Impact of Graft Mass on the Clinical Outcome of Kidney Transplants

Patients

Between April 1995 and June 2001, the weight of 1142 kidney grafts was collected prospectively and consecutively immediately before grafting. Our institution in Nantes and six other French transplantation centers (Nancy, Caen, Grenoble, Paris Bicêtre, Tours, and Strasbourg) participated in the study. Donors (n = 40) and recipients (n = 15) who were younger than 15 yr were excluded from the study, according to the growth-related changes in kidney and recipient size after transplantation. In addition, because some participating centers used bladder drainage for simultaneous kidney and pancreas grafts, we did not include these patients (n = 103) because their proteinuria measurements were uninterpretable. Patients in whom technical failure occurred, as defined as a definitive return to dialysis during the first 7 d after surgery, were also excluded (n = 8). Finally, data from one center (Paris Bicêtre) that initially participated in the weighing protocol but failed to provide patient follow-up were also excluded from the study (n = 12). Analysis thus was performed in the 964 remaining patients (including 7 living related donors) for whom all information was available. The mean follow-up period was 32 mo with a maximum of 94 mo. All patients received anticalcineurin inhibitors for maintenance therapy.

Kidney Weight

All participating centers were provided with the same electronic weighing scales (Maul; Gilbert Fourniture, France) with an accuracy of 1 g. The scales were located in the operating room. According to common procedure recommendations, kidney grafts were prepared first then weighed by the surgeon.

Parameters Studied

Clinical and biologic data concerning donors and recipients were obtained prospectively at the time of inclusion and centralized in the anonymous and coded study data bank located at Nantes University Hospital, in accordance with the local ethical committee requirements. Donor data include age, gender, creatinemia (μmol/L), and kidney weight (donor kidney weight [DKW] in g). Recipient data at the time of inclusion concerned age, gender, weight (recipient body weight [RBW] in kg), the ratio between DKW and RBW (DKW/RBW; in g/kg), the number of grafts, HLA-A-B-DR incompatibilities, anti–T panel reactive antibodies (%), and cold ischemia time (min). Moreover, at the time of each follow-up (3, 6, and 12 mo, and then every 12 mo until 48 mo), the following data were collected from recipients: Weight, creatinemia (μmol/L), creatinine clearance calculated according to the Cockroft formula {[(140 − age) × weight (kg)]/creatinemia μmol/L) × 1.04 for female or 1.23 for male recipient} (ml/min) and daily proteinuria (g/24 h). Also recorded were the occurrence of delayed graft function (as defined by the time interval, in days, to reach a Cockroft-calculated creatinine clearance >10 ml/min) (5), the occurrence of acute rejection episodes, and the time of death or definitive graft failure with return to chronic dialysis.

Statistical Analyses

The first objective of the study was to assess the effect of kidney and recipient mass at the allocation, defined by the DKW/RBW ratio, on graft function (Cockroft-calculated creatinine clearance [cCrCl]). To take into account interindividual variability, we first used a linear mixed-effect model (21). In this model, the effect of the DKW/RBW ratio was adjusted on all historical cofactors. As a preliminary study showed that the relationship between the cCrCl and the time interval from transplantation was linear with a break point at 6 mo, we then modeled the relation between the time and the clearance according to the two different slope before and after 6 mo. On the basis of biologic observations (change of mean of cCrCl at each data collected point), it was possible to identify three groups of patients according to three classes of ratio: DKW/RBW ratio (g/kg) <2, between 2 and 4, and ≥4. To allow for different cCrCl evolution in the three classes, we considered in the linear mixed-effect model an interaction between the time interval from transplantation and the DKW/RBW ratio.

The second aim was to assess the consequences of changes in graft filtration rate on the occurrence of kidney injury defined as a proteinuria. We built a Cox proportional-hazards model to determine which covariates and particularly DKW/RBW ratios were risk factors for the first occurrence of a proteinuria above a threshold of 0.5 g/24 h. Patients who died or underwent graft failure were censored. Finally, to study the impact and the rank of the DKW/RBW ratio on graft survival, we built a second Cox model whereby graft loss was defined as death or a definitive return in dialysis. For all tests, any variable that reached a P < 0.20 in the univariate analysis was introduced to the multivariate analysis. The Wald test was used to prove the effect of the variables. All data that were collected prospectively at the time of inclusion and during the follow-up were tested in each statistical model previously described. Finally, because a multicentric study implies that some clinical practice procedures and patient characteristics can vary between each center, each Cox proportional-hazards model was stratified on centers. The Software used was S-Plus 6.0 Professional.

Demographic Characteristics

Among the 964 patients included in the study, 88.7% were recipients of a first graft. Sixty-two percent were male recipients, and 68% received a kidney from a male donor. The mean HLA incompatibility was 3 ± 1.3, with only 12% of patients having 5 or fewer incompatibilities. Pregraft anti–T cell immunization was observed in 23% of patients. The mean delayed graft function was 5.4 d after surgery. Finally, 20% of patients developed an acute rejection episode during their follow-up period. The population characteristics are given in Table 1.

Kidney Weight Distribution According to Donor Gender and Age

The mean kidney weight ± SD of the whole population was of 202 ± 56 g (range 76 to 460). As expected, male kidneys were heavier than female kidneys with a mean weight of 216 ± 56 g (range 76 to 460) and 173 ± 42 g (range 94 to 335), respectively (P < 0.0001). In effect, 76% of female kidneys weighed <200 g as compared with only 44% for male kidneys. DKW distribution according to donor gender is illustrated in Figure 1A. Moreover, kidney weight increased significantly with donor age from 15 to 45 yr, independent of gender. However, after 45 yr of age, kidney weight remained stable, as shown in Figure 1B. Similarly, the mean recipient weight, which was used for the calculation of the DKW/RBW ratio, also increased with age (Figure 1C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

(A) Kidney weight distribution according to donor gender; 76% of female kidneys weighed <200 g. (B) Kidney weight distribution according to donor age. In both men and women, kidney weight increased with donor age until 45 yr of age and decreased thereafter. (C) Distribution of recipient weight according to age. Recipient body weight increased until 45 yr of age and remained stable thereafter.

Effect of DKW/RBW Ratio on Graft Function

Because body and kidney masses varied significantly between women and men as well as with age (Table 2), we took the option to analyze the effect of gender and the DKW/RBW ratio separately. On the one hand, despite a similar DKW/RBW ratio for male kidneys transplanted to male recipients (3.2 ± 1.2 g/kg; range 0.9 to 8.0) and for female kidneys transplanted to female recipients (3.1 ± 1 g/kg; range 1.2 to 6.9; NS), the 1-yr cCrCl was 20% higher for the first combination (cCrCl 63 ml/min) than for the second combination (52 ml/min; P < 0.0001). On the other hand, although the DKW/RBW ratio was highly different when male kidneys were given to female recipients (3.7 ± 1.3 g/kg; range 1.5 to 8.4) compared with female kidneys given to male recipients (2.6 ± 0.9 g/kg; range 1 to 5.6; P < 0.0001), the 1-yr graft function was similar for both with a cCrCl of 58.7 ± 23 ml/min (range 16 to 149) and 58 ± 19 ml/min (range 11 to 140), respectively, indicating that small female kidneys donated to large male recipients (i.e., a small DKW/RBW ratio) increased their filtration rate to fit their new body mass (Table 2). In contrast, large male kidneys that were donated to small female recipients (i.e., a high DKW/RBW ratio) were not associated with an increased function. Next, the effects of DKW/RBW ratios were studied, irrespective of donor and recipient gender. The mean 1-yr cCrCl was 54 ± 18 ml/min (range 15 to 108) for high DKW/RBW ratios ≥4 g/kg (n = 205; largest kidneys/smallest recipients) and increased by 10% for ratios between 2 and 4 g/kg (60 ± 20 ml/min; range 7 to 134; P < 0.02) and by 20% for small ratios <2 g/kg (n = 88; smallest kidneys/largest recipients; 69 ± 24 ml/min; range 25 to 149; P < 0.001). Moreover, a significantly higher mean cCrCl was observed at each follow-up interval (3, 6, 12, 24, and 36 mo), which continuously increased after 3 mo for the small ratios <2 g/kg and ratios between 2 and 4 but not for ratios ≥4 g/kg (Figure 2). The highest cCrCl observed for small ratios cannot be explained by an increase in weight of the patients of this group during the follow-up. In effect, only patients in a high DKW/RBW ratio group (≥4 g/kg) had a significant increase of weight after the graft (P < 0.001). In summary, whatever the recipient and donor gender, small kidneys that were donated to large recipients could adapt to the recipient body mass by increasing their function.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Kidney graft function in time according to the donor kidney weight/recipient body weight (DKW/RBW) ratio. The distribution of the mean Cockroft-calculated creatinine clearance at 3, 6, 12, 24, and 36 mo in each class of DKW/RBW ratios is shown. Small ratios <2 g/kg result in an increase in cCrCl at each data time point. When transplantations were performed with a high DKW/RBW ratio (≥4 g/kg), graft function remained stable in time.

Patients with a Low DKW/RBW Ratio Dramatically Increase Their Filtration Rate

This analysis was performed on a specific population that excluded patients with graft failure or those who died (n = 780). To test the hypothesis of graft filtration rate adaptation in conditions of small DKW/RBW ratios, we constructed a linear mixed-effect model. We showed that creatinine clearance evolved differently in the time before and after 6 mo according to the DKW/RBW ratios. As shown in Table 3, for DKW/RBW ratios <2 g/kg (smallest kidneys in largest recipients), creatinine clearance increased significantly by 2.38 ml/min every month until 6 mo (P < 0.0001) and by 0.27 ml/min thereafter (P < 0.0001). For DKW/RBW ratios between 2 and 4, cCrCl increased significantly by 1.14 ml/min every month for the first 6 mo (P < 0.0001) and by 0.11 ml/min per mo thereafter (P < 0.0001). However, for DKW/RBW ratios ≥4 g/kg, creatinine clearance did not change with time. Other covariates that were found to influence cCrCl in time after transplantation were donor and recipient age, which had a similar impact on graft function (P < 0.0001), cold ischemia time (P = 0.0093), and delayed graft function (P < 0.0001). Finally, at 12 mo, female recipients had a cCrCl 11 ml/min less than male recipients who received a male kidney (P < 0.0001) when they received a female kidney and 5.86 ml/min when they received a male kidney (P < 0.0001). To summarize, cCrCl was affected by donor and recipient gender and age, cold ischemia time, delayed graft function, and DKW/RBW ratio in time.

Low DKW/RBW Ratios Are Associated with Proteinuria (n = 909)

We next investigated whether the highest filtration rate observed with low DKW/RBW ratios resulted in deleterious consequences in terms of graft function and particularly the occurrence of proteinuria, the first quantifiable usual parameters that correlate with kidney graft injuries after hyperfiltration. We observed that recipient gender interfered significantly with donor age (P = 0.005); the results then were calibrated for donor age at, for instance, 45 yr (Table 4). One third of patients presented a proteinuria >0.5 g/24 h during the follow-up. Patients with the lowest DKW/RBW ratio (<2 g/kg) presented the highest (50%) and earliest occurrence (10.43 mo) of proteinuria compared with 33% and 13.44 mo for DKW/RBW ratios between 2 and 4 g/kg and 23% and 14.33 mo for DKW/RBW ratios ≥4 g/kg. The probability of the proteinuria >0.5 g/24 h occurring for the first time during the follow-up correlated with the DKW/RBW ratio. In effect, the risk of proteinuria occurrence increased 1.4-fold for the low ratios <2 g/kg, compared with DKW/RBW ratios between 2 and 4 g/kg (P = 0.03) and 2-fold compared with DKW/RBW ratios ≥4 g/kg (P < 0.001). Delayed graft function and the recipient’s being male were also found to be risk factors for the occurrence of proteinuria.

Graft Survival Analysis (n = 905)

Despite the effects of small DKW/RBW ratios on the filtration rate and proteinuria, no effect of DKW/RBW ratios per se on graft survival was observed according to Cox multivariate analysis at the end of the study survey (Figure 3). However, the risk of graft loss increased significantly with increasing donor age (in years; relative risk 1.02; 95% confidence interval 1.00 to 1.03; P = 0.019) and for six HLA incompatibilities (relative risk 3.10; 95% confidence interval 1.39 to 6.81 P < 0.0053). When the covariates that also were found to be significantly associated with proteinuria were tested for a possible synergistic effect with low DKW/RBW ratios (<2 g/kg; recipient gender and delayed graft function, see Table 4), no significant impact on graft survival was noted. Finally, no additional effect of other parameters classically associated with poor long-term graft survival (cold ischemia time, panel reactive antibodies, acute rejection episodes, and donor and recipient age) was observed.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Kaplan-Meier graft survival according to the DKW/RBW ratio. The DKW/RBW ratios have no impact on graft survival at the end of the study survey. One and 4-yr graft survival was 98 and 89% for patients with small DKW/RBW ratios <2 g/kg, 94 and 88% for patients with ratios between 2 and 4 g/kg, and 94 and 85% for patients with ratios ≥4 g/kg, respectively.