Response to Comments: MolDX: Plasma-Based Genomic Profiling in Solid Tumors

GO2 Foundation for Lung Cancer (formerly Lung Cancer Alliance and the Bonnie J. Addario Lung Cancer Foundation) is writing in support of the new proposed MoIDX: Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors local coverage decision (DL38043) which expands coverage for patients with advanced lung cancer and provides coverage for additional solid tumor cancers. We previously submitted comments in July 2017 in response to the LCD MolDX: Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Non-Small Cell Lung Cancer (NSCLC) (DL37338) and applaud Palmetto GBA’s efforts in monitoring evolving standards and testing and establishing medically appropriate coverage policies for the lung cancer community.

GO2 Foundation for Lung Cancer is the leading national patient support and advocacy organization dedicated exclusively to saving lives and advancing research by empowering the millions living with and at risk for lung cancer. It is estimated that more than 228,150 cases of lung cancer will be diagnosed this year in the United States representing 26% of cancer deaths (142,670). Lung cancer’s 5-year survival rate continues to stagnate at 18% today - levels documented over four decades ago.

GO2 Foundation for Lung Cancer is dedicated to reversing these trends and increasing survivorship. Since 2004, we have been advancing a more coordinated, comprehensive, and compassionate federal health policy strategy that elevates the unique needs of our entire community. We are strong proponents of health policy strategies that focus on patient-centered reforms that improve, not restrict, access to high-quality affordable care and do not hinder the doctor-patient relationship. This more targeted and individualized approach to restructuring our nation’s health system care will lead to improvements in quality of life and survivorship to those who matter most – the people it serves.

With this understanding, GO2 Foundation for Lung Cancer believes that coverage of plasma-based comprehensive genomic profiling (“liquid biopsy”) is critical for the advanced NSCLC expanded patient population and additional solid tumor cancers. We are writing to request that NSCLC patients be able to access plasma-based comprehensive genomic profiling if the physician thinks it is the most appropriate first-line test modality.

Recent publications have demonstrated that this test performed similarly to tissue testing in NSCLC patients and more NCCN mutations were found compared to first-line tissue-based tests. This data set is key, as it is challenging to obtain tissue from lung cancer patients and may result in delayed treatments in order to complete testing.

Multiple studies have shown that there is a serious unmet medical need with NSCLC patients that results in non- and under-genotyping. Our own data confirms this work. We previously commented in 2017, that less than 50% of patients with NSCLC who called our HelpLine at that time reported having received molecular testing, despite clear differences in potential first-line treatment options and the efficacy of those choices depending on testing results. Some of these patients were not offered testing and others did not have sufficient tissue from their biopsy to perform tissue-based molecular profiling.

We strongly believe in the era of precision/personalized medicine and that patients should have the option of comprehensive molecular profiling so that they, along with their caregivers and physicians can make the most informed treatment decision for their personal situation. Patients with similar stage cancers can often have dramatically different outcomes and patients need the additional information that molecular profiling can provide to inform their treatment decisions.

In many cases, tissue profiling is not available or appropriate and liquid biopsy is necessary for this molecular profiling. In other cases, there will be an unacceptable treatment delay waiting for possible tissue results or the amount of tissue sample warrants using that tissue for immunohistochemistry-based testing on protein biomarkers such as PD-L1 and using plasma for genomic studies. For these reasons, GO2 Foundation for Lung Cancer supports the coverage of this important technology and believes it is most appropriate for the provider to have the option to choose plasma-based comprehensive genomic profiling as a first line test since the new data demonstrates that this test discovers an equivalent number of treatable mutations as tissue testing.

Plasma-based comprehensive genomic profiling is often faster than tissue-based testing, allowing an expedited answer for a very sick patient population where time is valuable. However currently, the draft LCD requires that either tissue is infeasible or there are no actionable results from a CGP tissue test in order to cover NSCLC patients for plasma-based comprehensive genomic profiling. This workflow requires that a physician send out tissue if they are unsure if the tissue will be infeasible for CGP; if the tissue is subsequently insufficient this might delay genotyping for 3-4 weeks. During this wait, the physician may feel pressured to initiate some type of therapy. There is emerging data that putting patients on front-line treatments before realizing they have a targetable mutation and switching to a targeted therapy can increase adverse events for the patients.

As we know, invasive tissue biopsies can result in adverse events for many patients which causes a significantly increased cost to the system. Many patients in our community speak to the burden of complications from regular biopsies and the need for plasma-based testing. One such story has been profiled on our blog, where the patient wrote, “I would much prefer to have blood drawn, than a collapsed lung and eight days in the hospital,” referring to her previous complications with tissue-biopsy and subsequent liquid biopsy profiling at the time of recurrence. Validated liquid biopsy technologies should be an available option for doctors and patients to avoid the potential complications of re-biopsy and to remove a de facto barrier that hinders access to preferred and faster treatment pathways better tailored to a patient’s special circumstance.

All patients deserve access to technology that can inform their treatment choices. Coverage of comprehensive genomic profiling from liquid biopsies can help close this gap and improve their care.

Thank you for considering our comments and continuing to advance access to medically necessary services for patients with advanced lung cancers. We believe that coverage of comprehensive genomic profiling from liquid biopsies is a patient-centered approach that will help serve the interests and needs of the lung cancer community.

Thank you for the comment in support of the coverage decision.

Fight Colorectal Cancer represents thousands of colorectal cancer patients, survivors, caregivers, family and friends across the country. Access to timely, comprehensive biomarker testing is critical for our community to ensure that each individual patient receives the care that is best for them. Recent scientific advances have helped us to learn more about both the causes and risks for colorectal cancer, and in turn have helped patients and their care teams identify more targeted and effective treatment.

Over the past decade, we have seen a significant increase in the number of patients under the age of 50 who are diagnosed with colorectal cancer. For a variety of reasons that are still being researched, these early-age onset (EAO) patients are often diagnosed at an advanced stage, when the disease has metastasized and treatment becomes more difficult.

Whether diagnosed at a young age or later in life, late-stage colorectal cancer patients need access to appropriate diagnostic testing, like biomarker testing, to help them and their care team make the best treatment decisions. We would encourage MolDX to include a broad set of treatment codes for advanced stage colorectal cancer so that there are no barriers to patients receiving this critical testing.

In addition, eligibility for clinical trials is dependent, at least in part, on a patient’s biomarkers. Should a patient want to enroll in a clinical trial, they must receive biomarker testing prior to ensure that the trial they are on is appropriate for their tumor type.

For example, it’s well established that PD-1 inhibitor drugs are only effective for a small population of CRC patients with tumors that are microsatellite instable (MSI-H), information that can easily be attained by conducting biomarker testing. For a patient to fully understand and consider all of their treatment options, including clinical trials, it is critical that they have access to this information as part of their care planning.

Thank you for the comment in support of the coverage decision.

BloodPAC is a public-private consortium that develops standards and best practices, organizes and coordinates research studies through its members, and operates a data commons to support the liquid biopsy research community. Data from retrospective studies run by members, as well as studies BloodPAC organizes, are aggregated and contributed to the BloodPAC Data Commons (BPDC) to establish an open, publicly accessible data commons for the global liquid biopsy community.

Our mandate at BloodPAC is to accelerate the development of liquid biopsy assays to improve the outcomes of patients with cancer. We do this via an unprecedented collaborative consortium infrastructure of over 30 members comprised of industry, academia, and regulatory agencies.

We know that advanced diagnostic tests, and blood-based ones in particular, are critical to guiding physicians in making the most informed treatment decisions for patients suffering from cancer.

We are pleased that MolDX has proposed to expand access to comprehensive biomarker testing that may help lead to life-extending targeted therapies to patient populations that previously may not have had similar treatment options when limited by tissue-based comprehensive genomic profiling (CGP). Additionally, blood-based CGP has the potential to reduce the need for repeat biopsies in a patient population contending with taxing surgical procedures and multiple rounds of therapy.

At the same time, these biomarkers are only useful in patient management if the performance characteristics of the particular test(s) have been established by each laboratory seeking coverage. Such performance cannot be assumed or inferred from CLIA certification and state licensure. Therefore, we support the requirement for rigorous analytical validation and submission of an evidence dossier establishing such validation for individual tests to be eligible for coverage under DL38043.

As part of the Technical Assessment (TA) requirements, we ask that MolDX publicly define the criteria required for coverage to ensure Medicare beneficiaries with advanced cancer have access to the highest quality tests to inform their care.

Further, for complex alternations such as rearrangements or fusions which are more challenging to detect with NGS and more susceptible to stochastic variation we recommend a systematic demonstration of each assay’s genomic coverage technique as an accompaniment to analytical performance characterization. This would clarify and streamline the coverage pathway for labs with validated, blood-based NGS assays and their goal of providing beneficiary access to life-extending targeted therapies.

As we look to additional blood-based NGS assays demonstrating analytic and clinical validity in the future, we recommend that MolDX work with stakeholders to develop a broad, class-based LCD guidance, similar to those models described in draft LCDs (DL38045) and (DL38047).

In summary, BloodPAC supports proposed MolDX coverage framework set forth in DL38043 and its associated forms, provided a test will be covered only to the extent that its developer established that test’s analytical validity for the range of cancer indications for which coverage is sought.

Thank you for the comment. We have been seeking to use consensus to determine a threshold for coverage, but at this time to our knowledge, such standards remain in development. When such standards become available, we will evaluate them for adoption.

We have been advised that Guardant360^® has recently added the ability to detect MSI-H in all advanced solid tumor cancers. This may have important clinical utility for colorectal cancer patients, not only for assessment of familial risk, but also because MSI-H was recently FDA-approved for pan­-cancer treatment with pembrolizumab, providing further treatment options for patients.

Access to testing is critical for patients and physicians to make the optimal therapeutic choice and there are times when tissue is not available for further testing. Although colorectal tumor tissue is readily available via colonoscopy in early stage cancer, at the time of recurrence or metastasis a new biopsy would be required absent the availability of Guardant360^®.

It is our understanding that an additional review of the ICD codes is required for broader access to important diagnostic testing for colon and rectal cancer patients. As the oldest and largest patient advocacy organization solely dedicated to colorectal cancer patients and their families, we want to ensure patients have access to the most comprehensive, but also least invasive tumor testing. As such, we would encourage additional review of the ICD codes to ensure that the broadest ICD code set that physicians use to diagnose advanced cancer is covered to ensure appropriate access to care, for both for colon and rectal patients.

Thank you. We have reviewed the ICD-10 codes again, and we would like to note that ICD-10 codes contained in the coverage article may continue to be reviewed even following LCD finalization.

As a medical oncologist, I am writing in support of the pan-cancer (excepting primary brain) advanced solid tumor draft local coverage decision (LCD) proposed by MolDX entitled Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors (DL38043). Multiple studies have demonstrated that tumor genotyping can help identify actionable mutations, and thus potentially help in therapeutic selection, particularly for patients with metastatic disease.

Patients with metastatic breast cancer often have bone metastasis which makes it difficult to biopsy and there is increasing interest in plasma-based genotyping. In addition, the tumor can evolve over time making tissue genotyping of primary breast cancer less accurate than real-time blood-based genotyping in patients with metastatic disease, particularly for discovery of key acquired mutations such as ESR1 and HER2, including work by our own institution (Yu et al 2015, Medford et al 2018).

We have utilized Guardant360^® for clinical decision making at our institution and recently presented our experience at the American Society of Clinical Oncology (Vidula et al 2018). The retrospective study demonstrated that plasma-based genotyping (Guardant360^®) identified higher number of actionable mutations than tissue genotyping. Furthermore, patients undergoing matched therapy, based on plasma-based genotyping, had better overall survival compared to controls.

Thank you for the comment. We are pleased to hear that a practicing oncologist believes that our coverage decision may help improve outcomes in Medicare beneficiaries.

As an academic cancer center-based medical oncologist specializing in advanced colorectal cancer, I am writing this letter in support of the draft local coverage determination (LCD) proposed by MolDX entitled Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors (DL38043).

Molecular profiling of advanced colorectal cancer is critical for treatment selection. Identification of specific molecular alterations, such as mutations in KRAS, NRAS or BRAF, can predict lack of response to certain therapies, such as anti-EGFR antibodies. Furthermore, detection of specific molecular alterations, such as BRAF V600E mutations or microsatellite instability can identify patients who are likely to benefit from specific anti-cancer therapies.

Our research (Strickler et al, Cancer Discovery 2018) and the research of others has demonstrated that molecular profiling of plasma cell-free DNA, including Guardant360^® testing, can identify targetable molecular alterations with similar effectiveness as tumor tissue testing. Particularly when tumor tissue is not obtainable or of insufficient quantity, plasma cell-free DNA profiling can be critical in guiding treatment decisions. Thus, I offer my strong support for the proposed Local Coverage Determination: MolDX: Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors (DL38043).

Thank you for the comment. We are pleased to hear that a practicing academic oncologist believes that our coverage decision may help improve outcomes in Medicare beneficiaries.

On behalf of Foundation Medicine, Inc., I am pleased to submit comments regarding the above-captioned proposed local coverage determination (the “Draft LCD”). Foundation Medicine requests that the Proposed Local Coverage Determination (LCD) MolDX: GUARDANT360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors (DL38043) be generalized and written in a fashion similar to the proposed LCDs: Next-Generation Sequencing for Solid Tumors (DL38045) and MolDX: Next-Generation Sequencing Lab-Developed Tests for Myeloid Malignancies and Suspected Myeloid Malignancies (DL38047). This would provide consistency amongst the LCDs and allow other liquid biopsy tests to be evaluated and covered via the Palmetto GBA MolDX (‘MolDX’) Technical Assessment (TA) process.

Foundation Medicine appreciates recognition by the MolDX program that not all next-generation sequencing (NGS) based assays are equal. The continued differentiation between targeted NGS panels and comprehensive genomic profiling is important to ensure physicians and Medicare beneficiaries are aware of these differences. As part of the TA requirements, we ask that MolDX publicly define the thresholds that are required for coverage to ensure Medicare beneficiaries with advanced cancer have access to the highest quality tests to inform their care. Additionally, we believe it is important to understand and evaluate the bioinformatics processes the laboratory is using to interpret the sequencing data and generate the reports. Furthermore, we request that, if not already the case, the TA threshold be set at a level at least consistent with New York State Department of Health review standards as outlined by the U.S. Food and Drug Administration.

Thank you for the comment. We will attempt to maintain high standards for the coverage of genetic tests. We have added text to this LCD to allow for coverage of other liquid biopsies.

The proposed LCD is a major step toward increasing access to medically necessary genotyping for patients with advanced solid cancers. Currently, there are at least 16 cancer types for which national guidelines recommend that biomarkers guide treatment either with approved therapies or through clinical trials (www.nccn.org). The FDA approval in 2018 of two tumor type-agnostic, biomarker-driven therapies for patients with advanced cancer (larotrectinib for NTRK fusions and pembrolizumab for MSI-H) further emphasizes the need for all patients with a solid cancer to have appropriate genotyping to determine the optimal therapy to treat their disease.

We would like to clarify the evidence in one of the cited papers and request two minor adjustments to the proposed coverage policy in order to align better with the current published literature and the intended use of Guardant360^®. These items are (1) modification of the coverage language for newly-diagnosed non-small cell lung cancer (NSCLC) patients, and (2) inclusion of appropriate ICD codes to match the reasonable and medically necessary use cases for Guardant360^®.

1. Modification of the coverage language for newly-diagnosed NSCLC patients

The proposed LCD expands upon the previous LCD, L37699, by covering Guardant360^® testing in patients for whom comprehensive genomic profiling (CGP) through tissue testing did not find an actionable biomarker. This coverage is supported by clinical data that was published both before and after the release of the draft policy.

Since the publication of the proposed LCD, a prospective observational study has been published in which Guardant360^® genotyping was compared to standard-of-care tissue genotyping in 282 newly-diagnosed advanced NSCLC patients. This study demonstrated that Guardant360^® was as good as standard-of-care tissue genotyping in identifying patients eligible for biomarker-targeted therapy (27.3% vs. 21.3%, p < 0.0001 for non-inferiority). For biomarkers with associated FDA-approved targeted therapies (EGFR, ALK, ROS1, and BRAF), the concordance when both results were available was 98.2%, with 100% positive predictive value. Importantly, adding Guardant360^® to tissue-based genotyping algorithms in this study significantly increased the number of patients eligible for targeted therapy by 48% (60 to 89 patients), which reflects the use requested in our proposed revision.

Additionally, the study demonstrated that the turnaround time for Guardant360^® was 9 days compared to 15 days with tissue (p < 0.0001), an advantage that increased over the course of the study to a current turnaround time of around 7 days. This improvement is important, as turnaround time is critical to optimal treatment decisions since patients are often placed on non-targeted therapy when tissue-based genotyping results are delayed for several weeks.

Importantly the same study demonstrated that patients treated with biomarker-targeted therapy experienced an increase in overall survival compared to patients with the same biomarkers treated with non-targeted therapy [median survival of 18.6 months (n = 575) versus 11.4 months (n = 560), p > 0.001]. This finding is consistent with two other large NSCLC studies showing more than a doubling of overall survival when treatment is genomically targeted compared to empirical treatment with chemo- or immunotherapy.

Given these benefits, we propose altering the final bullet of the coverage policy that currently states:

– Tissue-based CGP is infeasible (i.e., quantity not sufficient for tissue-based CGP or invasive biopsy is medically contraindicated) or specifically in NSCLC; Tissue-based CGP has shown no actionable mutations.

To the following:

– Tissue-based CGP is infeasible (i.e., quantity not sufficient for tissue-based CGP or invasive biopsy is medically contraindicated) or specifically in NSCLC; Tissue-based CGP has shown no actionable mutations or has not been performed.

This change would enable physicians to make the decision on the best diagnostic method to employ for newly-diagnosed NSCLC patients while still reserving the ability to perform tissue CGP if cell-free DNA is negative.

2. Inclusion of appropriate ICD codes to match the reasonable and medically necessary use cases for Guardant360^®.

The ICD-10 codes listed in the current proposed LCD accurately cover the majority of reasonable and medically necessary Guardant360^® use cases. However, upon review of the Guardant360^® ordering patterns from treating physicians, we believe some codes may have been overlooked.

We used CMS’s coding transmittal for the recent NGS NCD (CAG-00450N) as a reasonable starting point for an ICD-10 crosswalk and discovered that 128 non-CNS ICD-10 codes were missing from the draft LCD. A few examples of such codes used by ordering physicians in caring for their patients include:

• C50.919: Malignant neoplasm of unspecified site of unspecified female breast
• C18.9: Malignant neoplasm of colon, unspecified
• C25.9: Malignant neoplasm of pancreas, unspecified

These ICD codes are some of the most common ones used by providers in ordering Guardant360^® for their respective cancer types, which is expected since Guardant360^® is intended to aid in therapy selection for metastatic patients. In such patients, the cancer has often spread to such a degree that the specific location of the primary tumor either cannot be determined or is deemed irrelevant by the treating physician. As a result, a broader ICD code is often employed. These patients are exactly those who require genotyping the most and should not be excluded from coverage.

In addition to the ICD-10 codes included in the NCD, 52 “secondary malignancy” codes are used to describe metastatic disease in the absence of a primary tumor. This use case is most often encountered when a primary tumor has been previously resected, and the patient is now presenting with metastatic recurrence. We feel that it is appropriate to include these codes as, again, these patients can benefit greatly from genotyping information for therapy selection.

Finally, 14 additional ICD codes were omitted from the LCD draft that describe either carcinoid tumors (also known as neuroendocrine tumors) or Kaposi’s sarcoma, both of which are solid tumors outside of the CNS that meet the disease criteria for treatment with larotrectinib or pembrolizumab. While Guardant360^® is uncommonly ordered in these indications, we feel that these diseases should be included in the comprehensive set of ICD-10 codes included for coverage.

Attached to this comment letter is a spreadsheet outlining the additional list of non-CNS ICD 10 codes that we recommend be included in the LCD.

3. Clarification of the Aggarwal et al. citation in the Summary of Evidence

The Aggarwal et al. citation here2 (reference #35 in the LCD) included a larger subset of patients for whom a mix of Guardant360^® and tissue testing were studied. However, due to various circumstances, tissue results were not available for a large portion of the studied population. We suggest inserting the following language outlined below in the Summary of Evidence:

A more recent study examined the clinical implications of using plasma-based testing in addition to tissue-based testing in 229 patients with NSCLC.35. Of the 229 patients in whom both tissue and plasma testing were ordered, the addition of plasma increased the percentage of patients eligible for targeted therapies from 21% (47/229) to 36% (82/229). For the 128 patients with successful tissue testing results (56% testing success rate), …

Thank you for the comment. We modified the LCD text as is suggested in subcomment 1. We added suggested codes in subcomment 2 to the coverage article. We also modified the review of the evidence to address the facts raised in subcomment 3.

On behalf of LUNGevity Foundation, the nation’s preeminent lung cancer nonprofit that funds research, provides education and support, and builds communities for the approximately 230,000 Americans diagnosed with lung cancer each year and the 538,243 Americans living with the disease, we appreciate the opportunity to submit our comments in response to the proposed Guardant360^® Local Coverage Determination, which was first posted on March 28, 2019.

As a leading patient advocacy group that represents the voice and interest of the national lung cancer survivor community by accelerating research to patients that is meaningful to them, empowering patients to be active participants in their care and care decisions, and helping remove barriers to access to high quality care, LUNGevity applauds Palmetto for providing a draft local coverage determination for the Guardant360^® test that, if finalized as proposed, would ensure new testing options are available for lung cancer patients. In this era of unprecedented scientific advancements for the treatment of lung cancer, particularly in the field of biomarker testing, liquid biopsy tests, such as those performed on a multi-analyte, next-generation sequencing (NGS) platform like the Guardant360^® test, are a new development that identify biomarkers predictive of response to particular treatments for patients in a non-invasive manner that can return test results more quickly than other technologies.

Non-small cell lung cancer (NSCLC) is the most common type of lung cancer, diagnosed in about 85 percent of people with lung cancer. The complex nature of this disease requires personalized management plans for patients. Since the discovery of the first epidermal growth factor receptor (EGFR) mutation in lung cancer in 2004, targeted therapies have become a major component of the treatment arsenal of NSCLC patients. Now at least 10 driver mutations in adenocarcinoma have been identified (EGFR, ALK, ROS, RET, ERB2/HER2 mutations, ERB2/HER2 amplifications, MET amplifications, MET mutations, TRK, BRAF, KRAS). In concert with the identification of an increasing number of targetable mutations is the development of novel, potent, and specifically targeted therapies. For example, at present, third generation EGFR11 tyrosine kinase inhibitors (TKIs) and anaplastic lymphoma kinase (ALK) TKIs12 are used in clinical practice. With the increased use of targeted agents has come the problem of acquired resistance, where cancer cells inevitably develop resistance to the targeted agent. The EGFR T790M is an excellent example of a resistance mutation that develops in patients treated with first- and second- generation EGFR TKIs. This mutation can be rapidly detected using a liquid biopsy test such as the Guardant360^® and the cobas EGFR Mutation Test v2. Lung cancer is now leading the field of precision medicine where research is rapidly progressing to (1) develop better targeted therapies that combat mechanisms of resistance, and (2) noninvasive sampling processes – such as liquid biopsies – that can monitor status of the resistance mutations (e.g., cobas EGFR Mutation Test v2), sequentially and in real time.

Current role of liquid biopsies in the clinical management of lung cancer:

The utility of liquid biopsies in the clinical management of lung cancer is unquestionable for several reasons. First, sample from a tissue biopsy may be inadequate to complete comprehensive biomarker testing. The Lung Cancer Mutation Consortium reports that only 48% of the starting 1542 subjects enrolled had comprehensive biomarker testing due to the issue of inadequate biopsy tissue. Second, tissue acquisition for a lung biopsy comes with significant complications. In a claims analysis conducted by Karve et al, 63% of patients complained of at least one complication (difficulty breathing, chest pain, or pneumothorax) following a lung biopsy. We applaud Palmetto’s decision to propose coverage for liquid biopsy-based testing for such patients. In her ASCO 2017 presentation on biomarker testing for lung cancer, LUNGevity Scientific Advisory Board (SAB) member, Dr. Alice Shaw from Massachusetts General Hospital, pointed out that liquid biopsies may help in (1) initial detection of targetable mutations in advanced-stage NSCLC at the time of diagnosis, (2) identification of acquired resistance mutations in patients who have relapsed on targeted therapies, and (3) monitoring response to targeted therapies and predicting outcome in advanced-stage NSCLC patient. As the science continues to evolve, the role of liquid biopsies in treatment decision-making for immunotherapies is also becoming evident. In a recent report, a positive correlation between co-occurring mutations in the KRAS and LKB1/STK11 genes and lack of response to immune checkpoint blockade in advanced-stage NSCLC was demonstrated. Platforms such as Guardant360^®’s liquid biopsy are able to detect such “exclusionary” changes. Furthermore, the utility of liquid biopsies to identify predictive biomarkers for immunotherapy, such as tumor mutational burden (TMB), is becoming increasingly established. Therefore, the application of liquid biopsy-based NGS testing will become increasingly important, especially in the lung cancer space, where biomarker testing will have to keep pace with the development of newer targeted therapies and immunotherapies.

A true comprehensive approach to biomarker testing:

Since the development of the first commercially available liquid biopsy approach (cobas EGFR Mutation Test v2), the science has considerably evolved. Current liquid biopsy technologies based on a NGS platform can detect multiple actionable mutations at the same time. In a recent study by Aggarwal and colleagues, in a sample of 323 NSCLC patients, 94 patients received a liquid biopsy test at the time of diagnosis (at the discretion of the treating physician). In this subset of patients, a targetable mutation was detected in 33%, suggesting that a comprehensive biopsy platform may be sufficient for treatment commencement. Among the remaining 229 patients who had concurrent plasma and tissue NGS or were unable to have tissue NGS, a therapeutically targetable mutation was detected in tissue alone for 21% of patients, whereas the addition of plasma testing increased this number to 36%. The detection of targetable mutations using only a liquid biopsy platform did not impact the clinical course of the disease. Patients in plasma only and plasma + tissue groups had similar outcomes to targeted therapies. These findings have been replicated in the recent Noninvasive vs. Invasive Lung Evaluation (NILE) study where the addition of a cell free DNA (cfDNA)-based NGS assay to tissue-based NGS increased the detection of an actionable mutation by 48% within a shorter turnaround time, including those with negative, not assessed, or insufficient tissue results.

However, it is important to keep in mind that all NSCLCs do not shed enough DNA that would enable detection by cfDNA-based technologies. Keeping this caveat in mind, we suggest that the optimal approach to receiving the most comprehensive picture of a patient’s tumor at the time of diagnosis (before the start of first line treatment) may be a dual NGS panel, using both a tissue- and a liquid-based approach. This dual approach would mitigate the risk of missing actionable mutations and also for the treating physician to commence treatment with the “first available” results.

Expanding the one test per primary cancer limit for NGS:

Access to high-quality, timely NGS testing (at diagnosis and at recurrence or progression) is instrumental for matching patients to the appropriate targeted therapy and advancing precision medicine.

LUNGevity recognizes that Palmetto is bound by the limitations in the NCD to only cover one test per primary cancer, and we acknowledge that the language in the draft LCD indicates that Palmetto seems to recognize the importance of testing at different time points during the course of a tumor. We appreciate Palmetto’s interpretation of the limitation under the NCD in a way that can benefit more Medicare beneficiaries, and we are committed to continue to work with CMS to expand this limitation of the NCD to allow for multiple tests per primary cancer —consistent with clinical evidence and practice—to achieve appropriate access for Medicare beneficiaries.

New evidence clearly establishes the value of multiple NGS tests in the duration of a patient’s treatment journey. An NGS panel at the time of diagnosis (primary cancer before first-line treatment is initiated) and subsequent NGS panels at recurrence/progression on first and subsequent lines of therapy fulfill similar and unique purposes.

NGS at diagnosis: An NGS panel at the time of diagnosis simultaneously checks for multiple clinically actionable mutations that help guide physicians to targeted therapies to treat NSCLC. This, in turn, helps timely matching of the patient to the right targeted therapy should a targetable mutation be present. The National Comprehensive Cancer Network (NCCN) guidelines recommend multiplex testing such as NGS platforms for making treatment decisions.

NGS at progression or recurrence: It is now well established that tumors evolve with time in response to targeted therapies. These new molecular alterations confer acquired resistance to targeted therapies and are responsible for progression or recurrence after a patient has received first-line targeted treatments. An NGS panel at the time of progression or recurrence helps identify these new mechanisms of resistance or tumor heterogeneity after treatment with a targeted agent, often independent of the original driver mutation detected at the time of diagnosis. In the recent FLAURA trial of first-line osimertinib in EGFR-positive NSCLC, NGS assays at the time of progression helped identify additional mechanisms of resistance such as a C797S mutation in the EGFR gene and mutations in the PIK3CA and the MET genes. Currently, drugs targeting MET amplification or PIK3CA are in clinical development and there is evidence suggesting that EGFR C797S is sensitive to first-generation EGFR inhibitors such as erlotinib. This suggests that an NGS panel is ideal for determining the next line of treatment for an NSCLC patient who has progressed on a targeted agent and reiterates the importance of multiple NGS panels in a patient’s lifetime.

As stated above, new mutations in NSCLC are being discovered very quickly and limiting access to one test per a patient’s lifetime for a single primary cancer may be detrimental to their treatment and could both prevent their physicians from identifying the accurate first- line targeted therapy that may save their life and impede access to subsequent lines of therapy.

One of the crucial benefits of NGS testing is allowing a complete molecular profile of the patient’s tumor before first-line treatment initiation and after treatment(s), and allowing novel classes of drugs to be offered to the patient as their tumor evolves. Offering an NGS panel at the time of diagnosis and at recurrence or progression also allows for identifying driver mutations that have drugs in clinical development both as first-line treatment options and at progression or recurrence, thereby allowing patients to be enrolled rapidly in clinical trials. This is especially crucial since NCCN guidelines suggest that clinical trials may often offer the best treatment option in first- and subsequent-line settings.

LUNGevity is grateful for the opportunity to comment on the above-captioned LCD and is eager to work with Palmetto to continue to ensure that patients have timely access to high-quality biomarker testing.

Thank you for the comment. At present, we feel that the evidence does not support simultaneous liquid biopsy and tissue genotyping, though we recognize that this is a rapidly evolving area of research. We will continue to monitor the evidence in this area for potential changes to this coverage decision. In regards to the matter of subsequent testing: as the commenter points out, we are covering repeat testing of the same primary tumor. However, progression of a malignancy on therapy would in many cases manifest as new malignant growth that would no longer be considered the same primary cancer.

The current national coverage decision (NCD) for tumor tissue NGS has been very helpful in completing genomic testing for NCCN guideline-recommended genomic alterations, including expanded RAS testing for exons 2,3,4 or the KRAS and NRAS genes, BRAF V600E, and ERBB2 (HER2) amplified colorectal cancer. Each of these biomarkers predict lack of response to cetuximab and panitumumab, anti-EGFR therapies that would otherwise be routinely used in colorectal cancer. This is important because these therapies have high associated costs and pose a 1%-2% risk of life-threatening anaphylactic reactions to those who take them. Even though over half of advanced colorectal cancers harbor these negative predictors of anti- EGFR monoclonal antibodies response, fewer than half of advanced colorectal cancer patients are assessed for them. Although colorectal tumor tissue is readily available via colonoscopy in early stage cancer, at the time of recurrence or metastasis the primary colon lesion has been resected, and biopsy of visceral metastases in liver, lung, or intra-abdominal lymph nodes is required. These repeat invasive biopsies can be often obviated with the non-invasive Guardant360^® test.

Not only are they negative predictors of response to anti-EGFR therapies, but BRAF V600E and ERBB2 (HER2) amplification respond to targeted therapies. Positive outcomes studies using Guardant360^® to detect BRAF V600E and ERBB2 (HER2) amplification have also been published.

NCCN guidelines have recommended assessment of microsatellite instability-high (MSI-H) in advanced colorectal cancer for people under age 50 since 2011, and in all ages since November 2016. However, a 2017 study reported that 72% of CRC patients are not tested for MSI-H. Guardant360^® has recently added the ability to detect MSI-H in all advanced solid tumor cancers. This has important clinical utility in CRC not just for assessment of familial risk, but also because MSI-H was recently FDA-approved for pan-cancer treatment with pembrolizumab.

Targetable fusions in NTRK, BRAF, ALK, ROS1, FGFR2 and FGFR3 have been identified in about 1% -2% of advanced CRC patients, and these are mutually exclusive of clonal BRAF, KRAS, and NRAS mutations. Guardant360^® detects ALK, ROS1, FGFR2, FGFR3 and NTRK1 fusions pan-cancer. NTRK fusions were found in CRC patients who responded to larotrectinib, an NTRK inhibitor and Guardant360^® was the liquid biopsy used in these studies.

When tissue is infeasible, unobtainable, or can’t be performed in a timely fashion, we utilize Guardant360^® for patient care in clinic. Please consider our strong letter of support for the proposed Local Coverage Determination (LCD): MolDX: Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors (DL38043).

Thank you for the comment. We are pleased to see that clinical and academic experts in the field of oncology believe that this expansion of coverage will have significant clinical utility in the treatment of cancer.

We are breast oncology specialists who work in academic cancer centers across the United States. We are writing in support of the pan-cancer (excepting primary brain) advanced solid tumor draft local coverage decision (LCD) proposed by MolDX entitled Guardant360^® Plasma- Based Comprehensive Genomic Profiling in Solid Tumors (DL38043).

The current national coverage decision (NCD) for tumor tissue NGS is a laudable step forward, however a majority of breast cancer patients with metastatic disease have lesions in bone (50%) which are difficult to biopsy and to extract DNA from, or brain (17%) and other deep visceral metastases such or lung (25%). Thus, advanced breast cancer poses significant challenges to genomic testing dependent on the site of tumor biopsy. Currently, there is published evidence for genomic assessment for targetable ERBB2 (HER2) amplification, ERBB2 mutations, germline and somatic BRCA1/2, ESR1 mutations, AKT1 mutations, and PIK3CA mutations in advanced breast cancer – all of these may be non-invasively assessed with Guardant360^®.

Advanced or recurrent breast cancer is not infrequently genomically different from archival tumor tissue from previous resection when the patient was stage I or II. For this reason, NCCN breast cancer guidelines call for a repeat biopsy to assess the genomic status of these recurrent or metastatic tumors. A cardinal example is change in HER2 status from negative in early primary tumors to positive in metastatic or recurrent tumors. This appears to occur in 5%-6% of HER2 negative patients, especially those with brain metastases. These metastases have been shown to be responsive to long established anti-HER2 therapies. Other targetable mutations acquired in advanced breast cancer and not present at initial presentation include ESR1 mutations which drive resistance to endocrine therapies and may require a change to selection estrogen receptor degraders and/or addition of CDK4/6 inhibitors such as palbociclib, or acquired somatic mutations in patients with germline BRCA1/2 mutations, which may predict sensitivity to PARP inhibitors.

Although rare, treatment of advanced breast cancer with NTRK fusions or microsatellite instability-high (MSI-H) has been shown to produce significant responses. Guardant360^® was the liquid biopsy used in both the phase I and phase II larotrectinib outcomes studies for NTRK fusions. Secretory breast carcinoma is an uncommon breast tumor but has a relatively high prevalence of NTRK fusions. Guardant360^® includes detection of MSI-H pan-cancer, including advanced breast cancer – in which it has reported out a dozen cases of MSI-H, representing a new and effective treatment opportunity.

We use Guardant360^® in our clinical practices to identify targetable genomic alterations when tissue is infeasible, unobtainable, or can’t be performed in a timely fashion. Please consider our strong letter of support for the proposed Local Coverage Determination (LCD): MolDX: Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors (DL38043).

Thank you for the letter. We are reassured to hear that oncologists specializing in breast disease believe that this test has clinical utility in the population that they treat.

Thank you for the opportunity to comment on MolDX draft LCD DL38043, which proposes limited coverage for Guardant360^® use in patients who have solid cancers of non-CNS origin.

We are writing in support of the draft local coverage decision (LCD) proposed by MolDX entitled Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors (DL38043). We are academic cancer center-based oncologists specializing in thoracic oncology.

The current national coverage decision (NCD) for tumor tissue NGS definitely has helped our patients with lung cancer get comprehensive genotyping. It is often infeasible to obtain tumor tissue biopsies in advanced lung cancer because of time urgency, difficult tumor location for biopsy, and especially insufficient needle biopsy tumor tissue quantity to test for all genomic biomarkers recommended in NCCN guidelines. This expanding list includes EGFR, BRAF, ERBB2 (HER2), and MET exon 14 skipping mutations, ALK, ROS1, RET, and NTRK fusions, and MET amplification) (www.nccn.org). Tissue feasibility is a key factor in patients not getting comprehensive genomic profiling in non-small cell lung cancer (NSCLC). The tissue failure rate can be even higher when using multiplexed single-gene tests or sequential gene testing protocols. This results in under genotyping of NSCLC patients being common with some studies showing that a minority of patients are tested for all genomic targets, and up to 1/3 are not even getting tested for the most common targetable mutations in the EGFR gene.

Targeted therapies for guideline-recommend genes in large NSCLC precision oncology studies have been shown to improve overall survival 2-3 fold over non-targeted treatment. Prospective studies have shown that Guardant360^® can improve biomarker testing rates in these advanced NSCLC patients both in academic and community settings. In addition, there are now over 20 published clinical utility studies that demonstrate that Guardant360^®-detected genomic targets in NSCLC predict response similarly to the same genomic targets detected from tissue biopsies.

A recent multicenter, prospective study has also shown that using Guardant360^® rather than tumor tissue biopsy as the first test to comprehensively genotype advanced NSCLC patients results in more patients getting completely genotyped for NCCN guideline-recommended targets, with 87% of targets identified with Guardant360^®-first versus only 67% using tissue first. A key result of the blood-first paradigm is that repeat invasive tissue biopsies can often be avoided. Thus, we recommend an edit (in bold below) to the proposed LCD so that oncologists have the option of ordering blood-first for comprehensive genotyping in advanced NSCLC:

– Tissue-based CGP is infeasible (i.e., quantity not sufficient for tissue-based CGP or invasive biopsy is medically contraindicated) or specifically in NSCLC; Tissue-based CGP has shown no actionable mutations or has not been performed.

This enables physicians to make the decision on the best test method to employ for newly-diagnosed NSCLC patients while still reserving the ability to perform tissue CGP if cfDNA is negative.

When tissue is infeasible, unobtainable, or can’t be performed in a timely fashion, we utilize Guardant360^® in newly diagnosed advanced NSCLC patients in our clinics. Also, we test at progression in patients who have not yet been completely genotyped or to look for targetable resistance mutations.

Thank you for the comment. At this time, we do not believe that the evidence yet supports plasma-based CGP when tissue is available, and our understanding is that tissue-based genomic profiling is still considered preferable when it can be performed. However, we recognize that this is an area of active research. We will continue to monitor ongoing research and recognize that we may need to revise this coverage decision in the future.

The current national coverage decision (NCD) for tumor tissue NGS has been very helpful in providing comprehensive genomic testing for NCCN guideline-recommended genomic alterations in advanced GEA, particularly to test for ERBB2 (HER2) amplified and microsatellite- high (MSI-H) tumors (www.nccn.org). However, tissue-based next-generation sequencing (NGS) is challenged by the need to repeat an invasive biopsy for tumor tissue when GEA recurs or becomes metastatic versus a liquid biopsy like Guardant360^®. Although gastric and gastroesophageal tumor tissue is readily available via endoscopy in early stage GEA, at the time of recurrence or metastasis the HER2 status of metastatic lesions has often changed from that of the primary archival biopsy sample, and biopsy of visceral metastases in liver, lung, or lymph nodes is required. Tumor heterogeneity for HER2 status is much greater in GEA than in breast cancer and poses a further challenge to small biopsies. In fact, we have shown that in almost 1/3 of GEA patients that there is a divergence in HER2 status between primary and metastases. Utilization of Guardant360^® in that same study showed that when plasma-based testing identified HER2 amplification discordant from primary GEA tissue, that the plasma result could be confirmed in 88% of patients by metastatic tissue biopsies.

These repeat invasive biopsies in advanced GEA can be often obviated with the non-invasive Guardant360^® test where outcomes studies have shown that plasma-detected HER2 amplification predicts anti-HER2 targeted therapy responses.

Testing for MSI-H is not always performed on primary GEA tissue which may be obtained by resection of an early stage tumor, and often we have to repeat this test on archival tissue or perform a new invasive tissue biopsy to get samples for testing. Guardant360^® has been shown to non-invasively detect MSI-H in GEA, and this serves as an important biomarker of pembrolizumab response. This same study also showed that almost 10% of tissue samples negative for MSI-H may be positive in a second tissue sample from the same patient. Thus, Guardant360^® may not only obviate the need for a repeat invasive tissue biopsy, but it may serve as a more global summary of MSI-H status in patients with tumor heterogeneity.

When tissue is infeasible, unobtainable, or can’t be performed in a timely fashion, we utilize Guardant360^® in newly diagnosed advanced GEA patients in our clinic. Also, we often use it to optimize treatment selection in later lines of therapy when tumor evolution or tumor heterogeneity may be a factor. Please consider our strong letter of support for the proposed Local Coverage Determination (LCD): MolDX: Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors (DL38043).

Thank you for the comment. We are reassured to receive support indicating clinical utility for expanding access of Guardant360^® to patients with gastric and gastroesophageal malignancies.

We are writing to provide input on the proposed MolDX local coverage determination (LCD) DL38043, “Guardant360^® Plasma-Based Comprehensive Genomic Profiling in Solid Tumors”. Overall, we feel this is an important and appropriate acknowledgement of the vital role that genomic information now plays in modern care across many tumor types; however, we have the following recommendations to improve the LCD.

First, we suggest minor revisions to the list of included ICD-10 codes. As the intent of this LCD mirrors that of the National Coverage Decision (NCD) CAG-00450N, which covers NGS-based genotyping for the same population, we suggest that all non-CNS ICD-10 codes listed there be included in this LCD. In our preliminary analysis, it appears that approximately 130 codes are listed in the NCD but not in the LCD. We also note 14 additional codes covering carcinoid tumors (neuroendocrine tumors, in pathological disease classifications) and Kaposi sarcoma that were not included; both of these disease groups are solid tumors that appear to be within the spirit of the LCD (e.g. meet FDA label requirements for larotrectinib’s NTRK fusion and pembrolizumab’s MSI-High pan-cancer approvals) and should be included. (C13.9, C46.0, C46.1, C46.2, C46.3, C46.4, C46.50, C46.51, C46.52, C46.7, C46.9, C7A.096, C7A.098, C7A.1)

Lastly, we note that a substantial number of secondary malignancy codes (used to denote sites of metastatic disease) are not covered. These are common manifestations of otherwise covered entities (e.g. resected primary with only a metastatic recurrence) and should be included.

Second, we suggest including patients that have not been tested in the indications for coverage. As currently written, the LCD’s coverage criteria does not cover testing in patients for whom tissue is technically possible, but where testing is not happening due to logistical or reimbursement barriers with obtaining or ordering tissue genotyping. This is an unfortunately common occurrence in the US’s patchwork system of health care, and we often see genotyping stymied in such circumstances. Liquid biopsy has the ability to help patients in this scenario by circumventing these logistical issues to provide the genotyping data necessary.

In summary, we applaud the foresight shown in the LCD and believe that by increasing access to genotyping information, it will improve the quality of care for a large number of Medicare beneficiaries. Moreover, its emphasis on quality and analytical rigor ensure that such information provided is accurate and protects patients.

Thank you for the comment. We have updated our covered ICD-10 codes to included coverage for all ICD-10 codes which have been transmitted as part Change Request 10878 to implement NCD 90.2. As regards the matter of tissue typing, Medicare covers a number of tests for genotyping tissue, so any Medicare beneficiaries who would be affected by this policy should have no reimbursement barrier to receiving tissue genotyping. Service to address logistical challenges are not a defined benefit category, and therefore we are unable to cover services for this intended use.

We are writing in support of the PROPOSED Local Coverage Determination (LCD): MolDX: Guardant360^® Plasma -Based Comprehensive Genomic Profiling in Solid Tumors (DL38127). Advances in molecular profiling of human malignancy have revolutionized the treatment of cancer, for some patients. The FDA has approved multiple novel gene-specific targeted therapies over the past few years, with dozens of new cancer therapies currently in development. In order to leverage the full potential of these targeted therapies, oncologists require timely and accurate molecular characterization of the patient's cancer to select the best possible therapy.

Biocept, Inc. is a San Diego-based provider of liquid biopsy testing services established in 1997. In the last 20 years, Biocept has developed minimally-invasive blood -based methods to detect predictive biomarkers (e.g., protein expression, gene fusions, amplifications, and oncogene mutations) from patients with cancer utilizing circulating tumor cell (CTC) and cell-free DNA (cfDNA) analysis. A significant portion of Biocept's molecular diagnostic tests, branded as Target Selector™, utilize next-generation sequencing (NGS) technologies to detect and analyze actionable biomarkers that are published in consensus guidelines and are deemed medically appropriate.

Despite the advancement of science, guidelines, positive medical policies, available targeted therapies and evidence of increased overall survival-a significant percentage of oncology patients do not receive a molecular profile and thus do not fully benefit from targeted therapy. This results in a pronounced gap in patient care as noted in the LCD. Finalizing the proposed LCD for Guardant 360^® liquid biopsy will help narrow the gap and provide patients the opportunity to seek personalized cancer care to achieve better outcomes.

To that end, Biocept encourages Palmetto GBA to consider expanding coverage in the LCD to other commercially available and clinically validated plasma-based comprehensive genomic profiling assays performed in CUA certified and CAP accredited clinical laboratories that meet pre-defined performance criteria. Such criteria may include common practice and generally accepted parameters utilized by professional organizations, such as CAP, AMP and the New York State Department of Health (NYSDOH). Facilitating patient access to assays like Biocept's Target Selector™ NGS Breast Panel and Target Selector™ NGS Lung Panel enables simultaneous identification of gene targets for which approved targeted therapies are available and clinical utility has already been established.

Thank you for the comment. We recognize that particular genetic biomarkers have been shown to predict responsiveness of cancer to particular therapies, which is the basis of our coverage for Guardant360^®, and that other liquid biopsies measuring the same biomarkers with similar levels of accuracy and precision would be expected to have the same clinical utility. As such, we have broadened the LCD to allow coverage of liquid biopsies that have similar performance in measuring genetic biomarkers.