DOI: https://doi.org/https://doi.org/10.57187/4469
Hepatocellular carcinoma (HCC) accounts for approximately 90% of primary liver cancer cases and a third of all cancer-related deaths globally [1, 2]. Studies have shown that in Switzerland there were 910 deaths caused by hepatocellular carcinoma in 2020 alone (age-standardised rate: 4.0 per 100,000) [3]. The primary risk factor for the development of hepatocellular carcinoma is cirrhosis caused by underlying chronic liver disease, such as metabolic dysfunction-associated steatotic liver disease (MASLD), hepatitis B virus (HBV) or hepatitis C virus (HCV) infection [1, 4–8]. However, as hepatocellular carcinoma is often asymptomatic in the early stages, timely diagnosis is rare [9]. As such, prognosis tends to be poor, as patients are often no longer suitable for surgical resection or liver transplantation by the time hepatocellular carcinoma is detected [9, 10]. Active surveillance of patients with certain high-risk aetiologies has been shown to improve overall survival by identifying tumours at an earlier stage and thereby enabling treatments to be initiated [11, 12].
Current guidelines recommend hepatocellular carcinoma surveillance with abdominal ultrasound (US) every 6 months in patients at risk of hepatocellular carcinoma. Research has shown that the diagnostic accuracy of US alone for the detection of hepatocellular carcinoma is poor, with the results of a meta-analysis demonstrating that US had a sensitivity of only 45% for the detection of early-stage hepatocellular carcinoma [13]. Several studies have suggested that serum biomarkers, such as α-fetoprotein (AFP) and protein induced by vitamin K absence or antagonist-II (PIVKA-II), have the potential to support a diagnosis of hepatocellular carcinoma [14, 15]. As such, European Association for the Study of the Liver guidelines recommend lifelong hepatocellular carcinoma surveillance with abdominal US, with or without AFP, in patients with cirrhosis, as long as they remain eligible for curative or disease-modifying treatment [16, 17]. The Swiss Association for the Study of the Liver recommends US, with or without AFP, for the surveillance of patients at risk of hepatocellular carcinoma, similar to practice guidance issued by the American Association for the Study of Liver Diseases [18, 19]. When used in combination, these surveillance strategies result in increased sensitivity for the detection of hepatocellular carcinoma compared with US alone, but with reduced specificity [20]. While the diagnostic value of AFP in isolation is debatable [21], combining AFP and PIVKA-II has been shown to improve detection of hepatocellular carcinoma compared with either biomarker alone [22].
GAAD (gender [biological sex], age, AFP, PIVKA-II]), a serum-based algorithm, is a novel in vitro diagnostic tool that combines gender (biological sex) and age with the serum biomarkers AFP and PIVKA-II (previously des-gamma carboxyprothrombin) to produce a semi-quantitative result. Notably, the GAAD algorithm has shown similar clinical performance for the diagnosis of hepatocellular carcinoma as the established GALAD algorithm, which combines the components of GAAD plus another biomarker, Lens culinaris agglutinin-reactive fraction of AFP (AFP-L3) [23].
Despite guideline recommendations, outside of tertiary centres, surveillance programmes are not always implemented owing to factors such as limited knowledge of guidelines in primary care, leading to some high-risk patients not being referred for surveillance. Additionally, patient-reported barriers to surveillance completion, such as scheduling difficulties and transportation concerns, have also been acknowledged [24]. Therefore, novel cost-effective surveillance strategies are needed to improve adherence to hepatocellular carcinoma surveillance. The objective of the current study was to perform a cost-effectiveness analysis of current hepatocellular carcinoma surveillance strategies in Switzerland, including the novel GAAD algorithm, in patients with compensated liver cirrhosis.
A micro-simulated Markov model was used to estimate the costs and performance of hepatocellular carcinoma surveillance strategies from the perspective of the Swiss healthcare system. The model was based on a recently published UK cost-effectiveness analysis in patients with compensated liver cirrhosis [25]. Each health state was associated with costs and health-related quality of life weights, with arrows representing transition probabilities (figure 1). A simulated cohort of 100,000 patients aged ≤75 years with compensated liver cirrhosis with any of the four underlying cirrhosis aetiologies (alcohol-associated liver disease [ALD], HBV, HCV or MASLD) were considered eligible for hepatocellular carcinoma surveillance and included in the model. The following surveillance strategies were evaluated: no surveillance, surveillance with US, surveillance with US+AFP and surveillance with GAAD alone. Active surveillance was performed at 6-month intervals. Patients were simulated individually, and outcomes were obtained for a lifetime horizon on a 6-month cycle. Disease stage was scored using the Barcelona Clinic Liver Cancer staging system, with stages 0/A representing early-stage hepatocellular carcinoma, and B/C/D representing late-stage hepatocellular carcinoma. In the simulation, patients started in the “compensated liver cirrhosis” state and either developed decompensated liver cirrhosis, developed early-stage hepatocellular carcinoma (Barcelona Clinic Liver Cancer stage 0/A undetected), which could progress to late-stage hepatocellular carcinoma (B/C/D undetected), or death (figure 1).
Model structure. *Treatment: liver transplantation, resection, radiofrequency thermal ablation, transarterial chemoembolisation, systemic therapy, best supportive care. Treatment for patients with decompensated liver cirrhosis liver transplantation or best supportive care only. Arrows indicate progression of patients through different states throughout the simulation. 1L: first-line; 2L: second-line; 3L: third-line; DCLC: decompensated liver cirrhosis; HCC: hepatocellular carcinoma.
Detection occurred in all stages via active surveillance or incidentally; incidental detection was assumed to be higher for late-stage hepatocellular carcinoma. The probability of progression from early- to late-stage hepatocellular carcinoma was predicted using expected tumour volume doubling times and assumed baseline tumour sizes. Adherence to hepatocellular carcinoma surveillance was assumed based on probability of attendance. Patients who developed decompensated liver cirrhosis either moved to the “waiting list” for liver transplantation, remained in the “decompensated liver cirrhosis” state until death or moved to the “hepatocellular carcinoma” state. The probability of tumour detection under surveillance was based on the probability of attendance at appointments, the diagnostic performance of the surveillance strategy used, and the probability of incidental/symptomatic detection. All patients with a positive test for hepatocellular carcinoma underwent confirmatory screening. After hepatocellular carcinoma diagnosis, patients received one of six possible first-line treatments: resection, orthotopic liver transplantation, radiofrequency ablation, transarterial chemoembolisation, best supportive care or systemic treatment. Patients eligible for orthotopic liver transplantation moved to the “waiting list” state, where the outcome was predicted according to stage-specific mortality. Based on calculated probability, patients could also receive any of the above treatments as second- or third-line therapy. Patients treated for hepatocellular carcinoma moved back into the “decompensated liver cirrhosis” state of the model (based on a survival probability for post-hepatocellular carcinoma treatment) unless they received orthotopic liver transplantation. The model assumed that no patient died with undetected hepatocellular carcinoma and that all patients diagnosed with and treated for hepatocellular carcinoma received palliative care in the last cycle of life.
Model parameters are reported in tables 1–3. Epidemiological and performance parameters are reported in table 1. Epidemiological parameters were estimated based on literature identified during a previous systematic literature review [25]. Adherence to surveillance was set at 52.0% (38.0–63.0%) for each surveillance strategy [26]. Performance of GAAD was based on data from the results of a multicentre analysis, in line with a previous UK-based cost-effectiveness analysis [23, 25]. Performance of US and US+AFP was based on the Tzartzeva et al. meta-analysis in 2018 [13]. Utility parameters (table 2) included health-related quality of life weights for all health states and palliative care, and they fell between 0 and 1, where 0 corresponded with death and 1 with perfect health. Utility parameters were sourced from published literature [27, 28]. Cost parameters (table 3) included costs for all surveillance strategies, together with treatment-related costs for compensated and decompensated liver cirrhosis, hepatocellular carcinoma confirmatory testing (true and false positive test results), symptomatic detection, follow-up and treatment. Further information on the management of false-positive results is reported in the appendix (page 2) and in appendix figure S1. All costs were obtained from Swiss sources: TARMED 2025 tariffs [29], the Analysis List [30], the Swiss Federal Statistical Office [31], publicly available data [32, 33] and published literature [34, 35]. Costs from published literature were not adjusted for inflation [34, 35].
Table 1Model parameters: epidemiology, surveillance and treatment.
| Parameters | Base case (minimum to maximum) | Source | |
| Age at start of surveillance in years (95% CI) | CLC (ALD) | 50.00 (34.81–73.23) | Base case: Kim et al., 2019 [38]. Range: Garay et al., 2024 [25] |
| CLC (HBV) | 50.00 (21.83–67.70) | Base case: Kim et al., 2019 [38]. Range: Garay et al., 2024 [25] | |
| CLC (HCV) | 50.00 (30.84–71.22) | Base case: Kim et al., 2019 [38]. Range: Garay et al., 2024 [25] | |
| CLC (MASLD) | 50.00 (43.81–84.19) | Base case: Kim et al., 2019 [38]. Range: Garay et al., 2024 [25] | |
| Upper age limit for surveillance in years (95% CI) | 75.00 (70.00–80.00) | Assumption based on UK analysis [25] | |
| Annual incidence of HCC, % (95% CI) | CLC (ALD) | 1.10% (0.99–1.21%) | Marot et al., 2017 [39] |
| CLC (HBV) | 3.35% (3.02–3.69%) | Thiele et al., 2014 [40] | |
| CLC (HCV) | 2.90% (2.61–3.19%) | Marot et al., 2017 [39] | |
| CLC (MASLD) | 3.10% (2.79–3.41%) | Marot et al., 2017 [39] | |
| Annual incidence of DCLC, % (95% CI) | CLC (ALD) | 7.30% (6.50–8.20%) | NICE guideline NG50 [41]; Fleming et al., 2010 [42] |
| CLC (HBV) | 5.00% (4.50–5.50%) | NICE guideline NG50 [41]; Dakin et al., 2010 [43] | |
| CLC (HCV) | 4.00% (3.60–4.40%) | NICE guideline NG50 [41]; Wright et al., 2006 [44] | |
| CLC (MASLD) | 3.80% (3.42–4.18%) | Estes et al., 2018 [45] | |
| DCLC-related parameters, % (95% CI) | DCLC after HCC treatment | 15.50% (13.95–17.05%) | Kondo et al., 2022 [46] |
| DCLC, listed for OLT | 3.00% (2.40–3.60%) | Assumption | |
| Annual incidental detection, % (95% CI) | HCC 0/A | 6.85% (0.00–23.25%) | Thompson et al., 2007 [47] |
| HCC B/C/D | 31.05% (0.00–65.15%) | ||
| Diagnostic performance by HCC stage in patients with cirrhosis: US, % (95% CI) | Sensitivity, early | 47.00% (33.00–61.00%) | Tzartzeva et al., 2018 [13] |
| Specificity, early | 91.00% (86.00–94.00%) | ||
| Sensitivity, all | 84.00% (76.00–92.00%) | ||
| Specificity, all | 91.00% (86.00–94.00%) | ||
| Diagnostic performance by HCC stage in patients with cirrhosis: US+AFP, % (95% CI) | Sensitivity, early | 63.00% (48.00–75.00%) | Tzartzeva et al., 2018 [13] |
| Specificity, early | 84.00% (77.00–89.00%) | ||
| Sensitivity, all | 97.00% (91.00–99.00%) | ||
| Specificity, all | 84.00% (77.00–89.00%) | ||
| Diagnostic performance by HCC stage in patients with cirrhosis: GAAD, % (95% CI) | Sensitivity, early | 67.40% (58.80–75.20%) | Subanalysis from Piratvisuth et al., 2023 [23], published in Garay et al., 2024 [25]. See appendix table S6. |
| Specificity, early | 86.60% (78.90–92.30%) | ||
| Sensitivity, all | 81.90% (76.90–86.20%) | ||
| Specificity, all | 86.60% (78.90–92.30%) | ||
| Treatment distribution for early-stage HCC detected, % (95% CI) | OLT | 5.34% (4.81–5.87%) | Kirstein et al., 2017 [48] |
| Resection | 22.33% (20.10–24.56%) | ||
| RFA | 39.32% (35.39–43.25%) | ||
| TACE | 18.93% (17.04–20.83%) | ||
| SYS | 3.88% (3.50–4.27%) | ||
| BSC | 10.19% (9.17–11.21%) | ||
| Treatment distribution for late-stage HCC detected, % (95% CI) | OLT | 1.18% (1.06–1.29%) | Kirstein et al., 2017 [48] |
| Resection | 19.69% (17.72–21.65%) | ||
| RFA | 10.99% (9.89–12.09%) | ||
| TACE | 28.12% (25.31–30.94%) | ||
| SYS | 14.06% (12.66–15.47%) | ||
| BSC | 25.96% (23.37–28.56%) | ||
| OLT waiting time (in 6-month cycles) | 3.00 | Schwab et al., 2024 [49] | |
AFP: α-fetoprotein; ALD: alcohol-associated liver disease; BSC: best supportive care; CI: confidence interval; CLC: compensated liver cirrhosis; DCLC: decompensated liver cirrhosis; GAAD: gender (biological sex), age, AFP, PIVKA-II; HBV: hepatitis B virus; HCC 0/A: hepatocellular carcinoma stage 0/A; HCC B/C/D: hepatocellular carcinoma stage B/C/D; HCV: hepatitis C virus; MASLD: metabolic dysfunction-associated fatty liver disease; OLT: orthotopic liver transplantation; RFA: radiofrequency thermal ablation; SYS: systemic therapy; TACE: transarterial chemoembolisation; US: ultrasound.
Table 2Model parameters: utility weights (QoL).
| Parameters | Base case (minimum to maximum) | Source |
| CLC | 0.75 (0.68–0.83) | McLernon et al., 2008 [27] |
| DCLC | 0.67 (0.60–0.74) | McLernon et al., 2008 [27] |
| HCC undetected | 0.75 (0.68–0.83) | Assumed asymptomatic and equal to CLC |
| Waiting list | 0.70 (0.63–0.77) | McLernon et al., 2008 [27] |
| OLT and post | 0.71 (0.64–0.78) | McLernon et al., 2008 [27] |
| Resection and post | 0.70 (0.63–0.77) | Lima et al., 2019 [28] |
| RFA and post | 0.76 (0.68–0.84) | |
| TACE and post | 0.65 (0.59–0.72) | |
| BSC and post | 0.40 (0.36–0.44) | |
| SYS | 0.76 (0.68–0.84) | |
| Palliative | 0.40 (0.36–0.44) |
BSC: best supportive care; CLC: compensated liver cirrhosis; DCLC: decompensated liver cirrhosis; HCC: hepatocellular carcinoma; OLT: orthotopic liver transplantation; QoL: quality of life; RFA: radiofrequency thermal ablation; SYS: systemic therapy; TACE: transarterial chemoembolisation.
Table 3Model parameters: costs.
| Parameter | Base case (minimum to maximum) | Source | |
| Surveillance costs, CHF (95% CI) | US | 121.00 (110.00–131.00) | Determined using the average from TARMED codes for non-radiologists (TARMED codes: 39.3250, 39.3800, 39.0020) and radiologists (TARMED codes: 39.3250, 39.3800, 39.0015) [29] |
| AFP | 17.40* | Determined via the official “List of analyses with tariff”; code 1034.00 [30] | |
| GAAD | 139.00 (111.20–166.80) | As GAAD is not reimbursed in Switzerland, the assumption was based on the medical value of GAAD and analogues based on the “List of analyses with tariff” [30] | |
| Confirmatory testing costs, CHF | CT | 203 | Determined using TARMED code (39.4080) [29] |
| MRI | 296 | Determined using TARMED code (39.5060) [29] | |
| Treatment costs, CHF (95% CI) | CLC (annual) | 2715.00 (1841.00–3806.00) | Blach et al., 2019 [35] |
| DCLC (annual) | 20,347.00 (16,561.00–24,517.00) | Pfeil et al., 2015 [34] | |
| True positive for HCC | 615.00 (553.50–676.50) | See appendix page 2 | |
| False positive for HCC | 824.00 (741.60–906.40) | ||
| Incidental diagnosis | 412.00 (370.80–453.20) | ||
| Follow-up after HCC, per cycle | 319.00 (287.10–350.90) | ||
| OLT, per operation | 125,102.00 (105,665.00–144,540.00) | Blach et al., 2019 [35] | |
| Post-OLT follow-up (year 1) | 19,323.00 (16,321.00–22,325.00) | ||
| Post-OLT follow-up (year 2+) | 19,323.00 (16,321.00–22,325.00) | ||
| Resection | 22,709.63 (18,935.31–27,936.06) | Office fédéral de la statistique, Section Services de santé, Switzerland [31] | |
| RFA | 19,586.34 (11,048.04–57,813.84) | ||
| TACE | 8617.70 (8013.19–9295.56) | ||
| BSC, per month | 1244.31 (1187.26–1295.86) | ||
| SYS, annual | 54,629.00 (49,395.00–59,863.00) | Average of yearly costs for lenvatinib (CHF 59,863) –Eisai Pharma AG [33] and sorafenib (CHF 49,395) –Bayer (Switzerland) Ltd [32] | |
* No variability assumed.
AFP: α-fetoprotein; BSC: best supportive care; CHF: Swiss franc; CI: confidence interval; CLC: compensated liver cirrhosis; CT: computed tomography; DCLC: decompensated liver cirrhosis; GAAD: gender (biological sex), age, AFP, PIVKA-II; HCC: hepatocellular carcinoma; MRI: magnetic resonance imaging; OLT: orthotopic liver transplantation; RFA: radiofrequency thermal ablation; SYS: systemic therapy; TACE: transarterial chemoembolisation; US: ultrasound.
In line with previous cost-effectiveness analyses for hepatocellular carcinoma surveillance [36, 37] and as such costs are not incurred by the Swiss healthcare system, indirect costs associated with diagnosis (such as productivity costs due to absenteeism) were omitted from the analysis.
A description of the software used to perform the analyses is reported in the appendix (table S1). A summary of the quality of evidence of the studies used for parameter inputs is reported in the appendix (table S2).
Primary outcomes were life years lived, quality-adjusted life years and total costs per patient and per cohort. The cost-effectiveness of each surveillance strategy was analysed through incremental cost-effectiveness ratios and net monetary benefit for a cost-effectiveness threshold of Swiss Franc (CHF) 100,000/quality-adjusted life year. Cost-effectiveness acceptability curves were generated for each comparison. Incremental cost-effectiveness ratios were defined as the ratio between the change in cost to change in effect of one surveillance strategy compared with that of another surveillance strategy. A discount rate of 3.5% was applied. The net monetary benefit was calculated using willingness-to-pay thresholds and quality-adjusted life years.
Multi-way sensitivity analyses were performed to assess joint parameter uncertainty and to build cost-effectiveness planes and cost-effectiveness acceptability curves. Due to computational burden, cost-effectiveness planes and cost-effectiveness acceptability curves were developed using cohorts with 10,000 patients/simulations each with 250 first-order samples. The probability distributions for multi-way sensitivity analyses were assumed following standard recommendations [50]. Tables 1–3 include the parameter ranges used for one-way sensitivity analyses. Net monetary benefits were displayed in tornado plots for a cost-effectiveness threshold of CHF 100,000/quality-adjusted life year; positive net monetary benefit indicated the results were cost-effective, and negative net monetary benefit indicated that the results were not cost-effective. The incremental net monetary benefit of each strategy, with 95% confidence intervals (CIs), was calculated using the simulations of the cohorts that were used for the cost-effectiveness acceptability curves.
Several scenario analyses were modelled, and a deterministic base case with 10,000 simulations/patients was calculated to serve as a reference for comparison with the scenarios. The following scenario analyses were modelled: (A) surveillance of patients with compensated liver cirrhosis, non-cirrhotic HBV and Metavir FIB-3, (B) surveillance of patients with FIB-3 only, (C) patients with cirrhotic MASLD only, (D) patients with FIB-3 MASLD only, (E) 10% increased adherence rate for GAAD, (F) 25% increased adherence rate for GAAD, (G) 50% increased adherence rate for GAAD, and (H) surveillance with GALAD. Additional parameters for scenario analyses are listed in appendix tables S3–S7. The performance of GALAD for early-stage (0/A) and any-stage hepatocellular carcinoma were derived from the meta-analysis of Guan et al. [51]. Given that performance estimates were not reported by hepatocellular carcinoma stage for cirrhotic patients in this study, a novel random effects meta-analysis was conducted using studies identified by Guan et al. [51], which comprised at least 90% of individuals with cirrhosis with early-stage [52–56] and all-stage hepatocellular carcinoma [52, 54–58]. The meta-analysis was conducted using MetaDTA: Diagnostic Test Accuracy Meta-Analysis v2.1.3. [59–61]. Forest plots for the sensitivity and specificity of GALAD in the meta-analysis are shown in appendix figures S2–S3, and sensitivity and specificity point estimates for each study are shown in appendix figure S4. The estimates for sensitivity and specificity for GALAD by hepatocellular carcinoma stage in cirrhotic patients can be found in appendix table S7. An explanation of cost parameters for GALAD is reported in the appendix (page 2).
A summary of the study design and base case results is reported in figure 2. In the base case analysis of 100,000 patients/simulations, the costs and quality-adjusted life years per patient, respectively, were CHF 43,493.61 and 5.893 for no surveillance, CHF 48,702.79 and 6.018 for US, CHF 49,980.60 and 6.042 for US+AFP, and CHF 49,983.10 and 6.048 for GAAD alone (table 4). In terms of quality-adjusted life years, GAAD was the preferred surveillance strategy, followed by US+AFP, US and no surveillance. Compared with no surveillance, incremental costs and quality-adjusted life years, respectively, were CHF 5209.18 and 0.125 for US; CHF 6486.99 and 0.150 for US+AFP; and CHF 6489.49 and 0.155 for GAAD, with incremental cost-effectiveness ratios below the willingness-to-pay threshold in the range CHF 41,509.01–43,321.81. Compared with US, incremental costs and quality-adjusted life years were CHF 1277.81 and 0.024, respectively (resulting in an incremental cost-effectiveness ratio of CHF 52,705.29), for US+AFP, and CHF 1280.30 and 0.030 (CHF 43,020.48) for GAAD. GAAD showed similar outcomes compared with US+AFP, with an incremental cost-effectiveness ratio of CHF 452.30 per quality-adjusted life year (incremental costs and quality-adjusted life years of CHF 2.49 and 0.006, respectively).
Figure 1Summary of study design and base case results. AFP: α-fetoprotein; CHF: Swiss francs; GAAD: gender (biological sex), age, AFP, PIVKA-II; HCC: hepatocellular carcinoma; ICER: incremental cost-effectiveness ratio; QALY: quality-adjusted life year; US: ultrasound.
An extended table of results detailing cost and health outcomes per cohort are reported in appendix table S8. Of the 100,000 patients with compensated liver cirrhosis in the simulation, 19,933 were considered to have developed hepatocellular carcinoma within their lifetime. In comparison with no surveillance, patients with hepatocellular carcinoma gained an average of 0.63 discounted quality-adjusted life years with US, 0.75 with US+AFP and 0.78 with GAAD. GAAD had the highest rate of early-stage hepatocellular carcinoma detection with surveillance (56%), followed by US+AFP (53%) and US (44%). US+AFP generated the highest number of false-positives (n = 123,687), followed by GAAD (n = 103,643) and US (n = 69,764). GAAD produced the fewest false-negatives (n = 5678), followed by US+AFP (n = 6361) and US (n = 10,378).
Table 4Base case results.
| No surveillance | US | US+AFP | GAAD | ||
| Cost per patient, CHF | 43,493.61 | 48,702.79 | 49,980.60 | 49,983.10 | |
| QALYs per patient | 5.893 | 6.018 | 6.042 | 6.048 | |
| Surveillance strategies compared with no surveillance | Incremental costs, CHF | – | 5209.18 | 6486.99 | 6489.49 |
| Incremental QALYs | – | 0.125 | 0.150 | 0.155 | |
| ICER per QALY, CHF | – | 41,509.01 | 43,321.81 | 41,798.74 | |
| Surveillance strategies compared with US | Incremental costs, CHF | – | - | 1277.81 | 1280.30 |
| Incremental QALYs | – | - | 0.024 | 0.030 | |
| ICER per QALY, CHF | – | - | 52,705.29 | 43,020.48 | |
| Surveillance strategies compared with US+AFP | Incremental costs, CHF | – | - | - | 2.49 |
| Incremental QALYs | – | - | - | 0.006 | |
| ICER per QALY, CHF | – | - | - | 452.30 | |
AFP: α-fetoprotein; CHF: Swiss franc; ICER: incremental cost-effectiveness ratio; QALY: quality-adjusted life year; WTP: willingness to pay.
ICERs in bold indicate that the surveillance strategy is cost-effective within the WTP threshold of CHF 100,000 versus the comparator.
Analyses of paired comparisons are shown in figure 3. Compared with no surveillance, all strategies were cost-effective at a cost-effectiveness threshold of CHF 100,000. Compared with US, GAAD was the most cost-effective strategy, followed by US+AFP. GAAD was also cost-effective compared with US+AFP at the cost-effectiveness threshold depicted. Strategies were evaluated in terms of total discounted quality-adjusted life years, costs and the estimated probability of cost-effectiveness for different willingness-to-pay thresholds (figure 4). One-way sensitivity analyses comparing all surveillance strategies are shown in appendix figures S5–S10. Performance of diagnostic modalities, age at the start of surveillance and annual incidental detection were the most influential variables. In particular, tornado plots showed that the performance of GAAD and comparators were the key variables contributing to the cost-effectiveness of GAAD (appendix figures S9 and S10). The incremental net monetary benefit with 95% CIs for each surveillance strategy are reported in appendix table S9.
Cost-effectiveness plane depicting the incremental costs and effects of each surveillance strategy versus comparator. AFP: α-fetoprotein; CE: cost-effectiveness; CHF: Swiss franc; GAAD: gender (biological sex), age, AFP, PIVKA-II; QALY: quality-adjusted life year; US: ultrasound.
Probability of being the most cost-effective option at a willingness-to-pay threshold of CHF 100,000. AFP: α-fetoprotein; CHF: Swiss franc; CE: cost-effectiveness; GAAD: gender (biological sex), age, AFP, PIVKA-II; US: ultrasound.
The results of the deterministic base case for patients with compensated liver cirrhosis, the reference case against which the results of all scenario analyses were compared, are reported in appendix table S10. Compared with no surveillance, incremental quality-adjusted life years were 0.130 for US, 0.165 for US+AFP and 0.166 for GAAD. Incremental cost-effectiveness ratios compared with no surveillance were CHF 41,792.57, CHF 41,810.82 and CHF 40,921.24 for US, US+AFP and GAAD, respectively. Incremental quality-adjusted life years compared with US were 0.035 for US+AFP and 0.037 for GAAD. Compared with US, the incremental cost-effectiveness ratios were CHF 41,877.95 for US+AFP and CHF 37,835.73 for GAAD. Compared with US+AFP, GAAD was the dominant strategy in terms of incremental cost-effectiveness ratio.
In scenario analysis (A), where surveillance was performed in a combined group of patients with cirrhosis, non-cirrhotic HBV or FIB-3, incremental quality-adjusted life years were fewer in comparison to the deterministic base case (appendix table S11) as the incidence of hepatocellular carcinoma was lower than for patients with compensated liver cirrhosis only. Additionally, when compared with no surveillance, the cost-effectiveness of all surveillance strategies decreased when surveillance was performed in this combined patient group (cirrhosis, HBV or FIB-3) versus the deterministic base case. For example, compared with no surveillance, the incremental cost-effectiveness ratios for GAAD were CHF 40,921.24 in the deterministic base case versus CHF 75,090.80. Compared with no surveillance, the most cost-effective strategy was GAAD (incremental cost-effectiveness ratio: CHF 75,090.80), followed by US (CHF 92,351.85) and US+AFP (CHF 96,974.25). Compared with US alone, the incremental cost-effectiveness ratios for US+AFP and GAAD were CHF 113,824.80 and CHF 23,809.37, respectively. Compared with US+AFP, GAAD remained the dominant strategy.
For scenario analysis (B), which included patients with FIB-3 only, the cost-effectiveness of surveillance strategies, in comparison with no surveillance, was lower versus the deterministic base case (appendix table S12). In this scenario, US+AFP and US were no longer cost-effective, whereas GAAD remained cost-effective but with a greater incremental cost-effectiveness ratio. Compared with no surveillance, GAAD was the most cost-effective (incremental cost-effectiveness ratio: CHF 88,976.53), whereas US and US+AFP had incremental cost-effectiveness ratios above the willingness-to-pay threshold (incremental cost-effectiveness ratio: CHF 110,263.67 and 119,699.52, respectively). GAAD was cost-effective, in comparison to US, in the FIB-3-only population (incremental cost-effectiveness ratio: CHF 22,472.61). In contrast, US+AFP was not cost-effective in comparison to US with an incremental cost-effectiveness ratio of CHF 158,680.66. GAAD remained the dominant strategy compared with US+AFP, with an incremental cost of CHF -921.82 and 0.002 incremental quality-adjusted life years.
In scenario analysis (C), patients with cirrhotic MASLD only were included. Incremental cost-effectiveness ratios for each strategy compared with no surveillance were lower than in the deterministic base case: CHF 37,969.44 for GAAD, CHF 39,005.08 for US+AFP and CHF 39,237.15 for US (appendix table S13). The cost-effectiveness of each strategy compared with US also further improved versus that of the deterministic base case. Compared with US, the incremental cost-effectiveness ratio was CHF 33,476.53 for GAAD and CHF 38,111.15 for US+AFP. GAAD was the dominant strategy compared with US+AFP in this scenario (incremental cost: CHF -83.84; incremental quality-adjusted life years: 0.004).
Scenario analysis (D) showed that when surveillance was performed in patients with FIB-3 MASLD only, the costs per patient decreased and quality-adjusted life years increased versus the deterministic base case (appendix table S14). Compared with no surveillance, GAAD was the most cost-effective strategy (incremental cost-effectiveness ratio: CHF 75,776.81), followed by US (CHF 90,362.51) and US+AFP (CHF 96,910.72). Similarly, compared with US, GAAD was the most cost-effective strategy (incremental cost-effectiveness ratio: CHF 31,095.91). In this scenario, US+AFP was no longer cost-effective compared with US (incremental cost-effectiveness ratio: CHF 122,967.28). GAAD was the dominant strategy compared with US+AFP (incremental cost: CHF -828.44; incremental quality-adjusted life years: 0.003).
In scenario analyses (E), (F) and (G), the cost-effectiveness of GAAD was analysed after various increases to the GAAD adherence rate: 10% (from 52.0% to 57.2%), 25% (from 52.0% to 65.0%) and 50% (from 52.0% to 78.0%). With each increase, quality-adjusted life years were greater in comparison with the deterministic base case (appendix tables S15–S17). Incremental cost-effectiveness ratios for GAAD compared with each strategy were greater compared with the deterministic base case as GAAD adherence rates increased.
In scenario analysis (H), where GALAD was included as an additional surveillance strategy, costs were slightly greater for GALAD than GAAD and quality-adjusted life years were similar (CHF 50,227.81 and 6.014 versus CHF 49,694.07 and 6.013, respectively) (appendix table S18). Compared with no surveillance, GAAD was the most cost-effective strategy (incremental cost-effectiveness ratio: CHF 40,921.24), followed by US (CHF 41,792.57), US+AFP (CHF 41,812.45) and GALAD (CHF 43,832.44). When compared with US, GAAD was the most cost-effective (incremental cost-effectiveness ratio: CHF 37,835.73), followed by US+AFP (CHF 41,885.60) and GALAD (CHF 50,840.89). GAAD was the dominant strategy compared with US+AFP. GALAD was no longer cost-effective compared with US+AFP and GAAD, with incremental cost-effectiveness ratios of CHF 177,927.32 and 474,692.26, respectively.
The prognosis at diagnosis of hepatocellular carcinoma is often poor, so effective surveillance is required to facilitate earlier detection and thus improve treatment options. This study used a simulated model to evaluate the cost-effectiveness of different hepatocellular carcinoma surveillance strategies in Switzerland, which, to our knowledge, is the first such study conducted in a Swiss healthcare setting.
The findings of this study indicate that, at a cost-effectiveness threshold of CHF 100,000, 6-monthly surveillance with the GAAD algorithm was cost-effective compared with US and US+AFP, the current surveillance strategies recommended in Switzerland [16, 18]. These results are similar to those from a comparable analysis reporting perspectives from the UK National Health Service, which found that GAAD alone was the most cost-effective strategy compared with US and US+AFP in cirrhotic patients [25]. The results of the UK analysis also indicated that 6-monthly hepatocellular carcinoma surveillance with US or US+AFP is unlikely to be cost-effective at a cost-effectiveness threshold of £30,000, but it becomes cost-effective at a cost-effectiveness threshold of between £31,000 and £34,000. However, our findings were not consistent with the results of another recent North American cost-effectiveness analysis that found that US+AFP was dominant compared to both US alone and no surveillance for the detection of hepatocellular carcinoma in patients with compensated cirrhosis [37]. Despite these differences, these studies highlight that although US+AFP and US alone are currently the most widely recommended strategies for hepatocellular carcinoma surveillance, additional cost-effectiveness studies to assess the role of existing and novel strategies for hepatocellular carcinoma surveillance in a diversity of healthcare settings are needed.
GAAD alone was the preferred surveillance strategy in terms of quality-adjusted life years gained, with potential drivers of this benefit including early detection of hepatocellular carcinoma and the associated shift from late-stage treatment (e.g. systemic) to early-stage treatment (e.g. orthotopic liver transplantation and resection). The average increase in quality-adjusted life years per patient for GAAD compared with each strategy was, as expected, small (0.006 compared with US+AFP, 0.03 compared with US alone and 0.155 compared with no surveillance), due to the fact that only a small fraction of patients who underwent surveillance developed hepatocellular carcinoma. However, for these patients, the increase in quality-adjusted life years was greater (0.03, 0.15 and 0.78 for GAAD compared with US+AFP, US alone and no surveillance, respectively).
When determining the cost-effectiveness of surveillance strategies, the incidence of hepatocellular carcinoma in each patient population considered for surveillance is of importance. In the scenario analyses, incremental cost-effectiveness ratios with each surveillance strategy were lower in patients with cirrhotic aetiologies, particularly cirrhotic MASLD. Incremental cost-effectiveness ratios were greater in the scenario with patients with compensated liver cirrhosis, HBV and FIB-3, compared with the deterministic base case, likely due to the high incidence rate of hepatocellular carcinoma in the cirrhotic population [1]. When the analyses were performed in patients with FIB-3, whose risk of hepatocellular carcinoma was lower than those with cirrhosis, incremental cost-effectiveness ratios were higher for all hepatocellular carcinoma surveillance strategies compared with base case, with only GAAD remaining cost-effective. Previous studies have shown that hepatocellular carcinoma risk can be stratified based on clinical variables, including aetiology and molecular profiling [62–64]. Therefore, determining the modality of surveillance based on individual hepatocellular carcinoma risk may be more cost-effective than using a single strategy across all patient populations. For example, psychiatric outpatients are more likely to develop hepatitis C or cirrhosis and may therefore benefit from scaling-up of hepatocellular carcinoma screening with GAAD [65]. Despite this variability in cost-effectiveness, GAAD remained the most cost-effective option compared with other strategies across all aetiology scenario analyses and the deterministic base case, likely due to the quality-adjusted life years gained as a result of early-stage hepatocellular carcinoma diagnosis with GAAD.
Another factor that significantly impacted cost-effectiveness was patient adherence to surveillance, as reflected in scenario analyses E–G. When adherence to GAAD was increased by 10%, 25% and 50%, the clinical effectiveness of GAAD improved. Although the overall cost-effectiveness of GAAD was slightly reduced when adherence was increased, GAAD remained cost-effective across all adherence rates. This reduction in overall cost-effectiveness was due to increased costs with greater adherence despite increased quality-adjusted life years.
While appropriate data for GAAD in overweight, obese and diabetic populations were not available at the time of the analysis, future research should investigate both the performance and cost-effectiveness of surveillance strategies in such populations. In particular, the increasing prevalence of non-cirrhotic MASLD in overweight populations is a point of interest. Moreover, a recent phase III cohort study demonstrated greater predictive value for the composite biomarker score, HES V2.0, which combines changes in AFP, AFP-L3 and PIVKA-II over 12 months with age, platelets, alanine transaminase and underlying disease aetiology, compared with GALAD and ASAP scores (age, sex, AFP and PIVKA-II) [66]. However, it should be noted that the cut-offs for GALAD used in this study are not those validated for clinical use and the sample size was relatively small [66]. As such, future research should aim to further validate the performance of the HES V2.0 strategy and once costs are available, its cost-effectiveness compared with other available strategies warrants investigation.
It is important to consider that data on the performance of GAAD were only available from studies of case-control design, which have inherent limitations that can result in the overestimation of test performance. These data can be considered less extensive and robust than the meta-analysis data available for the performance of the other surveillance strategies, such as US±AFP [67]. Nevertheless, we expect that the potential differences in performance estimations coming from the study designs were captured in the various sensitivity analyses. When all surveillance interventions were compared with no surveillance, changes in performance estimates did not play a relevant role (appendix figures S5–S7). Similarly, when compared with US, a lower sensitivity or specificity for GAAD, even below the 95% CI limit, did not change the conclusions. However, as GAAD and US+AFP have similar total costs and quality-adjusted life years, the results for this comparison were more sensitive to changes in performance estimates. Despite this, the 95% CIs around the incremental net monetary benefit values for each comparison were small, indicating a low level of variability in the data.
Although the base case probabilistic results suggest that GAAD was nearly cost-neutral compared with US+AFP, with a low incremental cost-effectiveness ratio per quality-adjusted life year of CHF 452, this result was dependent on performance data (appendix figure S10). GALAD, for which performance data from prospective cohort studies are available [68], has shown similar performance to GAAD [51]. Despite the inclusion of GALAD in scenario analysis H, GAAD remained the most cost-effective strategy compared with no surveillance, US and US+AFP. Although the inclusion of AFP-L3 in GALAD led to slightly higher quality-adjusted life years than GAAD, this was not cost-effective due to the extra cost of the additional biomarker test. These data indicate that GAAD is the preferred strategy over GALAD in terms of cost-effectiveness. However, as AFP-L3 is not available in Switzerland, it should be noted that the costs of GALAD were based on assumptions involving other available test prices, as detailed in the appendix (page 2).
There were several limitations associated with the analysis. Several parameters were obtained using non-Swiss databases or patients, including the survival curves stratified by hepatocellular carcinoma, which were taken from a German study [48]; the model only considered early- and late-stage hepatocellular carcinoma owing to a lack of data on the surveillance strategies by hepatocellular carcinoma stage, and a number of parameters were assumed to be equal irrespective of patient sex owing to a lack of data and for model simplification, including progression rates, incidence rates and statistics for survival curves. The absence of a single source providing a comprehensive performance estimate for all the strategies under comparison could also have created biases and potentially impacted the results. Additionally, in the scenario analyses, similar performance of US and US+AFP was assumed for both cirrhotic and non-cirrhotic disease. However, there is evidence to suggest that US performs better in non-cirrhotic aetiologies compared with cirrhotic aetiologies, due to decreased nodularity and better-defined margins, resulting in a clearer image [69]. Due to a lack of data and suitable sources for utilities, for some parameters, the most appropriate sources were slightly dated and based on viraemic HCV patients. TARDOC tariffs were not considered as they were not yet finalised at the time of the analysis. However, according to the deterministic sensitivity analysis, the cost elements concerned were not considered to have a substantial impact on the final results. Comparisons with GAAD+US were not included in the sensitivity analysis owing to the lack of robustness of the data available. Therefore, further investigation is needed for this comparison. Finally, external validity may be limited outside of a Swiss healthcare setting. However, the Swiss setting was selected due to the high standard of healthcare quality, comprehensive data availability and well-studied patient population, which enhances generalisability and comparisons to other healthcare settings. Furthermore, the conclusions were similar to those found in the UK-based cost-effectiveness analysis [25], indicating that these data are likely applicable to other healthcare systems.
Our analyses found that hepatocellular carcinoma surveillance is cost-effective in Switzerland, regardless of the modality used. Notably, GAAD was the most cost-effective strategy when compared with US alone. GAAD was nearly cost-neutral compared with US+AFP, with similar costs but higher quality-adjusted life years, although this was dependent on the performance data of both GAAD and the operator-dependent US combined with AFP. Further investigations are required to confirm these findings and to optimise surveillance of patients at risk of hepatocellular carcinoma.
The data supporting the conclusions of this article are publicly available and fully reported in the publication. Further data requests can be addressed to the corresponding author and data will be made available on reasonable request. The study protocol is described throughout the publication; however, any methodological questions can be addressed to the corresponding Methods author (Osvaldo Ulises Garay, MSc; Global Access & Policy, Roche Diagnostics International; Forrenstrasse 2, Switzerland, 6343 Rotkreuz ZG).
Author contributions: All authors made substantial contributions to the conception or design of the work, or acquisition, analysis or interpretation of data for the work; drafting the work or revising it critically for important intellectual content; provided final approval of the version to be published and agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
This analysis was funded by Roche Diagnostics. GAAD is a cost-effectiveness-marked digital tool used to aid in the diagnosis of early-stage hepatocellular carcinoma. Katie Williams and Nichola Cruickshanks of Springer Health+, Springer Healthcare Ltd., UK, provided medical writing support, which was funded by Roche Diagnostics International Ltd (Rotkreuz, Switzerland) in accordance with Good Publication Practice 2022 guidelines.
VK and LA were employed by Roche Diagnostics (Schweiz), AG, Switzerland, at the time of the study. VK owns stock in F. Hoffmann-La Roche AG. FG is the representative of Swiss hospitals at the Federal Drug Commission (Federal Office of Public Health) and is an expert for InnoSuisse, the Federal Agency for Innovation (Life Sciences section). CW is employed by Roche Diagnostics (Schweiz), AG, Switzerland and owns stock in F. Hoffmann-La Roche. OUG is employed by Roche Diagnostics International and owns stock in F. Hoffmann-La Roche. NG reports research funding from Gilead and Novo, congress funding from AbbVie and Gilead, and attends advisory boards for Roche Diagnostics and Novo Nordisk.
1. Llovet JM, Kelley RK, Villanueva A, Singal AG, Pikarsky E, Roayaie S, et al. Hepatocellular carcinoma. Nat Rev Dis Primers. 2021 Jan;7(1):6.
2. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021 May;71(3):209–49.
3. Rumgay H, Arnold M, Ferlay J, Lesi O, Cabasag CJ, Vignat J, et al. Global burden of primary liver cancer in 2020 and predictions to 2040. J Hepatol. 2022 Dec;77(6):1598–606.
4. Ebrahimi F, Hagström H, Sun J, Bergman D, Shang Y, Yang W, et al. Familial coaggregation of MASLD with hepatocellular carcinoma and adverse liver outcomes: nationwide multigenerational cohort study. J Hepatol. 2023 Dec;79(6):1374–84.
5. Girardin F, Hearmon N, Castro E, Negro F, Eddowes L, Gétaz L, et al. Modelling the impact and cost-effectiveness of extended hepatitis C virus screening and treatment with direct-acting antivirals in a Swiss custodial setting. Clin Infect Dis. 2019 Nov;69(11):1980–6.
6. Girardin F, Painter C, Hearmon N, Eddowes L, Kaiser S, Negro F, et al. Hepatitis C prevalences in the psychiatric setting: cost-effectiveness of scaling-up screening and direct-acting antiviral therapy. JHEP Rep. 2021 ;3(3):100279.
7. Herbst DA, Reddy KR. Risk factors for hepatocellular carcinoma. Clin Liver Dis (Hoboken). 2013 Jan;1(6):180–2.
8. West J, Card TR, Aithal GP, Fleming KM. Risk of hepatocellular carcinoma among individuals with different aetiologies of cirrhosis: a population-based cohort study. Aliment Pharmacol Ther. 2017 Apr;45(7):983–90.
9. Balogh J, Victor D 3rd, Asham EH, Burroughs SG, Boktour M, Saharia A, et al. Hepatocellular carcinoma: a review. J Hepatocell Carcinoma. 2016 Oct;3:41–53.
10. Benson AB, D’Angelica MI, Abbott DE, Anaya DA, Anders R, Are C, et al. Hepatobiliary Cancers, Version 2.2021, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2021 May;19(5):541–65.
11. Bordoni A, Cerny A, Bihl F, Alerci M, Mazzola P, Peverelli S, et al. Survival of hepatocellular carcinoma patients is significantly improving: a population-based study from southern Switzerland. Cancer Epidemiol. 2014 Dec;38(6):679–85.
12. Hong TP, Gow PJ, Fink M, Dev A, Roberts SK, Nicoll A, et al. Surveillance improves survival of patients with hepatocellular carcinoma: a prospective population-based study. Med J Aust. 2018 Oct;209(8):348–54.
13. Tzartzeva K, Obi J, Rich NE, Parikh ND, Marrero JA, Yopp A, et al. Surveillance imaging and alpha fetoprotein for early detection of hepatocellular carcinoma in patients with cirrhosis: a meta-analysis. Gastroenterology. 2018 May;154(6):1706–1718.e1.
14. Omata M, Cheng AL, Kokudo N, Kudo M, Lee JM, Jia J, et al. Asia-Pacific clinical practice guidelines on the management of hepatocellular carcinoma: a 2017 update. Hepatol Int. 2017 Jul;11(4):317–70.
15. Singh G, Yoshida EM, Rathi S, Marquez V, Kim P, Erb SR, et al. Biomarkers for hepatocellular cancer. World J Hepatol. 2020 Sep;12(9):558–73.
16. Sangro B, Argemi J, Ronot M, Paradis V, Meyer T, Mazzaferro V, et al.; European Association for the Study of the Liver. EASL Clinical Practice Guidelines on the management of hepatocellular carcinoma. J Hepatol. 2025 Feb;82(2):315–74.
17. Tacke F, Horn P, Wai-Sun Wong V, Ratziu V, Bugianesi E, Francque S, et al.; European Association for the Study of the Liver (EASL); European Association for the Study of Diabetes (EASD); European Association for the Study of Obesity (EASO). EASL-EASD-EASO Clinical Practice Guidelines on the management of metabolic dysfunction-associated steatotic liver disease (MASLD). J Hepatol. 2024 Sep;81(3):492–542.
18. Goossens N, Toso C, Heim MH. Management of hepatocellular carcinoma: SASL expert opinion statement. Swiss Med Wkly. 2020 Jul;150(3132):w20296.
19. Singal AG, Llovet JM, Yarchoan M, Mehta N, Heimbach JK, Dawson LA, et al. AASLD Practice Guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology. 2023 Dec;78(6):1922–65. doi: https://doi.org/10.1097/HEP.0000000000000466
20. Yang JD, Dai J, Singal AG, Gopal P, Addissie BD, Nguyen MH, et al. Improved performance of serum alpha-fetoprotein for hepatocellular carcinoma diagnosis in HCV cirrhosis with normal alanine transaminase. Cancer Epidemiol Biomarkers Prev. 2017 Jul;26(7):1085–92. doi: https://doi.org/10.1158/1055-9965.EPI-16-0747
21. Parikh ND, Mehta AS, Singal AG, Block T, Marrero JA, Lok AS. Biomarkers for the early detection of hepatocellular carcinoma. Cancer Epidemiol Biomarkers Prev. 2020 Dec;29(12):2495–503.doi: https://doi.org/10.1158/1055-9965.EPI-20-0005
22. Lee Q, Yu X, Yu W. The value of PIVKA-Ⅱ versus AFP for the diagnosis and detection of postoperative changes in hepatocellular carcinoma. J Interv Med. 2021 Mar;4(2):77–81.
23. Piratvisuth T, Hou J, Tanwandee T, Berg T, Vogel A, Trojan J, et al. Development and clinical validation of a novel algorithmic score (GAAD) for detecting HCC in prospective cohort studies. Hepatol Commun. 2023 Nov;7(11):e0317.
24. Wolf E, Rich NE, Marrero JA, Parikh ND, Singal AG. Use of hepatocellular carcinoma surveillance in patients with cirrhosis: a systematic review and meta-analysis. Hepatology. 2021 Feb;73(2):713–25.
25. Garay OU, Ambühl LE, Bird TG, Barnes E, Irving WL, Walkley R, et al. Cost-Effectiveness of Hepatocellular Carcinoma Surveillance Strategies in Patients With Compensated Liver Cirrhosis in the United Kingdom. Value Health. 2024 Dec;27(12):1698–709.
26. Zhao C, Jin M, Le RH, Le MH, Chen VL, Jin M, et al. Poor adherence to hepatocellular carcinoma surveillance: A systematic review and meta-analysis of a complex issue. Liver Int. 2018 Mar;38(3):503–14.
27. McLernon DJ, Dillon J, Donnan PT. Health-state utilities in liver disease: a systematic review. Med Decis Making. 2008;28(4):582–92.
28. Lima PH, Fan B, Bérubé J, Cerny M, Olivié D, Giard JM, et al. Cost-utility analysis of imaging for surveillance and diagnosis of hepatocellular carcinoma. AJR Am J Roentgenol. 2019 Jul;213(1):17–25.
29. Fédération des médecins suisses [Internet]. TARMED tariff online browser. 2025 [cited 2025 Sep 26]. Available from: https://www.tarmed-browser.ch/de
30. Federal Office of Public Health [Internet]. List of analyses with tariffs. 2023 [cited 2025 Sep 26]. Available from: https://www.bag.admin.ch/bag/de/home/versicherungen/krankenversicherung/krankenversicherung-leistungen-tarife/Analysenliste.html
31. Federal Statistical Office [Internet]. Health. 2023 [cited 2025 Sep 26]. Available from: https://www.bfs.admin.ch/bfs/en/home/statistiken/gesundheit.html
32. Bayer (Switzerland) Ltd [Internet]. NEXAVAR Filmtabl 200 mg. 2023 [cited 2025 Sep 26]. Available from: https://compendium.ch/product/1032677-nexavar-cpr-pell-200-mg
33. Eisai Pharma AG [Internet]. Lenvima®. 2023 [cited 2025 Sep 26]. Available from: https://compendium.ch/product/1304127-lenvima-kaps-4-mg/mpro#
34. Pfeil AM, Reich O, Guerra IM, Cure S, Negro F, Müllhaupt B, et al. Cost-effectiveness analysis of sofosbuvir compared to current standard treatment in Swiss patients with chronic hepatitis C. PLoS One. 2015 May;10(5):e0126984.
35. Blach S, Schaetti C, Bruggmann P, Negro F, Razavi H. Cost-effectiveness analysis of strategies to manage the disease burden of hepatitis C virus in Switzerland. Swiss Med Wkly. 2019 Mar;149(1112):w20026.
36. Kao SZ, Sangha K, Fujiwara N, Hoshida Y, Parikh ND, Singal AG. Cost-effectiveness of a precision hepatocellular carcinoma surveillance strategy in patients with cirrhosis. EClinicalMedicine. 2024 Aug;75:102755.
37. Parikh ND, Singal AG, Hutton DW, Tapper EB. Cost-effectiveness of hepatocellular carcinoma surveillance: an assessment of benefits and harms. Am J Gastroenterol. 2020 Oct;115(10):1642–9.
38. Kim HL, An J, Park JA, Park SH, Lim YS, Lee EK. Magnetic resonance imaging is cost-effective for hepatocellular carcinoma surveillance in high-risk patients with cirrhosis. Hepatology. 2019 Apr;69(4):1599–613.
39. Marot A, Henrion J, Knebel JF, Moreno C, Deltenre P. Alcoholic liver disease confers a worse prognosis than HCV infection and non-alcoholic fatty liver disease among patients with cirrhosis: an observational study. PLoS One. 2017 Oct;12(10):e0186715.
40. Thiele M, Gluud LL, Fialla AD, Dahl EK, Krag A. Large variations in risk of hepatocellular carcinoma and mortality in treatment naïve hepatitis B patients: systematic review with meta-analyses. PLoS One. 2014 Sep;9(9):e107177.
41. NICE guidelines [Internet]. Cirrhosis in over 16s: assessment and management. 2016 [cited 2025 Sep 26]. Available from: https://www.nice.org.uk/guidance/ng50/resources/cirrhosis-in-over-16s-assessment-and-management-pdf-1837506577093
42. Fleming KM, Aithal GP, Card TR, West J. The rate of decompensation and clinical progression of disease in people with cirrhosis: a cohort study. Aliment Pharmacol Ther. 2010 Dec;32(11-12):1343–50.
43. Dakin H, Bentley A, Dusheiko G. Cost-utility analysis of tenofovir disoproxil fumarate in the treatment of chronic hepatitis B. Value Health. 2010 Dec;13(8):922–33.
44. Wright M, Grieve R, Roberts J, Main J, Thomas HC. Health benefits of antiviral therapy for mild chronic hepatitis C: randomised controlled trial and economic evaluation. Health Technol Assess. 2006 Jul;10(21):1-113, iii. doi: .
45. Estes C, Razavi H, Loomba R, Younossi Z, Sanyal AJ. Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease. Hepatology. 2018 Jan;67(1):123–33.
46. Kondo T, Koroki K, Kanzaki H, Kobayashi K, Kiyono S, Nakamura M, et al. Impact of acute decompensation on the prognosis of patients with hepatocellular carcinoma. PLoS One. 2022 Jan;17(1):e0261619.
47. Thompson Coon J, Rogers G, Hewson P, Wright D, Anderson R, Cramp M, et al. Surveillance of cirrhosis for hepatocellular carcinoma: systematic review and economic analysis. Health Technol Assess. 2007 Sep;11(34):1–206.
48. Kirstein MM, Schweitzer N, Winter T, Lappas K, Graen N, Kunstmann I, et al. Patterns and challenges of treatment sequencing in patients with hepatocellular carcinoma: experience from a German referral center. J Gastroenterol Hepatol. 2017 Oct;32(10):1730–8.
49. Schwab S, Elmer A, Sidler D, Straumann L, Stürzinger U, Immer F. Selection Bias in Reporting of Median Waiting Times in Organ Transplantation. JAMA Netw Open. 2024 Sep;7(9):e2432415.
50. Briggs A, Claxton K, Sculpher M. Decision Modelling For Health Economic Evaluation. Oxford University Press; 2006. doi: https://doi.org/10.1093/oso/9780198526629.001.0001
51. Guan MC, Zhang SY, Ding Q, Li N, Fu TT, Zhang GX, et al. The performance of GALAD score for diagnosing hepatocellular carcinoma in patients with chronic liver diseases: a systematic review and meta-analysis. J Clin Med. 2023 Jan;12(3):949.
52. Chalasani NP, Porter K, Bhattacharya A, Book AJ, Neis BM, Xiong KM, et al. Validation of a novel multitarget blood test shows high sensitivity to detect early stage hepatocellular carcinoma. Clin Gastroenterol Hepatol. 2022 Jan;20(1):173–182.e7.
53. Chalasani NP, Ramasubramanian TS, Bhattacharya A, Olson MC, Edwards V DK, Roberts LR, et al. A novel blood-based panel of methylated DNA and protein markers for detection of early-stage hepatocellular carcinoma. Clin Gastroenterol Hepatol. 2021 Dec;19(12):2597–2605.e4.
54. Lambrecht J, Porsch-Özçürümez M, Best J, Jost-Brinkmann F, Roderburg C, Demir M, et al. The APAC score: a novel and highly performant serological tool for early diagnosis of hepatocellular carcinoma in patients with liver cirrhosis. J Clin Med. 2021 Jul;10(15):3392.
55. Singal AG, Tayob N, Mehta A, Marrero JA, El-Serag H, Jin Q, et al. GALAD demonstrates high sensitivity for HCC surveillance in a cohort of patients with cirrhosis. Hepatology. 2022 Mar;75(3):541–9.
56. Yang JD, Addissie BD, Mara KC, Harmsen WS, Dai J, Zhang N, et al. GALAD score for hepatocellular carcinoma detection in comparison with liver ultrasound and proposal of GALADUS score. Cancer Epidemiol Biomarkers Prev. 2019 Mar;28(3):531–8.
57. Caviglia GP, Abate ML, Petrini E, Gaia S, Rizzetto M, Smedile A. Highly sensitive alpha-fetoprotein, Lens culinaris agglutinin-reactive fraction of alpha-fetoprotein and des-gamma-carboxyprothrombin for hepatocellular carcinoma detection. Hepatol Res. 2016 Mar;46(3):E130–5.
58. Tayob N, Kanwal F, Alsarraj A, Hernaez R, El-Serag HB. The performance of AFP, AFP-3, DCP as biomarkers for detection of hepatocellular carcinoma (HCC): a phase 3 biomarker study in the United States. Clin Gastroenterol Hepatol. 2023 Feb;21(2):415–423.e4.
59. Freeman SC, Kerby CR, Patel A, Cooper NJ, Quinn T, Sutton AJ. Development of an interactive web-based tool to conduct and interrogate meta-analysis of diagnostic test accuracy studies: MetaDTA. BMC Med Res Methodol. 2019 Apr;19(1):81.
60. MetaDTA [Internet]. Diagnostic test accuracy mete-analysis v2.1.3. 2024 [cited 2025 Sep 26]. Available from: https://crsu.shinyapps.io/MetaDTA/
61. Patel A, Cooper N, Freeman S, Sutton A. Graphical enhancements to summary receiver operating characteristic plots to facilitate the analysis and reporting of meta-analysis of diagnostic test accuracy data. Res Synth Methods. 2021 Jan;12(1):34–44.
62. Goossens N, Bian CB, Hoshida Y. Tailored algorithms for hepatocellular carcinoma surveillance: is one-size-fits-all strategy outdated? Curr Hepatol Rep. 2017 Mar;16(1):64–71.
63. Ioannou GN. HCC surveillance after SVR in patients with F3/F4 fibrosis. J Hepatol. 2021 Feb;74(2):458–65.
64. Bengtsson B, Widman L, Wahlin S, Stål P, Björkström NK, Hagström H. The risk of hepatocellular carcinoma in cirrhosis differs by etiology, age and sex: A Swedish nationwide population-based cohort study. United European Gastroenterol J. 2022 Jun;10(5):465–76.
65. Girardin F, Tuch A, Eddowes L, Preisig M, Negro F. Scaling-up hepatitis C screening and treatment in Swiss outpatient psychiatric settings: A cost-effectiveness analysis. JHEP Rep. 2022 Mar;4(5):100464.
66. El-Serag HB, Jin Q, Tayob N, Salem E, Luster M, Alsarraj A, et al. HES V2.0 outperforms GALAD for detection of HCC: A phase 3 biomarker study in the United States. Hepatology. 2025 Feb;81(2):465–75.
67. Hou J, Berg T, Vogel A, Piratvisuth T, Torjan J, De Toni EN, et al., editors. GAAD and GALAD algorithmic scores demonstrate equivalent clinical performance for hepatocellular carcinoma (HCC) surveillance in prospective, multicenter, case-control studies. Proceedings of EuroMedLab; 2023 May 21-25; Rome, Italy.
68. Huang CF, Kroeniger K, Wang CW, Jang TY, Yeh ML, Liang PC, et al. Surveillance Imaging and GAAD/GALAD Scores for Detection of Hepatocellular Carcinoma in Patients with Chronic Hepatitis. J Clin Transl Hepatol. 2024 Nov;12(11):907–16.
69. Jamwal R, Krishnan V, Kushwaha DS, Khurana R. Hepatocellular carcinoma in non-cirrhotic versus cirrhotic liver: a clinico-radiological comparative analysis. Abdom Radiol (NY). 2020 Aug;45(8):2378–87.
The appendix is available in the pdf version of the article at https://doi.org/10.57187/4469.