Coverage Indications, Limitations, and/or Medical Necessity
This Medicare contractor will provide limited coverage for the Prospera™ donor-derived cell-free DNA test (dd-cfDNA) (Natera, Inc., San Carlos, CA) to supplement the evaluation and management of kidney injury and active rejection (AR) in patients who have undergone renal transplantation. It can inform decision making along with standard clinical assessments.
Criteria for Coverage
The Prospera™ assay is covered only when the following conditions are met:
- The patient has a renal allograft
- Physician-assessed pretest need to further evaluate patient for the probability of active renal allograft rejection
Summary of Evidence
In 2016, over 20,000 kidney transplants were performed in the United States. In addition, at least 80,000 surgical candidates were on dialysis while they waited for an available kidney.1 After transplantation, patients are put on immunosuppressant drug therapy and routinely monitored to prolong the survival of the donor kidney. The current standard-of-care for managing renal transplant patients has severe limitations: 20-30% of transplanted kidneys fail within five years and only 55% survive to ten years.2,3
The cost for managing a failed renal transplant patient may be 500% more than a patient with a functioning transplant.4 Meanwhile, current tools for diagnosing active allograft rejection (AR), the leading cause of graft failure, are inadequate. Early detection of AR has led to significant improvement in allograft survival in the first 12 months posttransplantation5. However, the traditional tools used for this are either invasive (biopsy) or inaccurate (serum biomarkers). This environment creates an unmet need for timely, sensitive, specific, non-invasive diagnostic tools to improve kidney transplant management.
The Prospera™ test detects dd-cfDNA in a recipient’s blood, which is elevated during AR due to increased cell death in the organ.6 Prospera™ is an effective, non-invasive method of assessing kidney allograft status with better performance than the current standard-of-care.6
Prospera™ Test Description and Performance
The Prospera™ assay is a cell-free DNA-based, next-generation sequencing assay that targets over 13,926 single-nucleotide polymorphisms (SNPs) to accurately quantify the fraction of dd-cfDNA in the transplant recipient’s blood, even in related recipient/donor pairs, without separate genotyping of either donor or recipient.6 The dd-cfDNA fraction in cfDNA can be measured with a turnaround time of up to 3 days; this turnaround time is necessary for the appropriate management of transplant recipients.
The clinical performance of Prospera™ was evaluated in a retrospective analysis of 217 samples from 178 unique transplant recipients whose kidney transplant was performed at the University of California at San Francisco Medical Center.6 The data demonstrated that the dd-cfDNA levels in patients with AR was significantly higher than patients where AR was not detected during biopsy.6
The amount of dd-cfDNA was significantly higher in the circulating plasma of the AR group (median=2.32%) compared with the non-rejection group (median=0.47%; P<0.0001).
Using a prospectively determined dd-cfDNA cutoff of 1%, the data showed the Prospera™ assay to have an 88.7% sensitivity (95% confidence interval [CI], 77.7%-99.8%) and 72.6% specificity (95% CI, 65.4%- 79.8%) for detection of AR. The area under the curve (AUC) was 0.87 (95% CI, 0.80-0.95). Based on a 25% prevalence of rejection in a clinically suspicious, at-risk population, the positive predictive value (PPV) was projected to be 52.0% (95% CI, 44.7%-59.2%) and the negative predictive value (NPV) was projected to be 95.1% (95% CI, 90.5%-99.7%).
Furthermore, the Prospera™ assay performance was assessed separately in a protocol biopsy (aka surveillance biopsy) population, which is expected to reflect performance when rejection is suspected but there is no overt clinical evidence of renal function deterioration (subclinical AR). This population showed detection of subclinical AR with 92.3% sensitivity (95% CI, 64.0%-99.8%), 75.2% specificity (95% CI, 65.7%-83.3%), and a 0.89 AUC (95% CI, 0.76-0.99). Based on a 15% prevalence of rejection in this monitoring population, the PPV was projected to be 39.7% (95% CI, 30.7%-48.7%) and the NPV was projected to be 98.2% (95% CI, 94.9%-100%). These data suggest that application of the Prospera™ assay in a clinical setting could potentially reduce unnecessary biopsies. Its demonstrated performance characteristics may allow use in the management of AR with immunosuppressants or the assessment of resolution of rejection episodes.
Median dd-cfDNA did not differ significantly between different types of rejection: antibody mediated rejection (ABMR; 2.2%), ABMR/T cell-mediated rejection (TCMR; 2.6%), or TCMR (2.7%) groups (P=0.855).
The analytical and clinical performance of the Prospera™ assay is summarized below.
The Prospera™ test is intended to supplement evaluation and management of kidney injury and AR in patients who have undergone renal transplantation. It may be used by physicians considering the diagnosis of AR, helping to rule in or out this condition when assessing the need for or results of a diagnostic biopsy. It should be considered along with other clinical evaluations and results, and may be particularly useful in patients with significant contraindications to invasive procedures.
Plasma collected in Streck Cell-Free DNA BCT® tubes
Slope = 1.0664 (95% CI 0.9416, 1.1912)
Intercept = 0.0008 (95% CI -0.0076, 0.0092)
R-squared = 0.9997 (95% CI 0.9997, 0.9998)
Slope = 1.0333 (95% CI 0.9241, 1.1425)
Intercept = -0.0001 (95% CI -0.0047, 0.0046)
R-squared = 0.9989 (95% CI 0.9986, 0.9990)
Accuracy was assessed using linear regression with respect to digital droplet PCR (ddPCR) (for CNV2 using probes specific for chromosome 1 as reference and chromosome Y as unknown) as a reference method. Three cell line-derived reference mixtures (1 related donor, 2 unrelated donors) at mixture fractions from 0.1% to 15% were run in minimum triplicates at 15, 30, 45 ng input DNA mass. The total number of measurements was 285 unrelated and 349 related.
15 ng minimum input tested in samples where input cfDNA concentration was measured.
Limit of Detection
0.15% unrelated donor
0.29% related donor
Limit of Detection (LoD) was assessed from 3 cell lines (2 unrelated, 1 related) reference panels at 15, 30, 45 ng input DNA mass, and 16 plasma-derived cfDNA mixtures at variable input mass, at mixture fractions 0.1, 0.3, 0.6%. Samples were run in minimum triplicates by two operators using two reagent lots, for a total of 168 (94 from Lot 1, 74 from Lot 2) and 220 (115 from Lot 1, 105 from Lot 2) measurements for unrelated and related donors, respectively. Samples from each reagent lot were evaluated using each of the two designs for environment (dfe) methods, unrelated and related donor. LoD values for each lot and each dfe method are calculated by using the Limit of Blank (LoB) values of the corresponding method. For each method, the final LoD is the maximum of Lot 1 and Lot 2 LoD values calculated with the corresponding method. Calculations followed the parametric method described in EP17A2 (see Appendix I [Technology Assessment of Similar Test: AlloSure LCD]).
Lower Limit of Quantitation
0.15% unrelated donor
0.29% related donor
Lower limit is assessed on the same sample set used for LoD, with a broader range of mixture fractions. Specifically, reference samples were tested at mixture fractions 0.1, 0.3, 0.6, 1.2, 2.4, 5, 10, 15%, and plasma-derived cfDNA mixtures were tested at mixture fractions 0.1, 0.3, 0.6, 1.2, 2.4, 5, 10%. Samples were run in minimum triplicates by two operators using two reagent lots, for a total of 381 (207 from Lot 1, 174 from Lot 2) and 412 (239 from Lot 1, 173 from Lot 2) measurements for unrelated and related donors, respectively. Lower limit is defined as the lowest value of donor fraction at which measurement CV (defined as the measurement standard deviation divided by the mean) is less than 20%. The requirement is satisfied over the entire tested range, for each reagent lot and dfe method, so the lower Limit of Quantitation (LoQ) is equal to the LoD by constraint that it cannot be less.
Upper Limit of Quantitation
Upper limit of quantitation = 15% for unrelated and related donors based on highest value tested.
Reference range defined as 0 to 1% based on previously published and approved technology using same analyte with corresponding patient population.6,7
Interference of excess ethanol carry-over and excess ethylenediaminetetraacetic acid (EDTA) on the multiplex PCR reaction was evaluated using Natera’s cfDNA protocol. Inhibitory effect on mmPCR reaction was observed at ethanol concentration of 5% and higher and EDTA concentration of 10 mM and higher. Visually hemolyzed samples are excluded from processing.
Clinical Performance: Validity
||Results (with 95% Confidence Intervals if applicable)
The amount of dd-cfDNA was significantly higher in the circulating plasma of the AR group (median=2.32%) compared with the non-rejection group (median=0.47%; P<0.0001). Median dd-cfDNA did not differ significantly between different types of rejection: ABMR (2.2%), ABMR/TCMR (2.6%), or TCMR (2.7%) groups (P=0.855). The data demonstrate that the Prospera™ assay can accurately discriminate AR from non-rejection across a range of pathologies, including both acute and chronic findings, in both the ABMR and TCMR groups.
Analysis of Evidence
(Rationale for Determination)
Level of Evidence
Quality of evidence – Moderate
Strength of evidence – Moderate
Weight of evidence – Low
The evidence is sufficient to support that Prospera™ provides a non-invasive assessment tool to assess for the presence of active allograft rejection. While the evidence does not indicate that the assay is a replacement for biopsies, the evidence does support that the assay performs sufficiently well to be covered if the patient’s treating clinician believes that a patient is need of an assessment for rejection that would, and he or she believes that the risk / benefit profile of this assay is superior to the risk / benefit profile of biopsy for that patient. Performance criteria demonstrate that the mmPCR-NGS method is able to discriminate AR from non-rejection (borderline rejection, stable transplant and other injury) status with an AUC of 0.87. This test performance is sufficient to help physicians rule in and rule out AR, in combination with other clinical factors. The evidence also supports that Prospera™ identifies both ABMR and TCMR, and it is validated to detect subclinical AR.
While the existing published evidence is sufficient to support coverage, the company has indicated that two post marketing studies are underway and will be submitted for peer-reviewed publication: a dd-cfDNA clinical utility randomized controlled trial which will study physician-reported current clinical practice and impact of Prospera™; and a dd-cfDNA registry study to further assess Prospera™’s impact on actual biopsy usage. As such, we anticipate ongoing evolution of the evidence and potential changes in accepted clinical practice, and we will continue to monitor the evidence for possible changes to coverage policy.