Hepatology highlights


  • Potential conflict of interest: Nothing to report.

Escape From HLA-B27 Restriction Is No Longer Fiction

Among the myriad of hypothesized mechanisms for HCV persistence, the role of HLA-restricted T-cell escape mutations has been controversial. In another of a series of elegant studies, the “Freiburg Fraternity” has convincingly shown that HLA-B27 restricted epitopes are the site of intense immune pressure and are important in HCV clearance. Many groups have looked for HLA association in HCV infection and some have been identified, but in the end these studies are unconvincing because they are just statistical associations. Neumann-Haeflin and coworkers have taken us a step further. It had been previously noted in the outbreak of hepatitis C among German women receiving contaminated Rh-immune globulin, that recovery from infection was significantly associated with HLA-B27. In the current study, the investigators identified 5 HLA-B27 restricted T-cell epitopes and showed that 5 of 6 women who had cleared HCV infection mounted strong CD8 responses to one epitope in the NS5B region (NS5B 2841) and not to any of the other B27-restricted epitopes. They then studied 8 HLA-B27+ patients who were chronically infected (only 1 in the Rh cohort). While T-cell responses to NS5B 2841 were more common in recovered subjects (5/6) than chronic patients (3/8), this difference was not significant (P = .14). Importantly those with chronic infection showed sequence variations in the dominant HLA-B27 2841 site that were far in excess of those observed in HLA-B27 negative subjects (P = .015), in the Los Alamos database (P < .0001) or in those that had recovered from infection. This result suggested that chronically infected patients had experienced selective pressure against this viral region and had evolved T-cell escape variants. This study thus nicely links two observations: (1) the association of recovery from HCV infection with HLA-B27, and (2) the evolution of critical T-cell escape mutations in patients with chronic infection. The authors also showed that T-cell lines from the chronically infected patients had functional impairments against the variant peptides compared to the wild-type as did intrahepatic CD8+ cells. These findings, together with prior studies in chimpanzees (PNAS 1995;92:2755, J Exp Med 2004;200:1593) and humans (J Exp Med 2005; 201:1741), reinforce a critical role for T-cell escape variants in persistent HCV infection and for the important role of certain HLA-types, particularly HLA-B27, in fostering immune-mediated recovery. (See HEPATOLOGY 2006;43:563–572.)

Building a Better Mousetrap: Conjugated Ribavirin in Murine Hepatitis

In seeking the cheese of better cure rates, Levy et al. have developed a clever mousetrap. The primary limitation of ribavirin therapy is a dose-dependent hemolytic anemia that occurs to some degree in 50% of treated patients. This study attempted to limit the exposure of red cells to free ribavirin and to target ribavirin directly to the liver. Their approach is founded on the fact that free hemoglobin is rapidly bound to haptoglobin and that the hemoglobin-haptoglobin complex is then taken up by Kupffer cells and hepatocytes trough the scavenger receptor CD163. The investigators developed a conjugate of acellular hemoglobin (Hb), haptoglobin (Hp) and ribavirin that they designated HRC 203. Multiple ribavirin molecules were conjugated to each hemoglobin molecule enhancing the potential for higher intracellular concentrations. Receptor-mediated binding and uptake of HRC 203 has already been demonstrated in model cells expressing Hb-Hp receptors and lysosomal enzymes have been shown to release bioactive ribavirin from HRC 203 in vitro. To this trap, now enter mice inoculated with mouse hepatitis-inducing coronavirus, MHV-3, that replicates in macrophages and then invades the liver. Viral infection is characterized by a massive pro-inflammatory response, severe hepatic necrosis and fulminant hepatic failure occurring within 3-4 days of infection leading to death in 80% to 90% of infected mice.

In initial studies, the investigators examined the effect of HRC 203 on viral replication in macrophages and isolated hepatocytes. Whereas free ribavirin had no effect on replication at the dose utilized, a lower dose of HRC 203 resulted in a 90% and 75% reduction in viral replication, respectively. Further, HRC 203 inhibited production of TNF-α twenty-fold compared to free ribavirin. These experiments served as prelude to assessing the effect of HRC 203 in experimentally infected mice. In studies involving 10 mice per group, all MHV-3 control mice had massive ALT elevations (ALTmax 7986 ± 325 IU/L) and all died within 3 days compared to ribavirin-treated mice (ALTmax 964 ± 128; 100% survival) and HRC 203-treated mice (ALTmax 848 ± 212; 100% survival). Although there was no significant difference in ALTmax or survival between free ribavirin and HRC 203, viral titers in the ribavirin group were more than twofold higher than in the HRC 203 group (P < .001). Further, ribavirin-treated animals had clinical symptoms, a fall in hematocrit and elevated BUN whereas HRC 203-treated mice were symptom free and maintained normal hematocrit and BUN.

So, are there any mouse holes in this study? Obviously, the main caveat is that these are indeed mice and not humans and that the virus being treated, while an RNA virus, is quite distinct from HCV. Nonetheless, the concept of conjugating ribavirin and facilitating receptor mediated targeting is quite sound. The implication from this study, and a previously highlighted study of escalating ribavirin dosage [HEPATOLOGY 2005;41:275–279] is that the more ribavirin, the better, if high concentrations can be achieved without limiting toxicity. Conjugation to normal human proteins with established macrophage and liver receptor mechanisms is a clever and rational approach that I believe will find its way into clinical trials. We may find further holes in this cheese once it is put to a definitive test in humans, but I am optimistic it will work. If not, I will shut my trap. (See HEPATOLOGY 2006;43:581–591.)

HBV: The Longitudinal That Follows the Horizontal

This study of Bortolotti et al. represents the final report of a 29-year longitudinal follow-up of a homogeneous population of HBV genotype D-infected Caucasian children from Padova, Italy. This long-term prospective follow-up provides important insights into HBV infection in children that was spread horizontally rather than from maternal transfer. It also addresses outcomes in a Caucasian population rather than the more frequent studies emanating from Asia. Overall, this study speaks to the relative benignity of HBV infection acquired in this manner at a young age. The investigators followed 91 HBsAg+/HBeAg+ and 8 HBsAg+/ HBeAg- children (Fig.). Among the 91 HBeAg+, 2 were lost to follow-up and the remaining 89 (100%) lost HBeAg and had anti-HBe seroconversion; of these 89, 4 had cirrhosis on presentation. Of the 85 without cirrhosis, 84 were classified as inactive carriers and only 1 had HBeAg-negative hepatitis. During a mean follow-up of 14.5 years, 4 of the 84 (4.7%) inactive carriers had a reactivation, 3 developing persistent, though generally low-grade hepatitis. Interestingly of the 4 with initial cirrhosis, 2 lost histologic features of cirrhosis as they reached adulthood. The other two developed small, resectable carcinomas and are still living with well-compensated cirrhosis. Of the 8 who were initially HBeAg-, 6 are inactive carriers and 2 have HBeAg-negative hepatitis. At the end of follow-up, 15/89 (17%) initially HBeAg+, not only lost HBeAg, but also HBsAg and HBV DNA suggesting true recovery. Thus, in this well defined, carefully followed long-term cohort, only 2% had a serious outcome over three decades, a percentage similar to that in children and young adults infected with HCV. Age at infection appears to be a critical determinant of outcome for both these agents. The unanswered question for both these agents is whether a stable, benign infection might exacerbate with the relative immunodeficiency of aging or with intercurrent immunosuppressive diseases or treatments. Also unanswered is whether one can acquire horizontal infection while standing up. (See HEPATOLOGY 2006;43:556–562.) 1.

Illustration 1.

Three Hard Looks at Liver Fibrosis

In this last quarter of HEPATOLOGY issues, three independent studies examined cofactors and predictors of progression of fibrosis, including the influence of coexistent schistosomiasis, steatosis and multiple factors associated with IVDU. Concomitant schistosomiasis and HCV has been shown previously to run a more rapid and severe histological course and to respond poorly to interferon. In this study, the calculated rate of progression of fibrosis was markedly greater in the coinfected population (0.6 ± 0.13) compared to HCV-monoinfected patients (0.1 ± 0.06; P < .001). Other points of clinical interest include: (1) HCV RNA levels were significantly higher in patients coinfected with S. mansoni (3.8×106 cop/mL vs. 1.6×106 cop/mL; P < .001); (2) at the end of follow-up >90% of coinfected patients had splenomegaly and esophageal varices compared to only 5% in HCV monoinfection; (3) progression of fibrosis to stage 3 or greater was seen in 82% of patients with schistosomiasis compared to none in the HCV monoinfected group. However, missing in this study is a third arm of patients who had only schistosomiasis. How many of these would have had severe outcomes, such as portal hypertension, even in the absence of HCV? Are the effects of HCV and schistosomiasis merely additive or is there is a synergistic effect that leads to an exponential progression of fibrosis? Nonetheless, these drastic outcomes make it appear that schistosomiasis may be the most detrimental cofactor in HCV fibrosis progression, surpassing alcohol, HBV, HIV and NASH.

Given this striking difference in progression of fibrosis, the investigators were able to compare predictors of fibrosis during this 10-year prospective study. The markers for fibrosis analyzed in both serum and liver were YKL-40, also known as human cartilage glycoprotein 39 (HC gp-39), N-terminal propeptide of collagen III (PIIINP), TGF-β, and TNF-α. Levels of these markers were comparable in both cohorts at entry and through the first 2 years of infection. Coinfected patients then had a sharp increase in serum YKL-40 and TGF-β beginning in the 3rd or 4th years of follow-up and these markers, particularly YL-40, correlated strongly with progression of fibrosis (r = 0.892) (Fig.). There was also a very strong correlation between serum levels of YKL-40 and TGF-β and hepatic mRNA expression confirming their hepatic origin and the utility of serum measurement. In the combined experimental and validation cohort analysis, only 2 patients without disease progression had increases in YKL-40 and only 2 patients with progressive disease failed to increase their YKL-40 levels (sensitivity and specificity, 96%). Overall, this is a very powerful study (indeed a helminth of a study) that makes a strong case for the use of YKL-40 and TGF-β as predictors of the progression of fibrosis. Because these markers are more specific for collagen deposition and extra-cellular matrix formation and because their validity was established prospectively, they are more intrinsically appealing than the many other indirect predictive markers that principally distinguish cirrhosis from mild disease, but do not adequately assess the stages leading to cirrhosis. (See HEPATOLOGY 2006;43:771–779.) 2.

Illustration 2.

Perumalswami et al. examined the computerized pathology database at NIH to assess the influence of steatosis on the progression of fibrosis in 494 patients with chronic hepatitis C, predominantly white males of genotype 1. Steatosis was observed in 40%, but only 9% were classified as moderate to severe steatosis. In a multivariate analysis, steatosis was significantly associated with older age, high BMI and ALT level, but not with fibrosis. Among 136 patients who had two liver biopsies, 40% had progressive fibrosis, but this progression did not correlate with the baseline level of steatosis (P = .58), nor did change in steatosis score correlate with change in fibrosis score (r2 = 0.0007, P = NS).. This lack of association between fibrosis and steatosis conflicts with previously published data, suggesting the effect of steatosis depends on the population studied, the severity of the steatosis and perhaps whether the steatosis is simply obesity related or part of the metabolic syndrome; in this study, less than 5% had diabetes, insulin resistance was not measured and alcoholism was infrequent. We are left to conclude either that steatosis per se is not relevant to the progression of fibrosis or that the fibrogenic influence of steatosis depends on the etiology and extent of the steatosis. Thus, when patients have an association between steatosis and fibrosis in cross-sectional studies, the steatosis may be a red herring, albeit a fat red herring. The interplay of HCV, obesity, the metabolic syndrome, hepatic steatosis, cytokines, oxidative stress, inflammation and fibrosis is very complex and a synthesizing model is not yet evident. Basically, the problem of fat is still very large and something to chew on. (See HEPATOLOGY 2006;43:780–787.)

Wilson and coworkers have taken another look at the Baltimore IVDU cohort to examine the progression of HCV-associated fibrosis in paired liver biopsies obtained over a mean interval of 4.2 years and an average 22-year estimated duration of infection. While the dogma has been that IVDUs will fare badly because of continued exposure, concomitant alcoholism and coexistent infections, particularly HBV and HIV, this cohort has done surprisingly well despite the fact that 82% were male, 27% HIV-infected, 4% HBV infected and 58% heavy drinkers. At baseline biopsy, only 10% had a modified Ishak fibrosis score of 3 or greater. After 4.2 years, 21% had progression of fibrosis (increase ≥ units), 7% to cirrhosis. Overall, after an estimated 26 years of infection, 81% had mild or no fibrosis and 72% had an HAI inflammatory score less than 5. The authors then focused on predictors of progression of fibrosis . Progression was observed more often in those with higher HCV RNA levels and higher ALT, AST and GTT levels at baseline. However, 28% of those with progression of fibrosis had normal aminotransferase and low HCV RNA levels. Progression of fibrosis was not associated with gender, age, lifetime or recent alcohol use, frequency of IVDU, steatosis or BMI. Perhaps most surprising, progression of fibrosis was not increased in those HIV infected, possibly explained by the fact that those with CD4 counts < 200/mm3 were excluded from study. After multivariate logistic regression analysis, only HCV RNA level >7.2 log copies/mL (OR = 3.04) and ALT >60 at baseline (OR = 2.7) were significant predictors of fibrosis outcome (P = .02 and .04, respectively). Published predictors of fibrosis, FibroSURE and APRI were predictably non-predictive. Of those with FibroSure or APRI scores predictive of progression, only 31% and 37.5%, respectively, developed significant fibrosis over the interval studied and PPVs were <18%. In contrast, these algorithms predicted insignificant fibrosis with >95% sensitivity, but even this might be spurious given the low overall prevalence/incidence of fibrosis in the cohort. The bottom line of this study seems to be that despite multiple factors favoring the development of fibrosis, this IVDU cohort had similar rates of severe fibrosis as the general HCV-infected population—approximately 20% after 25 years. This result might be because the negative factors were ameliorated by the relatively young age of the cohort and the high proportion of African Americans (96%). Such a balancing to the mean outcome might have been logically predicted, but logic is the only predictor that worked in this study. (See HEPATOLOGY 2006;43:788–795.)

Cyclosporin-A Comes Full Cycle

Cyclosporin A (CsA) is an 11 amino acid cyclic peptide that has potent immunosuppressive activity. Less recognized is the fact that it is a potent inhibitor of HIV replication and also has activity against schistosoma and plasmodium. Recently, CsA has also been shown to inhibit HCV replication in vitro and to enhance IFN therapy in small clinical trials. The problem with CsA as an anti-viral is its immunosuppressive effect. Now under the heading of “better living through chemistry” a non-immunosuppressive cyclosporin (DEBIO-025) has been developed. The primary modification in this compound is in the calcineurin binding domain, the region critical to its immunosuppressive effect. In a series of elegant studies involving several HCV replicons and an infectious clone, Paeshuyse et al. show that DEBIO-025 has potent inhibitory activity against HCV which significantly exceeds CsA. In in vitro studies with DEBIO-025 plus interferon, an additive, but not synergistic effect was shown consistent with their differing modes of action. In other studies DEBIO-025 has been shown to have anti-HIV activity that is 10-fold greater than CsA. It is thus tempting to think this drug might be particularly useful in treating combined HCV-HIV infections. The postulated mechanism of anti-viral activity of DEBIO and CsA is inhibition of the peptidyl-propyl cis/trans isomerase (PPIase) activity of cyclophilins. Sounds good to me. Whatever the mechanism, these in vitro data beg for a randomized, controlled clinical trial in which DEBIO-025 is combined with Peg-IFN or Peg-IFN plus ribavirin. Then we could truly find if we can fit a square Peg into a cyclic sporin. (See HEPATOLOGY 2006;43:761–770.)

Therapeutic Capsules

  • 65 HIV and HBV co-infected patients who received Tenofovir for at least six months during HAART were retrospectively analyzed; 80% were also receiving lamivudine (LAM) and 69% were LAM resistant. In 54 HBeAg+ subjects, HBV DNA declined a median of 4.56 log copies; among HBeAg negatives, the log reduction was 2.53. HBV DNA became undetectable in 30% and 82% of HBeAg+ and HBeAg- subjects, respectively. The tenofovir response was durable in 97%. Within the limitations of a retrospective study, Tenofovir showed efficacy against wild-type, presumed precore mutants and LAM-resistant HBV as an incidental bonus to HAART therapy for HIV. (See HEPATOLOGY 2006;43:548–555.)

  • In a multicenter study in Germany, 89 patients with acute hepatitis C were treated with 1.5 μg/kg peginterferon-α2b for 24 weeks beginning at a median of 76 days (range 14-150) after infection. In the total population, SVR was 71%, but that increased to 89% in the 65 (73%) who adhered to at least 80% of scheduled treatments. Thus, compliance was suboptimal, but sustained efficacy was higher than in treating chronic infection, though monotherapy would no longer be used for chronic infection. This study cannot address whether efficacy would have been higher had ribavirin been added. I still have the sense that acutely infected patients should be given 3-4 months to achieve spontaneous clearance prior to initiation of either mono or combination anti-viral therapy. (See HEPATOLOGY 2006;43:250–256.)

  • In a randomized trail of IFN-α2b (5MU daily x 4 weeks; then 5MU 3 times/wk × 28) plus ribavirin 1200 mg daily versus IFN alone for HBeAg+ chronic hepatitis B, no benefit was observed by the addition of ribavirin. 24 weeks post treatment, the combined loss of detectable HBV DNA and HBeAg seroconversion was between 20% and 25% in each treatment arm. Thus, the vast majority of HBV infected patients do not respond to IFN and ribavirin does not improve this outcome. (See HEPATOLOGY 2006;43:742–749.)