We consecutively included all patients with diagnosis of HVC infection between May 2016 and March 2017 who were going to start treatment with DAA. Whether sofosbuvir was used as part of the combination of drugs for the treatment or not was up to the decision of the treating physician, according to the EASL clinical practice guidelines in force during recruitment.15 Patients were excluded from the study if they were under 18 years of age or withdrew consent.

We collected the following baseline variables: demographic variables related to the epidemiology of the infection (genotype of the HCV, grade of liver by hepatic elastography, viral load, presence of cryoglobulinemia, MELD score), cardiovascular risk factors and diseases, concomitant treatment with antiarrhythmics, including betablockers, and other drugs that could potentially lengthen the QTc interval. The combination of antiviral drugs for the treatment of virus C was also collected.

All patients underwent 3 electrocardiograms (ECG): the first one was performed before starting the antiviral treatment (basal); the second, during treatment, 4 weeks after the start; and the third, at least 3 weeks after it was finalized, considering that the duration of treatment could be of 8, 12 or 24 weeks.

PageWriter TC30 Cardiograph (Philips, United States) device was used to obtain the ECG. Twelve lead electrocardiograms were obtained in supine at a speed of 25mm/s and a gain of 10mm/mV.

The following electrocardiographic variables were collected: type of rhythm, heart rate, duration of the PR interval (normal <200ms), duration of the QRS complex (normal <120ms), duration of the QT interval and corrected QT interval (QTc), normal <450ms in men and <470ms in women. All measurements were made manually on paper by an observer blinded to the patient's clinical data.

For the measurement of the ECG intervals, the mean of the longest measurement in precordial leads (preferably in lead II) and limb leads (preferably in lead V3) was obtained. The QT interval was defined as the distance from the beginning of the QRS to the intersection of a tangent drawn to the steepest last limb of the T-wave with the baseline as defined by Lepeschkin and Surawicz.16 The Bazet method was used to calculate the QTc interval given that it is the most widely used method to correct the QT interval, and that extreme heart rates, where its performance is worse, were not anticipated.

Patients were followed up by telephone calls and by searching in their medical history to determine admissions to the Emergency Department or consultations to their general practitioner for palpitations, dizziness, syncope, chest pain or other reasons.

The normal distribution of the variables was checked using the Shapiro–Wilk test for normality. Continuous variables are represented as mean and standard deviation (SD) or median and interquartile range. Dichotomous variables were compared with a 2-tailed t-test, or Fisher–Pitman permutation test when necessary, and in cases with more than two groups the analysis of variance (ANOVA), or multivariate analysis of variance (MANOVA) test were used. Categorical variables are expressed as number and percentage and were compared with a Chi squared test. For the analysis of quantitative variables in subgroups the nonparametric randomization test of Fisher–Pitman was used.

All tests were 2-sided, and differences were considered statistically significant at P-values <.05. Statistical analysis was performed with Stata V.12.0 (StataCorp, United States).

We included 115 patients in the study, of which 101 had a complete follow-up with the 3 ECG performed (Fig. 1). In 14 patients all the 3 ECG were not achieved for the following reasons: 2 patients died during follow-up; 2 moved to another region; in 2 patients follow-up was lost, and 7 did not attend the appointments to perform any of the ECG for other reasons. The baseline characteristics of patients are shown in Table 1.

Fig. 1.

Central illustration. ECG, electrocardiogram; HCV, hepatitis C virus.

Concomitant cardiovascular diseases were collected: only 1 patient (1%) had ischemic heart disease, 6.9% arrhythmias (all of them were supraventricular), 6% had some degree of valve heart disease, and 3 patients had a history of other heart diseases (1 patient had tako-tsubo syndrome, another dissecting aneurysm of ascending aorta treated with an ascending aortic graft, and another had had acute pericarditis).

According to the Vaughan–Williams classification of antiarrhythmic drugs, only 1 patient was on non-dihydropyridine calcium channel blockers (class IV), 6.95% took betablockers (class II), and no patients were on class I or III antiarrhythmic drugs.

Grade of liver fibrosis by hepatic elastography before starting treatment was determined as F0–F1 in 37.6% of patients, F2 in 33.7% of patients, F3 in 13.9%, and F4 in 13.9%. Before starting treatment, 81.1% had genotype 1 (16.8% had genotype 1a and 64.4% genotype 1b), 4% had genotype 2, 7.9% had genotype 3, and 7.9% had genotype 4. One patient had double genotype (1b and 4). Median MELD score was 7.24±3. Median viral load was 2,271,659.5±2,769,144.1UI/mL. Cryoglobulinemia was present in 25.7% of patients, and 18.8% had been previously treated against HVC. Treatment duration was 8 weeks in 23.8% of patients, 12 weeks in 72.3% and 24 weeks in 4%. At the time of the third ECG, 95% of patients had sustained viral response.

Of the 101 patients who completed the study, 60.4% received treatment with sofosbuvir in any of their combinations (sofosbuvir+ledipasvir 48.5%, sofosbuvir+daclatasvir 6.9%, sofosbuvir without another DAA 4%, sofosbuvir+ombitasvir, paritaprevir, ritonavir, dasabuvir 1%); 11.6% of patients treated with sofosbuvir also received ribavirin (n=7).

On the other hand, 36.6% received the combination of ombitasvir, paritaprevir, ritonavir, dasabuvir (5.4% of these patients also received ribavirin) and 3% received treatment with elbasvir plus grazoprevir (none of them associated with ribavirin). The pharmacological combinations are summarized in Fig. 2.

Fig. 2.

Distribution of pharmacological combinations of direct-acting antivirals. Data is given in percentage of the 101 patients that completed the follow-up. Dark grey represents patients without sofosbuvir and light grey represents patients with sofosbuvir.

In the basal ECG, 1 patient had atrial flutter with pacemaker-mediated ventricular stimulation and 2 had atrial fibrillation. The remaining patients were in sinus rhythm. One of them had frequent supraventricular extrasystole. The second ECG was performed 29.6 days on average after the start of treatment (standard deviation [SD], 8), and the third ECG was performed 111.4 days after the end of treatment (SD, 80.3). There were 2 patients who had sinus rhythm in the first ECG and presented atrial fibrillation in the second ECG. Mean heart rate, PR interval, QRS complex and QTc interval for each ECG are presented in Table 2.

There were no significant differences in heart rate, or PR interval between the first, the second and the third ECG, neither in the whole group, nor in patients under treatment with sofosbuvir or other drugs combinations. In the group of patients without sofosbuvir, we found a small but statistically significant difference in the QRS complex between the second and the third ECG (86.5ms vs 87.8ms vs 85.0ms; P=.033).

Conversely, the QTc interval was significantly prolonged between the first and the second ECG in the group under sofosbuvir, returning to basal values in the third ECG after finishing the treatment (415.3ms vs 420.8ms vs 411.2ms; P=.029). On the contrary, in the group without sofosbuvir, no significant differences were found in QTc duration between the 3 ECG (411.2ms vs 411.1ms vs 410.4ms; P=.964).

Only 1 female patient had a pathological QTc interval before treatment, which remained pathological during and after treatment. This patient received treatment without sofosbuvir.

In addition, there were 4 male patients in the sofosbuvir group who, having normal QTc values in the basal ECG, reached QTc pathological values during the treatment. In these 4 patients, the QTc interval in the third ECG returned to normal values. In the group without sofosbuvir there were no patients with pathological QTc lengthening during the treatment.

To further evaluate other potential factors that could affect the QTc interval, we compared the variations of the QTc interval in different subgroups (concomitant use of betablockers, drugs causing QTc prolongation, genotype and grade of liver fibrosis by hepatic elastography). All subgroups showed a tendency of enlarging the QTc in patients under sofosbuvir treatment, despite the difference remained only significant in patients without betablockers, with drugs causing QTc prolongation, and patients with grade F3–F4 of fibrosis. In patients without sofosbuvir there were no differences in any subgroup. Subgroup analysis is shown in Table 3.

The mean time of follow-up was 41.9±14 weeks. During treatment, in the group with sofosbuvir, 1 patient consulted the emergency room or his general practitioner due to palpitations, 1 due to dizziness and 1 due to chest pain. This last one was referred to the cardiology outpatient clinic. A Holter ECG was performed during HCV treatment where frequent ventricular premature ventricular beats and NSVT were observed. After finishing HCV treatment, and up-titrating betablockers, premature ventricular beats rate remained similar. Three patients also presented occasional palpitations but did not consult for it.

In the group without sofosbuvir, 1 patient consulted for dizziness and 1 due to chest pain. No studies were performed. In this group there were also 2 patients with occasional palpitations who did not consult. No patient presented syncope during the follow-up. The cause of death of the 2 patients who died during the follow-up was oncological. Both of them died several months after HCV treatment was finished. One of them had a large B-cell lymphoma that invaded the right atrium and caused complete atrio-ventricular block. The second had a glioblastoma multiforme.

This study has several limitations. First, our findings are based on punctual ECG measurements at certain times of treatment and after treatment, which did not allow us to determine the precise moment in which the QTc interval starts to prolong, and when these changes disappear. In addition, the absence of continuous monitoring prevents diagnosing the presence of non-sustained arrhythmias.

Secondly, due to the limited sample size, no definite conclusions can be drawn about whether the effect of QT elongation is due to any particular combination of drugs including sofosbuvir, nor if there are additive effects with other drugs that can potentially lengthen the QTc. Furthermore, there are new drug combinations currently available that were not marketed at the time of recruitment in this study, so we do not know the electrocardiographic effects of these combinations.

Finally, since it is an observational study, it is not possible to establish causal relationships between treatment with sofosbuvir and the rise in QTc duration, although the fact that the QTc returns to normal values after treatment fulfils another of the main pharmacological criteria of causality (reversibility). Thus, a prospective study with continuous monitoring in patients taking sofosbuvir would help to determine the appearance of QTc lengthening and the presence of adverse non-sustained arrhythmic events.