Low HBcrAg and HBsAg levels identify patients most likely to achieve sustained response after nucleos(t)ide analogue cessation: results from a global individual patient data meta-analysis (create study)

Abstract

Background and Aims: Sustained response is observed in a limited number of patients after cessation of nucleo(s)tide analogue (NA) therapy. We aimed to study the relationship between serum levels of hepatitis B core related antigen (HBcrAg) and HBsAg at treatment cessation and subsequent sustained virological response and HBsAg loss. Method: We performed an individual patient data meta-analysis of patients who discontinued NA therapy in compliance with EASL criteria. Patient data was acquired from 9 cohorts from Asia and Europe. HBcrAg and HBsAg were measured at treatment cessation. Studied endpoints included virological response (VR, defined as HBV DNA <2,000 IU/mL) and HBsAg loss at 24 and 48 weeks off-treatment. Retreatment was based on HBV DNA and ALT criteria, and retreated patients were considered non-responders. Results: We analysed 464 patients, of whom 317 (68%) were male and 75 (16%) were HBeAg positive at the time of treatment initiation. Median treatment duration was 294 weeks (IQR: 196–428), and median duration of consolidation therapy was 102 weeks (IQR: 56–165). Patients were treated with ETV/TDF/other in 61%/27%/12%. VR was observed in 234/464 (50%) at week 24 and 155/401 (39%) at week 48 post-treatment. HBsAg loss was observed in 17 (4%) patients. 216 patients were retreated during the follow-up period, with HBV DNA elevation (76%) and ALT flare (21%) the main indications. All resolved without sequelae after retreatment. VR at week 24 was observed in 74/113 (66%) of patients with HBcrAg <2 log at treatment cessation, compared to 44% in patients with HBcrAg >3 log (p = 0.001; figure 1). Similar results were observed at post-treatment week 48. HBsAg loss was observed in 12% of patients with HBcrAg <2 log at treatment cessation, versus 1% in patients with HBcrAg >3 log (p < 0.001). Similarly, VR at week 24 was observed in 21/26 (81%) of patients with HBsAg <10 IU/mL, versus 47% in patients with HBsAg >50 IU/mL (p = 0.001) at treatment cessation. Similar results were obtained when response was assessed at post-treatment week 48. HBsAg loss was observed in 27% of patients with HBsAg <10 IU/mL at treatment cessation, versus 2% in patients with HBsAg >50 IU/mL (p < 0.001). Findings were consistent when limited to the subset of patients who were HBeAg-negative at start of NA therapy. Conclusion: In this global individual patient data meta-analysis of patients who discontinued NA therapy, low HBcrAg (<2 log) and low HBsAg levels (<50 IU/mL) at treatment cessation were associated with highest rates of VR and HBsAg loss. These cut-offs could be used for an improved selection of patients for NA cessation.