Long-term follow-up of donor chimerism and tolerance after human liver transplantation

Authors


  • The authors thank Ian Ure for the English review and Romana Santorum for technical support.

Abstract

We aimed to quantify peripheral donor chimerism (DC) and to analyze its association with graft and recipient outcome. Forty-two liver transplant recipients and their respective donors were studied, providing a total of 148 posttransplantation serum samples. DC was assessed with real-time quantitative polymerase chain reaction (qPCR) to detect polymorphic markers. DC did not decrease with time post-transplantation and was higher in child recipients versus adults and in recipients of deceased donor liver transplants versus recipients of live donor liver transplants. Higher levels of DC were detected in Rh-positive blood group donors, in O blood group recipients versus A blood group recipients, and in recipients with hepatitis C virus versus recipients with alcoholic cirrhosis. High DC was associated with patients with organ damage due to recurrent disease and rejection. Stable, high levels of DC, in the absence of other major clinical events, may thus be a marker of transplantation tolerance, and this knowledge may help to tailor immunosuppressive treatment. In conclusion, qPCR is a useful technique for DC follow-up in liver transplantation, although the evolution of DC levels should be analyzed in accordance with the clinical outcome of the patient. Liver Transpl 15:581–591, 2009. © 2009 AASLD.

The quantification of hematopoietic chimerism after allogeneic stem cell transplantation is an important tool for monitoring posttransplant outcome.1, 2 Several studies suggest that a serial quantitative analysis of chimerism would enable early detection of patients at high risk of disease relapse, graft rejection, or graft-versus-host disease.3–6 The phenomenon of chimerism has also been observed in solid organ transplantation and in liver transplantation (LT), in which the occurrence of graft-versus-host disease is an infrequent but serious complication.7–9 The impact of donor chimerism (DC) on graft outcome after solid organ transplantation is, however, not universally accepted.

Limitations of methods previously used to detect DC, in that they are either semiquantitative10 or of low sensitivity [flow cytometry, polymerase chain reaction (PCR)/single-strand conformation polymorphism, Fluorescent In Situ Hybridization X/Y, or Short Tandem Repeats],11–13 could play a part in the controversy surrounding the impact of DC. Recently, real-time quantitative polymerase chain reaction (qPCR) methods using TaqMan (TM)1, 14–16 or hybridization (Hyb) technology17 have been developed that show higher sensitivity and accuracy and faster performance than previous techniques.

Starzl and Zinkernagel hypothesized that organ engraftment is a form of chimerism-dependent partial tolerance18–22 and that the completeness of this tolerance can be inferred from the amount of immunosuppression necessary to maintain stable function and structure of the graft.23, 24 Other studies support the idea that DC may be associated with long-term organ allograft survival but suggest that it is not essential for the induction or maintenance of the tolerant state.11 They also suggest that low-level (micro)chimerism is an epiphenomenon secondary to graft acceptance by T-regulatory lymphocytes or alternative mechanisms.25–29 Nonetheless, Zinkernagel et al.30 found that leukocyte chimerism, even at a microlevel, is a prerequisite for the perpetuation of allotolerance.

We performed a serial quantification of DC levels in serum samples after LT to analyze any association of DC levels with the clinical evolution of patients and grafts post-transplant. We detected bi-allelic single nucleotide polymorphisms (SNPs) using TM probes and insertion-deletion di-allelic polymorphisms using Hyb probes. Cell-free DNA in the serum samples was predominantly of a hematopoietic origin with minimal but indeterminate amounts of DNA from nonhematopoietic tissues.31–33

Evaluating DC levels with sensitive methods may help us understand the real value of chimerism. It may also help us to detect recurrent disease and early rejection or to monitor immunosuppression in LT.

Abbreviations

CE, clinical event; CI, confidence interval; CIN, calcineurin inhibitor; Cp, crossing point; CsA, cyclosporine A; DC, donor chimerism; DDLT, deceased donor liver transplantation; FK, tacrolimus; HBV, hepatitis B virus; HCV, hepatitis C virus; Hyb, hybridization; LDLT, live donor liver transplantation; LT, liver transplantation; OR, odds ratio; PCR, polymerase chain reaction; qPCR, quantitative real-time polymerase chain reaction; SD, standard deviation; SNP, single nucleotide polymorphism; TM, TaqMan; VC, variation coefficient.

PATIENTS AND METHODS

Characteristics of the Recipients and Donors

A total of 42 LT patients and their respective donors took part in this study. All patients underwent transplantation at the General Surgery Alimentary Tract and Abdominal Organ Transplantation Service of the Doce de Octubre University Hospital (Madrid, Spain) between June 2001 and January 2005. Informed consent was obtained from all donors/patients according to institutional guidelines.

The demographic data of the recipients and donors, the primary diagnosis and reason for LT, and the types of LT are summarized in Table 1. The final study was conducted with 39 LT recipients: 20 patients underwent live donor liver transplantation (LDLT), and 19 underwent deceased donor liver transplantation (DDLT). Thirty patients were adults, and 9 were children. Twenty-six patients were men, and 13 were women. There was gender mismatch between the recipients and donors in 14 cases, with 7 female patients receiving liver grafts from male donors and the remaining 7 males receiving liver grafts from female donors. All patients received standard immunosuppressive therapy, as indicated in Table 2.

Table 1. Characteristics of Liver Transplant Recipients
No.Sex/Age of Recipient at TransplantationCause of TransplantationSex/Age of DonorType of Liver TransplantationSplitRecipient/Donor ABO
  1. Abbreviations: DDLT, deceased donor liver transplantation; F, female; HBV, hepatitis B virus; HCV, hepatitis C virus; M, male; LDLT, live donor liver transplantation.

1M/53HCV cirrhosisM/25LDLTYes0/0
2M/53HBV cirrhosisM/19LDLTYes0/0
3F/3Hepatic fibrosisM/43LDLTYesA/A
4F/1Biliary atresiaF/35LDLTYes0/0
5F/2Biliary atresiaM/35LDLTYesA/A
6M/51Alcoholic cirrhosis and HCV cirrhosisF/23LDLTYes0/0
7M/2Byler syndromeM/26LDLTYesA/A
8M/57Septal cirrhosisF/43LDLTYesB/0
9M/54Alcoholic cirrhosisF/31LDLTYesA/A
10F/44Biliary cirrhosisM/32LDLTYesA/A
11M/55HCV cirrhosisM/23LDLTYesA/A
12M/1Biliary atresiaF/38LDLTYesA/A
13M/56HCV cirrhosisF/52LDLTYesA/A
14M/58Alcoholic cirrhosisM/23LDLTYes0/0
15M/56HBV cirrhosisM/24LDLTYesA/0
16M/56HBV cirrhosisF/26LDLTYesA/A
17F/10CryptogenicM/19LDLTYesB/0
18F/52HCV cirrhosisF/25LDLTYesA/A
19F/9CryptogenicF/34LDLTYesAB/B
20M/37Autoimmune cirrhosisM/26LDLTYes0/0
21M/55HCV cirrhosisM/56DDLTNoA/A
22M/51Alcoholic cirrhosisM/21DDLTNoA/A
23F/61HCV cirrhosisM/61DDLTNo0/0
24M/25HBV cirrhosisM/21DDLTNoB/B
25M/54Alcoholic cirrhosisM/40DDLTNoA/A
26F/57HCV cirrhosisF/83DDLTNo0/0
27M/54HBV cirrhosisM/69DDLTNo0/0
28F/38Alcoholic cirrhosisM/69DDLTNoA/A
29M/41HCV cirrhosisM/20DDLTNoB/B
30F/45HCV cirrhosisF/42DDLTNo0/0
31M/56Alcoholic cirrhosisM/34DDLTNoA/A
32M/39Alcohol and HCV cirrhosisM/53DDLTNoA/A
33M/57Alcoholic cirrhosisM/55DDLTNo0/0
34F/67Micronodular cirrhosisF/71DDLTNo0/0
35F/45Alcoholic cirrhosisM/24DDLTYesA/A
36M/48Alcoholic cirrhosisM/27DDLTNo0/0
37M/32Submassive necrosisM/73DDLTNoB/B
38M/69Deficit of alpha 1 antitrypsinF/33LDLTYes0/0
39M/12HCV cirrhosisM/3DDLTNoAB/0
Table 2. Donor Chimerism Results and Immunosuppression Therapy
Case No.Sex/Age of Recipient at TransplantEventTime of CEs (Months)Partial Liver GraftChimerism % (Months Post-Transplant) and Therapy/CINs/Antiviral
0–12 Months Post-TransplantAfter 12 Months Post-Transplant
  1. Abbreviations: antiviral, antiviral therapy of ribavirin and interferon; CE, clinical event; CIN, calcineurin inhibitor (tacrolimus or cyclosporine A); CsA, cyclosporine A; D, double therapy of tacrolimus or cyclosporine A and mycophenolate; F, female; FK, tacrolimus; HBV, hepatitis B virus; HCV, hepatitis C virus; M, male or monotherapy of tacrolimus or cyclosporine A; T, triple therapy of tacrolimus or cyclosporine A, mycophenolate, and prednisone.

14M/58Rejection0.6YesCE9.67 (2.5)2.89 (3)3.06 (4.5)1.17 (5.5)     
      T/FK/noT/FK/noT/FK/noD/FK/no     
7M/2Chronic rejection6.5YesCE6.78 (10)5.2 (16)       
      D/FK/noD/FK/no       
32M/39Recurrent HCV, rejection5.5No3.97 (0.5)7.53 (4)CE10 (10.5)      
     T/CsA/noT/CsA/no D/CsA/no      
27M/54Rejection3.5NoCE    0.7 (16)0.09 (23.5)   
          D/FK/noD/FK/no   
6M/51Rejection2 daysYesCE0.28 (5)1.78 (10)  0.03 (15)    
      D/CsA/noD/CsA/no  D/CsA/no    
24M/25Rejection1.5NoCE0.23 (3.5)0.17 (5.5)0.46 (7) 0.09 (13.5)    
      T/CsA/noT/CsA/noD/CsA/no D/FK/no    
20M/37Rejection26 daysYes0.021 (0.25)CE        
     T/FK/no         
13M/56Rejection, recurrent HCV5Yes0.00 (3)CE2.92 (9)  0.88 (13)    
     T/CsA/no T/CsA/no  D/CsA/no    
33M/57Cholangitis, rejection20No0.00 (0.5)20.28 (1)79.27 (4)0.00 (11) CE    
     T/CsA/noT/CsA/noT/CsA/noM/CsA/no      
34F/67Rejection0.5NoCE 0.19 (1)0.08 (5)0.00 (7)0.00 (8)0.22 (10)     
      D/FK/noD/FK/noD/FK/noD/FK/noD/FK/no    
8M/57Rejection, exitus13 daysYesCE0.12 (0.25)        
      T/FK/no        
38M/69Recurrent deficit of alpha 1 antitrypsin11Yes70.19 (1)89.53 (2)82.9 (3)79.72 (6)CE97.8 (17)    
     M/FK/noM/FK/noM/0/noM/0/no M/0/no    
29M/41Recurrent HCV2NoCE68.62 (4)41.1 (5)29.86 (8.5)22.27 (12)19.48 (17)51.95 (27)   
      T/FK/noD/FK/noM/FK/yesM/FK/yesM/FK/yesM/FK/no   
2M/53Recurrent HBV1YesCE19.32 (12)   33.9 (15)12.02 (18)19.1 (27)17.7 (35) 
      M/FK/yes   T/FK/yesT/FK/yesT/FK/yesT/FK/yes 
30F/45Recurrent HCV15No16.47 (4)71.91 (7.5)   CE22.12 (16)0.00 (20)0.00 (23) 
     D/FK/noD/FK/no    D/FK/noD/FK/yesM/FK/yes 
11M/55Recurrent HVC6Yes2.6 (0)3.1 (3)CE       
     T/FK/noT/FK/no        
21M/55Recurrent HCV1.5NoCE1.56 (4)1.66 (10)  11.24 (15)0.00 (19)0.00 (29.5)  
      D/FK/noM/FK/no  M/FK/noM/FK/noM/FK/no  
1M/53Recurrent HVC22Yes     CE 0.29 (21)1.35 (23.5)7.4 (28)1.88 (36)1.88 (36)
          D/FK/noD/FK/noM/FK/noM/FK/noM/FK/no
18F/52Recurrent HVC2Yes0.073 (0.5)CE0.05 (2)0.05 (3.5)      
     T/FK/no T/FK/noT/FK/no      
16M/56Recurrent HBV0.5YesCE0.03 (3.5)        
      T/FK/no        
23F/61Recurrent HCV4.5No0.00 (2.5)0.00 (4)CE0.00 (9)0.00 (12)10.24 (18)3.83 (22)   
     D/CsA/noD/CsA/no M/CsA/noM/CsA/noM/FK/noM/FK/no   
26F/57Recurrent HCV5No0.00 (2.5)CE0.00 (9.5)  8.31 (21)0.00 (22)0.00 (32)  
     T/FK/no M/FK/no  M/FK/noM/FK/noM/FK/no  
39M/12Cholestasis57No64.24 (3)69.21 (6)70.1 (9)79.11 (12) 64.42 (25)56.98 (39)CE  
     D/FK/noD/FK/noD/FK/noD/FK/no D/FK/noD/FK/no   
15M/56Cholestasis10 daysYesCE18.87 (1)5.38 (4.5)5.6 (8)      
      T/FK/noT/FK/noD/FK/no      
19F/9Cholestasis5 daysYesCE0.98 (0.25)        
      M/FK/no        
4F/1Thrombosis of liver arteries16Yes35.81 (11)    CE19.86 (17)16.91 (23)  
     M/FK/no     M/FK/noM/FK/no  
37M/32Necrosis1NoCE 0.00 (2)0.31 (3)0.00 (3.5)0.00 (5)0.00 (7)     
      M/FK/noM/FK/noM/FK/noD/FK/noD/FK/no    
31M/56No events No47.98 (0.5)18.61 (1.5)36.59 (3)6.61 (10.5) 22.46 (27.5)    
     D/FK/noD/FK/noD/FK/noD/FK/no D/FK/no    
25M/54No events No4.55 (0.75)0.00 (1.75)0.00 (3)0.66 (11) 0.00 (22)0.00 (25)0.00 (35)  
     D/FK/noD/FK/noM/FK/noM/FK/no M/FK/noM/FK/noM/FK/no  
28F/38No events No4.37 (1)4.72 (1.5)   0.77 (14.5)    
     D/FK/noD/FK/no   M/FK/no    
17F/10No events Yes4.03 (0.5)10.64 (1)7.17 (2)       
     M/FK/noM/FK/noM/FK/no       
22M/51No events No2.2 (1)0.36 (2.5)0.00 (3.5)  0.00 (15)0.03 (21)   
     T/CsA/noT/CsA/noT/CsA/no  D/CsA/noD/CsA/no   
9M/54No events Yes1.07 (1)0.5 (3)1.2 (9)  0.24 (20)0.19 (25)   
     T/FK/noT/FK/noD/FK/no  M/FK/noM/FK/no   
3F/3No events Yes0.59 (2.5)1.46 (4)        
     M/FK/noM/FK/no        
10F/44No events Yes0.57 (1)1.47 (3)1.36 (10)1.36 (10)      
     T/FK/noD/FK/noD/FK/noD/FK/no      
35F/45No events Yes0.45 (0.5)0.04 (2)0.00 (3)0.015 (6)0.00 (7)     
     D/FK/noD/FK/noM/FK/noM/FK/noM/FK/no     
36M/48No events No0.07 (1)0.00 (2)0.00 (3)       
     T/CsA/noT/CsA/noT/CsA/no       
5F/2No events Yes0.00 (3)0.00 (7)   0.00 (13)    
     M/FK/noM/FK/no   M/FK/no    
12M/1No events Yes0.00 (6.5)0.00 (11)4.24 (12)       
     M/FK/noM/FK/noM/FK/no       

Sample Collection and DNA Preparation

A total of 226 samples were studied. Seventy-eight were pretransplant samples from both the donor and the recipient, and 148 were posttransplantation samples. Samples were collected on the basis of the availability of donor samples and informed consent of the recipient in all LT procedures between June 2001 and January 2005. Serum samples were processed within 6 hours of collection, centrifuged at 3000 rpm for 10 minutes, and stored at −20°C. Genomic DNA was extracted from serum samples with the MagNAPure LC DNA isolation kit (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions.

Real-Time qPCR

qPCR was performed on a LightCycler 2.0 instrument (Roche Diagnostics) using TM or Hyb probes in combination with LightCycler FastStar DNA Master Hybridization probes (Roche Diagnostics). Sequences of the primers and probes (synthesized by TIB MolBiol, Berlin, Germany) used to amplify each polymorphic marker were obtained from previously published data.1, 17

Seven insertion-deletion di-allelic polymorphisms (GSTM, GSTT, SRY, Xq28, FVII, RhD, and rs4399) were detected with Hyb probes, and 14 SNPs (S1a, S1b, S4a, S4b, S5a, S5b, S7a, S7b, S8a, S8b, S9a, S9b, S11a, and S11b) were detected with TM probes.

PCRs were always carried out in a 10-μL volume, and 2 different PCR mixtures were used. The TM PCR mixture consisted of 2 μL of DNA, 1 μL of the LightCycler FastStar DNA Master Hybridization probe, 4 mM of MgCl2, 0.6 μM of each primer, and 0.2 μM TM probe. The Hyb PCR mixture consisted of 2 μL of DNA, 1 μL of LightCycler FastStar DNA Master Hybridization probes, 2.5 mM of MgCl2, 0.4 μM (for SRY polymorphism) or 0.5 μM (for GSTT1, GSTM1, Xq28, RhD, FVII, and rs4399 polymorphisms) of each primer, and 0.2 μM of each Hyb probe. In the TM PCR, the cycling conditions were 10 minutes at 95°C, 40 cycles of 95°C for 45 seconds, 60°C for 1 minute, and 72°C for 15 seconds. In the Hyb PCR, the cycling conditions were 95°C for 10 minutes, 40 cycles of 95°C for 10 seconds, 52°C (for GSTT1 and SRY polymorphisms), 55°C (for GSTM1 polymorphism), 58°C (for Xq28 polymorphism), 56°C (for RhD), 62°C (for FVII), or 60°C (for rs4399) for 10 seconds, and 72°C for 15 seconds. Melting curves of the PCR products were consistent with the expected products.

Genotyping and Chimerism Analysis

Donor and recipient genotypes were identified by real-time PCR (as described previously) before chimerism quantification. For each run and sample, the LightCycler software calculated the crossing point (Cp) value, and negative polymorphisms were defined as those with a Cp value over 36 or 38 according to the particular SNP-PCR. For chimerism quantification, a polymorphism was considered informative when it was positive on donor DNA and negative on recipient DNA. The chimerism analysis included 2 independent qPCRs, each performed in duplicate on post-LDLT DNA samples and on pre-LDLT donor DNA.

qPCR was always carried out with positive and negative controls. To control the variation between different assays and samples, a reference gene (human glyceraldehyde 3-phosphate dehydrogenase) was amplified simultaneously. For each experimental sample, duplicate Cp values were analyzed with the Pfaffl method.34 Thus, the relative amount of target DNA calculated (polymorphic positive marker) was based on qPCR efficiency and the Cp deviation of an unknown sample (post-LDLT or artificial chimeric DNA) versus a calibrator sample (donor DNA). It was expressed with respect to a reference marker (glyceraldehyde 3-phosphate dehydrogenase). The calibrator sample was considered to be 100% DC, and the relative amount obtained from the unknown sample was expressed as a percentage of donor DNA in the artificial or posttransplant sample.

Comparison of DC Quantification in Peripheral Blood and Serum Samples

To assess the robustness of the assay, DNA extracted from both peripheral blood and serum from healthy individuals (n = 6) was tested. Serum samples were processed immediately after collection and were centrifuged at 3000 rpm for 10 minutes, and genomic DNA was isolated as previously described. Whole blood samples were also processed immediately after collection, and genomic DNA was isolated with the same method.

To compare the assays of peripheral blood and serum, SNP-positive DNA was serially diluted (100%, 50%, 10%, 1%, and 0.1%) into SNP-negative DNA (from both peripheral blood and serum), and the DC level was quantified in these artificial samples. Each point in the dilution curve was measured in triplicate, and the standard deviation of triplicate Cp values was always lower than 0.3 Cp.

Statistical Analyses

Statistical analyses were performed with SPSS statistical software, version 11.0.1 (SPSS, Chicago, IL). To analyze associations between discrete and categorical numerical variables, the χ2 test and Fisher test were used. A comparison of the mean values of different groups was carried out with the Student unpaired t test, the Wilcoxon test, and the analysis of variance test. A P value less than 0.05 was considered to be significant.

To compare chimerism from peripheral blood and serum and to establish the degree of concordance, the Pearson correlation coefficient and an imprecision test were used. A forward stepwise regression logistic analysis was performed to find variables associated with DC and graft rejection. A high DC level was defined as the persistence of DC above the median value. Macrochimerism was defined as the persistence of >1% of circulating donor cells. The studied variables in the regression analysis were (1) recipient characteristics (age, adult or child, gender, ABO blood group, and Rh blood group), 2) donor characteristics (LDLT or DDLT, gender, ABO blood group, and Rh blood group), (3) immunosuppressive therapy (mono, double, or triple), and (4) graft outcome. A diagnosis of clinical events was established by biopsy.

A survival analysis was also performed. For univariate analysis, survival curves were calculated according to the Kaplan-Meier method, and differences between curves were evaluated with the log-rank test. The multivariate study was performed with a Cox proportional regression hazard model to identify factors that may have a significant independent influence on overall survival and rejection-free survival. The median value was used to define the higher or lower DC groups.

RESULTS

DC Quantification and Detection

An informative polymorphism between the recipient and donor was found in 39 of the 42 (93%) cases. In the remaining 3 cases, which involved family member donors, no informative polymorphism was detected. In DDLT, an informative polymorphism was established in 100% of cases. In all cases, DC was determined in serum samples with an approximate DNA amount 40 times less than that from whole peripheral blood.

The follow-up time post-transplantation to DC study was 46.9 months (range, 0–69.38 months). DC was detected in 111 of the 148 samples (75%; Table 2) with a median of 1.27% (range, 0%–97.85%). The median DC detected during the first 3 months after transplantation was 0.93% (range, 0%–89.53%); in samples between 4 and 12 months post-transplantation, it was 1.35% (range, 0%–79.72%); and in samples of more than 12 months after transplantation, it was 1.88% (range, 0%–97.85%; P = 0.784).

Sustained macrochimerism (>1%) was found in 14 LT recipients (35.9%), and in 6 of these (15.4%), DC was detected more than 2 years after LT. Sustained microchimerism (<1%) was detected in the 21 remaining recipients.

Comparison of DC Quantification in Peripheral Blood and Serum Samples

The median DNA concentration obtained from peripheral blood and serum was 41.9 ng/μL (range, 34.4–50.3 ng/μL) and 1.6 ng/μL (range, 0.4–3.7 ng/μL), respectively.

Correlation Between Blood and Serum Samples

The median target Cp/control Cp ratio in peripheral blood and serum was 1.07 [95% confidence interval (CI), 1.04–1.08] and 1.09 (95% CI, 1.07–1.13), respectively. There was no significant difference between the two (Wilcoxon test, Z-0.7207; P = 0.4711). The Pearson correlation coefficient between peripheral blood and serum assays was 0.83 (P < 0.0001).

Assay Imprecision

We determined the intra-assay variation coefficients (VCs) for DNA obtained from serial dilutions of peripheral blood and serum mixtures with different target polymorphism concentrations. The VCs for the SRY, S5B, S8B, and S9B polymorphism Cp in the DNA obtained from recipient peripheral blood containing 13, 4.6, 0.46, and 0.046 ng/μL mean donor DNA were 0.4%, 0.6%, 0.5%, and 0.8%, respectively. In the case of serum dilutions with 0.9, 0.2, 0.02, and 0.002 ng/μL mean donor DNA, the VCs were 0.6% and 1.16%, respectively for the 2 first serum dilutions. DC was not detected in the serum dilutions containing 0.020 and 0.002 ng/μL donor DNA. With SNP-positive DNA diluted in SNP-negative DNA, the detection limit was 0.04 ng/μL. Samples with less than 1 ng/μL SNP-positive DNA showed a relatively high VC (1.16%) in comparison with other concentrations (0.4%–0.8%).

DC and Immunosuppressive Therapy

There was no significant correlation between DC and immunosuppressive therapy for each transplant recipient during this study. Although there was an inverse correlation between immunosuppressor levels and the percentage of DC, there were no differences in the level of DC when patients receiving mono, double, or triple immunosuppressive therapies were compared.

The median and mean of DC detected in patients on 1 immunosuppressant drug were 0.45% and 7.13% (CI, 3.08–11.19), respectively, whereas with 2 drugs, they were 1.34% and 12.94% (CI, 5.59–20.28), respectively, and with 3 drugs, they were 0.57% and 10.7% (CI, 2.03–14.42), respectively (P = 0.469). DC was higher in patients receiving tacrolimus versus those on cyclosporine A (median DC, 1.35% and 0.076%, respectively; P = 0.079). However, no correlation was found between tacrolimus and cyclosporine A levels and DC.

DC and Recipient/Graft Characteristics

No correlation was found between DC and characteristics of the recipient patient, that is, gender, liver disease, ABO blood group, and Rh blood group (Table 3), apart from age (adult and child median DC, 1.22% and 5.72%, respectively; P = 0.023). DC levels tended to be higher in the O blood group liver recipients than in the A blood group liver recipients (2.24% and 1.27%, respectively; P = 0.095). No relationship was found between DC and graft characteristics, that is, gender, graft/donor gender match, partial liver graft, ABO blood group, and ABO matching (Table 3), although DC was significantly lower in Rh-negative grafts than in Rh-positive grafts (median DC, 1.27% and 2.24%, respectively, in univariate analyses; P = 0.047). DC levels tended to be lower in LDLT compared with DDLT (1.23% and 2.24%, respectively; P = 0.067).

Table 3. Results of DC with Respect to Liver Transplant Recipient Characteristics and Donor Characteristics
Characteristic of RecipientsGroupnMedian DC*Average DC*SDP
  • Abbreviations: DC, donor chimerism; HBV, hepatitis B virus; HCV, hepatitis C virus; SD, standard deviation.

  • *

    DC percentage of the donor DNA in the DNA from the recipient serum:

  • Quantification = equation image ÷ equation image

AgeAdult301.225.8810.030.023
 Child95.7221.6632.47 
SexMale261.9811.7821.770.287
 Female131.194.998.30 
CauseHCV112.7020.8430.390.202
 HBV45.107.669.67 
 Alcoholic cirrhosis100.706.1510.38 
 Other130.113.286.69 
 Alcoholic cirrhosis + HCV17.177.17  
ABO groupGroup O142.2413.2522.770.095
 Group A181.273.786.27 
 Group B50.189.3216.81 
 Group AB235.5635.5648.89 
Rh blood groupPositive192.1312.2220.990.383
 Negative201.346.9616.02 
Characteristics of DonorsGroupsnMedian DC*Average DC*SDP
SexMale262.708.9815.960.801
 Female130.9810.6023.60 
Gender match for recipient and donorYes252.7010.7116.590.600
 No141.187.4022.14 
Source of graftLiving201.234.246.760.067
 Deceased192.2415.0824.79 
Partial liver graftYes221.227.6818.250.488
 No172.2411.9019.21 
ABO groupGroup O182.3415.1724.410.174
 Group A171.273.416.26 
 Group B40.6110.0419.22 
 Group AB0    
Rh blood groupPositive222.2414.2423.220.047
 Negative171.273.416.26 
ABO group match for recipient and donorYes341.419.1418.940.743
 No51.1712.1117.29 

Although no variable was associated with the presence of macrochimerism (>1% DC), DC increased with the time post-transplantation. Time post-transplantation was the only variable associated with higher DC when a multivariable analysis was performed [odds ratio (OR), 1.128; 1.004–1.268; P = 0.042].

DC and Clinical Outcome

The median and mean DC of all the patients were 1.27% and 12.6%, respectively (range, 8.8%–16.4%) versus the median and mean DC from 27 samples in 12 patients with ongoing graft rejection, which were 0.39% and 5.11%, respectively (1.0%–11.2%). This difference was, however, not significant. The median and mean DC levels in 55 samples from 12 patients with no apparent clinical events were 1.17% and 12.32%, respectively, whereas the median and mean DC in 43 samples from 9 patients with recurrent liver disease were 2.70% and 19.11%, respectively. The median and mean DC in 11 samples from 3 patients with both recurrent disease and rejection were 5.47% and 11.57%, respectively, and in 12 samples from 8 patients with other events, the levels were 7.17% and 8.64%, respectively. There was macrochimerism in 3 (33%) of the 9 patients with graft rejection, in 8 (80%) of the 10 patients with posttransplantation recurrent liver disease, and in 7 (53.8%) of the 13 patients with no significant clinical events.

Three patients had very high DC (50%–80%) and a long posttransplant follow-up (Table 2). The first patient (case 38; DC median and mean, 70.19% and 89.53%, respectively) was a 69-year-old female with α1-antitrypsin deficiency who underwent LDLT. She presented with diabetes and, because of severe, multiple infectious complications, was only on mycophenolate mofetil as immunosuppressant therapy. The second patient (case 29; DC median and mean, 19.48% and 51.95%, respectively) was a 41-year-old male with hepatitis C virus (HCV) cirrhosis who underwent DDLT. Immediately after transplantation, he presented with recurrent HCV. His immunosuppressant therapy was tapered, and he received low doses of tacrolimus, interferon, and ribavirin for 2 years (the time during which the study was performed). The third patient (case 39; DC median and mean, 56.98% and 79.11%, respectively) was a 12-year-old male with HCV cirrhosis and Wilson's disease who underwent DDLT. He failed to follow medical instructions and received low and partial immunosuppressive treatment.

Two other patients on antiviral treatment (interferon and ribavirin or lamivudine) because of recurrent HCV or hepatitis B virus also had high DC levels. The first patient (case 2) had a DC median and mean of 12.02% and 33.9%, respectively, and the second (case 30) had a DC median and mean of 22.12% and 71.91%, respectively. However, there was no significant difference between the mean DC from patients who were receiving antiviral treatment and those who were not (DC median and mean, 18.4% and 14.46% versus 0.69% and 9.22%, respectively; P = 0.403). The 5 patients with high macrochimerism had no evidence of rejection at the time of DC quantification and were followed up for a median of 51 months post-transplantation (Fig. 1).

Figure 1.

Evolution of DC and levels of immunosuppression in 3 liver transplant patients with high macrochimerism and with no evidence of rejection at the time of DC quantification. Abbreviations: DC, donor chimerism; HBV, hepatitis B virus; HVC, hepatitis C virus.

In this multivariable study, no clinical event (rejection, recurrent liver disease, or drug toxicity) was associated with DC levels.

DC Impact on Survival

The median overall survival and median rejection-free survival were 51.24 and 45.40 months, respectively, in the univariate analysis. There was a tendency toward longer survival in patients with higher DC levels compared with those with lower DC levels (100% survival versus 89.7%), with a median follow-up of 44.92 months for the former and a median follow-up of 39.24 months for the latter. However, the difference was not significant (P = 0.0988).

Although there was a tendency for graft rejection in recipients with lower DC, the difference was not significant. The median rejection-free survival in recipients with lower DC was 44.48 months (range, 31.25–57.72 months), whereas the median rejection-free survival in recipients with higher DC was 55.18 months (44.17–66.9; P = 0.4439).

Several donor and recipient characteristics, such as age, gender, live or deceased donor, ABO blood group (O, A, B, or AB), Rh blood group (positive or negative), reason for LT (HCV, hepatitis B virus, alcoholism, or other), and mean donor level of peripheral chimerism were included in the Cox test. None of the variables demonstrated any influence on overall survival. On the other hand, Rh-positive recipient (positive/negative; OR, 0.07; CI, 0.013–0.385; P = 0.002), LDLT (LDLT/DDLT; OR, 0.161; CI, 0.035–0.743; P = 0.019), and recipient male gender (male/female; OR, 0.46; CI, 0.004–0.472; P = 0.010) had an adverse prognostic impact on rejection-free survival.

DISCUSSION

DC analysis by qPCR is feasible in most LT cases and has several advantages over previously employed methods such as flow cytometry, PCR/single-strand conformation polymorphism, Fluorescent In Situ Hybridization X/Y, and Short Tandem Repeats.11–13 It is a highly sensitive technique and can detect the presence of 0.1% to 0.01% donor DNA in the recipient.17 We found that the detection of SNP-positive DNA in SNP-negative DNA could be achieved with a donor DNA concentration over 0.04 ng/μL. We have confirmed that serum and whole blood show similar levels of DC, and so samples from different sources, depending on the amount of input DNA, can be analyzed. However, this method provides no data on cell lineage or the presence of regulatory T cells. With this technology, we found DC in most of the LT cases studied (93%).

Although the presence of microchimerism years after LT has been proposed to be a requisite for the maintenance of clonal exhaustion-deletion, which is achieved early after transplantation,18 there are conflicting views on the impact of DC.35, 36 Previous studies performed in peripheral blood using qualitative or semiquantitative methods for DC have yielded divergent results. In one study, the presence of low-level DC was not a sign of graft acceptance,37 whereas in others, such DC was associated with graft tolerance38 or immunosuppressive therapy.10, 37

Our data are compatible with the view that organ engraftment is a form of variable tolerance, usually incomplete, that requires different grades of maintenance immunosuppression. This is consistent with recent studies that have confirmed the importance of DC, induced by donor hematopoietic stem cell transplantation, in the tolerance of solid graft transplantation18, 39–42 and in cases in which sustained, heavy immunosuppressive treatment could be avoided. Occasionally, LT patients are able to avoid graft rejection while on no immunosuppressive treatment.40, 42

Surprisingly, in our multivariate analysis, the time after transplantation was the only variable associated with high DC. Of particular interest was the presence of exceptionally high levels of DC (over 50%) in 3 patients receiving very low immunosuppressive treatment and an unexpected absence of rejection. They had no evidence of graft-versus-host disease. Tolerance with similar mixed chimerism levels has been described in combined kidney and hematopoietic cell transplantation patients.41 Data from these 3 patients indicate that tolerance in selected patients could be a therapeutic objective and immunosuppression may be tailored according to DC,23 although our study found no correlation between immunosuppressant and DC levels.

Higher DC levels were found in child recipients, most of them 1 to 3 years old, than in adults. Higher DC levels could be associated with donor-specific tolerance to foreign antigens40 and may explain the lower rejection risk in this population in both LDLT and DDLT.

DC levels in adults were higher after DDLT than LDLT. This could reflect the balance between donor and recipient volumes and the higher percentage of foreign leukocytes infused with grafts from deceased males, although this needs to be confirmed.

An index of the evolution of the graft and a probable marker of graft lesions are the higher levels of DC detected in Rh-positive blood donor cases and in O blood group recipients. This could account for the longer rejection-free survival in the Rh-negative and ABO:A/A blood group recipients (donor:recipient) versus the Rh-positive and ABO:O/O blood group recipients. In all cases, transplantation was performed with a compatible ABO graft.

Persistently high DC was also found to be related to the injury of the graft (rejection, recurrent infection, drug toxicity, or other). Recipients with HCV had higher DC levels than recipients with alcoholic cirrhosis. Particularly in patients with HCV, higher levels of DC coincided with recurrent infection or acute rejection episodes, despite equal levels of immunosuppression, and with only a small number of patients who had started antiviral therapy. The DC levels in these cases could be due to the presence of increased amounts of circulating DNA from the liver31–33 and may reflect graft injury with increased permeability between the graft and the host.

Thus, qPCR is a useful technique for DC follow-up after LT. DC levels in LT patients are more difficult to interpret than in hematopoietic stem cell transplantation patients. This is because peripheral DC in LT reflects not only leukocyte DNA but also DNA of nonhematopoietic origin. The evolution of DC levels should be analyzed in accordance with the clinical outcome of the patient. Although we have studied only a relatively small number of patients, our results suggest that stable high DC levels, in the absence of other major clinical events, could be a marker of transplantation tolerance and may help to tailor immunosuppressive treatment.

Ancillary