This is the authors' abstract. We don't add key points for this paper.
Potential conflict of interest: Dr. El‐Serag received grants from Wako, Gilead, and Merck. See Articles On Pages 864, 875, AND 885 Hepatocellular Carcinoma Surveillance The attention to hepatocellular carcinoma (HCC) in the United States began with observations in the late 1980s of an unexpected increase in incidence, hospitalizations, and mortality in several national registries.1 HCC is currently the fastest rising cause of cancer‐related deaths in the United States. The presence of cirrhosis in most HCC cases, and the greatly elevated HCC risk among patients with cirrhosis, has led most, but not all, professional societies to recommend surveillance for HCC in patients with cirrhosis to enable the detection at an early stage where curative treatment can be applied. The main driver of the cost‐effectiveness of HCC surveillance as recommended by guidelines is the annual HCC risk (or incidence) in the at‐risk subgroups in the population. Against this backdrop, this issue of Hepatology presents three articles that tackle several aspects of HCC: (1) reliability of HCC coding in cancer registries; (2) the extent and determinants of HCC surveillance in the national Veterans Health Administration (VHA) system; and (3) patients' perspectives on the factors that influence their knowledge, attitudes, and barriers to HCC surveillance. Törner et al.2 compared three nation‐wide registries in Sweden where HCC reporting is mandatory (i.e., passive surveillance) and compared the Cancer Registry (CR), Cause of Death Registry, and the inpatient component of the National Patient Register. According to the Cancer Registry alone, the prevalence of HCC has decreased over time up to 2005 and coincided with changes in the method of diagnosis from autopsy early in the 1970s and 1980s to biopsy and noninvasive methods in the 1990s. Comparing HCC cases across the three registries, they found that 37%‐45% of HCC in Sweden were never reported to the CR. Thus, even in the “best” possible setting where health care coverage is universal and reporting cancer cases is required by law, HCC incidence and prevalence are likely underestimated, attributable to a combination of under‐reporting and the use of noninvasive methods of diagnosis. Given these findings, one may argue that HCC incidence figures coming from other publicly available databases in the United States provide conservative numbers. HCC has been increasing in the United States, and it is likely to be more frequent than currently reported or thought. One approach to overcome HCC under‐reporting in registries could be to engage health care providers and patients to practice HCC surveillance, and yet another approach is practice (or repractice) tissue‐based HCC diagnostic tests. Goldberg et al. explored factors associated with low HCC surveillance using VHA data.3 They identified 26,577 patients with cirrhosis defined by the presence of (1) inpatient or (2) outpatient International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) coding for cirrhosis, and evaluated the percentage of time up‐to‐date with surveillance (PTUDS) with an abdominal ultrasound (or contrast computed tomography, magnetic resonance imaging) scan. During follow‐up, 37.3% of patients with cirrhosis had a single abdominal ultrasound performed, and an additional 27.4% had other types of abdominal imaging. The median PTUDS was 10% (0%‐29%) and 16% (0%‐39%) using ultrasound or any imaging modality, respectively (100% PTUDS when the patient had regular imaging every 6 months). Given that procedure codes for ultrasound do not differentiate between surveillance and more common diagnostic purposes (e.g., abdominal pain), HCC surveillance ultrasound in this study is likely to be grossly overestimated. Yet, even with this overestimate, the proportion of patients with any abdominal ultrasound was only 62.7%. This misclassification hardly matters in making the point about HCC surveillance underutilization, which is clearly documented in many other studies.4 Although all of these studies were retrospective, and most were small and had methodological limitations, the totality and consistency of the evidence is overwhelming. Why is HCC surveillance low? HCC surveillance is a multistep process that requires recognition of liver disease risk factors and cirrhosis, awareness of surveillance guidelines, test ordering, as well as patient adherence with test completion, and repeating the same cycle 6 months later. The study by Goldberg et al.3 reported that the strongest predictor of HCC surveillance was the number of specialist visits (gastroenterology or infectious diseases) in the first year after cirrhosis diagnosis. Furthermore, ultrasounds that were ordered more than 3 months ahead of time were less likely to be performed. Patients with alcohol, nonalcoholic steatohepatitis, or cryptogenic cirrhosis had lower HCC surveillance rates than patients with hepatitis C virus (HCV). Finally, longer distance from the primary hospital was negatively associated with PTUDS. The likely misclassification of the purpose of abdominal imaging (surveillance vs. diagnosis) may have affected the findings about determinants of surveillance in unclear ways, and therefore one needs to regard these findings, and possibly the interventions that are based on these factors, with some skepticism. Nevertheless, most of the findings are similar to those in previous studies in and outside VHA.5 For example, Davila et al. reported similar results in 1,873 patients with HCC on cirrhosis identified in the linked Surveillance, Epidemiology, and End‐Results–Medicare claims. The likelihood of HCC surveillance increased in patients who had access to a specialist, new diagnosis of cirrhosis (within the first year), viral hepatitis (B or C), and living in urban areas or near academic institutions.7 Other studies indicated gaps in providers' knowledge as a possible reason for low HCC surveillance. Dalton‐Fitzgerald et al. surveyed 131 primary care physicians (PCPs) in Dallas County, Texas, and reported that physicians thought that clinical examination was enough for HCC surveillance.8 McGowan et al. found similar results in another primary care practice setting in North Carolina,9 where the ability of PCPs to identify at least one effective HCC therapy was independently associated with twice the odds of recommending HCC surveillance.9 What do patients think of HCC surveillance? In this issue, Farvadin et al. examined factors affecting surveillance among patients at high risk for HCC.11 Of 753 patients identified with cirrhosis in Dallas County, 541 (71.8%) agreed to participate in a survey using a questionnaire that explored patients' knowledge, barriers, and attitudes toward HCC surveillance. The findings showed major discord from the known evidence; for example, 48.6% thought that eating a healthy diet precluded need for HCC surveillance and 34% thought that HCC surveillance was not necessary if they had a normal physical exam and/or absence of symptoms. Scheduling, costs and transportation impacted 2‐3 of 10 patients. Although this is the largest behavioral survey of patients at HCC risk, it is limited by different methods of interview (384 patients completed the survey in clinic and 157 by telephone), 28.2% nonresponse, inherent subjectivity of self‐report, and limited generalizability to other patient groups. For example, in a previous patient survey conducted at the University of Michigan by the same investigators,12 knowledge of HCC surveillance was much higher than that in the current study and was positively associated with white race and college education. Similarly, Chen et al. in Taiwan surveyed 400 patients with HCV or hepatitis B virus13 and reported that household income was significantly and positively correlated with perceived HCC susceptibility, benefits, cues to action, and knowledge. There are two main nonmutually exclusive options to improve HCC surveillance: better implementation of existing surveillance methods and new and improved tests. The first option is to make the best of what we have (liver ultrasound and alpha‐fetoprotein) with interventions targeted at patients, providers, and systems to enhance knowledge of HCC risk factors and surveillance and treatment options, and improving coordination of care across different disciplines. For example, integration of decision support tools, such as clinical reminders with electronic health records, may help increasing provider knowledge about HCC surveillance. In one study. the use of reminders increased HCC surveillance rates by 51%, with an increase in HCC diagnoses.10 Whereas technical innovations in implementing current surveillance tests is important to maximize the use of what we have, HCC surveillance needs a spark in the methods and tests of surveillance, and a boost to the evidence of benefit (relative to harm and cost) for HCC surveillance. A systematic review concluded that there is very‐low‐strength evidence for the effects of HCC screening on mortality in patients with chronic liver disease.4 The existing observational data provide some evidence linking HCC surveillance with improved outcomes,14 but no such high‐quality evidence from randomized trials in patients with cirrhosis. High‐quality data, including from pragmatic clinical trials examining interventions on HCC surveillance and outcomes, as well as improving surveillance methods (new and novel biomarkers and imaging), are urgently needed.16 Otherwise, we may continue with the status quo of irregular and highly variable HCC surveillance. The burden of HCC is high, and will likely increase in the United States, and it may be higher than reported. Active surveillance for HCC is grossly underutilized among patients with cirrhosis; and this underutilization probably further exacerbates the under‐reporting (in addition to missing out on whatever benefits surveillance may have on patient outcomes). Studies have identified several actionable provider‐ and patient‐related factors that can be targeted by several interventions. However, we believe that the relative lethargy in this practice is attributed, to no small part, to lingering doubts about efficacy/harm/cost ratio of surveillance and skepticism about the current HCC surveillance tests and methods. We should make the best use of what we have and work to improve the methods and evidence base for HCC surveillance.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Hernáez et al. (2016) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: