Outcomes After Declining Increased Infectious Risk Kidney Offers for Pediatric Candidates in the United States

Abstract

BACKGROUND:

Kidneys from donors at increased risk for disease transmission (IRDs) confer substantial survival benefit in adults, yet the benefit of IRDs to pediatric candidates remains unclear in the context of high waitlist prioritization.

METHODS:

Using 2010-2016 OPTN data, we studied 2417 pediatric candidates (age <18) who were offered an IRD kidney that was eventually used for transplantation. We followed candidates from date of first IRD offer until date-of-death or censorship and used Cox regression to estimate mortality risk associated with IRD acceptance vs decline adjusting for age, sex, race, diagnosis, and dialysis time.

RESULTS:

Over the study period, 2250 (93.1%) pediatric candidates declined and 286 (11.8%) accepted an IRD offer; 119 (41.6%) of the 286 had previously declined a different IRD. Cumulative survival among those who accepted versus declined the IRD was 99.6% vs 99.4% 1yr and 96.3% vs 97.8% 6yrs post-decision (P = 0.1). Unlike the substantial survival benefit seen in adults (HR = 0.52), among pediatric candidates we did not detect a survival benefit associated with accepting an IRD (adjustedHR:_0.791.72_3.73, P = 0.2). However, those who declined IRDs waited a median 9.6 months for a non-IRD kidney transplant (11.2 months among those <6yrs, 8.8 months among those on dialysis). KDPI of the eventually accepted non-IRD kidneys (median = 13, IQR = 6-23) was similar to KDPI of the declined IRDs (median = 16, IQR = 9-28).

CONCLUSIONS:

Unlike in adults, IRDs conferred no survival benefit to pediatric candidates, although they did reduce waiting times. The decision to accept IRD kidneys should balance the advantage of faster transplantation against the risk of infectious transmission.

INTRODUCTION

Kidneys labelled as increased infectious risk for disease transmission (IRD) now represent almost 20% of the deceased donor kidney pool, driven largely by the opioid epidemic.^1,2 The main risk of accepting an IRD kidney is the potential for disease transmission; in the era of nucleic acid testing (NAT), the highest estimated risks of transmission are observed with donors who were injection drug users at 32.4 per 10,000 for HCV and 4.9 per 10,000 for HIV.^3-6 On the other hand, accepting an IRD kidney potentially leads to earlier transplantation, and therefore less waitlist morbidity and mortality. Therefore, the decision to accept an IRD kidney, or decline and wait for another offer, should balance the risk of potential disease transmission with the benefit of earlier transplantation.

In adults, we have shown that this balance clearly favors accepting an IRD kidney, which was associated with a 48% lower risk of death compared to declining that IRD kidney and waiting.⁷ Within five years of declining the IRD kidney, only 31% of adult candidates received a non-IRD kidney transplant (with a median KDPI of 52 after having declined an IRD with median KDPI of 21), and 20% died while waiting for a non-IRD kidney. However, the balance associated with IRD acceptance for pediatric candidates might be different from adults for two major reasons. First, pediatric candidates are highly prioritized under the current allocation system, and may not wait as long as adult candidates for another offer if they decline an IRD kidney.⁸ Second, waitlist mortality rates for pediatric candidates in general are very low, and thus, declining an IRD kidney and waiting for even an extended time is unlikely to result in their death.⁹

Given these unique considerations surrounding pediatric candidates, we used national registry organ offer data to characterize outcomes after declining an IRD kidney, and to estimate the survival benefit associated with accepting an IRD kidney versus remaining on the list and waiting for a non-IRD kidney.

METHODS

Data source

This study used data from the Scientific Registry of Transplant Recipients (SRTR) December 2018 public release. The SRTR data system includes data on all donor, wait-listed candidates, and transplant recipients in the US, submitted by the members of the Organ Procurement and Transplantation Network (OPTN), and has been described elsewhere.¹⁰ The Health Resources and Services Administration (HRSA), U.S. Department of Health and Human Services provides oversight to the activities of the OPTN and SRTR contractors. This study used deidentified data and was exempted by the Johns Hopkins School of Medicine Institutional Review Board (NA_00042871).

Study population

We studied 2,417 pediatric kidney-only transplant candidates who were actively listed in the United States and offered an IRD kidney that was eventually used for transplantation for an adult or pediatric patient between 12/25/2009 and 12/3½016. Pediatric candidates who declined an IRD offer and later accepted another IRD offer are included in both columns of Table 1. Pediatric candidates who were offered the IRD kidney in error (ie, after removal from the list, n = 3) were excluded from the analysis. We compared demographic (age, sex, race, blood type) and clinical characteristics (primary diagnosis, time on the waitlist, time on dialysis, preemptive listing, and previous transplant) of the pediatric candidates who accepted versus declined IRD offers using χ² tests for categorical variables and Wilcoxon rank sum tests for continuous variables.

Table 1.

Characteristics of Pediatric Candidates Who Accepted or Declined an Increased Infectious Risk Kidney Offer Between December 2009 and December 2016.

Pediatric Candidate Characteristic	Declined IRD Kidney (N = 2250)	Accepted IRD Kidney (N = 286)	P-Value
Age at offer, median (IQR)	13.0 (7.0, 16.0)	15.0 (10.0, 17.0)	<0.001
Age, categories			<0.001
<6	20.5%	13.3%
6–10	16.8%	13.6%
11–14	22.7%	22.4%
15–17	31.2%	36.4%
≥18	8.8%	14.3%
Male sex	58.3%	57.7%	0.8
Race/ethnicity			0.9
Caucasian	39.0%	37.4%
African American	21.9%	23.8%
Hispanic/Latino	32.5%	33.6%
Asian	4.5%	3.8%
Other	2.1%	1.4%
ABO			0.5
O	54.8%	57.0%
A	32.8%	29.0%
B	9.1%	10.8%
AB	3.3%	3.1%
Primary indication for transplant			0.8
Glomerular	8.1%	8.7%
Recurrent	13.9%	15.7%
Pyelo/Interstitial Nephritis	4.1%	3.1%
Hereditary	9.4%	8.4%
Autoimmune	2.7%	3.1%
Oncologic	0.8%	0.3%
Structural	25.6%	28.0%
Ischemic	1.9%	1.0%
Hypertension	1.2%	1.7%
Other	32.3%	29.7%
Preemptively listed	31.8%	28.3%	0.2
Dialysis for >1 year	43.9%	51.0%	0.02
Months on dialysis, median (IQR)	9.5 (0.0, 21.8)	12.6 (0.0, 26.4)	0.02
Months on waitlist, median (IQR)	4.1 (1.3, 10.6)	6.6 (2.1, 15.8)	<0.001
Previously transplanted	12.5%	11.2%	0.5

Mortality risk associated with IRD acceptance for pediatric candidates

To estimate the survival benefit of accepting an IRD kidney offer for pediatric candidates, we compared mortality between those who accepted and were transplanted with an IRD kidney versus those who declined an IRD kidney and remained on the waitlist. Pediatric candidates were followed from the date of decision until date of death or administrative censorship on 4/30/2018, irrespective of subsequent transplants or removal from the waitlist. In other words, even if a pediatric candidate later received a non-IRD DDKT or LDKT, or was removed from the waitlist for any reason, we followed them until eventual mortality or administrative censoring. Pediatric candidates who declined an IRD kidney but later accepted a different IRD kidney had a time-varying exposure; these candidates were followed from initial IRD decline, censored at the date of IRD acceptance, and then included in the risk set of candidates who accepted an IRD. We used Cox regression to estimate the relative hazard associated with IRD offer acceptance versus decline after adjustment for candidate age, sex, race, diagnosis, previous transplant, and time on dialysis prior to the decision. To determine whether mortality risk associated with IRD acceptance varied by patient characteristics, we tested the interaction between IRD acceptance and candidate age, blood type, preemptive listing (vs on dialysis), and previous transplants.

Outcomes following the decision to decline an IRD offer

To characterize the potential consequences of turning down an IRD offer, we followed pediatric candidates from the date of their initial IRD decline until eventual non-IRD DDKT, IRD DDKT, LDKT, removal from the waitlist, death, or administrative censoring on 4/30/2018. We estimated the cumulative incidence of each potiential outcome six years after IRD decline, accounting for competing risks.

Non-IRD DDKT following IRD decline

We estimated median time to non-IRD DDKT following IRD overall and stratified by candidate age and preemptive listing. Among the pediatric candidates who declined an IRD kidney and later received a non-IRD DDKT, we compared the KDPI of the declined and later accepted kidneys.

Statistical analysis

Confidence intervals are reported as per the method of Louis and Zeger.¹¹ We used a 2-sided α level of 0.05 to indicate a statistically significant difference. All analyses were performed using Stata/SE, version 15 (StataCorp).

RESULTS

Study population

Of 2417 individual pediatric candidates offered an IRD kidney at some point during our study period, 2250 (93.1%) declined at least one IRD kidney and 286 (11.8%) accepted and were transplanted with an IRD kidney. Those who accepted an IRD kidney were less likely to be very young (2.5% vs 4.9% were <2 years old) and more likely to be older adolescents (14.3% vs 8.8% were listed while pediatric but ≥18 at the time of IRD offer) than those who declined. Those who accepted an IRD had spent more time on dialysis (12.5 months [interquartile range (IQR): 0.0-26.4] vs 9.5 months [IQR: 0.0-21.8]) and on the waitlist than those who declined (6.6 months [IQR: 2.1-15.8] vs 4.1 months [IQR: 1.3-10.6]) (Table 1) prior to the IRD offer.

Of the 286 pediatric candidates who accepted an IRD, 119 (41.6%) had previously declined one or more IRD offers during the study period. There were 3 candidates who accepted an IRD offer after ≥14 previous IRD offer declines. Candidates who declined and later accepted an IRD had been on the waitlist for a median (IQR) of 3.4 months (1.0-10.6) and on dialysis for 10.8 months (0.0-20.7) at their initial IRD decline and on the waitlist for 10.1 months (4.8-19.3) and on dialysis for 14.6 months (0.0-27.6) at their eventual IRD acceptance. In contrast, those who accepted their first IRD offer without a previous IRD decline had been on the waitlist for 3.7 months (1.2-11.9) and on dialysis for 11.0 months (0.0-26.4) at the time of IRD acceptance. Median follow-up time from the date of first IRD decline for all included pediatric candidates was 3.98 years (IQR: 2.55-6.07) (3.59 years [2.57-5.62] among those who accepted and 4.00 years [2.53-6.11] among those who declined).

IRD kidney offers

There were a total of 7018 IRD offers from 1953 IRD donors made to pediatric candidates during the study period, and 286 acceptances (4.1%) from 248 donors. The percent of IRD offers accepted for pediatric candidates fluctuated over time from 5.8% in 2010, to a low of 3.2% in 2013, and finally to 4.0% in 2016. Accepted IRD kidneys were recovered from donors with median age of 23 (19-27) and median KDPI of 13 (IQR 5-23). Kidneys from the 1705 IRD donors that were declined by pediatric candidates had median KDPI of 20 (IQR 11-31) and were eventually used for transplant in 3052 adult recipients with median (IQR) age of 50 (IQR 39-61) at transplant.

Mortality risk associated with IRD acceptance vs decline for pediatric candidates

Cumulative mortality among pediatric candidates who accepted versus declined an IRD kidney offer was 0.4% vs 0.6% at 1-year and 3.7% vs 2.2% 6-years post-decision (P = 0.1) (Figure 1). After adjustment, we did not detect a survival benefit in pediatric candidates who accepted IRD offers; in fact, there was a trend towards increased risk (adjusted hazard ratio [aHR]: _0.791.72_3.73, P = 0.2) although this was not statistically significant.

Figure 1.

Cumulative mortality among pediatric candidates following the decision to accept or decline an increased-infectious risk donor (IRD) kidney offer

^aCandidates followed from date of IRD acceptance or initial IRD decline until date of death or administrative censoring on 4/30/2018, irrespective of subsequent transplants or removal from the waitlist.

However, the mortality risk associated with IRD acceptance varied by candidate preemptive status (p-value interaction term = 0.006). Among pediatric candidates who were on dialysis at the time of IRD offer, cumulative mortality for those who accepted (N = 205) vs declined (N = 1534) the IRD was 2.4% and 2.8% (2 and 21 deaths at six years, respectively) (Figure 2A). After adjustment, there was no additional mortality risk or survival benefit associated with accepting the IRD among candidates on dialysis at the time of IRD offer (aHR: _0.310.89_2.55, P = 0.8). Among preemptively listed pediatric candidates, however, cumulative mortality for those who accepted (N = 81) the IRD was actually higher than those who declined (N = 716): 7.0% vs 0.9% at six years (3 and 3 deaths by 6 years, respectively) (Figure 2A). After adjustment, the additional mortality among preemptive candidates who accepted the IRD remained statistically significant (aHR: _2.589.70_36.52, P = 0.001). To better understand this counterintuitive finding, we explored the characteristics of the preemptively listed candidates who accepted the IRD. The preemptive candidates who accepted the IRD were older (median [IQR] age: 15 [9-17] vs 13 [8-16], P=0.03) and had waited more months on the waitlist (Med [IQR]: 6.0 [1.9-15.9] vs 3.6 [1.2-10.4] months, P = 0.02) than those who decline the IRD, but again, differences in mortality remained after adjusting for these factors.

Figure 2.

Cumulative mortality among pediatric candidates following the decision to accept or decline an increased-infectious risk donor (IRD) kidney offer, stratified by preemptive^a status.

^aPreemptively listed candidates were not on dialysis at the time of the IRD offer.

^bCandidates followed from date of IRD acceptance or initial IRD decline until date of death or administrative censoring on 4/30/2018, irrespective of subsequent transplants or removal from the waitlist.

Outcomes following the decision to decline an IRD offer

Six years after their initial IRD offer decline, 68.6% of pediatric candidates had received a non-IRD DDKT, 13.0% had received an LDKT, 6.9% had received a different IRD DDKT, 0.8% had died on the waitlist, 8.4% had been removed from the waitlist, and 2.3% were still waiting (Figure 3). Among those removed from the waitlist, 2.4% were removed for improved condition and 5.3% for deteriorating condition. Cumulative incidence of death without a transplant varied slightly by candidate age (1.8% died among those <6 years old versus 0.0-1.0% in other age groups) (Table 2A). Cumulative incidence of LDKT after IRD decline varied slightly by candidate age (15.9% among those <6 years, 14.0% among those 11-14 years, and 7.1-12.7% in other age groups). After stratification by preemptive status, pediatric candiates who were listed preemptively had slightly higher cumulative incidence of LDKT (17.3% vs 11.0%) and lower cumulative incidence of non-IRD DDKT at six years when compared to those on dialysis at the time of IRD decline (61.8% vs 71.7%) (Table 2B).

Figure 3.

Outcomes following the first decline of an increased infectious risk (IRD) kidney offer among pediatric candidates.^a

IRD increased infectious risk donor; DDKT deceased donor kidney transplant; LDKT living donor kidney transplant

^aPediatric candidates followed from date of initial IRD decline until the next major outcome or administrative censoring on 4/30/2018.

Removal from waitlist includes transferred or transplanted at another center, condition improved, condition deteriorated, transplanted in another country, unable to contact, and other.

Table 2.

6-Year Cumulative Incidence^a of Outcomes Following the Decision to Decline an Increased Infectious Risk (IRD) Kidney Offer, Stratified by (A) Pediatric Candidate Age and (B) Preemptive^b Status.

(A) Stratified by Candidate Age.
	<6 Years (N = 462)	6–10 Years (N = 378)	11–14 Years (N = 510)	15–17 Years (N = 702)	≥18 Yearsc (N = 198)
Non-IRD DDKT	69.6%	70.6%	69.7%	66.7%	67.1%
LDKT	15.9%	12.1%	14.0%	12.7%	7.1%
IRD DDKT	4.6%	5.9%	7.8%	8.0%	7.3%
Died	1.8%	1.0%	0.0%	0.8%	0.7%
Removed	7.7%	6.9%	6.5%	9.8%	12.6%
Still waiting	0.5%	3.5%	2.0%	2.1%	5.2%
(B) Stratified by Candidate Preemptive Status.
	On Dialysis at Time of Offer (N = 1534)	Preemptively Listed (N = 716)
Non-IRD DDKT	71.7%	61.8%
LDKT	11.0%	17.3%
IRD DDKT	6.8%	7.0%
Died	1.0%	0.3%
Removed	8.0%	9.2%
Still waiting	1.5%	4.4%

DDKT deceased donor kidney transplant; IRD increased infectious risk donor; LDKT living donor kidney transplant.

^aPediatric candidates followed from date of initial IRD decline until the next major outcome or administrative censoring on April 30, 2018.

^bPreemptively listed; not on dialysis at the time of offer.

^cJoined the waitlist before the age of 18.

Non-IRD DDKT following IRD decline

Among pediatric candidates who refused an IRD offer, median time to non-IRD DDKT was an additional 9.6 months following the IRD decline; 54.2% received a non-IRD DDKT within one year of IRD decline, and 63.5% within 2 years. Stratifying by candidate age, for those ages <6, 6-10, 11-14, 15-17, and ≥18 years who declined IRDs, median time to non-IRD DDKT was an additional 11.2, 8.3, 9.7, 9.2, and 8.7 months following IRD decline. Stratifying by preemptive status, median time to non-IRD DDKT was 8.8 months for candidates who were on dialysis at the time of IRD decline and 12.7 months for candidates who were preemptively listed at the time of IRD decline. The 1491 (68.4%) pediatric candidates who later received a non-IRD DDKT accepted non-IRD kidneys with median (IQR) KDPI 12.7 (6.11-23.4) after having declined IRD kidneys with KDPI 16.0 (8.8-28.7); 20.9% received non-IRDs that were ≥10 KDPI units worse than the IRD they had previously declined.

DISCUSSION

In this national study of 2417 pediatric waitlisted kidney transplant candidates who were offered an IRD kidney that was eventually accepted by someone, there was no survival benefit associated with accepting an IRD kidney when compared to declining it and remaining on the waitlist (p=0.2). Long-term survival was excellent both for pediatric candidates who accepted and for those who declined the IRD kidneys, with six-year survival post-decision of 96.3% for those who accepted and 97.8% for those who declined an IRD kidney. Following an IRD decline, 0.8% of pediatric candidates overall died before receiving a transplant, and 1.8% of those <6 years old died before receiving a transplant. Pediatric candidates who declined an IRD waited a median of 9.6 additional months to undergo non-IRD DDKT (11.2 months among those <6 years old, 8.8 months among those on dialysis), and median KDPI of the eventually accepted non-IRD kidneys (13 [IQR: 6-23]) was similar to median KDPI of the declined IRDs (16 [IQR: 9-28]).

The lack of survival benefit for pediatric candidates is in contrast to our previous findings in adult candidates, where accepting an IRD kidney was associated with 33% lower risk of death from 1-6 months post-decision and 48% lower risk of death beyond 6 months post-decision.⁷ This may be the result of the relatively low waitlist mortality rate for pediatric kidney transplant candidates (1.4 deaths per 100 waitlist years in 2015-2016) and the higher priority of pediatric candidates on kidney transplant waitlists.^8,9 Between 66%-75% of pediatric candidates who declined an IRD kidney later received a non-IRD kidney transplant with median time to transplant of less than one year. In contrast, only 31% of adults received non-IRD DDKTs even at 5-years after IRD decline.⁷

Although we did not find a survival benefit for pediatric candidates on dialysis who accepted an IRD kidney, we did find that those on dialysis who declined IRDs had to wait a median of 8.8 additional months for a non-IRD transplant. This extension of time on dialysis is likely not without consequence, especially for children in whom ESRD is associated with impaired growth, decreased neurocognitive performance across a number of domains, and chronic school absenteeism.^12-17 On the other hand, there is a small but non-zero risk of disease transmission that parents and physicians must consider when deciding to accept an IRD kidney for a pediatric patient,^3,4,6 and pediatric candidates would have to suffer the consequences of this disease transmission for longer than their adult counterparts. For some candidates, the benefits of earlier transplantation (such as improved growth, freedom from dialysis) may outweight the low risk of infectious disease transmission. The balance likely depends on a combination of how the child is currently tolerating dialysis, parental preferences for ongoing dialysis, comfort with infection transmission risk, and physician judgement.

One particular pediatric patient group that we identified in this study were those ages <6 years old at the time of IRD offer. These candidates represented 20% of those who declined IRDs, and they experienced slightly higher risk of death following IRD decline (1.8% vs 0.0-1.0% among other age groups) and slightly longer median time to non-IRD DDKT following IRD decline (11 months vs 8-9 months among other age groups). While we observed equipoise in accepting an IRD kidney offer among pediatric candidates on dialysis in terms of mortality, the consequences of an additional year on dialysis, particularly for candidates <6, should be considered when faced with the decision to accept an IRD offer for young pediatric candidates. We also showed that preemptively listed pediatric candidates who accepted an IRD kidney experienced increased mortality compared to those who declined. However, this was driven by a very small number of deaths among those who accepted an IRD kidney (a total of 3 deaths). This may be the result of confounding by indication, and while our result was statistically significant, it should be considered in the context of a wide confidence interval (2.58-36.52).

Our work extends the recent work of Wrenn et al, who studied post-transplant mortality risk in 1922 pediatric kidney transplant recipients of whom 92 received an IRD kidney from 2004-2008, and found that IRD kidneys were not associated with increased risk of graft loss or mortality for pediatric recipients.¹⁸ In the present study, we studied the survival benefit of IRD acceptance on the waitlist among 2417 pediatric candidates from 2009-2017 and found no additional mortality risk associated with accepting them, although those who declined the offers waited a median of 9.6 additional months on the waitlist. An analysis of post-transplant mortality risk (such as Wrenn et al) assumes that pediatric candidates have simultaneous offers of IRD and non-IRD kidney offers, in which case the decision would be clear. By studying the survival benefit of IRD acceptance, we examined the consequences of the decision to decline the IRD kidney, where candidates must decide whether to accept a non-IRD kidney, or decline and wait for a non-IRD offer. Both studies, in conjunction with existing literature demonstrating similar post-transplant risk associated with IRD kidneys, support the argument for broader acceptance of IRD organs.^19-21

The limitations of our study merit consideration. We were unable to account for clinical characteristics that might have driven the decision to accept or decline an IRD but were not captured by OPTN data (ie, unmeasured confounders). This limitation might be particularly important for our analysis of preemptively listed pediatric candidates, as preemptive candidates who accepted an IRD might have been clinically different from those who declined. Additionally, we were unable to quantify the long-term effects of extended time on dialysis in pediatric candidates who did not accept the IRD kidney (eg, growth, neurodevelopment, and cardiovascular development) or study actual post-transplant transmission of infectious disease as these data are not collected by OPTN. Similarly, we are limited by the relatively short follow-up time (median 3.9 years); however, this is the longest follow-up time available with current organ offer data. Also, while the decision to accept an IRD kidney for a pediatric candidate likely varies according to the reason the donor was labeled IRD, these data are not collected by OPTN. Finally, a pediatric candidate with a potential living kidney donor progressing through evaluation might be more likely to decline an IRD kidney offer, however, this information is not available through OPTN data. While we were unable to account for these factors, the current analytical structure allows us to approximate the counterfactual patient who was offered and had the opportunity to accept an IRD. Previous survival benefit analyses were limited to using the waitlisted population as the comparison group; however, waitlisted candidates who were offered an IRD kidney are likely different from pediatric candidates who did not receive the IRD offer at all.

In conclusion, we have shown that pediatric candidates who accepted an IRD kidney did not have a survival benefit when compared to declining an IRD and waiting for a non-IRD kidney. However, candidates who declined an IRD kidney waited a median of 9.6 additional months for a non-IRD kidney, facing continued developmental, cognitive, and cardiovascular risks of ongoing ESRD in the meantime. Pediatric candidates, parents, and physicians should have an open discussion during the listing process and at the time of offer about whether declining an IRD kidney is worth extending time on dialysis for these candidates who can develop lasting consequences of continued ESRD.

ABBREVIATIONS

aHR

adjusted hazard ratio

DDKT

deceased donor kidney transplantation

ESRD

end-stage renal disease

HCV

hepatitis C virus

HIV

human immunodeficiency virus

IRD

increased risk for disease transmission/increased infectious risk

IQR

interquartile range

KDPI

kidney donor profile index

LDKT

living donor kidney transplant

OPTN

Organ Procurement and Transplantation Network

SCD

standard criteria donor

SRTR

Scientific Registry of Transplant Recipients