Clinical trials and their translation in hepatology: Past, present, and future

Since starting my life as a hepatologist in 1968, I have witnessed marked improvements in the design, conduct, and analysis of clinical trials—thanks to such pioneers as David Sackett and Gordon Gyatt, just two of the North American scientists devoted to studying clinical epidemiology and evidence-based medicine. The Cochrane Collaboration, first established in the United Kingdom, now with centers worldwide, has focused on systematic reviews of published clinical trials.1 This was a timely development, because randomized, controlled trials (RCTs) designed to evaluate new therapeutic agents for liver disease have multiplied, particularly over the last 15 years (Fig. 1). RCTs are needed to evaluate the efficacy of new drugs, procedures, dietary modifications, and so on. This process remains incomplete unless translated to healthcare providers. Published RCTs in hepatology. Total versus trials for ascites and for CHC (PubMed Medline). As a busy intern on a ward caring for 30 patients with liver disease, my armamentarium consisted of the following: vasopressin for “presumed” bleeding varices, lasix and aldactone for fluid retention, corticosteroids and azathioprine for autoimmune hepatitis, neomycin for hepatic encephalopathy, and a very small selection of antibiotics for sepsis. The only radiologic tests available were a flat plate of the abdomen, angiography, and splenic venography! There were no endoscopic procedures, aside from rigid sigmoidoscopy! The discovery of, and then testing for, hepatitis B2 and C3 identified many clinically silent, yet chronically infected, individuals. Some had received another diagnosis for their “hepatitis.” The scientists whose identification of hepatitis B and C revolutionalized hepatology were honored with a Nobel prize and the Lasker award, respectively. Their discoveries changed the focus for many scientists and put the pharmaceutical industry into “top gear.” Clinical trials in hepatitis B and C became big business (see Fig. 2) to the virtual exclusion of “investigator-initiated” trials in viral hepatitis. But, dilemmas remain, including the following: Who has chronic viral hepatitis?; Is it cost-effective to treat all infected?; and how do we pay for our successes? Knowing you have a potentially fatal condition, curable only at prohibitive cost, is, to me, unacceptable! Therapeutics for almost all liver diseases are entering the era of “designer” drugs, which interfere with specific targets, for example, the first antivirals targeted at the hepatitis C virus (HCV) itself or at cellular components essential to complete viral replication4 (Table 1). RCTs for CHB and CHC (1992-2008 in PubMed Medline). Many academics have adopted new research interests within hepatology, such as complex trial analyses (including meta-analyses), cost-effectiveness studies, quality analysis, and the development of management guidelines—all essential to translate the indications for new therapies to clinicians. The “takeover” by the pharmaceutical industry has translated new knowledge of antivirals to front-line physicians. However, there is no budget for the translation of investigator-initiated studies5 (mostly in liver failure)—hence the continued accumulation of such patients in the emergency room. Combining the results of RCTs provides the “power” to estimate the overall effect (i.e., good or bad). Because not all trials are conducted to the same standard, the inclusion of poorly designed or conducted studies may lead to misinterpretation of the results.6, 7 To translate specific findings to our patients, careful scrutiny of all factors relevant to patient outcome must be reported, as must adherence to recruitment criteria and/or results of the screening log (i.e., number approached of the total and proportion of those approached who consented—two items commonly found missing, but much needed to asses the generalizability of a study). Further analysis (e.g., race, percentage of those with symptomatic versus asymptomatic disease at baseline, severity of background liver disease, age, sex, comorbidities, outcome of previous treatments, drug interactions, and so on) is needed to relate the outcome to our patient population. When this information is omitted, in part because of publishers' length limitations, the trial data are inaccurately presented. Reexamination of trial data is possible now that all clinical trials must be registered online (www.clinical trials.gov), and all data generated are kept for 25 years after the study's completion. Great advances in our understanding of the treatment of liver disease have taken place during my academic career, in part because the science of designing and executing clinical trials has received great attention. A major “hidden” confounder of trials remains so long as there is no formal “reporting” system for publication of “negative” trials. Responsibility for this gross oversight—with potential to compromise patient safety—lies with both journals and investigators. The risk of subsequent patients receiving unhelpful, perhaps even toxic, therapies could easily be prevented by a requirement that all trial results be summarized—linked to the mandatory registration website. Healthcare professionals, whose job is to promote good health and reduce illness, need access to the “complete story.” All the data are needed, including safety and conflicts of interest, and clinically relevant measures of efficacy, both while on and following withdrawal from therapy. Results must be clinically meaningful, reproducible, and understandable. Only then can organizations such as the Cochrane Collaboration reliably estimate the benefit of new therapies. Approval of new drugs is a huge commitment for government agencies. Because many are for the same disease, “cost-effectiveness analysis” has become a “business of its own” used not only by treating physicians, but also health-insurance agencies and stockbrokers. In Britain, where government promises “universal” access to treatment, the National Institute for Health and Clinical Excellence was established partly to assess cost-effectiveness of new treatments and technologies and unify access across all health districts.8 The downside of limiting access to agents not shown to be cost-effective is their unavailability to specific individuals anxious for a reprieve from fatal illness (e.g., sorafenib), if only for a few months.9 Thereby, a conflict arises between what is cost-effective for a population and that which is not cost-effective, but still has seemingly tangible benefits for an individual. Though some of these expensive medications may make only minor differences in life expectancy, with time, these small, incremental advances may eventually lead to dramatic improvements in outcome (e.g., treatments for breast cancer). This review focuses first on some mistakes and/or misinterpretations of clinical trials since the 1970s that may have interfered with the production of reliable data (for the most part, from my own experiences). An RTC is the most rigorous assessment of a new agent's therapeutic effect. Randomization is now done in blocks of 4, 6, 10, and so on, depending on expected recruit numbers. When possible, both patient and investigator should be blinded to the randomization. In terms of clinical-trial expertise, I was very naïve in 1968, when part of my job was to monitor patients in follow-up in the 5-year trial of azathioprine for primary biliary cirrhosis (PBC). Neither single nor double blinding had been considered in the trial design, and no formal patient-evaluation process had been outlined! Fortunately, there were only 45 recruits—we had not calculated the sample size needed to show a difference in outcome (i.e., death) at 5 years! All documentation was with pen and paper. Every result had to be accurately transferred from many different sheets of paper, thereby limiting the reliability of the reporting—desktop computers did not exist back then. Twenty years later, I recognized that for our RCT to be truly double blind, the unpleasant taste of ursodeoxycholic acid (UDCA) should be “matched” with an unpleasant tasting, similarly shaped, and colored placebo, to prevent patients unblinding themselves when they “compared notes” at support-group meetings. Failure to “blind” a study effectively becomes particularly relevant and, possibly, very difficult, if the side effects of the treatment under study are marked and if the primary measure of outcome is “soft,” such as patients with PBC who want very much to have their fatigue or pruritus reduced. Only with close examination of the “strange” results of our crossover trial-design to evaluate the effect of ondansetron on fatigue in PBC did we “figure out” that the results were invalidated both by patient anticipation of benefit and the near-universal side effects of odansetron! In a more recently published trial in patients with autoimmune hepatitis (AIH), one outcome marker was the combined biochemical response and the change in sense of “well-being,” the latter of which may have been compromised by different dosing regimens for the two arms of the study.10 Patients with PBC tend to be more “informed” than most. My failure to take this into account taught me that before designing, writing, and one first of the My RCT of for because the patients were or were and few the In a more study of the effect of for chronic hepatitis C on system I that if measures of outcome take to or it is particularly to to for that small in can even the we are with some of the mistakes in when were not and/or may have been to in clinical The process of to be in a and for the But, in to me to Patients must to so many it is to that patient in their The process could be as particularly if may be that only to access to the need to side effect may from entering the thereby the benefits from the therapy. for this should not be just at the of of are the downside of what is very in the recruitment and of clinical This can be taken if we take and into need a of the process to that patients not only have to but that they are to and that autoimmune liver chronic hepatitis chronic hepatitis endoscopic hepatitis B hepatitis B hepatitis C for liver National Institute of and primary biliary primary randomized, controlled ursodeoxycholic in clinical trials may be close to such as a or can that is approached for Some may have knowledge of healthcare and/or are not in the of for were to recruitment of to a study on hepatitis that study be to the population in major But, the not understanding the process for may promote thereby limiting the study's There are to (e.g., and drug yet all can I that is not an under studies the academic a long efficacy in But, this is not to because they have available to a The to the of put at an than if they in a the study's yet may be in the of the their and their to clinically to be and following of the I of few that have been more than the The data from this study that outcome of chronic hepatitis B a to both hepatitis B and from is that individuals were not unless over this study not to individuals. to and the of the clinical of trials of the National Institute of and at the National of Health have clinical specific of liver disease across North (Fig. Their studies are designed and by academic and the of and/or their patients is studies are to by published and are designed to our to a of clinical that to a disease can be by as as In previous to treatment for of or more of was with both viral and a of disease to studies for one study to to of the one disease (e.g., and all of which are clinically but to the but factors that may patient and and/or The studies from the C to in clinically relevant from liver disease and clinical clinical findings and results were into one of the first combined measures of the first designed to outcome after for another combined measure of liver the for also to the of after is now to assess the for a in patients with liver A which may be even more is much to outcome in patients with we most in in In the most were only their liver disease was example, of patients with to the trials of in the all had liver over a it was possible to that those to had a than those who received years later, patients with with liver some are this diagnosis even they are This the the before we treat all we need to if the of asymptomatic is as as for those with symptomatic An when put to the in the 1970s by at because no patient to the There are data have all the of to that treatment may not be to those with asymptomatic there are many in that were not available before the endoscopic and to but a when used are very in or are to over because patient at be we have no good radiologic of outcome in primary and we on that may may both or of liver disease, including complete and liver as as the tests of such as and But, they have their (e.g., a marker for one disease may not be for another disease the same example, of in patients with PBC with is a reliable marker of good but this was found not to be so for patients with with analysis of the from in this latter trial that it was the in the that with as a reliable of outcome is but many drug review at in North As we the of liver with its of in terms of most because of size of on the we need truly of liver disease The of hepatic (e.g., or a of results (e.g., and so The and as are to be more reliable the more the but it is they reliably measure its or In to drug efficacy, drug is a A to all drug side effects from minor to major is particularly for drugs of possible or must to be even after because all are recognized many have received the the data should also analysis of individuals (e.g., effects of drug effect on and with such as may one to be the of a is it may the of drug as with from a trial to the of of response to treatment was in the with response I such is the of the not even as reliable to to or not to an with Patients with hepatitis with a result of as more recently a of or to treatment with corticosteroids with or after the first of Patients with ascites unless is if they are with they do not which is shown to In the same I primary with in patients with cirrhosis and fluid has been or Patients with hepatic or to be of studies as back as and as recently as benefit from a The the the these patients are to the liver they The benefit of the (e.g., which is not and only development of is but the treatment is the of continued to be is to be a patient for management of ascites who has had the need to in their In this era of of (i.e., of such a may be all that is for the radiologic of a the effect of management first to be to assess of ascites is possible with few side of I become one of the most of clinical research in hepatology. in this are to including what the are for an and what the effect of such as which itself My in this of is of the outcome of endoscopic to be the taken many and trials that endoscopic treatment, perhaps more with risk (e.g., death) than in My is, who is these (i.e., patient or and on what it that in the management of liver failure is is it be that most trials in this are investigator and their findings to be to the front-line be an are now at the of a very era of that the viral or the of viral The first two agents recently are and which both in patients and some with of with efficacy for is on There are and to both of these The trial of the registration trials were very so only which become for a very the both are and new management both of treating and their become more as access to these two agents becomes available to be in patients new But, both and are what patients of two antivirals to patients with CHC and one with to those with that an drug may be on the I do not it be long before we can with CHC that their can be at a studies the new are to show it be to treat all with CHC or just those with disease at diagnosis or for viral of the risk of liver failure in a the risk of but is for the first 5 years after an is also with a in of and this benefit of viral may relate to the of (Fig. a liver while still The of of the effect of the for hepatitis C may viral so that if just before they may prevent treatment for CHC are that it should be before the of cirrhosis at the very to all before the of liver This not the screening of individuals. for and for CHB may lead to of hepatitis B by of hepatic To of of those with CHB subsequent to treatment with is but not to those with A and and the for those who on the is with the of patients with those with most in the are to of the The benefit of the is the that no drug so been in of this However, patients in this trial had the of to if complete viral was not by so we do not the risk of drug on The of the trials of did not for complete follow-up of some patients after the first when those with or were from the and the of drug could not be reliably A subsequent study a very of at in the trials of both the antivirals for CHB should have been by the they for their need with both or no risk of drug and with efficacy all hepatitis B of in CHB after treatment is should be the number one viral of and when to those with liver disease on need that viral treatment or the risk of studies be very to do now that so many more antivirals have the with hepatitis B and/or C is the risk for whose are both in the United and in 5 and A study of screening versus be The data we have are from who were or who with But, unless chronic with hepatitis B or C is identified individuals as a of their must and screening of in the between from to and from to were for and percentage change in for from those an 5 all are are from the National with of hepatic by liver disease and liver disease may be the most for therapies. As the cellular for the of become treatment with specific could be first access to both and reliable screening to for such therapies. few the risk of hepatic in both and who are The outcome of PBC to have over the last a result of a of diagnosis and treatment with we do not the severity of disease at diagnosis of PBC or a in the need for liver for PBC is this is not the for a disease that The number of RCTs for those with biliary disease over the last years that has been to chronic biliary disease (see Fig. we have to those we to have so one can the of radiologic screening with for those at RCTs for biliary cirrhosis (PubMed Medline). The of PBC that a specific is these findings may the development of agents that interfere with the for this data on some specific to a new focus for be targeted There has been a in healthcare and in as a result of the of the are we need to the of the and of to a of be The focus of most hepatology since the for good been liver this was how and after many years of to have the of hepatology become recognized as a with its own in the United As a I very few of my patients a liver with years diagnosis and the of treatments, I to are for this change (e.g., PBC and chronic viral But, now that we have such treatments for some of liver disease, is it not to the cost-effectiveness of liver versus the of curable or liver the should a and its should no be by the need for liver the and I to be done of the need for their need to our both in healthcare and in the of across the including primary as as and hepatology. a to this or be by my chronic viral hepatitis is the in the of disease in the in where of the of were In most of the remains in North it is now we all that the to these two of chronic liver disease should be and Many in have been with we need to these two very we should how the of the was The from in the United and must all and healthcare to take the measures to reduce the risk of of the development of The of background liver disease remains and is a new research focus in medicine. is by the of new and in the need to we still patients with liver disease The knowledge we have as to be translated to many academic The major knowledge to be identification of individuals at risk of liver disease, the clinically of and, when liver failure a of for how this could be prevented or have shown that our are the result of of In a of and in primary only recognized that a was a marker of cirrhosis in a patient with A review was recently published of the quality for treatment in patients found to have we need to that many with cirrhosis are and their disease A new to knowledge translation was taken by the for Health in in specific I was to be to a in hepatitis C that from a very of (including are if their research are in the there is mandatory in online to their and their knowledge to their and was very to of all became in a of the hepatitis C so that across we now have in many different their research in hepatitis The hepatitis B has been available for close to 25 years and has been shown to have efficacy when at to has been shown when to in to reduce the of has this result not been to clinical in all and access to healthcare a In the it be to have the at the same as the combined for it to become both and A hepatitis C is a The of be an even to as our interests in to the liver disease is asymptomatic and may so for many we in that the we or at is should be the in and we not have a to all access to the many advances in the treatment of liver disease over the last reduce the of we an access to the translate the knowledge we have on and treatment more therapeutic to patients whose liver disease is at an reduce in the emergency with liver failure with or liver of these The of such to be in the I have found the last years of my in clinical research in hepatology to be than I the same be so for those who are just starting their This is my last to my on as I from to into a new The to all patients and who have the last years so The also to who the that the number of RCT published on PubMed

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.