Optimal Patient Selection for Simultaneous Heart-Kidney Transplant

A Modified Cost-Effectiveness Analysis

Brian Wayda; Xingxing S. Cheng; Jeremy D. Goldhaber-Fiebert; Kiran K. Khush

American Journal of Transplantation. 2022;22(4):1158-1168.

Abstract and Introduction

Abstract

Increasing rates of simultaneous heart-kidney (SHK) transplant in the United States exacerbate the overall shortage of deceased donor kidneys (DDK). Current allocation policy does not impose constraints on SHK eligibility, and how best to do so remains unknown. We apply a decision-analytic model to evaluate options for heart transplant (HT) candidates with comorbid kidney dysfunction. We compare SHK with a "Safety Net" strategy, in which DDK transplant is performed 6 months after HT, only if native kidneys do not recover. We identify patient subsets for whom SHK using a DDK is efficient, considering the quality-adjusted life year (QALY) gains from DDKs instead allocated for kidney transplant-only. For an average-aged candidate with a 50% probability of kidney recovery after HT-only, SHK produces 0.64 more QALYs than Safety Net at a cost of 0.58 more kidneys used. SHK is inefficient in this scenario, producing fewer QALYs per DDK used (1.1) than a DDK allocated for KT-only (2.2). SHK is preferred to Safety Net only for candidates with a lower probability of native kidney recovery (24%–38%, varying by recipient age). This finding favors the implementation of a Safety Net provision and should inform the establishment of objective criteria for SHK transplant eligibility.

Introduction

Kidney dysfunction is a common sequela of end-stage heart failure and afflicts many patients awaiting heart transplant (HT). Kidney function may recover after heart transplant alone (HT-only), particularly when acute and mild or moderate in severity.^[1] More chronic and severe kidney dysfunction may not be reversible after HT-only and often prompts consideration of simultaneous heart-kidney (SHK) transplant, the rates of which have doubled over the last decade.^[2]

This rise in SHK transplants has been facilitated by the current allocation system, in which all SHK candidates are prioritized above all candidates waiting for a kidney transplant alone (KT-only), with no standard criteria for SHK eligibility.^[3] In contrast, those with persistent kidney dysfunction after HT-only receive no such priority and, in the absence of a living donor, face the same expected wait time for a deceased donor kidney (DDK) as KT-only candidates.^[4] Such a policy produces an obvious incentive to favor SHK over HT-only whenever the reversibility of an HT candidate's kidney dysfunction is in question.

A similar situation existed in liver transplantation until a new allocation policy for simultaneous liver-kidney (SLK) transplant was introduced in August 2017.^[5] The SLK policy established a set of objective eligibility requirements for SLK and a "Safety Net" system, in which patients exhibiting irreversible kidney failure within the first year of liver transplant gain priority access to DDKs. Consistent with the results of a prior modeling study,^[6] this policy reduced kidney utilization for SLK, particularly among those with mild or moderate kidney dysfunction, with no adverse impact on kidney graft outcomes^[7] or liver transplant recipient survival.^[8]

From the standpoint of an individual SHK candidate, the incentive favoring SHK is harmless and potentially beneficial.^[9–14] But on a system-wide level, the liberal utilization of SHK transplant comes at a substantial cost. Each kidney used in the over 200 SHK transplants performed annually^[3] is one fewer available to the over 100,000 patients waiting for KT-only,^[4] for most of whom a kidney transplant will improve life expectancy and quality of life.^[15,16] Those KT-only candidates who "lose out" on a DDK due to its allocation instead for SHK suffer measurably worse outcomes.^[17]

The Final Rule of transplantation^[18] states that each organ should be allocated on the basis of medical judgment and not systemic biases.^[19] The allocation of a DDK to an HT recipient with reversible kidney dysfunction, and thus unclear benefit, ahead of a KT-only candidate with higher expected benefit reflects a systemic bias that violates this principle. To stem such "overuse" of kidneys for SHK, many have advocated for standardized SHK eligibility requirements^{[3,14,17,20–24]} and an accompanying Safety Net provision.^[20–22]

No studies have provided quantitative support for this policy. As a randomized controlled trial would be infeasible, we apply a decision-analytic model to quantify the trade-off between costs and benefits of SHK and to identify subgroups for which SHK is optimal from a societal standpoint.

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table 1. Model parameterization for base case and sensitivity analyses
Parameter	Base case value (plausible range^g)	Derivation/reference
Epidemiologic parameters
SHK benefit ratio^a	0.56 (0.45–0.80)	Represents a range of estimates from prior studies measuring the survival benefit of SHK. Base case, lower bound, and upper bound values are derived from Chou et al (2018),⁹ Gill et al (2008),¹² Schaffer et al (2014),¹³ respectively.
Probability of early^b death after SHK	9.0% (6.0–12.0%)	Reflects observed mortality after adult SHK transplants (2014–2018; n = 770)
Probability of early death after Safety Net KT	6.0% (4.0–8.0%)	Reflects observed mortality after adult kidney transplants (August 2017–February 2019; n = 370) occurring within one year of liver transplant; rationale is further detailed in Supporting Information (S4).
Relative risk of early death after re-transplant, compared to initial transplant of same organ	1.37 (1.25–1.49)	Barghash and Pinney (2020)³²
Probability of early graft failure after HT^c	0.3% (0.2–0.4%)	Reflects observed heart graft failure in a cohort including all (1) SHK transplants (2) HT among patients on dialysis pre-transplant (3) HT among patients with creatinine ≥2.3 pre-transplant (2014–2018; n = 1064)
Probability of early graft failure after kidney transplant^c	2.1% (1.4–2.8%)	Reflects observed kidney graft failure in a cohort including all (1) SHK transplants (2) kidney transplants occurring within one year of HT (2014–2018; n = 872)
Expected survival duration (years), conditional upon survival to 6 months post-transplant, by state^d
Thriving state (H+K+), average patient	17.7 (13.9–22.5)	Derived from observed post-HT survival among HT recipients surviving to 6 months post-transplant, as detailed in Methods. Sensitivity analyses were performed assuming a plausible range for H+K+ survival of 13.9 to 22.5 years, corresponding to old and young patients, respectively. Noting the small sample used to derive H+K- survival and resulting uncertainty, we assumed an especially wide plausible range (±67%).
Thriving state, young^e patient	22.5
Thriving stage, old^e patient	13.9
On long-term dialysis (H+K−)	3.5 (2.0–5.0)
In end-stage heart failure (H−)	0.25 (0.1–0.5)	Derived from reported mortality incidence after delisting from the HT waitlist due to clinical deterioration (VanderPluym et al 2014).³³ Given absence of data specific to the post-HT population, we assume a wide plausible range.
Quality of life parameters
QoL weight in thriving (H+K+) state	0.76 (0.68–0.84)	Based on Emin et al (2016),³⁴ Sharples et al (2006),³⁵ Saeed et al (2008),³⁶ Buendia et al (2011).³⁷ Given consistency of estimates across multiple studies, we employed a narrower (±10%) plausible range.
QoL weight in dialysis-dependent (H+K−) state	0.52 (0.44–0.60)	Dominguez et al (2011),³⁸ Held et al (2016)³¹
QoL weight in end-stage heart failure (H−) state	0.35 (0.30–0.40)	Holland et al (2010)³⁹
% decrease in QoL in first 6 months post-transplant^f	20% (13–27%)	Based on Butler et al (2003)⁴⁰; given the small sample size in this study, we employ a wider (±33%) plausible range.

Abbreviations: HT, heart transplant; KT, kidney transplant; OR, odds ratio; QoL, quality of life; SHK, simultaneous heart-kidney transplant.
^aRefers to the odds ratio: [odds of early death among SHK recipients]/[odds of early death among HT-only recipients with no renal recovery after HT]; defined in this way, a lower value for this parameter represents a greater benefit to SHK, in terms of mortality reduction.
^b"Early" death and graft failure refer to events occurring in the first 6 months post-transplant.
^cConditional on survival to 6 months post-transplant.
^dListed values refer to duration of survival beyond (i.e., not including) the first 6 months post-transplant.
^e"Young" and "old" refer to HT recipients in the lowest and highest age tertiles, respectively.
^fRefers to the decrement in QoL weight compared to that in the equivalent state after 6 months post-transplant.
^gPlausible range is assumed to be ±15% for quality of life parameters and ±33% for all other parameters, unless otherwise indicated.

Table 2. Outcomes of SHK compared to Safety Net strategy at different levels of reversibility and life expectancy
	Reversibility
	Low (10%)	Moderate (25%)	High (50%)
Absolute difference in 1-year mortality (SHK vs. Safety Net)	−9.8%	−8.1%	−5.4%
Additional kidneys used (SHK vs. Safety Net)	+0.24	+0.37	+0.58
Expected QALYs gained:
Young (18–49 years)	1.41	1.17	0.77
Average age (50–59 years)	1.18	0.98	0.64
Old (≥60 years)	0.98	0.81	0.54
Modified ICER (QALYs/kidney):
Young (18–49 years)	5.89	3.17	1.31
Average age (50–59 years)	4.94	2.65	1.10
Old (≥60 years)	4.11	2.21	0.92

Note: Each outcome is expressed as the difference between SHK and Safety Net on a per patient basis. "Reversibility" refers to the probability that a patient will have kidney recovery after HT-only. "Young" and "old" refer to life expectancy among HT recipients in the lowest and highest age tertiles, respectively.
Abbreviations: HT, heart transplant; SHK, simultaneous heart-kidney transplant.