1. Top of page
  2. Abstract

In 20% to 30% of infected individuals, hepatitis C virus (HCV) can cause cirrhosis and hepatocellular carcinoma, for which liver transplantation is the best treatment available. HCV re-infection is universal, and hepatitis disease recurrence occurs in most cases with a 30% probability of progression to graft cirrhosis at 5 years post-transplant. The immunological response to HCV involves natural killer (NK) cells and killer cell immunoglobulin-like receptors (KIRs), which specifically recognize human leukocyte antigen (HLA) class I antigens present on target cells. The effector functions of NK cells are influenced by inhibitory KIR interaction with self-HLA class I ligands, with HLA-C being the most predominant. This study examines the roles of KIR genotypes and their HLA ligands in both HCV disease recurrence and its progression. A total of 151 patients were included in the cohort, and their clinical details were recorded. Liver biopsies were used to define the absence/presence of recurrent hepatitis, the degree of fibrosis, and the progression to cirrhosis over a 10-year period. Mismatching of KIR–HLA-C ligands between donor-recipient pairs was associated with the recurrence of hepatitis (P = 0.008). The presence of KIR2DL3 in the recipient correlated with progression to liver fibrosis (P = 0.04). The mismatching of HLA-KIR ligands favored the progression of the recurrent hepatitis to fibrosis only in the presence of KIR2DL3 (P = 0.04). These preliminary results indicate that the KIR genotype and KIR–HLA-C ligand compatibility play roles in the recurrence and progression of hepatitis C disease in liver transplant recipients. Liver Transpl 15:390–399, 2009. © 2009 AASLD.

The most recent World Health Organization estimate of the prevalence of hepatitis C virus (HCV) infection is 2%,1 and 20% to 30% of HCV-positive subjects develop complications such as cirrhosis and hepatocellular carcinoma (HCC), for which liver transplantation is the best treatment available.2 Following liver transplantation, HCV re-infection is universal, and hepatitis occurs in most patients with a 30% cumulative probability of progression to graft cirrhosis at 5 years post-transplant.2–4

Several studies have demonstrated that the immunological response to HCV infection involves both the innate and adaptive immune systems and that natural killer (NK) cells play a key role in the innate antiviral immune response.5, 6 NK cells express killer cell immunoglobulin-like receptors (KIR) on their surface, which specifically recognize the human leukocyte antigen (HLA) class I antigens of the major histocompatibility complex present on the target cells. In vivo, the effector functions of NK cells are influenced by inhibitory receptors for self-HLA class I ligands HLA-A, HLA-B, and HLA-C, with HLA-C and Bw4 being the predominant ligands for most KIR receptors. The dimorphism in amino acids present in the alpha 1 helix of the HLA-C molecules is characterized by Ser77/Asn80 and Asn77/Lys80, which define distinct allotypes of HLA-C (C1 and C2, respectively). The alleles corresponding to the C1 allotype include HLA-Cw*01, HLA-Cw *03, HLA-Cw*07, and HLA-Cw*08, which interact with the KIR receptors 2DL2/3 and 2DS2. The alleles corresponding to the C2 allotype include HLA-Cw*02, HLA-Cw*04, HLA-Cw*05, and HLA-Cw*06, which interact with the 2DL1 and 2DS1 receptors.7, 8 KIRs 3DL1 and 3DS1 interact with HLA-Bw4, and KIR 3DL2 recognizes HLA-A3 and HLA-A11.9

A previous study concluded that these different KIR–HLA-C genotypes confer either susceptibility or resistance to infectious diseases10 and that homozygosity for HLA-C1 and KIR2DL3 may be advantageous for viral infections.11 Another study identified a protective effect of HLA-Bw4I80 and the KIR3DS1 gene against the development of HCC,12 which in association with cirrhosis is a common indication for liver transplantation. Weaker ligand binding with KIR2DL3 has been demonstrated, indicating weaker inhibition through this receptor,13 and a recently published study found that NK cell responsiveness corresponded with the KIR-HLA genotype.14 Previous studies have demonstrated an association with HLA-C antigens and clinical outcome in liver transplant recipients.15, 16 Moya-Quiles et al.15 found that a decrease in the number of rejection episodes correlated with a decrease in the number of HLA-C allele mismatches, and Bishara et al.16 concluded that recipients transplanted with organs from donors with no C epitope disparity had fewer rejection episodes. The number of rejection episodes does not seem to have an impact on HCV disease progression as indicated by earlier research.17 Less is known, however, about the role of NK cells and KIR-HLA interaction in liver transplant outcome with respect to HCV recurrence. This lack of data and the considerable pressure of organ shortages have prompted us to examine the possible role of KIR–HLA class I combinations in the recurrence of HCV and disease progression following liver transplantation. More specifically, we investigated the effects of KIR genes and HLA-C allotypes on the severity of the recurrent disease in a cohort of liver transplant recipients, taking advantage of the large sample size and the long follow-up time.


  1. Top of page
  2. Abstract

Study Cohort

A total of 151 consecutive donor/recipient pairs were included in the cohort for this retrospective study. The patients were a subset of a previously described cohort18 derived from 2 centers belonging to the North Italy Transplant Program19 who were transplanted between 1991 and 2001; 99 patients were transplanted in Milan, and 52 were transplanted in Padua. All patients from the previous cohort with either DNA or biopsy material available were included in this current study, and all patients involved gave informed consent. Ethical approval for the study was granted on the condition that the collected samples were screened for the purpose of assessing patients for transplantation.

All potential confounding factors such as other liver diseases and transplant-related disorders were excluded so that a well-defined homogeneous cohort remained, as described in our previous publication.1 All consecutive HCV-positive liver transplant recipients without B or delta coinfections and surviving at least 3 months were considered. The main clinical features of the patients from the 2 centers were recorded and are shown in Table 1. All patients were of Caucasoid Italian descent. Indications for transplantation were either end-stage cirrhosis (n = 109) or cirrhosis with HCC (n = 42). The number of patients with HCC post-transplant was also recorded (n = 4). In order to study the influence of acute rejection on HCV disease progression, the number of individuals who experienced acute rejection episodes was noted (n = 45). Significant alcohol consumption (>40 g/day for women, >80 g/day for men) was seen in 21 (14%) patients prior to transplantation, but only 3 patients (2%) resumed these levels of consumption post-transplant.

Table 1. Clinical Characteristics of Patients from the North Italy Transplant Program (Milan/Padua)
VariableAll Patients (n = 151)Milan (n = 99)Padua (n = 52)Milan Versus Padua P Value
  1. NOTE: Significant P values are bolded.

  2. Abbreviation: HCV, hepatitis C virus.

Gender (male/female)113/3873/2640/120.66
Mean follow-up period, months68 ± 3564 ± 3776 ± 300.06
Mean age at transplant, years51 ± 7.751.6 ± 7.350.4 ± 8.50.37
HCV genotype 190 (70%)61 (64%)29 (88%)0.014
Antilymphocyte antibody100 (66%)99 (100%)1 (2%)<0.0001
Azathioprine/mycophenolate112 (74%)99 (100%)13 (25%)<0.0001
Cyclosporine131 (87%)97 (98%)34 (65%)<0.0001
Tacrolimus20 (14%)2 (2%)18 (35%)<0.0001
Duration of steroids, months (range)4 (0–52)4.3 (0–52)4.5 (1–28)0.89
Steroid pulses45 (30%)27 (27%)18 (35%)0.0010
Antiviral therapy (pre-transplant)72 (48%)51 (52%)21 (40%)0.0004
Recurrent hepatitis127 (84%)79 (80%)48 (92%)0.46
Severe fibrosis (stages 4–6)30 (20%)17 (17%)13 (25%)0.25

Histological Evaluation

A minimum 1.5-fold increase in aspartate and/or alanine aminotransferase levels was the usual indicator for performing a liver biopsy examination. In addition, protocol biopsies were taken at 1, 3, 5, 7, and 10 years post-transplant when clinical indications of HCV recurrence were observed in the patient. The series of biopsies for each patient was reviewed over the study time period by 2 liver pathologists and scored according to the criteria of Ishak et al.20 in order to define the absence/presence of recurrent hepatitis, the degree of fibrosis, and the progression to cirrhosis. Recurrent hepatitis C disease was defined histologically from biopsy examinations performed because of clinical indications or in protocol biopsy specimens. Only patients showing a necroinflammatory grade above 5, with or without fibrosis, throughout the follow-up period were considered to have recurrent hepatitis. Patients progressing beyond stage 3 were considered to have severe recurrence.


Immunosuppressive regimens were different in the 2 centers (Table 1). In Milan, quadruple drug induction (rabbit antithymocyte globulins for 5 days, azathioprine for 1 month, corticosteroids for 1–6 months, and cyclosporine A) was used, followed by maintenance with long-term cyclosporine monotherapy, which was achieved within 6 months post-transplant in 85% of patients. Steroid withdrawal was anticipated between 1 and 3 months in more recent years. In Padua, double drug induction (cyclosporine A or tacrolimus and corticosteroids) was the standard protocol, whereas azathioprine or mycophenolate acid was added in cases of renal impairment with the aim of reducing the calcineurin-inhibitor dosage. Steroids were withdrawn within 6 months post-transplant in 93% of the patients.

Antiviral Therapy

A total of 72 patients with recurrent hepatitis underwent a treatment regimen of interferon-alpha 2b plus ribavirin for a minimum of 6 months. A histologic finding of a grading score higher than 6 and a stage score higher than 2 with or without raised aminotransferases were the usual criteria for considering combined antiviral therapy.

HCV Status

Patients were tested for anti-HCV antibodies with a second-generation enzyme-linked immunosorbent assay (ELISA II, Ortho Diagnostic System, Raritan, NJ), and HCV RNA was detected by nested reverse-transcription polymerase chain reaction (Amplicor HCV, Roche, Nutley, NJ) to determine the HCV genotype. HCV specificity was tested to determine with which viral strain individuals were infected and if differences in clinical outcome were influenced by differences in the viral strain.

HLA and KIR Genotyping

Genomic DNA was isolated from peripheral blood with Qiagen columns (QIAmp 96 spin blood kit, Qiagen GmbH, Hilden, Germany). Donor and recipient HLA-C typing was performed with sequence-specific oligonucleotide typing kits (Lipa, Innogenetics NV, Murex Biotech, Ltd., Dartford, UK). KIR allele genotyping was performed with Luminex-based reverse SSO typing (One Lambda, Los Angeles, CA). KIR haplotypes were assumed as family data were not available and were determined to be haplotype AA or B/x according to the presence of specific KIR genotypes as previously described.21

HLA-KIR ligand donor/recipient matching was divided into 3 groups based on the definitions of Tran et al.22 as follows: group 1 was a C1/2-Bw4 matched group in which the recipient and donor shared the same C1 (HLA-Casn80), C2 (HLA-Clys80), and Bw4 epitopes; group 2 was a C1/2-Bw4 mismatched compatible group in which the recipient and donor were mismatched for C1, C2, and Bw4 epitopes but the recipient's KIR epitope was not missing in the donor's HLA phenotype (eg, recipient with C2/C2, Bw4/Bw6; donor with C1/C2, Bw4/Bw4); and group 3 was a KIR C1/2-Bw4 mismatched incompatible group in which the donor's HLA phenotype did not include a C1, C2, or Bw4 epitope that was present in the recipient (eg, recipient with C1/C1, Bw6/Bw6; donor with C2/C2, Bw4/Bw4). KIR study groups 2 and 3, as described here, were also classified into a composite mismatched group (Gx) for additional analysis. The recipients were also classified into 2 groups with respect to the presence/absence of the KIR2DL3 receptor in the recipient and also in terms of matching for the HLA-C1/C2 group. Individuals in whom KIR2DL3 was absent or in whom KIR2DL3 was present but who were matched for the HLA-C group were classified as KIR2DL3 low reactive; that is, they had in theory greater NK cell inhibition and thus less NK cell activation. Individuals in whom KIR2DL3 was absent and who were mismatched for the C epitope were classified as KIR2DL3 reactive; that is, they had in theory less NK cell inhibition leading to greater NK cell activation.6, 14

Data Analysis

Allele frequencies were determined for the study populations for HLA-C and KIR and compared to the frequency of HLA-C and KIR genes in the North Italian population (, 24 Genotype and allele frequencies were contrasted between the patients with recurrent hepatitis C and those without it with the chi-square test with the level of significance set at P < 0.05 and 95% confidence intervals. For small numbers, Fisher's exact test was used. All tests were 2-sided. The measure of association used to determine the extent of risk associated with a particular genotype or phenotype was the odds ratio. P values were corrected for the number of comparisons according to Bonferroni.25 The patients with recurrent hepatitis were stratified into types of recurrence throughout the follow-up period (ie, the cumulative incidence of fibrosis stages at years 1, 3, 5, and 10), and univariate analysis was performed to test for associations between HLA class I KIR ligands, KIR genes, and disease progression. The following clinical variables were analyzed for a potential association with HCV progression and liver fibrosis outcome: donor age, recipient age at transplantation, donor gender, recipient gender, cold ischemia time, transplant center, incidence of HCC pre-transplant and post-transplant, presence of rejection episodes, HCV genotype, antiviral therapy, immunosuppressive therapy, and steroid treatment. Clinical factor and allele frequency analyses were performed with the Statistical Analysis System package, version 9.1.

All variables with a P value < 0.2 at univariate analysis were also analyzed with the multivariate logistic regression model.


  1. Top of page
  2. Abstract

Clinical Characteristics of the Patient Groups

Significant differences were observed between Milan and Padua for antiviral and immunosuppressive therapies, as shown in Table 1; however, there were no differences observed between the 2 centers in terms of clinical outcome, that is, recurrent hepatitis and severe fibrosis (P = 0.46 and P = 0.25, respectively).

HLA-C Allele and KIR Gene Distribution in Liver Transplant Recipients

The distribution of HLA-C alleles and KIR genes in our series of patients was not significantly different from that of a bone marrow donor population from Northern Italy (PHLA-C = 0.11 and PKIR genes = 0.83). The frequencies of the HLA and KIR alleles are shown in Table 2.

Table 2. HLA-Cw and KIR Phenotype Frequencies in HCV Liver Recipients and Population Controls
HLA-Cw Allele Frequencies
Cw AlleleHCV Liver RecipientsNorth Pavia, Italy*
Cw 010.02640.056
Cw 020.06250.068
Cw 030.05250.062
Cw 040.17200.204
Cw 050.05000.037
Cw 060.13550.112
Cw 070.25800.266
Cw 080.05250.031
Cw 120.06600.068
Cw 140.02650.031
Cw 150.03650.012
Cw 160.04300.044
Cw 170.01650.012
P = 0.107Pz = 151Pz = 81
KIR Phenotype Frequency (%)
KIR GeneHCV Liver RecipientsItaly*
P = 0.83Pz = 151Pz = 217

Histological Evaluation of Patient Liver Biopsy Samples

Of the 136 patients that showed clinical indications of HCV recurrence, 11 were not available for biopsy. Histologically proven recurrent hepatitis C [a necroinflammatory grade greater than 5 (G > 5)] was observed in all 125 biopsied individuals at a median of 9 months (range, 1–36) post-transplantation.

HLA-C Allele and KIR Gene Distribution in Patients with Recurrent Hepatitis

The analysis of HLA-C allele frequency and the presence of recurrent hepatitis in this study group showed that the allele HLA-Cw*02 was statistically significantly increased in HCV-recurrent patients in comparison with nonrecurrent patients (34% versus 10%, P = 0.03). There were no significant differences in frequency observed for any other HLA-C alleles.

Donor-recipient matching for HLA-KIR ligands was examined by a comparison of the group matched for HLA-C1 or HLA-C2 alleles and HLA-Bw (G1, n = 34) with the mismatched compatible group (G2, n = 49) and also with the mismatched incompatible group (G3, n = 68) with the development of recurrent disease (G > 5; G1, 75.76% versus G2, 91.23%, P = 0.06, and G1, 75.76% versus G3, 93.65%, P = 0.02). These statistically significant differences in clinical outcome indicate that the development of recurrent hepatitis is linked to the mismatching of donor/recipient KIR ligands as well as the absence of the recipient's KIR ligand in the donor liver (P = 0.02). Therefore, we created a composite mismatched group (Gx; ie, the combined group of donor-recipient pairs with HLA-KIR ligand mismatches: group 2 + group 3) and found that the frequency of patients with recurrent hepatitis (G > 5) was also increased in the Gx group in comparison with group 1 (Gx, 93.46% versus G1, 75.76%, P = 0.008). Figure 1 shows the distribution of HLA-KIR ligand groups G1 and Gx for 2 outcomes with respect to liver inflammation (no recurrence versus inflammation, G > 5). The analysis of the HLA-ligand mismatched group demonstrated that hepatitis recurrence correlated with HLA-C1 or HLA-C2 donor/recipient mismatch (P = 0.03) but not with HLA-Bw4 mismatch (P = 0.16).

thumbnail image

Figure 1. Liver transplant recipients were classified into 2 groups: those with no recurrence of liver inflammation (grade < 5) and those with inflammation (grade > 5). They were analyzed for recipient-donor killer cell immunoglobulin-like receptor ligand matching, that is, matched (G1) versus mismatched (Gx). P = 0.008. Abbreviation: HCV, hepatitis C virus.

Download figure to PowerPoint

A mismatch for the HLA-KIR ligand together with the presence of KIR2DL3 (KIR2DL3 reactive) was significantly associated with recurrent hepatitis in comparison with the KIR2DL3 low reactive group as defined previously (57.14% versus 32.14%, P = 0.02).

No association between the presence of KIR genes 2DL1 (P = 0.3), 2DL2 (P = 0.8), 3DL1 (P = 0.58), 2DS1 (P = 0.53), 2DS2 (P = 0.48), 2DS4 (P = 0.58), and 3DS1 (P = 0.9) and hepatitis recurrence was observed.

The presence of HCC before transplantation (P = 0.68) was not statistically significant. The number of patients with HCC post-transplantation (n = 4) was too small to test associations adequately. In patients with HCC pre-transplant, a protective effect of HLA-Bw4I80-receptor 3DS1 against HCV recurrence (P = 0.3) was not detected. The treatment of recipients with steroids had no significant effect on HCV disease progression (P = 0.37).

The following clinical variables were statistically significant following the univariate analysis and were included in the multivariate analysis: age of recipient at transplantation, female sex of the recipient, HCV genotype, KIR ligand mismatch, antiviral therapy, and immunosuppression (Table 3). The donor age (P = 0.7, median age = 37 ± 14.0), donor gender (P = 0.3), and cold ischemic time (P = 0.3, median time = 7 ± 4.0 hours) were not significant following univariate analysis. The multivariate analysis, however, showed only a significant difference between the matched (group 1) and the KIR ligand mismatched recipients (Gx) in relation to the recurrence of HCV in liver-transplanted patients (P = 0.003).

Table 3. Multivariate Analysis of Clinical and Immunological Variables in 151 HCV-Positive Recipients Without Recurrence (G < 5) or With Recurrence of Liver Inflammation (G > 5)
VariableNo Recurrence (n = 15)Recurrence (n = 136)P ValueCox Analysis
  1. NOTE: The significant P value is bolded.

  2. Abbreviations: CI, confidence interval; HCV, hepatitis C virus; HR, hazard ratio; KIR, killer cell immunoglobulin-like receptor.

Transplant center (MI/PD)4/1148/880.35HR, 3.16; 95% CI, 0.27–36.4
Age at transplantation53.3 ± 7.350.94 ± 7.130.72HR, 0.98; 95% CI, 0.89–1.09
Recipient sex (male/female)10/5103/330.28HR, 2.19; 95% CI, 0.52–9.27
HCV genotype 171%71%0.24HR, 0.39; 95% CI, 0.08–1.89
KIR ligand mismatch47%81%0.003HR, 8.02; 95% CI, 2.06–31.29
Antiviral therapy33.3%52.8%0.34HR, 1.89; 95% CI, 0.51–7.08
Cyclosporine/tacrolimus80%86%0.6HR, 0.47; 95% CI, 0.02–11.32
Steroid pulses40%28.8%0.11HR, 0.28; 95% CI, 0.06–1.38

HLA-C Allele and KIR Gene Distribution and Severity of Fibrosis

Recurrent hepatitis progressing beyond stage 3 was observed in 30 of 151 patients (20%). The 1-, 3-, 5-, and 10-year cumulative incidences of stage 4–6 fibrosis were 3%, 15%, 22%, and 36%, respectively. HLA-C allele frequencies were stratified into those that progressed from recurrent hepatitis to mild fibrosis (stages 0–3) and severe fibrosis (stages 4–6). A significant association between HLA-Cw*08 and the progression to severe fibrosis was observed (23% in recipients with severe fibrosis versus 7.4% in recipients with mild fibrosis, P = 0.02).

The frequency of patients with severe fibrosis was higher in the group of females versus the group of males (32% versus 16%, P = 0.03).

HCV type 1 was significantly associated with recurrent hepatitis progressing beyond stage 3 (88% of patients with stages 4–6 with HCV genotype 1 versus 66% of patients with stages 4–6 with the other non-1 HCV genotypes, P = 0.02).

No association between the HLA class I alleles A3, A11, C1, C2, and Bw (previously described as being the most influential9, 22) and the progression to fibrosis was observed. The analysis of KIR genotype distribution evidenced an association between the presence of KIR2DL3 in the recipient and progression to liver fibrosis (P = 0.04).

The mismatching of HLA-KIR ligands favored the progression of the disease to fibrosis (ie, stages 4–6) only in the presence of KIR2DL3 (P = 0.04). Because of the small number of patients who were HLA-C1–homozygous, the analysis of HLA-C1 homozygosity with the presence of KIR2DL3 and the development of fibrosis was not possible.

The study of KIR haplotypes indicated that the effect of KIR2DL3 was not associated with the presence of the AA KIR haplotype (P = 0.3).

Recipient gender, HCV genotype (1 versus others), KIR ligand mismatch, presence of KIR2DL3, donor age, and immunosuppression were included in the multivariate analysis (Table 4).

Table 4. Multivariate Analysis of Clinical and Immunological Variables for HCV-Positive Patients with Mild (Stages 0–3) or Severe (Stages 4–6) Fibrosis
VariableSevere Fibrosis (n = 30)Mild Fibrosis (n = 121)P ValueCox Analysis
  1. NOTE: Significant P values are bolded.

  2. Abbreviations: CI, confidence interval; HCV, hepatitis C virus; HR, hazard ratio; KIR, killer cell immunoglobulin-like receptor.

Recipient sex (male/female)18/1295/260.07HR, 0.39; 95% CI, 0.14–1.09
HCV genotype 188%70%0.009HR, 6.39; 95% CI, 1.6–25.74
KIR ligand mismatch83%77%0.34HR, 1.85; 95% CI, 0.53–6.27
Presence of gene KIR2DL3 in recipient100%76%0.04HR, 5.10; 95% CI, 1.06–24.53
Cyclosporine/tacrolimus90%86%0.96HR, 1.03; 95% CI, 0.23–4.51
Donor age ≥ 5011/1927/940.03HR, 3.10; 95% CI, 1.12–8.53

The multivariate analysis demonstrated that HCV genotype 1, donor age greater than 50 years, and the presence of the KIR2DL3 gene in the recipient were associated with more aggressive recurrent liver disease, that is, progression to severe fibrosis (P = 0.009, P = 0.03, and P = 0.04, respectively).

There was no association between viral persistence and clinical outcome nor any evidence of a protective effect with HLA-Cw*04 (P = 0.89), HLA-Cw*06 (P = 0.60), or HLA-Cw*07 (P = 0.39), as previously reported.5, 15, 26, 27

Analysis of Influence of Acute Rejection on HCV Disease Progression

The influence of acute rejection on HCV recurrence and the development of liver fibrosis was tested but found to have no significant effect (P = 0.7 and P = 0.36, respectively). The presence of KIR2DL3, presence of HLA-Cw*0727 in either the donor or recipient, and KIR haplotype (AA or B/x) in association with the occurrence of acute rejection were also analyzed and again were not statistically significant (P = 0.6, P = 0.3/P = 0.2, and P = 0.4, respectively).


  1. Top of page
  2. Abstract

Most liver allograft transplant losses are caused by recurrence of the primary disease rather than by rejection,28, 29 and in this respect, many contributing factors for progressive fibrosis have already been recognized.30 Nevertheless, the influence of the innate immunological system and the elements of this system on genetic susceptibility to both HCV recurrence and disease progression to cirrhosis in liver transplant recipients has never been investigated.

Recurrent HCV infection may begin in a manner similar to acute HCV infection,31 and the clearance and control of the virus may therefore depend on an effective and long-lasting NK, CD4+T, and CD8+T response. Although HCV virus per se is not cytopathic and its levels do not correlate with the degree of liver fibrosis, it has been demonstrated that it does induce the production of cytokines [interferon gamma (IFN-γ), interleukin 12 (IL-12), IL-18, and IL-15] from hepatocytes that activate the NK cells,32 which may result in liver damage. The active NK cells in turn secrete IFN-γ, which induces monokine-induced IFN-γ and IFN-γ–inducible protein 10, culminating in the activation and development of CD4+ and CD8+ T cells. In contrast, a virus-specific CD4+T cell response is often undetectable in patients with chronic liver disease.33

Several studies have been published in which the effects of KIR ligand incompatibility and NK cell responses in hematopoietic transplants were investigated,34, 35 but less is known about this in relation to solid organ transplantation.22 To our knowledge, this is the first study to evaluate the role of NK receptors and their ligands in HCV-infected transplant patients, and it has 4 major strengths: (1) it is based on a large unselected cohort of consecutive Caucasoid patients with long follow-up times, (2) HLA typing was performed by molecular techniques, (3) the hepatitis recurrence is defined by objective histologic criteria, and (4) protocol liver biopsies were available in order to assess disease progression.

Our data indicate that mismatching of KIR-HLA ligands between donor-recipient pairs in KIR2DL3-positive recipients is associated with the histologic recurrence of severe hepatitis and expand a previous study that demonstrated the importance of HLA-Cs on liver transplantation outcome.16 Our study, however, does not confirm a protective effect of HLA-Bw4I80-receptor 3DS1 against HCV recurrence. The highly polymorphic nature of KIR3DS1 and the peptide-specific manner of presentation by the molecule Bw4 might explain the different findings.36 Our study did not find an association between the occurrence of acute rejection and disease progression in accordance with a previous study.17 Our analysis of KIR2DL3, KIR haplotype, and HLA-Cw*07 as potential contributing factors for the occurrence of acute rejection showed no significant effect. Moya-Quiles et al.27 previously reported that HLA-Cw*07 has a protective effect against the occurrence of acute rejection. It is possible that the discrepant findings are due to the fact that Moya-Quiles et al.'s patient group consisted of liver transplantees as a whole, whereas our transplant recipient group consisted specifically of individuals with HCV pre-transplantation; also, the number of individuals who experienced acute rejection in our study group was small (n = 45). Our lack of a significant finding for the role of the use of steroids was again possibly due to the small number of individuals and also to the fact that in 93% of the recipients, steroids were withdrawn within 6 months.

It is well known that the major histocompatibility complex class I molecules bind and present endogenous antigens to CD8+ cytotoxic T cell molecules and are directly involved in the activation and inhibition of NK cells. The level of HLA-C expression regulates NK cell activity37 and thus represents a potential target for virally induced modulation of NK responses.

HLA-C matching, as shown in our study, recalls the concept of the dualistic effect of HLA matching, which on the one hand reduces the risk of rejection but on the other hand enhances immune-mediated injury.38 We hypothesize that HLA-C matching prevents the activation of the T lymphocytes, and this explains its importance in the development of HCV recurrence.

The importance of KIR2DL3 and NK cells in the progression to graft cirrhosis is evidenced by our study. In relation to this, Khakoo et al.6 demonstrated that the presence of the KIR2DL3 receptor in association with HLA C1C1 favors HCV clearance only in patients with a low viral load.

We postulate that in immunocompetent individuals, NK cells with receptor KIR2DL3 would have a protective effect during the first stage of the infection but not later when the viral load increases and the innate immune response is possibly overwhelmed.39–41

It has been shown in animal models that the depletion of NK cells before a hepatotropic viral infection leads to inhibition of a virus-specific T cell response as well as inhibition of liver injury.42 In 1 human study, it was demonstrated that during a persistent HCV infection, the virus faces diminished immune pressure over time, either from mutation to an immune-resistant sequence or from immunological exhaustion. This diminished immune response may be reflected by diminished inflammatory activity.43 When the HCV virus infects a patient, a weaker inhibitory receptor ligand such as KIR2DL3–HLA-C1 will be protective because it can more easily prevail by activating signals than a stronger inhibitory interaction such as KIR2DL2–HLAC1 or KIR2DL1–HLA-C2.13 This activation is important for virus clearance only during the infective phase; if the activation persists, it damages the liver because of NKR+CD8+T cell accumulation in the liver parenchyma.44, 45

With this hypothesis in mind, the absence of the KIR2DL3 receptor and matching for the HLA-KIR ligands in HCV-positive liver transplant recipients might have a protective effect for the evolution of HCV recurrence. Conversely, the NK cell receptor KIR2DL3 may modify cell activity in the presence of the mismatch, facilitating the progression to fibrosis. A recent study provides support for this hypothesis as it demonstrated a differential NK cell responsiveness to viral infections depending on the KIR-HLA genotype and showed that KIR2DL3-positive NK cells secreted more IFN-γ at an earlier time point following infection.14 Our preliminary data indicate that KIR2DL3-positive recipients would be better assigned a matched donor for the HLA-KIRs in order to reduce the risk of developing severe fibrosis after liver transplantation. Tran et al.22 showed a similar trend, in that renal recipient survival was better in matched individuals than in mismatched individuals and in incompatible individuals, although these finding were not statistically significant. Therefore, it would be potentially beneficial to the HCV-positive patients if the KIR genotype were determined pre-transplantation. We are currently undertaking a prospective study including functional cellular assays to test the previously discussed findings.

As far as the effect of posttransplant antiviral therapy is concerned, there are discrepant opinions at present. Several studies have demonstrated an increase in rejection episodes in recipients treated with IFN,46 whereas others deny the association.47 Our results indicate, however, that the presence of acute rejection is not associated with HCV progression. On the basis of our results, its appears that posttransplant antiviral therapy is effective only in the early stages of re-infection when, as previously reported, an increase in NK cells and higher NK activity can be observed.48 In the event of HCV recurrence, the activation of the defensive innate system produces an increase in inflammation with a possible evolution to severe fibrosis. Our data, however, confirm several studies that indicate that genotype 1b is harder to eradicate than genotypes 2 and 3 and that subjects with severe fibrosis are less responsive to the therapy.49–51 This is likely to be a result of a greater enhancement of the T cell response due to IFN treatment in genotype 2 and 3–infected patients versus genotype 1b–infected patients.52 In agreement with a previous study,4 our data show an association between donor age greater than 50 years and the development of severe fibrosis in the multivariate analysis. Another study reported an association between cold ischemic time greater than 10 hours and the development of HCV recurrence,53 but this was not confirmed by our group; however, it must be noted that for our liver transplants, the cold ischemic time was rarely greater than 10 hours.54

In terms of HLA matching alone, in a previous study we demonstrated that HLA-DRB1 mismatching is associated with the recurrence of histologic hepatitis and its progression beyond stage 3.18 Based on our previous results and current results, in our opinion, HCV-positive recipients should ideally be matched for HLA-C and DRB1 with their donors.

The presence of HLA-Cw*02 is significantly higher in patients without recurrence and shows a negative trend in patients with severe fibrosis. It is possible that HLA-Cw*02 presents HCV peptides that lead an effective immune response or that HLA-Cw*02 interacts with the virus or another host protein that favors viral clearance.

On the contrary, the frequency of HLA-Cw*08 is significantly higher in recipients who develop severe fibrosis. It can be postulated that subtle structural changes in HLA-Cw*08 molecules differentiate them from other HLA-C molecules, leading to greater specificity for peptide engagement and causing increased alloreactivity and risk for fibrosis.

In summary, the results presented in this study show that a disparity in HLA-C allotypes between the recipient and donor may increase the risk of recurrence of hepatic inflammation and support the importance of the KIR2DL3 receptor in the development of the disease. These are potentially important factors for the selection of the most appropriate HCV-positive recipient of liver transplantation and thus the optimum use of the limited organs available.

An extension of this study and confirmation of our findings will help to elucidate the role of HLA–KIR genotype combinations in the recurrence and progression of hepatitis C in liver transplant patients.


  1. Top of page
  2. Abstract