NGS Testing in Non–Small Cell Lung Cancer

J Clin Pathways. 2023;9(5):27-35. doi:10.25270/jcp.2023.09.01

The era of precision oncology has dawned, bringing the promise of more effective, better-tolerated, targeted therapies for many cancers. Precision oncology refers to the identification of characteristics (eg, biomarkers, molecular aberrations, etc) of cancers that can drive treatment decisions or forecast prognosis.¹ Precision oncol­ogy strategies are an important and growing component of cancer care. There are more than 90 US Food and Drug Administration (FDA)-approved targeted thera­pies available for use in eligible patients with cancer.² A recent oncology pipeline report showed that approximately 55% of all oncology clinical trials involved the use of biomarkers.2 Predictive biomarker testing, which can help identify patients who could benefit from targeted therapies, is a cornerstone of personalized medicine; it can lead to more rapid diagnosis while informing treatment decisions that could re­sult in improved patient outcomes and systemic efficiencies.^2,3 A recent clinical opin­ion indicated that, at a minimum, patients with metastatic or advanced cancer should undergo genomic sequencing if the presence of one or more specific genomic altera­tions can guide the use of, or exclusion from, targeted treatments for their disease. In addition, multigene panel–based assays should be used if more than one biomarker-linked therapy is approved for a patient’s disease.³

Although few question the importance of next-generation sequencing (NGS) testing in modern oncology care, the process of gathering primary molecular data, integrating them into electronic health records, and optimally using them as part of a clinical workflow remains far from seamless. Numerous challenges persist regarding data standards, interoperability, and operational inefficiencies. Clinicians often have a limited understanding of biomarker strategies and can experience difficulties in man­aging the growing amount of genomic knowledge required to care for patients.^2,4 As a result, many patients do not receive the most effective personalized treatments because of challenges associated with integrating predictive biomarker testing into clinical care.² Utilization of biomarker testing differs greatly across practice settings, tumor types, and biomarkers, with varying adherence to testing guidelines. Imple­mentation challenges are demonstrated in non–small cell lung cancer (NSCLC); even though more than 70% of these patients have tumors with biomarker alterations re­lated to therapeutic options, many still do not receive biomarker testing.² Further­more, while the cost of early comprehensive genomic profiling (CGP) was a barrier compared to more traditional testing methodologies, the costs are significantly reduced today.

For those patients who do receive biomarker testing, many patients with actionable results do not subsequently receive appropriate targeted therapies. Reports indicate that more than one-third of patients with cancer fail to receive targeted treatments because of suboptimal testing practices related to quality or sample management issues.²

The NGS Institute (NGSi) was created in 2022 to address the challenges associated with biomarker, liquid biopsy, and NGS/ CGP testing. These modalities are poorly understood by many stakeholders in cancer care. There is inconsistent reimburse­ment and inconsistent availability to patients, resulting in missed opportunities for treatment and management. The mis­sion of the NGSi is to promote education, research, and best practices in the appropriate use and reimbursement of com­panion, complementary, and continuity of care diagnostics, including biomarker, liquid biopsy, and NGS/CGP testing.

The NGSi Charter consists of 4 principles:

1. Curate educational and background materials for multiple stakeholders, including payers, employers, practices, physicians, and patients.
2. Develop transparent, consensus-based, patient-centric best practices in biomarker and NGS testing in collaboration with payers, employers, physicians/ practices, testing companies, and patient advocacy groups. The goal is to influence and align payment and administrative policies among payers and employers through principles for payment that address standards for ordering testing, use of test results in treatment selection, and frequency of testing.
3. Sponsor research to promote the appropriate use of biomarker and NGS testing.
4. Ensure that oncologists and patients have the best information available about the full range of effective treatment options, including clinical trials, for every patient.

The overall goals of the NGSi are to: (1) always be patient-centric in our discussions and efforts, and (2) balance clinical utility, testing efficiency, and cost to positively impact clinical outcomes.

Biomarker testing/profiling in NSCLC was chosen as the first area of focus. A subcommittee of various stakeholders was convened to examine the current state of NGS biomarker test­ing in NSCLC. Represented were providers, clinical pathway administrators, payers, employer groups, the laboratory indus­try, and the lung cancer advocacy community. The wide range of stakeholders was intentional in that each stakeholder brings a different perspective with respect to the utilization of NGS testing in the care of patients with NSCLC today. The views expressed and documented in this paper are those of our par­ticipants based upon their experience and expertise in the area of NGS/biomarker testing from the perspective of the patient and clinical outcomes. As the NGSi embarks on the effort to improve the utilization of biomarker testing in the NSCLC population, it is acknowledged that many other groups have similar ongoing efforts focused on the standards of practice representing quality care, irrespective of the setting—whether it be detection, establishing eligibility for a targeted therapy, promulgation of best practices in testing and matched thera­py selection, or disease monitoring. It is not the intent of the NGSi to duplicate ongoing efforts but rather to add value and possibly establish partnerships to improve utilization.

NSCLC was chosen as the first cancer type for addressing the underutilization of NGS/CGP testing. Studies examining practice-based data from more than 190 US hospital systems have estimated that only 65%-75% of patients with an action­able mutation receive targeted therapies.² Another study of pa­tients with NSCLC in the Veterans Affairs National Precision Oncology Program revealed that more than 30% of patients with highly actionable gene variants received chemotherapy rather than more effective, targeted treatments.²

In a study attempting to quantify the extent to which clinical practice gaps affect the treatment of patients with ad­vanced NSCLC (aNSCLC) in the US, data from a large popu­lation of newly diagnosed patients were analyzed to estimate the number of patients who could have, but did not, benefit from a personalized treatment. Of 144 486 newly diagnosed patients with NSCLC identified, 38 068 were determined to be patients with actively managed aNSCLC. The aNSCLC subset of patients were the focus, given that this is a subpopu­lation for whom there is little disagreement that biomarker testing should be conducted. The analysis focused on the in­dependent and cumulative impact of clinical gaps occurring during 7 discrete steps of the precision oncology pathway from diagnosis to treatment.² The gaps observed at each pro­cess step can be broadly summarized as follows:²

1. Biopsy referral: Initial solid or liquid biopsy was never performed.
2. Biospecimen collection: Challenges with biospecimen collection, including insufficient tissue or tumor cell content of initial biopsy or rebiopsy, inhibiting biomarker testing and its accuracy.
3. Biospecimen evaluation/pathology: Biospecimen tumor cell content was overestimated or used up in other pathology lab evaluations (immunohisto­chemistry [IHC], polymerase chain reaction, other) prior to NGS preparation, inhibiting biomarker testing and/or its accuracy.
4. Biomarker test ordering: Appropriate testing was not ordered, limited-scope gene panel was negative but used all available tissue, or treatment began before testing was ordered.
5. Biomarker testing performance: Biomarker testing provided inconclusive or false-negative results.
6. Test result reporting: As a result of turnaround time delays, treatment was initiated without consideration of test results.
7. Treatment decision: Targeted treatment was not selected despite positive test results, often attributed to lost reports, missed reports, or overly complex report formats.

The authors estimated that for every 1000 patients with newly diagnosed aNSCLC who are potentially eligible for targeted therapy, 497 patients were lost due to clinical gaps associated with getting biomarker testing. Another 147 bio­marker-positive patients did not receive the appropriate tar­geted therapy. As a result, an estimated 644 of every 1000 patients with newly diagnosed aNSCLC (64.4%) are not benefiting from precision oncology care options appropriate for their diseases and will likely have suboptimal outcomes. Simply restated, out of 1000 newly diagnosed patients with aNSCLC, only 356 patients benefited from precision oncol­ogy technology.²

NGS testing enables the simultaneous assessment of all 3 of the “must-test” genes in lung cancer (EGFR, ALK, ROS1) as well as each of the genes suggested for inclusion in larger pan­els (BRAF, RET, ERBB2 (HER2), KRAS, MET).⁵ (See supple­mental material for the Appendix: Practice Evolution to Current Genomic Biomarker Testing Recommendations.) In addition to small mutations, NGS assays can detect fusions/rearrangements and copy number changes in the targeted genes. NGS methods typi­cally require less input DNA (tissue sparing) and can accommo­date smaller samples with lower concentrations of malignant cells.⁵ However, National Comprehensive Cancer Network (NCCN) guidelines recommend a broad, panel-based approach and to consider RNA-based testing in patients with no iden­tified drivers. While RNA-based techniques require a higher-quality sample, they capture complex genomic alterations that are not easily discovered through DNA-based testing.^7,8

Reimbursement Barriers

As previously noted, utilization of broad comprehensive test­ing for biomarkers in NSCLC is suboptimal. This observation leads to the question of what barriers are in place that discour­age the use of comprehensive genomic profiling. Reimburse­ment will most likely rise to the top of the list of barriers, as adoption by payers was initially slow. Reimbursement is built upon the Healthcare Common Procedure Coding System (HCPCS), which represents two levels of coding. Billing for laboratory services falls under the HCPCS level I code set for procedures, commonly referred to as the Current Procedural Terminology (CPT) code set.⁹

In 2020, the Centers for Medicare & Medicaid Services (CMS) released its updated National Coverage Determination (NCD) 90.2 for NGS testing. The first NCD was published in 2018. In the current version, numerous modifications and refinements were made in response to the growing evidence of the clinical utility of NGS testing. The NCD 90.2 is only applicable to diagnostic lab tests using NGS for somatic (ac­quired) and germline (inherited) cancer, leaving the determi­nation for coverage for specific diagnostic lab tests using NGS for RNA sequencing and protein analysis to the local Medi­care Administrative Contractors.¹⁰

Pertinent to NSCLC, CMS has determined that NGS test­ing, as a diagnostic laboratory test, is reasonable and necessary and is covered nationally when performed in a Clinical Labo­ratory Improvement Amendments (CLIA)-certified labora­tory, when ordered by a treating physician, and when all the following requirements are met:¹⁰

A. Patient has:

I. either recurrent, relapsed, refractory, metastatic, or advanced stage III or IV cancer; and

II. not been previously tested with the same test using NGS for the same cancer genetic content; and

III. decided to seek further cancer treatment (eg, therapeutic chemotherapy).

B. The diagnostic laboratory test using NGS must have:

I. FDA approval or clearance as a companion in vitro diagnostic; and,

II. an FDA-approved or -cleared indication for use in that patient’s cancer; and,

III. results provided to the treating physician for management of the patient using a report template

to specify treatment options.

CMS will reimburse for NGS testing via a lab-specific Proprietary Laboratory Analysis (PLA) code, eg, 0022U, or via a nonspecified NGS testing code, eg, 81455, 81479. The reimbursement level for the PLA code will be specific to the CLIA-certified lab, whereas the reimbursement level for the nonspecified code will be based upon an actuarial analysis.¹¹

For the most part, private insurers have also responded to the growing evidence of the utility of NGS testing in NSCLC, although at a much slower pace compared to CMS. Prior to 2019, most private payers excluded the utilization of multi-gene molecular testing, eg, NGS testing of any type, be it for comprehensive or hotspot testing. Today, it appears that pri­vate insurers are following the coverage decision flexibility of CMS, utilizing the nonspecified panel code 81479 or 81445 as well as the PLA codes. However, it appears that many pri­vate insurers are limiting their coverage to panels of <50 genes (Payer Coverage Policy Review of Comprehensive Genomic Testing, author research, January 2022).

In short, prior to the initial 2018 NCD, payers were slow to adopt NGS testing, resulting in the practice of “stacked coding” in which the lab bills the payer for multiple single test codes that are in the claims system and payable. Utiliza­tion of the PLA code, or the nonspecific NGS testing code, should theoretically eliminate the practice of stacked cod­ing. However, there are more NGS laboratories conducting comprehensive genomic testing than there are labs with PLA codes. These are the labs that will utilize the nonspecific NGS testing code, eg, 81479 or 81455. If the reimbursement level for these nonspecific NGS test codes is not sufficient to cover the cost of the NGS testing, the laboratory will revert to the stacked code billing methodology to achieve a higher reim­bursement. The end result of this practice is that we may not have an accurate sense of the actual rate of NGS testing, and we will see CPT testing codes billed and reimbursed that may have no logical connection to the cancer being treated.

Provider Understanding

As previously noted, clinicians are faced with a limited under­standing of biomarker strategies and can experience difficul­ties in managing the growing amount of genomic knowledge required to care for patients.^2,4 Most oncologists practicing today completed fellowship training prior to the era of NGS-directed targeted therapy considerations for solid tumors. One issue here is that an oncologist’s knowledge and understand­ing is greatly dependent upon the mix of cancer types seen by the practice. A greater understanding of the implications of comprehensive genomic testing for NSCLC will be higher in a practice that has a significant NSCLC population. The im­portance of knowing and understanding the ramifications of comprehensive genomic testing in NSCLC is implied by the opinion that if an oncologist doesn’t know the 9 or 10 bio­markers for which there are FDA-approved targeted therapies for NSCLC, that oncologist is not providing optimal-quality care to patients.

It is important to know and understand the application of biomarker information; this was demonstrated in an analysis in which clinical data from electronic health record were linked with comprehensive genomic testing for 28 998 patients from 275 oncology practices across the US. Of these, 4064 patients with NSCLC were identified. The goal was to explore the as­sociation between biomarker genomics and clinical outcomes. The time frame for the analysis was January 1, 2011, to January 1, 2018. Of the 4064 patients identified with NSCLC, 871 pa­tients had an EGFR, ALK, or ROS1 alteration. Approximately two-thirds of the population identified with a biomarker al­teration went on to be treated with a targeted treatment option. Comparing the overall survival of the population receiving tar­geted treatments vs the population not receiving targeted treat­ments, the treated population had an overall survival of 18.6 months vs 11.4 months in the untreated population.¹²

It is crucial to conduct comprehensive genomic testing as the first step to formulating a treatment plan, as further confirmed by Schoenfeld et al. They observed that 15% of pa­tients treated with an immune checkpoint inhibitor followed by a tyrosine kinase inhibitor (TKI) developed a severe immune-related adverse effect (irAE). The irAE was not linked to any specific immune checkpoint inhibitor. Severe irAEs were most common among those who received the TKI within 3 months of a prior immune checkpoint inhibitor (24%), as compared with those who received the TKI at more than 3 to 12 months after an immune checkpoint inhibitor (13%). By contrast, no severe irAEs were identified among patients treated with the TKI followed by an immune checkpoint inhibitor. All patients with irAEs required steroids, and most required hospitaliza­tion.¹³ The administration of an immune checkpoint inhibitor in a patient with an EGFR alteration will jeopardize optimal treatment of the patient. Sequencing of therapy matters greatly in these patients. Thus, comprehensive genomic testing should be the first step to making a treatment decision. The second step should be to act on the reported findings.

Laboratory Barriers

There appear to be several different laboratory-related barri­ers. One barrier is obtaining an adequate tissue sample. This includes adequate quantity of tissue, microdissection where ap­propriate, and quality of tissue, with proper handling and time­ly preparation of tissue by knowledgeable surgical and pathol­ogy specialists. Tissue samples serve to support a diagnosis and are foundational for molecular testing. However, the quantity of available tissue often presents a key constraint for patients with advanced disease, for whom minimally invasive tissue bi­opsy typically returns small samples. It is recommended to use tissue biopsy techniques that deliver the greatest quantity and quality of tissue with the least risk to the patient. Sample pro­cessing should be managed according to biomarker testing re­quirements, because the tissue fixation methodology influences downstream nucleic acid, protein, and morphological analy­ses.¹⁴ The importance of obtaining an adequate tissue sample in NSCLC is twofold: first, the number of actionable molecular targets in NSCLC is higher than in other solid tumors. Second, most patients with NSCLC present with advanced disease, for which curative surgery is no longer feasible. Best-practice guidelines aim to conserve tissue and enable a complete mo­lecular diagnosis so that eligible patients may benefit from tar­geted therapy.¹⁴

A possible solution to inadequate tissue samples is the use of the “liquid biopsy,” a noninvasive diagnostic technique that includes testing of cell-free DNA (cfDNA), or circulat­ing tumor cells, from various body fluids, the most common of which is testing for cfDNA in blood-derived plasma. Liq­uid biopsy testing has both limitations and advantages com­pared to tissue biopsy. It should be considered as a comple­ mentary approach due to its lower sensitivity compared to tissue biopsies; however, liquid biopsy can help provide a molecular diagnosis, particularly when the tissue biopsy sample is insufficient or inadequate for molecular analysis.¹⁴ Our participants noted that both tissue and liquid biopsies could be done to have the data to support the rationale for first-line treatment, simply because the testing results will complement each other.^15,16

Another possible laboratory-related barrier to comprehen­sive testing in NSCLC stems from the reporting. Here again, the barrier is two-fold. First is the complexity of reporting. Several reporting pitfalls have been identified that hinder in­terpretation of test results.¹⁵ The complexity of reporting has increased with the growing number of clinically relevant biomarkers, and there is a need for standardization in report­ing format. Although multimarker panel reports may include information on potentially beneficial classes of treatment, the use of larger panels can identify variants of unknown signifi­cance, potentially complicating interpretation.^15,16

The second identified barrier to laboratory reporting is the inability of oncologists to fully interpret the results; this can be due to either the complexity of the report or to a lack of complete understanding or interpretation of the report due to its particular format. Our stakeholders suggest that a simple solution may be to report all the known essential biomarker results (eg, NCCN-guideline recognized or similar bench­marks) on a separate page from the remainder of the report, thus immediately bringing attention to those results that are the most meaningful. To address the issue of interpretation, key reporting criteria were proposed by the International Organization for Standardization in 2012. These criteria recommend that reports include an interpretation of the re­sults, with cautionary or explanatory notes.¹⁶ A recent obser­vational study of components currently present in NSCLC molecular pathology request forms and reports found that the reporting item considered most important by pathologists and/or molecular biologists and clinicians was the clinical interpretation of the test result; clearly, interpretation is the most important element of the report.^15,16

LUNGevity Foundation

One of our participating stakeholders is the LUNGevity Foundation. The work of LUNGevity spans across the en­tire spectrum of lung cancer care, with the goal of driving change for those with lung cancer today and in the future. While LUNGevity funds primary research, the organization also conducts patient-focused research to better understand the unmet needs related to lung cancer care, and it convenes multistakeholder meetings to streamline the research process and accelerate progress to patients. LUNGevity interfaces with community oncologists who are not NCSLC special­ists, and these oncologists confirm that the complexity of the molecular reporting is overwhelming. For community oncologists, the reporting complexity is a barrier and can influence a provider to opt out of ordering biomarker testing if they are unable to confidently discern how the results should impact patient treatment selection Our participants suggest that we should be a bit more forward thinking in NGSi’s approach to increasing testing utilization, which means we need to determine how to ensure that oncologists are making treatment decisions based on the test outcomes, whether this means developing a quality measure or whether it means cre­ating more education for nurses, nurse navigators, clinicians, and patients around the importance of testing and of waiting to treat based upon the results of the testing.

Another observation is the lack of standardization in both the ordering of tests and, as previously noted, the reporting of test results. This applies not only to the essential biomark­ers but also to additional biomarkers of interest that may be applicable to a clinical trial. In short, standardization of re­porting is needed for both actionable biomarkers associated with targeted drug therapy and also for emerging biomark­ers, given that clinical trials are often first-line options now, especially for rare mutations. Dufraing et al confirmed this in their findings, which identified several trends indicating the need for standardization of the test requesting and report­ing process in Belgium; this led to a proposal to develop a template for complete requesting and reporting, which could harmonize communication flows.¹⁵

In 2022, 65% of US workers with medical insurance were cov­ered under an employer self-funded health insurance plan.¹⁷ Self-funded plans are those in which employers choose to pay for some or all health services for employees directly rather than purchasing health insurance through a traditional insur­ance company. As such, self-funded employers contract for ser­vices through a third-party administrator (TPA) or through a traditional insurance company for administrative services only. The significance of this pertains to the employer’s understand­ing and awareness of the provision of necessary services with respect to the quality as well as the cost of health care services purchased. In some respects, self-funded employers have the ability to drive adoption, simply because they are not subject to state and federal regulatory oversight and can operate and make coverage decisions in ways that fully insured or govern­ment programs cannot.

The broad observation is that, in general, self-funded em­ployers are uninformed of the topics related to best practices in NGS testing, relying on the advice of their benefits broker or consultant, their TPA, and their health plan. When asked what their number one concern is with regard to their health care services, the response is usually cost. The quality of oncology care, or the quality of primary care, is rarely mentioned. To get employers, especially self-funded employers, interested in the quality of care, it would need to be demonstrated that im­ proving the quality of care will also improve the overall cost of care. A clear and concise message addressing the ability of best practices to impact quality and cost of care will provide a foun­dation for discussion and will gain the interest of employers.

As discussed by our participants, the fact that employer groups are not typically thinking about quality of care may be more of a reflection of health plans, consultants, care ad­ministrators, etc, not addressing it. The clinical complexity inherent to NGS may play a role in this. Delegated parties may simply not succeed at being able to contextualize what quality care is and what it needs to look like. Care managers will accept and adhere to treatment guidelines and recom­mendations, but the question remains: Do those guidelines really improve care and outcomes? The reasons for some of these unanswered questions are numerous and were addressed previously. However, health plans and benefits administrators need to do a better job of talking about treatment guidelines and recommendations. Ultimately, if employers are going to be asked to reimburse for more extensive testing, as well as for the higher-cost targeted therapies, we will need to dem­onstrate what the impact and benefits are. Is the quality of life improved with the targeted drugs vs with traditional chemotherapy, even if there is only an incremental benefit to survival? While we do not have a definitive answer to this question, the point to be made is that targeted therapies are generally better tolerated compared to standard intravenous chemotherapy. The biggest detriment to quality of life is un­controlled cancer, and it is very clear that the targeted thera­pies have high response rates that will control the cancer using mechanisms of action that generally will not directly impair normal cells. As one of our participants stated, the best way to improve a patient’s quality of life is to efficiently shrink the cancer with a favorable side effect profile, because their symp­toms improve and they feel better.

Thus, the ultimate goal is to control the overall total cost of care, both direct and indirect economic costs, and thereby improve value. While there may be an openness by employers to look at technologies, especially as they relate to being able to manage employees and their dependents longer term, they are also looking at the ability of the new technologies to create a better experience for their membership while also creating value. The discussion regarding new technologies is relatively simple when discussing technologies under FDA governance or services that are fairly well established and have been updated. However, it becomes a bit more difficult when discussing a new technology in which the quality of the services and the quality of the experience are less defined and measurable. In short, the value proposition is still being developed.

General Concerns Surrounding Comprehensive Genomic Testing and Gaps in Care

The greatest variability in NSCLC care is purported to occur within community oncology practices, which see the major­ity of patients with cancer. If true, the variability in practice patterns could result from these oncologists needing to treat a full range of solid and hemat logic malignancies, delivering much needed access to care but not being able to keep up with the current standard of practice for lung cancer care from a resource perspective. Academic centers have dedicated oncologists for most types of malignancies, including those who focus specifically in lung cancer and have the ability to stay current with research advances in lung cancer, transform­ing NSCLC into a genomic disease for which there are specific FDA-approved therapies.

The practice pattern variabilities among community on­cologists exist for a number of reasons, including the rapid development of different testing platforms that are technically complex and thus not well understood. For example, the previ­ously noted underappreciation of the value that plasma testing can bring to the clinical utility of testing. The most compre­hensive evidence for blood-based ctDNA analysis for molec­ular genotyping has stemmed from studies in aNSCLC. The NILE (Noninvasive vs Invasive Lung Evaluation) clinical trial (NCT03615443) prospectively analyzed the clinical utility of a plasma-based NGS test for first-line genotyping in patients with metastatic NSCLC as compared with tissue genotyping. In 282 patients analyzed, the investigated NGS assay detected clinically relevant NSCLC-associated biomarkers at a similar rate to standard-of-care testing and was deemed non-inferior. The combination of tissue-based and plasma-based genotyping and cfDNA resulted in a higher frequency of driver mutations identified compared to either method alone.^18,19

Another contributor to the complexity of genomic testing may very well lie with the pharmaceutical industry in general. With each novel targeted therapy that gets approved, a compan­ion diagnostic test (CDx) is also approved. That newly approved companion test is specifically focused on identifying a specific biomarker or mutation. A less expensive and broadly available methodology to achieve this biomarker goal is to use IHC as the pathology approach to an FDA-submitted CDx. Can this be cross-applied to NGS? Yes, but at a price. As an industry, what is needed is a call to action to consolidate testing into a compre­hensive testing strategy, thus making the ordering and report­ing of genomic testing simplified and broadly applicable to the CDx requirements of precision drug therapies. Meanwhile, ob­served variability in reporting, which will be discussed later in this article, needs to be addressed through an educational effort on the standard of care for comprehensive testing.

As previously touched upon, one big gap in reporting is the lack of consistency across naming and nomenclature in reports. The common labeling, or the name that’s used by on­cologists in the clinic and that patients are probably familiar with, may or may not be consistently shown in the report. The report may have only a string of letters and numbers to represent a gene, which can be confusing for a provider, par­ticularly a generalist, trying to understand and interpret the results. As previously noted, the inclusion of interpretation notes may be of assistance in this situation.

Any discussion regarding gaps and concerns with genomic testing, especially with multi-gene panel testing, is anchored upon three basic issues: (1) reporting testing noise, (2) analyti­cal validity, and (3) the perception by payers and employers of the test in terms of value or relative lack thereof. Reporting noise was previously addressed from a different perspective. Current multigene panels today report on more than just the essential biomarkers with approved targeted therapies, often due to their dual use in research or clinical trials matching. The testing noise can be minimized by reporting out the ac­tionable biomarker results on a separate page from the non-actionable biomarkers. This will allow oncologists to quickly identify the necessary information and rapidly move on to the decision process of acting on an appropriate targeted therapy, if this is an option.

Analytic validity—addressing whether the test will con­sistently report an accurate and reliable result—is the next issue. Analytical validity is a quality control metric to deter­mine the precision of the testing methodology, ie, whether the test will yield the same predictable result when run mul­tiple times. Payers must rely on College of American Patholo­gists, CLIA, and New York State certification (a higher bar of actual testing proficiency) to determine analytical validity. However, there are historical concerns around clinical valid­ity. With the diversity of laboratory-developed tests (LDTs) utilizing NGS platforms, there remains a question of consis­tent levels of accuracy being reported for targeted therapies.

Pfeifer et al conducted a pilot study of traceable refer­ence samples to measure NGS LDT performance among a cohort of CLIA-certified clinical laboratories.20 Wet and dry test samples were prepared in a controlled setting to assess the entire NGS test cycle. The pilot focused on KRAS and NRAS sequence variants, important for selection of patients with metastatic colorectal cancer likely to obtain therapeutic benefit with the EGFR inhibitor panitumumab. Twenty-one LDTs tested the wet samples, and 19 of the 21 LDTs tested dry samples. Of the laboratories that tested both the wet and dry samples, 7 of 19 (37%) laboratories correctly reported all variants, 3 of 19 (16%) had fewer than 5 errors, and 9 of 19 (47%) had 5 or more errors. Most errors were false nega­tives. In addition, because the NGS analysis of the reference samples by participating laboratories occurred from Decem­ber 2018 to March 2019, the variability in reporting accuracy would most likely not be due to a difference in the generation of the NGS platform. In short, the pilot showed differences among NGS LDTs with respect to accurately reporting ge­nomic alterations, both with respect to one another and to the published accuracy of an approved CDx. The differences in preanalytical variables, variant identification, and report­ing indicate that the LDTs of participating laboratories for identifying genomic alterations may not be interchangeable with an FDA-approved CDx for the identification of patients as candidates for targeted therapy.²⁰

The gap for employers is that the decisions with regard to coverage for health care services for employees and depen­dents are being made by individuals with little interest in the “science.” Health benefit decisions are made in consultation with benefits brokers and consultants, TPAs, or their health plans. Thus, the question the employers are asking is: “What education is being directed to these benefits advisors?” There is the potential that if employers understood the benefits of early NGS testing, they could take on the role of an active participant to educate payers. However, as noted, there is minimal under­standing by self-funded employers for this scenario to develop. First and foremost, as previously stated, the concern of the employer is the cost of providing health care services. While employers know that the cost of cancer care is expensive, they unfortunately do not have enough of a perspective to prioritize cancer care against other disease cost drivers. What it may take is finding individual(s) who care enough to ask, “What is the quality and the value of new technologies?”

CGP in a Vertically Integrated Healthcare Ecosystem

A clear trend in health care is the vertical integration of the national health insurance carriers as well as their purchasing of intermediaries in their supply and care-delivery chains. This includes, but is not limited to, CVS/Caremark owning Aetna, Cigna owning Express Scripts, and UnitedHealth Group own­ing Optum. These insurers, and others, continue to purchase not only pharmacy benefits management companies and spe­cialty pharmacy companies, they also purchase care-delivery entities. These range from primary care practices (ie, Oakstreet) to Home Health to skilled nursing facilities and can include on­cology practices and other services. In addition, there contin­ues to be a fair amount of horizontal integration across health insurance carriers, such as Anthem purchasing Blue Cross of Louisiana and other similar transactions. The “bulking up” of those entities, which determine a large measure of network participation contracts to providers of care and also determine payment terms and conditions on the delivery systems touched by their members, has not gone unnoticed by the delivery sys­tem and provider community. They are responding by also ag­gregating, both for economies of scale and for defensive con­tracting purposes.

What does this mean for the future of CGP and its appro­priate use in cancer care? It can be seen as both a promise and a peril. The federal government, beyond its regulatory role in ensuring fair competition in the business of the US, is tak­ing notice of this integration for another reason: the relaunch of the Cancer Moonshot, whereby a personalized medicine approach to clinical research in cancer can only be built out through a foundation of complete patient data, integrated across the patient journey. These data reside in both clinical and administrative repositories of care, exactly the places and companies where the aforementioned vertical and horizontal integrations are taking place.

When political energy meets private market dynamics, much progress can be made. However, this progress will likely need significant management and oversight. The promise of biomarker testing in cancer care can only come to fruition if our largest payer for cancer populations, Medicare, understands from a policy perspective the cost equation of testing strategies yielding effective therapy journeys across an insured popula­tion. These data are just becoming available at scale today, at the same time as testing production costs continue to decrease. Another concern in the value-proposition positioning of CGP in cancer care is the slowing of genomics lab innovation un­der contracting pressure of dominant health insurers outside of Medicare. Will advances continue if this becomes a commodity service, as we’ve seen other lab services become? Lastly, the cost of data acquisition is an important factor in creating the value story for CGP, and those costs are dropping as well in a man­dated, “interoperable” health care world. Much of this is to be determined, but clearly there is a role for both the public and private sectors to manage the genomics opportunity.

Next Steps

Based upon the discussions of the NGSi subcommittee on NGS testing in patients with NSCLC, the following questions/ action items were agreed upon:

1. Determining whether all NGS testing is the same

• Will all NGS platforms report to the same level of accuracy?
• If there’s some variability, is that variability clinically significant and warranted?
• Can a set of basic principles be developed to evaluate NGS platforms?

2. Identifying best practices

• Timing of the testing
• Utilizing tissue, liquid, or both samples for testing
• Application of the test results in treatment decision-making
• Understanding the limitations of testing
• Understanding the benefits and harms of the targeted therapy relative to existing treatment alternatives

3. Standardizing and simplifying reporting

• Standardizing the format to present essential information in language appropriate for general oncologists separate from research information

4. Using real-world evidence to create and demonstrate the value proposition

• Utilizing the biomarker results and track it to patient outcomes, quality of life, and overall cost of care
• Creating a clear picture of the true patient journey when biomarker information is utilized and the total cost of care is tracked. This can be compared to a control group not utilizing biomarker information appropriately to determine the cost differences
• This comparison forms the basis for the value proposition that can be used to educate payers, employers, brokers, benefits consultants, and providers

In summary, providers need more support to increase their overall comfort with best practices for integrating biomarker testing into their practice, including when to order testing, what tests to use (liquid or tissue), and how to apply results to patient treatment decisions. This will be made easier in part if the results reporting is standardized and simplified so that results can be more easily reviewed and utilized by oncologists. The uptake of NGS testing will generate real-world evidence, and analysis can be completed to determine its impact on out­comes, quality of life, and the total cost of care, thus creating the foundation for a value proposition. This value proposition can then be used to educate all stakeholders on the merits of utilizing comprehensive NGS biomarker testing upon diagno­sis of patients with NSCLC. Finally, all stakeholders are eager to ensure confidence around the accuracy of results.

This article has supplementary material, which can be accessed here.