User login
Dear colleagues and friends,
I write to introduce to you the new Perspectives section of GI & Hepatology News.
A more appropriate description is perhaps old-new, because Perspectives is the continuation and legacy of AGA Perspectives, the content of which has been consolidated into GI & Hepatology News. Perspectives will continue to feature the point/counterpoint expert debates about an important GI topic, which has historically been immensely popular with readers. In this edition, experts from Mayo Clinic and Cleveland Clinic discuss the pros and cons of universal multigene panel testing for colorectal cancer. These debates never end with the publication itself, and I hope they will continue to stimulate further thought and discussion. As always, I welcome your comments and suggestions for future topics.
–Charles I. Kahi, MD, MS, AGAF, is professor of medicine at Indiana University School of Medicine, Indianapolis. He is also an Associate Editor for GI & Hepatology News.
For everyone
By N. Jewel Samadder, MD, MSC
Traditionally, health care structure has been directed predominantly toward treatment rather than prevention. Advances in genomic medicine offer the opportunity to deliver a more personalized, predictive, and preventive strategy toward colorectal cancer. Approximately 150,000 men and women are diagnosed with colorectal cancer (CRC) every year in the United States.1 An estimated 10%-15% of these cancers are likely attributable to hereditary (germline) causes.2 Several genes are associated with an increased risk of developing CRC, and those of key interest include those for Lynch syndrome, MLH1, MSH2, MSH6, PMS2, EPCAM; adenomatous polyposis conditions (APC), MUTYH, POLE, POLD1, NTHL1; hamartomatous polyposis syndromes PTEN, SMAD4, STK11, and other rare cancer predisposition states where colorectal cancer is part of the phenotype, CHEK2 and TP532.
A universal strategy for multigene panel testing in all patients with CRC is an option versus the current strategy of guideline-based testing using family history and tumor features. In addition, the identification of germline alterations has substantial clinical implications including targeted therapies and future cancer prevention in the patient and relatives. This article will focus on the benefits of universal strategy for germline genetic evaluation in all patients with colorectal cancer.
The role and utility of current guideline-based testing
Given the therapeutic and prevention implications, the National Comprehensive Cancer Network (along with other professional organizations) has guidance on when patients with CRC should undergo genetic evaluation.3 Currently, these guidelines advocate an approach based heavily on family cancer history or utilizing colorectal phenotype based on the number and histology of polyps or tumor-based molecular features. Although family history is important for the diagnosis of hereditary CRC, the ability to accurately capture extended family cancer history in routine practice, from multiple generations and for different cancer types can be a challenge. The largest drawback of all such approaches is the focus on Lynch syndrome or only a few of the cancer predisposition syndromes. Recent studies have reported a substantial number (7%-10%) of CRC patients will have mutations in non–Lynch syndrome–associated genes and over half of these would be missed by using standard criteria for genetic evaluation.
Role of tumor-based screening approaches
More recently, health care institutions have begun to widely adopt “universal” tumor screening using microsatellite instability and/or immunohistochemistry (IHC) showing deficient expression of the mismatch repair proteins (MLH1, MSH2, MSH6, PMS2) to identify patients with colorectal or endometrial cancers that are likely to have Lynch syndrome. However, the sensitivity and specificity of IHC for Lynch syndrome ranges between 60% and 75% and there is considerable interobserver variation by pathologists in their interpretation.
Thus, both clinical guidelines (largely focused around family history and patient phenotype) and tumor molecular features will fail to identify a significant number of patients with inherited cancer predisposition.
Cost and availability of genetic testing
In the past, cost and availability of genetic testing were an impediment to such care. This has rapidly changed in the last few years. With modern next-generation sequencing technology and an ever increasing number of testing laboratories, the cost of genetic testing has dropped to below $500 and multigene panels can now test for dozens of genes in parallel offering comprehensive testing of genetic predisposition across multiple cancer types. The popularity of direct-to-consumer health-related genetic testing (with the inclusion of certain BRCA variants on these panels) has also fueled the public interest in cancer genetic testing.
Cancer prevention for family members
In individuals with CRC and hereditary cancer predisposition, implications for family members are clinically meaningful and include increased colorectal and extracolonic surveillance, consideration of risk-reducing hysterectomy, salpingo-oophorectomy, and bilateral mastectomy for colorectal, uterine, ovarian, breast, and other cancer prevention depending on the germline mutation.2 The goal of these intensive surveillance strategies is to either prevent the occurrence of cancer altogether or detect cancer at an earlier stage when cure is likely. Identifying these high-risk groups can thus play a significant role in our goal to reduce the burden of cancer in society.
Precision targeted treatment and chemoprevention
The treatment implications for patients with CRC and pathogenic mutations in the Lynch syndrome MMR genes are the best characterized and include response to immune checkpoint inhibitor therapy.4 Mismatch repair deficiency is highly predictive of response to immunotherapy in metastatic CRCs and led to expedited approval of both pembrolizumab and nivolumab monotherapies with disease control rates of 69%-77% with durable response and combination therapy with nivolumab and ipilimumab with likely even greater benefit. Multiple clinical trials are examining the role of immune checkpoint inhibitor therapy for first-line palliative treatment of MSI-high CRC (ClinicalTrials.gov ID NCT02563002; NCT02997228), adjuvant therapy (ClinicalTrials.gov ID NCT02912559), and even as potential chemoprevention in those with Lynch syndrome (ClinicalTrials.gov ID NCT03631641).
Long-term cancer prevention using a chemopreventive approach has long been a desire in the hereditary cancer community.5 The most well-studied group to date has been Lynch syndrome, where a large randomized clinical trial showed the effect of high-dose aspirin in decreasing the incidence of colorectal and other Lynch-associated cancers by nearly 60%.6 Similar smaller (earlier-phase) studies in familial adenomatous polyposis have suggested targeted chemoprevention options for the regression of colorectal or duodenal polyposis with COX inhibitors, EGFR inhibitors, DFMO (NCT01483144), and IL-23 blockade (ClinicalTrials.gov ID NCT03649971) may all be possible.
Cancer programs have already started to introduce genomic profiling (germline and tumor somatic) into the frontline care of their patients to help guide precision therapy approaches that optimize disease control, minimize side effects, and reduce risk of long-term recurrence.
The future
The approach to genomic profiling of cancer patients is rapidly changing because of the lack of sensitivity for the identification of these hereditary cancer predisposition syndromes utilizing current approaches focused on family history, clinical phenotype, and tumor features. The wide availability of low-cost/affordable multigene panel testing has implications for cancer therapy selection and cancer prevention. This supports establishing a universal approach to multigene panel testing of all patients with CRC.
It will be important for physicians of many different specialties – including gastroenterology and oncology – to become more adept in this changing landscape of genomic medicine and to work closely with the genetic counseling resources available in their communities to provide the best care for these high-risk cancer patients.
References
1. Siegel RL et al. CA Cancer J Clin. 2017;67:177-93.
2. Kanth P et al. Am J Gastroenterol. 2017;112:1509-25.
3. Gupta S et al. J Natl Compr Canc Netw. 2019;17:1032-41.
4. Ribas A, Wolchok JD. Science. 2018;359:1350-5.
5. Ramamurthy C et al. Surg Oncol Clin N Am. 2017;26:729-50.
6. Burn J et al. Lancet 2011;378:2081-7.
Dr. Samadder is a gastroenterologist in the division of gastroenterology and hepatology, Mayo Clinic, Phoenix. He is a consultant for Janssen Research & Development and Cancer Prevention Pharmaceuticals.
Not for everyone
By Carol A. Burke, MD, AGAF, and Brandie Heald Leach, MS
Multigene panel testing (MGPT) takes advantage of next-generation sequencing (NGS) a non-Sanger-based DNA sequencing technology which has revolutionized genomic research and clinical care because it can be run quickly, is lower cost than Sanger sequencing, can sequence an entire genome or exome, or specific genes of interest. Currently, cancer gene panels (disease specific or pan-cancer) are commonly utilized.
Approximately 10% of colorectal cancers (CRCs) are heritable because of a germline pathogenic variant (PV), most commonly in Lynch syndrome genes. Identification of patients with hereditary CRC is important because they are at greatest CRC and extracolonic cancer risk, benefit from aggressive cancer surveillance. and when indicated may need prophylactic surgery of at-risk organs, require multidisciplinary care, and may have at-risk family members who need testing.
Red flags regarding family cancer history may allow clinical inference as to the cause of CRC and direct who is offered germline testing. These include young age of cancer (age less than 50), synchronous or metachronous cancers, multiple relatives with CRC or extracolonic cancers, and cumulative lifetime numbers of adenomas or hamartomas. While overt clinical manifestations can be specific for predicting the causative gene defect, such as Amsterdam criteria for Lynch syndrome or numerous adenomas at a young age in familial adenomatous polyposis, overlap can occur between syndromes and single gene testing has its limitations. While family pedigrees with a phenotype that meets clinical criteria, such as Amsterdam II, can be very specific (although less sensitive) for predicting Lynch syndrome, or overt clinical manifestations such as 100 adenomatous polyps in an individual by the age of 40 is highly suggestive of familial adenomatous polyposis, overlap can occur between syndromes and single gene testing has its limitations.
The current standard of care for patients with CRC is germline testing after assessment of tumor mismatch repair (MMR) proficiency by microsatellite instability (MSI) testing and/or immunohistochemistry (IHC). Broadly, tumors that show high levels of MSI and or loss of expression of MMR proteins (not attributed to MLH1 promoter hypermethylation or double somatic mutations/loss of heterozygosity) are considered MMR deficient (MMRd) and suggestive of Lynch syndrome. MMRd directs treatment (immune check point inhibitors) and is a hallmark of Lynch syndrome as 95% of Lynch syndrome–related CRCs are MMRd.
The utility of MGPT in individuals with CRC can be inferred from two studies. In both, a 25-gene pan-cancer panel test was performed. In the first, 1,058 unselected individuals with CRC at a mean age of 56 were assessed regardless of MMR status; 9.9% were diagnosed with moderately (4.7%) or highly penetrant (5.2%) PV.1 In these individuals with CRC, 31% were diagnosed with Lynch syndrome and nearly all Lynch syndrome patients had MMRd tumors and met criteria for germline testing for Lynch syndrome; 22% of patients had other high-penetrance PV found, the majority lacking clinical features consistent with the PV. The second study,2 tested 450 patients with CRC diagnosed under the age of 50. Germline PV were detected in 16%. The majority of patients with an MMRd tumor were diagnosed with Lynch syndrome. Eight percent of patients with an MMR-proficient tumor had a PV detected. Nearly one-third did not meet clinical criteria for testing. Germline variants of uncertain significance (VUS) were noted in approximately 32% of patients in both studies. These data support the current standard of tumor assessment for MMRd, followed by Lynch syndrome germline testing as directed by IHC.
While MGPT for patients with CRC is feasible, the high rates of VUS, detection of moderate and low penetrance PV for which no clinical guidance exists, and dearth of evidence on penetrance and cancer risk attributable to incidentally found PV, need consideration. Prior to germline testing, patients and providers must understand potential testing outcomes, possible detection of incidental findings and VUS, and how each influence patient cancer risks and management. The commercial genetic testing companies accumulate information on VUS over time and reclassify the significance of the finding, but this process could take months to years. Providers ordering genetic testing must have a system to inform the patient when a VUS is reclassified.
Pre- and post-test genetic counseling, ideally by an individual with understanding of medical genetics, should be offered, including caveats, risks, benefits, and alternatives to germline testing, a plan for results disclosure, including to family members, and a plan for follow-up care. Patients with uninformative findings and VUS need to be followed as technology and research evolve. Patient preferences regarding genetic testing need to be considered. There still remains stigma and fear associated with genetic testing. Despite protections from the Genetic Information Non-Discrimination Act, many patients remain fearful of genetic discrimination. A genetic diagnosis comes with the burden that it reveals not only information about the patient’s risks, but potentially also his/her family members’ risks. These are valid patient concerns that need to be vetted and addressed.
Selection of correct testing strategy is important. A patient with a known PV in the family might benefit most from single-site analysis for the family mutation. For a patient with an affected relative who had negative genetic testing, additional genetic testing for that patients is unlikely to be beneficial. For a patient with no known PV in the family who meet genetic testing criteria, a cancer gene panel should be considered. However, guidance on which MGPT to order is lacking in professional guidelines and often left to the discretion of the provider and patient. Utilization of a “disease specific panel” (i.e., a panel of genes related to CRC risk) is useful for understanding the cause of the patient’s disease and guiding treatment, screening, and cascade testing while minimizing the number of VUS identified. Pan-cancer gene panels increase diagnostic yield, but include identification of PV in genes unrelated to phenotype or more poorly described risk and management recommendations and have a higher rate of VUS.
Finally, the cost of MGPT to the health care system needs to be considered. Despite dropping costs, the process of genetic counseling and testing remains expensive and will rise if and when testing is expanded to all patients with CRC.
MGPT is not for everyone.
References
1. Yurgelun MB et al. J Clin Oncol. 2017;35:1086-95.
2. Pearlman R et al. JAMA Oncol. 2017 Apr 01;3(4):464-71.
Dr. Burke is with the department of gastroenterology, hepatology, and nutrition, Sanford R. Weiss Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgical Institute, Cleveland Clinic; Ms. Leach is with the Center for Personalized Genetic Healthcare, Sanford R. Weiss Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgical Institute, Cleveland Clinic. Dr. Burke has no conflicts of interest, Ms. Leach serves on the advisory board of Invitae.
Dear colleagues and friends,
I write to introduce to you the new Perspectives section of GI & Hepatology News.
A more appropriate description is perhaps old-new, because Perspectives is the continuation and legacy of AGA Perspectives, the content of which has been consolidated into GI & Hepatology News. Perspectives will continue to feature the point/counterpoint expert debates about an important GI topic, which has historically been immensely popular with readers. In this edition, experts from Mayo Clinic and Cleveland Clinic discuss the pros and cons of universal multigene panel testing for colorectal cancer. These debates never end with the publication itself, and I hope they will continue to stimulate further thought and discussion. As always, I welcome your comments and suggestions for future topics.
–Charles I. Kahi, MD, MS, AGAF, is professor of medicine at Indiana University School of Medicine, Indianapolis. He is also an Associate Editor for GI & Hepatology News.
For everyone
By N. Jewel Samadder, MD, MSC
Traditionally, health care structure has been directed predominantly toward treatment rather than prevention. Advances in genomic medicine offer the opportunity to deliver a more personalized, predictive, and preventive strategy toward colorectal cancer. Approximately 150,000 men and women are diagnosed with colorectal cancer (CRC) every year in the United States.1 An estimated 10%-15% of these cancers are likely attributable to hereditary (germline) causes.2 Several genes are associated with an increased risk of developing CRC, and those of key interest include those for Lynch syndrome, MLH1, MSH2, MSH6, PMS2, EPCAM; adenomatous polyposis conditions (APC), MUTYH, POLE, POLD1, NTHL1; hamartomatous polyposis syndromes PTEN, SMAD4, STK11, and other rare cancer predisposition states where colorectal cancer is part of the phenotype, CHEK2 and TP532.
A universal strategy for multigene panel testing in all patients with CRC is an option versus the current strategy of guideline-based testing using family history and tumor features. In addition, the identification of germline alterations has substantial clinical implications including targeted therapies and future cancer prevention in the patient and relatives. This article will focus on the benefits of universal strategy for germline genetic evaluation in all patients with colorectal cancer.
The role and utility of current guideline-based testing
Given the therapeutic and prevention implications, the National Comprehensive Cancer Network (along with other professional organizations) has guidance on when patients with CRC should undergo genetic evaluation.3 Currently, these guidelines advocate an approach based heavily on family cancer history or utilizing colorectal phenotype based on the number and histology of polyps or tumor-based molecular features. Although family history is important for the diagnosis of hereditary CRC, the ability to accurately capture extended family cancer history in routine practice, from multiple generations and for different cancer types can be a challenge. The largest drawback of all such approaches is the focus on Lynch syndrome or only a few of the cancer predisposition syndromes. Recent studies have reported a substantial number (7%-10%) of CRC patients will have mutations in non–Lynch syndrome–associated genes and over half of these would be missed by using standard criteria for genetic evaluation.
Role of tumor-based screening approaches
More recently, health care institutions have begun to widely adopt “universal” tumor screening using microsatellite instability and/or immunohistochemistry (IHC) showing deficient expression of the mismatch repair proteins (MLH1, MSH2, MSH6, PMS2) to identify patients with colorectal or endometrial cancers that are likely to have Lynch syndrome. However, the sensitivity and specificity of IHC for Lynch syndrome ranges between 60% and 75% and there is considerable interobserver variation by pathologists in their interpretation.
Thus, both clinical guidelines (largely focused around family history and patient phenotype) and tumor molecular features will fail to identify a significant number of patients with inherited cancer predisposition.
Cost and availability of genetic testing
In the past, cost and availability of genetic testing were an impediment to such care. This has rapidly changed in the last few years. With modern next-generation sequencing technology and an ever increasing number of testing laboratories, the cost of genetic testing has dropped to below $500 and multigene panels can now test for dozens of genes in parallel offering comprehensive testing of genetic predisposition across multiple cancer types. The popularity of direct-to-consumer health-related genetic testing (with the inclusion of certain BRCA variants on these panels) has also fueled the public interest in cancer genetic testing.
Cancer prevention for family members
In individuals with CRC and hereditary cancer predisposition, implications for family members are clinically meaningful and include increased colorectal and extracolonic surveillance, consideration of risk-reducing hysterectomy, salpingo-oophorectomy, and bilateral mastectomy for colorectal, uterine, ovarian, breast, and other cancer prevention depending on the germline mutation.2 The goal of these intensive surveillance strategies is to either prevent the occurrence of cancer altogether or detect cancer at an earlier stage when cure is likely. Identifying these high-risk groups can thus play a significant role in our goal to reduce the burden of cancer in society.
Precision targeted treatment and chemoprevention
The treatment implications for patients with CRC and pathogenic mutations in the Lynch syndrome MMR genes are the best characterized and include response to immune checkpoint inhibitor therapy.4 Mismatch repair deficiency is highly predictive of response to immunotherapy in metastatic CRCs and led to expedited approval of both pembrolizumab and nivolumab monotherapies with disease control rates of 69%-77% with durable response and combination therapy with nivolumab and ipilimumab with likely even greater benefit. Multiple clinical trials are examining the role of immune checkpoint inhibitor therapy for first-line palliative treatment of MSI-high CRC (ClinicalTrials.gov ID NCT02563002; NCT02997228), adjuvant therapy (ClinicalTrials.gov ID NCT02912559), and even as potential chemoprevention in those with Lynch syndrome (ClinicalTrials.gov ID NCT03631641).
Long-term cancer prevention using a chemopreventive approach has long been a desire in the hereditary cancer community.5 The most well-studied group to date has been Lynch syndrome, where a large randomized clinical trial showed the effect of high-dose aspirin in decreasing the incidence of colorectal and other Lynch-associated cancers by nearly 60%.6 Similar smaller (earlier-phase) studies in familial adenomatous polyposis have suggested targeted chemoprevention options for the regression of colorectal or duodenal polyposis with COX inhibitors, EGFR inhibitors, DFMO (NCT01483144), and IL-23 blockade (ClinicalTrials.gov ID NCT03649971) may all be possible.
Cancer programs have already started to introduce genomic profiling (germline and tumor somatic) into the frontline care of their patients to help guide precision therapy approaches that optimize disease control, minimize side effects, and reduce risk of long-term recurrence.
The future
The approach to genomic profiling of cancer patients is rapidly changing because of the lack of sensitivity for the identification of these hereditary cancer predisposition syndromes utilizing current approaches focused on family history, clinical phenotype, and tumor features. The wide availability of low-cost/affordable multigene panel testing has implications for cancer therapy selection and cancer prevention. This supports establishing a universal approach to multigene panel testing of all patients with CRC.
It will be important for physicians of many different specialties – including gastroenterology and oncology – to become more adept in this changing landscape of genomic medicine and to work closely with the genetic counseling resources available in their communities to provide the best care for these high-risk cancer patients.
References
1. Siegel RL et al. CA Cancer J Clin. 2017;67:177-93.
2. Kanth P et al. Am J Gastroenterol. 2017;112:1509-25.
3. Gupta S et al. J Natl Compr Canc Netw. 2019;17:1032-41.
4. Ribas A, Wolchok JD. Science. 2018;359:1350-5.
5. Ramamurthy C et al. Surg Oncol Clin N Am. 2017;26:729-50.
6. Burn J et al. Lancet 2011;378:2081-7.
Dr. Samadder is a gastroenterologist in the division of gastroenterology and hepatology, Mayo Clinic, Phoenix. He is a consultant for Janssen Research & Development and Cancer Prevention Pharmaceuticals.
Not for everyone
By Carol A. Burke, MD, AGAF, and Brandie Heald Leach, MS
Multigene panel testing (MGPT) takes advantage of next-generation sequencing (NGS) a non-Sanger-based DNA sequencing technology which has revolutionized genomic research and clinical care because it can be run quickly, is lower cost than Sanger sequencing, can sequence an entire genome or exome, or specific genes of interest. Currently, cancer gene panels (disease specific or pan-cancer) are commonly utilized.
Approximately 10% of colorectal cancers (CRCs) are heritable because of a germline pathogenic variant (PV), most commonly in Lynch syndrome genes. Identification of patients with hereditary CRC is important because they are at greatest CRC and extracolonic cancer risk, benefit from aggressive cancer surveillance. and when indicated may need prophylactic surgery of at-risk organs, require multidisciplinary care, and may have at-risk family members who need testing.
Red flags regarding family cancer history may allow clinical inference as to the cause of CRC and direct who is offered germline testing. These include young age of cancer (age less than 50), synchronous or metachronous cancers, multiple relatives with CRC or extracolonic cancers, and cumulative lifetime numbers of adenomas or hamartomas. While overt clinical manifestations can be specific for predicting the causative gene defect, such as Amsterdam criteria for Lynch syndrome or numerous adenomas at a young age in familial adenomatous polyposis, overlap can occur between syndromes and single gene testing has its limitations. While family pedigrees with a phenotype that meets clinical criteria, such as Amsterdam II, can be very specific (although less sensitive) for predicting Lynch syndrome, or overt clinical manifestations such as 100 adenomatous polyps in an individual by the age of 40 is highly suggestive of familial adenomatous polyposis, overlap can occur between syndromes and single gene testing has its limitations.
The current standard of care for patients with CRC is germline testing after assessment of tumor mismatch repair (MMR) proficiency by microsatellite instability (MSI) testing and/or immunohistochemistry (IHC). Broadly, tumors that show high levels of MSI and or loss of expression of MMR proteins (not attributed to MLH1 promoter hypermethylation or double somatic mutations/loss of heterozygosity) are considered MMR deficient (MMRd) and suggestive of Lynch syndrome. MMRd directs treatment (immune check point inhibitors) and is a hallmark of Lynch syndrome as 95% of Lynch syndrome–related CRCs are MMRd.
The utility of MGPT in individuals with CRC can be inferred from two studies. In both, a 25-gene pan-cancer panel test was performed. In the first, 1,058 unselected individuals with CRC at a mean age of 56 were assessed regardless of MMR status; 9.9% were diagnosed with moderately (4.7%) or highly penetrant (5.2%) PV.1 In these individuals with CRC, 31% were diagnosed with Lynch syndrome and nearly all Lynch syndrome patients had MMRd tumors and met criteria for germline testing for Lynch syndrome; 22% of patients had other high-penetrance PV found, the majority lacking clinical features consistent with the PV. The second study,2 tested 450 patients with CRC diagnosed under the age of 50. Germline PV were detected in 16%. The majority of patients with an MMRd tumor were diagnosed with Lynch syndrome. Eight percent of patients with an MMR-proficient tumor had a PV detected. Nearly one-third did not meet clinical criteria for testing. Germline variants of uncertain significance (VUS) were noted in approximately 32% of patients in both studies. These data support the current standard of tumor assessment for MMRd, followed by Lynch syndrome germline testing as directed by IHC.
While MGPT for patients with CRC is feasible, the high rates of VUS, detection of moderate and low penetrance PV for which no clinical guidance exists, and dearth of evidence on penetrance and cancer risk attributable to incidentally found PV, need consideration. Prior to germline testing, patients and providers must understand potential testing outcomes, possible detection of incidental findings and VUS, and how each influence patient cancer risks and management. The commercial genetic testing companies accumulate information on VUS over time and reclassify the significance of the finding, but this process could take months to years. Providers ordering genetic testing must have a system to inform the patient when a VUS is reclassified.
Pre- and post-test genetic counseling, ideally by an individual with understanding of medical genetics, should be offered, including caveats, risks, benefits, and alternatives to germline testing, a plan for results disclosure, including to family members, and a plan for follow-up care. Patients with uninformative findings and VUS need to be followed as technology and research evolve. Patient preferences regarding genetic testing need to be considered. There still remains stigma and fear associated with genetic testing. Despite protections from the Genetic Information Non-Discrimination Act, many patients remain fearful of genetic discrimination. A genetic diagnosis comes with the burden that it reveals not only information about the patient’s risks, but potentially also his/her family members’ risks. These are valid patient concerns that need to be vetted and addressed.
Selection of correct testing strategy is important. A patient with a known PV in the family might benefit most from single-site analysis for the family mutation. For a patient with an affected relative who had negative genetic testing, additional genetic testing for that patients is unlikely to be beneficial. For a patient with no known PV in the family who meet genetic testing criteria, a cancer gene panel should be considered. However, guidance on which MGPT to order is lacking in professional guidelines and often left to the discretion of the provider and patient. Utilization of a “disease specific panel” (i.e., a panel of genes related to CRC risk) is useful for understanding the cause of the patient’s disease and guiding treatment, screening, and cascade testing while minimizing the number of VUS identified. Pan-cancer gene panels increase diagnostic yield, but include identification of PV in genes unrelated to phenotype or more poorly described risk and management recommendations and have a higher rate of VUS.
Finally, the cost of MGPT to the health care system needs to be considered. Despite dropping costs, the process of genetic counseling and testing remains expensive and will rise if and when testing is expanded to all patients with CRC.
MGPT is not for everyone.
References
1. Yurgelun MB et al. J Clin Oncol. 2017;35:1086-95.
2. Pearlman R et al. JAMA Oncol. 2017 Apr 01;3(4):464-71.
Dr. Burke is with the department of gastroenterology, hepatology, and nutrition, Sanford R. Weiss Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgical Institute, Cleveland Clinic; Ms. Leach is with the Center for Personalized Genetic Healthcare, Sanford R. Weiss Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgical Institute, Cleveland Clinic. Dr. Burke has no conflicts of interest, Ms. Leach serves on the advisory board of Invitae.
Dear colleagues and friends,
I write to introduce to you the new Perspectives section of GI & Hepatology News.
A more appropriate description is perhaps old-new, because Perspectives is the continuation and legacy of AGA Perspectives, the content of which has been consolidated into GI & Hepatology News. Perspectives will continue to feature the point/counterpoint expert debates about an important GI topic, which has historically been immensely popular with readers. In this edition, experts from Mayo Clinic and Cleveland Clinic discuss the pros and cons of universal multigene panel testing for colorectal cancer. These debates never end with the publication itself, and I hope they will continue to stimulate further thought and discussion. As always, I welcome your comments and suggestions for future topics.
–Charles I. Kahi, MD, MS, AGAF, is professor of medicine at Indiana University School of Medicine, Indianapolis. He is also an Associate Editor for GI & Hepatology News.
For everyone
By N. Jewel Samadder, MD, MSC
Traditionally, health care structure has been directed predominantly toward treatment rather than prevention. Advances in genomic medicine offer the opportunity to deliver a more personalized, predictive, and preventive strategy toward colorectal cancer. Approximately 150,000 men and women are diagnosed with colorectal cancer (CRC) every year in the United States.1 An estimated 10%-15% of these cancers are likely attributable to hereditary (germline) causes.2 Several genes are associated with an increased risk of developing CRC, and those of key interest include those for Lynch syndrome, MLH1, MSH2, MSH6, PMS2, EPCAM; adenomatous polyposis conditions (APC), MUTYH, POLE, POLD1, NTHL1; hamartomatous polyposis syndromes PTEN, SMAD4, STK11, and other rare cancer predisposition states where colorectal cancer is part of the phenotype, CHEK2 and TP532.
A universal strategy for multigene panel testing in all patients with CRC is an option versus the current strategy of guideline-based testing using family history and tumor features. In addition, the identification of germline alterations has substantial clinical implications including targeted therapies and future cancer prevention in the patient and relatives. This article will focus on the benefits of universal strategy for germline genetic evaluation in all patients with colorectal cancer.
The role and utility of current guideline-based testing
Given the therapeutic and prevention implications, the National Comprehensive Cancer Network (along with other professional organizations) has guidance on when patients with CRC should undergo genetic evaluation.3 Currently, these guidelines advocate an approach based heavily on family cancer history or utilizing colorectal phenotype based on the number and histology of polyps or tumor-based molecular features. Although family history is important for the diagnosis of hereditary CRC, the ability to accurately capture extended family cancer history in routine practice, from multiple generations and for different cancer types can be a challenge. The largest drawback of all such approaches is the focus on Lynch syndrome or only a few of the cancer predisposition syndromes. Recent studies have reported a substantial number (7%-10%) of CRC patients will have mutations in non–Lynch syndrome–associated genes and over half of these would be missed by using standard criteria for genetic evaluation.
Role of tumor-based screening approaches
More recently, health care institutions have begun to widely adopt “universal” tumor screening using microsatellite instability and/or immunohistochemistry (IHC) showing deficient expression of the mismatch repair proteins (MLH1, MSH2, MSH6, PMS2) to identify patients with colorectal or endometrial cancers that are likely to have Lynch syndrome. However, the sensitivity and specificity of IHC for Lynch syndrome ranges between 60% and 75% and there is considerable interobserver variation by pathologists in their interpretation.
Thus, both clinical guidelines (largely focused around family history and patient phenotype) and tumor molecular features will fail to identify a significant number of patients with inherited cancer predisposition.
Cost and availability of genetic testing
In the past, cost and availability of genetic testing were an impediment to such care. This has rapidly changed in the last few years. With modern next-generation sequencing technology and an ever increasing number of testing laboratories, the cost of genetic testing has dropped to below $500 and multigene panels can now test for dozens of genes in parallel offering comprehensive testing of genetic predisposition across multiple cancer types. The popularity of direct-to-consumer health-related genetic testing (with the inclusion of certain BRCA variants on these panels) has also fueled the public interest in cancer genetic testing.
Cancer prevention for family members
In individuals with CRC and hereditary cancer predisposition, implications for family members are clinically meaningful and include increased colorectal and extracolonic surveillance, consideration of risk-reducing hysterectomy, salpingo-oophorectomy, and bilateral mastectomy for colorectal, uterine, ovarian, breast, and other cancer prevention depending on the germline mutation.2 The goal of these intensive surveillance strategies is to either prevent the occurrence of cancer altogether or detect cancer at an earlier stage when cure is likely. Identifying these high-risk groups can thus play a significant role in our goal to reduce the burden of cancer in society.
Precision targeted treatment and chemoprevention
The treatment implications for patients with CRC and pathogenic mutations in the Lynch syndrome MMR genes are the best characterized and include response to immune checkpoint inhibitor therapy.4 Mismatch repair deficiency is highly predictive of response to immunotherapy in metastatic CRCs and led to expedited approval of both pembrolizumab and nivolumab monotherapies with disease control rates of 69%-77% with durable response and combination therapy with nivolumab and ipilimumab with likely even greater benefit. Multiple clinical trials are examining the role of immune checkpoint inhibitor therapy for first-line palliative treatment of MSI-high CRC (ClinicalTrials.gov ID NCT02563002; NCT02997228), adjuvant therapy (ClinicalTrials.gov ID NCT02912559), and even as potential chemoprevention in those with Lynch syndrome (ClinicalTrials.gov ID NCT03631641).
Long-term cancer prevention using a chemopreventive approach has long been a desire in the hereditary cancer community.5 The most well-studied group to date has been Lynch syndrome, where a large randomized clinical trial showed the effect of high-dose aspirin in decreasing the incidence of colorectal and other Lynch-associated cancers by nearly 60%.6 Similar smaller (earlier-phase) studies in familial adenomatous polyposis have suggested targeted chemoprevention options for the regression of colorectal or duodenal polyposis with COX inhibitors, EGFR inhibitors, DFMO (NCT01483144), and IL-23 blockade (ClinicalTrials.gov ID NCT03649971) may all be possible.
Cancer programs have already started to introduce genomic profiling (germline and tumor somatic) into the frontline care of their patients to help guide precision therapy approaches that optimize disease control, minimize side effects, and reduce risk of long-term recurrence.
The future
The approach to genomic profiling of cancer patients is rapidly changing because of the lack of sensitivity for the identification of these hereditary cancer predisposition syndromes utilizing current approaches focused on family history, clinical phenotype, and tumor features. The wide availability of low-cost/affordable multigene panel testing has implications for cancer therapy selection and cancer prevention. This supports establishing a universal approach to multigene panel testing of all patients with CRC.
It will be important for physicians of many different specialties – including gastroenterology and oncology – to become more adept in this changing landscape of genomic medicine and to work closely with the genetic counseling resources available in their communities to provide the best care for these high-risk cancer patients.
References
1. Siegel RL et al. CA Cancer J Clin. 2017;67:177-93.
2. Kanth P et al. Am J Gastroenterol. 2017;112:1509-25.
3. Gupta S et al. J Natl Compr Canc Netw. 2019;17:1032-41.
4. Ribas A, Wolchok JD. Science. 2018;359:1350-5.
5. Ramamurthy C et al. Surg Oncol Clin N Am. 2017;26:729-50.
6. Burn J et al. Lancet 2011;378:2081-7.
Dr. Samadder is a gastroenterologist in the division of gastroenterology and hepatology, Mayo Clinic, Phoenix. He is a consultant for Janssen Research & Development and Cancer Prevention Pharmaceuticals.
Not for everyone
By Carol A. Burke, MD, AGAF, and Brandie Heald Leach, MS
Multigene panel testing (MGPT) takes advantage of next-generation sequencing (NGS) a non-Sanger-based DNA sequencing technology which has revolutionized genomic research and clinical care because it can be run quickly, is lower cost than Sanger sequencing, can sequence an entire genome or exome, or specific genes of interest. Currently, cancer gene panels (disease specific or pan-cancer) are commonly utilized.
Approximately 10% of colorectal cancers (CRCs) are heritable because of a germline pathogenic variant (PV), most commonly in Lynch syndrome genes. Identification of patients with hereditary CRC is important because they are at greatest CRC and extracolonic cancer risk, benefit from aggressive cancer surveillance. and when indicated may need prophylactic surgery of at-risk organs, require multidisciplinary care, and may have at-risk family members who need testing.
Red flags regarding family cancer history may allow clinical inference as to the cause of CRC and direct who is offered germline testing. These include young age of cancer (age less than 50), synchronous or metachronous cancers, multiple relatives with CRC or extracolonic cancers, and cumulative lifetime numbers of adenomas or hamartomas. While overt clinical manifestations can be specific for predicting the causative gene defect, such as Amsterdam criteria for Lynch syndrome or numerous adenomas at a young age in familial adenomatous polyposis, overlap can occur between syndromes and single gene testing has its limitations. While family pedigrees with a phenotype that meets clinical criteria, such as Amsterdam II, can be very specific (although less sensitive) for predicting Lynch syndrome, or overt clinical manifestations such as 100 adenomatous polyps in an individual by the age of 40 is highly suggestive of familial adenomatous polyposis, overlap can occur between syndromes and single gene testing has its limitations.
The current standard of care for patients with CRC is germline testing after assessment of tumor mismatch repair (MMR) proficiency by microsatellite instability (MSI) testing and/or immunohistochemistry (IHC). Broadly, tumors that show high levels of MSI and or loss of expression of MMR proteins (not attributed to MLH1 promoter hypermethylation or double somatic mutations/loss of heterozygosity) are considered MMR deficient (MMRd) and suggestive of Lynch syndrome. MMRd directs treatment (immune check point inhibitors) and is a hallmark of Lynch syndrome as 95% of Lynch syndrome–related CRCs are MMRd.
The utility of MGPT in individuals with CRC can be inferred from two studies. In both, a 25-gene pan-cancer panel test was performed. In the first, 1,058 unselected individuals with CRC at a mean age of 56 were assessed regardless of MMR status; 9.9% were diagnosed with moderately (4.7%) or highly penetrant (5.2%) PV.1 In these individuals with CRC, 31% were diagnosed with Lynch syndrome and nearly all Lynch syndrome patients had MMRd tumors and met criteria for germline testing for Lynch syndrome; 22% of patients had other high-penetrance PV found, the majority lacking clinical features consistent with the PV. The second study,2 tested 450 patients with CRC diagnosed under the age of 50. Germline PV were detected in 16%. The majority of patients with an MMRd tumor were diagnosed with Lynch syndrome. Eight percent of patients with an MMR-proficient tumor had a PV detected. Nearly one-third did not meet clinical criteria for testing. Germline variants of uncertain significance (VUS) were noted in approximately 32% of patients in both studies. These data support the current standard of tumor assessment for MMRd, followed by Lynch syndrome germline testing as directed by IHC.
While MGPT for patients with CRC is feasible, the high rates of VUS, detection of moderate and low penetrance PV for which no clinical guidance exists, and dearth of evidence on penetrance and cancer risk attributable to incidentally found PV, need consideration. Prior to germline testing, patients and providers must understand potential testing outcomes, possible detection of incidental findings and VUS, and how each influence patient cancer risks and management. The commercial genetic testing companies accumulate information on VUS over time and reclassify the significance of the finding, but this process could take months to years. Providers ordering genetic testing must have a system to inform the patient when a VUS is reclassified.
Pre- and post-test genetic counseling, ideally by an individual with understanding of medical genetics, should be offered, including caveats, risks, benefits, and alternatives to germline testing, a plan for results disclosure, including to family members, and a plan for follow-up care. Patients with uninformative findings and VUS need to be followed as technology and research evolve. Patient preferences regarding genetic testing need to be considered. There still remains stigma and fear associated with genetic testing. Despite protections from the Genetic Information Non-Discrimination Act, many patients remain fearful of genetic discrimination. A genetic diagnosis comes with the burden that it reveals not only information about the patient’s risks, but potentially also his/her family members’ risks. These are valid patient concerns that need to be vetted and addressed.
Selection of correct testing strategy is important. A patient with a known PV in the family might benefit most from single-site analysis for the family mutation. For a patient with an affected relative who had negative genetic testing, additional genetic testing for that patients is unlikely to be beneficial. For a patient with no known PV in the family who meet genetic testing criteria, a cancer gene panel should be considered. However, guidance on which MGPT to order is lacking in professional guidelines and often left to the discretion of the provider and patient. Utilization of a “disease specific panel” (i.e., a panel of genes related to CRC risk) is useful for understanding the cause of the patient’s disease and guiding treatment, screening, and cascade testing while minimizing the number of VUS identified. Pan-cancer gene panels increase diagnostic yield, but include identification of PV in genes unrelated to phenotype or more poorly described risk and management recommendations and have a higher rate of VUS.
Finally, the cost of MGPT to the health care system needs to be considered. Despite dropping costs, the process of genetic counseling and testing remains expensive and will rise if and when testing is expanded to all patients with CRC.
MGPT is not for everyone.
References
1. Yurgelun MB et al. J Clin Oncol. 2017;35:1086-95.
2. Pearlman R et al. JAMA Oncol. 2017 Apr 01;3(4):464-71.
Dr. Burke is with the department of gastroenterology, hepatology, and nutrition, Sanford R. Weiss Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgical Institute, Cleveland Clinic; Ms. Leach is with the Center for Personalized Genetic Healthcare, Sanford R. Weiss Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgical Institute, Cleveland Clinic. Dr. Burke has no conflicts of interest, Ms. Leach serves on the advisory board of Invitae.