Kidney after Liver Transplantation Matched-Pair Analysis: Are Kidneys Allocated to Appropriate Patients to Maximize Their Survival?

Abstract

Background:

Kidney after liver transplantation (KALT) is the best therapeutic option for patients with end-stage renal disease after liver transplantation (OLT). New allocation policies prioritize kidneys to patients in renal failure within the first year following OLT. There is little data on how kidney quality, measured by kidney donor profile index (KDPI), impacts KALT survival outcomes.

Methods:

The UNOS database was queried for adult KALT recipients from 1988–2015 and compared to their paired kidney transplant alone (KTA) recipients. 745 pairs were stratified into three KDPI sub-groups and compared patient survival, graft survival and death-censored graft survival among matched-paired recipients.

Results:

Overall, KTA recipients had superior patient and graft survival compared to the KALT group. KTA patient survival was superior for all three KDPI sub-group analysis. KTA graft survival was superior compared to KALT recipients of KDPI 21–85% kidneys. Inferior graft half-life was observed in KALT vs KTA recipients with KDPI 21–85% and >85%.

Conclusions:

From a utilitarian perspective, it is important that kidneys are allocated to recipients that are able to maximize their benefit from the full life of the organ. In KTA recipients, graft quality correlates directly to graft survival. However, in KALT patients receiving the matched-pair kidneys of the KTA recipients, patient mortality, rather than kidney quality, dictates graft survival significantly. As allocation practices continue developing, utilization of expanded criteria kidneys that better match anticipated patient and graft survival should be strongly considered to maximize the benefits of limited resources for the greatest number of patients.

Introduction:

Since Model for End-Stage Liver Disease (MELD) was introduced in 2002, the relative risk of new onset end-stage renal disease (ESRD) in post-orthotopic liver transplant (OLT) patients has increased 21% compared to the pre-MELD era. The risk is estimated to be between 15 to 25% among OLT recipients.^1–3 In addition, the 5-year cumulative incidence of kidney failure, either ESRD or stage 4 chronic kidney disease, has recently been estimated to be as high as 18–22%.⁴ This incidence of ESRD among liver transplant recipients is the second highest among all non-renal solid organ transplants and is associated with high mortality rates.⁴

Unfortunately, as a by-product of current allocation schemes, there is a nation-wide increase in the average MELD score at the time of transplantation. As a consequence, patients wait longer on the transplant waiting list, increasing their risk for developing significant renal dysfunction and requiring renal replacement therapy (RRT). We have observed an increase in the number of patients in need of a simultaneous liver-kidney (SLK) or kidney after liver transplantation (KALT). This is expected to increase even further in years to come.^5,6

Renal failure associated with OLT can be present preoperatively or develop post-operatively. Potential contributors that have been identified as risk factors for the development of renal failure in cirrhotic patients include infections, volume depletion, hepatorenal syndrome (HRS), parenchymal nephropathy, calcineurin inhibitor-based immunosuppression, pre-transplant kidney dysfunction, peri-operative acute kidney failure, age, and hepatitis C.⁷ Liver transplant recipients with ESRD remaining on dialysis have a 2.5-fold increased risk of liver graft failure and a 3.6-fold increased risk of patient death compared to recipients of kidney transplants.^8,9

In June 2017, the policy for KALT allocation was updated to include specific “safey net” listing criteria to prioritize kidneys for patients who had a liver transplant and developed significant renal failure within the first post-operative year after liver transplantation.^10,11 The specific criteria include (a) requirement of dialysis for >60 days after liver transplantation or (b) continued renal dysfunction with a glomerular filtration rate ≤20 mL/min. However, recent changes in policy facilitate kidney allocation to patients that develop ESRD beyond the first year after OLT. This consideration could mitigate the high wait-list mortality for liver transplant recipients that develop kidney failure.^8,12 While it is clear that kidney transplantation offers superior survival over dialysis in patients who reach ESRD, how kidney quality impacts patient and graft survival after OLT remains unknown, as does the impact of this change on patients awaiting kidney transplants alone.^9–11

While the number of kidneys being transplanted has slowly increased in recent years, the demand for kidney transplantation has increased at a disproportionately higher rate.^13,14 With limitations on organ availability, allocation to maximize the benefit of each organ is imperative. Balancing utility, equity, and beneficence should be derived by critical analyses of the available evidence and translated to evidence-based policy that maintains an appropriate equilibrium.¹⁵

Analysis of the UNOS/OPTN database and SRTR registry has allowed us to summarize the national experience in regards to KALT and to establish hypotheses that cannot be answered with a single institutional database. To that end, we tested the hypothesis that KTA recipients will experience longer patient and graft survival over recipients of KALT. Secondly, we tested if KALT recipient patient and graft survival depends on the quality of the kidney allograft, similar to KTA patients.

Lastly, by systematically studying the impact of KDPI using a match pair analysis and comparing KALT recipients to KTA recipients with grafts from the same donor, we are hoping to understand how the projected-survival of the graft is affected by transplant into a recipient of a previous liver transplant. By analyzing the UNOS/SRTR registry our aim is that observations derived from this study will contribute to future policy changes that expand utilization of available grafts while establishing metrics for allocation appropriate to each graft to maximize the benefit of this limited resource.

Methods:

The UNOS/OPTN database was queried for adult KALT recipients from 1988–2015.¹⁶ Patients with prior liver transplants were included if they received their kidney after the age of 18 years and the matched kidney from the same donor was transplanted to a kidney alone recipient. From this group, patients were excluded if: a) they had any transplants prior to OLT, b) transplants other than KALT, and c) KDPI could not be calculated. These patients were retrospectively compared to their paired primary KTA recipients from the same donor group to evaluate patient survival, graft survival, and death-censored graft survival. The analysis was further stratified into three KDPI sub-groups: 1) KDPI <20%, 2) KDPI 21–85%, and 3) KDPI >85%. Since KDPI was introduced in December 2014, KDPI was retrospectively calculated using 2016 UNOS/OPTN guidelines for all patients included in the study.¹⁷ Estimated post-transplant survival (EPTS) score was calculated retrospectively using the 2018 guidelines.¹⁸ Matched-pair and Kaplan-Meier survival analysis were performed for both patient and graft survival. The log-rank test was used to compare survival among various cohorts. Cox proportional hazards regression analysis was performed to evaluate recipient characteristics using multivariable models and p<0.05 was considered statistically significant. All statistical analysis was performed using JMP Pro 2013 and SAS version 9.4 (Cary, NC).

Results:

Donor Characteristics:

Approximately 415,332 primary kidney transplants were performed between 1988–2015. Of these patients, 1,826 received a kidney transplant after OLT. After inclusion and exclusion criteria were applied, 745 donors were identified where one kidney went to a KALT transplant while the other went to KTA (Table 1). We used the same donor group to match the KTA and KALT recipient groups. KDPI was >85% in 10.5% of donor kidneys and the average KDPI was 58%. Average creatinine at the time of organ donation was 1.125 ± 1.11 mg/dL.

Table 1:

Kidney donor characteristics

All donors	N=745 (%)
Sex: Male	445 (59.7)
Age, y	42.8 ± 13.4
Race
White	569 (76.4)
Black	74 (9.9)
Hispanic	86 (11.5)
Other	16 (2.2)
Diabetes	54 (7.2)
Hypertension	192 (25.8)
Terminal creatinine, mg/dL	1.125 ± 1.11
BMI, kg/m2	27.7 ± 9.7
Hepatitis C-positive	120 (16.1)
CMV-positive	468 (62.8)
Cause of death
Anoxia	148 (19.9)
Stroke	277 (37.2)
Head Trauma	286 (38.4)
CNS tumor	9 (1.2)
Other	25 (3.4)
KDPI (average)	58%
Deceased after cardiac death	76 (10.2)

BMI: body mass index, CMV: cytomegalovirus, KDPI: kidney donor profile index

Recipient Characteristics:

Recipient demographics were compared between the KALT and KTA groups (Table 2). KALT group had a higher percentage of white recipients (75.8% vs. 45.9%, p<0.0001) and were older on average compared to KTA recipients (58 versus 54.2 years, p<0.0001). The overall prevalence of diabetes was similar between KALT and KTA recipients (p=0.1268). HCV and CMV were more common in the KALT group than the KTA group (36.2% and 70.6% vs. 17.3% and 61.4%, p<0.0001 and p=0.0005 respectively). Calcineurin toxicity was the most common reason for KALT recipients to require a kidney transplant (45.1%).

Table 2:

Kidney recipient characteristics

	KTA n=745 (%)	KALT n=745 (%)	P-value
Age at transplant, y	54.2 ± 11.96	58.04 ± 10.08	<0.0001
Sex: Male	485 (65.1)	512 (68.7)	0.1370
Race			<0.0001
White	342 (45.9)	565 (75.8)
Black	269 (36.1)	66 (8.9)
Hispanic	85 (11.4)	73 (9.8)
Other	49 (6.6)	41 (5.5)
BMI	28.09 ± 5.43	26.81 ± 5.07	<0.0001
Diabetes	292 (39.2)	321 (43.1)	0.1268
Hypertension	577 (86.1)	485 (48.5)	<0.0001
Hepatitis C-positive	126 (17.3)	255 (36.2)	<0.0001
CMV-positive	446 (61.4)	503 (70.6)	0.0005
Cold ischemic time, h	18.69 ± 9.10	18.37 ± 9	0.5091
Warm ischemic time, m	38.32 ± 24.51	36.98 ± 18.84	0.3498
Creatinine at transplant, mg/dL	8.32 ± 3.34	6.41 ± 2.52	<0.0001
Pretransplant dialysis	681 (91.4)	642 (86.3)	0.0012
Delayed graft function	194 (26)	143 (19.2)	0.0040
EPTS, %	39.62 ± 0.89	60.62 ± 0.89	<0.0001
Time on dialysis, y	3.31 ± 0.09	2.20 ± 0.09	<0.0001
Time on the wait list, y	1.96 ± 0.06	1.80 ± 0.06	0.1412
Time between OLT and KALT, y	N/A	8.32 ± 5.13
Diagnosis requiring kidney transplant			<0.0001
Diabetes	243 (32.62)	128 (17.18)
Hypertension	220 (29.53)	63 (8.46)
Cystic disease	49 (6.58)	19 (2.55)
Glomerulonephritis	86 (11.54)	65 (8.72)
Nephrotic syndrome	74 (9.93)	32 (4.30)
Calcineurin toxicity	N/A	336 (45.10)
Unknown/other	73 (9.80)	102 (13.69)

BMI: body mass index, CMV: cytomegalovirus, EPTS: estimated post-transplant survival, KALT: kidney after liver transplant

KTA: kidney transplant alone, OLT: orthotopic liver transplant

The EPTS score was 21% lower in the KTA group (39.62 ± 0.89%) vs the KALT group (60.62 ± 0.89%) (p<0.0001). Time on dialysis was 1.11 years longer in the KTA group vs the KALT group (p<0.0001). Time on the wait list was comparable between the two groups (p=0.1412). The average time between OLT and kidney transplants was 8.32 ± 5.13 years. KTA recipients had higher pre-transplant creatinine (8.3 vs. 6.4 mg/dL, p<0.0001) and were more likely than KALT recipients to undergo pre-transplant dialysis as well as develop delayed graft function (DGF), defined as dialysis in the first week after post-transplant (91.4% and 26% versus 86.3% and 19.2%) (p=0.0012 and p=0.004 respectively). Warm and cold ischemic times were comparable between the two groups (p>0.05).

Overall Patient Survival, Graft Survival and Death-censored Graft Survival:

Overall patient survival at 1, 5 and 10 years were superior in the KTA recipients (95.6%, 84.7%, and 72.2%) when compared to the KALT recipients (91.9%, 73.3% and 51.2%, p<0.0001) (Figure 1). Similarly, overall kidney graft survival was superior at 1, 5 and 10 years for the KTA group (91.8%, 71.8%, and 51.2%) compared to the KALT group (89%, 66.1% and 40.6%, p=0.0008). When overall graft survival was death-censored it was found to be comparable between the KTA and KALT recipients (p=0.5194).

Figure 1: Overall patient survival, graft survival and death censored graft survival.

Patient survival (a) was superior in the KTA group when compared to the KALT group at 1, 5 and 10 years (p<0.0001). The same was true for graft survival (b) (p=0.0008). Death-censored graft survival (c) was similar between the groups.

Patient Survival, Graft Survival and Death Censored Graft Survival by KDPI:

We compared KTA and KALT recipients stratified into three KDPI sub-groups: <20%, 21–85%, and >85% (Figure 2). Patient survival was superior for KTA vs. KALT for all KDPI groups. Patient survival for KTA recipients of KDPI <20% kidneys at 1, 5 and 10 years was 98.1%, 87.7% and 75% compared to KALT recipients with 94.4%, 79.2%, and 60.1% survival (p=0.0206). Similarly, for recipients of kidneys with KDPI 21–85% we observed that KTA had survival at 1, 5 and 10 years of 95.5%, 84.4%, and 71.6% compared to KALT recipients with 91.8%, 72.3%, and 50.3% survival (p<0.0001). In recipients of kidneys with a KDPI >85%, KTA recipients had 1, 5 and 10 year survival of 92.3%, 82.2%, and 72.4% vs. KALT recipients with 89.7%, 72.1%, and 45% survival (p=0.0059).

Figure 2: KALT vs. KAL patient survival, graft survival and DC graft survival by KDPI.

For KDPI <20%, KTA patients had superior patient survival (a), however graft survival (b) and death-censored graft survival (c) were similar between groups. For recipients of KDPI 21–85%, KTA patients had better patient (d) and graft survival (e), however death-censored graft survival (f) was similar. For recipients of kidneys with KDPI >85%, KTA patient survival was superior to KALT (g). Graft survival (h) and death-censored graft survival (I) were similar between groups.

Evaluation of graft survival revealed no statistical difference at 1, 5 and 10 years between KTA versus KALT for the KDPI <20% and >85% groups (p=0.1809 and p=0.0889). However, recipients of a kidney with a KDPI 21–85% in the KTA group had superior 1, 5 and 10 year graft survival (92.3%, 71.9%, 50.1%) compared to KALT (88.9%, 65.9%, 40.9%) (p=0.0068). Analyses of death-censored graft survival showed no statistical difference between KALT and KTA recipients at 1, 5 and 10 years for all KDPI groups (p=0.8990, p=0.4359, and p=0.6668 for KDPI <20%, 21–85%, and >85% groups respectively).

Matched-pair Analyses:

All KALT recipients were compared against their matched-paired KTA recipients from the same donor (Figure 3). Patient survival was 1.14 years longer for KTA vs KALT recipients in the KDPI <20% group, 1.25 years longer for KDPI 21–85%, and 1.54 years longer for KDPI >85% (p=0.0294, p<0.0001, p=0.0188 respectively). Graft survival was 0.589 years longer in KTA recipients when compared to KALT recipients for the KDPI 21–85% group (p=0.0090), however, there was no significant difference for the KDPI <20% or >85% groups (p=0.2012 and p=0.0638 respectively).

Figure 3: Matched pair analysis.

When recipients were matched by donor, KTA recipient patient survival was superior for all KDPI groups (p<0.05) (a – d). Graft survival for KDPI >20% and <85% was similar between groups (p>0.05) and superior in KTA recipients for the KDPI 21–85% (p=0.0090) (e – h).

Graft Half-life:

To determine renal graft half-life, we evaluated all patients who had a functioning kidney allograft at least one year after their kidney transplant (Table 3). Graft half-life was comparable for recipients of kidneys with a KDPI <20%. However, the half-life was significantly longer in the KTA group versus the KALT group for recipients of kidneys with a KDPI 21–85% and >85% (11.47 and 11.33 years vs. 8.98 and 6.47 years, p=0.0421 and p=0.0425 respectively).

Table 3:

Graft half-life

KDPI group	n	KTA	n	KALT	P-value
<20%	104	14.12	102	9.94	0.234
21–85%	516	11.47	497	8.98	0.0421
>85%	64	11.33	64	6.47	0.0425

KALT: kidney after liver transplant, KDPI: kidney donor profile index, KTA: kidney transplant alone

Cox Regression Analysis:

On multivariate analysis, several risk factors were identified that were associated with shorter patient and graft survival (Table 4). Major risk factors associated with shorter patient survival included recipient age (HR=1.037, p<0.0001), and previous liver transplant (HR=1.472, p=0.0185). Surprisingly, years on the wait list (HR=0.903, p=0.0249) and a diagnosis of cystic disease (HR=0.269, p=0.0293) requiring kidney transplant reduced the hazard ratio. Shorter graft survival was associated with increased years on dialysis (HR=1.057, p=0.0163) and inversely correlated with cystic disease requiring kidney transplant (HR=0.356, p=0.0109). Increased death-censored graft survival was associated with increasing age (HR=0.975, p=0.0004). Reduced death-censored graft survival was associated with longer CIT (HR=1.018, p=0.0160) and KDPI >85% (HR=2.284, p=0.0119).

Table 4:

Cox regression analysis for patient and graft survival

All recipients, n=1490	Patient survival			Graft Survival			Death-censored graft survival
Covariates	Hazards ratio	95% confidence interval	P-value	Hazards ratio	95% confidence interval	P-value	Hazards ratio	95% confidence interval	P-value
EPTS
Recipient age at kidney transplant	1.037	1.023–1.052	<0.0001	1.008	0.998–1.019	0.1212	0.975	0.926–0.989	0.0004
Previous liver transplant	1.472	1.067–2.030	0.0185	1.178	0.904–1.537	0.2256	0.989	0.661–1.478	0.9552
Years on dialysis	1.051	0.993–1.112	0.0884	1.057	1.010–1.106	0.0163	1.045	0.982–1.112	0.1618
Recipient diabetes	1.319	0.966–1.801	0.0818	1.274	0.969–1.674	0.0831	1.141	0.722–1.801	0.5727
Donor sex (male)	0.944	0.731–1.220	0.6616	0.929	0.749–1.153	0.5058	0.919	0.664–1.272	0.6109
Recipient sex (male)	1.292	0.979–1.706	0.0705	1.190	0.947–1.496	0.1359	1.088	0.777–1.524	0.6232
KDPI 21–85%	1.236	0.848–1.801	0.2699	1.174	0.868–1.588	0.2988	1.412	0.889–2.243	0.1435
KDPI >85%	1.285	0.762–2.169	0.3472	1.482	0.962–2.282	0.0742	2.284	1.200–4.346	0.0119
Years on wait list	0.903	0.826–0.987	0.0249	0.946	0.879–1.017	0.1324	0.963	0.864–1.073	0.4926
Donor CMV- positive	0.838	0.647–1.085	0.1791	0.892	0.718–1.107	0.3000	0.874	0.633–1.209	0.4166
Recipient Hepatitis C-positive	1.283	0.962–1.711	0.0903	1.102	0.861–1.410	0.4403	0.925	0.619–1.384	0.7052
Cold ischemia time	1.002	0.989–1.015	0.7970	1.009	0.999–1.019	0.0851	1.018	1.003–1.033	0.0160
Warm ischemia time	0.999	0.993–1.005	0.8183	0.998	0.993–1.002	0.3184	0.994	0.986–1.001	0.0849
Diagnosis requiring kidney transplant
Diabetes	1.016	0.679–1.520	0.9380	0.954	0.661–1.377	0.8016	1.051	0.559–1.976	0.8771
Hypertension	0.820	0.532–1.266	0.3712	0.924	0.643–1.326	0.6666	1.227	0.694–2.171	0.4819
Cystic disease	0.269	0.083–0.876	0.0293	0.356	0.160–0.789	0.0109	0.422	0.142–1.255	0.1207
Glomerulonephritis	0.683	0.397–1.178	0.1704	0.754	0.481–1.182	0.2182	0.993	0.510–1.935	0.9842
Nephrotic syndrome	0.695	0.359–1.345	0.2801	1.228	0.778–1.938	0.3789	1.738	0.906–3.334	0.0964
Unknown/other	0.937	0.601–1.461	0.7783	0.952	0.649–1.395	0.7993	1.145	0.635–2.067	0.6524

CMV: cytomegalovirus, EPTS: estimated post-transplant survival, KDPI: kidney donor profile index

Discussion:

Several studies in the literature evaluate the outcomes of SLK recipients compared to KTA recipients, but there is very little information on how KALT recipients perform post-operatively and compared to KTA recipients. The merit of our study lays on the fact that it is the first report evaluating long-term patient and graft survival of KALT recipients compared to their matched-pair KTA recipients. Overall, we found that KTA recipients have better patient and graft survival when compared to KALT recipients. Our analysis confirms findings by Gonwa et al. who previously reported longer patient and graft survival in deceased donor KTA compared to KALT recipients with similar death-censored graft survivals.¹ Their study was limited to 5 years of follow up, in which the impacts of quality of the graft on survival are not as apparent as those observed in our analysis at 10 years.

We compared patient and graft survival between KTA and KALT recipients stratified by KDPI (Figure 2). We found that patient survival was superior in KTA recipients when compared to KALT recipients for all KDPI groups. In terms of graft survival, we observed that KTA recipients had better graft survival in the KDPI 21–85% group, but performed similarly in the KDPI <20% and KDPI >85% groups. The interpretation of these findings are likely premature, since both the KDPI <20% and KDPI >85% groups had a much lower n-value (108 and 78 pairs of patients respectively) compared to the KDPI 21–85% groups (559 pairs of patients) and results may vary in the future when a larger group of patients is analyzed.

Likewise, for the death-censored analysis, all groups had similar graft survival. The differences in outcomes when graft survival was death-censored suggest that kidney grafts are surviving longer than their patients who received them. According to a study by Cassuto et. al, at 5 years post KALT, 64.3% of KALT recipients die with a functioning graft.¹⁹ These results are likely associated with the overall co-morbidities associated with recipients of previous liver transplantation, which offsets the benefit from less time on dialysis in the KALT recipients when compared to KTA recipients. A major factor that contributes to this is the immunosuppression required after an OLT. Calcineurin inhibitor toxicity was the most common cause of renal failure in the KALT group, accounting for almost half the KALT recipients (336 out of 745 patients).

The novelty of our study is in the matching of KALT recipients with the same donor KTA recipients (Figure 3), removing any bias that could be due to one group receiving better kidneys than the other. When we matched recipients by their donors, we observed that KTA recipients live longer than KALT recipients for all the different subgroup analyses based on KDPI and present significantly better graft survival in the KDPI 21–85% group. While trends toward better graft survival were observed in the KDPI <20% group and the KDPI >85% group, no statistical significance was observed. Final conclusions can only be derived when larger data for both sub-groups is analyzed in future studies. Comparing graft half-life (Table 3), KDPI 21–85% grafts lasted ~2.5 years longer for KTA recipients, and KDPI >85% kidneys lasted ~5 years longer, compared to KALT recipients. KDPI <20% kidneys lasted ~4 years longer in KTA recipients compared to KALT, but this was not statistically significant.

Lastly, we looked for common characteristics using Cox regression analysis that could be associated with our findings, including the components used to calculate their EPTS score. Similar to previously published data, we found that shorter patient survival was inversely-correlated with increasing age and having a prior liver transplant. Longer time on the wait list seemed to increase patient survival (HR=0.903) in this population. In our study, the time on the waitlist was similar between the two groups and less than two years. Hypothetically, we could envision having a priority score given to patients on the wait list, aiming to increase patient and graft survival in the KALT population. However, based on retrospective data, this would not be the case, since no difference in the KALT group was observed when comparison between those on the wait list for < 1 year versus > 1 year was performed (Supplemental figure 1). Shorter graft survival and death-censored graft survival was associated with longer time on dialysis, increased cold ischemia time and KDPI >85%^14,19–21.

There have been many statistical metrics designed to better assess trends in allocation outcomes and to compare alternative allocation systems among themselves including Life Years from Transplant, EPTS, and KDPI²². KDPI is currently used in kidney allocation and EPTS is used to allocate KDPI <20% kidneys. Both groups had an EPTS >20% and the KALT score was nearly 20% higher than in KTA, in part due to the previous liver transplant. Among the KDPI <20% group (Supplemental table 1), we observed an average EPTS of 34.69 ± 2.45 in the KTA group and 55.98 ± 2.52 in the KALT group (p<0.0001).

Since deterioration in graft function often precedes death-censored graft loss, our observations of correctable variables could be used as a signal to trigger interventions to ameliorate graft loss risk factors. Our analyses suggest that decreasing cold ischemia time and choosing kidneys with better KDPI would likely translate to improvement in graft survival for KALT recipients. However, an improvement in graft survival would be unlikely to improve patient survival in this population given the high number of KALT recipients who have a functioning graft at the time of death. This is similar to what has been seen in a matched-pair analysis performed by Choudhury et. al, which compared simultaneous heart-kidney transplant recipients to KTA recipients. They found that while there was no difference between 5-year graft survival, KTA recipients had a longer patient survival.²¹

The question of utility versus equity when determining organ allocation practices is an ongoing philosophical debate.^15,23,24 In gereral, to direct an organ to one patient means to take that organ away from another. This leads to the question, is it more important to get the optimal life out of each organ transplanted, or is it better to give preference to patients who have waited the longest or have a greater need / greater risk of dying? Allocation policies that favor patients at the greatest risk of dying may have unintended consequences for patients on the waiting list for a KTA. Systematic analysis of the data will allow us to develop and implement evidence-based allocation policy changes to mitigate these consequences.

To increase our resources would be ideal. With the current and the newly proposed OPTN liver allocation policy, in which acuity measured by MELD score is the determining factor to dictate a broader sharing of organs, trends toward transplanting patients with even higher MELD score is likely to occur. In our study, 45.1% of KALT recipients developed their initial renal failure secondary to immunosuppressive therapy. As the average MELD score at the time of transplantation increases and patients are waiting a longer period of time for their transplantation, we can expect a greater chronic kidney disease stage in that population, increasing the risk to worsen renal dysfunction, need for RRT, and the potential increase in the number of KALT in the future.

In a matched-pair analysis by Lloveras et. al, patients on dialysis were matched by their characteristics to recipients of kidneys donated by elderly deceased donors, who fall under the expanded donor criteria (ECD).²⁵ They found that receiving these ECD kidneys was superior to hemodialysis. This correlates to what has been found by Yunhua et. al, who compared dialysis in OLT recipients of both standard criteria and ECD kidneys.⁸ They found that patients remaining on dialysis had inferior patient survival and liver graft survival when compared to patients who received a KALT. They also observed no difference in death censored graft survival between ECD and SCD kidneys that in patients who received a KALT.

In combination with our data, we could hypothetically suggest to increasing the use of ECD kidney iutilization by prioritizing their use to those with matching marginal patient and graft survival for KALT recipients. While KDPI >85% kidney recipients do poorly when compared to KDPI <85% kidney recipients,²⁶ by utilizing higher KDPI kidneys in patients with prior liver transplants and shorter EPTS, we could allocate lower KDPI kidneys with a longer graft life expectancy to patients with a longer EPTS. Unfortunately, under the current system over half of ECD and KDPI >85% kidneys are discarded.^27,28

This study has multiple limitations. This is a retrospective study that covers long-term outcomes that may be affected by changes in allocation policies over time. The data used for this study was acquired from the UNOS/SRTR database and some data may be incomplete. KDPI was not implemented until 2014 and was calculated retrospectively based on the 2016 reference population, with additional allocation policy changes during this period. Specific patient medical records were not available and some patients were listed as lost to follow-up – we assumed that patients lost to follow up were alive with a functioning graft. Comparison of graft survival may be limited for the groups that received kidneys with a KDPI <20% and >85%. Lastly, based on limitations imposed in the design of the study, we were not able to compare recipients of kidneys with a KDPI >85% versus those that remain on the transplant waiting list.

In summary, our study evaluated long-term outcomes of KALT recipients compared to KTA recipients. This is the first study to use matched-pair analysis by kidney donor to evaluate long-term patient and graft survival. We demonstrated that patients who received a KALT have decreased long-term patient and graft survival when matched to the recipient of the other kidney. Patient survival in KALT recipients is significantly reduced by decreases in the quality of the kidney measured by KDPI. Additionally, graft survival was superior in KTA recipients compared to KALT recipients for KDPI 21–85% kidneys. Lastly, as allocation policies continue to improve, we need to reconsider the utilization, especially of ECD kidneys, and prioritization of organ allocation to maximize patient and graft survival in order to provide the most benefit of each organ to each patient.

Supplementary Material

Supplemental Figure 1. Supplemental Figure 1: Wait-list time <1 year versus >1 year.

When KALT and KTA patients on the waiting list for <1 year were compared to those on the waiting list for >1 year, we observed no difference in patient, graft or death censored graft survival (p>0.05).

Abbreviations:

ECD

expanded criteria donor

CMV

cytomegalovirus

EPTS

estimated post-transplant survival

ESRD

end-stage renal disease

GFR

glomerular filtration rate

HCV

hepatitis c virus

HRS

hepatorenal syndrome

KALT

kidney after liver transplant

KDPI

kidney donor profile index

KTA

kidney transplant alone

MELD

Model for End-Stage Liver Disease

OLT

orthotopic liver transplant

OPTN

organ procurement and transplantation network

RRT

renal replacement therapy

SLK

simultaneous kidney-liver

SRTR

scientific registry of transplant recipients

UNOS

united network for organ sharing