diseases. Genome Med. 2015;7:100.
6. Mauer CB, Pirzadeh-Miller SM, Robinson LD, Euhus DM. The integration of next-generation sequencing panels in the clinical cancer genetics practice: an institutional experience. Genet Med. 2014;16:407-12.
7. Tung N, Battelli C, Allen B, et al. Frequency of mutations in individuals with breast cancer referred for BRCA1 and BRCA2 testing using next-generation sequencing with a 25-gene panel. Cancer. 2015;121:25-33.
8. Walsh T, Casadei S, Lee MK, et al. Mutations in 12 genes for inherited ovarian, fallopian tube, and peritoneal carcinoma identified by massively parallel sequencing. Proc Natl Acad Sci USA. 2011;108:18032-7.
9. Desmond A, Kurian AW, Gabree M, et al. Clinical actionability of multigene panel testing for hereditary breast and ovarian cancer risk assessment. JAMA Oncol. 2015;1:943-51.
10. Dancey JE, Bedard PL, Onetto N, Hudson TJ. The genetic basis for cancer treatment decisions. Cell. 2012;148:409-20.
11. Axilbund JE. Panel testing is not a panacea. J Clin Oncol. 2016;34:1433-5.
12. Couch FJ, Hart SN, Sharma P, et al. Inherited mutations in 17 breast cancer susceptibility genes among a large triple-negative breast cancer cohort unselected for family history of breast cancer. J Clin Oncol. 2015;33:304-11.
13. Tung N, Lin NU, Kidd J, et al. Frequency of germline mutations in 25 cancer susceptibility genes in a sequential series of patients with breast cancer. J Clin Oncol. 2016;34:1460-8.
14. 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. 2015;17:630-8.
15. Ambry Genetics. Hereditary cancer panels: clinician guide. http://www.ambrygen.com/sites/default/files/pdfs/Hereditary_Cancer_Tabbed_Clinician_0914_final_web.pdf. Accessed February 27, 2016.
16. Easton DF, Pharoah PD, Antoniou AC, et al. Gene-panel sequencing and the prediction of breast-cancer risk. N Engl J Med. 2015;372:2243-57.
17. National Institutes of Health; National Cancer Institute. SEER cancer statistics review, 1975-2013. http://seer.cancer.gov/csr/1975_2013/. Accessed August 4, 2016.
18. Daly M, Pilarski R, Axilbund J, et al. NCCN clinical practice guidelines in oncology: genetic/familial high-risk assessment: breast and ovarian. Version 2.2015. J Natl Compr Canc Netw. 2016;14:153-62.
19. Gradishar WJ, Anderson BO, Balassanian R, et al. NCCN clinical practice guidelines in oncology: breast cancer. Version 1.2016. J Natl Compr Canc Netw. 2016;14:324-54.
20. Frey MK, Kim SH, Bassett RY, et al. Rescreening for genetic mutations using multi-gene panel testing in patients who previously underwent non-informative genetic screening. Gynecol Oncol. 2015;139:211-5.
21. Yurgelun MB, Allen B, Kaldate RR, et al. Identification of a variety of mutations in cancer predisposition genes in patients with suspected Lynch syndrome. Gastroenterology. 2015;149:604-13.
22. Green RC, Berg JS, Grody WW, et al. ACMG recommendations for reporting of incidental findings in clinical exome and genome sequencing. Genet Med. 2013;15:565-74.
23. Emory Genetics Laboratory. Emory Genetics Laboratory classification definitions. http://geneticslab.emory.edu/emvclass/EGLClassificationDefinitions.php. Accessed March 26, 2016.
24. Eggington JM, Bowles KR, Moyes K, et al. A comprehensive laboratory-based program for classification of variants of uncertain significance in hereditary cancer genes. Clin Genet. 2014;86:229-37.
25. Patient Crossroads. Prospective registry of MultiPlex testing (PROMPT). https://connect.patientcrossroads.org/?org=prompt. Accessed February 27, 2016.
26. Drohan B, Roche CA, Cusack JC Jr, Hughes KS. Hereditary breast and ovarian cancer and other hereditary syndromes: using technology to identify carriers. Ann Surg Oncol. 2012;19:1732-7.
27. Polubriaginof FC, Drohan B, Bosinoff P, et al. Implications of following the guidelines for genetic testing and MRI use for breast cancer. J Clin Oncol. 2014;32(suppl 5s):abstr 1549.
28. Acheson LS, Wiesner GL, Zyzanski SJ, et al. Family history-taking in community family practice: implications for genetic screening. Genet Med. 2000;2:180-5.
29. Wood ME, Kadlubek P, Pham TH, et al. Quality of cancer family history and referral for genetic counseling and testing among oncology practices: a pilot test of quality measures as part of the American Society of Clinical Oncology Quality Oncology Practice Initiative. J Clin Oncol. 2014;32:824-9.
30. Jolie A. My medical choice. New York Times. May 14, 2013 (A25).
31. Evans DG, Barwell J, Eccles DM, et al. The Angelina Jolie effect: how high celebrity profile can have a major impact on provision of cancer related services. Breast Cancer Res. 2014;16:442.
32. Evans DG, Wisely J, Clancy T, et al. Longer term effects of the Angelina Jolie effect: increased risk-reducing mastectomy rates in BRCA carriers and other high-risk women. Breast Cancer Res. 2015;17:143.
33. McCarthy AM, Bristol M, Fredricks T, et al. Are physician recommendations for BRCA1/2 testing in patients with breast cancer appropriate? A population-based study. Cancer. 2013;119:3596-603.
34. Brown KL, Hutchison R, Zinberg RE, McGovern MM. Referral and experience with genetic testing among women with early onset breast cancer. Genet Test. 2005;9:301-5.
35. Ruddy KJ, Gelber S, Shin J, et al. Genetic testing in young women with breast cancer: results from a Web-based survey. Ann Oncol. 2010;21:741-7.
36. Vig HS, McCarthy AM, Liao K, et al. Age at diagnosis may trump family history in driving BRCA testing in a population of breast cancer patients. Cancer Epidemiol Biomarkers Prev. 2013;22:1778-85.
37. Bellcross CA, Kolor K, Goddard KA, et al. Awareness and utilization of BRCA1/2 testing among U.S. primary care physicians. Am J Prev Med. 2011;40:61-6.
38. Guerra CE, Sherman M, Armstrong K. Diffusion of breast cancer risk assessment in primary care. J Am Board Fam Med. 2009;22:272-9.
39. Christianson CA, Powell KP, Hahn SE, et al. The use of a family history risk assessment tool within a community health care system: views of primary care providers. J Genet Couns. 2012;21:652-61.
40. Sabatino SA, McCarthy EP, Phillips RS, Burns RB. Breast cancer risk assessment and management in primary care: provider attitudes, practices, and barriers. Cancer Detect Prev. 2007;31:375-83.
41. Rosenberg SM, Ruddy KJ, Tamimi RM, et al. BRCA1 and BRCA2 mutation testing in young women with breast cancer. JAMA Oncol. 2016;2:730-6.
42. King MC, Levy-Lahad E, Lahad A. Population-based screening for BRCA1 and BRCA2: 2014 Lasker Award. JAMA. 2014;312:1091-2.
43. Robson ME, Bradbury AR, Arun B, et al. American Society of Clinical Oncology policy statement update: genetic and genomic testing for cancer susceptibility. J Clin Oncol. 2015;33:3660-7.
44. McPherson E, Zaleski C, Benishek K, et al. Clinical genetics provider real-time workflow study. Genet Med. 2008;10:699-706.
45. Armstrong J, Toscano M, Kotchjo, N, et al. Utilization and outcomes of BRCA genetic testing and counseling in a national commercially insured population. JAMA Oncol. 2015;1:1251-60.
46. Narod S. Genetic testing for BRCA mutations today and tomorrow—about the ABOUT study. JAMA Oncol. 2015;1:1225-6.
47. US Department of Health and Human Services Advisory Committee on Heritable Disorders in Newborns and Children. Recommended uniform screening panel. http://www.hrsa.gov/advisorycommittees/mchbadvisory/heritabledisorders/recommendedpanel/ Accessed August 18, 2016.
48. G2P: Research in Translational Genomics and Health Outcomes. The BabySeq project: genome sequence–based screening for childhood risk and newborn illness. http://www.genomes2people.org/babyseqproject. Accessed January 27, 2016.
49. AuthorMapper. BRCA testing from 1994. http://www.authormapper.com/search.aspx?q=BRCA1%20testing%20from%201994. Accessed February 28, 2016.
50. National Center for Biotechnology Information Genetic Testing Registry. http://www.ncbi.nlm.nih.gov/gtr. Accessed February 20, 2016.
51. Crane RM, Raymond B. Fulfilling the potential of clinical information systems. Permanente J. 2003;7:62-7.
52. Welch BM, Kawamoto K, Drohan B, Hughes KS. Clinical decision support for personalized medicine. In: Greenes R, editor. Clinical decision support: the road to broad adoption. 2nd ed. Burlington, MA: Elsevier; 2014. p. 383-413.
53. Ambry Genetics—search. http://www.ambrygen.com/search/node. Accessed April 10, 2016.
54. Myriad mySupport360. https://ms360.myriad.com. Accessed April 10, 2016.