Post-transplant complications, patient, and graft survival in pediatric and adolescent kidney transplant recipients at a tropical tertiary care center across two immunosuppression eras

Abstract

Background

We report pediatric PAKT patient and graft outcomes at a large tropical tertiary center spanning two transplant eras.

Methods

In this retrospective cohort study, all children ≤18 years who underwent kidney transplantation at our center between 1991 and 2016 were included. Data pertaining to their baseline characteristics, post-transplant events, and outcome were retrieved from transplant records and compared between transplant eras (1991-2005 and 2006-2016).

Results

A total of 139 children (mean age 15.2 ± 2.9 years) underwent PAKT during this period. The incidence of UTIs, CMV disease, BKVN, invasive fungal infections, new-onset diabetes after transplant, leucopenia, and recurrent NKD was higher in the 2006-2016 era (P <.001 for all), while 1-year cumulative BPAR was comparable (P =.100). Five-year graft and patient survival in the two eras were 89.9% and 94.2% (P =.365) and 92.1% and 95.3% (P =.739), respectively. Incidence of CMV disease, BKVN, graft loss, and death was lower in the calcineurin withdrawal group. Non-adherence accounted for 36% of graft loss; infections caused 43.7% of deaths. On multivariate Cox proportional hazards analysis, independent predictors for graft loss were UTIs and blood transfusion naïve status and for death were serious infections and glomerular NKD.

Conclusions

PAKT in India has excellent long-term graft outcomes, though patient outcomes remain suboptimal owing to a high burden of infections. Current immunosuppression protocols need to be re-examined to balance infection risk, graft, and patient survival.

Abbreviations

ABMR

Antibody-mediated rejection

ACR

Acute cellular rejection

ATG

Anti-thymocyte globulin

AUC

Area under the curve

BKVN

BK virus nephropathy

BPAR

Biopsy-proven acute rejection

CDC

Complement-dependent cytotoxicity

CMV

Cytomegalovirus

CNI

Calcineurin inhibitor

DSA

Donor-specific antibody

FSGS

Focal segmental glomerulosclerosis

HLA

Human leukocyte antigen

LMIC

Low- and middle-income countries

NAPRTCS

North American Pediatric Renal Trials and Collaborative Research

NKD

Native kidney disease

NODAT

New-onset diabetes after transplant

PAKT

Pediatric and adolescent kidney transplant

PCA

prednisolone, cyclosporine, azathioprine

PCR-SSOP

polymerase chain reaction-sequence-specific oligonucleotide probes

PCR-SSP

polymerase chain reaction-sequence-specific primers

PTLD

Post-transplant lymphoproliferative disorder

PTM

prednisolone, tacrolimus, mycophenolate

TB

tuberculosis

TDM

therapeutic drug monitoring

UTI

urinary tract infection.

1. Introduction

The first pediatric kidney transplant in India was performed at Christian Medical College, Vellore, in the year 1973.¹ However, despite subsequent advances in immunosuppression and surgical expertise, only a handful of centers in India perform >5 PAKTs per year.² We previously reported the outcomes of PAKT at our center spanning the years 1991-2005, during which time the majority of patients received PCA as their maintenance immunosuppression.³ This second report, spanning the years 1991 - 2016, seeks to provide a comprehensive overview of PAKT outcomes in the last 25 years, to determine whether there has been a change in infection profiles, graft, and patient outcome with the use of PTM-based immunosuppression.

2. Methods

In this retrospective cohort study, all children ≤18 years who underwent kidney transplantation at our center between January 1, 1991, and December 31, 2016, were included. Data pertaining to their baseline characteristics, post-transplant events, and outcome were retrieved from transplant records and the hospital information system. Ethical clearance was obtained from the Institutional Review Board and Ethics Committee of our institute, and the need for formal consent was waived as this was a retrospective study. All procedures performed were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

The aims and objectives of this study were to:

1. Describe the baseline clinical and demographic characteristics of the PAKT cohort

2. Compare the frequency of post-transplant medical (rejections, infections, and non-infectious) complications between patients belonging to different immunosuppression eras (1991-2005 and 2006-2016)

3. Compare graft and patient survival between the two eras

4. Describe the factors influencing graft and patient survival for the PAKT cohort as a whole

2.1. Donor selection protocol

For living donor transplants, we accepted only biologically related first- or second-degree donors. Biological relationship between the patient and the donor was ascertained by 50% matching at all HLA loci for first-degree relatives using PCR-SSP from 1991 to 2008 (Olerup SSP-HLA typing kit, CareDx Inc, USA) and PCR-SSOP from 2008 to 2016 (Tepnel LIFECODES HLA Typing kit, Immucor Diagnostics, WI, USA). Typing was done for A and B loci till 2007, with DR and DQ being added in 2008. DNA-based kinship testing was done for foreign nationals and second-degree relatives. After brain death, deceased donor kidneys were allocated to CDC crossmatch negative recipients as per the waitlist maintained by the state organ sharing network. Living donors were selected if they were aged 21-65 years and had an abbreviated Modification of Diet in Renal Disease eGFR of at least 80 mL/min/1.73 m². Donor kidney and vessel anatomy were elucidated using CT angiography. Donors, who were diabetic, who had a family history of diabetes, who were aged <40 years, who had HIV, hepatitis B or C infection, who were aged less than 50 years and hypertensive, or who had proteinuria >150 mg/day, were rejected.

2.2. Recipient evaluation

All children underwent a micturating cystourethrogram to rule out reflux nephropathy and underwent further urological evaluation if reflux or dysfunctional voiding was demonstrated. Hepatitis B and C infection was treated with antivirals or subcutaneous interferon therapy pretransplant, respectively. A CDC crossmatch was performed to rule out anti-HLA antibodies using neat, 1:2, and 1:4 dilutions of plain and dithiothreitol-treated patient serum and lymphocytes of the donor or patient (for auto-crossmatch). Standard and extended incubation with appropriate controls was carried out in each case. DSA screening was introduced in 2007 using ELISA (LAT-M, One Lambda, Inc, CA, USA), followed by Luminex crossmatch using donor lysate in 2008 (Tepnel LIFECODES DSA kit, Immucor Diagnostics, WI, USA) and Luminex single-antigen bead testing in 2013 (Tepnel LIFECODES LSA Class 1 and 2 kits, Immucor Diagnostics, WI, USA). We do not routinely test CMV serological status based on data from our center that found that 95% of recipients were D + R+.⁴

2.3. Surgical technique

Using an extra-peritoneal approach, the transplanted kidney was anastomosed to the most proximal iliac vessel using the end-to-end or end-to-side approach as appropriate. Uretero-neocystostomy was done by Roy Calne’s or Leadbetter-Politano technique, with or without DJ stenting.

2.4. Immunosuppression protocol

2.4.1. Induction protocol

1991-2005: No induction therapy was given for most of the period. 2006-2016: basiliximab (administered as IV infusion via peripheral venous access 1 hour before shifting to theater) 10 mg IV on Day 0 and Day 4 if recipient weighed <35 kg, 20 mg IV on Day 0 and Day 4 if the recipient weighed >35 kg. High immunological risk recipients (historical crossmatch positivity, second transplants, deceased donor transplants, etc) received single-dose ATG 3 mg/kg as IV infusion via central venous access over 6 hours pre- or peri-operatively for elective or deceased donor transplants, respectively.

2.4.2. Maintenance immunosuppression dosage protocols

Steroids: intravenous methyl prednisolone 15 mg/kg at clamp release followed by dexamethasone 0.5 mg/kg/day (Day 0-2), prednisolone 0.4 mg/kg/day (Day 3-60), 0.3 mg/kg/day (Day 61-90), 0.2 mg/kg/day (Day 91-120), and 0.1 mg/kg/day (Day 121 on-wards); azathioprine (1.5 mg/kg/day if >12 years of age, 1 mg/kg day if age <12 years or weight <35 kg); cyclosporine (8-12 mg/kg/day in two divided doses, tapered by 1 mg/kg/day every month to 4 mg/kg/day by 6 months in order to achieve a trough or C₀ concentration of 150-250 ng/ml in the first 3 months, 100-150 ng/ml from 3-6 months, and 75-100 ng/ml thereafter), tacrolimus (0.1 mg/kg/day in two divided doses to achieve C₀ 7-10 ng/mL in the first 3 months, 5-7 ng/ml from 3-6 months, and 3-5 ng/mL thereafter), and mycophenolate (mofetil salt at 40 mg/kg/day or sodium salt at 30 mg/kg/day to achieve an AUC of 30-60 mg.h/L). The CNI and anti-metabolite were started 5 days prior to the date of transplant for living donor transplants. We do not follow a steroid-free protocol.

2.4.3. Prophylaxis protocol

Co-trimoxazole prophylaxis was given for 6 months, and from 2005, valganciclovir prophylaxis was given for 3 months (patients receiving basiliximab induction, and after treatment for an acute rejection) or 6 months (patients receiving induction with ATG).

2.4.4. CNI withdrawal

Till 2007, withdrawal of CNIs for financial considerations after the first year of transplant was attempted in those who had not experienced an acute rejection, after which this practice was abandoned.

2.5. Recipient follow-up

Post-transplant, and prior to discharge from the hospital, patients were seen by a nurse educator and counseled regarding the follow-up protocol, dosage and timing of medications, general precautions to be followed, and the need for lifelong adherence with immuno-suppression and other medications. After discharge, patients were followed up in a dedicated transplant clinic thrice a week for the first two months, twice a week in months 3-4, once a week in months 5-6, at months 9, 12, 18, 24, and yearly thereafter. At each clinic visit, drug adherence was confirmed by the nurse educator.

2.6. Therapeutic drug monitoring protocol

The TDM protocol for tacrolimus at our center involves measuring C_o concentrations at Day 5 post-transplant, with the date of transplant considered as day 0. Thereafter, C_o concentrations are measured at least once a month for the first 6 months, 3 monthly for the next 6 months, 6 monthly during the second year, and at least once a year thereafter. For mycophenolic acid, 6- or 8-hour AUCs are measured using a limited sampling strategy as has been described previously.^5,6 MPA AUCs are measured at day 5 or 6 post-transplant and thereafter at least once in 3 months in the first year, 6 monthly in the second year, and at least once a year thereafter. If target levels are not achieved, additional TDM studies are carried out after appropriate dose adjustment.

2.6.1. Treatment of rejections

BPARs were classified as per the prevailing diagnostic criteria of that era. ACRs were treated with IV methylprednisolone 15 mg/kg/day for 3 days. Steroid-resistant rejections were treated with ATG 2.5 mg/kg/day (1991-2006) or 1.5 mg/kg/day (2007 till date) for 5-10 days. After the availability of C4d staining, which made the diagnosis of ABMRs possible, they were treated with plasmapheresis and replacement intravenous immunoglobulin (100 mg/kg) after each session with or without rituximab, based on clinical response. Biopsies were performed for indication and not per protocol. Due to variations in the criteria and armamentarium for diagnosis of rejections between the eras, only BPARs as whole were compared, without reference to the type of BPAR.

2.6.2. Surveillance for post-transplant infections

Due to financial considerations, our center did not have a policy of routine post-transplant surveillance for BK virus, CMV, or EBV infection during the study period, and all diagnoses were made based on clinical symptoms.

2.7. Definitions

With respect to treatment of BPARs, complete response was defined as creatinine returning to baseline, partial response as at least 50% improvement in glomerular filtration rate, and non-response as <50% improvement in GFR with the administered anti-rejection therapy.

Graft loss was defined as a GFR < 15 mL/min/1.73 m² for ≥3 months, transfer to dialysis or graft nephrectomy.

Infection events across both eras were analyzed as per the current prevailing diagnostic criteria as follows:

CMV disease and UTI were defined as per the American Society of Transplantation Infectious Disease Guidelines, 2013.^7,8 TB was defined as sputum, gastric juice aspirate, bronchio-alveolar lavage, bone marrow aspirate, body cavity fluid from a sterile site, or tissue biopsy positive for acid fast bacillus and confirmed to be M. tuberculosis based on culture (in Lowenstein Jensen medium or my-cobacteria growth indicator tube) or molecular techniques (Xpert MTB/RIF, Cepheid Inc, Sunnyvale, CA, USA) or pyrexia of unknown origin with non-contributory blood, urine, bone marrow, and tissue biopsy cultures that responded to empirical treatment with anti-tuberculous therapy. Invasive fungal infections were defined as culture of the organism from tissue or body cavity fluid obtained from a sterile site and/or histopathological evidence of fungal infection. Herpes labialis, primary varicella, and herpes zoster infections were diagnosed by classical skin lesions and confirmed on Tzanck smears, histopathology or PCR, as appropriate. Adenovirus infection was diagnosed on histopathology and confirmed with PCR.

BKVN was defined as histopathological evidence of polyomavirus infection and confirmed with one or more of the following: SV-40 immunohistochemical staining of biopsy tissue, BK viruria, or BK viremia.

NODAT was defined as per the International Consensus Guidelines for the diagnosis and management of NODAT.⁹ Leucopenia was defined as a total leukocyte count <4000/mm³ on two occasions at least 24 hours apart.

2.8. Statistics

Data were analyzed using the Statistic Package for Social Sciences version 15 (SPSS Inc Chicago, USA). Continuous variables were expressed as mean or median with interquartile range; categorical variables were expressed as frequency and percentage. Chi-square test was used to determine significance between proportions. Kaplan-Meir estimates were used to calculate 1-year cumulative BPAR rate and patient and graft survival at 1, 5, 10, and 15 years and multivariate Cox proportional hazards analysis to determine the factors associated with patient and graft loss. A P value <.05 was considered significant. Comparison of person-time incidence rates of post-transplant medical complications between the two eras was done using the OpenEpi: Open Source Epidemiologic Statistics for Public Health software, version 3.01, available from: https://www.openepi.com/PersonTime2/PersonTime2.htm.

3. Results

A total of 139 PAKTs were performed during the study period, comprising 5.8% of all transplants during that time (139/2395). Baseline characteristics and NKD of the recipients are described in Table 1. Compared to the 2006-2016 cohort, the 1991-2005 cohort had a fewer number of children who underwent preemptive transplantation or CAPD prior to transplant (2.2% vs 12.5% for both, P <.001), were significantly less likely to get induction therapy (7.7% vs 87.5%, P <.001) and valganciclovir prophylaxis (5.5% vs 72.9%, P <.001), and more likely to be on a PCA-based regime (90.1% vs 4.2%, P <.001), and have their CNI withdrawn (40.7% vs 10.4%, P <.001). Of the 7 deceased donor transplants that occurred, delayed graft function was seen in 4/7 (57%), with a median cold ischemia time of 8 hours.

Table 1. Baseline characteristics of recipients (n = 139).

Characteristics	n (%) or mean ± SD	Transplant era 1991-2005 (n = 91)	Transplant era 2006-2016 (n = 48)	P value
Mean age at transplant (years)					15.27 ± 2.98	15.20 ± 3.1	15.40 ± 2.8	.711
Patients aged ≤12 y (%)	22 (15.8)	16 (17.6)	6 (12.5)	.565
Males (%)	93 (66.9)	61 (67.0)	32 (66.7)	.965
Mean weight at transplant (kg)	39.06 ± 11.09	37.9 ± 10.6	40.9 ± 11.8	.142
Median weight at transplant (kg, IQR)	40 (31, 46)	41 (30, 45)	40 (34, 49)	.226
Pretransplant dialysis modality (%)
Hemodialysis	123 (88.5)	87 (95.6)	36 (75)	.001
CAPD	8 (5.8)	2 (2.2)	6 (12.5)
Preemptive	8 (5.8)	2 (2.2)	6 (12.5)
Median duration of dialysis (months, IQR)	3 (2,6)	3 (2, 15)	4 (2, 6)	.121
Pretransplant blood transfusions (%)	49 (35.3)	31 (34.1)	18 (37.5)	.719
Pretransplant bloodborne virus infection (%)	13 (9.3)	11 (12)	2 (4.2)
Hepatitis B	4 (2.8)	3 (3.2)	1 (2.1)	.684
Hepatitis C	9 (6.5)	8 (8.8)	1 (2.1)	.126
Pretransplant tuberculosis (%)	7 (5)	4 (4.3)	3 (6.2)	.634
Living-related transplant (%)	132 (94.9)	86 (94.5)	46 (95.8)	.692
HLA matching for ≥ 2 antigens (%)	131 (94.3)	85 (93.4)	46 (95.8)	.559
Immunologically high-risk transplants	6 (4.3)	2 (2.2)	4 (8.3)	.099
Historical crossmatch/DSA positive (%)	4 (2.8)	1 (1.1)	3 (6.2)
Re-transplant (%)	1 (0.7)	1 (1.1)	0 (0)
ABO incompatible kidney transplant (%)	1 (0.7)	0 (0)	1 (2.1)
Native kidney disease (n = 138)
Unknown	66 (47.5)	42 (46.2)	24 (50)	.665
Glomerular disease	39 (25.3)	28 (30.8)	11 (22.9)	.334
FSGS	11 (7.9)	8 (8.8)	3(6.3)
IgA nephropathy	7 (5)	3 (3.3)	4 (8.3)
Pauci-immune crescentic GN	4 (2.9)	3 (3.3)	1 (2.3)
Lupus nephritis	2 (1.4)	2 (2.2)	0 (0)
Anti-GBM disease	2 (1.4)	1 (1.1)	1 (2.3)
MPGN	2 (1.4)	2 (2.2)	0 (0)	.526
Other glomerular diseases	11 (7.9)	8 (8.8)	3 (3.3)
Urological causes	25 (18)	15 (16.5)	10 (20.8)
CAKUT	24 (17.4)	15 (16.5)	9 (18.7)
Calculous renal disease	1 (0.7)	0 (0)	1 (2.1)
Alport disease	2 (1.4)	2 (2.2)	0 (0)
Primary hyperoxaluria	1 (0.7)	1 (1.1)	0 (0)
HUS	2 (1.4)	2 (2.2)	0 (0)
Others	3 (2.2)	0 (0)	3 (3.3)
Relationship of donor (n = 129)
Mother	81 (60.4)	53 (61.6)	28 (58.3)	.393
Father	32 (23.8)	18 (20.9)	14 (29.2)
Sibling	11 (7.4)	9 (9.9)	2 (4.2)
Second degree	5 (3.6)	2 (2.2)	3 (6.2)
Donor age (years)	40.01 ± 9.5	39.78 ± 9.4	40.42 ± 9.5	.708
Induction received (%)	49 (35.3)	7 (7.7)	42 (87.5)	<.001
Basiliximab	43 (30.9)	7 (7.7)	36 (75.0)
Anti-thymocyte globulin	6 (4.3)	0	6 (12.5)
Maintenance immunosuppressiona
Pred + Aza	5 (3.6)	5 (5.5)	-	<.001
Pred + CyA +Aza	84 (60.4)	82 (90.1)	2 (4.2)
Pred + CyA +MPA	4 (2.9)	4 (4.4)	-
Pred + Tac +Aza	4 (2.9)	-	4 (8.3)
Pred + Tac +MPA	42 (30.2)	-	42 (87.5)
Valganciclovir prophylaxis (%)	40 (28.8)	5 (5.5)	35 (72.9)	<.001
CNI withdrawal per protocol (%)	42 (30.2)	37 (40.7)	5 (10.4)	<.001

Abbreviations: CAPD, continuous ambulatory peritoneal dialysis; GBM; glomerular basement membrane; MPGN, membrano-proliferative glomerulonephritis; CAKUT, congenital anomalies of the kidney and urinary tract; HUS, hemolytic uremic syndrome, Pred, prednisolone; Aza, azathioprine; CyA, cyclosporine A; MPA, mycophenolate; Tac, tacrolimus.

^aMaintenance immunosuppression regime as per initial assignment.

3.1. Biopsy-proven acute rejections (BPARs)

A total of 39 BPARs occurred in the study cohort (crude BPAR rate 28.1%). The Kaplan-Meir estimate of the 1-year cumulative BPAR rate was 18% (14.3% vs 25% in the first and second era, P =.100). Table 2 provides the crude incidence (events/number of persons at risk) of BPARs, anti-rejection treatment given, and response to treatment, by era. Crude incidence of BPAR did not vary between the two eras overall; however, the second era had a lower crude incidence of BPARs beyond the first year (2.7% vs 14.3%, P =.034). BPARs in the second era were more likely to show complete response to treatment (P =.020). Tables 3 and 4 compare the person-time BPAR incidence rate by era and CNI withdrawal status, respectively. The 2006-2016 cohort had a higher person-time incidence of BPARs in the first year post-transplant (Table 3), but not thereafter. The CNI withdrawal group had a lower crude and person-time incidence of BPARs in the first post-transplant year, but beyond the first year of transplant, there was no significant difference in BPAR incidence compared to the group in whom CNI was continued (Table 4). On chi-square analysis, we could not identify any risk factors for BPAR, including age, gender, type of donor, use of induction agent, immunosuppression protocol, NKD, CNI withdrawal, or transplant era (analysis not shown).

Table 2. Frequency of post-transplant infectious and non-infectious complications across transplant eras (n = 139).

Infection	n (%)	Median time tooccurrence (days)	Transplant era 1991-2005 (n = 91)	Transplant era 2006-2016 (n = 48)	P value
UTI	37 (26.6)	33	24 (26.4)	13 (27.1)	.928
Tuberculosis	12 (8.8)	133	10 (11.0)	2 (4.2)	.173
Pulmonarya	7 (58.3)
CNSa	1 (8.3)
Miliarya	1 (8.3)
Empirical treatmenta	3 (25)
CMV disease	22 (15.8)	123	14 (15.5)	8 (16.7)	.844
BKVN	10 (7.2)	134	0 (0)	10 (20.8)	<.001
HCV (de novo)	2 (1.4)	1167	2 (2.2)	0 (0)	.301
HBV (de novo)	1 (0.7)	3478	1 (1.1)	0 (0)	.466
Other viruses	13 (9.5)	86	11 (12.1)	2 (4.2)	.127
Herpes zostera	6 (46.2)	1976
Primary varicellaa	3 (23)	114
Herpes labialisa	3 (23)	75
Adenovirusa	1 (7.7)	32
Invasive fungal infections	5 (3.7)	139	3 (3.3)	2 (4.2)	.793
Candidaa	1 (20)
Cryptococcosisa	1 (20)
Histoplasmosisa	1 (20)
Rhizopusa	1 (20)
Acremoniuma	1 (20)
Leucopenia	49 (35.3)	-	32 (35.2)	17 (35.4)	.976
MPA relateda	15 (30.6)
Azathioprinea	15 (30.6)
CMV diseasea	10 (20.4)
Othersa	9 (18.4)
NODAT	14 (10.3)	-	8 (8.8)	6 (12.5)	.490
BPARs	39 (28.1)	240	26 (28.6)	13 (27.1)	.853
<3 mo	18 (12.9)		10 (11)	8 (16.7)	.427
3-6 mo	0 (0)		0 (0)	0 (0)	-
6-12 mo	7 (5)		3 (3.3)	4 (8.3)	.234
≤12 mo	25 (17.9)		13 (14.3)	12 (25)	.163
>12 mo	14 (10.1)		13 (14.3)	1 (2.1)	.034
Treatment of BPARs
MPb	33 (84.6)		25 (96.1)	8 (61.5)	<.001
MP + PLEXb	1 (2.6)		-	1 (7.7)
MP + PLEX+IVIGb	1 (2.6)		-	1 (7.7)
ATGb	2 (5.2)		1 (3.9)	1 (7.7)
IVIGb	2 (5.2)		-	2 (15.4)
Response to treatment
Complete responseb	14 (35.9)		6 (23.1)	8 (61.5)	.020
Partial responseb	7 (17.9)		6 (23.1)	1 (7.7)	.243
No responseb	18 (46.1)		14 (53.8)	4 (30.8)	.180
Recurrence of NKD	8 (5.7)	-	4 (4.3)	4 (8.3)	.090
FSGSa	4 (50)		2 (50)	2 (50)
IgA nephropathya	3 (37.5)		1 (25)	2 (50)
Primary hyperoxaluriaa	1 (12.5)		1 (25)	0 (0)
Graft loss	25 (18)	-	22 (24.2)	3 (6.3)	.009
Chronic rejection due to non-adherencea	9 (36)		9 (41)	0 (0)
CANa	8 (32)		8 (36.3)	0 (0)
Recurrencea	5 (20)		4 (18.2)	1 (33.3)
Surgical complicationa	1 (4)		0 (0)	1 (33.3)
Acute rejectiona	1 (4)		1 (4.5)	0 (0)
Recurrent graft pyelonephritisa	1 (4)		0 (0)	1 (33.3)
Death	16 (11.5)	-	12 (13.2)	4 (8.3)	.394
Infectiona	7 (43.7)		5 (41.6)	2 (50)
CVAa	3 (18.7)		3 (25)	0 (0)
PTLDa	1 (6.25)		0 (0)	1 (25)
Surgicala	1 (6.25)		0 (0)	1 (25)
Not knowna	4 (25)		4 (33.3)	0 (0)

Abbreviations: CNS, central nervous system; HCV, hepatitis C virus; HBV, hepatitis B virus; MPA, mycophenolic acid; CAN, chronic allograft nephropathy, CVA, cerebrovascular accident.

^aExpressed as a percentage of the total cases in the group.

^bPercentage calculated on total number of BPAR in the respective era.

Table 3. Person-time incidence rates of post-transplant infectious and non-infectious complications across transplant eras calculated as events per 1000 patient-years (n = 139).

	Overall	Transplant era 1991-2005 (n = 91)	Transplant era 2006-2016 (n = 48)	CMLE (95% CI)	P value
BPARs (n = 39)	51.6	47.1	62.9	3.56 (2.44-5.23)	<.001
≤12 mo (n = 25)a	198	161	264	2.93 (3.41-3.57)	<.001
>12 mo (n = 14)b	22.2	28.3	5.9	0.58 (0.22-1.34)	.222
UTI (n = 37)	48.9	43.7	62.9	3.81 (2.595-5.63)	<.001
Tuberculosis (n = 12)	15.9	18.2	9.7	1.478 (0.6556-3.1855)	.3280
CMV disease (n = 22)	29.1	25.5	38.7	4.151 (2.518-6.939)	<.001
BKVN (n = 10)	13.2	0	48.4	-	<.001
HCV de novo (n = 2)	2.6	3.6	0	0 (0-2.966)	.2791
HBV de novo (n = 1)	1.3	1.8	0	0 (0-9.238)	.5283
Other viruses (n = 13)	17.2	20	9.7	1.33 (0.5973-2.816)	.4605
Invasive fungal infections (n = 5)	6.6	5.5	9.7	5.321 (1.833-17.23)	.001
Leucopenia (n = 49)	64.8	58.2	82.3	3.698 (2.649-5.186)	<.001
NODAT (n = 14)	18.5	14.6	19.4	3.37 (1.705 - 6.757)	<.001
Recurrence of NKD (n = 8)	10.6	7.3	19.4	6.32 (2.812-15.3)	<.001
Graft loss (n = 25)	33	40	14.5	0.36 (0.08-1.10)	.078
Deaths (n = 16)	21.2	21.8	19.4	0.88 (0.24-2.64)	.872

Note: Incidence mentioned overall and for each era is calculated as events per 1000 patient-years where incidence rate = (events/total years at risk) x 1000. Total years of follow-up = 756.3 y; 1991-2005 = 549.7 y; 1996-2016 = 206.6 y.

Abbreviations: HCV, hepatitis C virus; HBV, hepatitis B virus; CMLE, conditional maximum-likelihood estimate of rate ratio (with transplant era 1991-2005 taken as the reference).

^aCalculated by summating the follow-up time contributed within the first 12 mo by individual patients.

^bCalculated by summating the follow-up time contributed beyond 12 mo by individual patients.

Table 4. Crude incidence of BPARs and person-time incidence rates of post-transplant infectious and non-infectious complications between CNI withdrawal and non-withdrawal group calculated as events per 1000 patient-years (n = 139).

	Overall	CNI withdrawal (n = 42)	CNI non-withdrawal (n = 97)	CMLE (95% CI)	P
Crude incidence of BPARs
All BPARs (n = 39)	39	9 (21.3)	30 (30.9)		.251
<3 mo (%)	18	3 (7.1)	15 (15.5)	-	.271
3-6 mo (%)	0	0 (0)	0 (0)	-	-
6-12 mo (%)	7	0 (0)	7 (7.2)	-	.102
≤12 mo (%)	25	3 (7.1)	22 (22.7)	-	.031
>12 mo (%)	14	6 (14.2)	8 (8.2)	-	.357
Person-time incidence of events
All BPARs (n = 39)	51.6	25.2	75.3	2.589 (1.672-4.107)	<.001
≤12 mo (n = 25)a	39.7	9.9	67.3	6.31 (3.35-12.91)	<.001
>12 mo (n = 14)b	22.2	19.8	24.5	1.22 (0.67-2.25	.515
UTI (n = 37)	48.9	39.1	57.7	1.335 (0.891-2.016)	.163
Tuberculosis (n = 12)	15.9	11.2	20.1	1.632 (0.786-3.529)	.193
CMV disease (n = 22)	29.1	19.6	37.6	1.795 (1.042-3.173)	.035
BKVN (n = 10)	13.2	2.8	22.6	10.32 (2.842-64.91)	<.001
HCV de novo (n = 2)	2.6	2.8	2.5	1.346 (0.200-11.32)	.774
HBV de novo (n = 1)	1.3	2.8	0	0 (0.0-1.539)	.106
Other viruses (n = 13)	17.2	5.6	27.6	5.025 (2.051-14.65)	<.001
Invasive fungal infections (n = 5)	6.6	5.6	7.5	1.436 (0.464-4.824)	.542
Leucopenia (n = 49)	64.8	50.3	77.8	1.4 (0.983-2.007)	.062
NODAT (n = 14)	18.5	16.8	20.1	1.056 (0.550-2.045)	.874
Recurrence of NKD (n = 8)	10.6	0	20.1	-	<.001
Graft loss (n = 25)	33.1	22.4	42.6	1.754 (1.055-2.979)	.029
Death (n = 16)	21.2	8.4	32.6	3.702 (1.765-8.553)	<.001

Note: Incidence mentioned overall and between CNI withdrawal and CNI non-withdrawal group is calculated as events per 1000 patient-years where incidence rate = (events/total years at risk) x 1000; total years of follow-up = 756.25 y; CNI withdrawal group = 357.7 y; CNI non-withdrawal group = 398.6 y.

Abbreviations: HCV, hepatitis C virus, HBV, hepatitis B virus, CMLE, conditional maximum-likelihood estimate of rate ratio with CNI withdrawal group as the reference.

^aCalculated by summating the follow-up time contributed within the first 12 mo by individual patients.

^bCalculated by summating the follow-up time contributed beyond 12 mo by individual patients.

3.1.1. Infectious and non-infectious complications

The most common post-transplant infectious complication was UTI (26.6%), 21.6% (8/37) of which were recurrent, while the most common non-infectious complication was leucopenia (35.3%), 61.2% drug-related. Two patients had post-transplant thrombotic microangiopathy, both due to CNI toxicity, and one patient had a PTLD, which was an Epstein-Barr virus-related high-grade B-cell non-Hodgkin’s lymphoma. The crude incidence of infectious and non-infectious complications differed between transplant eras only for BKVN, which was exclusively seen in the latter era (Table 2). However, there was a significantly higher incidence of UTI, CMV disease, BKVN, invasive fungal infections, leucopenia, NODAT, and recurrence of NKD in the 2006-2016 transplant era (Table 3) using the person-time incidence rate. The CNI withdrawal group had a lower person-time incidence rate of CMV, BKVN, other viral infections, and recurrence of NKD (Table 4). UTIs were more common in those with urological NKD (64% vs 18.4%, P <.001) and those who had DJ stenting (40.5% vs 10.8%, P <.001), but not those with CMV disease (31.6% vs 25.8%, P =.59). The former two factors did not vary between eras (P =.43 and 0.64, respectively). We could not identify any other risk factors for CMV disease, including BPARs, UTIs, age at transplant, valganciclovir prophylaxis, or induction therapy.

3.2. Graft and patient survival

The median duration of follow-up was 56.5 months (range 0-206 months). Three graft losses occurred in the first month posttransplant (recurrence of primary hyperoxaluria = 1, acute rejection with transmural arteritis = 1, air embolism with primary graft non-function = 1), the former two in the first transplant era. For the entire cohort, graft survival was 96.6%, 90%, 74.1%, and 55.8% for living related and 83.3%, 83.3%, 41.7%, and 41.5% for deceased donor transplants at 1, 5, 10, and 15 years, respectively (P =.253). Figure 1 compares the cumulative death-censored graft survival between eras, which was 94.8%, 83.9%, and 71.4% overall, 96.3%, 90.0%, and 81.3% for the 1991-2005 cohort, and 97.8%, 94.1%, and 90.0% for the 2006-2016 cohort, at 1, 5, and 10 years, respectively (P =.365).

Figure 1. Kaplan-Meir estimate of death-censored graft survival by era.

The cumulative graft loss was 18% (n = 25), the most common etiology being non-adherence (n = 9, 36%), followed by chronic allograft nephropathy (n = 8, 32%) and recurrence of native kidney disease (n = 5, 20%). Crude incidence of graft loss was higher in the 1991-2005 cohort (24.2% vs 6.3%, P =.009, Table 2); however, person-time incidence was not significantly different (P =.078, Table 3). The CNI withdrawal group had a lower person-time incidence of graft loss (P =.029, Table 4). On multivariate Cox regression analysis (Table 5), prior blood transfusions were protective (HR 0.14, 95% CI 0.03-0.55, P =.005), and UTIs (HR 7.17, 95% CI 2.51-20.43, P <.001) were associated with an increased risk of graft loss.

Table 5. Factors associated with graft survival on multivariate Cox regression analysis (n = 136^a).

Variable	Graft loss/total counts	Mean estimate of graft survival on KM curve			Univariate analysis			Multivariate analysis
		Mean estimate (95% CI) years	P value (log-rank Mantel Cox)		Hazard ratio (95% CI)	P value		Hazard ratio (95% CI)	P value
Overall
Total graft loss	25	12.973 (11.618 - 14.329)	-		-	-		-	-
Transplant era
1991-2005	20/89	12.702 (11.154 - 14.251)	.097		3.283 (0.746 - 14.436)	.116		3.010 (0.651 - 13.913)	.158
2006-2016	2/47	9.885 (9.078 - 10.691)
BPAR
Yes	10/38	11.019 (8.253 - 13.785)	.055		2.287 (0.982 - 5.326)	.055		2.423 (0.893 - 6.572)	.082
No	12/98	12.938 (11.838 - 14.039
BT
Yes	3/49	15.337 (13.367 - 17.307)	.068		0.336 (0.099 - 1.141)	.080		0.141 (0.036 - 0.552)	.005
No	19/87	11.600 (10.288 - 12.913)
IMS protocol
Pred/Cy/Aza	18/79	12.784 (11.173 - 14.396)	.202		2.786 (0.612 - 12.139)	.188		-	-
Pred/Tac//MPA	2/42	9.477 (8.479 - 10.475)						-
CNI withdrawal
Yes	8/42	14.237 (12.689 - 15.784)	.116		0.478 (0.192 - 1.189)	.112		0.360 (0.107 - 1.211)	.099
No	14/94	12.740 (10.767 -14.714)
UTI
Yes	11/37	11.236 (8.943 - 13.530)	.030		2.393 (1.034 - 5.536)	.042		7.172 (2.517 - 20.433)	>.001
No	11/99	14.417 (12.942 - 15.891)
NKD
Glomerular	10/47	9.413 (8.122 - 10.704)	.078		2.337 (0.940 - 5.810)	.068		2.533 (0.879 - 7.298)	.085
Others	12/89	14.140 (12.670 - 15.610)

Note: Other covariates analyzed: age at transplant (<12 y, ≥12 y), gender, induction therapy (yes/no), deceased donor (yes/no), preemptive transplant (yes/no), CMV disease (yes/no), HLA matches (<2 vs ≥2 antigen matches), and NKD (urological vs others).

Abbreviations: BT: blood transfusion, IMS: immunosuppression.

^aThree patients who experienced graft loss and/or died within 1 mo of transplant were excluded from the analysis.

Cumulative patient survival was 94.8%, 89.3%, 81.0%, and 79.3% overall; 98.4%, 92.8%, 83.6%, and 83.6% for living related; and 85.7%, 71.4%, 71.4%, and 71.4% for deceased donor transplants at 1, 5, 10, and 15 years, respectively (P =.065). Figure 2 compares the cumulative patient survival between eras, which was 97.5%, 93.8%, 91.4% for the 1991-2005 cohort, and 97.8%, 95.2%, 84.0% for the 2006-2016 cohort at 1, 5, and 10 years, respectively (P =.739). A total of 16 (11.5%) patients died, the most important cause being infections in 43.7% (n = 7, dengue hemorrhagic fever—1, brain abscess—1, fungal sepsis—1, septic shock, cause unknown—4). Two deaths occurred in the first month post-transplant due to intraoperative air embolism (n = 1) and cerebrovascular accident (n = 1), one in each transplant era. Table 2 provides information regarding the causes of death overall and by era. Crude (Table 2) and person-time incidence of death (Table 3) did not vary between transplant eras; however, the CNI withdrawal group had a significantly lower person-time incidence of death (P <.001, Table 4). On multivariate Cox regression analysis, only glomerular NKD (HR 4.77, 95% CI 1.27-17.94, P =.021) and serious infections, which was defined as a composite of UTI, CMV disease, TB, invasive fungal infections, and sepsis of undetermined etiology with septic shock (HR 6.17, 95% CI 1.33-28.64, P =.020), were significantly associated with a higher risk for death (Table 6).

Figure 2. Kaplan-Meir estimate of patient survival by era.

Table 6. Factors associated with patient survival on multivariate Cox regression analysis (n = 137^a).

		Mean estimate of patient survival on KM curve			Univariate analysis			Multivariate analysis
Variable	Deaths/total	Mean estimate (95% CI) years	P value (log-rank Mantel Cox)		Hazard ratio (95% CI)	P value		Hazard ratio (95% CI)	P value
Overall
Deaths	16	14.845 (13.704-15.986)	-			-			-
Transplant era
1991-2005	11/90	14.675 (13.322-16.029)	.500		0.1.555 (0.426-5.670)	.504
2006-2016	3/47	9.681 (8.831-10.531)
IMS protocol
Pred/Cy/Aza	9/80	14.850 (13.450-16.250)	.178		3.803 (0.472-30.620)	.209		3.737 (0.448-31.168)	.223
Pred/Tac//MPA	1/42	9.978 (9.455-10.502)
CNI withdrawal
Yes	3/42	15.807 (14.611-17.003)	.068		0.314 (0.085-1.158)	.082		0.379 (0.065-1.102)	.165
No	11/95	13.837 (11.895-15.779)
Serious Infectionsb
Yes	12/72	13.262 (11.436-15.087)	.010		5.741 (1.284-25.673)	.022		6.176 (1.332-28.646)	.020
No	2/65	16.438 (15.450-17.427)
NKD
Glomerular	11/48	9.231 (7.893-10.570)	<.001		7.123 (1.982-25.599)	.003		4.777 (1.271-17.946)	.021
Other	3/89	16.450 (15.567-17.243)
Donor
Living	12/130	15.010 (13.864-16.155)	.076		0.279 (0.062-1.259)	.097		2.487 (0.539-11.474)	.243
Deceased	2/7	8.712 (4.899-12.525)

Note: Other covariates analyzed: Age at transplant (<12 y, ≥12 y), gender, biopsy-proven acute rejection (yes/no), preemptive transplant (yes/no), induction therapy (yes/no), CMV prophylaxis (yes/no), NKD (urological vs others).

Abbreviation: IMS, immunosuppression.

^aTwo patients who died and experienced graft loss within 1 mo of transplant were excluded from analysis.

^bSerious infections: A composite of UTI, CMV disease, tuberculosis, invasive fungal infection, and sepsis of undermined etiology with septic shock.

3.3. Discussion

This study reports on a large PAKT cohort from a tertiary center located in southern India and seeks to compare outcomes across two eras—1991-2005 and 2006-2016 in order to explore the role that changes in immunosuppressive therapy may have played in this regard.

At the outset, the differences in these cohorts are stark and reflect the advances in transplantation that have taken place over the study period. Because of late referral, children in the earlier cohort were less likely to undergo a preemptive transplant. Induction therapy and valganciclovir prophylaxis had not yet become the norm, and PCA were the most commonly used immunosuppressive agents. CNI withdrawal was considered for patients who had stable graft function and no prior acute rejection in order to mitigate the effects of CNI toxicity and reduce out-of-pocket expenditure on maintenance immunosuppression, but after a series of graft losses in our adult cohort in the early part of the second era, this practice was abandoned.

The crude BPAR rate of 28.1% is comparable to that reported from other Indian transplant centers over the same time period^10–12 and did not vary between eras. The higher person-time incidence of BPARs in the first post-transplant year for the 2006-2016 cohort despite the use of induction therapy and a PTM-based regimen may appear surprising but is likely to be due to a more liberal biopsy policy, since a higher proportion showed complete response to therapy, indicating milder BPARs. The CNI withdrawal group had a lower early BPAR incidence because this was a prerequisite for CNI withdrawal. However, beyond the first post-transplant year, the BPAR rate showed an increasing trend in this group, though statistically the incidence was no different compared to the group in whom CNI was continued. An increase in BPAR rate after CNI withdrawal has also been reported by a meta-analysis of this strategy.¹³ Evidence suggests that CNI withdrawal is not appropriate for children with a prior history of acute rejection and those without evidence of CNI toxicity on the graft biopsy.¹⁴ TDM to insure adequate levels of the prescribed anti-metabolic agent is essential,¹⁴ as is rigorous follow-up in order to diagnose rejections early, both of which may be problematic in resource-constrained settings where access to care is delayed.

The type and crude incidence of various infections were broadly consistent with that reported from other Indian transplant centers,¹⁰ and mirrored the timetable for post-transplant infections in the tropics published by John GT et al¹⁵ However, using person-time incidence rate, we found that the incidence rate of UTIs, CMV disease, BKVN, and invasive fungal infections was higher in the 2006-2016 transplant era. This increase in incidence may be linked to the change in immunosuppression regime from prednisolone/azathioprine or PCA to PTM along with the use of induction agents, resulting in higher net immunosuppression. A previous study from our center found a higher risk of UTIs with DJ stenting, urological NKD, CMV disease, mycophenolate use, and acute rejections, but not with induction therapy.¹⁶ Urological NKD and DJ stenting, though associated with UTIs in this and other cohorts,¹⁷ were no different between eras; therefore, it is likely that mycophenolate use may be responsible for the increase in UTI incidence. The incidence of CMV disease, BKVN, and other opportunistic infections has been reported to be higher in pediatric kidney transplant recipients receiving a tacrolimus and mycophenolate-based regime^18,19 In addition, CMV, BKVN, and other viral infections had a lower person-time incidence in the CNI withdrawal cohort, in keeping with published literature.¹³ It is interesting that despite the more widespread use of valganciclovir prophylaxis in the last transplant era, CMV disease incidence increased. Published literature in the pediatric age group shows that prophylaxis does not reduce the incidence of CMV disease in high (D+/R-) or intermediate risk (D+/R+) groups,²⁰ and the median time to occurrence of CMV disease (123 days) in our cohort confirms that prophylaxis only delayed the onset of CMV disease. Leucopenia was expectedly more common in the second era due to the combined use of mycophenolate and valganciclovir prophylaxis,²¹ along with the increased incidence of CMV disease. NODAT has also been reported to be more common on tacrolimus-based, steroid-inclusive pediatric regimes.^22,23 The incidence of TB did not increase and may be reflective of the shift from cyclosporine to tacrolimus as the maintenance calcineurin inhibitor,²⁴ secular trends in incidence, and prevalence of TB in the general population, as well as improvement in nutrition and consequent immune status. Since the distribution of glomerular disease was no different between eras, we cannot explain the increase in recurrent NKD in the latter transplant era, apart from a probable change in biopsy practice. The lower incidence of recurrence in the CNI withdrawal group is explained by the fact that the CNI is not withdrawn in the presence of a recurrent glomerular disease like FSGS.

Overall five-year graft survival exceeded the NAPRTCS benchmark of 86.50% for living donor transplants and equaled it (83.24%) for deceased donor transplants²⁵ and was not significantly different between eras. Non-adherence was directly linked to 36% of graft loss in this study. Due to the retrospective design, we could not analyze non-adherence as a risk factor for graft loss because we were unable to confirm a bias in documentation of non-adherence (non-adherence is more likely to be documented in those who have suffered an acute rejection or graft loss). The fact that age was not a contributory factor to graft loss may point to the fact that in LMICs like India, non-adherence may be influenced by financial stress in addition to the behavioral constructs that have been extensively described in this age group. It has been estimated that 63%-90% of middle and lower socioeconomic class families avail of financial assistance to fund a pediatric kidney transplant and 45%-80% remain in debt post-transplant.²⁶ Currently, fully subsidized generic immunosuppression is only available to dependents of central government employees, or at government hospitals in a few states in southern India,²⁷ and extending this provision throughout the country may go a long way toward improving graft outcome. UTIs were independently predictive of poor graft survival due to the high proportion of recurrent UTIs. The association of pretransplant blood transfusions with improved graft survival is intriguing, and in line with the findings of a randomized control trial²⁸ and a 2012 technology assessment report by the Agency for Healthcare Quality and Review, drawn largely from studies published prior to 2000, that found a beneficial to neutral effect on acute rejections, 1 year, and maximum graft survival.²⁹ The mechanism underlying this effect is unknown and has been variably attributed to clonal deletion,³⁰ induction of anti-idiotypic antibodies,³¹ and T regulatory cells.³² Recent reports have struck a more cautionary note, highlighted the risk of allosensitization, de novo DSA generation, and increased waiting time for transplantation.^33,34 Although we are unable to confirm the effect of blood transfusions on pretransplant DSA status in our cohort, it is important to note that the risk of allosensitization with blood transfusions is not uniform, being higher in retransplants and multiparous females,³⁵ and may be negligible in an immunologically low-risk population of predominantly living-related donor pediatric transplant recipients, on adequate immunosuppression confirmed by TDM, and no other sensitizing events.

Three additional factors appeared to influence graft survival, though statistical significance was not achieved—BPARs, CNI withdrawal, and glomerular NKD. Two-thirds of BPARs in the first era, and one-third in the second era, showed partial or no response to treatment. Possible explanations include delay in referral, failure to diagnosis ABMRs because of non-availability of C4d staining and DSA testing for most of the study period, and non-adherence. Glomerular NKD is associated with recurrent disease, and CNI withdrawal would appear to be protective because it was selectively applied to an immunologically low-risk population, apart from reducing the risk of CNI toxicity and BKVN.

Five-year patient survival was much lower than the benchmark NAPRTCS value of 96.1% and 93.3% for living and deceased donor transplants, respectively, and did not vary across transplant eras. Infections were the main contributor to mortality, highlighting that even with TDM, graft function in the tropics is maintained at a cost. Children with glomerular NKDs enter into transplantation with a higher net immunosuppression burden and may be particularly prone to adverse outcomes.

3.3.1. The way ahead

Our data show that while patient and graft survival have remained unchanged, opportunistic infections have increased and may have blunted any beneficial effect on either parameter that may otherwise have been expected with the PTM regime. CNI withdrawal in the PCA era was associated with a lower incidence of CMV, BKVN, graft, and patient loss, but the trend toward increased BPARs post-withdrawal suggests this strategy is not without its risks. CNI withdrawal and replacement with mTOR inhibitors are not a financially viable solution in India, given the prohibitive costs involved.²⁷ The middle ground of CNI and anti-metabolite minimization may be the best path to follow, and prospective studies to determine the target CNI and mycophenolate concentrations to optimize patient and graft survival are urgently needed.

3.4. Strengths and Limitations

This study, spanning two immunosuppression eras over 25 years in a large PAKT cohort from India, gives a broad overview of the challenges that pediatric transplant centers in LMICs face while balancing the increased risk of rejection in young recipients within a highly infectious milieu. The comparison of complications across transplant eras highlights how graft and patient outcomes have been affected by changes in the therapeutic armamentarium. Person-time incidence rates are the preferred metric when comparing the relative frequency of events across different time periods with varying duration of follow-up, and its use to compare post-transplant complications between eras and between CNI withdrawal and continuation groups helped reveal changes in incidence trends that may not have been evident if crude incidence had been used.

The limitations of this study include its retrospective nature, the fact that it reports data from a single center and that transplant records were incomplete or missing for a small proportion of patients. Because DSA testing, C4d staining, and molecular testing techniques for BK virus infection were not available during the period from 1991 to 2005, ABMRs could not be diagnosed in that era, and our estimate of BKVN pertaining to that time may be erroneous. The Banff criteria for diagnosis of BPARs have undergone considerable modification over the study period and in order to allow a direct comparison between eras, we restricted our analysis to the incidence of BPARs as a whole, rather than the type of BPAR. We were unable to examine the role of non-adherence in promoting rejection and graft loss for the entire cohort because this entity can only be measured prospectively through the documentation of pill counts, prescription refill rates, or electronic medication monitoring systems such as mobile apps and bottle cap removal sensors.³⁶ The small proportion of children less than 12 years of age who were transplanted is another limitation; however, this is a systemic rather than a center-specific limitation. Because of delayed referral and non-biopsiable kidneys, there was a large proportion of children in whom the native kidney disease could not be ascertained, which may have limited our analysis of risk factors for graft survival. We could not demonstrate a significant difference in graft or patient survival between living and deceased donor transplants due to the small number of deceased donor transplants in the cohort. We could not report on post-transplant height gain as this was not routinely documented.

To conclude, we report that PAKT is associated with excellent long-term graft outcome, provided factors affecting medication adherence are actively addressed. Patient outcomes remain suboptimal due to a high infection burden, and immunosuppression concentration targets with the current PTM immunosuppression protocol may need to be re-examined in this light.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author [ATV], upon reasonable request.