Advertisement

Reply to: “HCC prediction post SVR: many tools yet limited generalizability!”

De novo HCC risk stratification after HCV cure: All roads lead to Rome?
  • Georg Semmler
    Affiliations
    Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria

    Vienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria
    Search for articles by this author
  • Elias Laurin Meyer
    Affiliations
    Institute for Medical Statistics, Center for Medical Statistics, Informatics and Intelligent Systems, Medical University Vienna, Vienna, Austria
    Search for articles by this author
  • Mattias Mandorfer
    Correspondence
    Corresponding author. Address: Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Waehringer Guertel 18-20, 1090, Vienna, Austria; Tel.: +43 1 40400 47440, fax: +43 1 40400 47350.
    Affiliations
    Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria

    Vienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria
    Search for articles by this author

      Linked Article

      • HCC risk stratification after cure of hepatitis C in patients with compensated advanced chronic liver disease
        Journal of HepatologyVol. 76Issue 4
        • Preview
          Hepatocellular carcinoma (HCC) is a major cause of morbidity and mortality in patients with advanced chronic liver disease (ACLD) caused by chronic hepatitis C who have achieved sustained virologic response (SVR). We developed risk stratification algorithms for de novo HCC development after SVR and validated them in an independent cohort.
        • Full-Text
        • PDF
        Open Access
      • HCC prediction post SVR: Many tools yet limited generalizability!
        Journal of HepatologyVol. 77Issue 4
        • Preview
          Despite attaining a sustained virological response (SVR), the risk of hepatocellular carcinoma (HCC) remains a significant concern in patients with chronic hepatitis C (CHC). The EASL guidelines advise HCC screening in a population with a high incidence of HCC, considering cost, expertise, treatment options, and rate of tumor growth.1 Accordingly, HCC screening is recommended in patients with CHC and >F3 fibrosis. Several prediction tools have been applied in various studies for HCC prediction; however, none is generalizable to the global population to date.
        • Full-Text
        • PDF
      To the Editor:
      We would like to thank Dr. Bhagat and colleagues
      • Bhagat N.
      • Verma N.
      • Singh V.
      HCC prediction post SVR in HCV patients: many tools yet limited generalizability!.
      for their interest in our study on de novo hepatocellular carcinoma (HCC) risk stratification in compensated advanced chronic liver disease (cACLD) patients who achieved sustained virologic response (SVR).
      • Semmler G.
      • Meyer E.L.
      • Kozbial K.
      • Schwabl P.
      • Hametner-Schreil S.
      • Zanetto A.
      • et al.
      HCC risk stratification after cure of hepatitis C in patients with compensated advanced chronic liver disease.
      The authors provided a summary of 11 selected risk stratification approaches and highlighted 5 points related to our study but also HCC risk stratification/screening in general.
      The multiplicity of available risk stratification tools underlines the high scientific interest in this clinically relevant issue. While some approaches follow beaten tracks (simple algorithms based on conventional statistical methods; e.g., Pons et al.
      • Pons M.
      • Rodríguez-Tajes S.
      • Esteban J.I.
      • Mariño Z.
      • Vargas V.
      • Lens S.
      • et al.
      Non-invasive prediction of liver-related events in patients with HCV-associated compensated advanced chronic liver disease after oral antivirals.
      ), complex artificial intelligence-based methods discovered alternative routes that are less obvious. However, several factors/variables are shared between models, indicating their key role in regard to the outcome de novo HCC: age and alpha-fetoprotein, next to surrogates of hepatic dysfunction (serum albumin) and liver fibrosis/portal hypertension (i.e., liver stiffness measurement [LSM] and platelet count as well as its derivatives, such as the fibrosis-4 score). It is important to note that variable selection is determined by their availability in retrospective datasets, indicating that their intersections – i.e., broadly available parameters – are likely overrepresented throughout the different risk stratification tools. Importantly, our derivation cohort of comprehensively characterized patients allowed for the selection of the best, rather than the best available predictors.
      As addressed by Bhagat et al.
      • Bhagat N.
      • Verma N.
      • Singh V.
      HCC prediction post SVR in HCV patients: many tools yet limited generalizability!.
      in their first point, ACLD was diagnosed by either hepatic venous pressure gradient measurement, LSM, or histology in our derivation cohort. However, this does not induce selection bias, as liver disease severity was staged by one or more of these methods in all patients undergoing antiviral therapy, due to implications for reimbursement and regimen selection. Moreover, the generalizability of our risk stratification approach to other predominately Caucasian patient cohorts was confirmed by extensive validation – which is of paramount importance in the context of predictive modeling, but has often been omitted by the other studies listed by Bhagat and colleagues.
      • Bhagat N.
      • Verma N.
      • Singh V.
      HCC prediction post SVR in HCV patients: many tools yet limited generalizability!.
      We fully agree about the underrepresentation of patients from Asia or Africa in our study and have repeatedly emphasized (see Discussion of our manuscript and correspondence
      • Semmler G.
      • Mandorfer M.
      Reply to: 'Risk stratification of hepatocellular carcinoma after hepatitis C virus eradication in patients with compensated advanced chronic liver disease in Japan'.
      ) the need for studies in other ethnicities/ethnically diverse populations.
      • Semmler G.
      • Meyer E.L.
      • Kozbial K.
      • Schwabl P.
      • Hametner-Schreil S.
      • Zanetto A.
      • et al.
      HCC risk stratification after cure of hepatitis C in patients with compensated advanced chronic liver disease.
      Second, we are well-aware of the dynamics of hepatic inflammation/portal hypertension
      • Schwabl P.
      • Mandorfer M.
      • Steiner S.
      • Scheiner B.
      • Chromy D.
      • Herac M.
      • et al.
      Interferon-free regimens improve portal hypertension and histological necroinflammation in HIV/HCV patients with advanced liver disease.
      ,
      • Mandorfer M.
      • Kozbial K.
      • Schwabl P.
      • Chromy D.
      • Semmler G.
      • Stättermayer A.F.
      • et al.
      Changes in hepatic venous pressure gradient predict hepatic decompensation in patients who achieved sustained virologic response to interferon-free therapy.
      – and thus, LSM – after SVR. However, the clinical utility of risk stratification tools is critically dependent on their robustness in regard to variations in care/data availability, e.g. the time point of LSM. Accordingly, we do not consider the differences in LSM time points between the derivation and validation cohorts as a limitation – it is rather a strength of our study, as the consistent predictive performance underlines the generalizability/robustness of our algorithms.
      Third, we want to emphasize that our score was intended to be an easy-to-use tool aiding clinical decision making and is as such not intended to be the basis for HCC surveillance policies. Similarly, while class imbalance is hard to avoid in the context of our research, we agree that advanced sampling methods applying balancing techniques and machine learning are promising options to increase the accuracy of classification
      • Fotouhi S.
      • Asadi S.
      • Kattan M.W.
      A comprehensive data level analysis for cancer diagnosis on imbalanced data.
      (at the cost of increased statistical complexity and clinical impracticability).
      Fourth, we agree that cost-effectiveness requires a critical appraisal and should ideally be evaluated in the context of specific health care systems, as the cost of screening and treatment options may vary substantially. While an HCC incidence threshold of 1.32%/year
      • Farhang Zangneh H.
      • Wong W.W.L.
      • Sander B.
      • Bell C.M.
      • Mumtaz K.
      • Kowgier M.
      • et al.
      Cost effectiveness of hepatocellular carcinoma surveillance after a sustained virologic response to therapy in patients with hepatitis C virus infection and advanced fibrosis.
      has been proposed to maintain an incremental cost-effectiveness ratio (ICER) below 50,000USD/quality-adjusted life year (QALY), the underlying model did not account for recent advances in HCC treatment.
      • Ioannou G.N.
      HCC surveillance after SVR in patients with F3/F4 fibrosis.
      Finally, the individual benefit of an early diagnosis of HCC critically depends on age and comorbidities – thus, a recent study
      • Mueller P.P.
      • Chen Q.
      • Ayer T.
      • Nemutlu G.S.
      • Hajjar A.
      • Bethea E.D.
      • et al.
      Duration and cost-effectiveness of hepatocellular carcinoma surveillance in hepatitis C patients after viral eradication.
      even proposed stopping HCC surveillance at the age of 70 in cirrhosis and 60 in advanced liver fibrosis, despite applying an ICER threshold as high as 150,000USD/QALY. However, the latter approach neglects profound variations in HCC risk within individual stages of liver fibrosis, as highlighted by the independent risk factors other than LSM that were identified in our study.
      • Semmler G.
      • Meyer E.L.
      • Kozbial K.
      • Schwabl P.
      • Hametner-Schreil S.
      • Zanetto A.
      • et al.
      HCC risk stratification after cure of hepatitis C in patients with compensated advanced chronic liver disease.
      Regarding the fifth and final point by Dr. Bhagat and colleagues,
      • Bhagat N.
      • Verma N.
      • Singh V.
      HCC prediction post SVR in HCV patients: many tools yet limited generalizability!.
      we acknowledge that regression coefficients rather than adjusted subdistribution hazard ratios may have been the preferred indicator for weighting of individual risk factors. Nevertheless, using the regression coefficients for assignment (importance: 1-1.5-2 vs. 1-2-3) together with a cut-off of 2.5 (instead of 3), the resulting classification of patients into risk groups would have yielded the same results as with the previous weighting (Fig. 1).
      Figure thumbnail gr1
      Fig. 1Cumulative incidence curves (using Fine & Gray competing risk analysis) of de novo HCC development.
      Cumulative incidence based on the post-treatment AFP/age/alcohol/LSM/albumin-derived strata (low-risk [0-2.5 points] vs. high-risk [3-7 points]) in patients with compensated advanced chronic liver disease in (A) the derivation cohort and (B) the validation cohort. Two points are assigned for FU-AFP ≥4.6 ngxml-1, 1.5 points for alcohol consumption above the threshold, 1.5 points for age ≥59 years, 1 point for FU-LSM ≥19 kPa, and 1 point for FU-albumin <42 gxL-1 (0 points if the respective criterion is not met). AFP, alpha-fetoprotein; FU, follow-up; HCC, hepatocellular carcinoma; LSM, liver stiffness measurement.
      All things considered, it appears that many roads lead to Rome (i.e., personalized management after SVR). However, while we have already established a highway (i.e., Baveno VII) for the individualization of portal hypertension surveillance, we fully acknowledge that the road to personalized HCC surveillance remains bumpy.

      Financial support

      No funding specific for this study was received.

      Authors’ contributions

      Analysis and interpretation of data (G.S., E.L.M., M.M.), drafting of the manuscript (G.S., E.L.M., M.M.), critical revision of the manuscript for important intellectual content (G.S., E.L.M., M.M.).

      Conflict of interest

      G.S. received travel support from Gilead. E.L.M. received grants from Novartis. M.M. served as a speaker and/or consultant and/or advisory board member for AbbVie, Gilead, Collective Acumen, Takeda, and W. L. Gore & Associates and received travel support from AbbVie and Gilead.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Supplementary data

      The following are the supplementary data to this article:

      References

        • Bhagat N.
        • Verma N.
        • Singh V.
        HCC prediction post SVR in HCV patients: many tools yet limited generalizability!.
        J Hepatol. 2022; 77: 1226-1228
        • Semmler G.
        • Meyer E.L.
        • Kozbial K.
        • Schwabl P.
        • Hametner-Schreil S.
        • Zanetto A.
        • et al.
        HCC risk stratification after cure of hepatitis C in patients with compensated advanced chronic liver disease.
        J Hepatol. 2022; 76: 812-821
        • Pons M.
        • Rodríguez-Tajes S.
        • Esteban J.I.
        • Mariño Z.
        • Vargas V.
        • Lens S.
        • et al.
        Non-invasive prediction of liver-related events in patients with HCV-associated compensated advanced chronic liver disease after oral antivirals.
        J Hepatol. 2020; 72: 472-480
        • Semmler G.
        • Mandorfer M.
        Reply to: 'Risk stratification of hepatocellular carcinoma after hepatitis C virus eradication in patients with compensated advanced chronic liver disease in Japan'.
        J Hepatol. 2022;
        • Schwabl P.
        • Mandorfer M.
        • Steiner S.
        • Scheiner B.
        • Chromy D.
        • Herac M.
        • et al.
        Interferon-free regimens improve portal hypertension and histological necroinflammation in HIV/HCV patients with advanced liver disease.
        Aliment Pharmacol Ther. 2017; 45: 139-149
        • Mandorfer M.
        • Kozbial K.
        • Schwabl P.
        • Chromy D.
        • Semmler G.
        • Stättermayer A.F.
        • et al.
        Changes in hepatic venous pressure gradient predict hepatic decompensation in patients who achieved sustained virologic response to interferon-free therapy.
        Hepatology (Baltimore, Md). 2020; 71: 1023-1036
        • Fotouhi S.
        • Asadi S.
        • Kattan M.W.
        A comprehensive data level analysis for cancer diagnosis on imbalanced data.
        J Biomed Inform. 2019; 90: 103089
        • Farhang Zangneh H.
        • Wong W.W.L.
        • Sander B.
        • Bell C.M.
        • Mumtaz K.
        • Kowgier M.
        • et al.
        Cost effectiveness of hepatocellular carcinoma surveillance after a sustained virologic response to therapy in patients with hepatitis C virus infection and advanced fibrosis.
        Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc. 2019; 17: 1840-1849.e1816
        • Ioannou G.N.
        HCC surveillance after SVR in patients with F3/F4 fibrosis.
        J Hepatol. 2021; 74: 458-465
        • Mueller P.P.
        • Chen Q.
        • Ayer T.
        • Nemutlu G.S.
        • Hajjar A.
        • Bethea E.D.
        • et al.
        Duration and cost-effectiveness of hepatocellular carcinoma surveillance in hepatitis C patients after viral eradication.
        J Hepatol. 2022; 77: 55-62