Implementing a Height-Based Rule for Allocation of Pediatric Donor Livers to Adults – A Liver Simulated Allocation Model (LSAM) Study

To the Editor:

In liver transplantation, adult women are more likely to die on the waitlist and less likely to receive a deceased-donor transplant compared to adult men.(1) Reasons for these disparities include: Incorporation of creatinine in the Model for End-Stage Liver Disease (MELD), candidate anthropometrics, and differences in access to transplant. While not enough to address all factors described, policy interventions aimed at improving donor-recipient size matching could be immediately implemented prior to more systematic allocation changes.

In an analysis of the Scientific Registry of Transplant Recipients (SRTR) database, weighting for candidate anthropometric measurements was found to increase waitlist mortality risks for women by 125.8%.(1) Candidate height has a disproportionate impact (even more so than MELD score) on a candidate’s success in receiving a transplant: In our previous analyses, we found that 166cm in adults is a height threshold below which waitlist mortality begins to increase.(2) Moreover, height-based interventions at this threshold would disproportionately benefit women as 72% of adult women and 9% of adult men on the waitlist had heights <166cm.(2) The allocation of pediatric donor grafts represents a potential area to implement this policy as adult women who received first-offers of pediatric grafts had similar waitlist mortality rates as men and more appropriate donor-recipient size-matching.(3)

Therefore, we hypothesized that a height-based rule for allocating pediatric grafts to adult candidates, after all pediatric candidates are exhausted, might help alleviate this sex-based disparity. To test this hypothesis, we used the Liver Simulated Allocation Model (LSAM) to simulate a height-based rule superimposed on the Acuity Circle allocation scheme.

Methods:

This study utilized SRTR data from 7/1/2013 through 6/30/2016 applied towards LSAM, version 2019. LSAM is the software package used by SRTR to perform event-based simulations of United States liver allocation.(4) The simulation time was three years with ten replications of each scheme. Settings were adjusted so that the total number of transplants remained similar under all simulations. Candidates whose disease trajectory was censored due to transplant had their trajectory imputed, to death or end of simulation, with those of similar candidates. Primary outcomes of interest were numbers of transplants and waitlist deaths.

To test the height-based rule for allocating pediatric grafts, we ran LSAM under two schemes:

1. Acuity Circles (AC): Current allocation scheme approved by OPTN in May 2019.

2. Acuity Circles+166cm (AC+166): Under this scheme, pediatric (<18 years old) donor livers were preferentially offered to adults (≥18 years old) with height <166cm prior to adults with height ≥166cm within each MELD score band (≥37, 33–37, 29–33, 15–29, <15) and 500 nautical miles after all pediatric (<18 years old) candidates were exhausted.

The AC+166 scheme maintains prioritization of children over non-Status 1 adults and only changes the order within which adults are prioritized. We define “gender gap” as the difference between the percentages of men and women who received transplants during the simulation period (number of transplants/number of total waitlist candidates).

Across the ten replications over three years of simulation, means for each primary outcome were calculated and presented unless otherwise stated. Results under AC+166 were compared to AC via paired t-tests. We considered p<0.001 as statistical significance as multiple tests were performed.

Results:

Each LSAM replication had an average 48,989 candidates: 36.4% were adult women. 71.8% of women and 9.8% of men had a height <166cm. Each LSAM replication had an average 21,972 donors, 10.0% were pediatric donors where our rule could be applied.

Compared to AC, AC+166 was projected to significantly increase the number of transplants in adults with height <166cm (5,742 vs. 5,844, p<0.001) and adult women (6,417 vs. 6,498, p<0.001). Conversely, the number of transplants in adults with height ≥166cm (12,370 vs. 12,261, p<0.001) and adult men (11,694 vs. 11,607, p<0.001) were projected to decrease significantly. Compared to AC, AC+166 resulted in no changes in the overall numbers of transplants for both adults (18,112 vs. 18,106, p=0.58) and children (1,843 vs. 1,841, p=0.70).

Compared to AC, AC+166 resulted in no statistically significant changes in waitlist deaths in adults with height <166cm (1,371 vs. 1,348, p=0.02), adult women (1,505 vs. 1,483, p=0.04), adults with height ≥166cm (2,510 vs. 2,525, p=0.03) and adult men (2,376 vs. 2,391, p=0.03). Compared to AC, AC+166 resulted in no changes in overall waitlist deaths for adults (3,881 vs. 3,874, p=0.44) and children (76 vs. 76, p=0.90).

In the AC scheme, the average numbers of transplants for adult women (6,417) and adult men (11,694) comprised 33.8% and 35.3% of all candidates, respectively, for a difference of 1.6%. In AC+166, adult women gained an average of 81 transplants while adult men lost an average of 87 transplants. These changes in the average numbers of transplants comprised 34.2% of adult women and 35.1% of adult men candidates under AC+166 for a difference of 0.9%. AC+166 eliminated 41.7% of gender gap in the percentages of adults who received transplants.

Conclusions:

We simulated a height-based rule on pediatric graft allocation to adults with a goal of closing disparities faced by adult women. While this does not address all factors contributing to sex-based disparities, this rule represents one that could be implemented immediately. In our AC+166 scheme, we narrowed the “gender gap” in transplant rates, by shifting size-appropriate grafts and increasing the number of women receiving transplants by 81. In practical terms, for women to close the “gender gap” or to achieve the average transplant rate for all adults, women would have to gain approximately 179 transplants. For women to match the transplant rates for adult men under the AC scheme, women would have to gain approximately 282 transplants. AC+166 effectively reduced this “gender gap” by an average of 41.7%: the percentage of women transplanted increased by 0.4% while that for men decreased by 0.3%.

Furthermore, our simulations showed AC+166 would result in no significant changes in waitlist deaths. While the percentage of women who received transplants did increase significantly, the lack of changes in waitlist mortality for women indicate that the pediatric grafts that end up being allocated to women may have taken place in areas with relatively low MELD scores. While this finding may be due to the maldistribution of pediatric transplant services in low MELD areas, this also confirmed that AC+166 is only a first step and further policy modifications may be necessary to have a substantial impact on waitlist mortality. Fortunately, our height-based rule was applied after pediatric candidates were exhausted to avoid harm to children, who were disadvantaged under region-based allocation.(5)

There are several limitations to this analysis. The SRTR data were from 2013 through 2016: Characteristics of patients who received transplants have shifted since 2016, resulting in a different case-mix of cirrhosis etiologies and demographics. Moreover, the LSAM accept/decline model was fit prior to AC and may not reflect current organ acceptance/rejection behaviors. Third, LSAM is not able to simulate split liver transplants, which is another avenue by which the gender gap could be closed that simultaneously helps children. By allocating “split-able” livers to pediatric candidates who only require the left lateral segment, the remaining right tri-segment would be a good match for smaller candidates who require a smaller donor liver. Lastly, the shift from region-based to AC allocation is projected to increase the proportion of pediatric grafts allocated to children from 46% to 77%.(5) Our height-based rule, therefore, applies for an even more limited number of pediatric grafts, limiting its potential impact.

Despite these limitations, our study demonstrates the utility of a height-based intervention to begin to achieve gender parity. Further modifications to AC, such as prioritization of adult donors <166cm for adult candidates <166cm within each MELD band, may close the gender gap even further.

Abbreviations:

MELD

Model for End-Stage Liver Disease

SRTR

Scientific Registry of Transplant Recipients

LSAM

Liver Simulated Allocation Model