Using the previously mentioned search criteria, 95 papers resulted.
Score exclusion criteria:
-
Prediction scores that depend on molecular and genetic risk factors which are expensive, not done routinely and were not available for patients in our cohort, like fat-genetic risk score [22], TLL1 [23, 24], IFNL3 [25, 26], MICA [27] and DEPDC5 [28].
-
Scores that needs data not routinely available in our dataset, as GGT (FIB4HCC score by Alonso López et al. [29] and Ganne-Carrié et al. [30]).
-
Scores that are depending on complicated mathematical methods like ElSerag et al. (HES) score [31] or the methods used for stratification of patients into risk groups was not fully reported in the published papers as Ioannou et al. score [32].
Accordingly, eleven scores and an algorithm were included in this study (Table 1).
Cohort characteristics:
The study included 3075 chronic HCV patients (1037 patients with F3 and 2038 with F4 stage) with SVR who met the inclusion criteria in our study between January 2014 and July 2019. Characteristics of the patients are shown in supplementary table 1. The mean observation period was 24.32 ± 12.14 months after the end of DAAs treatment (range 6–72 months). HCC developed in 212 cases during the study period. Characteristics of the patients according to the development of HCC are shown in supplementary table 2. Out of the 212 HCC cases identified during follow-up, 38 occurred during the first year of follow-up, 67 during the second year, 62 during the third year, and 26 during the fourth year, and 19 after the fourth year of follow-up.
Performance of eleven scores and an algorithm was compared and the results were presented in tables 2 and 3, and Fig. 1. Most scores stratified risk of HCC into three strata (low, intermediate and high), except two scores (Alonso López et al. TE-HCC score and Hiraoka et al. ADRES score) that stratified risk into four strata (very low, low, intermediate and high), and two scores (Hu et al. score and Abe et al. score) that stratified risk into two strata (low and high).
Sharma et al. THRI score:
Patients were classified into low risk group (766 patients, 24.9%), intermediate risk group (2072 patients, 67.4%) and high risk group (237 patients, 7.7%). HCC developed in 9 of patients belonging to low risk group with 5 years cumulative incidence of 0.51% (95% CI = 0.23–1.02), 165 belonging to intermediate risk group with 5 years cumulative incidence of 3.25% (95% CI = 2.73–3.84) and 38 belonging to high risk group with 5 years cumulative incidence of 6.47% (95% CI = 4.45–9.12). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was fair (0.6505). This score is considered statistically valid.
Watanabe et al. pre-treatment score:
Patients were classified into low risk group (676 patients, 22.0%), intermediate risk group (2077 patients, 67.5%) and high risk group (322 patients, 10.5%). HCC developed 6 of patients belonging to low risk group with 5 years cumulative incidence of 0.40% (95% CI = 0.10–1.09), 126 belonging to intermediate risk group with 5 years cumulative incidence of 2.42% (95% CI = 1.86–3.09) and 80 belonging to high risk group with 5 years cumulative incidence of 8.29% (95% CI = 5.95–11.25). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was fair (0.6406). This score is considered statistically valid.
Watanabe et al. post-treatment score:
Patients were classified into low risk group (1462 patients, 47.5%), intermediate risk group (97 patients, 39.9%) and high risk group (77 patients, 12.9%). HCC developed in 38 of patients belonging to low risk group with 5 years cumulative incidence of 1.10% (95% CI = 0.68–1.68), 97 belonging to intermediate risk group with 5 years cumulative incidence of 3.11% (95% CI = 2.33–4.08) and 77 belonging to high risk group with 5 years cumulative incidence of 7.58% (95% CI = 5.44–10.30). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was fair (0.6693). This score is considered statistically valid.
Hu et al. score:
Patients were classified into low risk group (1814 patients, 59.0%) and high risk group (1261 patients, 41.0%). HCC developed in 70 of patients belonging to low risk group with 5 years cumulative incidence of 1.70% (95% CI = 1.29–2.19), and 142 belonging to high risk group with 5 years cumulative incidence of 4.60% (95% CI = 3.80–5.51). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was fair (0.6158). This score is considered statistically valid.
Tani et al. score:
Patients were classified into low risk group (2178 patients, 70.8%), intermediate risk group (885 patients, 20.8%) and high risk group (12 patients, 0.4%). HCC developed in 94 of patients belonging to low risk group with 5 years cumulative incidence of 2.13% (95% CI = 1.73–2.59), 116 belonging to intermediate risk group with 5 years cumulative incidence of 6.48% (95% CI = 5.38–7.77) and 2 belonging to high risk group with 5 years cumulative incidence of 11.48% (95% CI = 1.93–37.93). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was fair (0.6388). This score is considered statistically valid.
Fan et al. aMAP score:
Patients were classified into low risk group (521 patients, 16.9%), intermediate risk group (1315 patients, 42.8%) and high risk group (1239 patients, 40.3%). HCC developed in 4 of patients belonging to low risk group with 5 years cumulative incidence of 0.35% (95% CI = 0.09–0.95), 38 belonging to intermediate risk group with 5 years cumulative incidence of 1.17% (95% CI = 0.79–1.68) and 170 belonging to high risk group with 5 years cumulative incidence of 5.40% (95% CI = 4.50–6.44). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was fair (0.6984). This score is considered statistically valid.
Hiraoka et al. ADRES score:
Patients were classified into very low risk group (504 patients, 16.4%), low risk group (1244 patients, 40.5%), intermediate risk group (1032 patients, 33.6%) and high risk group (295 patients, 9.6%). HCC developed in 6 of patients belonging to very low risk group with 5 years cumulative incidence of 0.54% (95% CI = 0.14–1.46), 36 of them belonging to low risk group with 5 years cumulative incidence of 1.18% (95% CI = 0.72–1.83), 96 belonging to intermediate risk group with 5 years cumulative incidence of 3.64% (95% CI = 2.72–4.79) and 74 belonging to high risk group with 5 years cumulative incidence of 9.08% (95% CI = 6.49–12.38). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was fair (0.7183). This score is considered statistically valid.
Abe et al. score:
Patients were classified into low risk group (2604 patients, 84.7%) and high risk group (471 patients, 15.3%). HCC developed in 136 of patients belonging to low risk group with 5 years cumulative incidence of 2.51% (95% CI = 2.00-3.10), and 76 belonging to high risk group with 5 years cumulative incidence of 6.43% (95% CI = 4.76–8.50). Log rank test for comparison of incidence curves is statistically significant (p = 0.013), and Harrell's c statistics was fair (0.5678). This score is considered statistically valid.
Shiha et al. GES score:
a - Pre-treatment score
Patients were classified into low risk group (1857 patients, 60.4%), intermediate risk group (719 patients, 23.4%) and high risk group (499 patients, 16.2%). HCC developed in 59 of patients belonging to low risk group with 5 years cumulative incidence of 1.66% (95% CI = 1.27–2.12), 71 belonging to intermediate risk group with 5 years cumulative incidence of 4.45% (95% CI = 3.50–5.57) and 82 belonging to high risk group with 5 years cumulative incidence of 7.64% (95% CI = 6.11–9.43). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was good (0.801). This score is considered statistically valid.
b - Post-treatment score
Patients were classified into low risk group (1764 patients, 57.4%), intermediate risk group (944 patients, 30.7%) and high risk group (367 patients, 11.9%). HCC developed in 46 of patients belonging to low risk group with 5 years cumulative incidence of 1.35% (95% CI = 1.00-1.79), 68 belonging to intermediate risk group with 5 years cumulative incidence of 3.49% (95% CI = 2.73–4.40) and 98 belonging to high risk group with 5 years cumulative incidence of 11.09% (95% CI = 9.05–13.45). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was good (0.828). This score is considered statistically valid.
Shiha et al. GES algorithm:
Patients were classified into low risk group (1574 patients, 51.2%), intermediate risk group (609 patients, 19.8%) and high risk group (892 patients, 29.0%). HCC developed in 37 of patients belonging to low risk group with 5 years cumulative incidence of 1.23% (95% CI = 0.88–1.67), 38 belonging to intermediate risk group with 5 years cumulative incidence of 2.93% (95% CI = 2.11–3.99) and 137 belonging to high risk group with 5 years cumulative incidence of 7.15% (95% CI = 6.03–8.43). Log rank test for comparison of incidence curves is highly statistically significant (p < 0.001), and Harrell's c statistics was good (0.832). This algorithm is considered statistically valid.
Chun et al. HCC-SVR score:
Patients were classified into low risk group (957 patients, 31.1%), intermediate risk group (1659 patients, 54.0%) and high risk group (459 patients, 14.9%). HCC developed in 31 of patients belonging to low risk group with 5 years cumulative incidence of 1.75% (95% CI = 1.21–2.45), 120 belonging to intermediate risk group with 5 years cumulative incidence of 3.52% (95% CI = 2.92–4.21) and 61 belonging to high risk group with 5 years cumulative incidence of 5.79% (95% CI = 4.47–7.39). Log rank test for comparison of incidence curves is not statistically significant (p = 0.053), and Harrell's c statistics was fair (0.6038).
Alonso López et al. TE-HCC score:
Patients were classified into very low risk group (490 patients, 15.9%), low risk group (848 patients, 27.6%), intermediate risk group (1236 patients, 40.2%) and high risk group (501 patients, 16.3%). HCC developed in 15 of patients belonging to very low risk group with 5 years cumulative incidence of 1.09% (95% CI = 0.44–2.27), 42 of them belonging to low risk group with 5 years cumulative incidence of 1.82% (95% CI = 1.10–2.85), 98 belonging to intermediate risk group with 5 years cumulative incidence of 2.90% (95% CI = 2.10–3.90) and 57 belonging to high risk group with 5 years cumulative incidence of 3.65% (95% CI = 2.37–5.39). Log rank test for comparison of incidence curves is not statistically significant (p = 0.374), and Harrell's c statistics was fair (0.5975).