Surveillance in HCC: Making the Most of What We Have Today

Abstract

Surveillance corresponds to the systematic and repeated action of a screening test during the time with the goal of improving survival [1]. In the hepatocellular carcinoma (HCC) realm, surveillance aims to reduce the risk of cancer-related death through the detection and treatment of HCC at an early stage. However, to recommend or not recommend surveillance in a specific population, it is crucial to integrate the incidence of a specific cancer together with the careful consideration of the competing risks for death and cost-efficacy analysis. As an example, for patients with decompensated liver cirrhosis not candidates for liver transplantation and untreatable for HCC due to liver function or comorbidities, the benefit of detecting an HCC vanishes since survival is dismal due to non-HCC-liver-related events (overdiagnosis) [2]. While the future of HCC surveillance probably leans toward personalised approaches (i.e., with the integration of new biomarkers, and with more sensitive techniques such as magnetic resonance [MR]), bi-annual abdominal ultrasound (US) with or without alpha-fetoprotein (AFP) [3] remains the cornerstone of current practice. Despite the strong recommendation of international guidelines [4-6], surveillance is underused, and this may depend on both physicians (i.e., lower rates for primary care doctors) and patients (low adherence) issues [7, 8]. Therefore, improving the training of both doctors and patients is a key objective in strategies aimed at enhancing surveillance adherence. A variety of approaches have been explored, such as the education of primary care physicians, nurse-led programmes, mailed outreach strategy, and EMR-led best practice alerts [9-12]. The results are heterogeneous but can achieve interesting results in nurse-led programmes, with 53% up to 80%–90% of adherence. However, these results should be confirmed in large-scale populations, and the availability of expert and dedicated nurses should be encouraged. Finally, it is to be noted that a widely expert opinion suggests that US surveillance should be performed by physicians with extensive expertise in liver US.

In the study by Brahmania et al. [13] in Liver International, the authors performed a retrospective study aiming to evaluate the impact of a region-wide automated recall program on adherence to HCC surveillance Specifically, in 2013 in Calgary (Canada) a diagnostic-image (DI) provider created an automated protocol-based surveillance strategy based on the software used for a breast cancer surveillance program using mammography for patients eligible for HCC screening.

A healthcare provider (gastroenterologist, hepatologist or primary care) was allowed to enrol patients in the program by submitting a completed one-page requisition with demographic characteristics, reason for screening, and the presence or absence of liver cirrhosis. Patients underwent biannual US, and if unreachable by the DI team twice, two different letters from both patient and physician were sent. The primary aim of the study was the retention rate to the surveillance program instead of HCC detection or HCC-related deaths. Here, acceptable surveillance was defined as at least one US within 1 year. If the authors should be frankly congratulated for putting in place such a program, the definition of acceptable surveillance seems suboptimal. The appropriate frequency of US was indeed proven years ago to be every 6 months. Specifically, a 3-month interval does not improve the detection of HCC > 1 cm, and an annual interval is associated with lower survival and HCC detection rates in comparison to the biannual interval. Therefore, assessing the legitimacy of the current program is challenging. In parallel, it is also worth noting that the definition of adherence, as well as the methods to measure it, are heterogeneous across the study in the literature. However, a pragmatic approach might be to calculate the number of US performed over the theoretical US for a specific patient in a determined timeframe. Brahmania et al. [13] included in the program a total of 7269 patients between 2013 and 2022, and the most common aetiology was hepatitis B virus (51%) and only 37% of the patients had liver cirrhosis. The fact that in most of the cases the indication of surveillance was not liver cirrhosis is surprising and cannot be explained only by the high proportion of hepatitis B virus. Of the whole cohort, 51.8% of the patients were considered retained in the surveillance program. This proportion may seem high in comparison to the pooled proportion of patients receiving adequate surveillance of 24%, with the lowest rate in the USA (17.8%) and highest in Europe (43.2%) [8]. Nonetheless, none of these numbers are comparable due to the non-homogeneous definition of adherence across the studies. A recent retrospective study in Spain revealed that 84% of the patients with known cirrhosis were diagnosed with HCC under a surveillance program [14]. In this sense, it is interesting to note a positive trend concerning the rate of HCC detected under surveillance since the previous one was around 47% [15]. In the study of Brahmania et al. [13] after a median follow-up of 1.89 years (IQR: 1.0–4.8), the median rate of US per year was 1.82 (IQR: 1.15–2.08). Therefore, the number of US seems high but should be relativised according to the short follow-up for a surveillance study. In addition, the same authors recognise some relevant limitations, such as the lack of prospectively recorded data on AFP, response to therapy in patients with viral aetiology, BMI, comorbidities, degree of portal hypertension, incidence of HCC, and overall survival.

The realm of surveillance in HCC is an absolutely evolving field of research for a variety of reasons. First, the change in the epidemiological landscape with the increase in patients with hepatitis C virus-cured and Steatotic Liver Disease (SLD) patients (with or without alcohol) mandates the careful assessment of HCC incidence in these specific groups. This information is needed to establish the benefit of surveillance programs for these patients. When combined with a revisited life expectancy (negative or positive depending on the groups), the competing risk of death, and the increased survival outcomes of HCC patients, it might mandate adjusting the current cut-offs and indications for surveillance. As an example, patients with SLD are at higher risk of death due to cardiovascular events and extra-hepatic cancers, and this could dilute the benefit of HCC screening in patients at risk. Second, the efficacy of the current tools has been questioned in terms of sensitivity. However, a recent well-designed randomised clinical trial reported a sensitivity of US alone of 77% for early detection [16], suggesting that when properly performed and registered, this technique is not bad at all. Finally, the available tools for HCC surveillance are underused independently of the definition.

However, it should be remarked that the optimisation of surveillance programmes faces several interrelated challenges: sustaining adequate adherence rates in established target populations, evaluating the potential expansion to patients with SLD, appropriately selecting candidates for advanced diagnostics (biomarkers and MR imaging), accurately identifying high-risk individuals, and delivering personalised surveillance approaches. Balancing these competing demands while ensuring programme sustainability remains a complex endeavour in current clinical practice (Figure 1).

While awaiting the development of an effective adjuvant therapy for HCC [17, 18], surveillance remains the cornerstone strategy for improving patient survival and continues to be the best method for reducing HCC-related mortality in high-risk populations.

Personalised surveillance approaches according to the individual risk together with techniques with higher sensitivity are a hopeful wish for the future. In the meanwhile, it is crucial to “Make the Most of What We Have Today” and all the programs enhancing the education and/or adherence are of utmost value. Of note, to properly measure if a strategy is effective, their evaluations should be appropriate as well. They should consider the benefit in terms of survival or HCC detection rate in a proper population or surveillance adherence in an adequate manner.

Interpretation of data and drafting of the manuscript (all authors); critical revision of the manuscript for important intellectual content (all authors). All authors approve the final version of the manuscript.

Marco Sanduzzi-Zamparelli received speaker fees from Bayer and AstraZeneca and travel grants from Bayer, BTG, Eisai, and Roche; Giuseppe Cabibbo participated in an advisory board and received speaker fees from Bayer, Eisai, Ipsen, AstraZeneca, MSD, Roche, and Gilead.