Genome Sequencing Studies Shake Up Views on the Cancer Mutation Landscape | OncLive

Genome Sequencing Studies Shake Up Views on the Cancer Mutation Landscape

July 2, 2019

Emerging findings from the most expansive studies of tumor genomes ever conducted are building a case for integrating all aspects of germline and somatic mutations into the analytical paradigm as a logical next step for precision oncology.

Emerging findings from the most expansive studies of tumor genomes ever conducted are building a case for integrating all aspects of germline and somatic mutations into the analytical paradigm as a logical next step for precision oncology.

Important revelations regarding myriad aspects of the cancer genome from the Pan-Cancer Atlas Project, the most comprehensive analysis conducted by The Cancer Genome Atlas (TCGA), have shed new light on clinical actionability of prognostic, diagnostic, and therapeutic significance.1-3

Table. Gene Mutations Targeted by FDA-Approved Companion Diagnosticsa9

Historically, assigning clinical actionability to oncogenic drivers has focused on the somatic mutations unique to cancer cells. Germline mutations contained within the heritable genome have been largely consigned to heritable cancer risk management. Now, genome sequencing study findings are fostering a growing appreciation of both the burden of germline mutations and their potential clinical actionability. Furthermore, in both germline and tumor genomes, specific mutations have been identified in discordant histological cancer types. Pan-Cancer analyses have also revealed genetic interactions between germline and somatic mutations, suggesting that the germline genome may have a significant impact on the somatic evolution of the tumor.4-6

Together, these groundbreaking studies underscore the importance of integrating germline and somatic mutation data, their interactions and their impact on the immune system and the microenvironment, and of continuing the trend toward tumor-agnostic clinical actionability.

Illuminating the Genomic Landscape

Cancer has long been considered a genetic disease, the result of DNA-damaging assaults on the cell that yield mutations in genes encoding proteins that play a role in certain hallmark processes that dictate malignant growth. The majority of cancers are thought to arise sporadically from the acquisition of somatic mutations that are unique to the tumor cells. Between 5% and 10% of cancers result from mutations contained within the heritable genome.7,8

The idea of exploiting these underlying mutations to therapeutically target cancer cells is the premise behind precision oncology. Over the past several decades, our ability to interrogate the tumor genome and identify oncogenic drivers has rapidly expanded thanks to cheaper and faster sequencing technology.

As a result, the precision oncology field has evolved from a handful of drugs designed to target aberrantly expressed proteins in cancer cells to an explosion in molecularly targeted and immune-based therapies, with the use of some now guided by specific genetic tests or next-generation sequencing panels.6,9-11 The FDA has approved 27 therapies for use with companion diagnostics that test for gene mutations (Table).9

Clinical trials are increasingly designed to be biomarker driven rather than based on tumor histology. The first drugs with tumor-agnostic indications were approved by the FDA in the past several years, with others likely to follow.12,13

Whole-exome sequencing (WES) is required to capture all mutations across an entire coding portion of the tumor genome. Although WES has not been broadly incorporated into clinical care, it has been central to our growing understanding of the mutational landscape of cancer, in addition to mechanisms of resistance and exceptional responses to targeted therapies that have been deemed clinical failures in unselected populations.14,15

Several large-scale, multicenter research projects have been established. TCGA is one of the most ambitious and successful to date, having collected data from over 11,000 patients with more than 30 types of cancer.16 One of its main goals is to provide a publicly available dataset to help inform better diagnosis and treatment of cancer.

Beginning with glioblastoma in 2008, each of the tumor types was characterized and the results published in high-profile journals. Although this facilitated significant progress in our understanding of the molecular drivers of these cancer types, the analytical and computational tools used in these studies varied widely and there has not always been agreement about candidate driver genes and mutations. With more than 400 terabytes of raw data collected, the cost and complexity of data storage hindered a uniform analysis across the dataset.1,17

A little over a decade after its inception, the TCGA project culminated in 2018 with the most comprehensive analysis of the collected data yet, the Pan-Caner Atlas, which sought to overcome these challenges. Multilab consortiums shared the cost and logistics of data storage, and novel tools were developed for uniform and unbiased analysis of 9423 tumor exomes from 33 cancer types. Germline mutations were analyzed in nearly 10,400 adult tumor samples.3-5,17 (Figure4,5).

In terms of somatic mutations, 299 cancer driver genes were identified, most of which had previously been described for the majority of cancer types. However, 59 were novel, either not having been described previously or, intriguingly, not having been described as a driver of the particular cancer type in which they were found. Among the driver genes, 142 were associated with a single cancer type, 87 had roles in 2 or more, and 20 were implicated in many tumor types.

The number of driver genes varied according to cancer type, and although overall there were more affected oncogenes than tumor suppressors, the relative ratio also varied across tumor types. Within these genes, they identified 9919 predicted cancer driver mutations, among which 3437 were unique.

Figure. Actionable Somatic and Germline Mutations in Pan-Cancer ATLAS Project4,5

Among the many instances where driver mutations were predicted in tumor types where the gene was not a known driver are EGFR mutations. Best known for their role as drivers of non—small cell lung cancer (NSCLC), EGFR mutations were identified in other tumor types, including head and neck squamous cell carcinoma, lung squamous cell carcinoma, stomach adenocarcinoma, and uterine corpus endometrial carcinoma. An overarching finding of the analysis was that cancers that originate in different organs share many commonalities at the molecular level, which lends further support to the current trend toward more tumor-agnostic approaches in clinical trial design.4

For germline mutations, an analysis of Pan-Cancer Atlas data identified high expression of 33 predisposition variants affecting activating domains of oncogenes, including MET, RET, and PTPN11. Variants of tumor suppressor genes were associated with low expression and loss of heterozygosity; variants in oncogenes correlated with high expression.5

From Identification to Actionability

The fundamental challenge of precision medicine is translating genome sequencing findings into direct or indirect action by improving diagnosis, prognosis or treatment—so-called clinical actionability.

Substantial resources have been devoted to optimizing bioinformatics techniques to connect the dots between identified mutations and potential actionability. In Pan-Cancer Atlas, actionable mutations, based on predicted sensitivity to FDA-approved therapies and those in ongoing clinical trials, as well as evidence from published clinical trials, were found in more than half of all tumor samples.4

Limited evidence of the clinical significance of many mutations and lack of a comprehensive database for cataloging evidence of clinical actionability are the most significant barriers to precision oncology. Current databases are not standardized and are maintained by a range of different entities.18

A number of efforts to address these barriers are ongoing. The European Society for Medical Oncology (ESMO) Translational Research and Precision Medicine Working Group recently developed the ESMO Scale of Clinical Actionability for Molecular Targets (ESCAT), defining 6 tiers of clinical evidence for molecular targets based on their implications for patient management.19

Germline Genome is Risky Business

Prior to the advent of high-throughput genome sequencing techniques, oncogenic drivers were frequently identified through genetic studies of individuals with inherited cancer syndromes. Germline mutations were identified that increase an individual’s lifetime risk of developing cancer.7,20

Germline mutations can dictate cancer risk in a number of ways: by affecting the function of proteins that repair DNA damage or by generating the “first hit” to a gene, leaving only 1 functioning copy that could drive cancer if a second mutation is acquired.

The potential therapeutic implications of germline mutations have been largely ignored, beyond their role in guiding prophylactic strategies and possible effect on the patient’s family history and lifestyle factors.15 The identification of actionable mutations has largely focused on the somatic genome.

The most high-profile exceptions are the breast cancer susceptibility genes BRCA1 and BRCA2. Germline loss-of-function mutations in these genes confer susceptibility to breast and ovarian cancers. The proteins they encode are involved in DNA repair through the homologous recombination (HR) pathway, which repairs lethal double-strand breaks (DSBs) in DNA; thus, BRCA1/2 mutations render tumors more dependent on alternative DNA repair pathways.

This drives sensitivity to platinum-containing chemotherapy drugs, which cause further DNA damage that is difficult for the BRCA1/2- deficient cancer cells to repair. It has also been therapeutically exploited in the development of targeted therapies that introduce additional DNA damage, a strategy known as synthetic lethality. This formed the basis for exploration of PARP inhibitors in germline BRCA1/2-mutated cancers, which have since become the standard of care in advanced ovarian cancer.6

Germline defects in DNA repair genes can lead to other heritable cancer syndromes, such as Lynch syndrome, which is characterized by an increased risk of colorectal (CRC), endometrial, and other cancers and results from germline mutations in genes involved in the mismatch repair (MMR) pathway, including MSH2, MLH1, MSH6, PMS2, and POLE.

MMR deficiency (dMMR) often fuels a phenotype known as microsatellite instability (MSI), which involves fluctuations in the length of repetitive sequences of DNA within the genome whose function is unknown. Tumors with high levels of defective MMR and MSI are often hypermutable and provoke a strong antitumor immune response, leading to the suggestion that they may be more sensitive to immunotherapy. This hypothesis has resulted in FDA approvals for pembrolizumab (Keytruda) in dMMR or MSI-H cancers regardless of the primary tumor site and nivolumab (Opdivo) for MSI-high or dMMR metastatic CRC.6,12

In pivotal studies of pembrolizumab, patients with dMMR CRCs (most of whom had Lynch syndrome) and non-CRC dMMR tumors (just under half of whom had Lynch syndrome) were compared with those with MMR-proficient tumors. Pembrolizumab dramatically improved response rates in patients with dMMR tumors, regardless of histological subtype. There were differences in response rates based on the presence of Lynch syndrome; all 6 patients with dMMR tumors not related to Lynch syndrome responded, whereas only 3 of 11 patients (27%) with tumors associated with the germline defects had a response (P = .009).21,22

Germline Findings Not Just Incidental

As genome sequencing has become more common, germline analysis has been used to provide a “normal” reference with which to compare the somatic mutational landscape. Assuming that the majority of germline mutations are not involved in the development of cancer, then true somatic mutations can be picked out if normal tissue is analyzed alongside the tumor.

Should a germline mutation in a gene that conveys an increased cancer risk be identified, it is referred to as an incidental finding. Due to ethical implications, these findings are currently reported only if the mutation is known to be pathogenic or is likely to be pathogenic. So-called variants of unknown significance are an increasing issue.23,24

More importantly, mounting evidence suggests that our assumptions about germline mutations in cancer are incorrect and that they are a much greater burden than anticipated. In one study of 1040 adults with cancer (mostly advanced cases), 17.5% had clinically actionable germline variants conferring cancer susceptibility; the mutations of 55% (101 of 182) of these patients would have been missed if testing were limited to guidelines based on family history.25

Analysis of germline mutations in the TCGA dataset was recently published. Pathogenic or likely pathogenic germline variants were identified in approximately 8% of patients, and a further 4.6% had possible pathogenic variants. BRCA1/2 represented 20.2% of the germline variants, and other HR-associated genes represented 26.6%. Additionally, 6% of pathogenic germline variants occurred in MMR genes.6

Studies in specific cancer types, including those traditionally thought to be predominantly sporadic in nature, have revealed a similar trend. In patients with advanced prostate cancer, a presumably pathogenic germline variant in 1 of 20 DNA repair genes was identified in 11.8% of cases, and higher rates of such mutations were found in metastatic compared with localized disease.26

The presence of germline mutations often could not have been predicted by family history, age at diagnosis or tumor type, and patients would not have undergone clinical testing based on current guidelines.25,27 Furthermore, both germline and somatic mutations are often found in discordant tumor types—that is, in tumor histologies where they are not currently predicted to play a significant role.4-6 This suggests that current germline sequencing may be missing significant clinically actionable information about tumors.

Integrated Genome Analysis

There is a growing appreciation of the potential clinical actionability of germline mutations. A recent study categorized the germline variants identified in the TCGA dataset into 3 tiers according to the level of evidence for their clinical actionability and found that 63.8% of the pathogenic and likely pathogenic variants were clinically actionable. This is defined as showing either sufficient evidence to gain regulatory approval (tier 1), a high level and quality of evidence but no current FDA approval (tier 2), or any evidence of clinical benefit for selecting a treatment based on the presence of the specific variant (tier 3).6

Germline mutations have been identified in some cancer-associated genes that were previously thought to only have acquired mutations. For example, germline mutations in EGFR have now been identified in patients with NSCLC, including mutations that seem to confer intrinsic resistance to EGFR inhibitors.28,29

Conversely, somatic mutations in genes more commonly associated with inherited risk have also been revealed. In the case of BRCA1/2, this has direct therapeutic benefits. There is now little doubt that the benefits of PARP inhibitors extend beyond patients with germline BRCA1/2 mutations. The identification of somatic BRCA1/2 mutations in a significant proportion of tumors, often without the presence of a germline mutation, has led to the ensuing approval of 3 PARP inhibitors in ovarian cancer with either germline or somatic BRCA1/2 mutations.6

There is also emerging evidence that germline and somatic mutations are intricately linked. Earlier analyses of TCGA data across 4034 tumor samples from 12 cancer types identified rare germline mutations in 114 cancer susceptibility- associated genes that were linked to increased somatic mutation frequency.30 This was confirmed in another recent analysis of approximately 6000 tumor samples, in which 412 genetic interactions were identified between germline mutations and major somatic events. Germline mutations may play an important role in the somatic evolution of a tumor.31

Immunogenomic analyses of the TCGA dataset have also uncovered revelations about the interactions between somatic and germline mutations and the antitumor immune response and tumor microenvironment.32

References

  1. Tomczak K, Czerwińska P, Wiznerowicz M. The Cancer Genome Atlas (TCGA): an immeasurable source of knowledge. Contemp Oncol (Pozn). 2015;19(1A):A68-A77. doi: 10.5114/wo.2014.47136.
  2. Blum, AE. TCGA to complete its final analysis: the PanCanAtlas. National Cancer Institute website. cancer.gov/about-nci/organization/ccg/blog/2017/tcga-pancan-atlas. Published January 9, 2017. Accessed May 29, 2019.
  3. Welcome to the Pan-Cancer Atlas. Cell Press website. cell.com/pb-assets/consortium/pancanceratlas/pancani3/index.html. Accessed June 4, 2019.
  4. Bailey MH, Tokheim C, Porta-Pardo E, et al. Comprehensive characterization of cancer driver genes and mutations. Cell. 2018;173(2):371-385.e318. doi: 10.1016/j.cell.2018.02.060.
  5. Huang KL, Mashl RJ, Wu Y, et al. Pathogenic germline variants in 10,389 adult cancers. Cell. 2018;173(2):355-370.e314. doi: 10.1016/j.cell.2018.03.039.
  6. Thavaneswaran S, Rath E, Tucker K, et al. Therapeutic implications of germline genetic findings in cancer. Nat Rev Clin Oncol. 2019;16(6):386-396. doi: 10.1038/s41571-019-0179-3.
  7. Nagy R, Sweet K, Eng C. Highly penetrant hereditary cancer syndromes. Oncogene. 2004;23(38):6445-6470. doi: 10.1038/sj.onc.1207714.
  8. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA Jr, Kinzler KW. Cancer genome landscapes. Science. 2013;339(6127):1546-1558. doi: 10.1126/science.1235122.
  9. List of cleared or approved companion diagnostic devices (in vitro and imaging tools). United States Food and Drug Administration website. www.fda.gov/medical-devices/vitro-diagnostics/list-cleared-or-approved-companion-diagnostic-devices-vitro-and-imaging-tools. Updated June 4, 2019. Accessed June 4, 2019.
  10. Schwartzberg L, Kim ES, Liu D, Schrag D. Precision oncology: who, how, what, when, and when not? Am Soc Clin Oncol Educ Book. 2017;37:160-169. doi: 10.14694/edbk_174176.
  11. Prasad V, Gale RP. What precisely is precision oncology and will it work? The ASCO Post. January 25, 2017. ascopost.com/issues/january-25-2017/what-precisely-is-precision-oncology-and-will-it-work/. Accessed June 4, 2019.
  12. Jørgensen JT. A paradigm shift in biomarker guided oncology drug development. Ann Transl Med. 2019;7(7):148. doi: 10.21037/atm.2019.03.36.
  13. Berry DA. The brave new world of clinical cancer research: adaptive biomarker-driven trials integrating clinical practice with clinical research. Mol Oncol. 2015;9(5):951-959. doi: 10.1016/j.molonc.2015.02.011.
  14. Subbiah IM, Subbiah V. Exceptional responders: in search of the science behind the miracle cancer cures. Future Oncol. 2015;11(1):1-4. doi: 10.2217/fon.14.204.
  15. Kumar-Sinha C, Chinnaiyan AM. Precision oncology in the age of integrative genomics. Nat Biotechnol. 2018;36(1):46-60. doi: 10.1038/nbt.4017.
  16. Wang Z, Jensen MA, Zenklusen JC. A practical guide to the cancer genome atlas (TCGA). Methods Mol Biol. 2016;1418:111-141. doi: 10.1007/978-1-4939-3578-9_6.
  17. Ellrott K, Bailey MH, Saksena G, et al; MC3 Working Group; Cancer Genome Atlas Research Network. Scalable open science approach for mutation calling of tumor exomes using multiple genomic pipelines. Cell Syst. 2018;6(3):271-281.e277. doi: 10.1016/j.cels.2018.03.002.
  18. Ramos EM, Din-Lovinescu C, Berg JS, et al. Characterizing genetic variants for clinical action. Am J Med Genet C Semin Med Genet. 2014;166C(1):93-104. doi: 10.1002/ajmg.c.31386.
  19. Mateo J, Chakravarty D, Dienstmann R, et al. A framework to rank genomic alterations as targets for cancer precision medicine: the ESMO scale for clinical actionability of molecular targets (ESCAT). Ann Oncol. 2018;29(9):1895-1902. doi: 10.1093/annonc/mdy263.
  20. Lolkema MP, Gadellaa-van Hooijdonk CG, Bredenoord AL, et al. Ethical, legal, and counseling challenges surrounding the return of genetic results in oncology. J Clin Oncol. 2013;31(15):1842-1848. doi: 10.1200/jco.2012.45.2789.
  21. Le DT, Durham JN, Smith KN, et al. Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade. Science. 2017;357(6349):409-413. doi: 10.1126/science.aan6733.
  22. Le DT, Uram JN, Wang H, et al. PD-1 blockade in tumors with mismatch-repair deficiency. N Engl J Med. 2015;372(26):2509-2520. doi: 10.1056/NEJMoa1500596.
  23. Kalia SS, Adelman K, Bale SJ, et al. Recommendations for reporting of secondary findings in clinical exome and genome sequencing, 2016 update (ACMG SF v2.0): a policy statement of the American College of Medical Genetics and Genomics. Genet Med. 2017;19(2):249-255. doi: 10.1038/gim.2016.190.
  24. Mandelker D, Zhang L. The emerging significance of secondary germline testing in cancer genomics. J Pathol. 2018;244(5):610-615. doi: 10.1002/path.5031.
  25. Mandelker D, Zhang L, Kemel Y, et al. Mutation detection in patients with advanced cancer by universal Sequencing of cancer-related genes in tumor and normal DNA vs guideline-based germline testing [erratum in JAMA. 2018;320(22):2381. doi: 10.1001/jama.2018.17511]. JAMA. 2017;318(9):825-835. doi: 10.1001/jama.2017.11137.
  26. Pritchard CC, Mateo J, Walsh MF, et al. Inherited DNA-repair gene mutations in men with metastatic prostate cancer. N Engl J Med. 2016;375(5):443-453. doi: 10.1056/NEJMoa1603144.
  27. Schrader KA, Cheng DT, Joseph V, et al. Germline variants in targeted tumor sequencing using matched normal DNA [erratum in JAMA Oncol. 2016;2(2):279. doi: 10.1001/jamaoncol.2015.6541]. JAMA Oncol. 2016;2(1):104-111. doi: 10.1001/jamaoncol.2015.5208.
  28. Ohtsuka K, Ohnishi H, Kurai D, et al. Familial lung adenocarcinoma caused by the EGFR V843I germ-line mutation. J Clin Oncol. 2011;29(8):e191-192. doi: 10.1200/jco.2010.31.4492.
  29. Yu HA, Arcila ME, Harlan Fleischut M, et al. Germline EGFR T790M mutation found in multiple members of a familial cohort. J Thorac Oncol. 2014;9(4):554-558. doi: 10.1097/JTO.0000000000000052.
  30. Lu C, Xie M, Wendl MC, et al. Patterns and functional implications of rare germline variants across 12 cancer types. Nature Communications. 2015;6:10086. doi: 10.1038/ncomms10086.
  31. Carter H, Marty R, Hofree M, et al. Interaction landscape of inherited polymorphisms with somatic events in cancer. Cancer Discov. 2017;7(4):410-423. doi: 10.1158/2159-8290.Cd-16-1045.
  32. Thorsson V, Gibbs DL, Brown SD, et al. The immune landscape of cancer. Immunity. 2018;48(4):812-830.e814. doi: 10.1016/j.immuni.2018.03.023.

x