If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Center for Global Health, Istituto Superiore Di Sanità (ISS), Rome, ItalyUniCamillus-Saint Camillus International University of Health Sciences, Rome, Italy
Infectious and Tropical Medicine Unit, Department of Public Health and Infectious Diseases, "Policlinico Umberto I" Hospital, Sapienza University of Rome, Rome, Italy
Internal Medicine and Hepatology Division, Department of Medicine, Surgery and Dentistry, “Scuola Medica Salernitana", University of Salerno, Baronissi, Italy
Prof. L. Chemello, Department of Medicine-DIMED, Padua University, University Hospital, Clinica Medica 5, Refering Regional Center for Liver Diseases, Via Giustiniani, 2 - 35128 Padova, Italy.
Severe liver disease markers assessed before HCV eradication are acknowledged to usually improve after the SVR. We prospectively evaluated, in the PITER cohort, the long-term HCC risk profile based on predictors monitored after HCV eradication by direct-acting antivirals in patients with cirrhosis.
Methods
HCC occurrence was evaluated by Kaplan-Meier analysis. Cox regression analysis identified the post-treatment variables associated with de-novo HCC; their predictive power was presented in a nomogram.
Results
After the end of therapy (median follow-up:28.47 months), among 2064 SVR patients, 119 (5.8%) developed de-novo HCC. The HCC incidence was 1.90%, 4.21%, 6.47% at 12-, 24- and 36-months from end-of-therapy, respectively (incidence rate 2.45/100 person-years). Age, genotype 3, diabetes, platelets (PLT)≤120,000/µl and albumin ≤3.5g/dl levels were identified as pre-treatment HCC independent predictors. Adjusting for age, the post-treatment PLT≤120,000/µl (AdjHR 1.92; 95%CI:1.06-3.45) and albumin≤3.5g/dl (AdjHR 4.38; 95%CI 2.48-7.75) values were independently associated with HCC occurrence. Two different risk profiles were identified by combining long-term post-therapy evaluation of PLT ≤ vs. >120,000/µl and albumin ≤ vs. >3.5g/dl showing a significant different HCC incidence rate of 1.35 vs. 3.77/100 p-y, respectively.
Conclusions
The nomogram score based on age, PLT and albumin levels after SVR showed an accurate prediction capability and may support the customizing management for early HCC detection.
]. The risk attributable to hepatitis C virus (HCV) infection is decreased by the achievement of a sustained virological response (SVR) after interferon (IFN) and DAA treatments. Still, clearance of HCV infection does not eliminate the risk of HCC occurrence when advanced fibrosis or cirrhosis was present before treatment [
]. Factors involved in these differences have mostly been analyzed retrospectively (i.e., length of pre- or post-treatment follow-up (FU), mean age, liver disease stage or biochemical surrogates, and characteristics of undefined nodules). Albeit factors relating to the stage of liver disease have often been linked to the higher risk of developing HCC, no apparent risk profile widely applicable in the clinical practice has emerged [
]. This may be partly explained because the above items were assessed before SVR, while it is now acknowledged that the disease severity parameters usually improve following SVR [
]. Since patients with cirrhosis remain susceptible to developing HCC after the SVR attainment, the identification of risk profiles applicable at the individual patient's level will help customize management policies.
Some studies have investigated the improvements in surrogate features of liver fibrosis and portal hypertension after HCV eradication, documenting different categories of HCC incidence and some predictive factors associated with an unfavorable outcome [
]. However, real-life data focusing on these potential predictors, when evaluated prospectively at long-term after achieving the SVR in real-life practice, are scarce.
Drawing from the experience of the PITER platform [
], which collected a prospective multicenter real-life cohort representative of HCV patients in care in Italy, we prospectively assessed potential predictive factors associated with HCC occurrence after achieving the SVR by DAA therapy in patients with cirrhosis. The final goal was to make available to clinicians a simple nomogram plotting the likelihood of an individual patient with advanced liver disease being HCC-free at specific time points during FU after SVR.
2. Patients and methods
PITER is an ongoing platform that enrolled and prospectively evaluated a representative cohort of chronic hepatitis C patients, since June 2014, from 60 specialized centers of liver and infectious diseases distributed throughout Italy [
Data were collected in dedicated electronic Case Report Forms (eCRFs), at different time points: pre-DAA treatment (at baseline) and during the long-term FU, according to each clinical center real-life practice.
Liver cirrhosis was defined either by liver biopsy (Metavir or Ishak score) or by Transient Elastometry using a liver stiffness measurement (LSM) cut-off >12.5 kPa [
]. In the lack of these procedures, it was agreed among clinical centers by the attendance of biochemical and/or clinical and instrumental features of liver cirrhosis, or portal hypertension. The Child-Pugh-Turcotte (CPT) class and the Fibrosis-4 (FIB-4) index, as non-invasive estimation tools of liver fibrosis, were generated automatically in the eCRF of each patient enrolled.
The presence of liver steatosis by abdominal ultrasound associated with hypertension or cardiovascular disease, type 2 diabetes, and BMI >25 kg/m2 were considered surrogate markers of the metabolic syndrome in patients with no alcohol consumption.
2.1 Inclusion and exclusion criteria
All consecutively enrolled patients diagnosed with liver cirrhosis who achieved the SVR by DAA therapy were evaluated. Patients with a compensated CPT-A or B class or with a previous decompensation history, but showing a stable clinical condition with compensation within the last 12-month period were included.
Patients with a previous diagnosis of HCC or pre-treatment unclassified nodular lesions, CPT-C class, orthotopic liver transplantation (OLT) and those on the OLT waiting list were excluded. The FU period of the study started from the EOT evaluation.
2.2 Hepatocellular carcinoma occurrence
HCC surveillance was performed in all patients by ultrasound (US) examination every six months. In addition, HCC was diagnosed by needle biopsy or by non-invasive criteria according to international guidelines [
]. Only HCC occurring after the EOT was considered as de-novo HCC.
2.3 Statistical analysis
Patients' main characteristics were reported as the median and interquartile range (IQR) for continuous or as proportions for categorical variables. Albumin level, PLT count and LSM were evaluated using clinically relevant threshold cut-offs. The Mann-Whitney U test, the Chi-squared test and the Wilcoxon matched-pairs signed-rank test were used as appropriate.
HCC occurrence was evaluated using Kaplan–Meier survival analysis. The log-rank test was used to identify differences in survival among groups.
Cox regression analyses were used to identify variables independently associated with HCC incidence adopting a stepwise forward selection. Gender variable was forced to be included in the model. PLT count ≤120.000/ µl was used, as a cut-off for HCC risk discrimination based on better performance associated with albumin level ≤3.5 g/dl in our study population, as also previously reported [
A further model was then constructed to explore the association of HCC incidence and variables related to the severity of liver damage individuated by the previous model and subjected to changes after achieving SVR. This predictive model was presented as a nomogram by a graphical representation of the probability of HCC free-survival for each patient at 1-, 3- and 5-years of FU, according to calculated points associated with each risk factor evaluated. The predictive accuracy of the models was measured through the concordance index (Harrell's c-index), quantifying the level of agreement between the predicted probabilities and the actual possibility of having the event of interest. The bootstrap resampling method was chosen to internally validate the predictive models' selecting a 500 repetition. The calibration curve was used to assess the goodness of the model fit. A p-value <0.05 was considered statistically significant. Statistical analyses were performed with STATA version 16.1 (StataCorp, College Station, TX, USA) and R software (version 4.1.2, Regression Modelling Strategies package. The R Foundation for Statistical Computing, Vienna, Austria).
2.4 Ethics
The study was conducted following the Declaration of Helsinki guidelines and the Good Clinical Practice principles. The study protocol was approved by the Ethics Committee of the Istituto Superiore di Sanità and of each participating center. Patients' data were evaluated anonymously, adopting generated codes. All patients gave their written informed consent to participate in the study.
3. Results
3.1 Pre-treatment characteristics and HCC incidence rate in SVR patients by DAA therapy
The study population consisted of 2064 patients with liver cirrhosis who achieved the SVR following the DAA treatment of whom 1368 (66%) were diagnosed by liver stiffness >12.5kPa and/or liver biopsy and 696 (34%) by biochemical and/or clinical or instrumental findings of cirrhosis or portal hypertension. Pre-treatment characteristics of patients, who achieved the SVR, followed up for a median time of 28.47 months (IQR 13.33-41.61 months) are shown in Table 1. One-hundred and nineteen patients (5.7%), who developed HCC, were older and reported more frequently clinical features of severe liver disease (lower PLT or albumin levels, LSM ≥20 kPa, FIB-4 >3.25 or showed a CPT class B or previous episode of decompensation), diabetes or a SOF-based treatment schedule use.
Table 1Pre-treatment characteristics of DAA successfully treated patients by HCC occurrence.
All patients did not have sign of liver decompensation at treatment start.
204
10.5
21
17.6
0.015
225
10.9
Treatment regimen
.
Ribavirin use
1119
57.5
76
63.9
0.174
1195
57.9
SOF-based treatment
1499
77.1
102
85.7
0.028
1601
77.6
Abbreviations: BMI, Body Mass Index; DAA, Direct Acting Antiviral; HBV, FIB-4, Fibrosis-4; Hepatitis B Virus; IQR, Interquartile Range; SOF, Sofosbuvir.
Inconsistencies are due to missing values.
p value Mann–Whitney rank-sum test.
p value Chi-square test.
§ All patients did not have sign of liver decompensation at treatment start.
The overall HCC-IR in patients who achieved SVR was 2.45/100 person-years (p-y). The 12-, 24- and 36-month cumulative HCC-IR was 1.90%, 4.21%, and 6.47%, respectively (Fig. 1A). The 12, 24- and 36-month HCC-IR in class CPT-A and CPT-B cases was 1.63%, 3.70%, and 5.97% and 3.41%, 7.13%, and 9.33%, respectively (log-rank test p=0.0227) (Fig. 1B).
Fig. 1Cumulative incidence of HCC development in patients with SVR after DAA therapy. (A) Cumulative incidence curve in patients with SVR. (B) Cumulative incidence curves in patients with SVR stratified for CPT-A and CPT-B class. Abbreviations: CPT, Child-Pugh-Turcotte; DAA, Direct Acting Antiviral; HCC, Hepatocellular Carcinoma; SVR, Sustained Virologic Response.
3.2 Pre-treatment predictive factors of HCC development
Pre-treatment variables independently associated with HCC development in patients who achieved the SVR are shown in Table 2. After adjusting for the confounding effect of each variable by stepwise forward selection: age (Adjusted Hazard Ratio (AdjHR) 1.06, 95%CI 1.04-1.09), HCV genotype 3 (AdjHR 4.27, 95%CI 2.22-8.23), PLT ≤120,000/µl (AdjHR 1.67, 95%CI 1.05-2.65), albumin ≤3.5 g/dl (AdjHR 2.51, 95%CI 1.62-3.88) and diabetes (AdjHR 1.75, 95%CI 1.10-2.76) were variables independently associated with HCC occurrence.
Table 2Pre-treatment predictive factors associated with HCC occurrence in DAA successfully treated patients at baseline. Univariate and multivariate analysis.
Cox forward stepwise selection. Gender variable was forced to be included in the model. Statistically significant hazard ratios and related 95% confidence intervals are reported in bold.
95% CI
p value
Age (increasing years)
1.04
1.02 - 1.06
<0.001
1.06
1.04 - 1.09
<0.001
Gender (ref. female)
1.32
0.91 - 1.91
0.139
1.39
0.89 - 2.17
0.143
BMI: overweight (ref. under-normalweight)
0.91
0.62 - 1.33
0.615
obese (ref. under-normalweight)
0.86
0.49 - 1.51
0.596
Alcohol use: current (ref. never)
1.38
0.76 - 2.48
0.290
past (ref. never)
1.36
0.89 - 2.08
0.157
HCV-genotype (3 vs others)
1.49
0.87 - 2.57
0.146
4.27
2.22 - 8.23
<0.001
Anti-HBc+
1.14
0.74 - 1.76
0.542
1.58
0.98 - 2.55
0.058
Previous Interferon treatment
1.11
0.78 - 1.60
0.563
Platelets count (ref. >120,000/µL)
1.95
1.30 - 2.90
0.001
1.67
1.05 - 2.65
0.029
Albumin level (ref. > 3.5 g/dL)
3.08
2.13 - 4.46
<0.001
2.51
1.62 - 3.88
<0.001
LSM (ref. < 20 kPa)
1.50
0.98 - 2.30
0.063
Past decompensation
1.74
1.09 - 2.80
0.021
Diabetes
1.56
1.05 - 2.31
0.027
1.75
1.10 - 2.76
0.017
Ribavirin use
0.82
0.56 - 1.20
0.300
SOF-based treatment
1.72
1.03 - 2.88
0.038
Abbreviations: BMI, Body Mass Index; CI, Confidence Interval; DAA, Direct Acting Antiviral; HCC, Hepatocellular Carcinoma; HCV, Hepatitis C Virus; HR, Hazard Ratio; LSM, Liver Stiffness Measurement; SOF, Sofosbuvir.
Cox forward stepwise selection. Gender variable was forced to be included in the model.Statistically significant hazard ratios and related 95% confidence intervals are reported in bold.
The cumulative HCC-IR according to pretreatment evaluation of PLT and albumin levels and by their combination are reported in Supplementary Fig. 1 (Panels A, B and C).
3.3 Prospective evaluation of HCC-risk predictors
Among patients who achieved SVR following the DAA therapy, those who have available paired pre- treatment and post-treatment data were evaluated prospectively. Pre-treatment characteristics by HCC occurrence are shown in Supplementary Table 1. Pre- and post-treatment LSM were available in 420 patients, 22 of whom developed HCC. As shown in Fig. 2A, the median pre-treatment LSM was similar in both groups of patients with or without the development of HCC (p=0.551), and a significant drop in LSM was observed after viral eradication both in cases with (p=0.024) or without (p<0.001) HCC occurrence. Patients who developed HCC during the FU showed significantly higher median post-treatment LSM compared with those who did not (18.5 vs. 12.0 kPa; p=0.016).
Fig. 2Pre- and post-treatment evaluation of liver disease severity factors in patients with SVR. Pre- and post-treatment LSM (A), platelets count (B) and albumin level (C) in patients with and without HCC occurrence. Values were reported as the median and interquartile range. Abbreviations: HCC, Hepatocellular Carcinoma; LSM, Liver Stiffness Measurement; SVR, Sustained Virologic Response.
Following HCV eradication, the PLT count increased significantly compared with the pre-therapy values in patients without HCC (p<0.001) but not in those who developed HCC (p=0.889) (Fig. 2B). Albumin levels increased in both groups of patients with or without the development of HCC during pre-and post-therapy evaluation (both, p<0.001) (Fig. 2C).
3.4 Post-treatment predictors of different HCC-risk groups
The cumulative HCC-IR at 12-, 24-, and 36-month in patients with post-treatment PLT count ≤120,000/µl was 2.81%, 6.40%, and 8.96%, respectively, which appeared significantly higher (log-rank test p<0.001) compared to patients with PLT count >120,000/µl (IR: 1.48%, 2.78% and 4.36%, at 12-, 24-, and 36-month, respectively) (Fig. 3A). The same profile was observed between groups with albumin levels <3.5 or >3.5 g/dl (IR at 12-, 24- and 36-month: 7.52%, 14.12%, and 21.49% vs. 1.30%, 3.16 %, and 4.72%, respectively; log-rank test p<0.001) (Fig. 3B).
Fig. 3Cumulative incidence of HCC according to different post-treatment characteristics. (A) Post-treatment platelets count (Platelet count ≤120,000/µl vs. Platelet count >120,000/µl). (B) Post-treatment albumin level (albumin level ≤3.5 g/dl vs. albumin level >3.5 g/dl). (C) Combination of post-treatment platelets and albumin values (Platelet count ≤120,000/µl and/or albumin level ≤3.5 g/dl vs. Platelet count >120,000/µl and albumin level >3.5 g/dl). Abbreviations: Alb, Albumin; HCC, Hepatocellular Carcinoma; Plat, Platelet.
By combining the two surrogate markers of liver disease severity (i.e., PLT count and albumin level) (Fig. 3C), the highest HCC-IR following HCV eradication was found in the group of patients with PLT count ≤120,000/µl and/or albumin level ≤3.5 g/dl (IR: 3.90%, 6.37%, 9.38%, at 12-, 24- and 36-month, respectively) compared to the group of patients categorized by both PLT >120,000/µl and albumin >3.5 g/dl (IR: 0.78%, 1.30%, and 2.04%, at 12-, 24- and 36-month, respectively; log-rank test p<0.001), who maintained the lowest risk of HCC occurrence during the FU compared to other combinations.
3.5 Prediction model for HCC free-survival probability after SVR
The nomogram was developed based on the multivariable Cox regression model. It included the variables evaluated post-therapy that were independently associated with HCC development such as: age (AdjHR 1.03; 95%CI 1.00-1.06), PLT count (AdjHR 1.92; 95%CI 1.06-3.45) and albumin level (AdjHR 4.38; 95%CI 2.48-7.75). It was assigned a score to each variable, and thus by summing up each point to obtain a total score and, drawing a vertical line from the total score (Total Point) up to the probability scale, at 1-, 3- and 5-year of overall survival at the bottom of the nomogram, it was possible to determine the probability of being HCC-free (overall survival probability) (Fig. 4). For example, a patient aged 65 years (55 points), with a PLT count ≤120,000/µl (37 points), and albumin level ≤3.5 g/dl (83 points) has a total scoring of 175 points; corresponding to a probability of being HCC-free survival at 1-year of about 90%, at 3-years ranging 80-75% and at the 5-year FU between 60-50%.
Fig. 4Nomogram for HCC prediction in patients with SVR. Multivariable Cox regression model (including age, post-treatment platelets count and albumin levels) and graphical representation as nomogram. For each risk factor draw a vertical line upwards and note down the corresponding points. Each score should be summarized ending up with a total score that corresponds to a predicted probability of 1-, 3- and 5- years of overall survival at the bottom of the nomogram.
The model showed a satisfactory prediction capability with a Harrell's c-index of 0.74 (95%CI 0.67-0.82). The Calibration Plot is reported in the Supplementary Material (Supplementary Fig. 2).
4. Discussion
To our knowledge, this is the first study focusing on the post-treatment predictive parameters of HCC occurrence at long-term FU in successfully DAA-treated patients with liver cirrhosis from a large ongoing prospective cohort.
In this analysis, we observed a cumulative HCC-IR of 6.47% at the 36-month FU after HCV eradication (2.45/100 p-y). This rate is somewhat higher than previously reported ones, ranging from 2 to 2.5% [
Factors associated with increased risk of de novo or recurrent hepatocellular carcinoma in patients with cirrhosis treated with direct-acting antivirals for HCV infection.
], and the different rates reported by several cohorts, compared to our data, may be explained by a different FU duration and, in particular, by different degrees of liver disease severity [
] among studies. In the PITER cohort, we reported a higher number of patients with features of severe liver disease (i.e., assessed by FIB-4 >3.25 in 67.2%; PLT ≤120,000/µl in 53.0%; LMS ≥20kPa in 47.5%; CPT-B in 14.9% and cases with previous liver decompensation in 10.9%), compared with some of the other cohorts described in literature [
Factors associated with increased risk of de novo or recurrent hepatocellular carcinoma in patients with cirrhosis treated with direct-acting antivirals for HCV infection.
ANRS collaborative study group on hepatocellular carcinoma (ANRS CO22 HEPATHER, CO12 CirVir and CO23 CUPILT cohorts) Electronic address: [email protected] Lack of evidence of an effect of direct-acting antivirals on the recurrence of hepatocellular carcinoma: data from three ANRS cohorts.
]. The peculiarity of our cohort was possibly related to the fact that in Italy between 2015 to 2017, the DAAs could only be prescribed for patients with progressive liver disease and previous resistance to IFN, but being in stable compensation at pretreatment evaluation and within the previous 12 months.
The pre-treatment factors independently associated with HCC occurrence by multivariate analysis were: older age, genotype 3, presence of diabetes, decreased PLT count and albumin level confirming previously reported data [
Factors associated with increased risk of de novo or recurrent hepatocellular carcinoma in patients with cirrhosis treated with direct-acting antivirals for HCV infection.
]. In Italy, genotype 3 is a prevalent genotype in younger patients, who had acquired HCV infection by drug use, less in older cohorts, that have acquired the infection by different transmission routes mainly infected by HCV genotype-1b. These last cases are the most representative group of patients with liver cirrhosis in Italy, and this could be one of the reasons for the lack of association between genotype 3 and HCC risk in other cohorts described previously. Age by itself is one of the main confounding factors and after age-adjusted Cox regression analysis in our study, genotype 3 became independently associated with HCC risk after viral eradication. Alcohol consumption, differently from other results [
], was not associated with HCC development after SVR, potentially explained by prevalent social alcohol use, rather than the presence of heavy drinkers in this relatively old-age patient cohort.
The comparison of albumin levels in available paired data collected pre- and post-treatment in both cases with or without HCC occurrence improved. However, it could be emphasized, that the median albumin level was fairly normal in patients who did not develop HCC, and in contrast, it appeared frankly pathological in patients who developed HCC. Furthermore, some features associated with portal hypertension, such as: thrombocytopenia, deterioration of sinthetic liver function, as reflected by low albumin levels and increased liver stiffness tend to persist in patients who developed HCC. Thus, the PITER cohort tends to depict two different categories of cases, those with main characteristics that are just apparently similar at pre-treatment data, but whose real picture can be unmasked after viral eradication, being associated to a higher HCC risk.
Our findings confirm that even at a 36-month FU period after SVR, a substantial residual risk of HCC development remained, particularly in those with pre-treatment severe liver damage and the presence of portal hypertension (i.e., in cases with albumin level <3.5 g/dl and/or PLT count <120,0000/µl) [
Factors associated with increased risk of de novo or recurrent hepatocellular carcinoma in patients with cirrhosis treated with direct-acting antivirals for HCV infection.
]. In these patients, the severity of liver disease was identified among the most critical predictors of de-novo HCC.
Considerable evidence and a meta-analysis support the use of LSM for predicting HCC-risk in patients who achieved SVR, and different cut-offs have been proposed, ranging from >20 or >30 kPa at pre-treatment, and >10 kPa at post-treatment evaluations, in relation to higher incidence of HCC [
Factors associated with increased risk of de novo or recurrent hepatocellular carcinoma in patients with cirrhosis treated with direct-acting antivirals for HCV infection.
Impact of liver-stiffness measurement on hepatocellular carcinoma development in chronic hepatitis C patients treated with direct-acting antivirals: a systematic review and time-to-event meta-analysis.
]. We did not have sufficient post-treatment LSM data to include the stiffness values in the risk evaluation modelling. However, the pre-treatment LSM ≥ 20 in the PITER cohort, was significantly associated with HCC development, as confirmed also by a recent HCC modeling risk [
]. Specifically, 60.7% vs. 46.8% of cases who developed HCC had pre-treatment LSM ≥20 kPa, respect to cases who did not (p=0.011). These cases showed a higher median stiffness value at 1-year FU compared to those who did not develop HCC (18.5 vs. 12.0 kPa; p=0.016), while a significant drop of LSM was observed after viral eradication both in cases without (p<0.001) or with HCC development (p=0.024). A similar pattern has been observed also in other studies [
Factors associated with increased risk of de novo or recurrent hepatocellular carcinoma in patients with cirrhosis treated with direct-acting antivirals for HCV infection.
], we could not find any change in LSM between pre-treatment and post-treatment as a predictor of HCC development, probably due to a low sample size with available paired data. However, pointing out the significantly lower drop in LSM among patients with HCC (-18.0% vs. -28.9% p = 0.005), the relation between LSM and switch-off of liver inflammation after viral eradication must be also considered, independently of HCC development [
]. In this cohort, HCC developed mainly among cases with LSM ≥ 20 kPa, when obtained within one year from EOT; however, in accordance with data of Pons et al [
], at post-treatment long-term FU after SVR, the LSM did not show an independent HCC predictive value.
The best predictor of HCC occurrence after SVR in our study was given by the association of two easy and accurate markers, undoubtedly associated with portal hypertension and liver function, as PLT count and albumin level, analyzed prospectively at long-term FU after SVR.
The highest HCC incidence in SVR cases was found in patients with PLT count ≤120,000/µl and/or albumin level ≤3.5 g/dl (IR 3.90%; 6.37% and 9.38%, at 12-, 24- and 36-months, respectively) compared with patients categorized by both PLT >120,000/µl and albumin levels >3.5 g/dl (IR 0.78%, 1.30% and 2.04%, at 12-, 24- and 36- months, respectively). The latest group maintained the lowest HCC risk during the FU compared to other combinations of these two parameters.
We followed the analysis based on the aforementioned markers, which combine indicators of hepatic function (i.e., serum albumin) and severe stage of liver disease (assessed with PLT count), evaluated during the FU after viral eradication, with the patient's age (a strong driver of carcinogenesis) [
]. The nomogram constructed in our study, using PLT and albumin values measured at long-term FU after EOT in patients who achieved the SVR, allows the estimation of the probability of being HCC-free over time (at 1-, 3- and 5-years of FU) in the real-life practice. Combining variables independently associated with HCC occurrence (age and PLT count ≤ vs. >120,000/µl and albumin level ≤ vs. >3.5 g/dl), we stratified two risk groups post-therapy, with an HCC-IR significantly different of 1.35 vs. 3.77/100 p-y. This developed model, calculated by nomogram maps, based on age, PLT, and albumin obtained post-therapy, showed an accurate prediction of overall survival probability to be HCC-free at 1, 3, and 5-years of FU in the individual patient.
Our data, as others recently reported, fails to identify a group with no risk in whom the HCC surveillance may be stopped, perhaps because we evaluated "as predictors" the biochemical tools strongly related to the progression of liver damage. Instead, we reinforce the need for very tight HCC surveillance in the risk group of patients with advanced liver disease, based, in addittion to the liver ultrasound, also in a periodical evaluation of PLT count and albumin level combination, which should be considered to drive the risk of HCC occurrence. Genetic determinants or algorithms based on other predictive variables, possibly identified by machine learning approaches, are warranted to refine the individualized prediction of HCC-risk ([
Major strengths of our study are the large patient cohort, the consecutive enrolment, the long-term prospectively evaluation , and the complete clinical characterization with specific essential real-life data obtained in tertiary health care centers not explicitly dedicated to HCC cure.
This multicenter ongoing life-practice patient enrolment could suffer from possible limitations, as the potential inclusion of cases with very early and undetectable liver nodules that cannot have been ruled out before the DAA therapy by the six-month US or other imaging surveillance. To limit this potential bias and to be more conservative in the analysis only HCC occurring after the EOT was considered de-novo HCC.
The potential heterogeneity among the network of clinical centers in terms of competence in diagnosing HCC could be another limitation of the study. However, we believe this limitation was overcome by the strict application of guidelines to diagnose HCC with adequate monitoring of cases by the participating centers.
The accurate monitoring of LSM during FU was lacking in some patients, making this tool not thoroughly evaluated in our analysis.
The real-world clinical practice of HCC surveillance, focusing on the risk stratification approach based only on two easy-accessible and accurate surrogate markers of liver function over time could be helpful. Further dedicated studies will be useful to fully confirm its utility in real-life clinical practice.
4.2 Conclusions
A scoring system, based on combined PLT and albumin values evaluated post-treatment, may be helpful to identify patients free of HCC risk after SVR. Using the produced nomogram, based on: patient age, albumin level ≤3.5 g/dl and PLT count ≤120,000/µl, performed at 1-, 2- and 3-year after SVR, could accurately identify the population with the highest HCC risk deserving of a careful long-term FU. Further dedicated studies are useful to fully confirm its utility in real-life clinical practice.
Conflict of interest
None declared.
Data availability statement
This study was approved by the Ethical Committee of Istituto Superiore di Sanità and the local Ethical Committees of the participating centers. By protocol, the property of the data is of participating clinical centers while Istituto Superiore di Sanità acts as a coordinating center for data management and analysis. Cumulative data are reported within the paper whereas each patient's data are not fully available and without restrictions for ethical reasons. Dr. LA Kondili ([email protected]) is in charge data management and the readers may contact her for specific data requests. She will provide the necessary ethical clearances for access to data. Interested readers may contact Carlo Petrini ([email protected]) from the Ethical Committee of Istituto Superiore di Sanità to request access to patient level data.
Acknowledgements
The authors wish to thank all PITER collaborating group and all participating centers, investigators and research staff (available in www progettopiter.it) who are involved in the study on a voluntary basis, for their time and effort. We also thank and Giampaolo La Terza (Medisoft Informatic Services) for Database maintenance and implementation. We additionally acknowledge Federica Magnani, Rosangela Duranti, Erika Olivieri, Alessandra Mattei for secretarial and administrative assistance.
This study was supported by Italian Ministry of Health (Grant number RF-2016-02364053 and Ricerca Corrente Annual Program 2023).
Salvatore Madonia (Department of Internal Medicine, Villa Sofia-Cervello Hospital, Palermo, Italy)
Donatella Ieluzzi (Liver Unit, University Hospital of Verona, Verona, Italy)
Gloria Taliani (Infectious and Tropical Medicine Unit, Department of Public Health and Infectious Diseases, "Policlinico Umberto I" Hospital, Sapienza University of Rome, Rome, Italy)
Lorenzo Badia (Clinic of Infectious Diseases and Microbiology Unit, Alma Mater Studiorum Bologna University)
Guglielmo Marco Migliorino (Department of Infectious Diseases, San Gerardo Hospital - ASST Monza, Monza, Italy)
Alessia Giorgini (Gastroenterology and Hepatology Unit, San Paolo Hospital, University of Milan, Italy)
Mario Masarone (Internal Medicine and Hepatology Unit, University of Salerno, Salerno)
Pierluigi Blanc (Infectious Disease Unit, Santa Maria Annunziata Hospital, Florence, Italy)
Valentina Cossiga (Gastroenterology Unit, Federico II University, Naples, Italy)
Martina DE Siena (Internal Medicine and Gastroenterology, Fondazione Policlinico Universitario Agostino Gemelli IRCCS, Rome, Italy)
Xhimi Tata (Center for Global Health, Istituto Superiore Di Sanità, Rome, Italy)
Maria Grazia Rumi (Hepatology Unit, San Giuseppe Hospital, Milan, Italy)
Luchino Chessa (Liver Unit, University Hospital, Monserrato, Cagliari, Italy)
Pietro Lampertico (Division of Gastroenterology and Hepatology, Foundation IRCCS Ca' Granda Ospedale Maggiore Policlinico, Milan, Italy)
Carlo Ferrari (Department of Medicine and Surgery, University of Parma, Unit of Hematology, Azienda Ospedaliero-Universitaria di Parma, Italy)
Ivan Gentile (Department of Clinical Medicine and Surgery, University of Naples Federico II, Napoli, Italy)
Giustino Parruti (Infectious Diseases Unit, Spirito Santo General Hospital, Pescara, Italy)
Leonardo Baiocchi (Hepatology Unit, University of Tor Vergata, Rome, Italy)
Alessia Ciancio (Gastroenterology Unit, Città della Salute e della Scienza of Turin, University Hospital, Turin, Italy)
Pietro Invernizzi (Division of Gastroenterology and Center for Autoimmune Liver Diseases, San Gerardo Hospital, Monza, Italy)
Alessandro Federico (Department of Hepato-Gastroenterology, University of Campania Luigi Vanvitelli, Naples, Italy)
Carlo Torti (Department of Medical and Surgical Sciences, Infectious and Tropical Diseases Unit, “Magna Graecia” University, Catanzaro, Italy)
Giulia Morsica (Department of Infectious Diseases, San Raffaele Hospital, Milan Italy)
Pietro Andreone (Department of Internal Medicine, University of Modena and Reggio Emilia, Modena, Italy)
Alessio Aghemo (Department of Gastroenterology, IRCCS Humanitas Research Hospital IRCCS, Rozzano, Italy)
Patrizia Popoli (National Center for Drug Research and Evaluation, Istituto Superiore Di Sanità, Rome, Italy)
Factors associated with increased risk of de novo or recurrent hepatocellular carcinoma in patients with cirrhosis treated with direct-acting antivirals for HCV infection.
ANRS collaborative study group on hepatocellular carcinoma (ANRS CO22 HEPATHER, CO12 CirVir and CO23 CUPILT cohorts)
Electronic address: [email protected] Lack of evidence of an effect of direct-acting antivirals on the recurrence of hepatocellular carcinoma: data from three ANRS cohorts.
Impact of liver-stiffness measurement on hepatocellular carcinoma development in chronic hepatitis C patients treated with direct-acting antivirals: a systematic review and time-to-event meta-analysis.
Inconsistencies are due to missing values.
00164-0&id=gr1.jpg){kind=link}
00164-0&id=gr2.jpg){kind=link}
00164-0&id=gr3.jpg){kind=link}
00164-0&id=gr4.jpg){kind=link}