Interleukin-28B polymorphism in hepatitis C and liver transplantation

Authors

  • Andres Duarte-Rojo,

    1. Division of Gastroenterology and Hepatology, University of Arkansas for Medical Sciences, Little Rock, AR
    2. Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN
    Search for more papers by this author
  • Matthew G. Deneke,

    1. Division of Gastroenterology and Hepatology, University of Arkansas for Medical Sciences, Little Rock, AR
    Search for more papers by this author
  • Michael R. Charlton

    Corresponding author
    1. Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN
    • Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First Street Southwest, Rochester, MN 55905
    Search for more papers by this author
    • Telephone: 507-266-7054; FAX: 507-266-1856


  • This work has been supported by US Public Health Service grants (National Institute of Diabetes and Digestive and Kidney Diseases grant RO1 DK069757-01 and General Clinical Research Center grant RR00585).

Abstract

The discovery of interleukin-28B (IL-28B) single-nucleotide polymorphisms has opened an important new area of research in liver transplantation (LT) for hepatitis C virus (HCV). Both recipient- and donor-derived IL-28B genotypes affect the post-LT treatment response, with sustained virological response (SVR) rates oscillating from >50% in homozygotes for the favorable allele (up to 90% when this is present in both the recipient and the donor) to <15% in homozygotes for the unfavorable allele and from 30% to 50% in heterozygotes. Other key posttransplant outcomes affected by the IL-28B genotype are the time to histological recurrence, HCV RNA and alanine aminotransferase levels, histological variables (including the rate of fibrosis progression), and hepatocellular carcinoma. Interactions between donor and recipient IL-28B genotypes are complex and may affect outcomes not directly related to HCV infections, such as acute cellular rejection (ACR) and metabolic diseases. A preferential allocation system in which livers from donors homozygous for the favorable allele are given to HCV patients might be postulated to improve SVR rates and post-LT outcomes in recipients with HCV infections (a 25% increase in SVR and an 8% decrease in mortality at 5 years). Although negative effects from this are difficult to predict, they could include an accelerated progression of fibrosis in patients with failed HCV eradication and an increase in ACR in non-HCV patients. Our knowledge of the precise role of IL-28B genotypes in the course of post-LT HCV is evolving, but existing knowledge suggests the possibility of exploring strategies that use IL-28B genotyping to reduce the impact of post-LT adverse outcomes. Liver Transpl 19:49–58, 2013. © 2012 AASLD.

The recent association of allelic variation in the interleukin-28B (IL-28B) gene with hepatitis C virus (HCV) eradication after antiviral therapy generated new insight into our understanding of the complex relationship between viral infections and the human immune system.1 The discovery suggested the importance of type III or lambda interferons (IFNs) for mounting an effective immunological response against HCV, either spontaneously or after stimulation with peginterferon and ribavirin. The initial report showing that patients homozygous for the C single-nucleotide polymorphism (SNP) at position rs12979860 of chromosome 19q (corresponding to 3 kb upstream of the IL-28B gene) were 2 times more likely to achieve a sustained virological response (SVR) than patients with either the CT or TT variant1 was rapidly followed by similar reports assessing SNPs within the same region of the IFN-λ gene as well as their association with spontaneous viral clearance after acute HCV infections.2-5

The study of IL-28B in HCV is particularly complicated for liver transplantation (LT) because recipients have 2 contributing sources of IL-28B genotypes: the recipient and the donor allograft. Thus, results from the nontransplant setting should not be extrapolated to the posttransplant setting. Several studies have assessed associations of IL-28B SNPs with the response to posttransplant antiviral treatment and with indices of HCV recurrence severity. The strong association of IL-28B with key posttransplant outcomes suggests the possibility of using IL-28B not only to target posttransplant antiviral therapy but also to optimize posttransplant outcomes because of its predictive usefulness, and we might even consider using IL-28B to guide allograft allocation in order to modify the clinical course of HCV recurrence.

IL-28B IN THE TREATMENT OF HCV INFECTIONS

In the nontransplant setting, the predictivity of the IL-28B genotype is greater than most factors associated with SVR, such as the pretreatment viral load, ethnicity, fibrosis stage, age, and metabolic factors.5, 6 Only the viral genotype and possibly intrahepatic interferon-stimulated gene (ISG) expression appear to be more predictive of SVR.7, 8 The majority of the studies evaluating the impact of IL-28B on rates of SVR in nontransplant patients have focused on patients infected with HCV genotype 1. In fact, the significance of IL-28B polymorphisms was first described in HCV genotype 1 patients.1-5 Although multiple loci for SNPs have been identified, the strongest evidence for improved rates of SVR in genotype 1 patients has been best described in association with the C allele at rs129798602, 6, 9 and the T allele at rs8099917.3, 4 Although less robust, there are data demonstrating an association between the IL-28B genotype and rates of SVR in patients with genotypes 2, 3, and 4 as well. Mangia et al.10 found an association between the rs12979860 genotype and the rate of SVR in patients with HCV genotype 2 or 3 who did not achieve a rapid virological response. Lindh et al.11 also found a higher rate of SVR in HCV genotype 2 or 3 patients with the rs12979860 CC genotype who did not achieve a rapid virological response, although this was observed only in patients treated for 24 weeks and not in their 12-week arm. In patients with HCV genotype 4, an association has been found between the rs12979860 genotype and the overall rate of SVR.12 Thus, the impact of IL-28B on SVR is not uniform among the HCV genotypes because it has the greatest impact on patients who are least likely to achieve SVR with peginterferon and ribavirin treatment, that is, patients with genotype 1 (likely 1a > 1b), who are closely followed by patients with genotype 4 and then patients with genotype 3, and it seems to have the least impact on patients with genotype 2.10, 12-14

In the LT setting, only the rs12979860 and rs8099917 SNPs have been investigated. In the former case, the C allele (versus the T allele) has been linked to a favorable antiviral response, whereas in the latter case, the T allele (versus the G allele) is predictive of SVR. The reported associations of IL-28B genotypes with SVR after antiviral therapy in posttransplant patients are summarized in Figs. 1 and 2. Fukuhara et al.15 studied 67 patients (59 with genotype 1 and 8 with genotype 2) for the rs8099917 SNP through liver biopsy samples from both recipients (n = 67) and donors (n = 41). Recipients who were homozygous for the favorable allele (TT) had a higher rate of SVR than recipients with the TG or GG genotype (54% versus 11%, P = 0.003). The same association was observed for the donor genotype and SVR (44% versus 9%, P = 0.025; Fig. 1A). Eurich et al.18 recently reported an analysis of European patients (133 with genotype 1 or 4 and 16 with genotype 2 or 3) that was also based on the genotype at the rs8099917 locus, which was characterized from blood samples (only for recipients). The magnitude of the effect of the IL-28B genotype on SVR was broadly similar to that reported in Japanese LT recipients (Fig. 1A), although in this cohort, none of the patients with the GG genotype achieved SVR. Most recently, Kawaoka et al.16 evaluated the impact of donor and recipient genotypes at the rs8099917 locus on SVR in a group of 20 patients treated for HCV genotype 1 after LT. In this group, the TT genotype in donors was found to be significantly associated with SVR in comparison with the TG and GG genotypes (73% versus 20%), whereas the TT genotype in recipients was not significantly associated with SVR in comparison with the others (64% versus 50%).16

Figure 1.

Variation in the SVR frequency with (A) the recipient rs8099917 genotype and (B) the recipient rs8099917 genotype and the donor rs8099917 genotype combined.

Figure 2.

Variation in the SVR frequency with (A) the recipient rs12979860 genotype and (B) the recipient rs12979860 genotype and the donor rs12979860 genotype combined.

A large HCV cohort from the Mayo Clinic studied the effect of rs12979860 SNP on SVR.22 This was recently updated to include 255 patients, with IL-28B genotyping available for 239 recipient and 241 donor liver biopsy samples.19 Antiviral treatment was administered to 120 patients (94 with genotype 1 or 4, 24 with genotype 2 or 3, and 2 with an unknown genotype). The rate of SVR was significantly higher in patients with the favorable genotype (CC) versus the other genotypes whether it was the recipient genotype (73% versus 49% for CT and 14% for TT, P < 0.001) or the donor genotype (53% versus 39% for CT and 13% for TT, P = 0.072; Fig. 2A). Lange et al.21 assessed the same IL-28B SNP in 91 patients with either blood or liver tissue. Forty-seven patients were treated for HCV (33 with genotype 1 or 4 and 14 with genotype 2 or 3), and in recipients with the CC genotype at rs12979860, there was a nonsignificantly higher rate of SVR (75% versus 50% for CT/TT, P = 0.118), although it became significant when the donor genotype was analyzed (88% versus 47% for CT/TT, P = 0.008; Fig. 2A). Most recently, Eurich et al.23 reported a separate analysis of posttransplant HCV patients based on the recipient rs12979860 genotype. The donor genotype was not assessed. Fifty of 139 patients (36%) treated for HCV after transplantation achieved SVR. The rate of SVR for recipients with the CC genotype was significantly higher than the rate for recipients with the CT or TT genotype (52% versus 29%). Donato et al.20 recently presented data on the rs12979860 SNP in 41 treated patients with genotype 1 or 4. The CC genotype from the donor was associated with more SVR in comparison with the CT and TT genotypes (62% versus 29%, P = 0.058), whereas a significant difference was not observed when the recipient SNPs were analyzed (70% versus 42%, P = 0.15).

Coto-Llerena et al.17 examined the impact of both rs8099917 and rs12979860 loci in 128 HCV genotype 1–infected LT recipients. The response to antiviral therapy in their study was significantly higher when recipients had the rs12979860 CC genotype versus the rs12979860 CT or TT genotype both before LT (100% versus 48%, P = 0.013) and after LT (59% versus 25%, P = 0.002). A similar picture was seen for the recipient rs8099917 genotype in terms of an early virological response before LT (90% for TT versus 46% for TG/GG) and an SVR after transplantation (47% for TT versus 23% for TG/GG). In their study, the impact of the donor genotype at both loci showed a trend toward improved rates of SVR with the favorable genotypes, but it was not statistically significant.

The aforementioned studies analyzed a possible synergistic effect on SVR when both the donor and the recipient have a favorable genotype instead of only one (a mismatch for the favorable genotype) or none. The results have been broadly, although not entirely, similar. A synergistic effect was reported by Fukuhara et al.15 for the end of treatment but not SVR, although a trend was observed for SVR. The Mayo Clinic study observed a marked synergy between donor and recipient IL-28B genotypes at the rs12979860 locus (the SVR rates were 90% when both the donor and the recipient had the CC genotype, 60% when only the recipient had the CC genotype, 47% when only the donor had the CC genotype, and 24% when neither had the CC genotype).19 Coto-Llerena et al.17 observed a similar synergy between the donor and recipient genotypes at both the rs12979860 locus (83% for CC in both versus 27% for other combinations) and the rs8099917 locus (56% for TT in both versus 26% for other combinations). Finally, synergism was also observed in the study by Kawaoka et al.16 because the presence of the TT genotype in both the donor and the recipient was associated with the highest rate of SVR (81%) and was the only independent determinant of SVR according to a multivariate analysis.

The effect of IL-28B polymorphisms on SVR for patients with different HCV genotypes has not been determined in the posttransplant population. Although some studies have included data on the HCV genotype in describing their patient populations, none have included the HCV genotype in the final data analysis of IL-28B genotypes. This is likely due to the relatively small numbers of treated patients in each of the studies. However, 1 study did perform a multivariate analysis to determine the degree of impact of various factors on SVR in posttransplant patients. In that analysis, the HCV genotype, the IL-28B genotype, and the patient's age were found to be independently associated with SVR, with the HCV genotype having the greatest impact [odds ratio = 10.3, 95% confidence interval (CI) = 1.9-55] and being followed by the IL-28B genotype (odds ratio = 2.4, 95% CI = 1.3-4.6).19

Thus, the IL-28B genotype is strongly predictive of SVR in LT recipients with chronic HCV, with a moderately favorable effect observed in patients who are heterozygous for favorable alleles. The effect is associated with both the recipient genotype and the donor genotype, and there is a synergistic effect when both the recipient and the donor bear the favorable genotype. Although there are differences in the results of the available studies, they are generally quite consistent; the differences are likely due to variations in the favorable allele frequency among the populations, in the proportions of patients with genotype 1 or 4 versus genotype 2 or 3 in each genotype group, in demographics (eg, ethnicity and body mass index), in antiviral treatment protocols (which are not described in detail), and in sampling and statistical power.

Notably, one factor that may introduce some variability into studies of transplant patients is the specimen sample used for the determination of the donor IL-28B genotype. The importance of sample selection was recently highlighted in a study by Coto-Llerena et al.,24 who reported that follow-up liver biopsy samples from LT recipients were not accurate for determining donor IL-28B rs12979860 SNPs. This was caused by CT genotype enrichment due to the detection of recipient alleles (more abundant in the T allele) in biopsy samples collected after transplantation versus those collected before or during transplantation (reperfusion). These findings were confirmed in a subsequent report using similar methods.25 Thus, only donor peripheral blood mononuclear cells or allograft biopsy samples up until the time of reperfusion should be used to study donor IL-28B SNPs in LT.

Abbreviations: ACR, acute cellular rejection; CI, confidence interval; DM, diabetes mellitus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; IFN, interferon; IL, interleukin; ISG, interferon-stimulated gene; LT, liver transplantation; SNP, single-nucleotide polymorphism; SVR, sustained virological response.

IL-28B IN THE SPONTANEOUS CLEARANCE OF HCV INFECTIONS

An association has been found between IL-28B genotypes and the rate of spontaneous clearance of HCV in nontransplant patients. In a study of more than 1000 patients with evidence of current or previous infections with HCV, the presence of the CC genotype at the rs12979860 locus was found to be associated with a markedly higher rate of spontaneous clearance of HCV in comparison with the CT and TT genotypes (53% versus 28%).2 The spontaneous clearance of an HCV infection is rare in the posttransplant setting, however, with the available literature limited to case reports. Only 2 reports of spontaneous clearance of HCV in posttransplant patients have included IL-28B genotyping, with 3 cases in all.19, 26 In all 3 cases, genotyping at the rs12979860 locus was performed. Two of the 3 cases involved recipients with the CT genotype, and 1 had the CC genotype. However, all 3 patients received organs from CC donors.

A consistent finding of studies evaluating the impact of the rs12979860 locus is that the frequency of the C allele in transplant recipients is lower than the frequency in non–HCV-infected donor livers (Table 1). This is in agreement with the reported association of favorable IL-28B genotypes with the spontaneous clearance of HCV and subsequent enrichment for the non-CC variants in patients with chronic HCV infections. Recipients with non-CC genotypes are also more likely to be prior nonresponders to IFN-based therapies before LT.17 Similar patterns of enrichment of the favorable T allele in transplant donors versus recipients at the rs8099917 locus have been reported in only 1 of 2 studies assessing this allele (Table 1). It is unclear why a difference was not observed in the Japanese population.

Table 1. Frequencies of the rs12979860 Locus C Allele in Transplant Recipients and Non-HCV Donors
StudyCountry of OriginC Allele (rs12979860)T Allele (rs8099917)
RecipientDonorRecipientDonor
Duarte-Rojo et al.19United States55%70%
Lange et al.21Germany/Switzerland57%67%
Coto-Llerena et al.17Spain50%72%63%82%
Fukuhara et al.15Japan72%73%

IL-28B BIOLOGY: IMPLICATIONS FOR VIRAL ERADICATION

The mechanism through which IL-28B SNPs affects HCV clearance has not been fully clarified. The type III IFN family consists of IL-28A (IFN-λ2), IL-28B (IFN-λ3), and IL-29 (IFN-λ1). They share many properties with type I IFNs (IFN-α), such as antiviral activity and signaling via the Janus kinase/signal transducer and activator of transcription pathway, although they use different receptors (IL-28Rα and IL-10RB) that are present primarily on epithelial cells.27, 28 Antiviral activity occurs both directly and indirectly through the modulation of both innate and adaptive immune systems, and some of the known functions of IFN-λ include the induction of ISGs, the maturation and differentiation of dendritic cells, the modulation of T helper 1 and T helper 2 immune responses, and the inhibition of regulatory T cells.29-32

Studies of viral kinetics have shown that the favorable IL-28B SNPs are associated with steeper decreases in HCV RNA (in both the first and second phases of viral clearance) after antiviral therapy, and this effect has been confirmed across genotypes.11 Although it is unclear whether the favorable SNPs induce greater levels of IFN-λ3,1, 3, 4, 15, 33, 34 they are consistently associated with decreased levels of intrahepatic ISGs.8, 33 Decreased baseline levels of ISGs portend a vigorous early antiviral response immediately after the beginning of treatment and are thus markers of viral eradication.35 This supposes that patients harboring the CC genotype are able to produce a greater increase in ISGs, which will result in more effective signaling pathways that control viral replication.36

Recent experimental evidence points to stronger hepatocyte induction of IFN-λ versus IFN-α immediately after an HCV infection, and this leads to robust ISG induction. This could subsequently disturb the IFN-α pathway and set the way for failed viral eradication. If this proposed model is in fact true, the favorable IL-28B SNPs should actually translate into a less effective IFN-λ pathway allowing less intense ISG induction and less blocking of IFN-α signaling.37 Upon failed spontaneous eradication, this will leave room for exogenous IFN-α to properly stimulate the immune system directing HCV eradication. Although this all seems somehow counterintuitive, the many pieces of the puzzle are just starting to disclose the final picture. The host-virus interaction, however, may be even more complex than what can be foreseen. There is now evidence that IL-28B SNPs can also influence viral adaptation and favor viral escape mutants linked to adverse outcomes.38 Future studies addressing IL-28B biology, particularly in the setting of posttransplant HCV, are eagerly awaited.

ASSOCIATIONS BETWEEN IL-28B AND BIOCHEMICAL AND HISTOLOGICAL VARIABLES

The time to the histological recurrence of HCV has been reported to be delayed in recipients with the rs12979860 CC genotype at 5 years, but it is not affected by the donor genotype.19 Although higher levels of HCV RNA have been described in nontransplant patients with the CC genotype at rs12979860, in the LT setting, a biphenotypic response appears to occur, and it depends on the donor genotype versus the recipient genotype: higher levels have been observed when the donor carries the favorable genotype in 2 studies,19, 21 and lower levels have been observed when the recipient is the one bearing the favorable genotype in another study18 (Table 2). The clinical significance of this finding is uncertain. Aminotransferases seem to mirror this pattern, with higher levels being described in donors with the CC genotype19, 21 and lower levels being described when this genotype occurs in the recipient.19 Lower levels of aminotransferases have also been described in recipients with the favorable (TT) genotype at rs8099917.18 Differences may exist in the timing of HCV RNA and aminotransferase measurements (taken at a fixed time point, at the time of recurrence, or as a peak of serial determinations) and perhaps explain some of the variability across reports. It is also notable that the analyses accounted for neither the body mass index nor the possibility of associated nonalcoholic fatty liver disease.

Table 2. Variations in the Frequencies of Posttransplant Outcomes With the rs12979860 and rs8099917 Genotypes
 rs12979860rs8099917 
 RecipientDonorRecipientDonor 
ParameterCCNon-CCCCNon-CCTTNon-TTTTNon-TTStudy
Fibrosis progression↓ or no effect↑ or no effectNo effectNo effectNo effectNo effectEurich et al.,18 Charlton et al.,22 and Duarte-Rojo et al.19
InflammationNo effectNo effectEurich et al.18 and Charlton et al.22
Alanine aminotransferaseEurich et al.,18 Charlton et al.,22 and Duarte-Rojo et al.19
HCV RNAEurich et al.,18 Charlton et al.,22 and Duarte-Rojo et al.19
Type 2 diabetes↑ (in TT)No effectNo effectAkuta et al.38
SteatosisNo effectNo effectVeldt et al.39
HCCLange et al.21 and Sadler and Williams36

The reported impact of IL-28B genotypes on posttransplant histology has been inconsistent (Table 2). Among studies evaluating the rs12979860 locus in posttransplant HCV patients, one report failed to show any difference in the progression to cirrhosis between favorable and unfavorable genotypes,21 another showed less fibrosis when the recipient had the favorable genotype but no differences in inflammation,19 and a third study showed no difference in the risk of developing advanced fibrosis between genotypes but did find that fibrosis developed more slowly in patients with the CC genotype.23 Another study that included HCV (40%) and non-HCV patients showed no differences based on the IL-28B genotype in the development of fibrosis 1 year after LT.40 The only study assessing the rs8099917 locus included only HCV patients and showed reduced inflammation when the recipient had the favorable TT genotype, but it found no association with fibrosis.18 One of the 2 studies that included routine protocol biopsies after LT in addition to biopsies performed for clinical reasons showed no significant effect of the rs12979860 genotype on the development of any fibrosis (Ishak grade ≥ 1) 1 year after LT.40 The other found fibrosis progression to Batts-Ludwig stage 2 or higher 1 year after transplantation in 12%, 25%, and 32% of the recipients with the CC, CT, and TT genotypes, respectively (P = 0.02), whereas the donor genotype was not associated with the stage of fibrosis.19 Larger histological analyses are needed.

Apart from the effects of IFN-λ on ISG induction, its role as a modulator of the adaptive immune system with proinflammatory properties suggests that it could facilitate lymphocytic infiltration of the allograft. Natural killer cells, which are abundant in the liver and are likely affected by IFN-λ, could also participate in fibrogenesis and serve as a link between the innate and adaptive immune systems.41-43 Because inflammation has been described more often in relation to the donor CC genotype with recurrent HCV, it is possible that the persistence of donor-derived natural killer, dendritic, and lymphocytic cells within the allograft could create a unique immunological environment favoring inflammation and fibrogenesis in the absence of viral eradication.

In a complex but striking case report, a 51-year-old Japanese male with the rs8099917 TG genotype underwent dual-graft living donor LT.44 The donor of the left lobe graft had the favorable TT genotype at rs8099917, whereas the donor of the right lobe graft had the unfavorable TG genotype. The patient was a virological relapser with respect to posttransplant antiviral therapy. Posttransplant biopsy samples were obtained from each graft 2 years after the antiviral therapy and relapse. The specimens showed minimal inflammation and no fibrosis in the graft from the TT donor and moderate inflammation and periportal fibrosis in the graft from the TG donor. More notably, quantitative polymerase chain reaction using RNA extracted from each graft revealed an absence of HCV RNA only in the graft from the TT donor. This case illustrates the potential impact of the donor IL-28B genotype on posttransplant outcomes and strategies for donor selection for HCV-positive recipients.

IL-28B AND ACUTE CELLULAR REJECTION (ACR)

Intuitively, it seems possible that the favorable IL-28B SNPs associated with a more vigorous immunological response to viral infections might also increase the risk of other immune phenomena such as ACR. This possibility was suggested in a study by Bitetto et al.,40 who assessed the risk of ACR in 251 consecutive patients undergoing LT (40% with cirrhosis due to HCV). During the first year after transplantation, the authors found a significantly lower risk of ACR in recipients with the CC genotype at rs12979860 (20.6% with CC, 34.1% with CT, and 47.8% with TT, P = 0.003). In a logistic regression analysis, only the rs12979860 genotype and cytomegalovirus reactivation were found to be predictive of ACR. Interestingly, the association was found to be very significant in patients treated with tacrolimus but not in those treated with cyclosporine. Also, the association was weaker in patients with HCV infections versus those with other etiologies of liver disease. The 2 published studies that have reported on the association between ACR and IL-28B SNPs in HCV-only cohorts found no significant variation in risk among the available genotypes for the rs12979860 locus (donor and recipient) or the rs8099917 locus (recipient).18, 21 Larger, more comprehensive studies of this issue are needed.

IL-28B AND GRAFT/OVERALL SURVIVAL

Three studies reported no association between rs12979860 SNPs and either graft or patient survival.19, 21, 45 However, the largest of these studies reported that patients with a CC genotype allograft and recurrent HCV were more likely to progress to a composite endpoint consisting of cirrhosis, retransplantation, and liver-related death than those with a non-CC genotype (hazard ratio = 2.59, 95% CI = 1.15-5.83). Although recipients with a CC genotype seemed to have an improved prognosis, this effect disappeared after patients were censored at the time of antiviral therapy, and this showed that the beneficial effect was indirectly given by its positive effect on viral eradication.19 It is likely that for an effect on overall survival to be observed (or excluded), larger cohorts, longer follow-up, or both are required. Available evidence points to a paradoxical effect of favorable IL-28B SNPs on graft survival where these could positively influence the graft and improve its prognosis if viral eradication is achieved (particularly for recipient IL-28B) or degrade it toward graft failure whenever HCV persists (particularly for donor IL-28B) and thus have a negative impact on survival.

IL-28B AND OTHER OUTCOMES

Fibrosing Cholestatic Hepatitis

Although the recurrence of HCV infections is nearly universal after LT, a minority of patients will develop a cholestatic form of hepatitis that rapidly progresses to advanced fibrosis and graft failure. Two studies have evaluated whether the IL-28B genotype plays a role in the risk of developing such cholestatic hepatitis. Graziadei et al.45 evaluated a cohort of 164 posttransplant HCV patients and assessed factors associated with the development of cholestatic hepatitis. rs12979860 genotyping was available for 151 of the recipients (92%) in this cohort but for only 53 of the donors (32%). The genotypes for the entire cohort were CC (24%), CT (62%), and TT (14%). However, in the subgroup of recipients with fibrosing cholestatic hepatitis (n = 20), the CC genotype was significantly less common (CC in 5%, CT in 70%, and TT in 25%). In a multivariate analysis, independent predictors of developing cholestatic hepatitis were the HCV viral load 2 weeks after transplantation and a non-CC recipient genotype at rs12979860. The donor IL-28B genotype was not found to be associated with outcomes, but this assessment was limited because of the poor availability of donor genotyping.

The Mayo Clinic cohort identified 12 cases of fibrosing cholestatic hepatitis in 11 patients.19 No association was identified between the recipient or donor genotype at rs12979860 and the incidence of cholestatic hepatitis. Unfortunately, the small samples of both studies limit the generalization of their results, and because of the rarity of this condition, multicenter studies should be sought.

Hepatocellular Carcinoma (HCC)

HCC is an increasingly common complication of cirrhosis, and LT is the most effective treatment option for patients with limited disease. Among patients undergoing LT for other reasons, previously undiagnosed HCC is often found in liver explants at the time of surgery. Because of the suspected role of tissue inflammation in the pathogenesis of HCC, it has been speculated that the IL-28B genotype may affect the risk for the development of HCC. Eurich et al.23 evaluated this possibility in a study of 167 patients with cirrhosis due to HCV who underwent LT; 61 were found to have HCC in hepatic explants. The recipient rs12979860 genotype was determined for all patients. The TT genotype was found to be markedly overrepresented in the HCC subgroup (55%) versus the total LT cohort (19%, P = 0.041). No association was found between the rs12979860 genotype and the severity of HCC (tumor size, histological grade, and presence of vascular invasion), but the TT genotype was associated with higher levels of alpha-fetoprotein among patients with HCC. In a multivariate regression analysis, the patient's age (P < 0.001) and the IL-28B genotype (P = 0.040) were the only factors independently associated with the risk for HCC.

In another study, Fabris et al.46 also evaluated the prevalence of rs12979860 genotypes in a cohort of 256 post-LT patients (41% with HCV); 85 had HCC in the liver explant at the time of surgery. In this cohort, the CC genotype was found to be associated with a significantly lower risk of HCC (46.2% with the CC genotype and 53.8% with a non-CC genotype in the non-HCC subgroup and 27.1% with the CC genotype and 72.9% with a non-CC genotype in the HCC subgroup, P < 0.005). This difference was found to be even greater in HCC patients with HCV (17.1% with the CC genotype and 82.9% with a non-CC genotype).

RISK OF DIABETES MELLITUS (DM) AND STEATOSIS

The development of type 2 DM, which is strongly predictive of posttransplant outcomes in patients with HCV infection,47 has been recently investigated in patients from the Mayo Clinic cohort without pretransplant diabetes (Table 2).39 In this study, 69 of the 221 included patients (31%) developed posttransplant DM. The TT genotype in the recipient was associated with posttransplant DM in a multivariate analysis (hazard ratio = 2.5, 95% CI = 1.2-5.4). Although the underlying mechanism is unknown, it can be speculated that the increased ISG levels observed in patients with the unfavorable genotype7-9, 33, 48 could potentiate pathways associated with insulin resistance in the context of an HCV infection, such as the suppressor of cytokine signaling family.49-51 Further analyses are needed to more definitively determine the effect of the IL-28B genotype on the risk for posttransplant DM.

The presence of hepatic graft steatosis in posttransplant patients has been associated with an increased risk of progression to advanced fibrosis52 (Table 2). IL-28B SNPs have been associated with the risk of developing hepatic steatosis in nontransplant patients with HCV.53, 54 Whether the IL-28B genotype is associated with the development of steatosis in posttransplant patients is less clear because no published studies have yet addressed this question. However, further characterization of the Mayo Clinic cohort showed a significantly increased prevalence of allograft steatosis with the donor TT genotype at the rs12979860 locus 1 year (38% with the CC genotype and 61% with the TT genotype, P = 0.047) and 4 years (63% with the CC genotype and 86% with the TT genotype, P = 0.02) after transplantation (unpublished data; Dr. Michael Charlton, Mayo Clinic, Feb 2012). This association persisted after corrections were made for the patient's body mass index, the HCV genotype, and coexistent type 2 DM. The severity of steatosis was also higher in allografts from TT donors. No association was found between graft steatosis and the recipient IL-28B genotype. As suggested in the nontransplant setting,53 it is possible that the effect of non-CC genotypes on increased ISGs could favor intrahepatic lipid deposition by suppressing lipoprotein lipase and limiting the conversion of very low-density lipoprotein to low-density lipoprotein. Unfortunately, there was no lipid analysis to provide some insight into the involved metabolic mechanisms. However, the effect was independent of DM, and it remains to be determined why recipient IL-28B favors DM, whereas donor IL-28B favors allograft steatosis.

POSSIBLE IMPLICATIONS FOR LIVER ALLOGRAFT ALLOCATION

The composite results of the reviewed studies raise the question whether the pretransplant determination of the IL-28B genotype has consequences for the further management of HCV patients with end-stage liver disease. Although patients with unfavorable IL-28B genotypes experience reduced SVR rates, the recipient IL-28B genotype is not an important predictor of patient or graft survival. There is no compelling basis for withholding LT from patients with unfavorable IL-28B genotypes. However, because more than half of long-term deaths and graft losses among HCV-infected recipients are related to the recurrence of HCV55 and because an SVR to posttransplant antiviral therapy confers a substantial survival benefit, the consistently reported superior SVR rates observed in patients with a favorable donor IL-28B genotype suggest that the preferential allocation of organs from donors with favorable IL-28B genotypes might improve the outcomes of LT in patients with HCV infections. Available technology would allow this practice in living donor LT, but low-cost technology for a rapid assessment of IL-28B genotypes during the short-timed process of allograft allocation would be needed for deceased donor transplantation.

When we consider the possible CC genotype pairs, the preferential allocation of organs from CC genotype donors could result in 33% of transplant recipients having both donor- and recipient-derived CC genotypes with an expected SVR rate of 90%, and the remaining non-CC recipient/CC donor pairs would achieve an expected SVR rate of approximately 47%.19 These figures are highly improved from the current SVR frequency of 35% to 40% observed at most experienced centers. An enhanced SVR frequency would be projected to improve posttransplant patient and graft survival.56 The downside is that according to recent findings, patients who are unable to clear the virus will have a higher risk for adverse outcomes (progression to cirrhosis, retransplantation, or liver-related death), and this will likely counterbalance the positive effect on SVR. The possible negative effect on the non-HCV transplant population will also need to be considered in light of the possible higher ACR rate observed with unfavorable IL-28B SNPs. Finally, the magnitude of any positive effect in the chronic HCV population will be diluted if treatment with direct-acting antiviral agents becomes a reality for LT recipients because the differences among the IL-28B genotypes are not so striking for direct-acting antiviral agents.57 Thus, the preferential allocation of CC genotype allografts to patients with chronic HCV seems unwarranted on the basis of the available literature.

SUMMARY

Although the study sizes have been relatively modest, recipient and donor liver IL-28B genotypes appear to be strongly and independently associated with IFN-based treatment responses in LT recipients. Interactions between donor and recipient IL-28B genotypes are complex and may affect outcomes unrelated to HCV infections such as inflammatory activity, fibrosis progression, the time to the occurrence of liver-related adverse outcomes, and metabolic diseases. Because the likelihood of achieving SVR after posttransplant antiviral therapy can range from very low to almost certain, the determination of the IL-28B genotype should be considered for donors and recipients. Further studies are needed to definitively determine the influence of IL-28B on ACR and overall survival. The allocation of favorable IL-28B genotype allografts to patients with chronic HCV, although an attractive possibility for improving posttransplant outcomes, cannot be recommended in light of the available evidence.

Ancillary