Outcomes of Single Pediatric Kidney Grafts: An Experienced Single-center Retrospective Study

Abstract

Background.

Kidney transplantation (KTx) is superior to dialysis but is limited by organ shortages. Kidneys from pediatric donors, whether single or en bloc, have excellent outcomes but are underutilized, with little consensus on splitting pediatric en bloc grafts.

Methods.

A single-center retrospective review of all deceased donor KTx using single kidneys from pediatric donors weighing ≤20 kg between 2005 and 2023 was performed. Recipients were compared based on donor weight: <15 and ≥15 kg. Data are presented as medians (interquartile range) and numbers (percentages).

Results.

Seventy-two recipients were identified, with 36 (50%) males and a median age of 48.3 y at KTx. Donor weight ranged between 7 and 20 kg, with 24 patients receiving a single graft from donors <15 kg and 48 having received a single graft from donors ≥15 kg. Kidney length measured 6–9 cm, with those in the <15 kg group having significantly shorter kidney length relative to those in the ≥15 kg group (7 versus 7.5 cm; P = 0.05). Both Kidney Donor Profile Index (60 versus 50; P = 0.002) and use of hypothermic machine perfusion (96 versus 63%; P = 0.002) were significantly higher in the <15 kg group. Primary nonfunction was significantly higher in the <15 kg group (12.5% versus 0.0%; P = 0.03). One graft was lost because of thrombosis in the ≥15 kg group. There was no difference in graft survival between weight groups, with 5-y graft survival approaching 91% for the entire cohort.

Conclusions.

Our center’s experience demonstrates that single pediatric kidney grafts from donors carry excellent long-term outcomes, especially in donors >15 kg. However, single grafts from donors <15 kg carried a higher risk of primary nonfunction and should be approached with particular caution regarding donor quality and cold ischemia time.

INTRODUCTION

Kidney transplantation (KTx) remains the optimal treatment for end-stage kidney disease, resulting in growing KTx waitlists amid organ shortages. Despite excellent long-term outcomes, pediatric deceased donor kidneys represent an underutilized resource, in part because of concerns regarding technical complications, hyperfiltration injury, and early graft loss.^1-5 Pediatric kidney grafts can be implanted as en bloc or single grafts, with little consensus on when to use either approach.

Donor body weight, kidney length, and age have each emerged as pragmatic surrogates for nephron mass and vascular caliber and have been used to guide decisions regarding en bloc versus split KTx. Bretan et al¹ proposed that grafts from donors <2 y old or weighing <15 kg should be transplanted en bloc, citing a graft survival rate of only 69% with split grafts from donors <2 y old. In contrast, Borboroglu et al² achieved a 2-y graft survival rate of 93% using length criteria of ≥6 cm when splitting en bloc kidneys from donors <2 y old. Multiple large national registry reviews of pediatric en bloc and solitary KTxs supported using solitary kidneys from donors either ≥10 or ≥20 kg, showing that the greater risk of graft failure when using solitary kidneys from smaller donors may be offset by greater total graft years to the recipient population. Registry analyses and single-center reports suggest that kidneys from donors above certain weight thresholds may be safely split and transplanted as single kidneys, but reported cutoffs typically range between 10 and 20 kg.^5-7 Many of these prior studies rely on national registry data, which may lack granular clinical detail and may be influenced by varying center-level selection bias.

The objective of this study is to characterize KTx outcomes using single pediatric grafts from donors weighing ≤20 kg at a high-volume transplant center and to identify donor factors associated with graft outcomes in this weight range.

MATERIALS AND METHODS

Study Design and Population

We conducted a retrospective cohort study of all deceased donor KTxs performed at UC Davis Medical Center between 2005 and 2023. We included all patients who received a single pediatric KTx graft from donors weighing ≤20 kg. Pediatric recipients, recipients of multiorgan transplants, and en bloc grafts were excluded. We compared donor characteristics, recipient characteristics, and overall outcomes between recipients of grafts from donors weighing <15 kg and recipients of grafts from donors weighing ≥15 kg. Data from recipients and donors were compiled from retrospective chart review and DonorNet.⁸ Informed consent was waived by the University of California, Davis Institutional Review Board because of it being a retrospective review.

Recipient Selection, Operation, and Immunosuppression

Recipients with a low body mass index (BMI), a low panel of reactive antibody (<40%), and patients with a low likelihood of disease recurrence were preferentially selected. The decision to transplant a pediatric en bloc graft as a single rather than an en bloc unit was made at the discretion of the transplanting surgeon, guided by a constellation of donor factors including donor status (donation after brain death versus donation after circulatory death [DCD]), terminal creatinine, kidney length, estimated cold ischemia time (CIT), donor-to-recipient weight matching, and intraoperative vascular anatomy. No formalized protocol governed this decision. All kidneys were implanted in the retroperitoneal space on the iliac vessels using standard technique. Ureteroneocystostomy was performed using either the Lich-Gregoir technique^9,10 or the Taguchi single-stitch technique.¹¹ Ureteral double-J stents were placed in all cases and were removed after 6 wk. All patients received induction immunosuppression with rabbit anti-thymocyte globulin and maintenance immunosuppression with tacrolimus and mycophenolate mofetil, along with a rapid steroid taper.

Outcomes

Primary nonfunction (PNF), vascular thrombosis, hospital length of stay, delayed graft function (DGF), creatinine at 6 mo and 1-y posttransplant, 1-y graft survival, and 1-y patient survival were evaluated. PNF was defined as dialysis dependence or failure to achieve adequate graft function (glomerular filtration rate >20 mL/min/m²) within 90 d posttransplant. DGF was defined as the need for dialysis within 7 d of transplantation.

Pathology

All transplant biopsies were assessed by a renal pathologist at UC Davis Medical Center. Biopsies were performed for clinical indications and were examined by hematoxylin and eosin, periodic acid-Schiff, and Masson trichrome stains. Analysis was performed according to the updated 2015 Banff classification.¹²

Statistical Analysis

Baseline donor, recipient, and transplant characteristics were compared between donors weighing <15 kg and donors weighing ≥15 kg. Categorical variables were presented as numbers with percentages, and continuous variables were presented as median and interquartile range (IQR). The Wilcoxon rank-sum test was used to compare continuous variables between the 2 donor weight groups. For categorical data, Fisher exact test was used. Statistical significance was determined at P values of <0.05 for all tests. Graft and patient survival rates were estimated using the Kaplan-Meier method from the date of transplant to the last date of follow-up. Patients without events were censored at the last follow-up time. Pairwise comparisons between groups were made using the log-rank test. All statistical analyses were performed using R (v4.4.0). Given the small number of outcome events, all analyses are descriptive and unadjusted. Multivariable modeling was not feasible. Findings should be considered hypothesis-generating and interpreted with caution with respect to causal inference.

RESULTS

Study Cohort

Donor Characteristics

A total of 72 recipients of single pediatric kidney grafts were included (Table 1). Median donor weight was 15.4 kg (IQR, 14–17.7 kg), with 24 (33%) grafts from donors weighing <15 kg and 48 (67%) grafts from donors weighing ≥15 kg. Donors ≥15 kg were also older (3 versus 2 y; P = 0.001), demonstrated longer kidney length (7.5 versus 7 cm; P = 0.01), had lower Kidney Donor Profile Index scores (50 versus 60; P = 0.002), and were less likely to be placed on hypothermic machine perfusion before transplant (62.5 versus 95.8%; P = 0.002). The median cold ischemic time for the cohort was 27.4 h, with no statistical difference between the 2 groups. There were 27 (37.5%) grafts from DCD donors, with no difference between both groups.

TABLE 1.

Donor demographics

Donor weight	<15 kg	≥15 kg	P
Sample size (%)	24 (33)	48 (67)	N/A
Age (y)	2 (1–3)	3 (2.8–5)	0.001
Sex (male)	13 (54.2%)	31 (64.4%)	0.45
Weight (kg)	14 (12.8–14)	17.1 (15.5–18)	<0.001
Kidney length (cm)	7 (6.5–7.3)	7.5 (7.0–8.0)	0.01
KDPI	60 (50–60)	50 (40–50)	0.002
DCD status	9 (37.5%)	18 (37.5%)	1
Terminal Cr (mg/dL)	0.4 (0.2–0.4)	0.3 (0.3–0.6)	0.32
Cold ischemia time (h)	28.2 (21.3–37.1)	26.2 (19.6–32.8)	0.34
Hypothermic machine perfusion	23 (95.8%)	30 (62.5%)	0.002

Cr, creatinine; DCD, donation after circulatory death; KDPI, Kidney Donor Profile Index; N/A, not applicable.

Recipient Characteristics

Recipients of kidneys from <15 kg donors were older (57.5 versus 49 y; P = 0.02), and there was a trend toward lower BMI (23 versus 24; P = 0.06) compared with recipients of kidneys from ≥15 kg donors (Table 2). This is reflective of the pretransplant selection tendency for lower BMI patients to receive kidneys from donors of lower weight. There was no difference in sex, history of diabetes, Estimated Posttransplant Survival score, prior transplants, or panel of reactive antibody between weight groups.

TABLE 2.

Recipient demographics

Donor weight	<15 kg	≥15 kg	P
Age (y)	57.5 (46–63.5)	49 (29.8–58.2)	0.02
BMI (kg/m2)	23 (21–24)	24 (22–27)	0.06
Sex (male)	12 (50%)	24 (50%)	1
History of diabetes	6 (25%)	14 (29.2%)	0.79
EPTS	41 (15–79)	29 (7–66)	0.32
Prior transplants	0 (0.0%)	3 (6.2%)	0.55
CPRA	0 (0–26.4)	0 (0–16.7)	0.68

BMI, body mass index; CPRA, calculated panel reactive antibody; EPTS, estimated posttransplant survival.

Graft Function and Donor Weight

We observed 4 cases of early graft loss in our cohort: 3 were PNF and 1 was vascular thrombosis (Table 3). All cases of PNF were within the <15 kg group (P = 0.03). The solitary episode of vascular thrombosis took place in the ≥15 kg group and was a left kidney with single vessels. The loss was attributed to venous thrombosis because of a technical error. The rates of DGF between <15 and ≥15 kg donors were not significantly different (29.2% versus 16.7%, respectively; P = 0.23). Creatinine in recipients with grafts from donors <15 kg versus those from ≥15 kg donors were not significantly different at 6 mo and 1-y posttransplantation (Table 3). All kidneys in this cohort measured ≥6 cm in length. The previously reported ≥6 cm threshold therefore did not discriminate outcomes within this series, and all 3 PNF grafts had kidney lengths of 7–8 cm.

TABLE 3.

Solitary kidney transplant outcomes

Donor weight	<15 kg	≥15 kg	P
Delayed graft function (%)	7 (29.2)	8 (16.7)	0.23
Primary nonfunction (%)	3 (12.5)	0 (0)	0.03
Hospital LOS (d)	5 (5–6)	5 (4–5)	0.16
Graft thrombosis (%)	0 (0.0)	1 (2.1)	1
Creatinine at 6 mo (mg/dL)	1.1 (0.9–1.3)	1.3 (1.0–1.5)	0.18
Creatinine at 12 mo (mg/dL)	1.1 (0.9–1.3)	1.1 (0.9–1.3)	0.75

LOS, length of stay.

Among all recipients, there was a median follow-up period of 6.7 y. The median follow-up period between both groups was not found to be significantly different and was 6.8 and 6.4 y in the <15 and ≥15 kg donor weight categories, respectively (P = 0.62). Overall estimated 1-y graft survival was 94.4% (95% confidence interval [CI], 85.9%-97.9%). Estimated 1-y graft survival rates for kidneys from donors <15 and ≥15 kg were 87.5% (95% CI, 74.3%-100%) and 97.9% (95% CI, 93.9%-100%), respectively. Estimated 3-y graft survival was 83.1% (95% CI, 68.0%-98.2%) and 95.6% (95% CI, 89.7%-100%) for the <15 and ≥15 kg groups, respectively. Estimated 5-y graft survival was 83.1% (95% CI, 68.0%-98.2%) and 90.6% (95% CI, 81.8%-99.4%), respectively, with the survival curves converging over longer follow-up. During the total follow-up period, graft survival between groups was not significantly different (log-rank P = 0.40; Figure 1). There were no deaths in the first year of follow-up. Overall estimated patient survival at 1 y was 100% and 98% for the <15 and ≥15 kg donor groups and did not differ significantly (log-rank P = 0.78; Figure 2).

FIGURE 1.

Single pediatric kidney graft survival.

FIGURE 2.

Single pediatric kidney transplant recipient survival.

Early Graft Loss Cohort

There were 3 recipients who suffered from PNF. Median donor weight was 12 kg (IQR, 10.9–14.3 kg), with all weighing <15 kg (Table 4). The 3 PNF cases had numerically longer median CIT (32 versus 26.5 h), although this comparison was underpowered given the small event count (P = 0.31). CIT across the full cohort ranged from 5.3 to 53.6 h, and several grafts with CIT exceeding 50 h functioned successfully. The contralateral (mate) kidney from 2 of the 3 PNF donors was identifiable in our dataset. In the first case, the mate kidney was implanted with approximately 6 fewer hours of cold ischemia (25.4 versus 31.1 h), had no DGF, and remains functional at nearly 8 y of follow-up, suggesting that CIT may have been a contributing factor in the PNF outcome. In the second case, the PNF kidney had a CIT exceeding 47 h. Its mate (CIT 39.9 h) developed DGF and ultimately failed at 2.3 y from causes attributed to recipient factors. The mate kidney from the third PNF case was discarded and not transplanted. Histologic changes of the failed grafts are shown in Figure 3. Widespread tubular atrophy and interstitial fibrosis were present in each of the 3 grafts that failed because of PNF. There was also evidence of mesangial expansion, commonly seen in pediatric grafts with hyperfiltration injury. There was 1 graft lost in the ≥15 kg group because of vascular thrombosis. The graft was from a 2-y-old DCD donor, weighing 18 kg, Kidney Donor Profile Index of 61% with normal renal function, and a kidney length of 7.5 cm.

TABLE 4.

Primary nonfunction cases

	Donor age (y)	Donor weight (kg)	Kidney length (cm)	DCD	HMP	Terminal Cr	KDPI	CIT	Recipient BMI
Case 1	<1	10.9	7	No	Yes	2.9	85	32	21
Case 2	1	12	7	Yes	Yes	0.75	77	47.2	27.6
Case 3	2	14.3	8	No	Yes	0.59	64	31.1	20.4

BMI, body mass index; CIT, cold ischemia time; Cr, creatinine; DCD, donation after circulatory death; HMP, hypothermic machine perfusion; KDPI, Kidney Donor Profile Index.

FIGURE 3.

Histology of early graft failure. A, Light micrographs show the development of significant mesangial widening because of mesangial matrix accumulation (PAS, ×400). B, Segmental glomerulosclerosis with hyalinosis and adhesion to the adjacent Bowman capsule, indicating focal segmental glomerulosclerosis (PAS, ×400). C, Glomeruli showing global sclerosis alongside mesangial widening, one exhibiting segmental glomerulosclerosis (PAS, ×200). D, Development of interstitial fibrosis and tubular atrophy (trichrome, ×200). PAS, periodic acid-Schiff.

DISCUSSION

Utilization of single pediatric kidneys from small donors has been effective at expanding the donor pool, yet outcome data in this weight range remain limited. We report our single-center experience with 72 single kidney grafts from pediatric donors weighing <20 kg. Our data demonstrate excellent graft survival with a median follow-up exceeding 6 y. A higher risk of PNF was observed in the smaller donor group (<15 kg). This finding is based on a smaller number of events and should be interpreted as hypothesis-generating pending confirmation in larger series.

There are a few reports of solitary KTx from small pediatric donors, and reported outcomes are varied. Results from earlier reports, often before 2000, showed poor outcomes, with reported 1-y graft survival rates of 48% and 67% for solitary KTxs from donors <2 y old and donors 2 to 6 y old, respectively.¹³ Bretan et al¹ similarly reported a 69% graft survival rate in solitary KTxs from donors 1 to 2 y old. Preventing episodes of acute rejection was a key step toward improving outcomes with these small kidneys, as acute rejection was reported in 80% of these solitary KTxs in these earlier studies.¹³ Subsequent improvements in immunosuppression have also taken place for both en bloc and solitary pediatric KTxs.^1,13,14 With these improvements in immunosuppression and the incorporation of risk-mitigating selection strategies for recipients (low recipient weight and low immunologic risk) and grafts (donor weight, donor age, graft length and hypothermic pulsatile perfusion), outcomes have improved significantly in more recent years, as has been the experience at our high-volume center. Using kidney length ≥6 cm as a selection criterion, a 2004 study reported 93% 2-y graft survival rates with solitary kidneys from donors <2 y old. Interestingly, this group had a mean donor weight of 14 kg.² A later 2010 study reported similar results in a series of 7 solitary pediatric KTxs that had a mean donor weight of 15 kg.¹⁵ Lastly, a similarly timed 2008 study reported 1-y graft survival of 86% for 14 solitary KTxs from donors weighing <15 kg, an outcome remarkably comparable to our survival rate of 87.5% among donors of similar weight.¹⁶

Despite excellent long-term graft outcomes, pediatric kidney grafts suffer from a low recovery and high discard rate. In addition, there is little consensus regarding when to use these grafts as en bloc versus single grafts. The obvious paradox stems from expanding the donor pool by splitting smaller donors at the expense of a higher early graft loss rate, so determining the best weight or anatomic parameter is needed. Based on our center’s experience, it appears that donor weight, rather than kidney length, is a more practical and suitable parameter to use when splitting en bloc kidneys. Despite using grafts ≥6 cm in length and selecting low-risk recipients, 3 grafts in our series from donors <15 kg failed with PNF. Moreover, thrombosis was not a factor in this weight category. A higher risk of graft thrombosis with small pediatric kidneys has been reported in past studies and has been a significant factor limiting their utilization. Studies cite a rate of thrombosis between 7% and 14% when using single kidneys from donors with either mean weights of 14 kg or <15 kg, respectively.^2,16 In contrast, our series of 72 solitary KTxs resulted in only 1 thrombotic event, which interestingly took place in a donor of >15 kg. Instead, after histological analysis, graft failure in the <15 kg was related to hyperfiltration and factors intrinsic to these small kidneys. Notably, the CIT in these 3 cases was prolonged (>30 h). The contralateral (mate) kidney from the first PNF donor, implanted with approximately 6 fewer hours of cold ischemia, had no DGF and remains functional at nearly 8 y, raising the possibility that CIT contributed to the PNF outcome. In the second PNF case, the mate kidney ultimately failed from causes attributed to recipient factors, underscoring the difficulty in attributing PNF to any single variable. Several grafts in this cohort with CIT exceeding 50 h functioned successfully (CIT range, 5.3–53.6 h), highlighting that CIT is 1 of several interacting risk factors. Graft failure in small pediatric grafts is likely multifactorial, reflecting the interplay of prolonged cold ischemia, donor quality, and graft-to-recipient size mismatch.

Despite the 3 early graft losses in our smaller donor cohort, 21 patients had excellent graft outcomes. The mate kidney analysis is consistent with CIT as a contributing factor, although a definitive causal attribution is not possible given the small event count. This highlights that when single grafts from donors weighing <15 kg are considered, a comprehensive assessment of donor factors including DCD status, terminal creatinine, anticipated CIT, and donor-to-recipient weight matching is essential. Future prospective studies incorporating these variables are needed to establish better selection guidelines for this weight range.

Donor weight is a simple, objective parameter that is readily measured. Unlike kidney length, donor weight can be assessed before procurement and may facilitate earlier allocation decisions, potentially reducing CIT. In our cohort, all kidneys measured ≥6 cm in length. Kidneys measuring <6 cm were likely transplanted en bloc at our institution, explaining their absence from this series. The ≥6 cm criterion previously reported in the literature therefore did not discriminate outcomes within our single-graft cohort. All 3 PNF grafts had kidney lengths of 7–8 cm. Kidney length can also be variably measured because of the type of preservation solution, varying amounts of perinephric adipose tissue, the presence of adrenal glands, and generalized organ edema at procurement. Together, these observations support donor weight as a more practical initial parameter, although prospective validation is needed before formal incorporation into allocation protocols.

The limitations of our study stem from its retrospective design and the single high-volume center nature, which may limit generalizability. The number of PNF events is small, precluding multivariable adjustment, making all analyses unadjusted and descriptive. The study is insufficiently powered to detect the observed differences in DGF (29% versus 17%) between weight groups. Formal power calculations indicate that more than 160 patients per group would be required to detect a difference at 80% power with a 2-sided alpha of 0.05. Nonsignificant findings for DGF, creatinine, and graft survival should therefore be interpreted as inconclusive rather than evidence of equivalence. Recipients were preferentially selected based on low BMI, low immunologic risk, and low likelihood of disease recurrence, with graft splitting performed at surgeon discretion. This introduces selection bias that limits generalizability. Finally, our study did not include an en bloc comparator group, thus limiting direct comparison of single versus en bloc transplantation outcomes.

CONCLUSIONS

In this single-center experience, single kidney grafts from pediatric donors weighing ≥15 kg demonstrated excellent long-term outcomes, with 5-y graft survival approaching 91%. These findings are consistent with a 15 kg donor weight threshold as a practical starting point for decisions regarding splitting of en bloc pediatric kidneys. Donor weight is objective, measurable before procurement, and may facilitate earlier allocation decisions. However, this is a hypothesis-generating, single-center finding and confirmation in larger multicenter studies with en bloc comparator groups is required before formal allocation recommendations can be made. Caution is warranted when using grafts from donors weighing <15 kg, and individual donor factors, including CIT, DCD donor status, and terminal creatinine, should be carefully evaluated.