Our colleague Dr. Greg Gores declared HEPATOLOGY as a home for liver cancer manuscripts and described how this hot area of liver disease has grown in recent years. Clearly, liver cancer has grown from a minor topic to become one of the most competitive realms in our field. Several groups are investing efforts and resources to produce results, get grant support, and publish manuscripts. This enthusiasm affects both laboratory and clinical research, with major emphasis placed on the need to translate results from the bench to the bedside. The fancy current term is “translational research” but this concept has always been present: the objective of research in biomedical sciences is to improve human health, by either preventing or treating diseases. Not surprisingly, when more activity takes place in any given field, the number of ongoing studies and proposals increases. This translates into a larger number of manuscripts written to report the data obtained in such investigations. Increased attention to the field of one's own interest is always pleasant, but this causes a tight competition for manuscripts accepted in the journal that the authors feel is best suited for publishing their results. Obviously, if HEPATOLOGY presents itself as a home for liver cancer manuscripts (and data are there to support this statement), a steadily increasing number of articles fitting in this category are being submitted for peer review. Because of the limited capacity of the journal, an increase in submissions lowers the acceptance rate (this is now around 17% for original research manuscripts submitted to HEPATOLOGY). This law is common to all competitive journals, and in order to provide fair decisions to the submitting authors, a stringent policy that is consistent and selects only those contributions that provide relevant information must be in place. The material must be both novel and interesting, because it is not enough to be novel but not interesting or interesting but not novel.
This demand for first-class content does not mean that HEPATOLOGY should only accept breakthrough data. In clinical research, such results would come only from large double-blind placebo-controlled trials assessing new agents for any disease. Such robust information is key to define standards of care and prepare practice guidelines, but major developments in understanding a disease or treating it may also be derived from expert mining of databases, case-controlled analysis, or prospective cohort investigations. The latter is easily exposed by the seminal Mazzaferro manuscript about liver transplantation for small hepatocellular carcinomas (HCCs).1 The article changed practice on a worldwide basis and is one of the most frequently quoted articles in the HCC field. Hence, major information with clinical impact will still come from any sort of optimally conducted investigation. Indeed, novel diagnostic and therapeutic approaches have to be tested at different phases of development. There is a need to run pilot investigations prior to large-scale assessments and not infrequently, innovation in concepts and understanding of a disease comes from small-scale, investigator-initiated studies. Because several medical specialties converge in the field of liver cancer, it is clear that in some instances the information provided will fit better into a highly specialized journal. However, if the data are novel and interesting, and may influence patient management, the likelihood for acceptance in HEPATOLOGY is much increased. It is worth noting that confirmation of novel insights that influence patient management is also important. Not everything is tested within a randomized study, and prior to fully accepting any new diagnostic or therapeutic strategy, repeated confirmation by additional investigators is always demanded (again, the Mazzaferro criteria for liver transplantation becomes a brilliant example). Nevertheless, the request for well-designed and controlled studies with the least risk of bias possible should not be neglected. The controlled experimental approach is fully established in bench research and it should be equally in place in clinical investigation. Guidelines have been developed to report data from diagnostic investigations (www.cancerdiagnosis.nci.nih.gov/assessment/progress/remark.htm), clinical trials (www.consort-statement.org), and meta-analyses (www.prisma-statement.org/), among other study types, and such homogenization of effort will translate into improved strength of any clinical investigation. Laboratory studies are usually performed under very controlled conditions, because this is well accepted to produce valid experiments with relevant messages.
Clearly, not all investigations result in a major new concept or changes in our understanding of the biologic mechanisms of a disease or its therapy. In some instances, the data in cells or experimental models are merely one of the bricks that allow us to fortify concepts and data, even if the results are negative or deny the initial hypothesis. Ultimately, one step after the other, a relevant breakthrough develops and, not infrequently, this is achieved by another group of researchers and perhaps even a number of years later. Hence, what defines a relevant study deserving acceptance?
The main drivers are the research question leading to the study proposal and the methodology. The question faced has to be relevant and timely, either as a novel proposal or as a needed validation of prior findings. After this first step, the key points in any investigation are the methods applied and to what extent they are adequate to properly answer the questions that have been raised. It is common to divide research into basic and clinical, but the division is better done between good and poor research; this distinction is defined by the quality of methods applied. Rejection of experimental manuscripts because of inadequate/nonvalidated methodology is not discussed, but frequently, the request for robust methodology in clinical research is not well accepted. The current methodology to follow for HCC research has been extensively detailed in a recent document resulting from an American Association for the Study of Liver Disease Endpoints Conference.2 Readers are referred to this article for further information. This commentary will not discuss the different measures that must be taken into account to ensure an unbiased outcome in any clinical investigation. Instead, it emphasizes that one of the major endpoints of clinical research, and more specifically in the field of liver cancer, is the need to provide tools for clinical decision-making. The central part of physician life involves diagnosis, staging, prognostic prediction, and treatment recommendations. In any of these aspects, it is important to distinguish between probabilities leading to predict/suspect HCC or guess outcome according to any specific profile, and reliable criteria to confirm the diagnosis and recommend therapy according to evidence-based data. In the era of genomic profiling and search for biomarkers for prognostic prediction and tailored treatment, a large proportion of investigations seek to establish a statistically significant correlation between any molecular characteristic and any relevant clinical aspect. Unfortunately, in most instances, the data offered by the investigations are relevant, but indeed represent small pieces of a big puzzle that will take a long time to consolidate. Some of the data described may have reached statistical significance, but the lack of any mechanistic insight may prevent us from distinguishing valuable input from a random association without major relevance. As such, the initial clinical impact will be limited and seldom will the data help in changing the current management of the patients, but merely pave the road for future research. Hence, statistical significance has a value, but by itself does not imply clinical significance or important knowledge.
To properly balance all the parameters leading to acceptance or rejection, all editorial boards rely on the input of expert reviewers who generously contribute time to maintain the peer review system. Reliable and timely feedback is instrumental for a fair management of manuscripts. Indeed, the whole review process is based on these anonymous and frequently poorly rewarded individuals. Their task does not end when the review of a manuscript is forwarded. Manuscripts may be rejected in an initial journal and be submitted thereafter to a second or third editorial venue, with or without incorporation of the criticisms raised at prior reviews. If ultimately published, the expert reviewers that evaluated the manuscript in prior versions still can exert their responsibility by detecting scientific misconduct. This is not a frequent scenario, but active positioning by reviewers has recently been the trigger leading to the retraction of some positive therapeutic trials that were published after their design was rewritten after the fact to gain more scientific strength. Again, expert review and control of a specific field of interest is the basis for the success of the current editorial strategy to select the optimal contributions among all those submitted for consideration.
We hope all these comments may help the reader understand the editorial process within our scientific journal and the final fate of a submitted manuscript. Knowing that only one among six submissions will be ultimately accepted makes authors of accepted articles proud, but will not diminish the disappointment of those receiving a rejection decision. Nevertheless, it is worth stating that we are all authors of manuscripts and we all suffer the same competitive scrutiny during our professional life. Surely, being on both sides of the thin red line primes our commitment to delivering a balanced and fair decision-making process targeting the identification of novel and interesting data.