Among the 487 living donors, mean donor age was 40 years (range 20 to 65), 88% were white, and 62% were biologically related (). Of the 487 donors, 486 had baseline, pre-donation laboratory results, and 466 (94%) had laboratory results from at least one post-donation time point (235 from 4 or more time points, 123 from 3, 76 from 2, and 32 from 1 time point).
| Table 1Characteristics of 487 Liver Transplant Donors at Time of Evaluation |
and depict the distributions of laboratory values over time from donation through year 4. The greatest changes occurred in the immediate postoperative period, as shown in the week 1 levels, and most values approached baseline levels by three months. However, statistically significant differences in median values compared to baseline were noted for up to one year for AST, alkaline phosphatase, total bilirubin, INR, and albumin. Platelet counts were significantly lower than pre-LT at each of the follow-up intervals (). The average reduction in platelet count from baseline to year 4 was 16%.
and show scatter plots of the laboratory values at year 1 plotted against baseline values for each patient. Points below the equivalence line indicate patients whose one year values were lower than baseline and points above the line indicate one year values that were higher than baseline. For ALT and INR, the distribution is equally displaced around the equivalence line, suggesting that there was no overall shift in values between baseline and one year. show the week 1 and month 1 INR values plotted against baseline values; INR increased substantially at week 1, but was nearly back to normal at month 1. The variances of the distributions of AST (p<0.0001), alkaline phosphatase (p<0.0001) and bilirubin (p=0.09) were larger at one year compared to baseline, mostly due to more outliers at one year. Albumin showed a modest shift downward, but no difference in variance between baseline and one year (p=0.39). show the week 1 and month 1 albumin values plotted against baseline values; albumin decreased substantially at week 1, and recovered only partially by month 1. The most striking changes occurred in the platelet count, where there was an obvious downward shift in the platelet distribution post-donation, and a significant decrease in variance (p=0.004) at one year compared to baseline (). Compared to baseline, 89.4% of patients had lower platelet counts at one year. Donors with a platelet count ≤150 (×1000/mm3) at one year had had a significantly lower mean platelet count at evaluation (189±32) than the remainder of the cohort (267±56, p<0.0001).
To further investigate the relationships shown in and , comparisons were made of abnormal laboratory values at baseline and at the first follow-up visit that occurred one year or more post donation (), testing the likelihood of becoming abnormal when previously normal versus becoming normal when previously abnormal. Significantly more abnormal values were found at one year or later for AST, albumin, and platelet count, and nearly significant results were found for alkaline phosphatase and total bilirubin (p=0.07 and 0.08, respectively)
| Table 2Abnormal laboratory values among donors with values both at baseline (pre-donation) and at first follow-up at least one year later. |
For all seven laboratory measures, covariates tested for their ability to predict the difference between the one year and baseline value included age, sex, weight at baseline, weight change between baseline and one year, the baseline value itself, and interactions between these variables and the baseline value. Donors tended to gain weight after transplant (mean=1.0 kg, standard deviation=5.7, p=0.02). A 10-year older age was associated with higher values at one year for ALT (b=1.9 IU/L, p=0.011), AST (b=1.7 IU/L, p=0.001), and AP (b=4.5 IU/L, p=0.0003), and with lower values for albumin (b= −0.06 g/dl, p=0.006). Women had lower values at one year for bilirubin and albumin, by an average of −0.13 mg/dl (p=0.0008) and −0.11 g/dl (p=0.030), respectively. Baseline values were highly predictive of values at one year for all laboratory measures, as is obvious in and (p<0.05 for all; p<0.0001 for most). Weight change from baseline was significantly predictive of AP (b=0.64/kg, p=0.013), and was close to significant for ALT (b=0.30/kg, p=0.082) and bilirubin (b= 0.006/kg, p=0.086). No other factors were predictive of the change in platelets from baseline to one year.
We further tested the correlation between platelet counts and WBC, which would be expected if the drop in platelet counts were related to portal hypertension. We found significant (all p<0.0001) correlations at all time points tested (through year 3), most substantially larger than the correlation seen at baseline (r=0.23): week 1 (r=0.32), month 1 (r=0.40), month 3 (r=0.22), year 1 (r=0.35), year 2 (r=0.45), year 3 (r=0.43).
Because the higher graft weight to total liver volume ratio (GW/TLV) has been associated with both larger spleen volume and lower platelet counts following donation,
17 we tested the correlation between GW/TLV and the platelet count at each post-donation time point divided by the baseline count. The early post-donation correlations were significant, but not at one year or beyond (week 1, r = −0.16, p=0.011; month 1, r = −0.19, p=0.004; month 3, r = −0.21, p=0.003; year 1, r = −0.09; p=0.29; year 2, r = −0.03, p=0.83; year 3, r = −0.08, p=0.60). We also tested GW/TLV in the year 1 platelets regression model reported above, but it was not significant (p=0.853). However, in similar models for months 1 and 3, GW/TLV was a significant predictor of a reduction in platelets. For an increase of 0.10 in GW/TLV, platelets were lower on average by 12.6 (×1000/mm
3) at month 1 (p=0.013) and by 7.0 (×1000/mm
3) at month 3 (p=0.013). We note that left lobe donors (n=24) had lower GW/TLV (p<0.0001) and higher platelet counts at week 1 (p=0.029) than right lobe donors. Platelet counts were not significantly different at subsequent time points between left and right lobe donors.
Additional laboratory tests were analyzed, but are not reported in detail because the findings were less likely to be clinically relevant: WBC, ferritin, HGB, SCR, and BUN. Median values of these tests followed similar trends to the laboratory medians discussed above. SCR, BUN, and HGB median values all dropped significantly from baseline to week 1 and recovered gradually to baseline values by year 1 (SCR, BUN) or year 2 (HGB). WBC peaked significantly at week 1 then returned to baseline values by month 1. Ferritin did not stabilize until year 2: it peaked at week 1, then descended below the baseline value through month 3 and slowly increased to approach the baseline value by year 2.