Context is Everything

Multiorgan transplant remains a vexing problem. National policies must be liberal enough to protect patients from harm due to dialysis after primary organ transplant, yet restrictive enough so as not to disadvantage patients waiting for kidney transplant alone. The 2017 simultaneous liver kidney (SLK) guidelines were heralded as a reasoned approach to ensure equipoise in abdominal transplantation.¹ The logic is simple, just because a kidney can be allocated to a liver transplant recipient doesn’t mean it ought to be based solely on physician preference. Rather, data would provide guardrails this allocation and allow for a more rational distribution of kidney allografts among the growing number of patients who may benefit from them. Such a simple policy change, however, has had myriad effects-both measured and unmeasured. Indeed, challenges remain in optimizing this policy for all patients with liver and kidney disease, and also extending its guidance to thoracic organs.²

One of the challenges is identifying patients inadvertently disadvantaged by the 2017 SLK policy, and patients with sustained acute kidney injury (AKI) may be one such group. Tanaka et al examined this question with a retrospective analysis of OPTN data undertaken to determine the effects of the 2017 SLK policy change on outcomes of patients with liver failure and sustained AKI.³ The authors utilized a dual approach of instrumental variable analysis and propensity scoring in an attempt to estimate the effect of the 2017 allocation policy on post-transplant survival of a variety of sub-populations. They designated the date of listing (before vs after national policy) as an instrumental variable in an attempt to account for practice-related factors and combined this method with propensity score matching to construct similar cohorts for comparison. The authors specifically looked at race, age and sex as drivers of differential policy benefit and examined the impact of a race-neutral estimated GFR.

The analysis revealed no evidence of significant overall decrease in posttransplant survival among most patients with liver failure. In fact, fewer than 10% of patients with sustained AKI who underwent LTA required subsequent kidney transplant at 1 year. Yet, Black patients who were no longer eligible for SLK after the 2017 policy were found to have significantly diminished survival which appear to reflect the use of a race-based estimation of renal function. Their analysis revealed that use of a race neutral formula would have allowed for SLK listing for 66 additional Black patients and eliminated the race-based difference in mortality. There were also non-statistically significant trends toward a survival decrements for women and older adults (age ≥ 60 years) who were only eligible for LTA under the new criteria.

The authors here demonstrate that a policy change that, on its face, was race neutral, indeed had disparate impacts by race. While convincing, it is important to note that the authors face limitations in their analysis. Instrumental variable methods are used to estimate causal relationships consistently when explanatory variables are correlated with error terms and minimize the chances of their findings being due to confounding. However, it is impossible to determine which patients listed in the post-policy time period (for whom an SLK would theoretically improve survival) would have received that kidney in a pre-2017 environment. Though the authors use propensity scores to account for this, these are limited in that they assign a probability of receipt of SLK while in actuality receipt of a second organ is a binary outcome. Although potentially disserved by sample size, the question of accurate estimation of renal function is salient in these populations as well. More generally, all models are limited by their assumptions and covariates. There may be unmeasured but vitally important variables that further explain some of the variation in the data.

Overall, the findings support the idea that AKI does not necessarily mandate listing for SLK. They also add to the published literature reporting that the 2017 SLK criteria reduced the number of SLK nationally without creating a commensurate rise in post-LTA mortality.⁴ Yet, fundamental questions related to patient selection for multiorgan transplant remain to be answered. These include identifying which forms of renal dysfunction are reversible in patients with end stage organ disease, and discerning which clinical strategies can be employed to extend native renal function before, and mitigate renal injury during and speed renal recovery after transplant. Importantly, the interaction between time to renal recovery and decreased survival due to dialysis may be different for heart, lung and liver recipients and these differences should be reflected in national policy. Finally, this study is a word of caution with regard to constructing clinical-or national-practice in the absence of societal or cultural context. The various layers of inequity can be difficult to parse and any attempt to rectify one wrong may inadvertently create another, particularly when predicated upon unmeasured assumptions and imperfect systems. As nearly all complex modern systems fit this categorization, any attempt at improvement must intentionally be tested against the assumption that it does not privilege one group against another or, even worse, consolidate that privilege among an already favored group.

Abbreviations

AKI

acute kidney injury

OPTN

organ procurement and transplantation network

SLK

simultaneous liver kidney transplant

LTA

liver transplant alone