Response to Comments: MolDX: AlloSure® or Equivalent Cell-Free DNA Testing for Kidney and Heart Allografts

The following comment was submitted to Noridian:

We support the policies for blood-based donor DNA detection as a medically necessary test to monitor organ rejection. We note that many recent LCDs from MolDX include text stating that tests with similar indications and performance will also be covered. For example, LCD ARV7 Prostate Biomarker states, "Other assays that detect nuclear AR-V7 in peripheral blood, reporting presence or absence of this biomarker, are also considered reasonable and necessary for the indications above if they also have similar analytical performance for detecting as Oncotype DX AR-V7 Nuclear Detect. Analytical performance will be assessed by MolDX through the technical assessment process." Similarly, LCD Signatera colon cancer test for disease recurrence states, "Another MRD test may also be covered for the above indications upon submission to MolDX if on technical assessment the test displays acceptable analytical and clinical validity."

There are already multiple tests in this clinical area, and there will likely be new entrants on a rolling basis. Therefore, donor DNA transplant tests should have the same language regarding new tests in the future, as seen in many other current MolDX LCDs. New tests will still be closely reviewed by MolDX, but the proliferation of cumbersome lists of LCDs on the same topic can be avoided. We recommend language for "other molecular tests" as new tests may include analytes not in the predicate test or may include parameters like gender or age. As in other LCDs, coverage would apply to similar (but not identical) tests for the same indication following the same coverage reasoning.

Recommendation 1. We propose and recommend that AlloSure Cell Free DNA Testing, should be renamed Organ Transplant Cell Free DNA Testing. Even if the name is not changed, we believe that alternate test language can still be added, as occurs in Signatera and Minimal Residual Disease Testing for Colorectal Cancer which has a branded LCD name but includes alternate test language in the body of the LCD.

Recommendation 2. We recommend this language addition: "Another molecular test may also be covered for the above indications upon submission to MolDX if on technical assessment the test displays acceptable analytical and clinical validity.”

Thank you for your thoughtful comments. We agree with the general concept that policy should not be limited to a specific provider but provide guidance for coverage for services that are reasonable and necessary. If subsequent providers perform similar services with similar performance as the predicate test, they should be similarly covered under the same policy. We will make the recommended changes to the policy to reflect this premise.

The following comment was submitted to Palmetto GBA:

CareDx appreciates this opportunity to comment on the above-referenced MolDX proposed local coverage determination (LCD) for AlloSure use in both kidney transplantation and heart transplantation. The purpose of this comment letter is to provide evidence for the utility of using AlloSure Heart independent of AlloMap in heart transplantation.

AlloSure for use in kidney transplantation was covered by MolDX in October 2017 in a policy specific to the performance in kidney transplantation, associated with Z-Code ZB11X. In the spring of 2019, we submitted a dossier seeking coverage of AlloSure in heart transplantation, with clinical validation and utility based primarily on Khush et al., Am J Transplant 2019. The current draft LCD includes coverage for both kidney transplantation and heart transplantation in a single policy. In addition, this draft policy covers AlloSure in heart transplantation (associated with code ZB8DL) only when used in conjunction with AlloMap. The draft LCD does not consider the use of AlloSure in heart transplantation independent of AlloMap.

The draft LCD references Crespo-Leiro et al., 2015 as a publication that demonstrates higher AUC when AlloMap and AlloSure Heart are used together. However, AlloMap and AlloSure Heart each have independent validity and utility. AlloMap clinical validity and utility has been demonstrated in many peer reviewed publications since 2006, including a landmark clinical utility study, Pham et al., NEJM 2010. As mentioned above, independent validity for AlloSure Heart has been published in Khush et al., 2019, in addition to Crespo-Leiro et al., 2015. Additionally, New York State has granted conditional approval for AlloSure Heart independent of AlloMap (Project ID 63259).

Independent use of AlloSure Heart is reasonable and necessary for numerous scenarios. Those defined in this comment include: early post transplantation; high dose corticosteroid maintenance immunosuppression; rejection treatment, especially with steroids; when the major concern is antibody mediated rejection; and patients specifically at risk for cytomegalovirus infection. Each of these is addressed below.

Early Post Transplantation

AlloMap is FDA-cleared for use in heart transplant recipients ≥55 days post-transplant. The AlloMap score is derived from gene expression measurement in immune cells that interact with the transplanted heart. The expression of some of the genes in the AlloMap signature and therefore the AlloMap score can be influenced by treatments or other factors in the first two months post transplantation. There remains significant unmet need to evaluate patients for rejection earlier than 55 days post-transplant. With the 30-day mortality rate around 10% this is a time period with significant risk of rejection (Luckraz et al., 2005).

AlloSure Heart is a measure of donor-derived cell-free DNA (dd-cfDNA), an indicator of graft injury. While dd-cfDNA is elevated above baseline very early post-transplant, presumably by the transplant surgery and ischemia-reperfusion injury, it resolves to a baseline level by day 14 post-transplant. Research studies have demonstrated that dd-cfDNA can be measured successfully in the first two months post-transplantation, including De Vlaminck et al., 2014, making it an important tool in the early post-transplant period.

In the D-OAR study and SHORE registry, data have been collected that characterize AlloSure Heart used in heart transplant recipients earlier than 55 days post-transplant. These data (Table I) show that the distribution of AlloSure Heart results is similar between two groups representing the days before and after the 55 day cutoff. These data have been submitted as an abstract for publication in Am J Transplant and presentation at to the American Transplant Congress 2020 (Teuteberg 2020).

Table I. Descriptive Statistics of AlloMap and AlloSure Heart Scores Stratified by Days Post Transplant Category

High Dose Corticosteroid Maintenance Immunosuppression

A subset of the genes in the AlloMap gene expression test are influenced by high levels of corticosteroids used for maintenance immunosuppression (Starling et al., 2006). Therefore, clinicians are cautioned that systemic corticosteroid dosage of greater than 20 mg/day of prednisone may artificially decrease the AlloMap score. High doses of steroids can be used by some centers for several months post-transplant in a subset of heart transplant recipients where there is concern for rejection. In these patients there is an obvious concern for rejection and an unmet need to non-invasively identify rejection. With 15-20% of recipients dying within 1 year of surgery, having a noninvasive tool to quantify allograft injury as part of their surveillance has been shown to be very valued (Wilhelm et al., 2015).

Since AlloSure Heart is a measure of allograft injury, not recipient immune cell gene expression, steroids do not influence the result. Data from the D-OAR study and SHORE registry (Table II) show that there is no influence of steroid dose on the AlloSure Heart result. These data have been submitted as an abstract for publication in Am J Transplant and presentation at to the American Transplant Congress 2020.

Table II. AlloSure Heart measurements (dd-cfDNA, %) stratified by steroid dose, combined data from D-OAR and SHORE.

Rejection Treatment

Following treatment for a diagnosis of rejection, physicians are concerned with whether the treatment has been successful. Information regarding success or failure of the treatment will dictate whether there is a need for more or different treatment.

AlloMap, as a measure of gene expression in the immune cells, may be directly impacted by treatments given for rejection (often very high dose steroids) even if the treatment has not sufficiently impacted the rejection process. Therefore, clinicians are cautioned that the performance characteristics of AlloMap have not been established for patients who have received treatment for rejection within the past 21 days.

Since AlloSure Heart is a real-time indicator of allograft injury, it can be used to identify the impact of treatment without concern that the test is impacted by the treatment. Data supporting this use have appeared in an abstract and presented at the International Society of Heart and Lung Transplantation Annual Meeting (Stypmann et al., 2015).

Antibody-Mediated Rejection

The AlloMap gene expression signature has strong performance in acute cellular rejection. However, antibody mediated rejection (AMR) has been shown to have less influence on the AlloMap score.

Despite aggressive treatments, AMR is associated with a sustained increased risk of mortality and with development of cardiac allograft vasculopathy (CAV) and resulting graft dysfunction (Clerkin et al., 2016). With CAV affecting over a third of all allografts, using AlloSure Heart to detect and quantify this injury as early as possible has high value (Ramzy et al., 2005).

In kidney transplantation there is significant association between AlloSure and AMR (Bloom et al., 2017). In heart transplantation, data published by Khush et al., 2019 demonstrate a significant association with AMR in a high-risk population. Clinicians may choose to use AlloSure Heart and not AlloMap in situations where the prior risk for ACR is low and the risk of AMR is high. This is often a concern immediately post-transplant when the steroid dose is also still higher than 20mg/day. Two recent single center analyses of patients with AMR diagnosed beyond one year post-transplant demonstrated 50-53% mortality within one-year (Coutance et al., 2015 and Hodges et al., 2012).

Cytomegalovirus Infection

Cytomegalovirus infection has been shown to influence the AlloMap score in patients not experiencing rejection (Fujita et al., 2017). In situations where the risk of CMV infection is greater than the risk of infection, clinicians may choose to avoid using AlloMap. In these patients, there is a need for non-invasive testing for graft rejection. One example of a patient at high risk for CMV infection is a patient negative for CMV who receives a heart from a donor that was positive for CMV. In these situations, the clinician may choose to use AlloSure Heart to identify rejection. Data from the D-OAR study has been evaluated to determine whether CMV infection impacts the AlloSure Heart result. These data show that CMV does not impact the AlloSure Heart result in situations where the AlloMap score is impacted by CMV. These data have been submitted as an abstract for publication in J Heart and Lung Transplant and presentation at to the International Society of Transplantation 2020 (Kanwar et al., 2020). They have also been prepared as a manuscript for submission of a peer-reviewed publication.

Summary

The data presented and publications referenced in this comment address the use of AlloSure Heart independent of AlloMap. AlloMap has an extensive record of publications and the value of AlloMap for non-invasive means to detect allograft rejection status is well established.

The use of AlloSure Heart in conjunction with AlloMap provides demonstrated performance to discriminate rejection that is greater than either individual test. However, there is significant evidence supporting the use of AlloSure Heart in heart transplantation independent of AlloMap defined testing protocols. Multiple clinical situations exist where the independent use of AlloSure Heart would be needed and warranted to provide critical value in heart transplantation patients (Table III). These include early post transplantation, high dose steroid immunosuppression, following rejection treatment, high risk for antibody-mediated rejection, and specific risk for CMV infection.

We recommend modifying the draft LCD to include independent use of AlloSure Heart in heart transplantation, while retaining coverage for use of AlloSure Heart in conjunction with AlloMap.

The flexibility to use both AlloMap and AlloSure Heart together, AlloMap individually, or AlloSure Heart individually, enables optimal patient management and the flexibility for the physician to personalize the surveillance to the situation most valuable for each patient.

Table III. Summary of Clinical Scenarios for Independent Use of AlloSure Heart

Recommendations for Coverage

CareDx suggests coverage for AlloSure as follows:

Coverage Indications, Limitations, and/or Medical Necessity

This Medicare contractor will provide limited coverage for tests performed using the AlloSure donor-derived cell-free DNA assay (CareDx, Inc., Brisbane, CA) for the following indications:

1. AlloSure Kidney is covered to assess the probability of allograft rejection in kidney transplant recipients with clinical suspicion of rejection and to inform clinical decision- making about the necessity of renal biopsy in such patients at least two weeks (14 days) post-transplant in conjunction with standard clinical assessment.
2. AlloSure Heart is covered to assess the probability of allograft rejection in heart transplant recipients with clinical suspicion of rejection and to inform clinical decision-making about the necessity of a heart biopsy in such patients at least two weeks (14 days) post-transplant in conjunction with standard clinical assessment.
3. AlloSure Heart is covered when used in conjunction with AlloMap to assess the probability of allograft rejection in heart transplant recipients with clinical suspicion of rejection and to inform clinical decision-making about the necessity of a heart biopsy in such patients at least 55 days post-transplant in conjunction with standard clinical assessment.

Coding Table

Thank you for your comments and suggestions. The strongest evidence considered in this policy included the use of both AlloSure^® and AlloMap^®, which resulted in the coverage determination presented in the draft policy. Based on your comments, there are other potential indicated uses of AlloSure^® independent of AlloMap^®. Some of these uses, such as for identifying rejection with contaminant CMV infection, early rejection post day 14 of treatment, or high corticosteroid use, are apparently still under investigation and we may only evaluate published evidence. We will continue to evaluate evidence as it is published in peer-reviewed publications.