Hepatology highlights

A New Paradigm for Fibrosis Formation: Great Deductule Reasoning

Clouston and coworkers in Australia have written a provocative and important manuscript that presents a new paradigm for fibrosis evolution in the liver. Their hypothesis is that a fundamental defect underlying fibrosis is hepatocyte replicative arrest due to a variety of insults including toxins, viral infections and steatosis. Hepatocyte replicative arrest in turn stimulates an alternate proliferative pathway involving hepatic progenitor cells (HPCs) that can differentiate into either hepatocytes or bile ductular epithelium. HPCs reside predominantly in the periphery of the portal tract and are in close proximity to bile ductular cells. In this hypothesis, activation of the alternate pathway leads to a ductular reaction that expresses cytokines that attract and activate stellate cells leading to collagen deposition at the portal tract interface. To test this hypothesis, the authors used immunohistochemistry and image analysis of paraffin liver sections to quantify the degree of replicative arrest, the number of HPCs, the extent of ductular reaction and stellate cell activation in 115 liver biopsies from HCV RNA positive patients and from 15 normal livers. These parameters were assessed against the severity of fibrosis and steatosis. A ductular reaction, identified by cytokeratin-7 reactivity, was very common in HCV infected livers, was located at the portal interface and was significantly associated with fibrosis stage (P <.0001). The mean number of HPCs was 10 fold higher in HCV-infected than control livers (P < .0001) and these were located in close proximity to the ductular reaction. HPC numbers also correlated with fibrosis. Steatosis also significantly correlated with HPC numbers and the bile ductular reaction, but these changes could occur in the absence of steatosis. In a multivariate analysis, interface hepatitis was independently associated with both HPC and the bile duct reaction. Lastly, hepatocyte replicative arrest, as measured by p21 expression, was associated with HPC expansion (P = .002), increased BMI (P < .001) and lobular inflammation (P = .005).1

In summary, these data support the hypothesis that a relative proliferative arrest induced by viruses, steatosis and other insults, drives an alternate regenerative pathway and a periportal ductular reaction that then triggers a profibrogenic reaction at the portal tract interface. It does not, however, prove this mechanism is directly related to fibrosis, as it did not measure the signals generated that might bring stellate cells and myofibroblasts to the periportal area. Nonetheless, this study potentially answers two important “whys” (1) why does an infection of hepatocytes within lobules produce fibrosis predominantly in the portal area? (2) why and how does steatosis lead to periportal fibrosis? What the study doesn't answer is, why didn't I think of this first? (See HEPATOLOGY 2005;41:809-818)

Radical Ribavirin Regimen Reveals Robust Responses

Although ribavirin alone does not inhibit HCV replication, as an adjunct to interferon therapy, it has an unexpected and remarkable effect on enhancing the rate of sustained virological response (SVR). Now, Lindahl et al. test the hypothesis that a higher dose of ribavirin, individualized to drug levels in a given patient, could have an even more profound effect. The initial ribavirin dose was calculated from a pharmacokinetic formula that took into account renal function as assessed by creatinine clearance. The authors had previously shown that renal function was of greater importance than body weight in determining ribavirin clearance and effective blood level. In this study, they sought to achieve a steady state ribavirin concentration of >15 μmol/L, about twice the level achieved with standard therapy. Ten previously untreated patients with chronic hepatitis C (genotype 1) and high viral load were treated with a combination of peg-IFN plus individualized high-dose ribavirin for 48 weeks; dose adjustments were based on plasma levels measured by HPLC every 2-4 weeks. The median initial calculated ribavirin dose was 1520 mg/day (range 1,200-2,200), but less than desired plasma levels led to gradual dose escalation such that by 24 weeks the mean daily dose was 2540 mg/day (range 1,600-3,600) resulting in a near target ribavirin concentration of 14.7 μmol/L (range 7.8-22). The highest dose given was 4000mg/day. Such doses make one gasp, but no patient was removed from study because of ribavirin side effects, though the tolerance for side effects was very high. All patients received prophylactic erythropoietin. In two patients, hemoglobin levels decreased below 8 gm/dL requiring blood transfusions on two occasions each. Five other patients had hemoglobin levels that fell to between 8 and 10 gm/dL leading to minor dose reductions in two. Fatigue, nausea and dermatological problems were considered more frequent and severe than usually observed with standard therapy and capacity to work was reduced in all patients.

Was the anemia and relative incapacitation worth it? It appears so: 9 of 10 patients became HCV RNA negative 24 weeks after cessation of therapy, an SVR far higher than has been observed for genotype 1 patients with high viral load. Although there are caveats to interpretation of these data, particularly small sample size and the lack of controls, it is quite striking that these response rates could be achieved with existing drugs by using a dose-escalation strategy. It had seemed that such high response rates would have to await the addition of new therapeutic agents combined into treatment cocktails. Clearly, this approach requires much closer monitoring of patients and the willingness to accept the inherent risks of blood transfusion. If validated in larger trials, this scheme will not be everyone's cup of tea, but it could be avidly swallowed by many. Overall, the authors present a rational and innovative approach that could greatly increase efficacy rates and would certainly be worth attempting in patients who have failed conventional therapies. Let the trials begin! (See HEPATOLOGY 2005;41:275-279)

Fitting a Square Peg-Interferon Into a Round Transplanted Liver

Reinfection of the transplanted liver is universal in HCV infection and can lead to accelerated fibrosis progression and ultimate graft failure. It is thus rational to consider anti-viral therapy in the transplant setting. The problem with this strategy is that by the time patents have advanced to end-stage liver disease, they have usually proven to be nonresponsive to antiviral therapy. Second, the side effects of interferon could greatly complicate posttransplant management and equally important, the immunomodulatory effects of interferon might enhance graft rejection. Chalasani and the Pegasys Transplant Study Group report on two prospective randomized control trials using pegylated interferon monotherapy (180 ugm/wk) within 3 weeks of transplantation (prophylaxis trial) or 6-60 months after OLT (treatment trial). 54 patients participated in the prophylaxis trial with 26 assigned to treatment and 28 to no treatment; only 15 in each group completed study. Although the treated group had better virological responses that reached significance at some study intervals, in the end, only 2 patients (8%) had an SVR in the treated group compared to none in the untreated group (P = .14). Biochemical response did not differ between the groups. The inflammatory grade and fibrosis stage increased in both groups during the course of the study and although the increase was slightly less in the treated group there was no statistically significant difference.

The results in the treatment trial were a little better. In this trial, treated patients had a significantly higher virological response rate during treatment and at the end of follow-up, but the SVR was still quite low (12% treated vs. 0% untreated; P = .03). There was no significant advantage to treatment in biochemical or histological response, though 76% had improved HAI compared to only 10% in the untreated arm. In the prophylaxis arm, the number of adverse events (AE), serious AEs, rejection episodes and deaths were similar in treated and untreated subjects. The proportion withdrawing from study was also similar in the two groups. In the treatment trial, 3 patients, all in the treatment arm had severe AEs and two additional treated patients died, but none of these AEs were attributed to peg-IFN. Four patients in the peg-IFN group and none in the untreated group had an episode of acute rejection (P = .11).2

These marginal efficacy data are fodder for different spins. While one can claim interim benefit of treatment, the bottom line is that the SVR rates were very low, the biochemical benefit was nil and there was no sustained histological improvement. Although no significant detrimental effects of treatment were observed, this remains a difficult treatment in the transplant setting and one with potential detriment to the graft. As good as peg-IFN is, I believe this is a square peg that doesn't fit in this transplant hole. Nonetheless, additional therapeutic trails are warranted, including those with peg-IFN plus ribavirin, but at the present time the evidence does not support routine antiviral therapy in the liver transplant setting. (See HEPATOLOGY 2005;41:289-298)

HCV Envelope is Opened to a New Role in Carcinogenesis

Because HCV does not integrate into the host genome, its role in the causation of hepatocellular carcinoma (HCC) is thought to be secondary to the induction of cirrhosis and the mitotic events that accompany compensatory hepatic regeneration. The question has always remained as to whether HCV could play a more direct role, even in the absence of insertional mutagenesis. Kamegaya and coworkers thus explored the carcinogenic role of HCV proteins using a transgenic mouse model. The effect of HCV core and HCV core-E1-E2 transgenes was studied in a diethynitrosamine (DEN)-induced murine cancer model. HCC developed between 24 and 32 weeks in all DEN-treated animals. The mean number of adenomas and HCCs/mouse did not differ significantly between HCV transgenic mice and nontransgenic mice. However, the mean size of adenomas and HCCs was significantly higher in HCV core-E1-E2 transgenics than in core transgenics (P = .008) or non-transgenics (P = .003). Further, using a Ki67 immunohistochemical proliferation index (PI), it was shown that proliferation was significantly higher in tumor than nontumor tissue, but in HCC tissue there was no difference in the PI of transgenic versus non-transgenic mice. In contrast, the apoptotic index (AI) was much lower in core-E1-E2 transgenics than in core transgenics or nontransgenic controls (P < .001 for all comparisons). These observations were confirmed in transfected HepG2 cells where the AI was again significantly lower in cells transfected with core-E1-E2 compared to core constructs alone. These data implicate E1 and/or E2 in the suppression of apoptosis.

The implications of these experiments are that HCV envelope proteins may play a direct role in HCC evolution by inhibiting apoptosis rather than by stimulating proliferation. In these studies, HCV proteins appeared to serve as a second hit in a chemically induced genetic injury. While unproved by this model, it is intriguing to consider that in humans, HCV may play a dual role (cirrhosis induction/apoptosis inhibition) in a multi-hit process of hepatocarcinogenesis. Other factors on the hit list might include generation of reactive oxygen species by steatosis and iron overload. The concept of a dual role for HCV in HCC causation is an important advance in our thinking about this devastating outcome of HCV infection. (See HEPATOLOGY 2005;41:660-667)

Truth Versus Pseudo-Truth?

The development of HCV pseudoparticles (HCVpp) constructed by simultaneous transfection of 293T cells with constructs for a retroviral core, an HCV envelope and a retroviral carrier of green fluorescence protein (GFP) has opened the door to studying HCV cell entry and inhibition of entry by neutralizing antibodies. In this study, Lavillette, et al generated a panel of full-length HCV E1-E2 glycoprotein clones representative of genotypes 1-6 and then measured their ability to infect a variety of cell lines and their requirement for specific receptors. Several interesting observations were made: (1) highly infectious titers (>5x10⁴) were observed for genotypes 1a, 1b, 2a, 2b, 4, 5 and 6; however, genotype 3 infectivity titers were lower by at least a log; (2) all liver cell lines, including Huh 7, PLC/PRF/5 and Hep3B were infected by HCVpp of each of the genotypes; Hep G2 cells, that do not constitutively express CD81, were not infected by any genotype; (3) importantly, none of the non–hepatic cell lines were infected by HCVpp including many cancer cell lines and T and B lymphoid cell lines. Similarly, human T and B primary cells could not be infected by HCVpp suggesting that lymphoid cells are probably not an important reservoir or replication site for HCV; (4) by inhibition of the proton pump necessary for acidification of endosomes, it was shown that HCV cell entry proceeds by endocytosis in a pH-dependent manner and that this path was conserved across genotypes; (5) CD81 was shown to be an essential receptor in that nonpermissive HepG2 cells could be infected by all genotypes after CD81 was expressed by transfection; conversely permissive HuH7 cells could be blocked 70%-99% by anti-CD81 antibody; (6) using silencing RNA interference assays, Scavenger Receptor Class-B Type 1 (SR-B1) was also shown to be important for cell entry of all genotype HCVpp; however, the genotypes were differentially SR-B1 dependent with HCVpp-1a being most affected by RNA silencing of SR-B1; (7) while some antibodies from HCV carriers have broad neutralizing capability, others neutralize in either a genotype or strain-specific manner.

In sum, although it appears that both CD81 and SR-B1 are required for HCV cell entry, they are not sufficient since CD81 and SR-B1 transfected CHO cells were unable to support infection by any HCVpp genotype nor were CD81/SR-B1- expressing HeLa cells. There is thus a hepatocyte-specific piece(s) of the HCV receptor complex yet to be discovered. These “pseudo” particles may or may not provide the “real” truth, but certainly they provide new insights into both cell entry and HCV neutralization. Hopefully, these findings can be translated into therapeutic immune globulins and drugs that work by receptor blockade. (See HEPATOLOGY 2005;41:265-274)

T-Regs to the Regscue

Not to be outdone by HCV, the hepatitis B virus has its own T-reg story, and the story is virtually identical. Stoop et al, found that the proportion of T-regs (CD4+, CD25+, CD45RO+) was significantly higher (5.4%) in PBMCs of patients with chronic HBV infection than in healthy controls (3.2%; P = .003) or those with resolved HBV infection (2.9%; P < .03). This was confirmed by measuring the relative FoxP3 (T-reg transcription factor) mRNA levels wherein chronic carriers had mean levels of 204 compared to 10.9 in healthy controls (P < .001). Fox P3 mRNA levels were 193-fold higher in CD4+CD25+ than CD4+CD25- cells supporting the relative specificity of CD25 for identifying T-regs. Depletion of CD4+CD25+ cells from PBMC of chronic HBV carriers resulted in significantly stronger proliferation upon HBcAg stimulation than nondepleted PBMC and both proliferation and IFN-γ responses could be diminished in a dose dependent manner by feed-back of CD4+CD25+ cells. HBeAg+ patients had a higher percentage of T-regs in PBMC than HBeAg- patients. This is of interest because HBeAg has been reported to induce tolerance in the murine model and this might be one mechanism of such tolerance. T-regs did not correlate with viral load or ALT level.

Hence, as in HCV infection, T-regs may play an important role in creating a diminished T-effector cell response to HBV and thus contribute to persistent infection. These T-regs may be induced, at least in part, by HBeAg. This study did not examine T-reg dynamics within the liver nor examine the mechanism of T-cell suppression. In-vitro studies of HCV and other agents, have suggested that the suppressive effect of T-regs involves a cell-to-cell contact mechanism in which TGFβ may play an important role. It has also been suggested in other studies That T-regs can be induced through repetitive stimulation of T-cells by high concentrations of antigen over prolonged intervals. Certainly HBV infection is characterized by high levels of surface antigen and sometimes HBeAg, but is this the chicken or the egg? Does high-level antigen induce T-regs or do the suppressive effects of T-regs lead to less viral containment and higher antigen levels? The former possibility is favored by the fact that HBV-specific T-regs do not correlate with higher viral load.

Lastly, these investigators have previously shown that T-regs can be induced by immature dendritic cells (DC) and that patients with HBV had impaired DC maturation. In turn, increased T-regs are capable of inhibiting maturation of DC. Thus, there may be a self-perpetuating regulatory loop wherein DC and T-regs feed off each other leading to HBV tolerance and chronic disease. All these possibilities are making me loopy and unregulated, so I will end here. (See HEPATOLOGY 2005;41:771-778)

Mini-Highlights

• A retrospective-prospective study in Japan traced 143 anti-HCV+ patients presumably infected by transfusion at the time of surgery for pulmonary TB; 145 TB patients without anti-HCV served as controls. Overall, 6% of HCV+ patients died of liver disease compared to none among controls. Of these 8 deaths, 7 were from HCC; these patients had been transfused at age 25-41 and developed HCC at age 65-75. The overall risk of liver-related death after a mean duration of 34 years is relatively low and commensurate with other studies of this type. However, the increasing risk of HCC with age raises the dilemma of whether this simply reflects duration of infection or a late acceleration of disease progression due to age-related immunodeficiency. (See HEPATOLOGY 2005;41:819-825)

• In a nationwide cohort of 27,150 HCV infected Swedish patients with an observation time of 122,272 person-years, the risks of non-Hodgkin lymphoma and multiple myeloma were significantly increased in those HCV-infected >15 years (observed/expected ratio 1.89 and 2.54, respectively). This large-scale analysis supports prior observations in smaller cohorts. (See HEPATOLOGY 2005;41:652-659)

• Plasmacytoid dendritic cells (pDCs) are the major endogenous IFN-α producing cells in the body and are important in the innate immune response. pDCs were found to be significantly reduced in acute hepatitis C, were immature and produced little IFN-α compared to healthy controls. pDC were inversely correlated with ALT. Less pronounced pDC malfunction was observed in chronic hepatitis C. It is probable, but not proved, that HCV itself suppresses pDC function as another of its mechanisms of immune escape. This study provides rational for therapies, such as CpG oligos, that have the capacity to stimulate pDCs to produce IFN-α. (See HEPATOLOGY 2005;41: 643-651)