Study protocol for FASTRAK: a randomised controlled trial evaluating the cost impact and effectiveness of FAST-MRI for HCC suRveillance in pAtients with high risK of liver cancer

Abstract

Introduction

The surveillance of hepatocellular carcinoma (HCC) using semi-annual liver ultrasound (US) is justified in patients with cirrhosis. In this context, US has a low sensitivity (<30%) for the detection of HCC at the very early stage (ie, Barcelona clinic liver cancer (BCLC) 0, uninodular tumour <2 cm). The sensitivity of abbreviated liver MRI (AMRI) is reported to exceed 80%, but its use is hampered by costs and availability. Our hypothesis is that AMRI used as a screening examination in patients at high risk of HCC (>3% per year) could increase the rates of patients with a tumour detected at an early stage accessible to curative-intent treatment, and demonstrate its cost-effectiveness in this population.

Methods and analysis

The FASTRAK trial is a multicentre, randomised controlled trial with two parallel arms, aiming for superiority and conducted on patients at high risk for HCC (yearly HCC incidence >3%). Randomisation will be conducted on an individual basis with a centralised approach and stratification by centre. After inclusion in the trial, each patient will be randomly assigned to the experimental group (semi-annual US and AMRI) or the control group (semi-annual US alone). The main objective is to assess the cost/quality-adjusted life year and cost/patient detected with a BCLC 0 HCC in both arms. A total of 944 patients will be recruited in 37 tertiary French centres during a 36-month period and will be followed-up during 36 months.

Ethics and dissemination

The FASTRAK trial received ethical approval on 4 April 2022. Results will be disseminated via publication in peer-reviewed journals as well as presentation at international conferences.

Trial registration number

Clinical trial number (ClinicaTrials.gov) NCT05095714.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Abbreviated liver MRI is probably superior to ultrasound for hepatocellular carcinoma surveillance but is impacted by higher costs and low availability.

• This randomised controlled trial tests a new risk-based strategy comparing these two surveillance modalities.

• The primary end point is based on the calculation of incremental cost-utility ratio using costs evaluation and quality-adjusted life years.

• The duration of the trial will not allow the evaluation of long-term medico-economic end points which will be estimated over a lifetime using modelling.

Introduction

Hepatocellular carcinoma (HCC) is the sixth most common cancer and the fourth leading cause of cancer-related mortality in the world.¹ In 85% of patients, HCC occurs in the context of cirrhosis, the most common causes being excessive alcohol consumption, metabolic syndrome or infection with hepatitis B virus (HBV) and hepatitis C virus (HCV).² Patients with cirrhosis are recommended to be included in surveillance programmes involving semi-annual ultrasound (US) examinations to detect HCC eligible for curative-intent treatment (hepatic resection or percutaneous ablation). This practice is cost-effective when the yearly HCC incidence is >1.5%.³

French data have reported a median survival of 12 months for all stages of HCC.⁴ These statistics reflect the fact that only 20% of patients diagnosed with HCC eventually have access to a first-line curative-intent treatment, which is the only guarantee of extended survival (approximately 60% at 5 years). Early detection of HCC is, therefore, a major concern given the increasing incidence of the disease, especially due to the rise of metabolic syndrome related to obesity.⁵ Thus, evaluating the effectiveness, feasibility and cost-effectiveness of new methods for early detection of this tumour is crucial in this context.⁶

US has been reported to have relatively low sensitivity (estimated around 25%) for detecting very early stage HCC (ie, Barcelona clinic liver cancer (BCLC) stage 0: single tumour <2 cm).^7–11 Hepatic MRI has been assessed as a screening tool for HCC.¹² Its performance for detecting very early stage BCLC 0 HCC exceeds 80% in retrospective studies. However, due to the higher cost compared with US, it has been estimated that its use in screening would only be cost-effective in case of an annual HCC incidence >3%.¹³ The use of expensive and time-consuming MRI can be optimised through the use of ‘fast MRI protocols’ or ‘abbreviated MRI’ (AMRI), that is, short acquisition protocols (<10 min) based on sequences with the best detection sensitivities (Se >83%),¹⁴ and potentially without contrast injection.^{13 15}

We hypothesise that a cost-effective use of AMRI for screening in high-risk patients with HCC would increase the rates of very early stage HCC detection, allocation to curative-intent treatment and improved survival.

Methods and analysis

The FASTRAK trial is a multicentre, randomised controlled trial with two parallel arms, aiming for superiority and conducted on patients at high risk for HCC (yearly HCC incidence >3%) comparing surveillance using semi-annual US and AMRI versus semi-annual US alone. The main objective is to assess the cost/quality-adjusted life year (QALY) and cost/patient detected with a BCLC 0 HCC in both arms. A total of 944 patients will be recruited in 37 tertiary French centres during a 36-month period and will be followed-up during 36 months.

Patient and public involvement

Patients and public were not involved in the design of the trial. However, the FASTRAK trial is part of the GENIAL project (‘Understanding Gene ENvironment Interaction in ALcohol-related hepatocellular carcinoma’, HORIZON-MISS-2021-CANCER-02, https://www.genial-project.com) funded by the European Union (grant agreement ID: 101096312). In the setting of this project, patients will be included in scientific and ethical processes and in communication activities, including on the FASTRAK trial. A specific task milestone is added in each work package (WP) to track patient involvement.

Selection of sample

The improvement of HCC screening programmes will come with increased costs and must consider the evolving epidemiology of chronic liver diseases and HCC. This is particularly relevant for patients infected with HCV or HBV and with sustained virological response in the case of HCV or maintained virological response in the case of HBV. In these patients, the incidence of HCC has significantly decreased following the widespread implementation of direct-acting antiviral treatments or nucleos(t)ide analogues.^16–18 Large prospective cohorts of patients with cirrhosis in Europe and the USA have reported an overall annual incidence of HCC between 1.5% and 2.5% in these patients, which is similar to the incidence observed in patients with non-viral causes of cirrhosis, whether related to alcohol or metabolic syndrome.^{10 19} In this context, identifying patients with non-viral causes of cirrhosis or those with virological cure/control who have a high incidence of HCC and for whom strengthening screening programmes would be cost-effective is crucial to improve the prognosis of this cancer.

A key parameter in the cost-effectiveness calculation is the incidence of HCC in the population under consideration for screening because it determines the total expenditure on AMRI, the implementation of curative-intent treatment and overall survival in at-risk patients for this cancer. Based on data from French multicentre prospective cohorts of patients with cirrhosis enrolled in surveillance programmes (HCC 2000, ANRS CO12 CirVir, CIRRAL and ANRS CO22 Hepather),10 11 20 21 we have developed several prediction models stratifying for the risk of HCC.^{22 23} The most recent one²⁴ developed in >4000 patients with cirrhosis is a ‘universal’ algorithm applicable regardless of the cause of cirrhosis that takes into account the decrease in HCC incidence in patients with viral cirrhosis due to advances in antiviral treatments^{16 17} (table 1). This score helps identify patients with an annual HCC incidence over 3% (score >9) for more precise targeting during inclusion in screening-focused clinical trials. This high-risk population represented 35% of the whole studied population.

Table 1.

HCC risk score in French patients with cirrhosis²⁴

Variables	aSHR (95% CI)	P value	Coefficient (ln(SHR))	Score (×3.5)
Age (years)		<0.001
≤60	Reference			0
60; 65	1.92 (1.21 to 3.06)	0.006	0.6548859	2
>65	2.96 (2.02 to 4.32)	<0.001	1.084074	4
Male sex	1.86 (1.23 to 2.79)	0.003	0.6180754	2
Platelet count ≤120/mm3	1.81 (1.25 to 2.62)	0.002	0.5935341	2
GGT >1.5 ULN	2.16 (1.46 to 3.20)	<0.001	0.7692455	3
Total bilirubin >12 µmol/L	1.63 (1.13 to 2.36)	0.010	0.4897013	2
AFP >5 ng/mL	2.53 (1.75 to 3.65)	<0.001	0.9268662	3

APF, alpha-fetoprotein; aSHR, adjusted standard hazard ratio; GGT, gamma-glutamyl transferase; HCC, hepatocellular carcinoma; ULN, upper limit of normal.

In parallel, we conducted a health economic evaluation using a previously published Markov model, taking into account French costs.²⁵ For an incidence rate exceeding 3%, MRI surveillance was shown to be cost-effective in the French context with an incremental cost‐effectiveness ratio of €15 447/life years.²⁴

Based on these analyses, we designed the FASTRAK trial to assess the cost/QALY and cost/patient detected with a very early stage (BCLC 0) HCC by semi-annual liver US and AMRI, compared with conventional semi-annual monitoring by liver US alone in patients with cirrhosis followed-up in tertiary centres and an anticipated HCC incidence >3%.

Selection criteria

Inclusion criteria

(1) Histologically proven Child-Pugh A or B cirrhosis or cirrhosis unequivocally suggested by non-invasive examinations; (2) no evidence of HCC on imaging within the last 3 months; (3) hepatic parenchyma assessable by US; (4) non-viral cirrhosis or controlled/cured viral B/C cirrhosis; (5) estimated annual HCC risk >3% using the aforementioned clinical scoring system; (6) signed informed consent.

Exclusion criteria

(1) Child-Pugh C score; (2) active HBV or HCV; (3) estimated annual HCC risk <3%; (4) contraindications for undergoing MRI; (5) non-echoic liver.

Recruitment methods

Patients with cirrhosis will be recruited during hepatology consultations. They will be informed about the study by the principal investigator of each centre or one of the co-investigators, and informational materials will be made available to inform the subjects in waiting areas or during interviews (posters, information leaflets).

Interventions

Semi-annual hepatic ultrasound

A non-invasive examination focused on the liver parenchyma to identify hepatic nodules that may indicate HCC in the context of cirrhosis.

Semi-annual abbreviated (or fast) MRI and liver US

AMRI without contrast injection, using diffusion-weighted sequences, fat saturation T1-weighted and T2-weighted sequences associated with liver US.

Sample size calculation

Concerning the clinical efficacy criterion, based on published data, it is anticipated that approximately 28%–30% of HCC cases will be detected at a very early stage with the usual US strategy, while an expected 60% detection rate is anticipated for the AMRI strategy. With a bilateral alpha risk of 5%, the inclusion of 944 patients with high-risk HCC (n=821+15% expected follow-up losses) followed for 3 years with an expected annual HCC incidence of ≥3% (n ≥74 expected detected HCC cases) will enable the identification of a significant difference with a minimum power of 80%.

From a medical-economic perspective, only one study has been published¹³ regarding costs and quality of life expressed in QALYs associated with using liver MRI compared with US for semi-annual surveillance of patients with high-risk HCC. This study, conducted in South Korea, primarily with patients with HBV-related cirrhosis, allows us to anticipate a cost difference of approximately €800 (±€150) and a QALY difference of 0.04 (±0.15) over a 3-year timeframe. Based on this, and using Glick’s formula²⁶ assuming a maximum willingness to pay per QALY of €100 000 (which is conservative, as an article from 2020 suggests values of €150 000–€200 000 for France), the most unfavourable possible correlation between cost and effect (rho=−1), and a bilateral alpha risk of 2.5% (Bonferroni approach for the double medical-economic criterion analysis according to EQ5D or EORTC-QLQC30), the sample size calculation described earlier for clinical efficacy will enable the validation of the cost-utility ratio with a minimum power of 80%.

To this aim, 944 patients will be recruited in 37 tertiary French centres during a 36-month period and will comprise a 36-month follow-up (table 2).

Table 2.

Recruitment and timelines

Number of anticipated patients	944
Number of recruiting centres	37
Inclusion period (months)	36
Follow-up (months)	36
Number of patients/centre	25.51
Number of patients/centre/month	0.70
First included patient	October 2022
End of trial	October 2028

Outcomes

Main research objective

The main objective of the research is to evaluate the cost per QALY for patients detected with a BCLC 0 HCC through semi-annual surveillance using liver US and AMRI compared with traditional semi-annual surveillance with liver US alone.

Secondary research objectives

• Compare the percentage of very early stage HCC detected based on the surveillance strategy.

• Confirm the metrological properties of AMRI for non-invasive detection of very early stage HCC <2 cm.

• Evaluate the modalities and determinants of access to first-line curative-intent treatments.

• Assess the determinants of overall survival and recurrence-free survival after curative-intent treatment.

• Describe patient acceptability and compliance with the surveillance programme and assess its determinants.

Research design

Primary outcome measure

The primary outcome measure is the incremental cost per QALY ratio. This will assess the medical-economic efficiency by measuring the difference in the total costs for each study arm and the difference in quality of life using the EQ-5D-5L and French values. The primary outcome measure is the difference in costs divided by the difference in QALYs.

Secondary outcome measures

The secondary outcome measures of the research include at a 36-month timeframe:

• Percentage of very arly BCLC 0 HCC detected every 6 months.

• Sensitivity, specificity, positive and negative predictive values for diagnostic methods.

• Proportions of treatments performed with curative intent.

• Overall survival and progression-free survival.

• Follow-up of surveillance programmes, including the scheduled appointments and examinations.

Experimental design

Figure 1 shows the trial design. This is a multicentre, randomised controlled trial with two parallel arms, aiming for superiority and conducted in high-risk patients for HCC. Randomisation will be conducted on an individual basis with a centralised approach and stratification by centre. After inclusion in the trial, each patient will be randomly assigned to the experimental group (semi-annual liver US and AMRI) or the control group (semi-annual liver US alone). These examinations will be performed at the expert centre responsible for patient follow-up.

Figure 1.

FASTRAK trial design. BCLC, Barcelona clinic liver cancer; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; QALY, quality-adjusted life year; US, ultrasound.

A patient will be considered free of nodules during each semi-annual visit if neither the US nor the AMRI detects any nodules. If either of the two examinations detects a nodule, it will undergo a characterisation process following the recommendations of Western scientific societies,³ using a combination of contrast-enhanced imaging and liver biopsy. The diagnosis of HCC will be definitively determined in each centre during a multidisciplinary discussion meeting.

The experimental group follows the standard management (semi-annual liver US) with the addition of semi-annual fast-MRI (or AMRI) of the liver. The control group follows the standard management of semi-annual liver US alone.

Research timeline

Table 2 and figure 2 show patients recruitment details and timelines.

Figure 2.

Participant timeline.

• Maximum duration between selection and inclusion: 1 month.

• Duration of the inclusion period: 36 months.

• Duration of each subject’s participation: 36 months.

• Total research duration: 72 months.

• Timing of randomisation: inclusion visit.

• Duration and follow-up procedures for subjects in case of premature termination: until the planned 36 months of follow-up.

Data analysis plan

General principles

Descriptive statistical analyses will be conducted to assess the randomisation groups regarding demographic and clinical characteristics, rates of detected HCC and associated tumour characteristics. Continuous variables will be presented as means (±SD) or medians (with IQRs) depending on the normality of variable distributions as assessed by the Shapiro-Wilk test, while categorical variables will be presented as counts (%).

As appropriate, comparative analysis of binary parameters, including the primary outcome, will be performed using the χ² test or Fisher’s exact test. In addition to univariate analyses, multivariable logistic regression models will be constructed to identify other potential predictive factors associated with the detection of early stage HCC and/or potential effect-modifying factors by introducing interaction terms between the randomisation group variable and covariates corresponding to prespecified subgroups (eg, based on viral aetiology B or C or non-viral aetiology of cirrhosis).

Comparative analysis of quantitative outcome values will be conducted using independent samples t-test or the non-parametric Mann-Whitney U test, depending on conditions and the normality of variable distributions. Multiple linear regression models will be tested to identify independent determinants.

Analyses related to the time-to-event outcomes based on censored follow-up data will rely on the log-rank test and the construction of survival curves using the Kaplan-Meier method (univariate analysis). Multivariate Cox proportional hazards models will be constructed for overall survival (all-cause mortality), progression-free survival (progression of HCC or all-cause mortality) and models based on the Fine and Gray regression method (competitive risks) for the analysis of cumulative HCC incidence taking into account the competitive risk associated with death.

The confirmatory evaluation of the diagnostic performance of AMRI for the detection of BCLC 0 HCC will be based on the calculation of the following diagnostic indices: positive predictive value (true positives (TP)/(TP+false positives)) and, depending on the criteria used to define true negatives and false negatives, negative predictive value, sensitivity, specificity, accuracy, Youden’s index and positive and negative likelihood ratios.

The primary outcome analysis will be conducted on the intention-to-diagnose (ITD) population. Additional analyses of the primary outcome and all secondary outcome analyses will be performed on the ITD and per-protocol (PP) populations without major protocol deviations to describe the patients excluded from the PP population and assess the impact on the ITD analysis and the robustness of the results. The ITD population will consist of patients randomised into one of the two study arms who have not withdrawn their consent. The ITD population will be analysed according to the initial allocation arm. The PP population will consist of patients included in the study who did not have major protocol deviations, including modifications, lack of compliance or the incorrect inclusion of patients who did not meet all the inclusion/exclusion criteria.

The analyses of the primary and secondary outcomes will be performed at a two-sided alpha level of 5%. All analyses will be conducted using Stata V.16.1 (StataCorp, College Station, Texas, USA) and R V.4.0.2 (R Foundation, Vienna, Austria) within the Department of Public Health at Henri Mondor Hospital, under the responsibility of Professor EA.

Medical-economic evaluation analyses

The primary end point is the incremental cost-utility ratio, calculated as the ratio between the difference in costs and QALYs between the two strategies. It will be estimated at 3 years based on trial data and over a lifetime using modelling.

• Trial-based evaluation

Costs will be estimated over the trial 3-year period for the entire population of patients included, from the perspective of the healthcare system. Patient-level data on resource use will be collected via the study electronic case report form (eCRF), complemented if necessary by queries in the hospitals’ computerised records and claims database. Resources will be valued using the prices relevant to the social health insurance for tests and professional services and the severity-adjusted diagnosis-related group costs for hospital admissions.

Health-related quality of life will be collected using EQ-5D5L self-administered questionnaires. Questionnaires will be filled at each follow-up visit up to 3 years. The utility values will be attributed to the time period corresponding to mid-point between data collection. We anticipate that the effect of transfusion will decline rapidly after 3 months.

The difference in 3-year QALYs will be estimated as the difference in the area between the utility curves for the two treatment groups.

A secondary effectiveness criterion is the percentage of BCLC 0 HCC detected. The results will be presented as a mean and 95% CI. Deterministic and probabilistic sensitivity analyses will be performed for this effectiveness criterion.

Costs, life-years and QALYs will be presented as means with 2.5%–97.5% bootstrapped intervals. Between-group comparisons of costs will be performed using the bootstrap t-test. Between-group comparisons of effects will be performed using non-parametric testing. Costs and QALYs are discounted 2.5% yearly as requested by French guidelines.

The trial-based cost-effectiveness ratios are defined as the difference in costs/difference in QALYs and difference in costs/difference in HCC detected.

The uncertainty in these ratios will be measured by:

1. Deterministic analysis, particularly regarding adherence, the percentage of feasible curative-intent treatments and the type of curative-intent treatment implemented.

2. Probabilistic analysis using clinical trial data at 3 years.

To perform probabilistic sensitivity analyses on clinical trial data, resampling with replacement (bootstrap: 5000 resamples) will be conducted. The average cost and effectiveness in each arm will be calculated for each sample and presented on a cost-effectiveness plane. Acceptability curves will also be generated, showing the probability that AMRI use is cost-effective as a function of the societal willingness to pay.

• Model-based extrapolation

We will use a standard Markov model, previously developed for other economic evaluation of HCC surveillance.²⁵

Costs inputs for the model will use the state specific costs estimated for our previous analysis and updated. State-specific utilities will be extracted from the literature and reflect as much as possible French or Wester European values. We will use trial data to estimate the earlier transition probabilities (eg, nodule to HCC) and literature data for later transitions. New states will be added based on state-of-the-art treatments and international recommendations. The model outcome is the incremental cost-utility ratio.

Plausible parameter distributions will be determined from the literature to perform probabilistic sensitivity analyses on model data. In particular, costs will be modelled using a gamma distribution, and probabilities will be modelled using a beta distribution. Survival will be extrapolated using exponential, Weibull, lognormal and Gompertz distributions, with parameters determined from survival data during the trial. The distribution of survival parameters will follow the resampling of patient data from the trial.

• A budget impact analysis at 5 years will extrapolate the study results with assumptions about service choices, patient adherence and epidemiological data. The primary data sources for the BIA will be trial results. National statistical information will be used to estimate the size of the target population. The cost of illness will be based on the yearly calculation of the social health insurance, hospital claims data combined with the population size. Alternative scenarios chosen from the perspective of the decision-maker will be presented.

Ethics and dissemination

Table 3 shows the Trial registration data. National Ethics Committee: Comité de protection des personnes Sud-Est VI (registration number: 2021-A03037-34, date of approval: 4 April 2022).

Table 3.

Trial registration data

Data category	Information
Primary registry and trial identifying number	ClinicaTrials.gov: NCT05095714
Secondary identifying numbers	Promotion research code: APHP21098/N° IDRCB: 2021-A03037-34
	National Ethics Committee: Comité de protection des personnes Sud-Est VI (registration number: 2021-A03037-34, date of approval: 4 April 2022)
Date of registration in primary registry	27 October 2021
Source(s) of monetary or material support	The trial is funded by a grant from Programme de Recherche Médico-Économique (PRME) 2020 (Ministry of Health)
Sponsor	The trial sponsor is Assistance Publique-Hôpitaux de Paris (APHP), Delegation for Clinical Research and Innovation
Contact for public/scientific queries	Pierre Nahon, MD, PhD Service d’Hépatologie, Hôpital Avicenne, 93 000 Bobigny, France Email: pierre.nahon@aphp.fr
Public title	FAST-MRI for HCC suRveillance in pAtients with high risK of liver cancer (FASTRAK)
Scientific title	FASTRAK: a randomised controlled trial evaluating the cost impact and effectiveness of FAST-MRI for HCC suRveillance in pAtients with high risK of liver cancer
Countries of recruitment	France
Health condition(s) or problem(s) studied	HCC surveillance in high-risk patients with cirrhosis (annual incidence >3%)
Intervention(s)	Intervention: semi-annual AMRI without contrast injection, using diffusion-weighted sequences, fat saturation T1-weighted and T2-weighted sequences associated with liver US
	Control: semi-annual liver US
Key inclusion and exclusion criteria	Inclusion criteria: (1) Child-Pugh A or B cirrhosis; (2) hepatic parenchyma assessable by ultrasound; (3) non-viral cirrhosis or controlled/cured viral B/C cirrhosis; (4) estimated annual HCC risk >3%. Exclusion criteria: (1) Child-Pugh C score; (2) active hepatitis B or C; (3) estimated annual HCC risk <3%; (4) contraindications for undergoing MRI; (5) non-echoic liver
Study type	Interventional
	Multicentre randomised controlled trial with two parallel arms aiming for superiority
	Primary outcome: incremental cost per QALY ratio

AMRI, abbrevaiated MRI; HCC, hepatocellular carcinoma; QALY, quality-adjusted life year; US, ultrasound.

Project structure and participant engagement

This project will involve university hospitals with cirrhosis surveillance programmes for HCC. Each centre will have a lead hepatologist and a lead radiologist as principal investigators. Only hepatologists will have the authority to include patients. Inclusion and follow-up will take place during specialised consultations while screening examinations will be conducted in outpatient care settings.

The principal investigator or a designated collaborating physician, listed on the function delegation form and trained in research, will inform individuals participating in the research to obtain their consent. A hepatologist will provide the information during the selection visit (conducted as part of medical care). The patient will have a necessary and sufficient reflection period (with a maximum duration of 1 month corresponding to the period between the selection visit and inclusion) before deciding on participation.

The hepatologist will inform the patient about the study’s objectives, constraints and foreseeable risks. An information sheet and a consent form will be provided. Before any specific research-related tests, the patient will receive answers to their questions from the hepatologist. If the patient consents to participate, the patient and the hepatologist will write their names and sign the consent form. A copy of this form will be given to the patient, the hepatologist will keep another and the trial sponsor will retain a third.

All medical information will be entered and monitored in an eCRF (the CleanWeb Telemedicine Technologies 2007 eCRF developed by the research unit at Henri Mondor University Hospitals (URC-HMN)). All changes in the protocol will be the focus of an amendment and communicated to all centres.

Description of measures taken to reduce and avoid biases

Subject identification: in this research, subjects will be identified using a unique reference format: centre number (three numeric positions)—person’s selection order in the centre (four numeric positions)—initials of last name—initials of first name. This unique reference will be maintained throughout the research. Recruiting patients already enrolled in HCC screening programmes, who are presumably compliant, will help limit the number of lost-to-follow-up cases (estimated at 15% in sample size calculations).

Randomisation: patients meeting all selection criteria and providing consent will be randomised using the eCRF by the investigator during the inclusion visit. The research unit at URC-HMN will develop the CleanWeb Telemedicine Technologies 2007 eCRF and integrate the randomisation list. This list will be generated by the study statistician’s computer under the responsibility of Professor EA. Randomisation will be individualised using a 1:1 allocation ratio, centrally managed and stratified based on the centre. These measures aim to ensure a systematic and unbiased selection and allocation process for subjects participating in the research. The unique identification system and the randomised allocation method contribute to minimising potential biases and ensuring the integrity of the study’s findings.

Method for addressing missing, unused or invalid data

Any missing or invalid data will be systematically sought for verification in the patient’s medical record. In addition to analyses of complete cases without missing data for the primary outcome, sensitivity analyses will be conducted using several methods to handle missing data. These methods include last observation carried forward, worst-case scenario analysis and multiple imputation techniques using chained equations. Additionally, no plans exist to replace subjects who prematurely exit the study.

Summary of anticipated benefits and known risks for individuals participating in the research

Risks and constraints: the potential benefits for patients undergoing the proposed AMRI in addition to routine US include improved clinical outcomes, a more cost-effective approach to managing early detected HCC and potential enhancements in the healthcare system’s efficiency. However, the primary risk or constraint involves the requirement for an additional biannual AMRI, which could pose challenges for patients, especially those prone to claustrophobia or discomfort during MRI procedures.

Publication rules

The results of this study will be published in peer-reviewed journals, as well as presented at national and international conferences. Pertinent results will be shared with participating institutions prior to publication.

Affiliation mention for APHP promoted projects

• If an author has multiple affiliations, the order in which institutions are listed (APHP, University, INSERM, etc) does not matter.

• However, if the research is funded through an internal call for proposals from APHP, the primary affiliation should be ‘APHP’.

• Each of these affiliations should be identified by an address separated by a semicolon (;).

• The institution APHP should appear as ‘APHP’ first in the address, followed precisely by APHP, hospital, department, city, postal code and France.

Mention of APHP sponsor (DRCI) in manuscript acknowledgements: ‘The sponsor was Assistance Publique-Hôpitaux de Paris (Delegation for Clinical Research and Innovation)’.

Mention of the funder in manuscript acknowledgements: ‘The study was funded by a grant from Programme de Recherche Médico-Économique (PRME) 2020 (Ministry of Health)’.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s).