The Effects of Sustained Virological Response to Direct-Acting Anti-Viral Therapy on the Risk of Extrahepatic Manifestations of Hepatitis C Infection

Hashem B. El-Serag; Israel C. Christie; Amy Puenpatom; Diana Castillo; Fasiha Kanwal; Jennifer R. Kramer


Aliment Pharmacol Ther. 2019;49(11):1442-1447. 

In This Article

Materials and Methods

The US Department of Veterans Affairs (VA) operates the largest medical system in the United States with 153 medical centres, 745 community-based outpatient clinics and several other supportive-care facilities grouped into 21 Veterans Integrated Service Networks throughout the United States. The VA Corporate Data Warehouse was our primary study data source. Corporate Data Warehouse data exist on the VA Informatics and Computing Infrastructure platform in a relational database. It includes all laboratory test results and pharmacy data as well as inpatient and outpatient utilisation in the form of International Classification of Diseases, Ninth and Tenth Editions (ICD-9 and 10) (from October 2015) diagnoses codes and Vital Status.

Study Population

We identified a study cohort of patients with active HCV and treated with DAAs, by applying the following inclusion criteria: (a) at least one positive serum HCV RNA or genotype test from 1 October 1999 to 30 December 2016; (b) at least 18 years at the time of the HCV index date; (c) DAA treatment without simultaneous interferon, defined by sofosbuvir, simeprevir, ledipasvir/sofosbuvir, dasabuvir/ombitasvir/paritaprevir/ritonavir, daclatasvir or elbasvir/grazoprevir with or without ribavirin received from any VA nationwide pharmacy with courses beginning between 1 September 2012, to 1 July 2016; and (d) had laboratory testing required to verify SVR status. The date of the first DAA prescription was considered as the DAA index date, and extrahepatic manifestation outcomes data were evaluated through 1 November 2017, the end of the study period. This allowed at least 1 year of follow-up after treatment completion and SVR status was ascertained to determine the presence of extrahepatic conditions. We examined only the first course of DAA treatment; subsequent DAA courses were not considered.

We defined SVR by undetectable HCV RNA in all follow-up HCV RNA tests at least 4 weeks after the end of treatment (SVR4). We chose SVR4 because of its high correlation with SVR12 in previous studies.[9] If there was at least one positive HCV RNA result any time posttreatment, we classified the patient as not achieving SVR. Those with unknown SVR status because of a lack of RNA testing were excluded from all analyses. Prior HCV treatment experience was defined as filled prescription of interferon-based regimen with or without ribavirin or first-generation DAA (ie boceprevir, telaprevir) prior to index date.

The outcomes of interest were mixed cryoglobulinaemia, glomerulonephritis, lichen planus, porphyria cutanea tarda, NHL and diabetes, each defined by having at least one inpatient or outpatient ICD-9/10 diagnosis. Diabetes was defined as two or more outpatient or one or more inpatient codes or at least one VA prescription for oral hypoglycaemic drugs or insulin. Patients were identified as having a prevalent case if the ICD-9/10 code of extrahepatic manifestations was present any time before or 6 months after the DAA index date. Incident cases were defined as those occurring at least 6 months (ie specifically 26.07 weeks) after DAA index date after exclusion of patients with prevalent extrahepatic conditions. All patients were followed up until their last visit to a VA medical facility, death or 1 November 2017.

Potential confounders defined in the cohort included age at the DAA index date, gender, race/ethnicity, HCV genotype (as separate binary variables indicating a history of type 1a, 1b, 2, 3, 4, and 5 or 6), healthcare utilisation and comorbidities. The VA utilisation variable is defined as the number of outpatient visits to a VA medical facility within 1 year before and 1 year after DAA index date as separate binary variables (high and low), indicating greater and lower than the median number of visits. Other covariates included a history of alcohol and/or drug abuse, hypertension, hepatitis B virus (HBV), HIV co-infection and cirrhosis and its complications (ascites, variceal bleeding, hepatic encephalopathy or hepatorenal syndrome) any time before the DAA index date based on inpatient or outpatient ICD-9/10 diagnoses codes. Diagnoses codes used to define these conditions were obtained from the Agency for Healthcare Research and Quality Clinical Classifications Software (Table S1).

Data Analysis

We examined the effects of SVR following DAA treatment on risk of developing extrahepatic manifestations in HCV+ patients. The outcomes were incident cases diagnosed during follow-up, as defined above. We calculated the incidence rates by dividing number of outcomes by total person-years by SVR status for each condition as well as incidence rate ratios (IRR) and their corresponding 95% confidence intervals (CI), using unconditional maximum likelihood and normal approximation. We determined whether attainment of SVR after DAA affects the risk of extrahepatic manifestation outcomes, as measured by the incidence of these outcomes. We fitted multivariable Cox proportional hazards regression models with HCV treatment having two levels—treated with SVR and treated without SVR. We performed unadjusted, minimally adjusted and fully adjusted models. Minimally adjusted models included age, sex and race/ethnicity, and fully adjusted models included these variables in addition to preexisting alcohol abuse, drug abuse, HIV, hypertension, cirrhosis, cirrhosis complications, HBV, pre- and postutilisation data, HCV treatment experience and HCV genotype.

We conducted two sensitivity analyses to test the robustness of our results. First, we redefined the outcomes of interest using a more specific definition of two outpatient or one inpatient diagnosis codes. Second, we restricted the analysis to patients who had data available to determine SVR12 as this definition is used more frequently in clinical trials.

An alpha of 0.05 was used as the criteria for statistical significance. All statistical analyses were conducted using R version 3.5.1.[10]