Resumen de Cancer incidence related to pharmacological treatment of hepatitis c virus infection
Background & Aims: Chronic infection by hepatitis C virus (HCV) is a well-known cause of morbi-mortality, by causing liver cirrhosis and eventually hepatocellular carcinoma (HCC). First treatments aimed to eradicate HCV were interferon (IFN) based regimes, generally associated to ribavirin; these were poorly tolerated and thus were used only in very fit patients. Later, direct-acting antivirals (DAA) replaced IFN based regimes and provided a very high rate of HCV eradication with good tolerability, allowing a wide use in all types of patients. In routine care, after treatment patients are generally discharged and often lost to follow-up. Whether they may experience cancer later on is unknown, and some concerns on increased cancer risk after treatment despite virus eradication have been raised. An observational retrospective study was designed with the aim to compare the incidence of cancer between patients receiving antiviral treatment for HCV infection and matched controls.
Methods: We carried out a population-based study using real-world data sources of linked healthcare registries from the Catalan Health System (ICS), including patients treated for HCV infection between 2012 and 2016 with either interferon, usually combined with ribavirin, (IFN), IFN followed later on by DAA (IFN-DAA), or with DAA only (DAA), and their matched controls. Since treatments were not concurrent in time, but used at different times and in different types of patients, propensity score matching (PSM) of HCV patients with concurrent comparable controls was carried out for each group (IFN, IFN-DAA and DAA). Poisson regression models were used to determine the annual cancer incidence and the rate ratios (RR) between HCV-treated patients and controls. Hazard ratios (HR) from Cox proportional hazard models were estimated. To account for potential information and selection biases, a number of sensitivity and subgroup analyses were carried out.
Results: Estimated cancer incidences per 100,000 person-years (95% confidence intervals [CI]) were 596.1 (482.5-671.4) cases for IFN, 1255.3 (947.9-1662.2) cases for IFN-DAA, and 1552.0 (95% CI 1380.1-1745.3) for DAA. A modestly increased cancer risk as compared to matched controls was found for IFN- DAA (RR 1.77, 95% CI 1.27-2.46) and for DAA (RR 1.90, 95% CI 1.66-2.19), but not for IFN (RR 1.11, 95% CI 0.92-1.32). In DAA-treated patients, the cancer risk was increased mostly in the subgroup of patients with cirrhosis, and attributable to HCC.
Discussion: A slight increase in the incidence of cancer has been observed in patients treated for HCV infection shortly after completion of their treatments. The study was observational and used data already available in administrative and clinical databases, so that there is limited information available for thorough adjustments allowing to control for potential biases. Thus, we cannot confirm whether the observed increase is related or not to the pharmacological effect of the antiviral agents, since treatments were not used simultaneously nor in the same types of patients, and results cannot be completely adjusted for indication biases, so that residual confounding may be still substantial. However, an increased cancer rate has yet been observed in patients once cured of their HCV infection, thus suggesting that after treatment completion they should not be discharged and lost to follow-up, but should undergo systematic follow-up screening for oncological diseases instead.
Conclusions: In general, treated HCV patients showed a slight increase in overall cancer incidence than matched controls without HCV infection and the risk was notably higher for HCC. Whether this increased risk is related to HCV infection, pharmacological treatment or any unidentified confounder requires further research, but in all cases continued monitoring after DAA treatment for early detection of cancer seems advisable, especially in cirrhotic patients.
