SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Clinical Aspects of Fibrosis
  4. Basic Science Advances in Fibrosis
  5. Antifibrotic Therapy
  6. Cell-Based Therapy
  7. Concluding Remarks
  8. Acknowledgements
  9. REFERENCES

The third American Associated for the Study of Liver Diseases (AASLD)–sponsored Single Topic Conference on hepatic fibrosis was held in June 2006. The conference was both international, with 6 countries represented, and cross-disciplinary, linking the basic molecular and cellular biology of fibrogenic cells to clinical trial design for emerging antifibrotic therapies. The specific goals of the conference were: (1) to consolidate knowledge about the natural history of fibrosis; (2) to clarify potential endpoints and markers; (3) to emphasize new antifibrotic targets developed on the basis of advances in basic science; and (4) to understand current critical issues pertaining to clinical trial design. Given the tremendous growth of the field and the constraints of a 2-day format, the selection of speakers was a challenge. A number of topics not included in the oral presentations were featured at poster sessions, lending breadth and depth to the meeting as a whole. Surprising new themes emerged about molecular, clinical, and regulatory aspects of the field, and a consensus emerged that hepatic fibrosis has matured into an integrated discipline that promises to significantly improve the prognosis of patients with fibrosing liver disease. (HEPATOLOGY 2007;45:242–249.)

The Fibrosis Single Topic conference has tracked the field's growth from its inaugural meeting in 1989 to the 2000 meeting and now the 2006 meeting. The 1989 conference emphasized the biochemistry of extracellular matrix and its role in determining cellular behavior, encapsulating what could be called the era of matrix biology.1 In June 2000 clinical aspects of fibrosis emerged as part of the era of stellate cell biology. Intracellular signaling of stellate cells, in particular cytokine and nuclear receptor pathways of gene regulation, were active areas.2 This year's conference acknowledged the explosion of new data, which has framed the field as the era of the integrated wound healing response, incorporating a far broader spectrum of topics.

Clinical Aspects of Fibrosis

  1. Top of page
  2. Abstract
  3. Clinical Aspects of Fibrosis
  4. Basic Science Advances in Fibrosis
  5. Antifibrotic Therapy
  6. Cell-Based Therapy
  7. Concluding Remarks
  8. Acknowledgements
  9. REFERENCES

Natural History, Genetic Determinants, and Mouse Models of Fibrosis.

The conference opened with discussion about the natural history of fibrosis. Detlef Schuppan, (Beth Israel Deaconess Hospital) emphasized that elevated hepatic venous wedge pressure associated with fibrosis may indicate a poorer outcome, at least in patients with HCV following liver transplantation,3 although it is not clear whether the amount of fibrous tissue correlates with liver function. It also is not known how frequently fibrosis progresses or what clinical features predict its progression.

Progression of fibrosis has been best studied in patients with HCV infection. In addition to well characterized cofactors including age, sex, and alcohol use, coinfection with Schistosoma mansoni or the human immunodeficiency virus (HIV) accelerates progression by as much as 2- to 6-fold.4

Genetic modulation of fibrogenesis is emerging as a key issue. Genetically distinct mice may develop the same extent of injury following hepatic insult, yet they can have divergent fibrogenic responses. However, the evidence in humans that specific genetic polymorphisms regulate fibrogenesis is limited and contentious (see ref.5 for review). Hongjin Huang (Celera Diagnostics) and collaborators have performed a whole genome scan in patients with HCV infection to identify genetic determinants of fibrosis progression.6 They first identified more than 400 functional single-nucleotide polymorphisms (SNPs)6 using a data set including nearly 1,500 subjects and have more recently reported 7 SNPs that are associated with fibrosis progression. Although the identities or functions of the individual SNPs were not revealed, data were presented suggesting that a cirrhosis risk score could be developed. Many questions arose. For example, what are the mechanistic links of specific SNPs to fibrosis? What is the future predictive value of such a genetic risk score? Can environmental risk factors be added to help better predict risk? Also, what is the contribution of epigenetics to this combined risk?

A better understanding of the natural history of fibrosis progression and regression may be possible by using suitable animal models, a topic addressed by Hitoshi Yoshiji (Nara Medical University). The ideal model would (1) duplicate the features of human diseases, (2) exhibit gradual and discrete progression of pathologic changes, (3) have high reproducibility and low mortality, (4) display dynamic features including reversibility, analogous to human disease, and (5) parallel the pathophysiology of human disease. A growing number of genetic mouse models complement the traditional models of liver injury. These include transgenic mice, some of which overproduce specific proteins with temporal and spatial specificity, and knockout mice, which can be engineered (using cre-lox technology) to abrogate gene expression in a time- and tissue-specific manner.

Fibrosis Regression — Evidence and Underlying Mechanisms.

The reversibility of fibrosis in animal models has been confirmed by observations in humans, fueling the enthusiasm for antifibrotic therapies.7Nezam Afdhal (Beth Israel Deaconess Hospital) emphasized the different extents to which fibrosis regression may be tracked, ranging from histologic regression alone to regression associated with functional improvement (e.g., reduced portal pressure) to regression associated with reduced clinical events and prolonged survival. Future trials should provide evidence of improved long-term clinical outcomes, particularly in those with either sustained clearance of HCV or suppression of HBV and HDV. Preliminary findings from the CO-PILOT study using PEG-interferon-α2 maintenance therapy for advanced HCV fibrosis indicate reduced incidence of clinical events in those with portal hypertension (Afdhal, unpublished), but longer-term outcomes are awaited.

Substantial insights into how fibrosis might regress in these patients have emerged from animal models. John Iredale (University of Edinburgh) emphasized the roles of myofibroblast-like cells and macrophages in orchestrating the resolution of fibrosis in the healing liver. Tissue inhibitor of metalloproteinase-1 (TIMP-1) represents a critical switch because it both inhibits matrix proteases and promotes survival of fibrogenic cells, in part through induction of the antiapoptotic protein Bcl-2. Resolution of fibrosis is associated with reduced TIMP-1 expression.8 Potential substrates of metalloproteinases as TIMP-1 levels decline include N-cadherin as well as collagen I. Tissue transglutaminase-mediated crosslinking of fibrillar collagens may also control matrix resorption,9 rendering the hepatic scar resistant to degradation, especially as the local milieu becomes increasingly hypocellular. Whereas TIMP-1 expression is localized to activated stellate cells/myofibroblasts, hepatic macrophages are an important source of proteases, including MMP-13.10 Notably, resolution of fibrosis in CCl4-induced liver injury is attenuated in mice lacking MMP-13 (Iredale, unpublished), yet fibrosis is accelerated in these knockout animals with biliary fibrosis.11 The discrepant findings from these 2 models may indicate that the same proteases serve different roles depending on the type and duration of injury.

Assessment of Fibrosis.

Development of noninvasive markers of fibrosis is a high priority, as this will accelerate clinical trial development by clarifying endpoints (discussed later in this article). Robert Fontana (University of Michigan) reviewed current performance characteristics of both serum assays and those requiring only routine lab tests (e.g., APRI). Whereas current tests can reliably identify those with either minimal or very advanced disease, there remains a substantial fraction of patients whose indeterminate values preclude accurate fibrosis staging.12, 13 It remains possible that one or more serum panels may prove more predictive of clinical events than the static information from a liver biopsy, but such longitudinal data are not yet available.

A noninvasive test generating considerable interest is the Fibroscan, which sends a shear wave through the liver from a percutaneous probe, and then an ultrasound transducer within the device calculates shear velocity as a measure of matrix stiffness. As reviewed by Marianne Ziol (Hopitaux de Paris, Paris 13 University), Fibroscan can readily distinguish cirrhosis from precirrhotic stages of fibrosis (see ref.14 and its references), with the exact cutoff values for establishing cirrhosis dependent on the etiology of liver disease. Ongoing longitudinal studies are assessing its accuracy in tracking fibrosis progression and regression in patients on HCV therapy, with early data indicating its ability to distinguish nonresponders from sustained responders (Grando-Lemaire, unpublished). The noninvasive nature of the technology, combined with the “instant” results it provides, has made it appealing to patients, with studies under way in the United States to complement those reported from Europe.

New solutions to the problem of fibrosis assessment may come from advanced technologies that include microarrays and proteomics. Emanuel Petricoin (George Mason University) emphasized that detection of low-abundance serum proteins, which comprise less than 1% of the total proteome, may hold the key to distinguishing different disease states and responses to therapies.15 Such low-abundance proteins are typically bound to carrier proteins, in particular albumin; thus, methods that elute proteins away from their carriers in a controlled fashion represent a potential trove of diagnostic information. Reverse-phase microarrays, which measure phosphorylation levels of hundreds of signaling molecules at once from source material,16 offer tremendous sensitivity in quantifying low-abundance molecules. This technology is now being applied to the study of tissues from patients with NASH to identify biomarkers that may define disease pathogenesis, clinical subsets, and potential treatments.

Basic Science Advances in Fibrosis

  1. Top of page
  2. Abstract
  3. Clinical Aspects of Fibrosis
  4. Basic Science Advances in Fibrosis
  5. Antifibrotic Therapy
  6. Cell-Based Therapy
  7. Concluding Remarks
  8. Acknowledgements
  9. REFERENCES

The identification of hepatic stellate cells as key effectors of the liver's inflammatory response rather than simply targets of inflammation represents a recent conceptual advance. A major intracellular mediator of that response is NFκB and its related signaling intermediates, including Iκκ. Studies presented by David Brenner (Columbia University) and others17, 18 have clarified the importance of this pathway in regulating stellate cell survival. The role of NFκB was uncovered in part by analysis of the mechanism of action of gliotoxin, a fungal metabolite that selectively drives stellate cell apoptosis, associated with increased mitochondrial depolarization.19 NFκB induces the expression of inflammatory genes in activated stellate cells. In addition, NFκB protects cells from apoptosis by induction of a large number of antiapoptotic genes, some of which have been identified in stellate cells. Agents that induce apoptosis in activated stellate cells typically inhibit Iκκ and reduce NFκB activity, which, in addition to gliotoxin, includes sulfasalazine, nerve growth factor, and specific NFκB antagonists including NFκB essential modulator (NEMO) and Iκκ inhibitors.20 In addition, proteasome inhibition, which blocks NFκB activity by increasing the half-life of its inhibitors, is also antifibrotic.21 However, inhibiting NFκB activity alone does not always induce apoptosis in activated stellate cells in culture and thus convergence with other pathways may be necessary to provoke cell death.

Stellate cells also participate in inflammatory signaling through expression of Toll like receptor-4 (TLR4),22 part of a family of “pattern recognition receptors” driving the innate immune response.23 When a ligand such as LPS binds to TLR4, it induces an intracellular signaling pathway, including activation of NFκB. TLR4 is expressed on both Kupffer and stellate cells, and its genetic deletion reduces macrophage infiltration, injury, and fibrosis in TLR4 knockout animals with experimentally induced liver damage. Interestingly, gut sterilization also reduces liver injury and fibrosis after biliary occlusion, underscoring the physiologic importance of the signaling induced by LPS binding to TLR4 in the liver's response to an injury stimulus. The role of endogenous TLR4 ligands in the liver remains to be identified but might include HSP 60 and 70, hyaluronan, and/or high-mobility group box 1 protein (HMGB1).

Whereas apoptosis of stellate cells should be antifibrotic, hepatocyte apoptosis appears to be a profibrogenic stimulus. Studies by Gregory Gores (Mayo Clinic) have demonstrated that engulfment of apoptotic bodies drives stellate cell fibrogenesis,24 which is blocked by the microtubule inhibitor nocodazole. Clinical trials of small-molecule antagonists of caspases are currently under way.25 This approach is a form of “targeted chemotherapy” for liver fibrosis, which, when combined with agents that promote selective apoptosis of stellate cells, represents a 2-pronged attack on fibrogenesis by manipulation of apoptotic signaling in parenchymal and fibrogenic cells, respectively. Specificity in targeting these pathways will be essential to ensuring safety and efficacy.

In the kidney, conversion of tubular interstitial epithelial cells to fibroblasts through epithelial mesenchymal transition (EMT) is a well-established determinant of renal fibrosis26 and is also associated with carcinogenesis.27 However, evidence for EMT in the liver has been less convincing, and studies using cultured hepatocytes have been confounded by the recognition that hepatocyte isolates contain a small fraction of stellate cells. Michael Zeisberg (Beth Israel Deaconess Hospital) has used genetic lineage tracing in which animals expressing a Cre-recombinase driven by an albumin promoter can mark hepatocytes through beta-galactoside expression when crossed with Rosa26 reporter mice. With this approach, it was possible to identify in situ some cells expressing both beta-galactosidase and FSP-1, a fibroblast protein, following liver injury. The results of these studies suggest that up to about 40% of resident FSP1-positive fibroblasts could be a result of EMT in liver fibrosis. BMP-7, which inhibits fibrosis in liver models, may prevent or reverse EMT by antagonizing TGFβ1 signaling.28

Antagonism of fibrogenic mediators is a major area of interest, and connective tissue growth factor (CTGF, also known as CCN2) is an attractive target to block because, unlike TGFβ1, this molecule is neither immunomodulatory nor growth suppressive. David Brigstock (Ohio State University) is defining the molecular interactions of CTGF with stellate cells in hopes of developing targeted CTGF antagonists.29 In activated stellate cells, CTGF production is driven by constitutive mechanisms as well as a TGF-β−dependent pathway. Specific receptors on the stellate cell surface including integrins mediate CTGF effects including proliferation and matrix production. These findings encourage the use of specific antagonists to block only those effects of CTGF mediated by stellate cells while avoiding unwanted collateral activity. CTGF-overexpressing transgenic mice show enhanced susceptibility to developing liver fibrosis and are being used to define mechanisms of CTGF action in vivo.

The similarity of stellate cells and adipocytes has intrigued investigators for many years. Their many common features include storing fat, being responsive to adipogenic mediators, and being regulated by similar signaling molecules including PPARγ.30Hidekazu Tsukamoto (University of Southern California) highlighted quiescent stellate cell/adipocyte shared regulatory pathways, including Wnt signaling.31 Dr. Tsukamoto has also explored the paradox that fat is beneficial to the function of stellate cells but damaging to hepatocytes, demonstrating that activated stellate cells appear to be relatively insulin resistant.

Alterations in T-cell subsets and activity may regulate stellate cell activity,32 but the bivalent Th1-versus-Th2 paradigm in immunologic responses is now viewed as overly simplistic, with many more than two T-cell subsets underlying a highly complex, orchestrated response. Studies of schistosomiasis by Thomas Wynn (National Institute of Allergy and Infectious Disease) provide an important paradigm for how these intersecting pathways may regulate fibrosis.33 In animal models, interleukin-13 has emerged as a key mediator because it increases TGFβ1 and MMP expression by macrophages, whereas interleukin-4 has a limited role. Only one study has investigated other liver diseases; it examined the activity of IL-13 in cultured stellate cells.34

Although T-cell subsets are implicated in liver fibrosis, the composition of these subsets is distinct from those of circulating blood. Wajahat Mehal (Yale University) emphasized the distinct characteristics of hepatic T cells, including a high percentage with activation markers and a high basal apoptotic rate. In addition, a large fraction of NK and NK-T cells also characterizes the hepatic immune system, and NK cells may be important for killing stellate cells in liver injury models,35, 36 such that NK cell depletion or inactivation may augment fibrosis. He further suggested that pathways inhibited by HCV to counteract the antiviral innate immune system may interfere with mechanisms intended to limit liver fibrosis. One example is the ligation of CD81 by the HCV E2 protein, which could inhibit NK cell function and enhance survival of activated stellate cells.

The B lymphocyte is also implicated in hepatic fibrosis. Studies by Tatiana Novobrantseva and colleagues (performed at Biogen Idec) underscore the possible role of this cell type, which makes up as much as 50% of the entire lymphocyte pool in the liver.37 Hepatic B cells most closely resemble those of the spleen based on their cell surface marker phenotype and are derived from bone marrow. Mice genetically deficient in B cells (JH−/− strain) have the same extent of injury following acute administration of CCl4 but clear it more rapidly, raising the possibility that B cells could interact with fibrolytic pathways in resolution of liver fibrosis; this pathway is independent of antibody function.37

Antifibrotic Therapy

  1. Top of page
  2. Abstract
  3. Clinical Aspects of Fibrosis
  4. Basic Science Advances in Fibrosis
  5. Antifibrotic Therapy
  6. Cell-Based Therapy
  7. Concluding Remarks
  8. Acknowledgements
  9. REFERENCES

At the 2000 conference, antifibrotic therapy was on the horizon, and there was consensus on some targets, particularly TGFβ. The pharmaceutical industry was beginning to pay attention, given the potentially enormous market for a safe and effective antifibrotic treatment. Acknowledged challenges included the slow pace at which fibrosis proceeds in most diseases, with the prospect of long trials with large numbers of patients and consequently high expense. Another area of uncertainty was the attitude of regulatory agencies toward antifibrotic drugs. Would “hard” endpoints (e.g., transplant, death) be required as proof of efficacy, or would “surrogate” endpoints (e.g., biopsy evidence of diminished fibrosis, improvement of a clinical disease index, or reduced portal hypertension) be acceptable? Also, what would be the regulatory stance toward a drug that has no demonstrable effect when used alone but is effective in combination with another drug—as with ribavirin for hepatitis C treatment? These were some of the questions on the agenda for the 2006 meeting.

Emerging Targets for Antifibrotic Treatment.

The renin-angiotensin system (RAS) regulates not only aldosterone secretion and vasoconstriction but also fibrogenesis. The RAS is up-regulated in rats with liver injury, and inhibitors of angiotensin-converting enzyme (ACE) or blockers of the angiotensin receptor (AR) reduce experimental hepatic fibrosis. Conversely, infusion of angiotensin II to rats activates hepatic stellate cells in situ and augments the fibrotic response to injury (see ref.38 and references within). Ramon Bataller (Hospital Clinic, Barcelona) emphasized the rationale for human trials of anti-RAS agents. A few small trials have tested the AR blocker losartan or ACE inhibitors. In patients with nonalcoholic steatohepatitis (NASH), oral losartan produced histological improvement in 4 of 7 patients,39 similar to findings in a retrospective trial of patients who had undergone liver transplantation for HCV cirrhosis40; larger trials are underway. RAS blockers are safe for prolonged administration and should be well tolerated by patients in whom antiviral regimens for HCV are either ineffective or poorly tolerated.

The fibrogenic and growth inhibitory activities of the TGFβ family were well established at the time of the 2000 conference. However, there were concerns related to the disruption of the homeostatic functions of TGFβ by TGFβ antagonists, including growth control. Against this background, Rebecca Wells (University of Pennsylvania) discussed her work on the interrelationship of matrix stiffness, stellate cell activation, and TGFβ signaling. In cultured stellate cells, integrins such as α5β1 (a fibronectin receptor) recognize matrix constituents and are mechanotransducers, transmitting information to the cell about matrix “stiffness.”41 The latter is a significant determinant of stellate cell activation in culture. Interestingly, the time course of culture activation differs for Smad2 and Smad3, two key signaling molecules downstream of the TGFβ receptor, suggesting that they subserve different functions, because overexpression of Smad2 in stellate cells reduces proliferation, whereas Smad3 enhances conversion to the activated phenotype. Thus, a selective small-molecule inhibitor of Smad3 merits exploration as a novel antifibrotic agent.

Development of Fibrosis Therapy: the Pharma Perspective.

The study of hepatic fibrosis over the past 20 years has uncovered many potential, novel therapeutic targets. Indeed, in the view of some participants at this meeting, we have enough targets. What we need now is a coherent strategy for translating the available experimental information into clinically useful products.

M. Michelle Berrey (GlaxoSmithKline) reviewed the parallel histories of drug development for human immunodeficiency virus (HIV) and HCV. Both represent areas of high unmet need requiring combination therapy, with slow progression mandating trials of long duration. In both HCV and HIV, a reduction in viral level is a key endpoint of treatment, whereas endpoints of antifibrotic treatment must be defined, preferably using noninvasive measures. At GlaxoSmithKline, a PPARγ agonist (GSK262570) is in a phase 2b trial for use as an antifibrotic. This proof-of principle study is being conducted in patients with HCV genotype 1 and Ishak scores of 2-4; the primary endpoint is reduction of α-SMA on liver biopsy, with secondary endpoints of fibrosis and steatosis on paired biopsies.

Hugo E. Gallo-Torres (U.S. Food and Drug Administration [FDA]) provided an FDA perspective on what makes a convincing case for a new antifibrotic. This perspective has been influenced somewhat by the experience with antivirals for hepatitis B and C, where suppression of viral replication (and hence, viral load) is a dynamic endpoint. However, histology is still considered the most valuable because the FDA believes histological improvement is the most direct and stringent way to evaluate a drug's effect on progression of a disease, short of measuring late and infrequent clinical events. Surrogates such as viral load, transaminases, fibrosis markers, and imaging have not yet demonstrated good correlation with histological outcomes. Beyond the results with individual endpoints, for a convincing demonstration of efficacy it is vital that all the data for a new drug—histological, biochemical, and clinical—point in the direction of the drug as being beneficial.

Surrogate endpoints may reflect the pathobiology of fibrosis but are of unknown prognostic value. Validation is needed, such as correlation with improved histology or with a change in the natural history of the disease. That said, surrogates have a place in clinical studies. For example, they may be useful in diseases that are serious or life-threatening or when an association with improved outcomes is anticipated. Surrogate markers may also refine dose-ranging (phase 2) studies. FDA encourages the use of surrogate markers as secondary endpoints in treatment trials to establish their validity and reliability. An important caveat is that a surrogate marker that proves accurate for one candidate drug may not be accurate for a drug with a different mechanism of action. Further, the surrogate marker may not predict the time frame of the disease response.

The future treatment of fibrosis in viral hepatitis may consist of multidrug regimens, in which an antifibrotic is given together with one or more antivirals. For nonviral conditions, a drug combination that modulates two aspects of fibrogenesis (e.g., stellate cell activation and fibrolysis) may find a place. When it comes to new drug entities designed to be part of combination therapies, the FDA requires a biological rationale and safety data for the individual components; evidence that each component is effective by itself is not required.

Marielle Cohard-Radice (Novartis Pharmaceuticals) reviewed current challenges in taking preclinical data to human trials. Specifically, are liver injury models in mice sufficiently robust to model human disease? Are facets of the injury response in rodents, including the immune system, comparable to those in humans? Are the pathogenic mechanisms of liver injury—ranging from oxidant stress to RAS activation—of comparable importance in both species? Answers to these questions are critical, particularly in the face of the recent apparent failures to translate preclinical findings to clinical use (for example, administration of IL-10 or interferon-gamma).

A second challenge concerns the appropriate structure for clinical trials of new antifibrotics. Many trials to date have been underpowered; some have not been placebo controlled or blinded. When histological improvement is the endpoint, biopsy size has not been standardized. Trials with histology as the primary endpoint should be powered in a way that recognizes the number of inadequate biopsies and the related error in interpretation. Current studies are attempting to address some of these issues. For example, a trial of the AR blocker irbesartan employs a double-blind placebo-controlled design with 200 subjects over 2 years (details at www.clinicaltrials.gov).

Cell-Based Therapy

  1. Top of page
  2. Abstract
  3. Clinical Aspects of Fibrosis
  4. Basic Science Advances in Fibrosis
  5. Antifibrotic Therapy
  6. Cell-Based Therapy
  7. Concluding Remarks
  8. Acknowledgements
  9. REFERENCES

There is consensus that the best antifibrotic therapy is to eliminate the cause of the injury. Refinement of cell-based therapies for fibrosis has been accelerated by the development of mouse models of acute hepatic failure. In these models, hematopoietic cells (including hematopoietic stem cells) seemed to repopulate the liver, whereas resident hepatocytes underwent necrosis or apoptosis. The hematopoietic stem cells appeared capable of transdifferentiation into hepatocytes; however, the repopulation was later explained by the transplanted cells having instead fused with hepatocytes. Holger Willenbring (University of California, San Francisco) discussed the potential use of therapeutic cell fusion, with myelomonocytic cells (such as macrophages), rather than hematopoietic stem cells, fusing with hepatocytes42 in acute liver failure models; however, fusion was very low in the absence of injury. A current goal is to render macrophages more fusogenic while targeting them to the liver. Meanwhile, the debate over fusion versus transdifferentiation continues.

It has been suggested that hepatic nonparenchymal cells have a hematopoietic origin. Recent studies by Stuart Forbes (University of Edinburgh), suggest that some fibrogenic cells in liver, which are distinct from typical stellate cells, arise as bone marrow–derived “fibrocytes.” When mice received a bone marrow transplant that was traceable (e.g., male marrow into female mice), cells from the transplanted marrow were identified in the recipient mouse liver and in fibrotic bands from humans with liver injury.43 The assimilation of marrow-derived myofibroblasts does not appear to involve cell fusion.

Concluding Remarks

  1. Top of page
  2. Abstract
  3. Clinical Aspects of Fibrosis
  4. Basic Science Advances in Fibrosis
  5. Antifibrotic Therapy
  6. Cell-Based Therapy
  7. Concluding Remarks
  8. Acknowledgements
  9. REFERENCES

A remarkable synthesis of new ideas and refined concepts emerged from the conference. Although they are enumerated in the following list, it is not intended to be all-inclusive but rather to reflect some key themes.

  • 1
    A growing number of cellular and molecular signals are involved in hepatic fibrosis. These include B cells, NK and NKT cells, adenosine, CTGF, and toll-like receptors. Their relative contribution to matrix accumulation and degradation is unclear. Despite the importance of platelets in vascular injury and repair, they have received very little notice. A recent article implicating serotonin in hepatic regeneration,44 combined with evidence of serotonin receptors on stellate cells,45 could reignite interest in this cell type.
  • 2
    The physical features and structure of the extracellular matrix, in particular its stiffness, affect the cellular biology of fibrogenesis and may help to refine its diagnosis. Biochemical correlates of this stiffness remain to be clarified. In particular, does the stiffness reflect different extents of crosslinking/reversibility and/or edema and inflammation? Does it help to explain the more rapid reversibility of fibrosis in rodents than in human beings?
  • 3
    Our appreciation of the cellular and molecular heterogeneity of fibrogenic responses continues to broaden. Multiple fibrogenic cell types represent variable sources, as well as the epithelial-mesenchymal transition in liver. Are these responses disease-specific, with implications for the treatment of fibrosis and regeneration? Studies of gene array profiles suggest that fibrogenic pathways in different etiologies are far more similar than different (D. Brenner, unpublished observations), although subtle yet important distinctions in cellular and cytokine responses may be critical for defining early interventions.
  • 4
    Combinations” is the new buzzword. This reflects growing enthusiasm for combinations of diagnostic markers, therapies, and endpoints. Driven by the example of the HIV paradigm, we are increasingly comfortable with the concept of therapeutic “cocktails.” What is less certain is whether diagnostic and therapeutic combinations will correlate with and improve clinical outcomes, respectively.
  • 5
    As before, we continue to draw lessons from other tissues and disease paradigms, but the analogies are new. These include adipose tissue, T cells, and HIV disease. More information is needed on disease-specific mechanisms of pathogenesis. Moreover, are lessons gleaned from studies of schistosomiasis relevant to parenchymal liver diseases?
  • 6
    Biopsy is not perfect, but it is still the best way to assess fibrosis, at least for now. It remains to be established whether the use of mRNA or cellular activation markers from biopsy tissue will allow more precise quantification of fibrosis or reduce the sampling variability that diminishes the biopsy's value.
  • 7
    From a therapeutic perspective, NFκB may be more realistic as a therapeutic target than previously thought. There is an increasing ability to target specific features of NFκB signaling, for example, its epigenetic regulation by modulating upstream regulators including Iκκ.46 What will the impact be of NFκB manipulation on inflammatory cells and hepatocytes, and can such approaches be targeted to stellate cells?
  • 8
    Bone marrow is a new player and may have therapeutic potential. However, the signals and cellular subsets underlying bone marrow recruitment are still obscure. Efforts to use bone marrow as therapy are exciting but also pose considerable risk until these mechanistic questions are addressed.
  • 9
    New technologies of genomics, proteomics, and glycomics are inspiring and can be applied to liver fibrosis. It remains unclear, however, whether these technologies will prove superior to standard, less high-tech methods. There is still an urgent need for novel noninvasive methods that assess dynamic, early changes in fibrogenesis and fibrolysis as indicators of antifibrotic drug efficacy. Such methods would boost enthusiasm for conducting proof-of-concept antifibrotic clinical trials.
  • 10
    Almost all the key stakeholders in the study of hepatic fibrosis are engaged in a dialogue to develop fibrosis diagnostics and therapies. However, sufficient public awareness and advocacy are still lacking. It seems timely to engage the public in this exciting new area, as well as to form a clinical trials network, in a manner similar to successful efforts in the treatment of other major diseases.
  • 11
    Clinical trials are beginning, cooperation is excellent, and progress is likely. As these trials progress, we need to continue refining trial endpoints on the basis of emerging diagnostic technologies. Those interested in drug development, particularly pharmaceutical companies, are encouraged to “stay the course” and maintain patience and persistence in attacking this problem, which will require sustained commitment in order to yield maximum progress.

Participants unanimously agreed that 6 years is too long to wait for the next Fibrosis Single Topic Conference in view of both the accelerating progress in this arena and the significant number of important topics that could not be covered. These include the role of oxidant stress in fibrosis, the participation of stellate cells in regeneration, and hepatic progenitor responses, as well as signaling molecules newly connected to fibrosis, including serotonin, cannabinoids, ADAMTS-13, integrin-linked kinase, and adipokines.

A final point to emphasize is the extraordinary collegiality and cooperation among investigators in this field. There was a sense of shared mission and friendship, which added a satisfying personal dimension for all the attendees. This constructive climate has surely contributed to the substantial successes to date and to the many more yet to come.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Clinical Aspects of Fibrosis
  4. Basic Science Advances in Fibrosis
  5. Antifibrotic Therapy
  6. Cell-Based Therapy
  7. Concluding Remarks
  8. Acknowledgements
  9. REFERENCES

The authors express their gratitude to all the speakers and poster presenters who presented their unpublished data, which led to a unique and stimulating meeting. We also gratefully acknowledge the support of the AASLD and its highly professional staff, in particular Melissa Parrish, Janeil Klett, and Eunice Chege. We have almost exclusively cited very recent references because of space limitations, but we encourage readers to seek additional information from the reference lists in the cited articles.

REFERENCES

  1. Top of page
  2. Abstract
  3. Clinical Aspects of Fibrosis
  4. Basic Science Advances in Fibrosis
  5. Antifibrotic Therapy
  6. Cell-Based Therapy
  7. Concluding Remarks
  8. Acknowledgements
  9. REFERENCES