“Pardon the Interruption”: Comments on the International Liver Cancer Association Fourth Annual Conference, 2010. Is There Really No Role for Liver Biopsy Interpretation in Liver Cancer? σ

HCC, hepatocellular carcinoma; ILCA, International Liver Cancer Association. A s I write my final Associate Editor Commentary, I would first like to acknowledge the distinct privilege and honor it has been to be a member of the Associate Editor Board for the “Lindor” years of HEPATOLOGY. I have enjoyed my service to the American Association for the Study of Liver Diseases, as well as both the pathology and liver disease communities. Frankly, I will not, however, miss the difficult choices that are necessary in this position. Writing this Commentary, however, largely by the encouragement and invitation of coauthor Gregory Gores, has not been one of the difficult choices of my career! For some background: The International Liver Cancer Association (ILCA) held its fourth annual conference in September in Montreal, Canada. As an invited member for the first meeting of the society, and with a long-standing interest in the pathology of liver cancer, I had only missed one other meeting to date. I was looking forward to presenting a poster that might stir discussion and interest in the outcome of HCC after transplant with certain histopathologic/immunohistochemical findings that were previously reported to be poor prognosticators, but that in our series had not correlated with adverse outcome. However, after the first day of the presentations, as excellent as they were, the message was clear: -omics, gene arrays, micro-RNAs, and biomarker assay development were “in”, and further, most appeared not to involve the “services” or expertise of prior histopathologic evaluation of the tissues on which they were based. The primary concern can be illustrated with these facts: ILCA is a multidisciplinary organization. This year, the Secretary–Treasurer reported the largest attendance of all prior meetings: 600 registrants from more than 16 countries. Of these 600 participants, nine (1.5%) were pathologists. Having never looked at these attendance records before, I am not aware when the apparent attrition of attendees from our specialty occurred, but I am concerned it will only accelerate unless the role of the liver pathologist is once again resurrected for the value provided. A major, fundamental question is whether molecular diagnostics can or even should fully supplant careful histopathologic examination of a radiographically characterized lesion in the liver. In presentations and publications related to molecular studies or “-omics” of liver cancer, the question that never seems to be clearly asked (or answered) is this: “Exactly what cancer is being studied?” International groups of pathologists have gathered, studied, and published findings over the past decade, yet it is apparent that nonpathologists are not aware of our growing knowledge of the many forms of liver cancer. The concept that there are simply two types of primary liver cancer, hepatocellular carcinoma (HCC) and cholangiocarcinoma, is no longer valid. Not all primary carcinomas that arise in cirrhosis are HCC, and as the burden of metabolic/obesity-related liver disease grows, we are increasingly aware that not all HCC develops in a background of fibrosis/cirrhosis. Within a given form of liver cancer, there may be significant tumoral inhomogeneity, some of which can be observed by light microscopy and some of which requires immunohistochemical characterization. For instance, not all HCCs are uniform in architectural pattern, cytologic features, inflammatory response, and stromal environment, and thus cannot be graded easily as well differentiated, moderately differentiated, or even poorly differentiated according to the standard Edmondson-Steiner classification. Some liver cancers are not even typical HCC, but are composed primarily of small spindled cells, are difficult to characterize as epithelial by conventional immunohistochemical markers for HCC, and best labeled simply as “primary liver cancer”, and their outcome may be considered similar to that of other undifferentiated primary carcinomas. We now recognize liver carcinomas can also be characterized as having arisen from the Hering canal (progenitor) cells (so-called cholangiolocarcinoma). These tumors have a striking histologic appearance of altered cord-like and antler-shaped structures within a dense stromal background, often reminiscent of an aberrant ductal plate, and differ in many ways from conventional cholangiocarcinoma. If these tumors and their stromal components are never carefully documented histologically, results of the sophisticated molecular -omics, micro-RNA studies will continue to be heterogenous, unfocused, and essentially irreproducible. We are, we fear, losing a valuable opportunity to correlate genotypic information with histopathologic phenotype of the plethora of cancers in these studies. In most published manuscripts on molecular signatures of HCC, the cancer phenotype presented is limited to size and number of lesions, presence of microvascular invasion, serum alpha-fetoprotein levels, and, of course, disease recurrence or survival. Detailed histopathology is commonly missing from such articles and, yet, has much to offer. An example of this lack of attention to details of histopathology at the recent ILCA meeting was the luncheon workshop for Molecular Markers of HCC. Both moderators made a point of stating early on in the meeting that “liver biopsies need to be done on all liver cancer”. Why, they asked? They went on to answer: To be able to study predictors (molecular), as well as prognosis (molecular). Also noted, however, in the seminar was the diversity of (molecular) platforms for these highly sophisticated studies, the plethora of genetic, inflammatory, metabolic pathways and data being generated, and our current need, therefore, for highly sophisticated, costly, and center-specific techniques for analysis of large numbers of cases and datasets. Yet, no discussion whatsoever occurred of the possible presence or significance of correlations or differentiating histopathologic subtypes. It seems unfortunate that the history of our discipline in hepatology is rapidly being lost to the memories of the young generation. It is worth noting the remarkable observations made purely by histochemical techniques (long predating the knowledge of immunohistochemistry) of the divergence and plasticity of the cells arising within the liver toward either malignant “hepatocellular” carcinoma or “adenocarcinoma” by one of the elder statesmen of liver pathology, Valeer Desmet, in work published only in his Ph.D. thesis, as quoted by Fausto et al.1 in the textbook, The Liver. As noted in discussions with one of the other nine attending liver pathologists at the ILCA 2010 meeting, to date, liver cancer is now the only cancer of a large organ for which only imaging characteristics but no positive tissue diagnosis is required by regulatory agencies, either for definitive therapy or experimental treatment protocols. How long this can last is unknown. Although this approach may be well-supported by current evidence-based medicine, there is still much to be learned about early stage liver cancer by careful histopathologic evaluation. Indeed, the authors of this commentary predict that histopathology and established techniques of microscopic analysis may well provide more important biomarker information and help with a personalized medicine approach to this cancer than will large, 1000-gene expression signatures. Liver biopsy takes specialist interpretation. Yes, it does. That has been shown in the literature time and again. Expertise is required in all aspects of advanced liver disease diagnosis and management, is it not? Investigators of the “-omic approaches” are themselves specialists in what they do, and indeed should consider prospectively seeking consultation with expert liver histopathologists, much as they do with biostatisticians in performing these studies. We cannot talk our patients into (another) biopsy. This discussion is tautologic. If the commitment exists for the importance of tissue acquisition prior to potentially life-threatening treatment, or enrollment into an experimental treatment protocol, request for review of a prior biopsy or discussions with patients can occur. However, without commitment from the treating physician, there certainly will be no a priori request made by a patient! Expert liver histopathology is not always available locally. Unfortunately, in many hospitals, and indeed in many centers today, this truly is a problem. In fact, the current focus in many medical practices in the United States is on large-volume, high-throughput gastrointestinal pathology; the emphasis thus remains on simple questions (Is it cancer or not?), and there is a shortage of trained pathologists focused on a career in liver diseases. This has resulted in a problem in many sites, yet those of us in the field realize there remain many opportunities for clinical and basic investigation in liver disease, and that working as a respected member of a multidisciplinary team with surgeons, hepatologists, radiologists, and basic science investigators is rewarding. For clinicians, however, without pathology analysis reports that yield precise and informative details, the request for analysis of tissue obtained by invasive means will (and should) dissipate. Clearly, pathologists focused on liver disease need to continue to work to share our enthusiasm and bring our younger colleagues into such a career. We can do so, however, only if there is a future for them in it.