Response to Comments: MolDX: Envisia, Veracyte, Idiopathic Pulmonary Fibrosis Diagnostic Test

I am the Director of the Advanced Lung Disease Program and Medical Director of the lung transplant program at Inova Fairfax Hospital in Falls Church, Virginia. I regularly diagnose patients with interstitial lung disease and manage these patients with various therapies, including antifibrotics in those with IPF. I have served on multiple committees, including FDA advisory boards and steering committees for clinical trials in IPF. I am a paid consultant to Veracyte and am also involved in their clinical trial, BRAVE, that was used to train and validate the Envisia classifier.

I want to express my support for your determination that Envisia is medically necessary and reasonable for Medicare patients with a HRCT of the chest demonstrating a probable UIP pattern and/or Indeterminate for UIP.

My first recommendation is to simplify your criteria for "Situations in which Envisia should not be used." I recommend you delete the second sentence in bullet number two; "In particular, Envisia should not be used if any of the following findings are present on HRCT," and delete the list of radiographic features that follow. In my clinical experience, there is rarely one single HRCT feature that results in us categorically defining a patient as being "consistent with a non­lPF diagnosis." It is the collective evaluation of a patient's demographics, clinical history, environmental exposures, and the HRCT features that lead to an overall assessment. Additionally, several of the features listed as exclusion criteria for Envisia use can be present, in some form or degree, in patients with an indeterminate UIP or probable UIP pattern. Agreement can be limited as to the extent and importance of these exact features on HRCT among radiologists as well as ILD clinicians.

A second recommendation would be to allow the use of the Envisia test in select patients who have "CT features most consistent with non-IPF diagnosis". IPF is a disease of the elderly and the older the patient, the greater the likelihood that any ILD is IPF, especially if other causes have methodically been ruled out through a thorough history, physical examination and connective tissue disease panel. Indeed, there is data to suggest that up to 50% of elderly patients (for example>70 years) who have such a HRCT pattern might still have UIP once subjected to lung biopsy. In such a scenario where the pretest likelihood of IPF is very high but the CT is inconsistent, the Envisia test might have a role since these patients might warrant bronchoscopy anyway to elucidate the cause of their ILO. If the Envisia is positive in these circumstances, it might be sufficient to tip the diagnosis to IPF and thereby forego the need for a surgical lung biopsy.

In conclusion, I support finalizing your coverage determination for the Envisia test, and simplifying the situations in which Envisia should not be used to the categorical diagnosis of a typical UIP pattern on HRCT only.

Please see below for response.

I am writing to provide comments on Proposed Local Coverage Determination (LCD): MolDX: ENVISIA, Veracyte, Idiopathic Pulmonary Fibrosis Diagnostic Test (DL37919).

I am a pulmonologist in practice at Columbia University Irving Medical Center in New York City where I am co-director of the interstitial lung disease program. I have considerable expertise in the clinical care of adults with idiopathic pulmonary fibrosis. By way of disclosure, I have received consulting fees from Veracyte in the past, and I continue to have a consulting arrangement with Veracyte for which I receive no fees. I will also disclose that I provided guidance by telephone to Jim Almas during the development of the draft LCD.

I have the following comments:

1. I believe that the listing of specific CT findings is too restrictive. Clinicians should be left to decide if a CT scan does or does not meet Fleischner CT criteria. In particular, I suggest removing the following text:

"Envisia should not be used if any of the following findings are present on HRCT:

1. more than mild ground-glass opacities present in areas away from reticulation
2. more than mild mosaic ground-glass attenuation
3. more than mild expiratory gas trapping pattern
4. peribronchovascular-predominant distribution of disease (as compared to a subpleural- predominant distribution)
5. Substantial upper lobe predominance
6. Unilateral or focal findings
7. Subpleural sparing
8. Diffuse nodules
9. Diffuse cysts"

2. I believe the following statement is ill-informed and misleading: "Patients with an HRCT suggestive of definitive UIP for whom the clinician would recommend testing in lieu of surgical lung biopsy to definitively confirm histopathologic UIP." This sentence cannot be applied in clinical practice. I strongly suggest deleting this sentence or changing the work "testing" to "IPF treatment.”

Please see below for response.

As the Chief Scientific and Medical Officer at Veracyte, Inc, South San Francisco, CA, it is my pleasure to provide you with the enclosed comments regarding the Proposed Local Coverage Determination (LCD): Envisia, Veracyte, Idiopathic Pulmonary Fibrosis Diagnostic Test (DL37857).

I support Palmetto's proposal to provide coverage for Medicare beneficiaries for the Envisia Genomic Classifier testing for patients with interstitial lung disease whom are suspected of idiopathic pulmonary fibrosis.

My enclosed comments are intended to specifically:

1. Express my support for the MolDx Program and finalizing the Envisia Draft LCD
2. Provide recommendations regarding the Criteria for Coverage

I would like to thank you and the Palmetto MolDx team for your thoughtful review of the literature regarding interstitial lung diseases and Envisia Genomic Classifier testing. I believe the Envisia classifier will improve the safety and efficacy of the diagnosis of interstitial lung disease and idiopathic pulmonary fibrosis for Medicare beneficiaries.

Please see below for response.

I practice at the University of Chicago where I am an Associate Professor of Radiology, Section Chief of Thoracic Radiology, and the Interim Vice Chair for Quality in the Department of Radiology. In my practice, I specialize in cardiopulmonary imaging, and have expertise in the interpretation of chest CT of patients presenting with interstitial lung diseases. I am published researcher with interests in the imaging features that present when patients harbor an interstitial lung disease.

I want to complement you on your accurate and complete review of the diagnosis and management of patients presenting with interstitial lung disease with idiopathic pulmonary fibrosis in the differential diagnosis. The imaging and pathologic diagnosis of this disease is complex and highly variable from clinician to clinician, leading to suboptimal patient outcomes. The addition of diagnostic tools such as Envisia should help to improve standardization of the diagnosis and treatment of patients suspected of IPF.

I support your decision that the Envisia Genomic Classifier is medically reasonable and appropriate for Medicare beneficiaries presenting with “Probable UIP” and “Indeterminate for UIP” as described in the 2018 Fleischner Society White Paper on the Diagnostic Criteria for Idiopathic Pulmonary Fibrosis. These are the cases that challenge clinicians, including myself, the greatest in clinical practice, and where a genomic classifier will most significantly improve patient outcomes.

Based on my own published research, I strongly recommend that you revise the section titled “Situations in which Envisia should not be used” within the Criteria for Coverage. It is appropriate to suggest under the first bullet that Envisia should not be used when a “Typical UIP” pattern on HRCT is observed as defined by the 2018 Fleischner Society whitepaper.

However, I recommend you revise the second bullet by deleting the statement I have struck through below:

2. “CT features most consistent with non-IPF diagnosis” on HRCT as defined by the 2018 Fleischner Society White paper. In particular, Envisia should not be used if any of the following findings are present on HRCT:

• More than mild ground-glass opacities present in the areas away from reticulation
• More than mild mosaic ground-glass attenuation
• More than mild expiratory gas trapping pattern
• Peribronchovascular-predominant distribution of disease (as compared to subpleural-predominant distribution)
• Substantial upper lobe predominance
• Unilateral or focal findings
• Subpleural sparing
• Diffuse nodules
• Diffuse cysts

The 2018 Fleischner Society White paper identifies these features as being “most consistent with a non-IPF diagnosis.” The authors did not intend to imply that these features NEVER appear in the presence of UIP and a clinical IPF diagnosis.

I was the lead author of one of the most significant studies evaluating HRCT and histologic features present in patients by UIP diagnostic category in the journal CHEST2. In Table 3 from that study (attached), several of the features listed as criteria to exclude the use of Envisia were present in patients with “Probable UIP” and “Indeterminate for UIP.” For example, while substantially more ground-glass opacities were present in the Inconsistent with UIP group (95%), ground-glass opacities were found in 23% of patients Indeterminate for UIP, and in 7% of patients with Probable UIP.

I believe the 2018 Fleischner Society White Paper provides an excellent framework for clinicians to make a categorical assign a likelihood of UIP from which the criteria for Envisia coverage can be derived. As you point out in the background of your literature review, the coexistence of multiple radiographic features is what makes this diagnosis so challenging. Explicitly excluding patients from Envisia testing due to the presence of a single HRCT feature may result in patients not receiving the optimal treatment.

In summary, I am supportive of your draft Local Coverage Determination for the Envisia Genomic Classifier. And I recommend you limit the “Situations in which Envisia should not be used” to the clinician’s categorical determination of “Typical UIP” or “CT Features most consistent with a non-IPF diagnosis,” and not explicitly exclude coverage in the presence or absence of any single HRCT feature.

I appreciate your consideration of my comments.

Thank you all for your comments. Because all comments are in regards to the same principle topics, they are all addressed concurrently here. We hope this policy will help Medicare beneficiaries have access to necessary and reasonable testing that may prevent unnecessary invasive procedures.

In regards to the comments regarding restricting use of Envisia to prohibit its use with radiological evidence that is not consistent with a diagnosis of IPF, we agree that the language used in the draft LCD may be too restrictive and not result in the intended utilization of the test. We have now revised the LCD to clarify that individual radiological findings do not individually preclude the coverage of the test, but maintain that testing should be restricted to suspected IPF. We therefore preclude coverage when HRCT suggests a “non-IPF diagnosis” or a “Typical UIP” pattern.

Regarding the use of Envisia for patients who have HRCT findings consistent with “non-IPF diagnosis,” the data would have to clearly show benefit for the use of this test in the Medicare patient population to expand coverage to this category.

Lastly, as identified in one comment, unclear language suggesting testing and histologic confirmation in lieu of this test with definitive IPF HRCT has been removed from the coverage policy.