Reimbursement Landscape for Molecular Testing in Non-Small Cell Lung Cancer

, , , , ,
Evidence-Based Oncology, February 2018, Volume 24, Issue 2

This review assesses the current molecular testing landscape for non–small-cell lung cancer in the United States.

ABSTRACT

Introduction: The identification of oncogenic genomic alterations and the development of matched targeted therapies have made molecular testing an increasingly important approach to treat non—small cell lung cancer (NSCLC). However, little information is available concerning use of molecular testing in clinical practice and about coverage of these novel tests.

Areas covered: In particular, clinical guidelines and consensus recommendations, currently available molecular tests along with their associated advantages and disadvantages, the use of molecular testing in clinical practice, and current managed care coverage policies.

Commentary: The landscape for molecular testing in NSCLC is evolving rapidly. Although targeted therapy in patients with specific oncogenic genomic alterations is associated with superior outcomes, the use of molecular testing in clinical practice is hindered by several factors, including long turnaround times and tissue sample requirements. Clinical guidelines support the use of broad molecular testing in NSCLC, but most US health plans cover testing only for a limited number of genomic alterations. Nevertheless, as testing technology improves and targeted therapies become more available, molecular tests are expected to eventually become the standard of care in NSCLC treatment.Introduction

LUNG CANCER IS THE leading cause of death from cancer worldwide.1 In the United States, more than 225,500 new lung cancer cases and 155,870 lung cancer—related deaths were expected in 2017, making lung cancer the second most prevalent cancer among both men and women.2 Non—small-cell lung cancer (NSCLC) is the most common form of lung cancer, representing approximately 85% of all cases.3 NSCLC is characterized by a number of genomic alterations (mutations, rearrangements, and amplifications), and these alterations are responsible for initiating and maintaining tumor growth through constitutive activation of oncogenic signaling pathways.4 Oncogenic genomic alterations (hereafter referred to as driver mutations) in lung cancer have been identified in genes encoding EGFR, ALK, ROS1, BRAF, MET, RET, and HER2, among others.5

The development of therapies targeting known driver mutations has permanently altered the treatment landscape of NSCLC.6 The use of targeted therapies in patients harboring specific genomic alterations for which a specific therapy was developed is associated with improved treatment response and survival.6 Indeed, in recent studies, patients with NSCLC who received targeted therapy were shown to have a higher overall response rate—both in first- and second-line settings7—as well as prolonged progression-free7 and overall survival compared with patients not receiving targeted therapy.7,8 Currently, FDA-approved targeted therapies for NSCLC are available targeting EGFR, ALK, ROS1, and BRAF.6,9-11 Other therapies are in development for targets such as MET, HER2, and RET.12-14

Targeted therapies first approved for other tumor types have also demonstrated clinical benefit in alterations existent in NSCLC. For example, the combination of dabrafenib (Tafinlar) and trametinib (Mekinist) was initially approved for BRAF V600E mutations in metastatic melanoma but now is also approved for BRAF V600E mutations in NSCLC.10,11,15 BRAF mutations (primarily V600E) are present in approximately half of all cases of metastatic melanoma16,17 but in just 1% to 4% of patients with NSCLC, with V600E being the most common variant.7,18

Similar to other NSCLC subtypes, patients with BRAF-positive NSCLC receiving targeted therapy experience improved clinical outcomes as demonstrated by the interim results of a phase II trial.19,20 Based on the results of this ongoing trial, the dabrafenib/trametinib combination received breakthrough therapy designation, followed by FDA approval for the treatment of BRAF V600E—positive NSCLC.15

In addition, National Comprehensive Cancer Network (NCCN) guidelines for NSCLC have recently been updated to include BRAF mutation testing in the standard set of biomarkers that should be assessed for patients with NSCLC, with the recommendation to use dabrafenib plus trametinib in first-line therapy for BRAF V600E—mutant metastatic NSCLC.21

The identification of novel driver mutations defining clinically relevant molecular subtypes of NSCLC has made molecular testing and subtyping an increasingly important diagnostic tool. Molecular testing using next-generation sequencing (NGS) technology has emerged as a tissue- and time-efficient testing approach as it allows an entire panel of genotypes to be tested simultaneously, typically requiring a small tissue sample.22 Accordingly, several clinical guidelines now endorse the use of broad molecular testing to identify actionable driver mutations for which targeted agents may be available.21,23

In light of the accumulating evidence for the value of molecular testing in NSCLC, it is important to understand current patterns in molecular testing in lung cancer in US clinical practice, particularly the use of multiplex testing by NGS. Testing in actual clinical practice may be particularly important in NSCLC and for patients with relatively rare genomic alterations, given the increasing number of available targeted therapies. Considering the need to identify the appropriate targeted therapy for the right NSCLC patients, and the risk of running out of tissue in sequential testing modalities, multiplex testing may be even more important in NSCLC because of its many less-common actionable driver mutations/alterations, such as BRAF mutations. Anecdotally, patient access to molecular testing for both established and emerging NSCLC driver mutations also varies, with medical coverage policies that may not reflect current scientific and medical consensus in this rapidly changing area. To address this knowledge gap and better understand the current molecular testing landscape for NSCLC in the United States, this targeted literature review included the following 4 objectives:

1. Describe published clinical guidelines and consensus recommendations related to the use of molecular testing in patients with NSCLC.

2. Describe molecular diagnostic tests currently available in the United States for the detection of the BRAF mutations, their use in clinical practice, and their associated

advantages and disadvantages from the point of view of both patients and physicians.

3. Describe current managed care policies regarding the coverage of molecular testing for NSCLC.

4. Identify policies and barriers regarding the use of molecular testing in clinical practice, and the implications and ramifications they present to the molecular

testing landscape for NSCLC.

Methods

Data Sources

To identify relevant information regarding the 4 study objectives listed above, a targeted literature review was conducted using the following data sources: 1) MEDLINE and EMBASE (via Ovid); 2) published abstracts from the American Society of Clinical Oncology (ASCO); (3) published treatment and diagnostic oncology guidelines from ASCO, the NCCN, and the College of American Pathologists (CAP)/International Association for the Study of Lung Cancer (IASLC)/Association for Molecular Pathology (AMP); 4) published or otherwise publicly

available care pathways including diagnostic testing for NSCLC; 5) medical policies describing coverage for molecular diagnostic tests for NSCLC tumor samples that include BRAF mutations; 6) grey literature, including pharmaceutical, molecular diagnostic and managed care industry websites, white papers, trade press, and newsletters (eg, PinkSheet, GreySheet, GenomeWeb, Oncology Times, Evidence-Based Oncology™, Managed Care magazine), describing relevant molecular tests and their coverage and reimbursement; and 7) ad hoc Internet and PubMed searches.

Search Strategy

Peer-reviewed articles

Peer-reviewed articles identified during the targeted literature review were selected based on their potential relevance to the 4 study objectives. The search focused on articles published between January 1, 2011, and June 6, 2016, and was limited to English articles focusing on the United States. The search strings used to conduct the target literature review in Ovid MEDLINE and EMBASE contained the terms BRAF, B-RAF, or B RAF and 1) carcinoma, non—small-cell lung/ or non–smallcell lung cancer$.mp. or NSCLC.mp. or ((lung neoplasms/ or bronchial neoplasms/ or carcinoma, bronchogenic/) and (adenocarcinoma/ or adenocarcinoma, bronchioalveolar/ or carcinoma, large cell/ or carcinoma, squamous cell/)) or 2) lung or NSCLC or non–small-cell lung cancer.

Clinical guidelines

The 3 main US oncology guidelines for the diagnosis and treatment of NSCLC were selected a priori: ASCO, NCCN, and CAP/IASLC/AMP guidelines. Searches for current guidelines were conducted initially in June 2016 and updated in November 2017.

Molecular diagnostic tests

Molecular diagnostic tests that detect the BRAF V600E mutation and other genomic alterations using NGS were selected based on Internet searching. To be eligible for inclusion, the tests were required to be marketed for diagnostic use in lung cancer, commercially available in the United States, and used or produced by large central/national laboratories, molecular diagnostics specialty companies, or academic laboratories. It should be noted that, because of the nonsystematic nature of the search, the tests that were selected for this study are not necessarily representative and/or inclusive of all the tests that are currently available on the US market.

Payer medical policies

Payer medical policies (ie, coverage polices) were identified by searching the websites of small and large US healthcare plans. Searches for publicly available medical policies were conducted initially in June and July of 2016.

Results and Discussion

A total of 73 articles relevant to the study objectives were identified and selected: 1 was related to objective 1,24 59 to objective 2,25-82 8 to objective 3,25,30,31,83-87 and 11 to objective 4.25,26,31,59,62,63,83,86,88-90 A total 19 currently available molecular tests for the detection of the BRAF V600E mutation were selected and reviewed: 3 were from large central/national laboratories, 14 from molecular diagnostics specialty companies, and 2 from academic laboratories. A total of 16 healthcare plans were selected and their

medical policies regarding the coverage of molecular tests for patients with NSCLC were reviewed. The results obtained by reviewing the above selections are presented below for each of the 4 study objectives.

Clinical guidelines and consensus recommendations for molecular tests in NSCLC

The 2016 draft guidelines from CAP/IASLC/AMP support BRAF testing in NSCLC (Table 1).91,92 More specifically, they recommend molecular testing be performed to identify genomic alterations in BRAF, MET, KRAS, HER2, and RET, either initially or when routine EGFR, ALK, and ROS1 tests are negative.91,92 It should be noted that the CAP/IASLC/AMP guidelines were first published online in 2013. Revised 2016 draft recommendations were anticipated for publication in early 2016, however, as of November 2017 updated guidelines have not been published.

NCCN guidelines continue to support broad molecular profiling (Table 1), and they recommend testing for ALK gene rearrangements and EGFR mutations (category 1 for both) in the NSCLC algorithm for patients with nonsquamous NSCLC or NSCLC not otherwise specified so that patients with these genetic abnormalities can receive effective treatment with targeted agents such as ceritinib, erlotinib, gefitinib, afatinib, and crizotinib. The NCCN guidelines also recommend testing for ROS1 rearrangements (category 2A) as well as for BRAF V600E mutations for patients with metastatic NSCLC. These guidelines also state that other driver mutations and gene rearrangements (ie, driver events) are being

identified, such as RET gene rearrangements, high-level MET amplification or MET exon 14 skipping mutation, and HER2 (also known as ERBB2).Targeted agents are available for patients with NSCLC who have these other genetic alterations, although they are FDA approved for other indications.21

ASCO guidelines date back to 2014, when the ASCO staff reviewed and endorsed the 2013 CAP/IASLC/AMP guidelines (Table 1). At that time, the CAP/IASLC/AMP guidelines only addressed the use of molecular testing for the selection of patients with lung cancer with genomic alterations in EGFR and ALK.93

There are a variety of NSCLC care pathways. Anthem Cancer Care Quality Program Treatment Pathways do not specify which protocols should be used for molecular testing. Overall, care pathways were not publicly available and mostly focused on chemotherapy, targeted therapy, and supportive care regimens.24

Available Molecular Tests for the BRAF V600E Mutation: Practical Advantages and Disadvantages

Available molecular tests

A description of the characteristics of a selection of currently available molecular tests for the detection of the BRAF V600E mutation in patients with NSCLC is presented in Table 2. BRAF tests are available both as single analyte tests and as part of multigene panels.

Besides BRAF, a growing number of lung cancer panels also assess genomic alterations in several other genes, including ALK, EGFR, HER2, KRAS, MET, RET, and ROS1. To identify BRAF mutations, various test technologies are used, with detectable classes of genomic alterations varying with the tests.

As illustrated in Table 2, some tests detect only point mutations while others are more comprehensive and detect genomic alterations such as insertion and deletions (indels), chromosomal rearrangements, and copy number alterations. One implication of this variability across tests is that not all the actionable—and thus treatable–driver mutations in NSCLC are identified by all available tests. Not surprisingly, the cost of comprehensive panel testing appears to be substantially higher than the cost of single nucleotide polymorphism tests.

Sample requirements for available tests vary, as summarized in Table 3. While most tests require formalin-fixed paraffin-embedded tissue, some also accept a blood/liquid biopsy sample, or purified DNA. Several tests have been developed/validated specifically for lung cancer, while other tests may be applied to any solid tumor or are specific to hematologic malignancies.

Test turnaround time from sample collection to availability of results varies widely across tests, ranging from 1 to 4 days to 4 to 6 weeks (Table 4). Test performance also varies and, in addition, is not reported consistently. Similarly, the information available for each test was often incomplete and not reported consistently. For instance, the sensitivity, specificity, and sequencing depth and coverage of the tests were rarely provided.

Very little was found regarding testing patterns in clinical practice. In particular, no information directly related to the clinical practice of BRAF testing was found. The information available was related only to EGFR and did not encompass all types of genomic alterations. Anecdotal evidence suggests a rapid increase in the availability of comprehensive genomic profiling tests and their use by physicians in treatment selection. As a case in point, Foundation Medicine conducted more than 8000 FoundationOne and FoundationHeme NGS tests in the third quarter of 2015, a 25% increase from the previous year. In addition, a global survey conducted by Kantar Health between December 2014 and January 201525 found that overall 81% of newly diagnosed patients with stage IIIb/IV NSCLC received testing for EGFR prior to first-line therapy; this percentage was lower among patients treated by US and European oncologists (77%).94 Practical advantages and disadvantages Comprehensive genomic profiling tests that assay and detect various types of driver mutations have substantial advantages for patients and physicians (Table 2). As long as the information provided by these tests is actionable—and thus has clinical utility—clinical outcomes are likely to improve. Several studies have investigated the utility of molecular testing in patients with NSCLC by comparing different tests and methods.32,43,70,73-76,79,80,95,96 In 1 of these studies, evidence for the utility of targeted NGS assays was obtained by comparing the information obtained from a single gene assay and NGS assays. The study showed that 50-gene panel assays were able to identify at least 1 actionable gene variant in almost twice as many specimens than single gene assays.96

However, the practical implications of the potentially useful clinical information provided by molecular testing remain unclear, as no studies quantifying the benefit of improved test performance for patients and/or payers were identified. However, time to results for all actionable genomic alterations and technical improvements related to diminished sample requirements with a 50-gene panel may have substantial advantages for both patients and physicians (Table 3). Newer tests are becoming more efficient in detecting NSCLC genomic variants and they require gradually smaller amounts of sample tissue. This is particularly important given that the limited amount of tissue typically available from a lung biopsy needs to be used in multiple histological and pathological tests, including resampling, following a diagnosis of NSCLC.

Importantly, tests can now be performed using samples that require less invasive procedures, such as a liquid biopsy. Several articles assessed the use of these less invasive procedures.33-35,37,42,44,45,51,53,67,77-79,97-100

Although liquid biopsy has been validated for EGFR testing,101 its reliability for NSCLC panel tests compared with direct analysis of tumor tissue has not been established.

A fast test turnaround time may be critical to inform clinical decision making (Table 4). Reported turnaround times varied widely. No studies assessing real-world turnaround times or the impact of turnaround time on clinical outcomes were identified. However, it is obvious that assessing 50 genes at once will be faster than analyzing multiple genes in sequence, as is common in many laboratories.

Update

As of December 2017, other approved NGS tests included Oncomine Dx as well as the broader FoundationOne CDx (F1CDx) and Memorial Sloan Kettering Cancer Center’s Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT), which are both approved to detect mutations in more than 300 genes in any solid tumor type, including NSCLC, melanoma, and breast cancer, enabling the identification of patients who may benefit from at least 15 different FDA-approved targeted therapies.

US Managed Care Policies

In terms of coverage, managed care policies regarding molecular testing in NSCLC vary considerably (Table 5). Most of the medical policies identified in this review cover only ALK and EGFR testing. For example, a number of Blue Cross Blue Shield medical policies are similar and usually consider only ALK and EGFR testing as medically necessary. A few plans, such as Health Net, Inc, and CMS, cover testing for genomic alterations in BRAF, ALK, EGFR, HER2, KRAS, MET, RET, and ROS1. NCCN guidelines are typically cited in medical policies as the reason some tests are deemed medically necessary and others are not.

Although most plans align medical policies with anatomical tumor location, some payers such as Health Net, Inc, UnitedHealthcare, Aetna, and several CMS Local Coverage/Medicare contractors have issued medical policies covering comprehensive genomic profiling of tumors using NGS. Standard guidelines for coverage of molecular diagnostic tests have been proposed by the Center for Medical Technology Policy, but they have not generally been put into practice. Similarly, in 2011, the Molecular Diagnostic Services (MolDX) Program was created to establish clear expectations for coverage and reimbursement of molecular diagnostic tests.102 Based on this review of managed care policies for diagnostic tests, adherence to MolDx recommendations on the part of diagnostic developers and application of these guidelines by payers is not apparent.

Current Molecular Testing

Landscape in NSCLC The technology to detect genomic alterations continues to improve, and several studies have shown that NGS-based assays are capable of more

precisely detecting a wider range of alterations than are standard non-NGS tests.103-105 These findings underscore not only the greater efficiency of NGS testing in the detection of genomic alterations, but also the importance in identifying the right patients who could benefit from targeted therapies.103,104

Nevertheless, there is currently a wide variation in clinical practice for molecular testing in NSCLC. A variety of tests with different characteristics, sample requirements, and reimbursement levels are available on the market, making it challenging for physicians to select the most appropriate test for their patients. In addition, the complexity of genomic information provided by the tests creates substantial challenges in interpreting the results of a test. The growing number of identified genetic variants and the increasing technical complexity of molecular tests are likely to exacerbate this problem.

While great strides have been made in advancing molecular diagnostics, several hurdles still need to be overcome to make molecular testing a routine tool for diagnostic workup of patients with NSCLC. In a global survey conducted,25,106 the main reasons by oncologists for not testing for genomic alterations in EGFR included insufficient tumor tissue and long turnaround time. More specifically, because of the long turnaround time, 26% of US physicians made their treatment decisions before test results were made available.94 Sequential single gene testing can leave an insufficient amount of tissue to analyze additional genomic alterations, an issue seen by oncologists as an important barrier to testing.106

To address the issue of limited tissue availability, CAP/IASLC/AMP guidelines recommend liquid biopsy/circulating tumor DNA (ctDNA) assay be used for EGFR testing when tissue is insufficient for molecular testing. These guidelines also state, “Pathologists and laboratories should utilize tissue-sparing techniques to preserve tumor tissue for diagnosis and to enable subsequent lung cancer biomarker testing.”107 In addition, NCCN guidelines recommend broad molecular testing. Regarding ctDNA, the NCCN guidelines state, “Recent data suggest that plasma genotyping (also known as liquid biopsy or plasma biopsy) may be considered instead of tissue biopsy to detect whether patients have T790M; however, if the plasma biopsy is negative, then tissue biopsy is recommended, if feasible.”21 This tissue-sparing approach is being used by physicians across different tumor subtypes, substantially increasing the number of tests that can be conducted for each patient.

Although some form of molecular testing is covered by most health plans, uncertain reimbursement may limit its use in clinical practice. According to test manufacturers, payment for covered molecular diagnostic tests is inconsistent and does not reflect the value of the information provided. For instance, some Clinical Laboratory Improvement Amendments laboratories describe low payment levels and limited coverage for molecular diagnostic tests, especially panel tests.61

In addition, private health plans may not reimburse tests that are not priced by Medicare and, in some instances, match Medicare prices that are below the actual costs of performing the test. For broad molecular profiling tests that include hundreds of genes, individual contracts between health plans and test manufacturers may overcome these limitations. As a case in point, UnitedHealthcare and Foundation Medicine recently reached an agreement according to which UnitedHealthcare will cover the FoundationOne test for patients with metastatic stage IV NSCLC.108

Consortia of test manufacturers may also help establish the value of comprehensive genomic profiling. In 2015, Thermo-Fisher, Illumina, Eli Lilly, Celgene, and Roche/Genentech committed to provide their competence and funds to Molecular Evidence Development Consortium—a nonprofit organization that aims to establish standards for molecular tests and build the clinical utility evidence around targeted treatment strategies.106 Moreover, statutory changes to Medicare may contribute to supporting value-based pricing for diagnostic tests. To this end, the implementation of a law that seeks to establish a market-based payment system for molecular tests—the Protecting Access to Medicare Act of 2014—is ongoing.

Medical policies establishing coverage for molecular testing face a tremendous challenge in keeping pace with technological advancement in molecular diagnostics. According to NextGen DX’s market analysis, more than 60,000 unique molecular diagnostic testing products are presently on the market and 8 to 10 new tests are estimated to be launched daily.109 The large—and increasing—number of available tests makes setting health plan coverage policies especially challenging, particularly given that the evidence supporting the clinical value of all genes in a comprehensive genomic profile is limited. Additionally, the medical policies of most health plans are updated far less frequently than the NCCN guidelines, and thus they are unlikely to keep up to date with the latest guideline recommendations. This is possibly the reason why some tests are not classified as medically necessary despite being listed in the most recent NCCN guidelines.

Establishing consistent coverage may depend on clear demonstrations of the value of molecular testing. One recent study compared the values of multiplex and sequential testing and concluded that sequential testing “is very inefficient especially with respect to the time it takes to complete testing,…[to] the total cost, and…to the amount of tissue necessary to complete testing.”32 This targeted literature review did not identify any studies quantifying the benefits of multiplex testing or evaluating the clinical and economic outcomes associated with BRAF testing in NSCLC, either alone or in the context of a multiplex/panel test. More studies measuring the economic value of molecular testing and comparing different types of tests are needed for each genomic alteration. With such a vast number of existing testing options, information on the comparative effectiveness and cost-effectiveness of available tests will be crucial to help physicians select the most clinically appropriate test and assist health plans in making more informed coverage and reimbursement decisions.

CONCLUSIONS

The landscape for molecular testing in NSCLC is evolving rapidly, mostly due to significant technological advances that capture actionable information about disease subtypes with increasingly accurate results. Although treating patients with NSCLC who have driver mutations with appropriate targeted therapy is associated with superior outcomes, the use of molecular testing in clinical practice appears to be limited. The use of molecular testing may be hindered by several factors, including long turnaround times to generate test results and limitations on the availability of tumor tissue. In addition, although several clinical guidelines support the use of broad molecular testing in patients with NSCLC, most health plans only cover tests to identify genomic alterations in ALK and EGFR. Based on the information identified in our search of medical policies, only a few health plans extend their coverage to other genomic alterations in targets such as BRAF, HER2, KRAS, MET, RET, and ROS1.

For References, see eAppendix at ajmc.com/journals/evidence-basedoncology/2018/february-2018 ABOUT THE AUTHORS:

Dave Nellesen, PhD, is vice president of the Analysis Group, Inc, in Menlo Park, California. Katherine Dea, MSc, and Annie Guerin, MSc, are employed by the Analysis Group in Montreal, Canada. Kenneth W. Culver, MD, is an executive director at Novartis Oncology; Alex Mutebi, PhD, is associate director of Health Economic and Outcomes Research at Novartis; and Anand Dalal PhD, MBA, BSPharm, is employed in Health Economics and Outcomes Research at Novartis; all are employed in East Hanover, New Jersey.

CORRESPONDING AUTHOR:

Dave Nellesen, PhD

1010 El Camino Real, Ste 310, Menlo Park, CA 94025.

E-MAIL: Dave.Nellesen@analysisgroup.com

FUNDING SOURCE:

This study was funded by Novartis Pharmaceuticals Corporation. Dr Culver, Dr Mutebi, and Dr Dalal are employees of Novartis Pharmaceuticals Corporation and may own stock or stock options. Dr Nellesen, Ms Dea, and Ms Guerin are employees of Analysis Group, Inc. which has received consultancy fees from Novartis Pharmaceuticals Corporation. Writing assistance was provided by Cody Patton and Cinzia Metallo, PhD, employees of Analysis Group, and was funded by Novartis Pharmaceuticals Corporation.