First Transplant Factors and the Influence on Kidney Retransplant Outcomes Authors: Ali Sherazi, Karthik Tennankore, Amanda Jean Vinson Published: 2026 ABSTRACT AbstractRationale & ObjectiveRisk factors for adverse outcomes after first transplant have been established, but less is known about predictors of risk at the time of retransplant, including the impact of first transplant characteristics. Therefore, in this study, we aimed to examine how first transplant factors may affect subsequent transplant.Study DesignRetrospective cohort study.Setting & ParticipantsWe analyzed data from adult patients in the United States (Scientific Registry of Transplant Recipients) who underwent second kidney transplant (2000-2017).ExposureFirst transplant characteristics (age, human leukocyte antigen match, panel-reactive antibody [PRA], first transplant graft survival time, and time from failure of first transplant).OutcomeDeath-censored graft loss (DCGL) after repeat transplant.Analytical ApproachMultivariable Cox proportional hazard models.ResultsThis study included 23,688 patients who underwent repeat transplant. DCGL risk was increased for patients with higher PRA at first transplant (adjusted HR [aHR], 1.16; 95% CI, 1.00-1.34 for PRA >80% vs <20%) and decreased for recipients aged 18-30 years at first transplant (aHR, 0.87; 95% CI 0.76-0.99 vs <18 years). Compared with preemptive retransplant, patients requiring a short duration of dialysis after the first transplant failed (<1 month) had the highest risk of DCGL (aHR, 2.36; 95% CI, 2.06-2.71). First transplant survival time was also associated with risk of DCGL (aHR, 0.56 [95% CI, 0.48-0.65] for <1 month; aHR, 0.90 [95% CI, 0.80-1.02] for 1-3 years; aHR, 0.70 [95% CI, 0.62-0.78] for 3-7 years; aHR, 0.50 [95% CI, 0.44-0.57] for >7 years [relative to survival time 1 month-1 year]). Human leukocyte antigen mismatch at first transplant was not significantly associated with DCGL at retransplant.LimitationsPotential for selection bias; patients who died before retransplant or became ineligible for repeat transplant were excluded.ConclusionsFirst transplant PRA >80%, age 18-30 years, time between a first transplant failing and retransplant, and first transplant survival duration are significantly associated with DCGL at kidney retransplantation.Index Words: Graft loss, kidney transplant, prediction, repeat transplant, retransplant Kidney transplantation is a life-saving procedure for individuals with kidney failure. With improvements in immunosuppression and disease management,1 patient survival after transplant has improved, with many patients now outliving their kidney allografts, resulting in an increased proportion of patients requiring retransplant over time.1 Therefore, an examination of factors influencing graft survival for those who are retransplanted is increasingly important. Risk factors for kidney graft loss with the first transplant are well established and include both donor and recipient characteristics. Known factors influencing primary graft survival include donor organ quality (eg, as assessed by the Kidney Donor Profile Index), recipient variables such as age, race, sex, comorbidity burden, and body mass index), and immunological factors like panel-reactive antibody (PRA) levels and human leukocyte antigen (HLA) mismatching (MM).2, 3, 4, 5, 6, 7, 8 Having had a prior transplant has been independently associated with increased risk of subsequent graft failure.9 Although increased sensitization and surgical complexity are felt to contribute to the higher risk demonstrated with repeat versus initial transplant,10 other factors associated with the first transplant have not been studied in detail. Prolonged dialysis and waiting times after a failed transplant have been associated with inferior retransplant outcomes,11, 12, 13 emphasizing the need for timely retransplantation for eligible candidates. Additionally, studies have shown that recipients whose initial graft survival was short (between 30 days and 5 years) face significantly poorer outcomes in subsequent transplants,12,14 with younger recipients at the time of first transplant and recipients older than 50 years at the time of retransplant facing additional challenges.15,16 Meta-analyses suggest that poor HLA matching at the time of initial transplant may contribute to increased sensitization, complicating subsequent transplant attempts.17 Furthermore, allograft nephrectomy post failure has been proposed to affect outcomes, with some studies indicating increased risks of delayed graft function (DGF), acute rejection, and retransplant graft loss,18,19 whereas others report no significant difference.20,21 However, existing studies have been limited by small sample size, have focused on single variables, and have included inadequate adjustment for confounding factors, and/or compared dissimilar populations (those with a first versus subsequent transplant), demanding further comprehensive research. Given the somewhat limited understanding of first transplant factors associated with subsequent retransplant outcomes, this study aims to bridge the knowledge gap by collectively evaluating the effect of first transplant factors (including age, HLA MM, PRA levels at first transplant, duration of first allograft survival, and the time between the first transplant failure and retransplantation) on retransplant outcomes. Methods Subject Selection This study included adult patients who underwent a second living or deceased donor kidney transplant in the United States between January 1, 2000, and December 31, 2016, identified using the Scientific Registry of Transplant Recipients (SRTR). We excluded patients aged <18 years, those undergoing simultaneous or sequential transplant, and those without a prior history of or with >1 prior kidney transplant. Patients with a prior transplant but no details regarding first transplant characteristics were also excluded. Data Collection Baseline characteristics were compiled for all patients undergoing retransplant including characteristics of the retransplant (donor and recipient sex, age, and race/ethnicity, recipient cause of original kidney failure, HLA MM, PRA, comorbid conditions (including coronary artery disease, peripheral vascular disease, diabetes, and hypertension), donor type (deceased vs living), warm ischemia time, and cold ischemia time. In addition, characteristics of the first transplant were collected, including recipient age, HLA MM, and PRA at time of first transplant, first allograft survival duration, and time since the first transplant failed. We used multiple imputation with chained equations and 10 imputations to account for missing data. Exposure The exposures of interest were characteristics of the first kidney transplant including age (<18 [reference], 18-30, 30-50, >50 years), HLA MM (0 mismatches [reference], >0 mismatches), PRA at time of first transplant (<20% [reference], 20%-80%, and >80%), first allograft survival duration (<1 month, 1 month-1 year [reference], 1-3 years, 3-7 years, and >7 years), and time since the first transplant failed (preemptive retransplant; no interval dialysis [reference], <1 month, 1 month-1 year, 1-3 years, 3-5 years, and >5 years). In separate analyses, if the association between death-censored graft loss (DCGL) and first allograft survival and/or time since the first allograft failed was linear, we planned to also examine these as continuous rather than categorical variables (with preemptive kidney transplant treated as occurring within 0.1 days). Outcome The primary outcome was DCGL, defined as return to dialysis or preemptive retransplantation in patients with a second kidney transplant. Secondary outcomes included all-cause graft loss (ACGL; the composite of graft loss or death) and DGF (defined as need for dialysis in the first week after transplant). Statistical Analysis Descriptive statistics were reported for all patients at the time of their first and repeat transplants. The duration the first allograft survived as well as time since the first transplant failed were captured with characteristics available at the time of retransplant. Medians and interquartile ranges (IQRs) were calculated for both first and retransplant continuous variables, whereas counts and percentages were reported for categorical variables. Significant differences between first and retransplant characteristics were identified using the Pearson χ2 test or the Wilcoxon rank-sum test as appropriate. Analyses In our primary analysis, multivariable Cox proportional hazards models were used to evaluate the independent association between first transplant characteristics (age, HLA MM, PRA, first transplant graft survival time, and time from failure of first transplant) and DCGL after retransplant, additionally adjusting for factors and demographics at retransplant including donor and recipient age at retransplant, race, sex, comorbid conditions (coronary artery disease, hypertension, peripheral vascular disease, diabetes), native kidney failure diagnosis, PRA and donor type. In a secondary analysis, we similarly examined the outcome of ACGL. Finally, we used multivariable logistic regression to examine the independent association between first transplant characteristics and DGF, after accounting for the other patient factors present at the time of retransplant. Assessment of Predictive Capacity of Prior Transplant Factors To evaluate whether incorporating first transplant characteristics (age, HLA MM, PRA, first transplant graft survival time, and time from failure of first transplant) improved the ability to predict DCGL at retransplant, we assessed model fit (using the Akaike Information Criterion [AIC]) and discrimination (using Harrell’s C-index) in a base model (adjusting for only patient factors present at retransplant and not prior transplant factors) and in the updated model (adjusting for both current patient factors at retransplant and prior transplant details). The AIC is used to test the relative fit of nonnested models for predicting an outcome in a given data set. It offers a trade-off between model fit and model simplicity (penalizing models for including covariates with little predictive capacity) with the best model having the lowest AIC among all other models.7,22 To compare AICs between models, models must include the same population; therefore, patients with missing first transplant details were excluded from both models. This study used data from the SRTR. The SRTR data system includes data on all donors, wait-listed candidates, and transplant recipients in the United States, submitted by the members of the Organ Procurement and Transplantation Network. The Health Resources and Services Administration of the US Department of Health and Human Services provides oversight to the activities of the Organ Procurement and Transplantation Network and SRTR contractors. The data reported here were supplied by the Hennepin Healthcare Research Institute as the contractor for the SRTR. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy of or interpretation by the SRTR or the US Government. All statistical analyses were performed using Stata (Stata Corp). A P value < 0.05 was considered statistically significant. The study protocol was submitted to the Nova Scotia Health Research Ethics Board, who deemed the study met requirements outlined in the Tri-Council Policy Statement for an REB review exemption (REB FILE #: 1028871) and that written consent was waived. Results Baseline Characteristics After relevant exclusions, a total of 23,688 people underwent a second kidney transplant over the study period and were included for analysis. Cohort derivation is shown in Figure S1. Baseline characteristics present at the time of second kidney transplantation are shown in Table 1, along with characteristics for those same patients at the time of their first transplant. There were significant differences between first transplant and retransplant characteristics. Median age at transplant was significantly younger at first (34 years [IQR, 24-45]) than second (45 years [IQR, 36-55]) transplant, P < 0.001. Peak PRA was lower with the first than the second transplant (2% vs 59%, P < 0.001), and comorbidity burden was greater at the time of retransplant (coronary artery disease, 2.7% vs 5.3%; peripheral vascular disease, 0.9% vs 4.0%; diabetes, 11.5% vs 18.9%; and hypertension, 55.6% vs 73.8%; all P < 0.001). By definition, the cause of kidney failure for all retransplant recipients was a prior failed graft; however, original causes of native kidney failure are shown in Table 1. The leading cause of original kidney failure was glomerulonephritis (38.6%). Missing data are listed in the Table 1 footnote; although missingness was substantial for some variables (eg, coronary artery disease missing in 34.6% of patients at first transplant and 20.0% at retransplant; hypertension missing in 30.0% of patients at first transplant and 14.3% at retransplant), it was less for those variables that were ultimately included in multivariable modeling, and missing data were treated with multiple imputation with chained equations. Table 1Baseline Characteristics of Study ParticipantsFactorFirst TransplantN = 23,688Subsequent TransplantN = 23,688Recipient sex Male13,958 (58.9%)13,960 (58.9%) Female9,730 (41.1%)9,728 (41.1%)Recipient age (y)34 (24-45)45 (36-55)HLA mismatch 01,847 (7.8%)3,321 (14.0%) 11,370 (5.8%)1,182 (5.0%) 23,401 (14.4%)2,280 (9.6%) 35,845 (24.7%)4,343 (18.3%) 44,592 (19.4%)5,235 (22.1%) 54,235 (17.9%)5,094 (21.5%) 62,055 (8.7%)2,138 (9.0%)PRA group 0%-20%20,626 (87.1%)8,701 (36.7%) 20%-80%1,444 (6.1%)5,732 (24.2%) >80%389 (1.6%)3,974 (16.8%)Peak PRA (%)2 (0-10)59 (6-92)Recipient race White16,795 (70.9%)16,719 (70.6%) Black5,713 (24.1%)5,715 (24.1%) Other1,180 (5.0%)1,254 (5.3%)Coronary artery disease633 (2.7%)1,262 (5.3%)Peripheral vascular disease220 (0.9%)947 (4.0%)Diabetes mellitus2,717 (11.5%)4,469 (18.9%)Hypertension13,178 (55.6%)17,483 (73.8%)Kidney failure diagnosis Glomerulonephritis9,138 (38.6%)9,036 (38.2%) Diabetes2,246 (9.5%)2,358 (10.0%) Polycystic kidney disease1,616 (6.8%)1,611 (6.8%) Hypertension/renovascular3,969 (16.8%)4,413 (18.6%) Other4,827 (20.4%)4,948 (20.9%)Donor sex Male12,187 (51.5%)13,037 (55.0%) Female11,501 (48.5%)10,651 (45.0%)Donor age (y)39 (27-49)38 (25-48)Donor race White19,698 (83.2%)19,967 (84.3%) Black3,242 (13.7%)2,947 (12.4%) Other707 (3.0%)772 (3.3%)Donor type Deceased13,093 (55.3%)16,470 (69.5%) Living10,595 (44.7%)7,218 (30.5%)Warm ischemic time (min)30 (15-44)35 (28-46)Cold ischemic time (h)12.3 (1-22)14 (4.4-21)Time from transplant failure Preemptive—3,526 (14.9%) <1 mo—1,962 (8.3%) 1 mo-1 y—2,616 (11.0%) 1-3 y—4,832 (20.4%) 3-5 y—3,410 (14.4%) >5 y—7,342 (31.0%)Survival time of first transplant <1 mo—1,651 (7.0%) 1 mo-1 y—1,643 (6.9%) 1-3 y—3,789 (16.0%) 3-7 y—7,370 (31.1%) >7 y—9,235 (39.0%)Note: Values are n (%) or median (interquartile range). Missing: HLA mismatch (n = 343 [1.4%] first transplant; n = 95 [0.4%] retransplant); PRA (n = 1,229 [5.2%] first transplant; n = 5,281 [22.3%] retransplant); donor race (n = 41 [0.2%] first transplant; n = 2 [0.0%] retransplant); coronary artery disease (n = 8,197 [34.6%] first transplant; n = 4,725 [20.0%] retransplant); peripheral vascular disease (n = 7,909 [33.4%] first transplant; n = 995 [4.2%] retransplant); diabetes (n = 6,726 [28.4%] first transplant; n = 443 [1.9%] retransplant); hypertension (n = 7,101 [30.0%] first transplant; n = 3,377 [14.3%] retransplant); cause of kidney failure (n = 1,892 [8.0%] first transplant; n = 1,322 [5.6%] retransplant); warm ischemia time (n = 6,818 [28.7%] first transplant; n = 10,731 [45.3%] retransplant); cold ischemia time (n = 3,660 [15.5%] first transplant; n = 2,586 [10.9%] retransplant; donor age (n = 3 [0.0%] first transplant).Abbreviations: HLA, human leukocyte antigen; PRA, panel-reactive antibody. Baseline Characteristics of Study Participants Note: Values are n (%) or median (interquartile range). Missing: HLA mismatch (n = 343 [1.4%] first transplant; n = 95 [0.4%] retransplant); PRA (n = 1,229 [5.2%] first transplant; n = 5,281 [22.3%] retransplant); donor race (n = 41 [0.2%] first transplant; n = 2 [0.0%] retransplant); coronary artery disease (n = 8,197 [34.6%] first transplant; n = 4,725 [20.0%] retransplant); peripheral vascular disease (n = 7,909 [33.4%] first transplant; n = 995 [4.2%] retransplant); diabetes (n = 6,726 [28.4%] first transplant; n = 443 [1.9%] retransplant); hypertension (n = 7,101 [30.0%] first transplant; n = 3,377 [14.3%] retransplant); cause of kidney failure (n = 1,892 [8.0%] first transplant; n = 1,322 [5.6%] retransplant); warm ischemia time (n = 6,818 [28.7%] first transplant; n = 10,731 [45.3%] retransplant); cold ischemia time (n = 3,660 [15.5%] first transplant; n = 2,586 [10.9%] retransplant; donor age (n = 3 [0.0%] first transplant). Abbreviations: HLA, human leukocyte antigen; PRA, panel-reactive antibody. Association Between First Kidney Transplant Characteristics and DCGL The adjusted associations between first kidney transplant characteristics and DCGL at the time of retransplant are shown in Table 2. Recipients aged 18-30 years at first transplant were at a lower risk of DCGL than when the first transplant occurred before 18 years (adjusted hazard ratio [aHR], 0.87; 95% confidence interval [CI], 0.76-0.99; P = 0.04), with no differences by age at first transplant otherwise. Relative to a PRA <20%, PRA >80% at the time of first transplant was also associated with an increased risk of DCGL at retransplant (aHR, 1.16; 95% CI, 1.00-1.34; P = 0.04). There were no associations between HLA MM at first transplant and subsequent DCGL with retransplant. Relative to a first allograft surviving 1 month-1 year, there was a U-shaped association between first allograft survival and risk of subsequent DCGL with retransplant; lowest risk was among those with a first allograft surviving <1 month or >7 years (aHR, 0.56 [95% CI, 0.48-0.65] for <1 month; aHR, 0.90 [95% CI, 0.80-1.02] for 1-3 years; aHR, 0.70 [95% CI, 0.62-0.78] for 3-7 years; and aHR, 0.50 [95% CI, 0.44-0.57] for >7 years). Finally, the interval between the failure of the first transplant and subsequent retransplant was also significantly associated with graft outcomes. Relative to preemptive retransplant, transplants performed <1 month after the first allograft failure were at the highest risk for DCGL (aHR, 2.36; 95% CI, 2.06-2.71), with an overall increasing risk of DCGL with retransplant from 1 month after failure until >5 years after failure, when the risk began to improve again (aHR, 1.12 [95% CI, 0.97-1.31] for those transplanted 1 month-1 year after failure; aHR, 1.34 [95% CI, 1.18-1.53] for 1-3 years after failure; aHR, 1.43 [95% CI, 1.25-1.64] for 3-5 years after failure; and aHR, 1.27 [95% CI, 1.12-1.45] for retransplant >5 years after failure). Given the nonlinear association between first allograft survival and time from first allograft failure with DCGL, these first transplant variables were ultimately not examined continuously. Table 2Adjusted Associations Between First Kidney Transplant Characteristics and Death-Censored Graft Loss After a Second Kidney TransplantVariableHR (95% CI)PAge (y) <18Ref 18-300.87 (0.76-0.99)0.04 30-500.96 (0.80-1.15)0.65 >501.11 (0.84-1.47)0.45HLA match at first transplant 0 mismatchesRef >0 mismatches0.92 (0.81-1.04)0.19PRA at first transplant <20%Ref 20%-80%1.03 (0.93-1.13)0.62 >80%1.16 (1.00-1.34)0.04First transplant survival time 1 mo-1 yRef <1 mo0.56 (0.48-0.65)<0.001 1-3 y0.90 (0.80-1.02)0.09 3-7 y0.70 (0.62-0.78)<0.001 >7 y0.50 (0.44-0.57)<0.001Time from transplant failure PreemptiveRef <1 mo2.36 (2.06-2.71)<0.001 1 mo-1 y1.12 (0.97-1.31)0.13 1-3 y1.34 (1.18-1.53)<0.001 3-5 y1.43 (1.25-1.64)<0.001 >5 y1.27 (1.12-1.45)<0.001Note: Model was adjusted for first transplant factors including recipient age, HLA mismatch status, PRA, first transplant survival time, time from first transplant failure, and retransplant factors including recipient age, donor age, recipient race, donor race, recipient sex, donor sex, comorbid conditions (coronary artery disease, hypertension, peripheral vascular disease, diabetes), PRA at retransplant, donor type, and native kidney failure diagnosis.Abbreviations: CI, confidence interval; HLA, human leukocyte antigen; HR, hazard ratio; PRA, panel-reactive antibody; Ref, reference. Adjusted Associations Between First Kidney Transplant Characteristics and Death-Censored Graft Loss After a Second Kidney Transplant Note: Model was adjusted for first transplant factors including recipient age, HLA mismatch status, PRA, first transplant survival time, time from first transplant failure, and retransplant factors including recipient age, donor age, recipient race, donor race, recipient sex, donor sex, comorbid conditions (coronary artery disease, hypertension, peripheral vascular disease, diabetes), PRA at retransplant, donor type, and native kidney failure diagnosis. Abbreviations: CI, confidence interval; HLA, human leukocyte antigen; HR, hazard ratio; PRA, panel-reactive antibody; Ref, reference. Assessment of Predictive Capacity of Prior Transplant Factors To assess whether accounting for prior transplant characteristics improved model fit and discrimination for predicting DCGL, we assessed the AIC and Harrell’s C-index for a multivariable model that accounted only for patient factors present at retransplant (not prior transplant details) and subsequently, a multivariable model accounting for both prior transplant characteristics and the same baseline patient factors present at retransplant. Based on study design, all patients required a prior transplant to be included in the study; therefore, we did not adjust for prior transplant status (yes/no) in the baseline model. Including prior transplant details improved the AIC from 49,600.5 to 49,306.8, indicating better predictive capacity while maintaining parsimony. Model discrimination was also improved from a C-index of 0.642 in the model that did not include prior transplant variables to 0.674 in the model that also adjusted for first transplant characteristics. Association Between First Kidney Transplant Characteristics and Secondary Outcomes: ACGL First transplant characteristics were also significantly associated with ACGL (Table S1). Similar to the analysis of DCGL, age (aHR, 0.80 [95% CI, 0.72-0.89] for 18-30 years and aHR, 0.81 [95% CI, 0.71-0.93] for 30-50 years, relative to age <18 years) and PRA (aHR 1.14, 95% CI 1.02-1.27 for PRA >80%, relative to PRA <20%) at first transplant (but not HLA MM) were associated with subsequent ACGL with retransplant. The association between first transplant survival duration (aHR, 0.49 [95% CI, 0.43-0.55] for <1 month; aHR, 0.99 [95% CI, 0.90-1.09] for 1-3 years; aHR, 0.83 [95% CI, 0.76-0.91] for 3-7 years; and aHR, 0.71 [95% CI, 0.65-0.78] for >7 years, relative to a survival time of 1 month-1 year) and the interval between failure of the first transplant and subsequent retransplant (aHR, 2.69 [95% CI, 2.47-2.94] for those retransplanted <1 month after failure; aHR, 0.99 [95% CI, 0.89-1.10] for 1 month-1 year after failure; aHR, 1.18 [95% CI, 1.08-1.29] for 1-3 years after failure; aHR, 1.24 [95% CI 1.13-1.37] for 3-5 years after failure; and aHR, 1.19 [95% CI, 1.09-1.30] for retransplant >5 years after failure, relative to preemptive retransplant) were both significantly associated with the risk of ACGL with the same patterns as observed for DCGL. DGF The adjusted associations between first kidney transplant characteristics and DGF at the time of retransplant are shown in Table S2. Age, PRA, and HLA MM at time of first transplant were not associated with risk of DGF. Relative to a first allograft survival duration of 1 month-1 year, those with a first allograft surviving >7 years had a lower risk of DGF (adjusted odds ratio [aOR], 0.78; 95% CI, 0.67-0.90), with no other significant associations. Finally, the interval between failure of the first transplant and subsequent retransplant was strongly associated with risk of DGF (aOR, 3.64 [95% CI, 2.95-4.50] for retransplant performed in <1 month; aOR, 3.08 [95% CI, 2.49-3.79] for those transplanted 1 month-1 year after failure; aOR, 4.04 [95% CI, 3.35-4.86] for 1-3 years after failure; aOR, 4.59 [95% CI, 3.80-5.54] for 3-5 years after failure; and aOR, 4.64 [95% CI, 3.87-5.58] for retransplant >5 years after failure, relative to those undergoing preemptive re-transplant). Discussion In this large cohort study, we demonstrate for the first time, to our knowledge, that a constellation of factors present at the time of first kidney transplant affect subsequent transplant outcomes. These include recipient age and PRA levels at first transplant, duration of first allograft survival, and the time between first transplant failure and retransplantation, each of which was independently associated with the risk of DCGL and ACGL after retransplantation. Notably, HLA match status at first transplant was not associated with subsequent transplant outcomes. Importantly, we also demonstrate that accounting for these first transplant factors improved the prediction of DCGL, as evidenced by increased model discrimination (C-index improved from 0.642 to 0.674) and better model fit (lower AIC from 49,600.5 to 49,306.8), indicating improved predictive capacity of the model while maintaining parsimony. These findings suggest that first transplant characteristics should be considered when evaluating risk profiles for candidates at the time of retransplantation. Although previous studies have largely focused on individual risk factors from the first transplant, the present study is the first to comprehensively examine multiple first transplant characteristics as an aggregate and shows that these collectively enhance the ability to predict outcomes at the time of retransplant. These findings suggest that incorporating a more holistic view of patients’ transplant history, rather than isolated variables, should be considered when examining risk at the time of retransplantation. Our study aligns with previous research indicating that younger age at first transplant is associated with higher risk of graft loss upon retransplantation.15,23 We show that patients aged 18-30 years at first transplant experience significantly less risk with subsequent transplant than those aged <18 years. It is known that younger recipients experience more allograft rejection than older recipients; this has been largely attributed to immunosenescence in older individuals with a waning innate and adaptive immune response with advancing age.24 Nonadherence to immunosuppressive therapy, which is more prevalent among younger patients, may also contribute to this greater rejection risk and reduced first allograft survival in younger than older recipients.23,25 Whether increased retransplant risk attributable to young age at first transplant is due to increased incidence of first transplant rejection or through another mechanism requires further study, although notably, the age-related risk was independent of differences in PRA at first transplant and retransplant. Although increased sensitization is known to associate with graft loss at the time of first transplant,10 higher PRA levels at the time of first transplant have also been associated with an increased risk of graft loss after retransplant.9,25 However, these studies have been small scale with PRA measurement most often occurring at the time of first allograft failure and/or nephrectomy.26 To our knowledge, no prior study has examined PRA preceding the first transplant and how that may influence subsequent retransplant outcomes. In our study, a PRA>80% at first transplant remained independently associated with both DCGL and ACGL at retransplantation, even after adjusting for updated PRA at retransplant. In addition to recipient age and PRA at first transplant, the length of time the first allograft survived was also significantly associated with outcomes after retransplant. Both very short (<1 month) and very long (>7 years) first graft survival times were associated with better retransplant outcomes compared with intermediate durations. This U-shaped relationship may reflect that early graft loss is often due to nonimmunologic factors and possibly technical losses, which may have less bearing on retransplant outcomes, whereas longer graft survival may indicate a population of patients more compliant with immunosuppression, less comorbid, and more health conscious, possibly favoring better outcomes in subsequent transplants.12,16,27 Reasons for this require further study. Finally, the time between first graft failure and retransplantation was also noted to be associated with retransplant outcomes. Preemptive retransplantation (before the need for dialysis after the first graft failed) was associated with the lowest risk of graft loss, consistent with prior studies emphasizing the benefits of minimizing dialysis exposure at the time of first transplant failure and regrafting.11, 12, 13,28,29 Delays in retransplantation may lead to increased sensitization and accrual of dialysis-related comorbidity burden, adversely affecting graft survival.13 These findings are similar to those observed for transplant-naïve patients, for whom a longer dialysis vintage before transplant is an established risk factor for worse graft survival.30,31 Contrary to several earlier studies suggesting that HLA MM at first transplant may contribute to poorer outcomes upon retransplant due to increased sensitization risk,12,17,23 we did not find a significant association between HLA MM at first transplant and either DCGL, ACGL, or DGF at retransplant. This may be partly explained by the adjustment for PRA levels at retransplant in our models, which likely captures the immunologic risk conferred by prior MM. Also, since our study included only those who underwent second transplant, it is possible that patients with poor HLA matching at the time of first transplant who became very highly sensitized after their first graft failure could not undergo retransplant on account of prohibitory antibodies, and therefore would not be included here. Improved immunosuppressive protocols and desensitization techniques over the past decade may also mitigate the risks associated with HLA MM.25 The improvement in predictive performance of our models by including first transplant characteristics underscores their clinical relevance. The increased C-index and lower AIC suggest when they are accounted for, first transplant factors add valuable information beyond the currently considered recipient and donor characteristics. Incorporating these variables into risk assessment tools could enhance the selection and management of retransplant candidates, potentially leading to personalized immunosuppression strategies and better allocation of donor kidneys. It is worth mentioning that patients whose first graft lasted longer than 7 years were at lower risk for DGF, and patients who underwent non-preemptive retransplant after a failed first graft were at a consistently higher risk for DGF than preemptive retransplant recipients. The strong benefit of preemptive retransplant (vs retransplant after an interval of dialysis) likely mirrors the lower risk of DGF in patients undergoing first preemptive (vs non-preemptive) kidney transplant, largely attributable to residual renal function and fewer dialysis-related comorbid conditions. Other characteristics of the first transplant were not associated with risk of subsequent DGF at the time of retransplant. Our study has several strengths, including being the first to examine the association between an aggregate of first transplant characteristics and subsequent retransplant outcomes and the relative impact on ability to predict retransplant graft loss when these first transplant factors are considered versus when they are not. However, limitations exist. As a retrospective study using registry data, there is potential for misclassification and miscoding of variables, although such errors are likely to occur at random. Importantly, patients who died before retransplantation were excluded, which may introduce selection bias; those with the highest-risk first transplant characteristics might be underrepresented due to mortality before retransplant listing. Similarly, as discussed above, patients who were not able to receive a repeat transplant (whether due to prohibitive sensitization or comorbidity burden) were also excluded; this too can result in a biased population. We lacked granular data on reasons for first graft loss and adherence to immunosuppressive therapy, which could also influence retransplant outcomes. In particular, younger age and higher PRA at first transplant were both associated with worse outcomes at the time of retransplant, both of which may be expected to increase rejection risk with first transplant. Identifying whether a first transplant lost to allograft rejection is associated with worse subsequent retransplant outcomes is a point of future study. Likewise, although we were able to adjust for HLA matching at the time of first and repeat transplant, we were unable to examine the effect of repeated HLA MM between first and second transplant organs, which may have implications for later rejection episodes. Data were not available to characterize early graft loss (<1 month) as primary nonfunction, hyperacute rejection, or technical losses and whether the patient underwent subsequent graft nephrectomy—all of which may differentially impact subsequent graft outcomes. Notably, although early graft loss may be caused by hyperacute or accelerated rejection, the most common causes are nonimmunologic (most commonly vascular thrombosis or ureteric operative complications),32,33 and we hypothesize it is these technical losses that drive the signal of lower risk in patients with early graft loss. Finally, the generalizability of our results to health care systems outside the United States may be limited due to differences in transplantation practices and access. Future research should aim to validate our findings in other populations and explore strategies to mitigate the risks associated with unfavorable first transplant characteristics. Prospective studies assessing interventions to reduce sensitization, such as novel desensitization protocols25 and improved access to timely retransplantation, especially preemptive transplant and retransplant, are warranted. Additionally, investigating the effect of adherence-enhancing interventions in younger patients may be beneficial.23,34 As the number of patients requiring retransplantation continues to increase, understanding and addressing factors influencing graft survival is becoming increasingly important. In conclusion, our analysis demonstrates for the first time that a constellation of first transplant characteristics including recipient age, PRA levels, graft survival duration, and time to retransplant are significant predictors of graft loss after kidney retransplantation. Including these factors in risk assessment models improves the prediction of adverse outcomes, highlighting their importance in the evaluation and management of retransplant candidates. This is of particular importance as the number of patients requiring and undergoing retransplant continues to increase over time. Article Information Authors’ Full Names and Academic Degrees Ali Sherazi, MD, Karthik Tennankore, MSc, MD, and Amanda Jean Vinson MSc, MD. Authors’ Contributions Conceptualization and study design: data acquisition: data analysis/interpretation: statistical analysis: supervision or mentorship: AV designed the study and conducted all statistical analyses. AS wrote the initial draft with feedback from AV and KT. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. Support None. Financial Disclosure AV has done consultant work for Paladin Labs Inc and Takeda Pharmaceuticals, and KT has received funding from Otsuka for an investigator-initiated study. The remaining authors declare that they have no relevant financial interests. Data Sharing Data are available through the Scientific Registry of Transplant Recipients with appropriate permissions. Peer Review Received December 8, 2025. Evaluated by 2 external peer reviewers, with direct editorial input from an Associate Editor and the Editor-in-Chief. Accepted in revised form March 13, 2026. Footnotes Complete article and author information provided before references. Supplementary File (PDF). Figure S1: Flow diagram for cohort derivation Table S1: Adjusted Associations between First Kidney Transplant Characteristics and All-Cause Graft Loss Following a Second Kidney Transplant Table S2: Adjusted Associations between First Kidney Transplant Characteristics and Delayed Graft Function Following a Second Kidney Transplant Supplementary Material Supplementary File (PDF) Figure S1; Table S1-S2. References Associated Data Supplementary Materials Supplementary File (PDF) Figure S1; Table S1-S2.