Factors influencing the failure of interferon-free therapy for chronic hepatitis C: Data from the Polish EpiTer-2 cohort study

Machine learning prediction models for clinical management of blood-borne viral infections: a systematic review of current applications and future impact

BACKGROUND

Machine learning (ML) prediction models to support clinical management of blood-borne viral infections are becoming increasingly abundant in medical literature, with a number of competing models being developed for the same outcome or target population. However, evidence on the quality of these ML prediction models are limited.

OBJECTIVE

This study aimed to evaluate the development and quality of reporting of ML prediction models that could facilitate timely clinical management of blood-borne viral infections.

METHODS

We conducted narrative evidence synthesis following the synthesis without meta-analysis guidelines. We searched PubMed and Cochrane Central Register of Controlled Trials for all studies applying ML models for predicting clinical outcomes associated with hepatitis B virus (HBV), human immunodeficiency virus (HIV), or hepatitis C virus (HCV).

RESULTS

We found 33 unique ML prediction models aiming to support clinical decision making. Overall, 12 (36.4%) focused on HBV, 10 (30.3%) on HCV, 10 on HIV (30.3%) and two (6.1%) on co-infection. Among these, six (18.2%) addressed the diagnosis of infection, 16 (48.5%) the prognosis of infection, eight (24.2%) the prediction of treatment response, two (6.1%) progression through a cascade of care, and one (3.03%) focused on the choice of antiretroviral therapy (ART). Nineteen prediction models (57.6%) were developed using data from high-income countries. Evaluation of prediction models was limited to measures of performance. Detailed information on software code accessibility was often missing. Independent validation on new datasets and/or in other institutions was rarely done.

CONCLUSION

Promising approaches for ML prediction models in blood-borne viral infections were identified, but the lack of robust validation, interpretability/explainability, and poor quality of reporting hampered their clinical relevance. Our findings highlight important considerations that can inform standard reporting guidelines for ML prediction models in the future (e.g., TRIPOD-AI), and provides critical data to inform robust evaluation of the models.

Changes in characteristics of patients with hepatitis C virus-related cirrhosis from the beginning of the interferon-free era

BACKGROUND

Nearly 290000 patients with chronic hepatitis C die annually from the most severe complications of the disease. One of them is liver cirrhosis, which occurs in about 20% of patients chronically infected with the hepatitis C virus (HCV). Direct-acting antivirals (DAAs), which replaced interferon (IFN)-based regimens, significantly improved the prognosis of this group of patients, increasing HCV eradication rates and tolerability of therapy. Our study is the first to assess changes in patient profile, effectiveness, and safety in the HCV-infected cirrhotic population in the IFN-free era.

AIM

To document changes in patient characteristics and treatment regimens along with their effectiveness and safety profile over the years.

METHODS

The studied patients were selected from 14801 chronically HCV-infected individuals who started IFN-free therapy between July 2015 and December 2021 in 22 Polish hepatology centers. The retrospective analysis was conducted in real-world clinical practice based on the EpiTer-2 multicenter database. The measure of treatment effectiveness was the percentage of sustained virologic response (SVR) calculated after excluding patients lost to follow-up. Safety data collected during therapy and the 12-wk post-treatment period included information on adverse events, including serious ones, deaths, and treatment course.

RESULTS

The studied population (n = 3577) was balanced in terms of gender in 2015-2017, while the following years showed the dominance of men. The decline in the median age from 60 in 2015-2016 to 57 years in 2021 was accompanied by a decrease in the percentage of patients with comorbidities and comedications. Treatment-experienced patients dominated in 2015-2016, while treatment-naive individuals gained an advantage in 2017 and reached 93.2% in 2021. Genotype (GT)-specific options were more prevalent in treatment in 2015-2018 and were supplanted by pangenotypic combinations in subsequent years. The effectiveness of the therapy was comparable regardless of the period analyzed, and patients achieved an overall response rate of 95%, with an SVR range of 72.9%-100% for the different therapeutic regimens. Male gender, GT3 infection, and prior treatment failure were identified as independent negative predictors of therapeutic success.

CONCLUSION

We have documented changes in the profile of HCV-infected cirrhotic patients over the years of accessibility to changing DAA regimens, confirming the high effectiveness of IFN-free therapy in all analyzed periods.

Decade of optimizing therapy with direct-acting antiviral drugs and the changing profile of patients with chronic hepatitis C

Chronic infection with the hepatitis C virus (HCV) remains a major health problem affecting approximately 58 million people worldwide. In the era of interferon (IFN)-based regimens, patients particularly infected with genotypes 1 and 4 achieved a low response rate. The implementation of direct-acting antivirals changed the landscape of HCV treatment. The increase in effectiveness provided us with the hope of eliminating HCV as a significant public threat by 2030. In the following years, there was an observed improvement in the treatment of HCV with genotype-specific regimens and highly effective pangenotypic options that are the most recent stage of the revolution. The optimization of therapy was accompanied by changes in the patient profile from the beginning of the IFN-free era over time. Patients treated with antiviral therapies were younger in successive periods, less burdened with comorbidities and comedications, more frequently treatment-naïve and had less advanced liver disease. Before the IFN-free era, specific subpopulations such as patients with HCV/HIV coinfection, those with a history of previous treatment, patients with renal impairment or with cirrhosis had lower chances for a virologic response. Currently, these populations should no longer be considered difficult to treat. Despite the high effectiveness of HCV therapy, there is a small percentage of patients with treatment failure. However, they can be effectively retreated with pangenotypic rescue regimens.

Are There Still Difficult-to-Treat Patients with Chronic Hepatitis C in the Era of Direct-Acting Antivirals?

Difficult-to-treat populations with chronic hepatitis C (CHC), in the era of interferon treatment, included patients with liver cirrhosis, kidney impairment, treatment-experienced individuals, and those coinfected with the human immunodeficiency virus. The current study aimed to determine whether, in the era of direct-acting antivirals (DAA), there are still patients that are difficult-to-treat. The study included all consecutive patients chronically infected with hepatitis C virus (HCV) who started interferon-free therapy between July 2015 and December 2020 in the Department of Infectious Diseases in Kielce. The analyzed real-world population consisted of 963 patients, and most of them were infected with genotype 1 (87.6%) with the predominance of subtype 1b and were treatment-naïve (78.8%). Liver cirrhosis was determined in 207 individuals (21.5%), of whom 82.6% were compensated. The overall sustained virologic response, after exclusion of non-virologic failures, was achieved in 98.4%. The univariable analysis demonstrated the significantly lower response rates in males, patients with liver cirrhosis, decompensation of hepatic function at baseline, documented esophageal varices, concomitant diabetes, body mass index ≥25, and previous ineffective antiviral treatment. Despite an overall very high effectiveness, some unfavorable factors, including male gender, genotype 3 infection, liver cirrhosis, and treatment experience, significantly reduce the chances for a virologic response were identified.