Abstract
PURPOSE
The Veterans Health Administration (VHA) is the largest cancer care provider in the United States, with the added challenge of serving more than twice the percentage of patients with cancer in rural areas than the national average. The VHA established the National Precision Oncology Program in 2016 to implement and standardize the practice of precision oncology across the diverse VHA system.
METHODS
Tumor or peripheral blood specimens from veterans with advanced solid tumors who were eligible for treatment were submitted for next-generation sequencing (NGS) at two commercial laboratories. Annotated results were generated by the laboratories and independently using IBM Watson for Genomics. Levels-of-evidence treatment recommendations were based on OncoKB criteria.
RESULTS
From July 2016 to June 2018, 3,698 samples from 72 VHA facilities were submitted for NGS testing, of which 3,182 samples (86%) were successfully sequenced. Most samples came from men with lung, prostate, and colorectal cancers. Thirty-four percent of samples were from patients who lived in a rural area. TP53, ATM, and KRAS were among the most commonly mutated genes. Approximately 70% of samples had at least one actionable mutation, with clinical trials identified as the recommended option in more than 50%. Mutations in genes associated with a neuroendocrine prostate cancer phenotype were expressed at increased frequency among veterans than in the general population. The most frequent therapies prescribed in response to NGS testing were immune checkpoint inhibitors, EGFR kinase inhibitors, and PARP inhibitors.
CONCLUSION
Clinical implementation of precision oncology is feasible across the VHA health care system, including rural sites. Veterans have unique occupational exposures that might inform the nature of the mutational signatures identified here. Importantly, these results underscore the importance of increasing clinical trial availability to veterans.
INTRODUCTION
Genomic profiling of tumors is increasingly used to guide therapeutic decisions in oncology. The use of US Food and Drug Administration (FDA)–approved therapies that target specific tumor mutations requires genomic analysis to demonstrate mutation expression in tumor tissue or circulating tumor DNA (ctDNA). Eligibility for clinical trials evaluating novel targeted therapies increasingly requires genomic profiling of tumors to ensure the presence of the targeted genomic alteration. Conversely, therapeutic resistance to certain FDA-approved targeted therapies can be predicted by the presence of specific mutations. The development of scalable, cost-effective, high-throughput next-generation sequencing (NGS) technologies has made tumor molecular profiling practical for clinical care.
The Veterans Health Administration (VHA) is the largest integrated health care system in the United States and the largest provider of cancer care in the country. Approximately 50,000 new cancer cases are reported annually through the Veterans Affairs Central Cancer Registry.1 Veterans receive cancer care in VHA facilities across 152 hospitals and 1,400 clinics nationwide.
CONTEXT
Key Objective
What were the findings of molecular testing of advanced solid-tumor malignancies performed through the Veterans Health Administration National Precision Oncology Program?
Knowledge Generated
From July 2016 to June 2018, 3,182 samples were sequenced, of which 34% were from patients in rural areas. Approximately 70% of samples had at least one actionable mutation, with clinical trials identified as the recommended option for more than 50%. Mutations in genes associated with a neuroendocrine prostate cancer phenotype were expressed at increased frequency among veterans than in the general population.
Relevance
Our findings demonstrate the feasibility of implementing a large-scale precision oncology program across heterogenous and geographically dispersed systems such as the VHA, including rural settings with limited pathology infrastructure. Veterans remain a unique population in terms of demographics and exposure history; comparison between veteran and civilian tumor mutational profiles may shed light on tumorigenesis and potential etiologies for unidentified mutational signatures.
With a common, systemwide electronic medical record system, the VHA represents an ideal health care system to establish a nationwide, clinical-genomic precision oncology program and database. However, there are potential obstacles to implementing a precision oncology program across the VHA that are not shared by large academic cancer centers. For example, VHA facilities that provide cancer care vary from large medical centers affiliated with academic institutions that possess significant oncology expertise and infrastructure, to small rural clinics that may lack the resources to process tissue for molecular analysis. In addition, 36% of VHA patients with cancer live in rural areas, compared with the national average of 14%.2
The VHA established the National Precision Oncology Program (NPOP) in 2016 to address these challenges.3 The NPOP was tasked with creating guidelines for genomic testing of clinical tumor samples across the VHA health care system; a particular emphasis was placed on making NGS technologies available to veterans in rural areas, facilitating access to FDA-approved targeted therapies and immune checkpoint inhibitors, and increasing participation in clinical trials. Here, we discuss the genomic profile of the 3,182 samples that were successfully sequenced through the NPOP and the targeted or immunotherapeutic drugs administered on the basis of molecular testing.
METHODS
Patients
Genomic testing was recommended for patients with advanced solid tumors who were medically eligible to receive additional treatment with a targeted or immunotherapeutic drug. Tumor sequencing was performed as part of standard clinical oncology care; therefore, no institutional review board–approved research protocol or informed consent was required. Veterans with rural-urban commuting area codes greater than one, based on their ZIP code, received a rural designation, as defined by the VHA Office of Rural Health.
Tumor Sequencing
Formalin-fixed, paraffin-embedded tumor sections or peripheral blood specimens for ctDNA testing were sent to one of two Clinical Laboratory Improvement Amendments–certified genomic laboratories: Personalis (Menlo Park, CA) or Personal Genome Diagnostics (PGDx; Baltimore, MD). Molecular testing was performed using one of five panels (PGDx CancerSELECT 88, 125, and 203; Personalis ACE CancerPlus; and PGDx PlasmaSELECT 64) that included between 64 and 203 cancer-associated genes with a large degree of overlap among the panels. Samples submitted to Personalis were sequenced at 500-fold coverage, and those submitted to PGDx were sequenced between 1,000- and 1,500-fold coverage to improve the sensitivity of testing and identification of low-frequency mutated alleles. Differences in DNA sequence coverage were based on specifications of the particular assay. The panels were constructed to identify base and missense substitutions, insertions and deletions, copy number variations, and selected gene rearrangements and fusions. Microsatellite instability (MSI) status, where available, was based on NGS results (rather than immunohistochemistry or polymerase chain reaction). Tumor mutational burden (TMB) was defined and calculated using IBM Watson for Genomics (WfG) (Armonk, NY) methodology (Appendix) and was only available to providers through VHA precision oncology consultation. Hematoxylin and eosin–stained slides were analyzed by a pathologist at the testing laboratories to verify tumor histology and the percent tumor content in each sample. Additional details of these commercial tests can be found on the PGDx and Personalis Websites.
Levels of Evidence for Actionable Mutations
Annotated clinical reports of mutations identified by NGS testing were provided by PGDx and Personalis. On the basis of data in the variant call files, IBM WfG provided a second annotated report that included FDA-approved targeted therapies and immunotherapies and available clinical trials. The levels of evidence that supported treatment recommendations were based largely on OncoKB criteria.4
RESULTS
Patient Characteristics
Between July 2016 and June 2018, 3,698 tumor and blood samples from 72 VHA facilities were submitted through the NPOP for NGS testing. The number of samples sent for molecular testing steadily increased over the 2-year reporting period (Fig 1A). Of the samples submitted for NGS testing, 35% were from patients in rural areas, which is consistent with the higher percentage of VHA patients with cancer who live in rural areas, compared with the rest of the nation.2
FIG 1.
Overview of the Veterans Health Administration (VHA) National Precision Oncology Program (NPOP) cohort. (A) Number of tumor samples sent for next-generation sequencing testing, reported per month since program inception in 2016. (B) Age distribution of patients included in VHA NPOP cohort. (C) Most common cancer diagnoses of those patients who underwent NGS testing of tumor samples. H&N, head and neck.
The demographics of the patients tested in the NPOP were consistent with a recent report from the VA Central Cancer Registry,5 with men comprising 97% of VHA patients with cancer and a median age of approximately 72 years (Fig 1B). Adenocarcinomas of the lung, colon, and prostate were the three most frequent tumor types tested (Fig 1C), at 18%, 8%, and 29%, respectively, which is consistent with their prevalence among VHA patients with cancer.5
Molecular Analysis
The quantity and quality of DNA obtained from tumor biopsy specimens can vary depending on the tumor type, biopsy site, and type of biopsy performed. Fine-needle biopsies are more likely to yield inadequate DNA content compared with biopsy specimens from larger-bore core needles.6,7 Needle biopsy specimens of lung cancer, which often have significant stromal tissue, and bone metastases are more likely to be associated with a low yield of tumor DNA, often insufficient for NGS testing.8,9 Of the 3,698 samples submitted for molecular analysis, 3,182 (86%) had sufficient quantity and quality of tumor DNA to be successfully sequenced. Samples were tested on one of five cancer gene panels, with 51% and 32% analyzed on the 125 and 181 cancer gene panels, respectively. The ctDNA assay (PlasmaSelect 64) comprised 4% of the total samples submitted for sequencing (Appendix Table A1). The average turnaround time was 19 days.
The most frequently mutated gene in our patient population, regardless of tumor type, was TP53. Mutations in TP53 were identified in 66% of all sequenced samples. ATM was the second most frequently mutated gene, occurring in 28% of samples, followed by KRAS mutations (27%). CREBBP, TSC2, and KMT2A were the next most commonly mutated genes (Appendix Table A2). A heat map showing the frequency of allelic variants in the commonly mutated genes across different tumor type is shown in Fig 2. These findings were consistent with reports in other genomic databases (eg, The Cancer Genome Atlas, Memorial Sloan Kettering Integrated Mutation Profiling of Actionable Cancer Targets) where there was a high frequency of TP53 mutations in lung, esophageal, colorectal, and breast cancers.10,11 Unexpectedly, in prostate cancers sequenced through the NPOP, mutations in SMARCA4, CREBBP, TSC2, KMT2A, and NOTCH1 were more prevalent than mutations in AR, PIK3CA, and KRAS.12-15 Other tumor types had expected mutation frequencies, including glioblastoma, in which EGFR mutations, notably the EGFRvIII mutation, are common.16
FIG 2.
Frequency of the 20 most common gene variants displayed across tumor types. Color signifies increased frequency from light to dark. H&N, head and neck.
We also identified 10 NTRK gene rearrangements. These included ETV6-NTRK3 fusions in patients with non–small-cell lung cancer, papillary thyroid cancer, and adenocarcinomas of the colon and breast (Appendix Table A3). NTRK gene rearrangements are clinically relevant with the recent FDA approval of larotrectinib, an NTRK inhibitor for treatment of tumors expressing NTRK gene rearrangements.17
Treatment Recommended and Prescribed
Clinical annotations of NGS results were provided by PGDX and Personalis, in addition to a separate WfG analysis. The percentage of actionable mutations and their classification on the basis of levels of evidence derived from the WfG report are shown in Fig 3A. The breakdown of treatment recommendations on the basis of levels of evidence in each of the most frequent tumor types is shown in Fig 3B. Overall, an actionable mutation was identified in 70% of the samples sequenced. Approximately 9% of tumors expressed a mutation for which there was a corresponding FDA-approved targeted drug or immunotherapy (levels 1 and 2A). A level 2B treatment recommendation for off-label use of a targeted drug or immunotherapy in a different tumor type from the FDA-approved indication was made in 9% of the cases. Most samples (approximately 52%) expressed one or more actionable mutations for which there was no FDA-approved therapy, either on or off label. Under these circumstances, a clinical trial was the recommended treatment option (levels of evidence 3A and B). Not surprisingly, we found that tumors such as melanoma, for which there are several FDA-approved targeted therapies on the basis of a relatively high frequency of BRAF mutations, had the highest percentage of level 1 treatment recommendations (Fig 3B).
FIG 3.
Clinical actionability of mutations discovered by next-generation sequencing testing. (A) Actionable alterations were annotated using Watson for Genomics (WfG) according to OncoKB levels of evidence: Food and Drug Administration (FDA)-recognized predictive biomarkers (level 1), standard-of-care biomarkers predictive of response to FDA-approved therapies (level 2), and biomarkers predictive of response to therapeutic agents currently under investigation (level 3), with levels 2 and 3 further categorized by whether evidence existed for the particular tumor type in question (2A, 3A) or a different tumor type (2B, 3B). The distribution of the highest level of evidence across all patient samples is displayed. (B) Distribution of highest level of evidence of actionable mutations by tumor type. H&N, head and neck.
A total of 136 patients received the FDA-approved targeted therapy or immunotherapy recommended in the annotated clinical report, either on- or off-label, on the basis of the NGS results. Programmed cell death protein 1 (PD-1)/programmed death-ligand 1(PD-L1) targeted immune checkpoint inhibitors were the most frequently prescribed treatment after NGS results. Of the targeted therapies administered after NGS testing, 36 were anti-PD-1 or PD-L1 antibodies. Many of the patients who received immune checkpoint inhibitors had advanced-stage lung adenocarcinoma, for which checkpoint inhibitors are FDA approved as a first- or second-line therapy without requiring NGS testing. Therefore, it was difficult to determine whether NGS testing influenced their use in patients with advanced-stage lung cancers. The second most commonly prescribed targeted therapy in response to NGS results were EGFR tyrosine kinase inhibitors.
In terms of specific tumor types, 53 of the patients treated with a targeted therapy or PD-1/PD-L1 immune checkpoint inhibitor after NGS testing had metastatic lung adenocarcinoma (Fig 4). Twenty of these patients received an FDA-approved EGFR tyrosine kinase inhibitor (eg, erlotinib, osimertinib) on the basis of the presence of exon 19 deletions or an exon 21 missense mutation (L858R) that leads to constitutive activation of EGFR. Nine patients were treated with crizotinib or alectinib, including five with an EMLA4-ALK translocation, and two each for a ROS1 translocation and MET splice variants, respectively. Two patients with BRAF V600E mutations received BRAF-MEK inhibitors (Fig 4).
FIG 4.
Clinical annotation and recommendations of non–small-cell lung cancer specimens on the basis of next-generation sequencing results. (A) Distribution of actionable mutations (ie, eligible for on- or off-label use of a Food and Drug Administration–approved targeted therapy or immune checkpoint inhibitor on the basis of NGS test results) demonstrated by NGS. (B) Targeted and immunotherapy drugs prescribed after NGS testing. Data are reported as the name of gene or drug (no.). (*) DNA repair genes (eg, BRCA1, BRCA2, ATM, ATR). (†) Mismatch repair (MMR) genes (eg, MLH1, MSH2, MSH6, PMS2). MSI-H, microsatellite instability–high; TMB, tumor mutational burden.
Thirteen patients with advanced castration-resistant prostate cancer were treated with targeted therapies or immune checkpoint inhibitors on the basis of NGS findings. Seven patients with pathogenic mutations in DNA repair genes were treated off-label with an FDA-approved PARP inhibitor, including four patients with BRCA1 and BRCA2 frameshift mutations and three patients with pathogenic ATM and ATR mutations. Four patients with MSI-high status, high TMB, or pathogenic mutations in specific DNA mismatch repair gene (eg, MLH1) were treated with an immune checkpoint inhibitor (Fig 5).
FIG 5.
Clinical annotation and recommendations of prostate cancer specimens on the basis of next-generation sequencing (NGS) results. (A) Distribution of actionable mutations (ie, eligible for on- or off-label use of an Food and Drug Administration–approved targeted therapy or immune checkpoint inhibitor on the basis of NGS test results) demonstrated by NGS. (B) Targeted therapy and immunotherapy drugs prescribed after NGS testing. Data are reported as the name of the gene or drug (no.). (*) DNA repair genes (eg, BRCA1, BRCA2, ATM, ATR). (†) Mismatch repair (MMR) genes (eg, MLH1, MSH2, MSH6, PMS2). TMB, tumor mutational burden.
Of the 27 patients with colorectal cancer whose treatment was selected according to the NGS results, four were treated with a PD-1 checkpoint inhibitor, on the basis of MSI-high status or high TMB. EGFR antibodies were also frequently used in colorectal cancers that lacked the pathogenic mutations in KRAS, NRAS, and BRAF that predict for therapeutic resistance to the anti-EGFR antibodies cetuximab and panitumumab (Fig 6).
FIG 6.
Clinical annotation and recommendations of colorectal cancer specimens on the basis of next-generation sequencing (NGS) results. (A) Distribution of actionable mutations (ie, eligible for on- or off-label use of Food and Drug Administration–approved targeted therapy or immune checkpoint inhibitor on the basis of NGS test results) demonstrated by NGS. (B) Targeted therapy and immunotherapy drugs prescribed after NGS testing. Data are reported as the name of the gene or drug (no.). (*) Mismatch repair (MMR) genes (eg, MLH1, MSH2, MSH6, PMS2). MSI-H, microsatellite instability–high.
DISCUSSION
From the inception of the program, we faced the challenge of implementing precision oncology in the backdrop of a higher percentage of VHA patients with cancer living in rural areas compared with the national average. Health care disparity among patients living in rural areas is not unique to the VHA. Recent published data from the Centers for Disease Control and Prevention indicate rural areas in the United States have higher cancer mortality rates than do urban areas, despite lower adjusted cancer incidence rates.18 Moreover, although adjusted death rates for all cancer sites are decreasing, the decrease is at a slower pace in rural areas, further widening the rural-urban disparity in cancer-related mortality.18 One factor contributing to the widening gap in cancer death rates seems to be limited access to health care and cutting-edge technologies in rural areas. We made a concerted effort to enroll rural VHA facilities in the NPOP and worked closely with pathologists at rural sites to improve the selection of samples for NGS testing to improve testing success rates. The result of these efforts was an 86% success rate for sequencing samples across the VHA system, comparable to rates reported in other large-scale characterization studies performed largely at academic centers.11,19-22 We implemented blood-based ctDNA assays to enable rural sites with limited pathology infrastructure to participate in the program through a simple blood collection. As a consequence, 35% of samples submitted for analysis and 39% of patients receiving targeted therapies agents were from rural areas. Furthermore, 4.5% of patients tested using ctDNA assay were treated with targeted therapies, as compared with 4.3% of patients who received tumor tissue–based assays. The NPOP experience, therefore, may inform strategies to maximize participation of patients with cancer who live in rural areas in precision medicine programs in other health care systems.
We found numerous similarities in the overall mutational pattern among the veteran population to previous molecular profiling studies performed in the general population. For example, TP53 was one of the most commonly mutated genes in the VHA cancer population, consistent with previous large-scale molecular profiling efforts outside the VHA.11,19,23-25 KRAS mutations were identified in more than one-fifth of our patients, with the majority occurring in GI malignancies. These findings are clinically significant, because pathogenic KRAS, NRAS, and BRAF mutations are associated with therapeutic resistance to FDA-approved antibodies and small molecules targeting EGFR in colorectal and non–small-cell lung cancers, respectively.26 Mutations in APC, PIK3CA, CREBBP, and several other genes have been associated with oncogenesis in more common malignancies such as colorectal, lung, and breast carcinomas. Although no targeted therapies are currently approved for these mutations, PI3K inhibitors are showing promising clinical activity in breast cancers that express activating PIK3CA mutations.27 Similarly, therapies are in development that specifically target colorectal cancers that express mutant APC.28 Mutations in ALK, ROS1, and RET already have corresponding targeted therapies that are FDA approved in several malignancies.29
Analysis of our data showed several differences in both overall and specific frequency of gene variants. ATM mutations were identified in almost one-quarter of all patient samples sequenced through the NPOP, which seems to be a higher rate than those reported in other non-VHA studies.11,24 ATM plays a key role in repair of DNA double-strand breaks. Mutations in ATM are associated with numerous cancers, best described in hematologic malignancies, but also various solid tumors. However, the relatively increased rate of pathogenic ATM mutations seen in our patient population, particularly those with gastric or neuroendocrine tumors, may provide a potential therapeutic target, because preliminary clinical data suggest benefit from the use of PARP inhibitors in ATM-deficient cancers.30 ATM and mutations in other DNA mismatch repair genes have been linked to germline mutations. Although we did not sequence normal tissue as part of the NPOP, patients whose tumors expressed potential germline mutations were offered genetic counseling.
The molecular architecture of prostate cancers sequenced through the NPOP seems to differ from that previously reported in large, non-VHA, prostate cancer genomic databases. As shown in Fig 5, there were higher frequencies of mutations in KMT2A, NOTCH1, SMARCA4, TSC2, and CREBBP among veterans, compared with The Cancer Genome Atlas prostate data.12 Although these genes may not share a common functional theme, they are expressed at an increased frequency in neuroendocrine prostate cancers.13 For example, the frequency of SMARCA4 mutations is reportedly less than 5% in prostate adenocarcinomas.14,15 In contrast, SMARCA4 mutations have been reported in approximately 15% of neuroendocrine prostate cancers.13-15 KMT2A is a member of the lysine methyltransferase 2 (KMT2) gene family.31 Whereas other KMT2 family members (KMT2C, KMT2D) are frequently mutated in a variety of solid tumors and hematologic malignancies, KMT2A mutations are infrequently found in prostate adenocarcinomas (less than 5%). However, in neuroendocrine prostate cancers, the frequency of KMT2A mutations has been reported to be 14%, with similar findings for DNMT2 and CREBBP.13 Transformation from prostate adenocarcinoma to a neuroendocrine phenotype has been reported in approximately 17% of patients, coincident with disease progression.32 That process is thought to be multifactorial, including specific gene rearrangements and epigenetic changes that were not tested here.13 Although patients whose tumors were sequenced almost universally had advanced prostate cancer, the majority of samples sent for sequencing were from the original archived prostate biopsy specimens rather than metastatic sites of disease. Therefore, we speculate that these genomic features suggestive of a neuroendocrine phenotype may have been present before initiation of androgen-deprivation therapy.
Overall differences in the genomic signatures in our patient population compared with previous non-VHA studies are likely to be multifactorial in etiology. As previously posited, veterans represent a demographically unique population; the vast majority of patients reported here were men, which skews findings toward profiles associated with common male cancers (ie, lung, prostate, colorectal). Veterans also differ from the general population in their occupational exposure history (eg, Agent Orange), which increases risk of specific tumor types and may lead to a distinct mutational profile, compared with sporadic cases of similar histology. Additional investigation directly comparing the mutational signatures of tumors sequenced through the NPOP to those of patients in the general population may be crucial to link particular occupational or environmental exposures to mutations expressed in their tumors.
Importantly, we have shown that implementation of NGS testing influenced the treatment of veterans across the country. Our patients had a similar percentage of actionable mutations for which there was an FDA-approved targeted drug or immune checkpoint inhibitor levels of evidence 1 and 2A), compared with those reported from Memorial Sloan Kettering (Fig 3A).11 Consistent with the findings from the Memorial Sloan Kettering study, the majority of tumors tested through the NPOP expressed actionable mutations for which there was no FDA-approved therapy.11 If eligible, these patients should be considered for a clinical trial, which underscores the need to increase access to oncology clinical trials for veterans.
In conclusion, our findings shed light on some unique aspects of the demographics and mutational patterns seen in the veteran patient population. Our successful implementation of an NPOP demonstrates the feasibility of delivering timely, cutting-edge precision oncology care across the large, heterogeneous, and geographically dispersed VHA system. In the process, we successfully overcame the challenges of an increased percentage of patients with cancer who live in rural areas, compared with the general population, as well as operating within the confines of a fixed-budget government health care system, which may serve as a model for implementation of molecularly driven oncology care in rural and underserved communities across the United States.
ACKNOWLEDGMENT
This project is a clinical/operational endeavor under the heading of the Veterans Health Administration (VHA) National Precision Oncology Program, and our manuscript is a reporting of the results (at 2 years) of this clinical program. Similar to funding for other aspects of veteran clinical care through the VHA (eg, laboratory testing, imaging, drugs/therapies), funding for the program is provided by national congressional budgeting (either as a separate program or under general Veterans Health Administration hematology/oncology funding), and no outside (eg, industry, pharmaceutical) funding sources are used for the program. Details of this funding are publicly available as a matter of Congressional budgeting. Establishment of VHA National Precision Oncology Program: Jill E. Duffy, Bradley Hintze, Maulik Shukla, Thomas S. Brettin, Neil R. Conrad, Hyunseung Yoo, Christopher Guertin, Jane A. Looney, Michael J. Kelley, Neil L. Spector.
APPENDIX
Methods: Assessment of Tumor Mutational Burden
Tumor mutational burden (TMB) was defined as the number of coding, nonsynonymous mutations divided by the target region, which is estimated as the size of the genome (in Mbps) covered by the assay. For the gene mutation panels from which data are reported in this article and for which TMB was calculated, region of interest and threshold values for high TMB are as follows: Personalis ACE: region of interest, 0.25 Mb, threshold, 100 mutations/Mb; PGDx CancerSelect125: region of interest, 0.16 Mb, threshold, 56 mutations/Mb.
TABLE A1.
Frequency of the 10 Most Common Tumor Types in Samples Submitted for Circulating Tumor DNA-Based Assays
TABLE A2.
Ten Most Common Variant Genes Detected Across Tumor Types
TABLE A3.
NTRK Fusion Mutations Detected by Next-Generation Sequencing Testing Through the National Precision Oncology Program
Footnotes
Supported by the Veterans Health Administration National Precision Oncology Program (P.J.P.).
AUTHOR CONTRIBUTIONS
Conception and design: Jill E. Duffy, Maulik Shukla, Thomas S. Brettin, Neal R. Conrad, Hyunseung Yoo, Christopher Guertin, Jane A. Looney, Vishal Vashistha, Michael J. Kelley, Neil L. Spector
Collection and assembly of data: Bradley Hintze, Neil L. Spector
Data analysis and interpretation: Pradeep J. Poonnen, Bradley Hintze, Neil L. Spector
Manuscript writing: All authors
Final approval of manuscript: All authors
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/po/author-center.
Pradeep J. Poonnen
Other Relationship: IBM
Christopher Guertin
Employment: Walmart
Stock and Other Ownership Interests: Walmart
Vishal Vashistha
Other Relationship: IBM
Michael J. Kelley
Consulting or Advisory Role: AstraZeneca, Eisai, IBM Japan
Research Funding: Bavarian Nordic, Novartis, AstraZeneca, Bristol-Myers Squibb
Other Relationship: IBM
Neil L. Spector
Stock and Other Ownership Interests: Eydis Bio, Bessor Pharma
Research Funding: Immunolight
Patents, Royalties, Other Intellectual Property: I am on a patent related to my work with the company Immunolight, and I am listed on a patent through Eydis Bio
No other potential conflicts of interest were reported.
REFERENCES
- 1.Zullig LL, Jackson GL, Dorn RA, et al. Cancer incidence among patients of the US Veterans Affairs Health Care System. Mil Med. 2012;177:693–701. doi: 10.7205/milmed-d-11-00434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.US Department of Agriculture, Economic Research Service Population, income, food insecurity, education, and employment. https://data.ers.usda.gov/reports.aspx?ID=17854
- 3.US Department of Veterans Affairs Office of Research and Development Precision Oncology Program. https://www.research.va.gov/research_in_action/Precision-Oncology-Program.cfm
- 4.Chakravarty D, Gao J, Phillips SM, et al. OncoKB: A precision oncology knowledge base. JCO Precis Oncol. doi: 10.1200/PO.17.00011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zullig LL, Sims KJ, McNeil R, et al. Cancer incidence among patients of the U.S. Veterans Affairs Health Care System: 2010 update. Mil Med. 2017;182:e1883–e1891. doi: 10.7205/MILMED-D-16-00371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Schneider F, Smith MA, Lane MC, et al. Adequacy of core needle biopsy specimens and fine-needle aspirates for molecular testing of lung adenocarcinomas. Am J Clin Pathol. 2015;143:193–200, quiz 306. doi: 10.1309/AJCPMY8UI7WSFSYY. [DOI] [PubMed] [Google Scholar]
- 7.Łukasiewicz E, Ziemiecka A, Jakubowski W, et al. Fine-needle versus core-needle biopsy - which one to choose in preoperative assessment of focal lesions in the breasts? Literature review. J Ultrason. 2017;17:267–274. doi: 10.15557/JoU.2017.0039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Richter PD, Ilacqua J. Correlation between biopsy type and insufficient tissue availability for biomarker testing in five solid cancer types. J Clin Oncol. 2017;31(suppl 15):e22136. [Google Scholar]
- 9.Wu JS, Goldsmith JD, Horwich PJ, et al. Bone and soft-tissue lesions: What factors affect diagnostic yield of image-guided core-needle biopsy? Radiology. 2008;248:962–970. doi: 10.1148/radiol.2483071742. [DOI] [PubMed] [Google Scholar]
- 10.Kandoth C, McLellan MD, Vandin F, et al. Mutational landscape and significance across 12 major cancer types. Nature. 2013;502:333–339. doi: 10.1038/nature12634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. doi: 10.1038/nm.4333. Zehir A, Benayed R, Shah RH, et al: Mutational landscape of metastatic cancer revealed from prospective clinical sequencing of 10,000 patients. Nat Med 23:703-713, 2017 [Erratum: Nat Med 23:1004, 2017] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cancer Genome Atlas Research Network The molecular taxonomy of primary prostate cancer. Cell. 2015;163:1011–1025. doi: 10.1016/j.cell.2015.10.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Beltran H, Prandi D, Mosquera JM, et al. Divergent clonal evolution of castration-resistant neuroendocrine prostate cancer. Nat Med. 2016;22:298–305. doi: 10.1038/nm.4045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6:pl1. doi: 10.1126/scisignal.2004088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Cerami E, Gao J, Dogrusoz U et al. The cBio cancer genomics portal: An open platform for exploring multidimensional cancer genomics data. Cancer Discov 2:401-404, 2012. [DOI] [PMC free article] [PubMed]
- 16. doi: 10.1016/j.cell.2013.09.034. Brennan CW, Verhaak RGW, McKenna A, et al: The somatic genomic landscape of glioblastoma. Cell 155:462-477, 2013 [Erratum: Cell 57:753, 2014] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Drilon A, Laetsch TW, Kummar S, et al. Efficacy of larotrectinib in TRK fusion-positive cancers in adults and children. N Engl J Med. 2018;378:731–739. doi: 10.1056/NEJMoa1714448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Henley SJ, Anderson RN, Thomas CC, et al. Invasive cancer incidence, 2004-2013, and deaths, 2006-2015, in nonmetropolitan and metropolitan counties - United States. MMWR Surveill Summ. 2017;66:1–13. doi: 10.15585/mmwr.ss6614a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Meric-Bernstam F, Brusco L, Shaw K, et al. Feasibility of large-scale genomic testing to facilitate enrollment onto genomically matched clinical trials. J Clin Oncol. 2015;33:2753–2762. doi: 10.1200/JCO.2014.60.4165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mody RJ, Wu YM, Lonigro RJ, et al. Integrative clinical sequencing in the management of refractory or relapsed cancer in youth. JAMA. 2015;314:913–925. doi: 10.1001/jama.2015.10080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Uzilov AV, Ding W, Fink MY, et al. Development and clinical application of an integrative genomic approach to personalized cancer therapy. Genome Med. 2016;8:62. doi: 10.1186/s13073-016-0313-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Harris MH, DuBois SG, Glade Bender JL, et al. Multicenter feasibility study of tumor molecular profiling to inform therapeutic decisions in advanced pediatric solid tumors: The individualized cancer therapy (ICAT) study. JAMA Oncol. 2016;2:608–615. doi: 10.1001/jamaoncol.2015.5689. [DOI] [PubMed] [Google Scholar]
- 23.Oberg JA, Glade Bender JL, Sulis ML, et al. Implementation of next generation sequencing into pediatric hematology-oncology practice: Moving beyond actionable alterations. Genome Med. 2016;8:133. doi: 10.1186/s13073-016-0389-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hovelson DH, McDaniel AS, Cani AK, et al. Development and validation of a scalable next-generation sequencing system for assessing relevant somatic variants in solid tumors. Neoplasia. 2015;17:385–399. doi: 10.1016/j.neo.2015.03.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Parsons DW, Roy A, Yang Y, et al. Diagnostic yield of clinical tumor and germline whole-exome sequencing for children with solid tumors. JAMA Oncol. 2016;2:616–624. doi: 10.1001/jamaoncol.2015.5699. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Slebos RJ, Kibbelaar RE, Dalesio O, et al. K-ras oncogene activation as a prognostic marker in adenocarcinoma of the lung. N Engl J Med. 1990;323:561–565. doi: 10.1056/NEJM199008303230902. [DOI] [PubMed] [Google Scholar]
- 27.Alpelisib extends PFS in PIK3CA-mutant breast cancer. Cancer Discov. 2019;9:6–7. doi: 10.1158/2159-8290.CD-NB2018-146. [DOI] [PubMed] [Google Scholar]
- 28.Zhang L, Theodoropoulos PC, Eskiocak U, et al. Selective targeting of mutant adenomatous polyposis coli (APC) in colorectal cancer. Sci Transl Med. 2016;8:361ra140. doi: 10.1126/scitranslmed.aaf8127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Asao T, Takahashi F, Takahashi K. Resistance to molecularly targeted therapy in non-small-cell lung cancer. Respir Investig 57:20-26, 2019. [DOI] [PubMed]
- 30.Choi M, Kipps T, Kurzrock R. ATM mutations in cancer: Therapeutic implications. Mol Cancer Ther. 2016;15:1781–1791. doi: 10.1158/1535-7163.MCT-15-0945. [DOI] [PubMed] [Google Scholar]
- 31.Rao RC, Dou Y. Hijacked in cancer: The KMT2 (MLL) family of methyltransferases. Nat Rev Cancer. 2015;15:334–346. doi: 10.1038/nrc3929. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Aggarwal R, Huang J, Alumkal JJ, et al. Clinical and genomic characterization of treatment-emergent small-cell neuroendocrine prostate cancer: A multi-institutional prospective study. J Clin Oncol. 2018;36:2492–2503. doi: 10.1200/JCO.2017.77.6880. [DOI] [PMC free article] [PubMed] [Google Scholar]