HCC is often diagnosed at an advanced stage, the stage where therapeutic options are limited and mostly palliative. In contrast, the patients who are diagnosed at an earlier stage are candidates for curative treatments such as surgical resection, radiofrequency and liver transplantation. The screening of cirrhotic patients by ultrasound (every 6 months) is recommended by a large majority of experts and specialized societies to detect and treat HCC at an early stage. However, the impact of screening on patients’ survival remains controversial due to its evaluation by the methodological shortcomings of studies. Among the methodological uncertainties, the lead-time bias which corresponds to an extension of the follow-up time which is due to an earlier screening has not often been taken into account in studies that evaluated the impact of screening. The modeling approach is an attractive option for the evaluation of screening because the performance of a randomized controlled trial comparing the screened and the unscreened patients is impossible for ethical reasons. In this work, we developed a model of HCC progression from the date of diagnosis until death. This model takes into account the patient\\\\\\\\\\\\\\\\'s age, the awareness of HCV-status and the main prognostic factors for HCC in compensated and decompensated cirrhosis due to HCV. The model was supplied by several types of data to set the probabilities of transitions in it and validate its output. In the first step, in order to avoid an overestimation of the benefit of screening, it was necessary to calculate the lead-time bias and to include it in the calculation of the survival of screened patients. In a second step, the model evaluated the impact of screening for HCC in patients with compensated HCV-related cirrhosis and aware of their HCV-status. In the context of compensated cirrhosis with known HCV-status, our study showed that HCC screening performed in currently existing practices (rate of access to screening = 57%; effectiveness of screening corresponding to 42% of patients diagnosed at an early stage) improves the survival of patients with HCC, with an increase in life expectancy (LE) of 11 months and reduced HCC mortality at 5 years by 6% compared to the absence of screening (rate of access to screening = 0%; 19% of patients who had not been screened for HCC are diagnosed at an early stage). It showed the importance of the two variables, the rates of access to HCC screening and the effectiveness of screening on the survival of patients. Regarding the screening performed in currently existing practices: a) a scenario of increase of the rate of access to HCC screening from 57% to 97% which increases the LE by 7 months and reduces HCC mortality at 5 years by 5%; b) a scenario of increase of the effectiveness of HCC screening from 42% to 87% corresponding to that observed in the randomized trial CHC-2000 (optimal screening) which increases the LE by 14 months and reduces mortality at 5 years by 9%; c) a scenario combining an increase of the effectiveness of HCC screening from 42% to 87% and an increase of the rate of access to HCC screening from 57% to 97% which increases the LE by 31 months and reduces mortality at 5 years by 20%. This study emphasizes the need for strict application of rules of HCC screening in order to optimize its effectiveness to diagnose HCC at an early stage. This work suggests that experts should focus their recommendations on the effectiveness of screening. Such recommendations could lead to discuss the experience and qualification of operators and the quality of the park ultrasound used to standardize homogenize the quality of screening. Finally, we observed in a preliminary work that the optimal choice of the correction method for calculating the lead time bias should take into account the tumor progression from an asymptomatic stage to a symptomatic stage, which differs from a cancer to another.
