SEARCH

SEARCH BY CITATION

Keywords:

  • Kidney transplantation;
  • living donor transplantation;
  • marginal donors;
  • recipient outcome;
  • systematic review

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

Older individuals or those with medical complexities are undergoing living donor nephrectomy more than ever before. Transplant outcomes for recipients of kidneys from these living expanded criteria donors are largely uncertain. We systematically reviewed studies from 1980 to June 2008 that described transplant outcomes for recipients of kidneys from expanded criteria living donors. Results were organized by the following criteria: older age, obesity, hypertension, reduced glomerular filtration rate (GFR), proteinuria and hematuria. Pairs of reviewers independently evaluated each citation and abstracted data on study and donor characteristics, recipient survival, graft survival, serum creatinine and GFR. Transplant outcomes for recipients of kidneys from older donors (≥60 years) were described in 31 studies. Recipients of kidneys from older donors had poorer 5-year patient and graft survival than recipients of kidneys from younger donors [meta-analysis of 12 studies, 72% vs. 80%, unadjusted relative risk (RR) of survival 0.89, 95% confidence interval (CI) 0.83–0.95]. In meta-regression, this association diminished over time (1980s RR 0.79, 95% CI 0.65–0.96 vs. 1990s RR 0.91, 95% CI 0.85–0.99). Few transplant outcomes were described for other expanded criteria. This disconnect between donor selection and a lack of knowledge of recipient outcomes should give transplant decision-makers pause and sets an agenda for future research.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

Many transplant centers have expanded their selection criteria for living donors in response to long wait times for deceased donor kidneys, the success of living donation and a greater emphasis on donor autonomy (1). Individuals with historic contraindications to kidney donation are now accepted as living ‘expanded criteria donors’. They have also been referred to in the literature as donors with ‘medical complexities’ (2), ‘isolated medical abnormalities (IMAs)’ (3), ‘asymptomatic urinary abnormalities’ (4), ‘marginal’ or ‘at incremental risk’ (2,5,6). These donors are older, obese or have conditions such as hypertension, a reduced glomerular filtration rate (GFR), proteinuria or hematuria at the time of donation. Studies examining donor outcomes for individuals with these criteria have recently been summarized (3).

In deceased donation, the quality of the donated organ has significant implications on graft outcomes. Similarly, a better understanding of how a recipient's outcome is impacted by receipt of an organ from a living expanded criteria donor would improve donor selection and transplant decision-making. We conducted this review to comprehensively assemble all relevant literature on this topic. We also aimed to better understand study results, reasons for observed differences between studies, and to identify current gaps in knowledge to guide future research and clinical practice.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

We conducted this systematic review and meta-analysis according to a prespecified protocol, which was guided by the Meta-analysis of Observational Studies in Epidemiology consensus statement (7).

Study selection

Living expanded criteria donors were defined as having any of the following prior to the time of donation: (1) age > 60 years old; (2) obesity: BMI either >30 kg/m2 or >35 kg/m2; (3) hypertension: >140/90 mmHg or on blood pressure medication; (4) low GFR (as defined by the authors); (5) proteinuria: either ≥150 mg/day or ≥300 mg/day or (6) microscopic hematuria. With the exception of lower GFR, we defined the donor complexities according to clear-cutoff points (i.e. BMI >30 kg/m2) to facilitate generalizability to routine care. We included full-text cohort studies published in any language if they examined at least 20 living kidney donors, with 10 or more being expanded criteria donors. We excluded smaller sized studies from this review to avoid less reliable estimates. Due to the growing interest of the impact of donor GFR on recipient outcome, we relaxed the criteria regarding the number of donors necessary in each study examining GFR. All studies were required to assess at least one of the following recipient outcomes a year or more after transplant: patient survival, graft survival, serum creatinine or GFR. Improvements in graft survival have been nominal since the start of the cyclosporine era in the 1980s (8). Thus, to limit secular effects of changing immunosuppression regimens, we only accepted studies that recruited recipients after 1980.

Identifying relevant studies

We systematically searched MEDLINE, PreMedline, Experta Medica (EMBASE), Cochrane library, BIOSIS Previews, the ISI Science Citation Index Expanded, Google Scholar, Elsevier's scientific search engine SCIRUS, Cochrane Renal Group's Register and SCOPUS electronic databases, from the inception of each source to June 2008.

The search strategy included, but was not limited to, the terms ‘kidney’, ‘renal’, ‘graft’, ‘transplant’, ‘donor’, ‘living’, ‘age’, ‘GFR’, ‘serum creatinine’, ‘obesity’ and ‘hypertension’. These keywords were used in combination and with their corresponding subject headings. The selected subject headings were exploded in order to include all narrower subheadings. The search strategy was tailored for each database and was performed with the help of an experienced information specialist.

We performed cross-reference searches and citation tracking using the ISI Science Citation Index and SCOPUS, and used the ‘related articles’ features in PubMed, Elsevier's SCIRUS, and Google Scholar. An additional search for relevant reviews was also conducted. The reference lists of all reviews and pertinent articles were evaluated to identify additional articles meeting our inclusion criteria.

All citations were downloaded into Reference Manager, version 11.0 (Thomson ISI Research-Soft, Philadelphia, PA).

Article eligibility criteria

Pairs of reviewers (YI/NS/AY/JJ/AG) independently evaluated the eligibility of each citation. Ineligible articles were excluded on the basis of title or abstract. Any potentially relevant citation was retrieved in full text for more detailed evaluation. Pairs of reviewers independently evaluated the eligibility of all English language articles; disagreements were resolved by consensus. For all non–English-language articles, a single reviewer evaluated eligibility with the aid of a translator. When the same group of donors was described in multiple publications, we reviewed all of the publications and cited the article with the most complex donors, or the longest study duration if the sample size was equal.

Data abstraction

After pilot-testing the data abstraction forms, data was extracted from each eligible article independently by pairs of reviewers (YI/NS/AY/JJ/AG). Differences were resolved by consensus. The following information was abstracted from each study: study characteristics, methodological quality, baseline donor and recipient characteristics, and recipient outcomes. Quality appraisal of the studies was guided by Hayden et al. (9). When a study described more than one medical complexity, data on each medical complexity were abstracted independently. The primary outcome was the composite of patient and graft survival, also referred to as uncensored graft survival (graft considered failed if a recipient died). In instances where we could not discern if graft survival was death censored, we assumed graft survival was uncensored for recipient death unless a primary author indicated otherwise. Graft survival was considered death censored when we could determine censoring from the survival value, or the terms ‘functional graft survival’ or ‘graft function rate’ were used. When available, we abstracted both short- (1 year) and long-term (5 years) outcomes. We attempted to contact primary authors of all included studies to confirm data, to clarify interpretations or to provide additional data. Where continuous data were not reported in text but displayed in graph form, we abstracted information from the graph and asked the primary authors to verify the values. Events were estimated from Kaplan-Meier graphs when no other information was available. It is recognized that this tends to overestimate the number of events but would not influence the association between the presence of medical complexity and outcome (10). Recipient serum creatinine and GFR levels were not considered beyond 1 year, as reported results did not account for competing events of recipient death or graft failure.

Statistical analyses

Reviewer agreement on study eligibility was quantified using the kappa statistic (11). For study-level meta-analysis, the I2 statistic was calculated to determine the level of between-study heterogeneity (12). The I2 statistic describes the percentage of total variation across studies that is due to heterogeneity rather than chance; I2 values of 25%, 50% and 75% were indicative of low, moderate and extreme heterogeneity, respectively (12).

Meta-analyses of the relative risk (RR) for dichotomous outcomes between donors with and without expanded criteria were performed for 2 × 2 table count data. We also abstracted associative measures reported in the primary studies that had been adjusted for other factors. The number of recipients eligible for assessment (i.e. the denominator) could be appreciably different in later years than at cohort inception, and this frequently was not reported. For meta-analysis of longer-term outcomes, we used numbers reported at cohort inception if no other data were provided. When the outcome of death-censored graft survival was considered in unadjusted 2 × 2 table count data, those who died were excluded from both the numerator and the denominator of the cells. Although some studies reported GFR standardized to body surface area, others did not. When describing combined results we referred to the unit as mL/min, as it was the most commonly used metric. Some studies did not report standard deviation for average serum creatinine values. Following the guidelines for handling missing variance data, we used algebraic recalculation to recover the missing standard deviations when possible (13,14). The upper limit of reported p-values was used to obtain conservative estimates. When algebraic recalculation was not possible, or estimates from authors and nonparametric summaries were not available, we used single imputation (substitution) based on the following criteria: year of publication, country of study population and sample size. Sensitivity analyses were performed for this choice of imputation using the maximum of other reported standard deviations, and the results were not qualitatively different.

Publication bias was assessed by funnel plot and Egger's test for funnel plot asymmetry (15). Egger's test was used to determine whether standardized effect estimates were associated with the precision of a study (i.e. SD of effect size). If so, this suggests that small studies are only published if the effect estimates are large, instead of being published regardless of the effect estimate and whether it is statistically significant (publication bias). Exploratory meta-regression analyses were used to assess the effect of study-level baseline characteristics on the association between medical complexity and outcome observed across studies (16). The explanatory ability of each univariate meta-regression model was summarized as the proportion of between-study variability, using the SAS macro (SAS statistical software, SAS Institute, Inc., Cary, NC) and methods described in Berkey et al. (17). We considered the association between outcomes and mean/median recipient age for older and younger donors. The following study-level factors were also examined: study design (prospective vs. retrospective cohort), mean length of follow-up, mid-point year of donation, English-speaking country of recruitment (defined by national language), percentage of female donors, female recipients, black donors, black recipients, donors genetically related to recipients and mean number of HLA mismatches. Statistical significance was inferred by the proportion of variability explained by the model and from the size of the residual variance; residual (or between-study) variance was estimated using an iterative procedure outlined in Berkey et al. (17,18). Best-fit lines in meta-regression were produced using generalized estimating equations (SAS procedure, PROC GENMOD, SAS statistical software). Details on generating the best-fit lines and 95% confidence intervals (CIs) are described elsewhere (19).

Meta-analyses were conducted using Review Manager 5.0 (Cochrane Collaboration, Oxford, England). Statistical analyses were performed using SAS version 9.1 (SAS Institute, Inc., Cary, NC). Results were graphed in R 2.0.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

Study selection

We screened 13 928 citations, from which 343 full-text articles were retrieved for detailed evaluation, and 36 studies met our criteria for review (5,20–54). Reasons for study exclusion are described in Figure 1. The chance-corrected agreement between two independent reviewers was excellent for the assessment of full-text article eligibility (κ= 0.85).

image

Figure 1. Selection of studies.

Download figure to PowerPoint

Description of included articles

The 36 studies were from 13 countries (Table 1–3). Some studies described more than one expanded criteria in donors (31,45), resulting in 31 studies on older donors (a total of 3058 older donors and 13 926 younger donors) (5,20–38,40–50), seven studies examining donor GFR (two on donors with a GFR < 80 mL/min [a total of 54 donors with GFR < 80 mL/min and 442 with a GFR ≥ 80 mL/min] (31,39) and five using other GFR cutoffs) (5,45,51–53), and two studies on hypertensive donors (42 hypertensive donors, 124 controls) (45,54). In some cases, the same donor had multiple expanded criteria (5,31,45). Donors with the criteria of obesity, proteinuria or microscopic hematuria were not described in any study.

Table 1.  Characteristics of studies with expanded criteria donors who were older
ReferencePrimary countryYears of donationNo. of expanded criteria donorsNo. of standard criteria donorsAge of expanded criteria donors (years)Age of standard criteria donors (years)
  1. … Unreported.

  2. — Not applicable.

  3. 1We accepted these studies with an age cutoff of >65 or >66 because the average ages of those donors and their recipients were essentially the same as the other studies.

  4. 2Supporting information provided by the author.

  5. 3Included study because follow-up time and date of publication suggest earliest date of donation was after 1980.

20Italy1983–2002 39 193>60<60
21USA1987–1995404298161–7646–60
22USA1996–20005675131>6046–60
23Norway1983–1988 62 170>60<60
24USA1990–2005 64≥6040–59
25Germany1983–2002 19 125>60<60
26USA1995–2003280191260–6455–59
27Greece1990–1996 14  19>60<60
28Japan1980–1983 41None>661
29Macedonia1991–1999 28  50>65124–59
30USA1996–2002 42  42>60<40
31Japan1983–1992 49 114≥60<60
33Greece1984–1988 49  65>60≤60
32Greece1986–1996161 174≥60<60
34India1989–1993 21  25>60<45
 5India1988–2001 82None>60
35Austria1984–1991 22  57>6030–60
36Yugoslavia1987–1992 50  99>60<60
37Serbia & Montenegro1987–1999115 15860–8534–59
38Germany1996–2005 35 158>60≤60
50Japan1982–2002 28  5160–70<60
40Norway1994–2004 56 2892>65150–64
41Turkey3 44  44>60<60
42Japan1982–1995241 518>60<60
43Israel1980–1993 44 267>651<65
44India1989–1997 24 350>60≤60
45India1996–2004 46None>60
46Japan1982–1998 33  48>6050–59
47Japan1983–1998343 749≥60≤59
48Turkey1988–1995 30  38>60<55
49Germany1995–2004 25  99>60<60
Table 2.  Characteristics of studies with expanded criteria donors with a low glomerular filtration rate (GFR)
ReferencePrimary countryYears of donationNo. of expanded criteria donorsNo. of standard criteria donorsGFR of expanded criteria donorsGFR of standard criteria donorsMethod of measuring GFR
  1. GFR = glomerular filtration rate; skGFR = single-kidney GFR.

  2. 1Assumed to be skGFR due to very low donor GFR.

  3. Included study because follow-up time and date of publication suggest earliest date of donation was after 1980.

51USA1997–2003121 126iGFR < 110 mL/miniGFR > 110 mL/min125I-iothalamate urinary clearance
31Japan1983–199228124GFR ≤ 80 mL/minGFR > 80 mL/minThiosulphate, phenolsulphopthalein, p-aminohippuric salt clearance
 5India1988–200110No controlsskGFR<45 mL/min1No controlsUnreported
52Serbia & Montenegro1988–20013832skGFR < 50 mL/minskGFR > 50 mL/min99mTc-DTPA (technetium-labeled diethylene-triamine-pentacetate)
39Sweden1985–199726318GFR < 80 mL/minGFR ≥ 80 mL/min51-Crom Ethylenediamine tetraacetic acid (EDTA) clearance
53USA1998–20026653skGFR<55 mL/min/1.73 m2skGFR>55 mL/min/1.73 m2125I-iothalamate urinary clearance
45India1996–2004 8No controls‘low GFR’No controlsUnreported
Table 3.  Characteristics of studies with expanded criteria donors with hypertension
ReferencePrimary locationYears of donationNo. of expanded criteria donorsNo. of donors without hypertensionDefinition of hypertension
45India1996–200418NoneTreated with medication and a subsequent BP < 140/90 mmHg
54USA2001–200224124Awake ambulatory blood pressure monitoring > 135/85 mmHg and clinic/RN BP > 140/90 mmHg

All studies were published between 1992 and 2007 and covered a period of donation from 1980 to 2005. Thirty-five articles were published in English and one in Japanese (50). The studies were primarily from Europe (Germany and Greece: three studies each; Norway and Serbia & Montenegro: two studies; Austria, Italy, Macedonia, Sweden, Yugoslavia: one study), followed by Asia (Japan: six studies; India: four studies; Turkey: two studies; Israel: one study), and North America (USA: eight studies). We successfully contacted 25 authors who wrote 28 of the studies. Twelve authors kindly supplied additional data for 14 studies (20–23,25,26,29,30,36,37,39,40,51,53).

The donor and recipient characteristics in each study are described in Table 4. The methodological quality of the included studies is presented in Table 5. Three studies did not have any donor controls (i.e. standard criteria donors) (5,28,45). Two studies were conducted in a prospective fashion (40,54), while the remainder were retrospective. The proportion of recipients lost to follow up was available for seven studies and ranged from 0% to 19% (24,25,30,36,40,51,53). The outcomes examined in each study are presented in Table 6.

Table 4.  Donor and recipient characteristics
Reference% Female donors% Genetically related to recipientDonor age (age ± SD (range))Recipient age (age ± SD (Range))Follow-up period (mean ± SD (range)), years
Expanded criteria donorsStandard criteria donorsExpanded criteria donorsStandard criteria donorsExpanded criteria donorsStandard criteria donorsExpanded criteria donorsStandard criteria donorsExpanded criteria donorsStandard criteria donors
  1. Values in table reported as: mean ± standard deviation (Range).

  2. …Unreported.

  3. —Not applicable.

  4. 1Supporting information provided by the author.

  5. 2Median follow-up.

Donor age > 60
206967908764 ± 4 (60–72)47 ± 8 (19–59)36 ± 11 (18–64)31 ± 13 (9–66)
21(61–76)
22
2366 (60–81)41 (22–54)
24
256865797364 ± 3 (60–68)44 ± 9 (15–59)42 ± 15 (11–65)33 ± 15 (4–66)3.1 ± 3. 1 (0.3–12)14.9 ± 5.21
26667647 ± 13
27
2881100 71 (66–80)44 (17–62)3
29  100 100 69 (65–81)50 (24–59)
304343506765 ± 4 (…)33 ± 551.345.8
317.92 
3369 (61–80)50 (26–60)
326882100 100 68 ± 6 (60–82)51 ± 7 (24–59)40 ± 8 (19–58)30 (18–57)(0.8–4)(0.8–4)
3464 (60–71)31
 54563 ± 31.8
3534283.3 ± 0.6 (0.3–7)
365820100 66 ± 4 (60–85)48 ± 4 (34–59)27 ± 7
375070100 100 66 ± 550 ± 639 ± 828 ± 83.6 (2.8–5.5)3.6 (2.8–5.5)
386062606365 ± 449 ± 750 ± 1338 ± 14(4-…)(4-…)
50(60–70)4.3 (0.1–9.1)4.3 (0.1–9.1)
4070 ± 4156 ± 4154.4 ± 15146 ± 141
416157100 100 66 ± 446 ± 7(0–11.5)(0–11.5)
426167 36 ± 633 ± 105.2 ± 2.35.2 ± 2.3
4369 ± 150 ± 6 (1–13.8)(1–13.8)
4442(28–46)(12–62)
455962 ± 3 (60–71)
4665 ± 4 (60–78)55 ± 3 (50–59)35 ± 929 ± 61.8 ± 0.1 (0.5–4.2)1.8 ± 0.1 (0.5–4.2)
47 
48834765 ± 440 ± 1532 (19–45)29 (12–52)
49766546 ± 15 (27–71) 2.12.9
Donor with lower GFR
5179143166.7177144 ± 9141 ± 10146 ± 13142 ± 1212121
31
 53.3 ± 0.6 (0.3–7)
39
5257 ± 953 ± 1134 ± 7.532 ± 105.826.32
5373143165176144 ± 9141 ± 9147 ± 12144 ± 1312121
45
Donor with Hypertension
452.522.52
5457.25041.4 ± 1 (21–74)53.4 ± 2 (35–71)0.8 (0.3–1.2)0.8 (0.3–1.2)
Table 5.  Methodological quality of included studies
ReferenceCohort study designRepresentative sample% Lost to follow upDonor factors clearly describedRecipient outcomes and method of measurement clearly describedConfounders addressed
  1. … Unreported

  2. 1Supporting information provided by author.

Donor age > 60
 20RetrospectiveYesYesYesNo
 21RetrospectiveYesYesYesNo
 22RetrospectiveYesYesNoNo
 23RetrospectiveYesYesPartiallyNo
 24RetrospectiveYes0YesNoYes
 25RetrospectiveYes18YesYesYes
 26RetrospectiveYesPartiallyPartiallyYes
 27RetrospectiveYesNoYesNo
 28RetrospectiveNoYesYesNo
 29RetrospectiveUnclearYesNoNo
 30RetrospectivePartially191YesYesPartially
 31RetrospectiveYesNoNoNo
 33RetrospectiveYesYesNoNo
 32RetrospectiveYesYesYesNo
 34RetrospectiveNoYesYesNo
  5RetrospectiveNoNoNoNo
 35RetrospectiveYesYesYesYes
 36RetrospectiveYes51YesNoYes
 37RetrospectiveYesYesYesNo
 38RetrospectiveYesYesYesYes
 50RetrospectiveYesYesNoUnclear
 40ProspectiveYes01YesYesYes
 41RetrospectiveYesYesYesNo
 42RetrospectiveYesYesYesYes
 43RetrospectiveYesYesYesNo
 44RetrospectiveUnclearYesYesYes
 45RetrospectiveYesNoNoNo
 46RetrospectiveYesYesYesNo
 47RetrospectiveYesNoNoYes
 48RetrospectiveUnclearYesNoYes
 49RetrospectiveYesYesYesNo
Donor with lower GFR
 51RetrospectivePartially1.9YesYesYes
 31RetrospectiveYesNoNoNo
  5RetrospectiveNoNoNoNo
 52RetrospectiveUnclearYesYesYes
 39RetrospectiveYesYesYesYes
 53RetrospectiveNo1.7YesYesYes
 45RetrospectiveYesNoNoNo
Donor with hypertension
 45RetrospectiveYesNoNoNo
 54ProspectiveYesYesYesYes
Table 6.  Recipient outcomes assessed across studies of complex living kidney donor
ReferenceGraft survival (censoring unspecified)Uncensored graft survivalDeath- censored graft survivalPatient survivalSerum creatinineGFR
  1. x = reported data.

Donor age >60
 20x     
 21 x    
 22 x    
 23x   x 
 24 xxx  
 25  xxxx
 26 x   x
 27x     
 28 x xx 
 29x x x 
 30x   x 
 31    x 
 33  xx  
 32 x xx 
 34    xx
  5x     
 35x  x  
 36x  xx 
 37 xxxx 
 38  xxx 
 50x   x 
 40  x   
 41   xx 
 42 x x  
 43x   x 
 44x     
 45  xx  
 46  x   
 47 x    
 48 x x x
 49x   x 
Donor with lower GFR
 51     x
 31    x 
  5     x
 52 xxx  
 39  x   
 53     x
 45     x
Hypertensive Donors
 45    xx
 54    xx

Donors 60 Years of Age or Older

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

Definition of older age

Older donors ranged from 60 to 85 years of age and younger donors ranged from 31 to 55 years of age (4,29). Recipients who received older kidneys were older than those who received younger kidneys (based on nine studies, pooled estimates of 38 years [95% CI 36–40] vs. 31 years [95% CI 28–33]). Recipients of older and younger kidneys had a similar likelihood of receiving a kidney from a woman or a genetically related donor.

Composite outcome of recipient and graft survival

Most reported estimates were unadjusted for possible confounders. In meta-analysis, recipients of kidneys from older donors had slightly poorer 1-year survival (patient and graft) than those receiving kidneys from younger donors (91% vs. 94%, unadjusted RR of survival: 0.97, 95% CI 0.94–1.00, I2= 55%, 15 studies, in total 10 355 younger donors and 1812 older donors) (20–23,27,29,30,35–37,43,44,47,48,50). Recipients of kidneys from older donors had poorer 5-year survival than recipients of kidneys from younger donors (Figure 2A, 72% vs. 80%, unadjusted RR of survival: 0.89, 95% CI 0.83–0.95, I2= 71%, 12 studies, in total 10 704 younger kidneys and 2061 older kidneys) (21–23,29,32,35–37,42–44,47). The association between donor age and outcome adjusted for relevant factors such as recipient age was examined in three studies (Table 7) (26,35,47). Recipients of kidneys from younger donors had better outcomes than kidneys from older donors (p < 0.05).

image

Figure 2. Meta-analyses of 5-year outcomes. (A) Composite outcome of graft and recipient survival. (B) Death-censored graft survival. (C) Recipient survival.

Download figure to PowerPoint

Table 7.  Adjusted hazard ratio (HR) for risk of recipient death and graft loss, comparing recipients of older and younger kidneys
ReferenceAdjusted HR95% Confidence intervalpFactors adjusted for
  1. 1Multivariable analyses adjusted for the factors listed.

2611.281.52–1.10Recipient age 50–54, 55–69, 60–64, 65–69, 70+, female gender, HLA match, cause of end-stage renal disease, comorbid conditions (congestive heart failure, peripheral vascular disease, ischemic heart disease, cancer), duration of dialysis, race (black, other)
3513.130.01Recipient age, age match, HLA B+DR mismatch, panel reactive antibodies, number of transplants, family relationship, warm ischemia time
4712.56<0.0001Recipient sex, donor sex, HLA-DR mismatches, HLA-AB mismatches, ABO compatibility, kidney weight/body weight ratio, donor type (related/unrelated), acute rejection episodes

Additional analyses

Death-censored graft survival: In meta-analysis recipients of kidneys from older donors had poorer 1-year death-censored graft survival than recipients of kidneys from younger donors (94% vs. 99%, unadjusted RR of survival: 0.96, 95% CI 0.93–1.00, I2= 0%, four studies) (25,37,38,46). There was a trend of poorer 5-year graft survival for recipients who received older kidneys that was not statistically significant (Figure 2B, 70% vs. 87%, unadjusted RR of survival: 0.88, 95% CI 0.72–1.07, I2= 87%, four studies) (25,37,38,46). A single study examined the association between donor age and death-censored graft survival adjusted for recipient age, recipient sex, donor sex, and number of HLA-DR mismatches (Table 8) (40). Recipients who received a kidney from an older donor had a poorer survival than those who received a kidney from a younger donor (p < 0.05).

Table 8.  Adjusted hazard ratio for risk of death-censored graft loss, comparing recipients of older kidneys (>65 years of age) to those receiving younger kidneys (18–49 years of age).
ReferenceTime periodAdjusted HR95% Confidence intervalpFactors adjusted for
  1. 1The comparison between recipients of kidneys from older donors (>65 years) to recipients of younger kidneys (18–45 years) is similar for various time periods (3 months to 5 years and ≥3 months).

40≥5 years15.091.55–16.72<0.05Recipient age, donor gender, acute rejection episode, steroid resistant rejection, preemptive transplantation

Recipient survival: In meta-analysis recipients of kidneys from older donors had similar 1-year survival to recipients of kidneys from younger donors (unadjusted RR of survival: 0.99, 95% CI 0.96–1.01, I2= 0%, 5 studies) (25,35–38). Trends in 5-year survival were not appreciably different (Figure 2C, 90% vs. 92%, unadjusted RR of survival: 0.97, 95% CI 0.94–1.01, I2= 37%, 7 studies) (25,32, 36–38,41,42). The association between donor age and recipient survival adjusted for relevant factors was examined in one study (Table 9) (36). Recipients who received a kidney from an older donor had poorer survival than those who received a kidney from a younger donor (p = 0.02).

Table 9.  Adjusted hazard ratio (HR) for risk of recipient death, comparing recipients of older and younger kidneys
ReferenceAdjusted HR95% Confidence intervalpFactors adjusted for
  1. 1Supporting information provided by author.

363.411.17–6.3310.021Donor–recipient age difference, recipient age, dialysis duration, delayed graft function, acute rejection episode, graft function1

Recipient serum creatinine and glomerular filtration rate: In meta-analysis, recipients of kidneys from younger donors had a lower 1-year serum creatinine than recipients of kidneys from older donors (Figure 3A, weighted mean difference: −24 μmol/L[−0.3 mg/dL], 95% CI −31 to −17 μmol/L, I2= 52%, 9 studies, in total 1010 younger kidneys and 548 older kidneys) (5,23,25,30–32,37,38,41). In meta-analysis, recipients of younger kidneys had a statistically higher 1-year GFR than recipients of older kidneys, but the absolute difference was small (Figure 3C, weighted mean difference: 3 mL/min, 95% CI 1.1–4.8, I2= 88%, 4 studies, in total 2100 younger kidneys and 350 older kidneys) (25,26,34,48).

image

Figure 3. Meta-analysis of recipient kidney function 1 year after transplantation. (A) Difference in serum creatinine (μmol/L) between recipients of younger and older kidneys. (B) Differences in GFR (mL/min) between recipients of younger and older kidneys.

Download figure to PowerPoint

Publication bias and meta-regression: There was no publication bias for the primary outcome of 5-year recipient and graft survival (Egger's test p = 0.83). The exclusion criteria used in some studies contributed to between-study heterogeneity. One study excluded recipients who either died in the first year or never had a functioning graft (36). Several studies limited their study population to first time graft recipients, who tend to have better outcomes compared to individuals with a history of previous failed transplants (24–26,36,37,48,49). These exclusions likely inflated the longer-term graft survival rate, although the association between medical complexity and outcome should not have been greatly impacted.

Sparse data precluded random-effects meta-regression to examine the impact of the following study-level characteristics on the relationship between donor age and 5-year recipient and graft survival: study design (prospective vs. retrospective cohort), mean length of follow-up, percentage of female donors, female recipients, black donors, black recipients, donors genetically related to recipients and mean number of HLA mismatches. Whether a country was English speaking or not did not influence the results.

Mid-point year of recruitment was associated with study results and explained 31% of the between-study variability (Figure 4, meta-regression plot). Five studies were conducted in the 1980s, seven studies in the 1990s, and none in the 2000s. The poorer composite outcome of patient and graft survival for recipients of older kidneys compared to recipients of younger kidneys diminished with time (1980s RR of survival: 0.79, 95% CI 0.65–0.96, I2= 68%; 1990s RR of survival: 0.91, 95% CI 0.85–0.99, I2= 73%).

image

Figure 4. Relative risk of graft survival by era. Recipients of kidneys from younger donors had better 5-year patient and graft survival than recipients of kidneys from older donors, but this association was less prominent over time across the studies. Year of study was defined by the mid-point year of follow-up; this characteristic explained 31% of the between-study variability. The area of each circle is proportional to the number of patients in each study. Curve is best-fit line from weighted meta-regression. See ‘Methods’ section.

Download figure to PowerPoint

Donors with a Low GFR

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

In total, seven studies examined recipient outcomes for donors of kidneys with a low GFR (as defined by the primary authors, Table 2). Two studies followed 54 donors with a GFR < 80 mL/min and 442 donors with a GFR > 80 mL/min (31,39). Five studies varied in their GFR cutoff, defining low GFR as follows: iothalamate GFR < 110 mL/min, GFR < 45 mL/min (measurement technique unspecified), single-kidney iothalamate GFR < 55 mL/min/1.73 m2, single-kidney 99mTc-DTPA GFR <50 mL/min/1.73 m2 or simply the term ‘low GFR’. Recipients of kidneys from a donor with a higher GFR tended to have better graft survival, patient survival and kidney function (Table 10). No studies examined the impact of GFR on outcome, adjusted for other relevant factors.

Table 10.  Studies of recipient outcomes from living donors with a higher and lower GFR
ReferenceDefinition of lower donor GFRLength of follow- up years1Outcome reportedRecipient outcome for lower GFR donorsRecipient outcome for higher GFR donorsp-Value
  1. CI = Confidence interval; HR = hazard ratio; NS = not significant; N/A = not applicable; NR = not reported; skGFR = single-kidney GFR.

  2. skGFR-–single-kidney GFR, which was assumed to mean the measured GFR for both kidneys was divided by two, unless specified otherwise. Poggio et al. (53) assumed skGFR was proportional to kidney volume, and calculated value based on donated kidney volume.

  3. 1Some studies report results at multiple lengths of follow-up. For studies reporting (change in) recipient eGFR or serum creatinine, we chose to report 2-year outcomes. For patient and graft survival, we chose to report 5-year outcomes and where 5 years was unavailable, 10-year outcomes.

  4. 2Comparing low GFR donor recipients to higher GFR donor recipients (95% CI).

  5. 3Author did not specify if this was single-kidney GFR-–it was assumed to be skGFR due to the low value.

Lower GFR defined as <80 mL/min
 31GFR < 80 mL/min/1.73 m2 (thiosulphate clearance)2Recipient serum creatinine (mg/dL)1.701.62NR
 39GFR < 80 mL/min/1.73 m25Death-censored graft survival (%)7688NR
 39GFR < 80 mL/min/1.73 m25Death-censored graft loss hazard ratio (95% CI)22.28 (1.18, 4.38)NR
Lower GFR defined by another cutoff
 51iGFR < 110 mL/min2Recipient eGFR (mL/min/1.73 m2), mean (SD)53 (16)60 (21)<0.01
 53skGFR < 55 mL/min/1.73 m22 51 (17)64 (17)<0.01
 45‘low GFR’2.2 Change in recipient eGFR (mL/min/1.73 m2)13(No control group)N/A
 5skGFR < 45 mL/min32 15.2(No control group)N/A
 52skGFR < 50 mL/min/1.73 m25Recipient and graft survival (%)93.192.9NS
 52skGFR < 50 mL/min/1.73 m210 Death-censored graft survival (%)7585NR
 52skGFR < 50 mL/min/1.73 m25Recipient survival (%)93.197.1NS

Hypertensive Donors

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

Two studies examined 42 hypertensive donors and 124 controls (Table 3) (45,54). Loss to follow up and outcomes of graft or patient survival were not reported. In the first study, hypertension was described as “hypertensive donor treated with medication (current blood pressure < 140/90)” (45). Two and a half years after transplant, the serum creatinine was < 124 μmol/L (<1.4 mg/dL) in 10 of the 18 recipients who received kidneys from a hypertensive donor, while the remaining 8 recipients had values < 221 μmol/L (<2.5 mg/dL) (45). In the second study, hypertension was defined as awake ambulatory blood pressure monitoring >135/85 mmHg and blood pressure measured in clinic by a nurse >140/90 mmHg (54). The 1-year serum creatinine of recipients of kidneys from normotensive donors was 154 ± 9 μmol/L (1.74 mg/dL), compared to 145 ± 9 μmol/L (1.64 mg/dL) from hypertensive donors (p not significant). The 1-year GFR of recipient of kidneys from normotensive donors was 54.2 ± 2 mL/min/1.73 m2 compared to 49 ± 4 mL/min/1.73 m2 from hypertensive donors (p not significant) (54).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

Every year more individuals with expanded criteria are accepted as living donors than ever before. Of the 9319 living kidney donations reported in the American Organ Procurement and Transplantation Network (OPTN) / United Network for Organ Sharing (UNOS) database from July 2004 to December 2005, 368 (4.0%) donors were over the age of 60 (of which 86 [0.9%] were over the age of 65), 956 (10.3%) had hypertension, 2785 (29.9%) had an eGFR < 80 mL/min per 1.73 m2 (of which 392 [4.2%] had an eGFR < 60 mL/min per 1.73 m2) and 1194 (12.8%) had a BMI > 30 kg/m2 (of which 250 [2.7%] had a BMI > 35 kg/m2) (P. Reese, personal communication, 2008) (1). We comprehensively reviewed available evidence on the outcomes of recipients who received kidneys from expanded criteria living donors. Two key findings emerged. First, despite all donors going through a rigorous evaluation process to confirm good health, recipients of a kidney from a younger donor had appreciably better outcomes than recipients of a kidney from an older donor. This is a well-documented finding in deceased donation. However, 8 of the 12 studies examining the composite outcome of graft and recipient survival in living donation reported no significant difference between kidneys from older and younger donors (22,32,35,37,42–44,47). Only through visual presentation and meta-analysis did this association become more apparent (Figure 2A). This association was also evident in three studies where confounding factors such as recipient age were accounted for in the analysis (26,35,47). It should be noted that this association appeared to diminish with era and was less prominent in studies conducted in the 1990s than those conducted in the 1980s (Figure 4); to our knowledge, this is also a new observation which should be considered in future studies. Second, there was a striking paucity of data for most of the defined expanded criteria. The number of studies involving donors with a low GFR or hypertension was limited. Studies of donor obesity, proteinuria or microscopic hematuria were nonexistent. This disconnect between the number of individuals with expanded criteria being accepted as donors and a lack of knowledge of recipient outcomes should give the transplant community pause and sets a clear agenda for future research.

Strengths and limitations of the review

This is the first systematic review and meta-analysis to examine the impact of living kidney donor age and other expanded criteria on recipient outcomes. The donor criteria examined in this study were chosen because they are evident prior to nephrectomy. Understanding the impact of these factors can aid in informed consent and kidney donor selection. Previous narrative reviews and clinical practice guidelines have described the effect of living donor age on recipient outcome (1,55–60). However, they did so less extensively or only examined ‘old for old’ donation. We used a number of measures to reduce the chance of missing relevant studies in our synthesis. An experienced information specialist helped execute our search of electronic databases. We evaluated reference lists of all relevant articles and completed cross-reference searches with citation tracking in various databases and search engines. We screened nearly 14 000 records and reviewed over 340 full-text articles. We performed the article identification, selection and data abstraction independently and in duplicate to minimize any potential biases arising from the subjectivity in these tasks. We also received additional information from authors that was not available in the original reports.

The results of any review are inherently limited by the quality of the included studies. We summarized these methodological limitations. Some studies did not report all relevant recipient outcomes. Others did not specify whether graft survival was censored for death. Uncensored graft survival considers death with a functioning graft to be graft loss and would underestimate survival in older recipients. Essential information about study design, patient demographics, inclusion/exclusion criteria and statistical analysis often went unreported. The results of some outcomes differed markedly across the studies making meta-analysis less reliable. For this reason, we also conducted meta-regression to better understand the reasons for observed differences between the studies.

The decision to accept a medically complex donor and future research

The Amsterdam Forum report provides recommendations for an international standard of care regarding the responsibility of the transplant community for the live kidney donor (61). An upper age limit for a live kidney donor was not discussed or defined. The guidelines do suggest individuals with GFR < 80 mL/min, hypertension, body mass index > 35 kg/m2 or proteinuria > 300 mg/day should generally be precluded or discouraged from donation. The reason to preclude such donors with these criteria lies in the substantial uncertainty for long-term health risks to the donor for an uncertain degree of recipient benefit. Yet, more individuals with expanded criteria are being accepted as donors than ever before.

Despite all living donors going through a rigorous selection process to confirm good health, in this synthesis recipients of kidneys from older donors had worse outcomes than recipients of younger donor kidneys. The choice of accepting an older donor may be influenced by the availability of other donors or the wish to save a young donor for future need. Decision-making is also informed by the alternate treatment options (remaining on dialysis until the availability of a kidney from a younger deceased donor), where future comparative research is needed. If graft survival proves lower than expected for kidneys received from older living donors, this may cause some families to reevaluate the decision to proceed with this treatment option. In this synthesis, recipients of older kidneys were on average 7 years older than recipients of younger kidneys (38 vs. 31 years of age), which may partially explain the associations observed in unadjusted analyses. However, recipient survival was not significantly better for recipients of younger kidneys compared to older donor kidneys, suggesting the increased incidence of graft loss seen with older kidneys may be a function of kidney quality (62). For example, renal senescence and the inability of an older kidney to fully accommodate the insults that occur during and after transplant may partly explain these findings (63,64). The living donor age–recipient outcome relationship should be confirmed in future research, particularly as this association was less prominent in studies conducted in the more recent era (Figure 4, meta-regression). It is possible modern immunosuppression and supportive care has reduced demands on the transplanted kidney. Furthermore, the functional demands and required duration of an older kidney may be less for an older than younger recipient. This may make an older individual a better living donor for some recipients and not others.

Finally, more clarification is needed as to whether it is the living donor's older age per se, or subtle conditions co-occurring with older age, that influence outcome. Until such time when new data becomes available, transplant professionals must continue to judiciously counsel expanded criteria potential donors and their recipients. This includes a thorough discussion of the available treatment options, the anticipated outcomes and highlighting current uncertainty where it exists.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References

We thank Mr. Dariusz Gozdzik, Ms. Patricia Hizo-Abes, Dr. Michiya Kawai and Ms. Gloria Valbuena for translating foreign language articles. We also thank the primary authors of the included studies who generously confirmed and provided information for this review.

This review was supported by the Multi-Organ Transplant Program of the London Health Sciences Centre, London, Ontario, Canada. Mr. Yiannis Iordanous was supported by the Schulich School of Medicine Summer Research Training Program (SRTP). Ms. Ann Young was supported by a Canada Graduate Scholarship from the Canadian Institutes of Health Research and a Schulich Graduate Scholarship from the University of Western Ontario. Dr. Amit Garg was supported by a Clinician Scientist Award from the Canadian Institutes of Health Research.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Donors 60 Years of Age or Older
  7. Donors with a Low GFR
  8. Hypertensive Donors
  9. Discussion
  10. Acknowledgments
  11. Conflict of Interest Statement
  12. References