US payer budget impact of a microarray assay with machine learning to evaluate kidney transplant rejection in for-cause biopsies

Abstract

Aim

This study evaluates the economic impact to US commercial payers of MMDx-Kidney used in conjunction with histologic evaluation of for-cause kidney transplant biopsies.

Materials and methods

An Excel-based model was developed to assess the cost impact of histology plus MMDx-Kidney versus histology alone for the evaluation of potential rejection in kidney transplant patients who receive a for-cause biopsy. Different model time periods were assessed, ranging from 1 to 5 years post-biopsy. A targeted literature review was used to identify parameter estimates, validated by two external clinicians with expertise in managing kidney transplant rejection. A sensitivity analysis was conducted to evaluate the relative impact of key clinical and cost parameters. In particular, the model identified the magnitude of MMDx-Kidney’s impact on graft failure from rejection that would be required for MMDx-Kidney to be cost-neutral.

Results

By more accurately characterizing rejection, MMDx-Kidney is estimated to increase antirejection treatment costs by $1,126 per test. Nevertheless, a break-even analysis shows that the costs of MMDx-Kidney and anti-rejection medication, as well as the costs associated with an increase in the number of patients with functioning transplants, may be offset by reductions in costs associated with graft failure (i.e. costs of hospitalizations, dialysis, and repeat transplants) over 5 years, assuming MMDx-Kidney reduces annual graft failure from rejection by at least 5%. For the base case, with a 25% relative reduction in annual rate of graft failures from rejection, MMDx-Kidney increases overall costs incurred in the first year of the model but starts generating savings by the second year of the model.

Conclusions

Compared with histologic evaluation of for-cause kidney transplant biopsies alone, the use of MMDx-Kidney in conjunction with histologic evaluation improves the diagnoses of graft dysfunction and may have the potential to generate overall savings from reductions in rejection-related graft failure.

Keywords:

• Kidney biopsy
• transplant rejection
• graft failure
• mRNA
• gene
• budget
• cost

JEL Classification Codes:

• I15
• I1
• I
• I10

Introduction

According to the Organ Procurement and Transplantation Network (OPTN), 22,817 kidney transplants were performed in the United States in 2020¹. Kidney transplantation has grown steadily by 20% in the last 5 years alone¹. This increase is driven by the rise in eligible deceased-donor kidney transplants to over 77% of all kidney transplants from 40% in 2015¹. While greater opportunities for transplant benefit patients, deceased-donor transplants present an increased risk of rejection; approximately 60% of kidney graft failures in the US today are due to chronic or acute rejection². Currently, 21% of deceased-donor kidney transplant recipients and 14% of living-donor kidney transplant recipients experience graft failure within 5 years of transplant.

Potential treatment options following kidney graft failure are dialysis or another kidney transplant. The median wait-time for a kidney transplant recorded in 2013 was 49.2 months, due to the limited availability of donor kidneys³. Patient survival rates with dialysis are also poor when compared to survival rates for patients with a functioning kidney transplant⁴. The median overall survival after kidney graft failure is estimated to be 3 years⁵.

The economic burden of graft failure in kidney transplant recipients is substantial. Based on a 2017 US transplant cohort, graft failure from rejection imposes a cumulative economic burden to Medicare of $698 M, amounting to roughly half of the total economic burden of graft failure². According to the United States Renal Data System (USRDS), at the individual patient level, a functioning transplant cost Medicare $28,106 per year in 2018, compared to $126,771 for dialysis in the year of graft failure and $136,696 for a repeat transplant in the year of the initial graft failure³.

Optimal therapy directed at preserving transplant viability for as long as possible requires accurate, timely assessment of the cause of a failing graft. Histologic biopsy interpretation and measurements of serum creatinine, glomerular filtration rate (GFR), and proteinuria are the current gold standards for assessing renal allograft function after kidney transplant, but these tools often do not accurately detect whether allograft dysfunction derives from injury, antibody-mediated rejection (ABMR), T-cell-mediated rejection (TCMR), or ABMR and TCMR combined (mixed rejection).

Grades of rejection are defined by the Banff Classification of Allograft Pathology – established in 1991 and updated regularly as additional research informs consensus, with the most recent criteria set at the 2019 meeting⁶. Guided by a histology assessment based on the Banff criteria, pathologists exhibit a 30% error rate when evaluating kidney allograft dysfunction⁷. In addition to inadequate detection of rejection type, clinical interpretation of histology is subjective and, thus, remains prone to inter- and intra-observer variability^8–15. Further, as the tissue becomes more abnormal, the likelihood that the pathologist’s assessment departs from Banff criteria increases¹⁶. The Banff guidelines themselves can be a limitation, too, as classifications in the guidelines are often ambiguous^6,17,18. Lastly, the sample type required by histology may also be a limitation. For example, histology requires 10 glomeruli and two arteries, but the medulla contains none of the required glomeruli; thus, histology alone cannot assess rejection in the medulla of biopsies¹⁹.

The Banff 2019 meeting report advocates for the implementation of molecular diagnostics in conjunction with histology to improve the evaluation of renal allograft rejection and prognosis of graft failure, particularly when Banff classifications are ambiguous (i.e. ABMR in the absence of donor-specific antibodies). More specifically, the report names Molecular Microscope MMDx-Kidney (MMDx-Kidney) as a “more pathogenesis-driven way” to improve diagnostic capabilities⁶.

MMDx-Kidney uses microarrays to measure messenger ribonucleic acid (mRNA) levels and profiles 1,494 genes in fresh tissue transplant biopsies. With predefined algorithms, MMDx-Kidney assesses the probability of rejection or injury based on observed patterns in gene expression in the biopsy. In particular, MMDx-Kidney is intended to complement all available diagnostic information for more precise evaluation and management of TCMR, ABMR, acute injury, chronic injury, atrophy fibrosis (AF), arteriolar hyalinosis (AH), acute kidney injury (AKI), and disturbance (inflammation). MMDx-Kidney diagnoses abnormalities in histologically problematic kidney transplant biopsies in all donor types, unrelated and related, living and deceased⁸. In addition, unlike histology, which generates binary or semi-quantitative grades (0, 1, 2, or 3), MMDx-Kidney provides continuous results, which are especially critical when a biopsy has values near boundaries.

In particular, MMDx-Kidney is intended to be used with for-cause biopsies (also known as indication biopsies), which are defined as biopsies conducted outside pre-determined intervals. Such biopsies are more likely to present with rejection than protocol or surveillance biopsies, and, thus, MMDx-Kidney is more likely to provide needed guidance when used with for-cause biopsies²⁰.

Accurate diagnoses of dysfunctional kidney grafts may allow for optimized immunosuppressive treatment and improved outcomes, including graft survival. Based on our extensive review of published literature, this paper is the first to explore the economic impact associated with a reduction in rates of graft failure with MMDx-Kidney. The budget-impact model assesses the potential overall costs or savings associated with the use of MMDx-Kidney for a US commercial payer compared to the current standard of care: histopathology evaluation.

Methods

Study design

To estimate the economic impact to a US commercial payer of histology plus MMDx-Kidney versus histology alone following a for-cause biopsy in kidney transplant patients, a budget impact model was developed in Excel. The dynamic model was intended to allow easy examination of several scenarios, including varying assumptions about the impact of MMDx-Kidney on graft failure from rejection. The target population comprises kidney transplant recipients receiving a for-cause biopsy. The care pathway that the model follows is illustrated in Figure 1.

Figure 1. Care pathway of patients with for-cause renal biopsy.

In the base case of the model, patients presenting with symptoms of abnormal kidney graft activity receive a for-cause biopsy and are allocated to the following outcome categories:

1. Functioning initial transplant;

2. Graft failure and re-transplant in the same year;

3. Graft failure and dialysis in the same year;

4. Death with a functioning graft; and

5. Death after graft failure.

In the base case scenario, the model assumes MMDx-Kidney reduces the annual rate of graft failure from rejection by 25% in relative terms (e.g., from a 5.5% graft failure rate in the first year post-biopsy to 4.1%). While the magnitude of MMDx-Kidney’s impact is an assumption, a recent MMDx-Kidney study has shown that adding ABMR scores to the conventional assessment improves the prediction of 3-year graft failure by 23%²¹. The model calculates undiscounted savings or losses in five separate model duration scenarios, ranging from 1 to 5 years after a patient receives a for-cause biopsy.

Model development

A focused literature review was used to develop a clinical care pathway following the initial for-cause biopsy, as well as treatment costs for rejection and outcomes post-biopsy. The estimates identified as the baseline parameters are from peer-reviewed publications, the United States Renal Data System (USRDS) database, Scientific Registry of Transplant Recipients (SRTS)/OPTN database, and Centers for Medicare and Medicaid Services Fee-for-Service (CMS FFS) data. The identified parameters were further verified through consultation with two practicing physicians experienced in the diagnosis and management of kidney transplant rejection (G.G. and C.L.) to ensure values reflected real-world experience.

Model assumptions

Given current limitations in longitudinal testing and outcomes data, the base-case model relies on several simplifying assumptions. First, patients receive a single diagnosis after exhibiting symptoms of graft dysfunction that results in a for-cause biopsy. Consequently, the model reflects one set of antirejection treatments per patient diagnosed with rejection and does not account for additional for-cause biopsies or multiple uses of MMDx-Kidney per patient in the time span of the model. Second, patients who start a certain type of dialysis after a graft failure do not switch to another type of dialysis in the time frame of the model. Third, while the model identifies patients who receive a second transplant in the year of graft failure, patients with graft failure who start dialysis do not receive a second transplant in the time frame of the model. Fourth, outcomes of death, functioning transplant, and graft failure occur after patients prescribed therapy for rejection have fully completed their treatment. Fifth, while the model identifies patients who do not receive a rejection diagnosis, the model does not specify the number of patients who are non-adherent to immunosuppression therapy or who have graft injury or scarring.

Outside the base case, the assumption regarding the number of MMDx-Kidney tests per patient is adjusted in a single scenario: an evaluation of the number of MMDx-Kidney tests required to eliminate savings over 5 years, given a 25% reduction in annual graft failure from rejection with the test.

Kidney rejection diagnosis and treatment

Based on the prospective INTERCOMEX study of kidney transplant biopsies in North America and Europe, Table 1 presents the percentage distribution of patients by diagnosis status with the standard of care (histology) versus histology plus MMDx-Kidney²². With MMDx-Kidney, 54% of indication biopsies in the study showed “no-rejection” and 46% showed “rejection”. MMDx-Kidney further classified rejection biopsies by rejection type.

Table 1. Percentage distribution of diagnoses after for-cause kidney biopsy.

Diagnosis after for-cause biopsy	SoC^a (histology) (%)	SoC and MMDx-Kidney^a	References/Notes
Mixed rejection	3	5	^21
No rejection	56	54
Active ABMR	20	30
ABMR suspicious and transplant glomerulopathy	5	3
Suspicious (borderline) for acute T-cell-mediated rejection	8	1
T-cell mediated rejection	8	7

^aThe total distribution for SoC and SoC and MMDx-Kidney is 100%.

^bThe percentage of mixed rejection is calculated by subtracting all other diagnoses from 100%.

Abbreviation. SoC, standard of care.

While all patients diagnosed with TCMR, mixed rejection, and suspicious/borderline for acute TCMR receive treatment, 80% of the patients with an ABMR diagnosis and 20% of patients with an ABMR suspicious and transplant glomerulopathy diagnosis receive treatment, based on the experience of two clinical experts (G.G. and C.L.). Treated patients are assigned diagnosis-specific therapies derived from peer-reviewed sources and confirmed by clinical practice (G.G. and C.L.) (Table 2 and Supplementary Table S1). Based on an analysis of hospital claims for the ICD-10 T86.11 code for transplant rejection, approximately one third of treated patients are assumed to receive antirejection therapies in an inpatient setting, with the rest receiving outpatient treatment²³. Patients without rejection do not receive any additional treatment in the model.

Table 2. Antirejection treatment percentage distribution and costs for outpatient visit.

Treatment	% Distribution	Cost to Payer	CPT/HCPCS codes
Treatment cost for ABMR
Plasmapheresis	100	$11,031	(CPT: 36514, 96365, 96366)
IVIg	100	$2,653	(CPT: 96365, 96366; HCPCS: J1560)
Methylprednisolone (steroids)	100	$1,297	(CPT: 96365, 96366; HCPCS: J1040)
Bortezomib	15	$3,586	(CPT: 96365, 96366; HCPCS: J9044)
Rituximab	5	$3,860	(CPT: 96365, 96366; HCPCS: J9312)
Treatment cost for suspicious ABMR
Plasmapheresis	100	$11,031	(CPT: 36514, 96365, 96366)
IVIg	100	$2,653	(CPT: 96365, 96366; HCPCS: J1560)
Methylprednisolone (steroids)	100	$1,297	(CPT: 96365, 96366; HCPCS: J1040)
Bortezomib	15	$3,586	(CPT: 96365, 96366; HCPCS: J9044)
Rituximab	5	$3,860	(CPT: 96365, 96366; HCPCS: J9312)
Treatment cost for TCMR
Methylprednisolone (steroids)	100	$1,297	(CPT: 96365, 96366; HCPCS: J1040)
rATG	20	$20,102	(CPT: 96365, 96366; HCPCS: J7511)
Suspicious (borderline) for TCMR
Methylprednisolone (steroids)	100	$1,297	(CPT: 96365, 96366; HCPCS: J7511)
Mixed Rejection
Plasmapheresis	100	$11,031	(CPT: 36514, 96365, 96366)
IVIg	100	$2,653	(CPT: 96365, 96366; HCPCS: J1560)
Methylprednisolone (steroids)	100	$1,297	(CPT: 96365, 96366; HCPCS: J1040)
Bortezomib	15	$3,586	(CPT: 96365, 96366; HCPCS: J9044)
Rituximab	5	$3,860	(CPT: 96365, 96366; HCPCS: J9312)

Abbreviations. ABMR, antibody-mediated rejection; IVIG, intravenous immune globulin; TCMR, T-cell mediated rejection; rATG, rabbit antithymocyte globulin; CPT, Current Procedural Terminology; HCPCS, Healthcare Common Procedure Coding System.

Clinical outcomes

The proportion of patients with functioning transplants, mortality, graft failure from rejection, and graft failure from other causes over 5 years was estimated using the 30-year simulation model developed by Sussell et al². (Table 3). Further, the relative proportion of annual mortality in patients with graft failure versus patients with a functioning graft was estimated from an assessment of USRDS data reported by Kaplan and Meier-Kriesche²⁴.

Table 3. Rates of outcomes 5 years after for-cause kidney biopsy.

Patient outcome	Value (%)	References/Notes
Annual mortality		^2
Year 1 after for-cause biopsy	2.5
Year 2	2.6
Year 3	2.6
Year 4	2.7
Year 5	2.8
Annual graft failures (rejection)
Year 1 after for-cause biopsy	5.5
Year 2	5.6
Year 3	5.8
Year 4	5.9
Year 5	6.1
Annual graft failures (no rejection)
Year 1 after for-cause biopsy	4.0
Year 2	4.1
Year 3	4.2
Year 4	4.3
Year 5	4.4
Kidney replacement therapies after graft failure
Hemodialysis after graft failure	84	^23
Peritoneal dialysis after graft failure	6
Second transplant after graft failure	10

Costs

Cost estimates following for-cause kidney biopsies include: (1) the cost of MMDx-Kidney, (2) the cost of antirejection treatment in all settings, and (3) the average annual costs associated with each of the clinical outcomes, including renal replacement therapy (dialysis or re-transplant).

Cost of MMDx-Kidney test

The cost per MMDx-Kidney test was assumed to be $3,159.42 based on Medicare reimbursement rates for CPT Code 0088 U: “transplantation medicine (kidney allograft rejection), microarray gene expression profiling of 1,494 genes, utilizing transplant biopsy tissue, algorithm reported as a probability score for rejection”, as published in the CMS CY 2022 Clinical Lab Fee Schedule Final Payment Determination^25,26.

Cost of outpatient antirejection treatment

The outpatient costs for each antirejection treatment, including administration costs, were estimated using the CMS FFS schedule and based on the recommended dosage and length of treatment for each medication (Table 2). The total Medicare costs were then multiplied by an adjustment factor of 126% to reflect the approximate amounts incurred by commercial payers relative to Medicare²⁷.

Cost of inpatient antirejection treatment

For patients treated in the hospital inpatient setting, Medicare costs associated with three Diagnosis-Related Group (DRG) codes for “other kidney and urinary tract diagnoses” (698, 699, 700) were identified using CMS Fiscal Year (FY) 2021 Final Rule Tables and the CMS ICD-10-CM/PCS MS-DRG Definitions Manual^28,29. All three DRG codes map to the International Classification of Disease (ICD-10) CM T.86-11 code for the diagnosis of kidney transplant rejection in general; the ICD-10 code is not specific to the type of rejection (ABMR versus TCMR versus mixed). Payment distribution across the three DRG codes was calculated using hospital claims data from the Definitive Healthcare database. The weighted average of the Medicare costs was multiplied by a factor of 189% to reflect the amounts incurred by commercial payers³⁰.

Cost of outcomes

Table 4 reports the annual costs of each outcome following a for-cause biopsy. Data from USRDS models were used to estimate annual costs for surviving patients within each outcome group. These models calculate costs from ESRD Medicare inpatient, outpatient, skilled nursing facility, hospice, home health, physician/supplier, durable medical equipment, and Part D claims data from 2018. To estimate Medicare costs associated with death with graft function and death after graft failure, 2014 economic data from the OPTN/SRTR database were used³¹. To reflect commercial payer costs, the Medicare costs were multiplied by an adjustment factor of 146%, based on a weighted average of inpatient and outpatient adjustment factors^27,30. All costs were inflated to 2021 costs³².

Table 4. Annual costs of outcomes following for-cause kidney biopsy.

Outcome	Cost to payer^a	References/Notes
Functioning transplant: patients who have a functioning kidney transplant	$44,609	^3
New hemodialysis: patients who continue to hemodialysis in year of kidney graft failure^b	$203,742
New peritoneal dialysis: patients who continue to peritoneal dialysis in year of kidney graft failure^b	$172,150
Second transplant: patients who continue to new transplant in year of kidney graft failure	$216,961
Continuing hemodialysis: patients receiving hemodialysis after year of kidney transplant failure	$147,911
Continuing peritoneal dialysis: patients receiving peritoneal dialysis after year of kidney transplant failure	$124,976
Functioning second transplant: patients with new transplant after year of kidney transplant failure	$44,609
Death with graft function: patients who die with functioning graft	$210,033	^30
Death after graft failure: patients who die after graft failure	$288,589

^aAll costs have been adjusted to commercial values and inflated to reflect amounts in 2021.

^bCalculated from overall cost of new dialysis from the United States Renal Data System database.

Sensitivity analysis

Multiple one-way sensitivity analyses were conducted using a relative change of ±10% from the base-case values of model input parameters. Parameters that demonstrate the highest impact on base case values are presented.

Results

Base case budget impact analysis

The average antirejection treatment costs for biopsies diagnosed with rejection with the standard of care versus MMDx-Kidney plus the standard of care are provided in Table 5. In the base case scenario, MMDx-Kidney increases total rejection treatment costs by $1,126 per test as a result of identifying more patients with ABMR and mixed rejection in the first year of the model.

Table 5. Antirejection treatment costs per MMDx-Kidney test.

Diagnosis	SoC (histology)	SoC and MMDx-Kidney	Savings per MMDx-Kidney test^a
ABMR	$2,633	$3,949	($1,316)
ABMR suspicious and TG	$165	$99	$66
TCMR	$645	$564	$81
Suspicious (borderline) for TCMR	$426	$53	$373
Mixed rejection	$494	$823	($329)
Total rejection treatment	$4,362	$5,488	($1,126)

^aItalics indicate MMDx-Kidney is cost-additive.

Figure 2 presents the savings with MMDx-Kidney by year in a 5-year model, given the assumption that MMDx-Kidney may reduce annual graft failure from rejection by 25% after a for-cause biopsy, in relative terms. In year one, the costs increase by $3,159 per patient for MMDx-Kidney, $1,126 per patient for anti-rejection treatment, and $613 per patient for the management of functioning transplants, while savings of $2,509 and $283 per patient occur because of fewer dialysis cases and fewer repeat transplants, respectively. From years two through five, overall savings are observed as a result of the reduced rates of graft failure; the decrease in the costs associated with patients receiving renal replacement therapy (dialysis and repeat transplants) outweighs the increase in costs from additional patients with functioning kidney transplants.

Figure 2. Savings per MMDx-kidney test by year.

Overall, assuming one MMDx-Kidney test per patient, MMDx-Kidney results in an estimated undiscounted savings of $19,271 per test over a 5-year period, with positive savings starting with models of a 2-year duration (Table 6).

Table 6. Savings per MMDx-Kidney test with different model durations, assuming a 25% annual Impact of MMDx-Kidney on graft failure from rejection.

Time frame of model	Savings per MMDx-Kidney test^a
1 Year	($2,106)
2 Years	$1,339
3 Years	$6,050
4 Years	$12,028
5 Years	$19,271

^aItalics indicate MMDx-Kidney is cost-additive.

Over a 5-year timeframe, reductions in costs associated with graft failure (i.e. costs of hospitalizations, dialysis, and repeat transplants) completely offset the increased costs of antirejection treatment, functioning transplants, and a single-use of MMDx-Kidney per patient, assuming MMDx-Kidney reduces annual graft failure from rejection by at least 5%. This break-even point is illustrated in Figure 3.

Figure 3. Break-even analysis: relationship between overall savings per MMDx-Kidney test over 5 years and MMDx-Kidney’s impact on graft failure from rejection. The orange line indicates the minimum percentage in graft failure reduction that will offset increased costs associated with MMDx-Kidney. The blue, line represents the savings per test at increasing percentages of reduction in graft failure from rejection with MMDx-Kidney.

Scenario analysis: number of MMDx-Kidney tests

If the model assumes MMDx-Kidney can be used multiple times per patient, one could potentially use MMDx-Kidney six times over 5 years before the cost of MMDx-Kidney ($3,159.42) eliminates the estimated savings from the test, assuming a 25% reduction in annual graft failure from rejection with MMDx-Kidney. This simplified assessment assumes MMDx-Kidney neither further improves the patient’s clinical trajectory beyond the 25% impact nor reduces the need for additional for-cause biopsies. This analysis also does not include the cost of the for-cause biopsy, as all costs are measured after the first for-cause biopsy. In this simplified scenario, the model continues to assume that each patient diagnosed with rejection receives one set of antirejection treatments over the 5 years of the model and that patients who start a certain type of dialysis after a graft failure do not switch to another type of dialysis in the time frame of the model. The model also continues to assume that patients with graft failure who start dialysis do not receive a second transplant in the time frame of the model.

Sensitivity analysis

Results from the one-way sensitivity analysis indicate the parameters with the greatest impact on savings are the following: the annual cost of maintaining hemodialysis in the years after graft failure, the cost associated with hemodialysis in the year of graft failure, the percentage of patients receiving hemodialysis after graft failure, the annual cost of a functioning initial transplant, and the rate of graft failure from rejection in the year after for-cause biopsy. The absolute value of the changes in savings per MMDx-Kidney test over 5 years range from $724 to $1,717 with these inputs (Figure 4).

Figure 4. Sensitivity analysis: top five input parameters with the most impact on the change in savings per test. ^aThe change in savings per MMDx-Kidney test is defined as the difference between (a) the savings per test over 5 years between the MMDx-Kidney plus the standard of care and the standard of care alone in the base case and (b) the savings per test over 5 years between MMDx-Kidney plus the standard of care and the standard of care alone with the adjusted input parameter defined in the sensitivity analysis. ^bContinuing hemodialysis cost: annual cost to payers for patients receiving hemodialysis after the year of kidney transplant failure. ^cNew hemodialysis cost: annual cost to payers for patients who continue to hemodialysis in the year of kidney graft failure. ^dHemodialysis percentage: percentage of patients with graft failure who receive hemodialysis. In this analysis, the percentage of patients receiving peritoneal dialysis remains constant; the change in hemodialysis percentage directly impacts the percentage receiving a second transplant after graft failure. ^eFunctioning transplant: annual cost to payers for patients who have a functioning kidney transplant. ^fRate of graft failure from rejection: standard of care rate of graft failure from rejection in year 1 after the for-cause biopsy. In this analysis, the model applies MMDx-Kidney’s relative percentage improvement in graft failure in the base case (25%) to this adjusted rate.

Discussion

MMDx-Kidney in conjunction with the standard of care (histopathologic evaluation of kidney biopsies) produces savings to commercial payers within 2 years because of its ability to clarify rejection classifications and reduce costs associated with rejection-related graft failure, including re-transplants and dialysis. Although treatment of active ABMR may not always result in complete and prolonged response, several recent studies have provided evidence that MMDx-Kidney improves clinical decisions and avoids over-treatment with immunosuppression therapy^9,21,33–36.

While MMDx-Kidney’s 25% impact on the annual rate of graft failure from rejection is an assumption in the base case model, a break-even analysis indicates that use of MMDx-Kidney will generate positive savings over 5 years, provided MMDx-Kidney delivers a minimum of only a 5% reduction in annual graft failure resulting from rejection. By reducing graft failure rates and the need for subsequent dialysis or second transplant, MMDx-Kidney lowers the overall costs commercial payers incur within 2 years of the initial for-cause biopsy. Overall, the savings per MMDx-Kidney test over 5 years is estimated to be $19,271 in the base case model.

This study is the first to estimate the economic impact associated with the use of MMDx-Kidney, and the study’s principal strength is the use of well-documented assumptions derived from peer-reviewed literature and validated by leading clinical experts.

There are nevertheless several limitations to our study. First, the model may not capture the complete range of events in the care pathway for kidney transplant patients undergoing for-cause biopsies. To address this limitation, clinicians reviewed input parameters that were derived from a literature review, and appropriate sensitivity analyses were performed to assess the impact of uncertainty on interpretation of results. Second, given the paucity of longitudinal testing and outcomes data, the model reflects one MMDx-Kidney test used in conjunction with a for-cause biopsy, followed by one set of antirejection treatments per patient diagnosed with rejection (i.e. one diagnosis per patient, with no additional for-cause biopsies). Although MMDx-Kidney can benefit patients who receive multiple for-cause biopsies⁹, the literature suggests the number of for-cause biopsies per kidney transplant patient is low in the 5 years following for-cause biopsy. For example, the total number of for-cause kidney biopsies at a transplant center in Belgium was 1,443 over a 20-year follow-up period, while 752 patients received at least one for-cause biopsy, suggesting each of these patients received 1.9 for-cause biopsies, on average³⁷. Similar rates of for-cause biopsies per patient can be derived from French studies^38,39. Third, the model does not explicitly include the cost of adverse events related to antirejection medication. A literature search for possible adverse events associated with antirejection treatment identified heterogeneity in adverse event reporting, low adverse event rates in general, and a lack of clear control data^40,41. Consequently, MMDx-Kidney’s net impact on medication-related adverse events is unclear at this time; adverse events may rise with an increase in ABMR treatment following the use of MMDx-Kidney, while other adverse events may decrease as physicians are able to identify more accurately those patients who would not benefit from treatment. Fourth, the analysis does not account for indirect costs, quality-of-life, and impacts to patients diagnosed without rejection; consequently, the values presented likely reflect a conservative estimate of the impact of MMDx-Kidney^42,43. For example, when MMD-Kidney discounts histologic rejection, expensive immunosuppressive drugs may be avoided and quality of life for patients may improve.

As more data become available, further research is needed to evaluate the economic impact of MMDx-Kidney on diagnosing injury and on immunosuppressive management. Additional research can be conducted to demonstrate the overall reduction in mortality and graft failure, including graft failure from causes other than rejection, associated with the use of MMDx-Kidney. It will also be helpful to evaluate the economic impact of MMDx-Kidney in the context of new treatments such as tocilizumab, which may become more widely used.

Conclusion

While a mainstay in the evaluation and management of kidney transplant dysfunction, histopathology based on for-cause biopsies provides insufficient information on the true state of kidney transplant grafts. The diagnostic uncertainty is further exacerbated by limitations in evaluation criteria and treatment guidelines, which increase the burden of management of suspected graft rejection on physicians and patients. Correct diagnosis of the underlying cause of observed graft dysfunction is critical for optimal immunosuppressive management, lowering the risk of graft failure and its corresponding costs.

By integrating machine learning with 1,494 unique gene profiles, MMDx-Kidney provides an opportunity to improve patient management, particularly for lesions classified by histology as ambiguous. Through more accurate diagnosis of rejection, MMDx-Kidney may facilitate better treatment options and reduce the total costs incurred by payers for post-transplant graft failure and the need for subsequent dialysis or replacement transplant. In sum, this budget impact model demonstrates MMDx-Kidney can be an effective complement to histology when performed in patients with suspected kidney dysfunction receiving a for-cause biopsy after transplant.

Transparency

Declaration of funding

This study was funded by a research grant from Thermo Fisher Scientific.

Declaration of financial/other relationships

The analysis was funded by Thermo Fisher Scientific Corporation; one co-author (CL) is currently employed by Thermo Fisher. Boston Healthcare Associates, Inc. (now a Veranex company) received consulting fees from Thermo Fisher for developing the model on which this manuscript is based. Co-authors LF; SM; SFM; and TFG were employed by Boston Healthcare Associates, Inc. (now a Veranex company) at the time of this research.

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgements

No assistance in the preparation of this article is to be declared.

Data availability statement

The datasets generated during and/or analyzed during the current study may be available from the corresponding author on reasonable request.