En-Bloc Kidney Transplantation in the United States: An Analysis of United Network of Organ Sharing (UNOS) Data from 1987 to 2003

Transcription

1 American Journal of Transplantation 2005; 5: Blackwell Munksgaard Copyright C Blackwell Munksgaard 2005 doi: /j x En-Bloc Kidney Transplantation in the United States: An Analysis of United Network of Organ Sharing (UNOS) Data from 1987 to 2003 Vikas R. Dharnidharka a,, Gary Stevens b and Richard J. Howard c a Departments of Pediatrics, b Biostatistics and c Surgery, University of Florida College of Medicine, Gainesville, FL, USA Corresponding author: Vikas R. Dharnidharka, vikasmd@peds.ufl.edu With increasing donor organ shortages, en-bloc kidney (EBK) transplantation is an alternative to utilize very young or very old donor age cadaver kidneys for transplantation. Several single-center series have reported excellent graft survival (GS). We sought to determine national level registry-based patterns for GS and determine adjusted hazard ratios (AHR) for graft loss after EBK versus single kidney (SK) cadaver transplants. Data reported to UNOS from 1987 to 2003 were analyzed using PHREG (SAS version 8.1) statistical procedures. Proportional hazards models were constructed that included multiple donor, recipient and surgical variables. Of the 2160 EBK transplants reported, 77% were from donors < 5 years of age. EBK transplants had superior GS to SK transplants, when donor age was restricted to < 5 years (AHR 0.708, p < 0.001). GS at 1, 3 and 5 years post-transplant was superior with EBK (85%, 76% and 71%) versus SK (81%, 68%, 63% and p < at all time points). EBK transplants from very young donors were associated with a significantly lower rate of delayed graft function than SK transplants (17.9% versus 23.4%, p < 0.001). National registry data suggest that EBK transplants present a viable option for transplantation of very young donor kidneys. Key words: Kidney transplantation, en bloc, dual kidney transplantation, graft survival, delayed graft function, graft thrombosis Received 9 October 2004, revised 12 January 2005 and accepted for publication 27 January 2005 Introduction Pediatric age donor kidneys were initially thought to be most suited for transplantation to pediatric aged recipients, in the belief that optimal-sized kidneys would be a better fit in the abdominal cavity and grow in concordance with the recipient s growth. However, analysis of data by the North American Pediatric Renal Transplant Cooperative Study, demonstrated a marked increase in risk for graft loss due to thrombosis in pediatric recipients of very young donor kidneys (1). These results led to a change in practice patterns and kidneys from very young donors were no longer transplanted to pediatric recipients. Additional results demonstrating other surgical complications in these donor kidneys and poor results due to inadequate nephron mass (2 5) led to a decline in their usage for adult recipient transplantation even after recovery. The recent exacerbations in donor kidney shortages in the Unites States,along with improvements in surgical technique, have led to increased popularity of young donor kidneys being transplanted into adult recipients in en-bloc fashion (6), i.e. where both kidneys are harvested along with the major blood vessels (aorta, inferior vena cava) in one piece and the vascular anastomosis is performed between the larger allograft vena cava and aorta (rather than smaller renal vessels) to the recipient vessels. First reported in a series of patients in 1972, several recent reports from single-center series have demonstrated excellent graft survival(gs) for en-bloc kidney (EBK) transplants in comparison to single kidney (SK) transplants (7 17). Kidneys may also be transplanted in dual or double sequential fashion, where the anastomoses are still performed between allograft renal vessels to recipient iliac vessels and the main advantage relates to increased functioning renal mass since both donor kidneys are given to the same recipient (18 24). To our knowledge, no studies to date have compared the results of EBK to dual kidney (DK) transplants. In this study, we sought to determine national level registry-based patterns for GS of EBK versus single cadaver kidney and dual kidney cadaver transplants and determine adjusted hazard ratios (AHR) for graft loss between 1513

2 Dharnidharka et al. EBK, SK and DK transplants, particularly at very young donor age. Materials and Methods Data for this study were obtained from the United Network of Organ Sharing database and covered the period from October 1, 1987 to November 1, Standard Transplant Analysis (STAR) files in SAS format for transplant and follow-up were obtained and analyzed using SAS version 8.1 statistical software (Cary, NC) and its proportional hazards program PHREG. Transplant type as reported by the individual center to UNOS includes the variables single, double or en-bloc as required fields for transplant type for all cadaver kidney transplants. The variable double in this field represents dual kidney transplants. We then constructed multiple proportional hazards models to calculate AHR for graft loss and constructed Kaplan-Meier probability plots for time-varying adjusted GS. The proportional hazards model assumes that the ratio of the hazards (an approximation of risk ) for a particular outcome at different values of covariates (i.e., predictor variables ) does not change over time. The unadjusted hazard ratio is the ratio of the hazard attributed to one value of a covariate compared to another value of the covariate. The AHR for a covariate is the above ratio adjusted for the influence of other variables included in the multiple regression model. If h(t Z) is the hazard rate for an individual with covariate vector Z, then the basic proportional hazard model is given by h(t Z) = h 0 (t)c(b Z), where h 0 (t) is an arbitrary baseline hazard function, b is the vector of parameters and C(b Z)isaknown function. The most common model for C(b Z) is ( p ) C(b Z) = exp b k Z k. k=1 In this model, the individual parameters, b k, are called the adjusted hazard ratios. AHR values greater than 1.0 imply a significantly higher risk of graft loss if p-value < 0.05 (see Tables 3 and 4). Variables included in these analyses were extensive and included a) donor age, gender, race; b) recipient age, gender, race; c) mismatches between HLA alleles (categorized as zero, 1 3 and 4 6 mismatches, respectively), racial mismatches (African American race to other, other to African American, other to other), percent recipient panel reactive antibody (PRA); d) cold ischemia time in hours, warm ischemia time in minutes, use of perfusion pump or not; e) graft thrombosis and dialysis in the first week post-transplant (used as a surrogate for delayed graft function); f) immunosuppressive induction therapy use (from among any of ALG, ATG, ATGAM, Thymoglobulin, OKT3, daclizumab, basiliximab, campath). Results Demographic trends In the time period analyzed, there were a total of cadaver kidney transplants reported to UNOS. Of these, 2160 were EBK and 448 were DK kidney transplants, the remainder being SK transplants. Table 1 depicts the donor age distribution of the EBK and DK transplants. Donor age was less than 5 years in 77% of all EBK transplants, with the next highest percentage seen in donor age > 60 years (8.56%). In contrast, almost all DK transplants were performed in older donor ages, greater than 60 years (55%), Table 1: Distribution of percentages of en-bloc kidney (EBK) and double kidney (DK) transplants by donor age group Donor age group (years) EBK (n) EBK% DK (n) DK% 0 5 years > 60 years Total Table 2: Distribution of percentages of en-bloc kidney (EBK) and double kidney (DK) transplants by transplant year Transplant En-bloc EBK Double DK year kidney (n) percentage kidney (n) percentage Total years (19%) or years (12%). A small percentage of DK transplants were performed, where donor age was less than 5 years (7.3%, n = 33). Table 2 depicts the distribution of EBK and DK transplants by year of transplant. EBK transplants have been performed since inception of the UNOS registry but 61.8% of all EBK transplants have been performed since 1996 onwards, indicating a rising popularity in recent years. DK transplants were not reported to the UNOS registry at all prior to 1996 and are, thus, a relatively recent procedure. In this study, we have focused on an analysis of EBK transplants from very young donors only. The analysis of EBK transplants from older donors is the subject of a separate study. Impact of transplant type on graft loss Initially, we constructed a proportional hazards model that included all variables listed in the Methods section and did not differentiate between transplant type or donor age. In this model, results for which are shown in Table 3, 1514 American Journal of Transplantation 2005; 5:

3 UNOS Data Analysis of En-Bloc Kidney Transplants Table 3: Adjusted hazard ratios (AHR) for graft loss in all cadaver kidney transplant recipients, all donor age groups and transplant types, n = 116, 467 Variable AHR P-value Recipient age (per unit year increase) <0.001 Donor age (per unit year increase) <0.001 Recipient gender male Donor AA to other race recipient <0.001 Donor other race to AA recipient <0.001 Donor other to other race <0.001 Cold ischemia time (per unit hours increase) <0.001 Warm ischemia time (per unit minute increase) <0.001 PRA (per unit percent increase) <0.001 HLA mismatch (zero, 1 3, 4 6) <0.001 Transplant type (en bloc versus other types) Use of perfusion pump Graft thrombosis occurrence <0.001 Dialysis first week post-transplant <0.001 transplant type (EBK versus other types) was associated with a slightly increased adjusted hazard ratio for graft loss (AHR 1.181, p = ), when accounting for the effects of other variables such as graft thrombosis, delayed graft function, AA race, cold ischemia time, HLA mismatch, etc., all of which have been reported in many studies to be associated with worse GS. It is important to note that the PHREG program provides adjusted hazard ratios for each variable, thus allowing for the effects of transplant type (SK, EBK or DK) to be judged even in the presence of the other variable effects. However, in view of the unique donor populations involved in EBK transplants, we proceeded to construct a separate proportional hazards model that restricted donor age to less than 5 years. Table 4 depicts the AHR and p-values for several variables in a model, where donor age was restricted to less than 5 years and comparison of transplant type was performed between SK and EBK. As shown, EBK Table 4: Adjusted hazard ratios (AHR) for graft loss in cadaver kidney transplant recipients, donor age < 5 years, en-bloc kidney (EBK) versus single kidney (SK) transplant, n = 3957 Variable AHR p-value Recipient age (per unit year increase) Recipient gender male Donor AA to other race recipient Donor other race to AA recipient Donor other to other race Cold ischemia time (per unit hours increase) Warm ischemia time (per unit minute increase) PRA (per unit percent increase) HLA mismatch (zero, 1 3, 4 6) Transplant type (en bloc) <0.001 Use of perfusion pump Graft thrombosis occurrence <0.001 Dialysis first week post-transplant <0.001 Use of induction therapy transplantation resulted in a significant decrease in AHR for overall graft loss (0.708, p < 0.001), even after controlling for the dominant effects of graft thrombosis and delayed graft function and multiple other variables listed in Table 4. Figure 1 depicts the Kaplan-Meier probabilities for adjusted graft survival as determined from the PHREG model. At all time points, GS with EBK was higher than GS with SK in this young donor age group. As shown in Table 5, the specific estimates for GS at time points 1 year, 3 year and 5 year post-transplant were significantly higher for EBK than SK, when donor age was less than 5 years. Frequency of graft thrombosis and delayed graft function based on transplant type Since the technical aspects of EBK are intimately related to avoiding a small anastomotic vessel size, we compared the incidence of graft thrombosis and delayed graft function (DGF) between the different transplant types and across the donor age groups. Among transplants from donors less than 5 years of age, DGF was reported in 23.44% of SK transplants (537/2291) versus 17.89% of EBK transplants (300/1677) versus 21.21% of DK transplants (7/33; p = between groups). Among donors less than 5 years of age, graft thrombosis was reported in 3.1% of SK transplants (71/2291) versus 4.11% of EBK transplants (69/1677). No graft thrombosis events have been reported from 33 DK transplants when donor age was less than 5 years (p < between the groups). Discussion Our analysis of UNOS data shows that EBK transplants are associated with significantly lower hazard for early graft loss in comparison to SK transplants in an adjusted proportional hazards analysis. In transplants from very young donors, EBK clearly offers superior GS to SK transplants. These results are in accord with a prior analysis of UNOS data for GS and donor age performed by Bresnahan et al, in which EBK transplants from donors less than 5 years of age had better GS than SK transplants (25) and significantly lower odds ratio for graft loss of Our analyses extend the work done in that study, particularly since 62% of all EBKs in the UNOS registry have been performed after their study time period of Many single-center studies have documented good GS with EBK transplants (7 15,26 29). In these studies, 1-year GS has ranged from 76 89%, 2-year GS from 77 80% and 5-year GS from 60 89% (8 10,12 14,25). Our study results are in accord with these single-center GS data. In the present study, we have also analyzed for the effects of other many other variables that are known to impact graft outcomes. The superior GS seen with the use of EBK transplant type with the very young donor under 5 years of age appears to be unrelated to a reduction in graft American Journal of Transplantation 2005; 5:

4 Dharnidharka et al. Figure 1: Kaplan-Meier probability plot of adjusted graft survival for en bloc or single kidney transplant where donor age is less than 5 years. Table 5: Graft survival (GS) probability estimates for en-bloc kidney (EBK) transplants compared to single kidney (SK) transplants Survival probability Lower 95% Upper 95% p-value Year Type estimate confidence interval confidence interval for difference 1 year GS Single < Enbloc years GS Single < Enbloc years GS Single < Enbloc thrombosis events, since the incidence of this complication was not clinically significantly different in SK versus EBK transplant. The superior GS with EBK transplants when donor age was less than 5 years may therefore, be partially due to the lower DGF rate seen in EBK transplants versus SK transplants. Since UNOS does not collect data on surgical details such as use of vicryl mesh usage, we are unable to determine if torque of renal vasculature may have led to complications in EBK transplants. Since the focus of our study was graft survival, we have not analyzed for comparisons of renal function. UNOS collects data on serum creatinine values at follow-up, from which it is possible to construct GFR estimates using conversion equations such as Cockroft-Gault or MDRD. However, when pediatric sized kidneys are used in adult recipients, it is unclear if a pediatric equation such as Schwartz formula or the adult conversion equations should be used to estimate GFR. Several previous single-center studies have shown either superior or equivalent Cockroft-Gault estimated GFR values or lower serum creatinine values with EBK transplants versus SK when donor ages were very young (9,10,13). We also could not analyze for center effects in relation to EBK, since data provided by UNOS to outside individual (non-unos or SRTR) investigators have center identifiers removed at the time of data transfer. In summary, our analysis of EBK transplants from a large multi-center national database suggests that national EBK results are excellent in terms of graft survival. In conjunction with the data from single-center studies, EBK transplants should be the preferred type of transplant, when 1516 American Journal of Transplantation 2005; 5:

5 UNOS Data Analysis of En-Bloc Kidney Transplants donor age is less than 5 years. In the current era of severe organ donor shortage, use of en-bloc technique allows for valuable utilization of deceased donor kidneys that might otherwise be discarded. Acknowledgments This study was based on data reported to UNOS from October 1, 1987 to November 1, 2003 and represents the opinions of the authors only. The authors thank Ms. Sarah Taranto at UNOS for providing the data used in this analysis. This study was presented in abstract form at the American Society of Nephrology 2004 annual meeting. References 1. Singh A, Stablein D, Tejani A. Risk factors for vascular thrombosis in pediatric renal transplantation: a special report of the North American Pediatric Renal Transplant Cooperative Study. Transplantation 1997; 63: Hayes JM, Novick AC, Streem SB et al. The use of single pediatric cadaver kidneys for transplantation. Transplantation 1988; 45: Satterthwaite R, Aswad S, Sunga V et al.outcome of en bloc and single kidney transplantation from very young cadaveric donors. Transplantation 1997; 63: Terasaki PI, Gjertson DW, Cecka JM, Takemoto S, Cho YW. Significance of the donor age effect on kidney transplants. Clin Transplant 1997; 11: Modlin C, Novick AC, Goormastic M, Hodge E, Mastrioanni B, Myles J. Long-term results with single pediatric donor kidney transplants in adult recipients. J Urol 1996; 156: Meakins JL, Smith EJ, Alexander JW. En bloc transplantation of both kidneys from pediatric donors into adult patients. Surgery 1972; 71: Chinnakotla S, Leone JP, Taylor RJ. Long-term results of en bloc transplantation of pediatric kidneys into adults using a vicryl mesh envelope technique. Clin Transplant 2001; 15: Gomez Vegas A, Blazquez Izquierdo J, Perez Contin MJ et al. [En bloc renal transplant from infant donors to adults]. Arch Esp Urol 1998; 51: El-Sheikh MF, Gok MA, Buckley PE et al. En bloc pediatric into adult recipients: the Newcastle experience. Transplant Proc 2003; 35: Sanchez-Fructuoso AI, Prats D, Perez-Contin MJ et al. Increasing the donor pool using en bloc pediatric kidneys for transplant. Transplantation 2003; 76: Strey C, Grotz W, Mutz C et al. Graft survival and graft function of pediatric en bloc kidneys in paraaortal position. Transplantation 2002; 73: Ruff T, Reddy KS, Johnston TD et al. Transplantation of pediatric en bloc cadaver kidneys into adult recipients: a single-center experience. Am Surg 2002; 68: Borboroglu PG, Foster CE, 3rd, Philosophe B et al. Solitary renal allografts from pediatric cadaver donors less than 2 years of age transplanted into adult recipients. Transplantation 2004; 77: Merkel FK. Five and 10 year follow-up of En Bloc small pediatric kidneys in adult recipients. Transplant Proc 2001; 33: Varela-Fascinetto G, Bracho E, Davila R et al. En bloc and single kidney transplantation from donors weighing less than 15 kg into pediatric recipients. Transplant Proc 2001; 33: Wright FH, Jr., Banowsky LH, Floyd M, Kothmann R, Houser E. Single pediatric donor kidneys for adult recipients: the San Antonio experience. In: Cecka JM, Terasaki PI, eds. Clin Transpl. Los Angeles: UCLA Tissue Typing Laboratory; 1995: pp Shapiro R, Vivas C, Scantlebury VP et al. Suboptimal kidney donors: the experience with tacrolimus-based immunosuppression. Transplantation 1996; 62: Bunnapradist S, Gritsch HA, Peng A, Jordan SC, Cho YW. Dual kidneys from marginal adult donors as a source for cadaveric renal transplantation in the United States. J Am Soc Nephrol 2003; 14: Infante B, Stallone G, Schena A et al.[double renal transplant. Retrospective analysis of data on the patient population with double kidney transplantation in the setting of the AIRT]. G Ital Nefrol 2003; 20: Dietl KH, Wolters HH, Marschall B, Kisters K, Heidenreich H, Senninger N. [Results of double kidney transplantation from elderly donors]. Langenbecks Arch Chir Suppl Kongressbd 1998; 115: Dietl KH, Wolters H, Marschall B, Senninger N, Heidenreich S. Cadaveric two-in-one kidney transplantation from marginal donors: experience of 26 cases after 3 years. Transplantation 2000; 70: Jerius JT, Taylor RJ, Murillo D, Leone JP. Double renal transplants from marginal donors: 2-year results. J Urol 2000; 163: Remuzzi G, Perico N, Gotti E, Ruggenenti P, Rota G, Locatelli G. The kidney transplant program at the Bergamo Center. Clin Transpl 2000; Remuzzi G, Grinyo J, Ruggenenti P et al. Early experience with dual kidney transplantation in adults using expanded donor criteria. Double Kidney Transplant Group (DKG). J Am Soc Nephrol 1999; 10: Bresnahan BA, McBride MA, Cherikh WS, Hariharan S. Risk factors for renal allograft survival from pediatric cadaver donors: an analysis of united network for organ sharing data. Transplantation 2001; 72: Hiramoto JS, Freise CE, Randall HR et al. Successful long-term outcomes using pediatric en bloc kidneys for transplantation. Am J Transplant 2002; 2: Bretan PN, Koyle M, Singh K et al. Improved survival of en bloc renal allografts from pediatric donors. J Urol 1997; 157: Ratner LE, Flye MW. Successful transplantation of cadaveric enbloc paired pediatric kidneys into adult recipients. Transplantation 1991; 51: Nghiem DD, Schlosser JD, Hsia S, Nghiem HG. En bloc transplantation of infant kidneys: ten-year experience. J Am Coll Surg 1998; 186: s7. American Journal of Transplantation 2005; 5: