The World Health Organization estimated that 257 million people were living with chronic hepatitis B virus (HBV) infection in 2015, leading to 0.88 million deaths every year worldwide due to HBV-related end stage liver diseases, such as decompensated cirrhosis, liver failure, and hepatocellular carcinoma (HCC) (1, 2). China has changed from a highly endemic to an intermediate endemic area for HBV infection in the past three decades (3, 4). Hepatitis B surface antigen (HBsAg) prevalence among people aged 1 ~ 29 years declined 46% during 1992 to 2006 (from 10.1 to 5.5%) (5) and 52% during 2006 to 2014 (from 5.5 to 2.6%) (4). There are approximate 70 million HBsAg carriers with 5 ~ 6% prevalence at present in China (1). Thus, China still has the world’s largest burden of HBV infection and will be the major contributor toward the global elimination of hepatitis B by 2030 (1, 6).

Effective inhibition of HBV replication by antiviral therapy delays or prevents the progression from HBV-related compensated cirrhosis to decompensated cirrhosis, leading to the further reduction in the risk of incidence of HBV-related liver failure and HCC (7). Moreover, HBsAg loss or seroconversion, which is an indicator of functional cure, reduces the occurrence of liver cirrhosis and HCC (8) and improves the long-term outcomes than those who remain HBsAg positive (9–11). Currently, there are two main therapeutic recommendations for chronic hepatitis B (CHB) patients: treatment with nucleos(t)ide analogs (NAs) [including entecavir (ETV), tenofovir (TDF), tenofovir alafenamide (TAF), and tenofovir amibufenamide (TMF)] or pegylated interferon-α (PEG-IFN-α) (12–15). Although NAs could effectively inhibit HBV replication, the cumulative rate of HBsAg loss was low, which is even comparable with that of untreated patients (16). PEG-IFN-α reveals both antiviral and immunomodulatory activity with sustained response after a finite course of therapy, resulting in a higher rate of both HBsAg loss and HBsAg seroconversion (17, 18). Patients with lower baseline HBsAg and HBV DNA level (19), rapid reduction of HBsAg during treatment (20), as well as the CC and TT genetic polymorphisms of interleukin-28B (21) are more likely to respond effectively to PEG-IFN-α therapy.

For hepatitis B e antigen (HBeAg)-positive CHB patients who had achieved HBV DNA inhibition and low HBeAg level in response to ETV therapy, switching to a finite course (48 weeks) of PEG-IFN-α-2a robustly elevated the rates of HBeAg seroconversion and HBsAg clearance (“OSST trial”) (22). Moreover, HBeAg seroconversion and HBsAg loss were sustained in most patients during off-treated 1 year follow-up (23). For CHB patients who achieved virological response and HBeAg loss to a previous NA treatment, switching to PEG-IFN-α-2a led to high rate of HBsAg loss at 48 weeks (14.4%) and 96 weeks (20.7%) post PEG-IFN-α-2a therapy (“New Switch study”) (24). For HBeAg-positive CHB patients, PEG-IFN-α-2a add-on to on-going ETV strategy led to significantly more decline in HBV DNA, HBsAg, and HBeAg level and higher proportion of HBeAg loss (25). These studies were mainly focused on sustained virological and serological response to PEG-IFN-α therapy in CHB patients. Several previous reports also showed that PEG-IFN-α was also effective and safe for patients with HBV-related fibrosis (26, 27). However, the application of PEG-IFN-α in patients with HBV-related compensated liver cirrhosis remains to be explored.

HBsAg seroclerance decreased the risk of hepatic decompensation in CHB patients (28), and was associated with a lower risk of late recurrence of HBV-related HCC (29). Cost-effectiveness analyses also revealed that earlier implementation of expanded antiviral therapy could decrease HBV-related complications and deaths in compensated cirrhosis patients with low-level viremia and even in untreated minimally active CHB patients, which contributed positively to individual clinical benefits and national healthcare budgets (30–32). Thus, although the medicine expense for PEG-IFN-α is higher than NA treatment, it could not only strongly inhibits viral replication and reduces HBsAg level, but also eliminates progression to end-stage liver diseases, leading to the down-regulation of both direct and indirect costs as well as potential long-term benefits. Furthermore, the immunomodulatory property of PEG-IFN-α could enhance natural killer cell activation (33, 34) and restore viral specific CD8⁺ T cell response (35), revealing particularly advantageous in cirrhotic patients.

Therefore, we conducted a real-world observational prospective study to investigate the safety and efficacy of PEG-IFN-α therapy for patients with HBV-related compensated cirrhosis.

To the best of our knowledge, this is the first report regarding the efficacy and safety of PEG-IFN-α-based therapy to patients with HBV-related liver cirrhosis. All patients confirmed the diagnosis of liver cirrhosis based on the imaging assessments, including the evidence of morphological changes of the liver, formation of liver nodules, and portal hypertension. However, these patients did not suffer with decompensated cirrhosis. PEG-IFN-α-2b treatment not only induced the virological response with undetectable serum HBV DNA in most of the patients, but also strongly down-regulated HBsAg levels (the median level reduction from 1,668 IU/mL to 227.2 IU/mL) at 48 weeks post-therapy. Importantly, administration of a finite course of PEG-IFN-α-2b did not accelerate the disease progression for cirrhosis, which presented as comparable LSM, thickness and length of spleen, as well as diameter of portal vein post PEG-IFN-α-2b therapy. No patients progressed to end-stage liver diseases during 48 weeks of PEG-IFN-α-2b treatment. The overall safety of PEG-IFN-α-2b application was good. The most common symptoms of AEs were fever, fatigue, weight loss, and alopecia, which were similar to those in CHB patients and inactive HBsAg carriers during PEG-IFN-α treatments (22, 24, 38–40). The changes of liver function and blood routine test were also presented the similar trends to those in chronic HBV infected patients in response to PEG-IFN-α therapy (22, 24, 38–40). Only three patients discontinued therapy due to severe AEs, including severe hyperthyroidism, continuous thrombocytopenia, and grade 1 ascites. These severe AEs were recovered after stopping PEG-IFN-α therapy and receiving symptomatic treatments. Collectively, 48-week PEG-IFN-α-2b-based therapy was safety, and indicated moderate efficacy for controlling chronic HBV infection in patients with HBV-related compensated liver cirrhosis.

PEG-IFN-α therapy could reach high rate of virological and serological response to patients with chronic HBV infection. The OSST trial revealed 72.0% (59/82) CHB patients achieved undetectable serum HBV DNA (<1,000 copies/ml) after switching from ETV to 48-week PEG-IFN-α-2a therapy, while the rate of virological response was 97.8% (90/92) in patients with continuous ETV therapy. Moreover, 8.5% (8/94) CHB patients achieved HBsAg clearance, and 4.3% (4/94) reached HBsAg seroconversion after switching to PEG-IFN-α-2a. However, no patients obtained HBsAg loss or seroconversion with continuous ETV treatment (22). The New switch study also demonstrated that switch from NAs to PEG-IFN-α-2a showed sustained HBV DNA inhibition (<200 IU/mL). 14.4% (22/153) patients obtained HBsAg loss at 48 weeks post-PEG-IFN-α-2a therapy, but the serological response rate for HBsAg clearance (20.7%, 31/150) did not robustly increased for a 96-week treatment (24). Wu et al. showed that PEG-IFN-α-2a therapy elevated the HBsAg clearance rate from 1.9% (2/104) in NAs monotherapy group to 37.4% (34/91) in add-on group (38). Wen et al. revealed that more than 50% CHB patients achieved HBsAg loss in both PEG-IFN-α-2b monotherapy and NAs add-on PEG-IFN-α-2b group (41). Furthermore, both PEG-IFN-α-2b monotherapy and PEG-IFN-α plus adefovir dipivoxil treatment results in more than 40% of HBsAg loss and more than 30% of HBsAg seroconversion in inactive HBsAg carriers (39, 40, 42, 43). Herein, we showed that 77.78% (42/54) patients with HBV-related compensate cirrhosis had undetectable serum HBV DNA (<50 IU/mL) at 48 weeks post-PEG-IFN-α-2b therapy. The rate of virological response was also comparable between treatment-naïve patients (PEG-IFN-α-2b monotherapy, 77.33%, 22/30) and NAs-experienced patients (PEG-IFN-α-2b plus NAs therapy, 83.33%, 20/24). This rate was similar to the previous reports regarding to the PEG-IFN-α treatment to HBV DNA suppression in CHB patients (22, 24, 38). We also show that although there were clear trends in HBsAg reduction, only PEG-IFN-α-2b add-on therapy to ongoing NAs could statistically reduce HBsAg level at 48 weeks post therapy. This might partly due to the fact that NAs-experienced patients had lower baseline HBsAg level than treatment-naïve patients, although this difference failed to achieve statistical significance. The limited enrolled patients might also contribute to the certain differences might not reach significance. Moreover, the present data revealed that although HBsAg level was gradually decreased during PEG-IFN-α-2b-based therapy, only 5.56% (3/54) patients reached HBsAg loss and 3.70% (2/54) patients obtained HBsAg seroconversion, which were lower than the rates in the published literatures in CHB patients and inactive HBsAg carriers (22, 24, 38–40, 42, 43). The key differences between the current study and other reports were shown in Table 3. In our opinions, the following three reasons might contribute to this difference. Firstly, we used a 48-week finite course of PEG-IFN-α-2b because prolonged PEG-IFN-α-2b treatment might induce the incidence of AEs and lead to the acute decompensation in compensated liver cirrhosis. Secondly, baseline HBsAg level less than 1,500 IU/mL was associated with high rate of HBsAg loss in response to PEG-IFN-α-2b therapy (22, 24, 38). The median baseline HBsAg level was 1,668 IU/mL, and 53.70% (29/54) patients had baseline HBsAg level higher than 1,500 IU/mL. We also observed that the three patients who achieved HBsAg loss also had moderate low level of baseline HBsAg (14.80 IU/mL, 60.74 IU/mL, and 512.1 IU/mL, respectively). Thirdly, the circulating and hepatic immune environment changed during the progression from CHB to liver cirrhosis (44, 45). The enrichment and elevation of exhausted immune cells in the circulation and liver microenvironment might reduce the responsiveness to PEG-IFN-α in HBV-related cirrhotic patients (44, 45).

Long-term HBV suppression by TDF could lead to the regression of fibrosis and cirrhosis (46). Papatheodoridis et al. retrospectively analyzed 147 HBeAg-negative CHB patients with or without IFN-α treatment and found that fibrosis regression rate in patients underwent IFN-α therapy was remarkably higher than that in untreated patients (17.5% vs 4%) according to the histological assessments, and the effect of fibrosis regression was mainly observed in patients who achieved sustained biochemical responses (47). Buster et al. also revealed that PEG-IFN-α-2b induced higher rate of HBeAg seroconversion, HBV DNA inhibition, and improvement of liver fibrosis in HBeAg-positive CHB patients with advanced fibrosis (26). However, Chen et al. showed that fibrosis regression rate was similar between ETV monotherapy (68%, 32/47) and PEG-IFN-α-2a add-on (56%, 60/108) in CHB patients with pre-treatment biopsy-proven Ishak fibrosis score 2, 3, or 4 after 78 weeks of therapy (27). In consistent with the findings by Chen et al. (27), our present data also indicated that 48-week PEG-IFN-α-2b might induce neither regression nor progression of liver cirrhosis based on Fibroscan and sonography assessments, which presented the comparable LSM, thickness and length of spleen, and diameter of portal vein between baseline and 48 weeks post-PEG-IFN-α-2b-based therapy regardless of monotherapy or add-on strategy in patients with HBV-related compensated liver cirrhosis. The ALB level also maintained stable during the observational period. In several patients, we also performed CT or MRI scans 48 weeks post therapy. The degree of liver cirrhosis did not present significantly changes in response to PEG-IFN-α-2b treatment. Further histological assessments should be performed to confirm the results.

We observed good safety profiles in HBV-related compensated liver cirrhosis in response to PEG-IFN-α-2b therapy. No unexpected severe AEs were reported. A 38-year old male patient suffered with grade 1 ascites 40 weeks post PEG-IFN-α-2b add-on ongoing TAF treatment. The patient felt mild abdominal distension, and received sonography assessment for a 1.3 cm depth of liquid dark area. This indicated grade 1 and a small amount of ascites. We also re-analyzed the CT scan of the patient in the baseline, and found a small amount of fluid accumulation around the spleen. Although the aminotransferase levels of this patient maintained stable during treatment, we still discontinued PEG-IFN-α-2b therapy and continued TAF therapy. The ascites disappeared 4 weeks post PEG-IFN-α-2b withdrawal. The liver function remained stable and no decompensated symptoms were reported during the following-up period. It was also well accepted that hepatic decompensation and HCC progression was uncommon but not eliminated even in patients with HBV-related cirrhosis receiving antiviral therapy (48, 49). Thus, it was still could not confirm the association between PEG-IFN-α-2b and hepatic decompensation in this patient. We also found the ALT elevation induced by PEG-IFN-α-2b, which might reflect immune clearance of HBV. However, ALT elevation more than five times of ULN was only found in four patients and no patients discontinued PEG-IFN-α-2b therapy due to ALT flare. Moreover, the down-regulations of WBC, neutrophils, platelets, and HGB levels were found during the treatment period. It was interesting that the recovery of netrophils and platelets count seemed faster in PEG-IFN-α-2b add-on strategy, indicating PEG-IFN-α-2b plus NAs treatment might have slighter affection to blood routine test.

There were several limitations in the current study. Firstly, we only enrolled 54 patients with HBV-related compensated cirrhosis in a single center. Large scale, multi-center cohort study should be performed to confirm the present findings. The histological assessments for liver cirrhosis before and after PEG-IFN-α-2b therapy should also be investigated. Secondly, we only observed the 48-weeks of PEG-IFN-α-2b treatment for enrolled patients. For liver cirrhosis patients, longer follow-up would be valuable to assess the sustainability of the virological response and whether the liver function improvements are maintained. The off therapy response and HCC progression should also be monitored for a long-term follow-up period. Thus, the long-term studies and innovative combination therapies should be performed in further research.

In summary, a 48-week finite course of PEG-IFN-α-2b therapy was well-tolerated, and reduced HBsAg level without accelerating disease progression in patients with HBV-related compensated cirrhosis. Our findings might help inform optimal therapeutic strategy for HBV-related compensated cirrhosis, leading to fewer incidences of end-stage liver diseases. Importantly, integrating personalized approaches is pivotal for PEG-IFN-α-2b-based therapy in clinical practice, which will provide a comprehensive and forward-looking perspective during the treatment of HBV-related cirrhosis.