Study population
This study included 107 patients with a diagnosis of an oncohaematological disease. Most patients were male (57%) and Italian (88.8%), with a median (IQR) age of 66 (58‐77) years (Table 1). The most frequent haematological malignancy was non-Hodgkin lymphoma (NHL) (36.5%), followed by acute myeloid leukaemia (AML) (15.9%), multiple myeloma (MM) (14.0%), and chronic lymphocytic leukaemia (CLL) (12.1%) (Table 1). Most patients received rituximab-containing chemotherapy (39.2%), while 40 (37%) patients underwent HSCT as part of the therapy course. Among them, 32 (80%) underwent allogeneic HSCT.
At HBV baseline screening, before starting the immunosuppressive regimen, patients were HBsAg-negative and anti-HBc-positive. Most of them (79.3%) were also positive for anti-HBs (median [IQR] titer: 152 [47–976] mIU/ml) (Table 1). At the end of the study evaluation, 61/107 (57%) patients completed the recommended course of anti-HBV prophylaxis (Table S1).
Occurrence of HBV-R
During the entire patients’ monitoring (median [IQR duration] time: 44 [31–56] months), HBV-R was observed in 17/107 (15.8%) patients, with a 5-year cumulative risk of 27.4% (Table 2 and Fig. 1A). The 5-year cumulative risk of HBV-R was significantly higher in patients undergoing HSCT (without statistically significant differences between autologous and allogenic HSCT: 54.3% vs. 51.1%) than in those receiving chemotherapies with or without rituximab (Fig. 1B).
Notably, most cases of HBV-R (11/17, 64.7%) occurred after completing anti‐HBV prophylaxis (range: 1–27 months after prophylaxis completion). In the remaining 35.3%, HBV-R occurred while receiving antiviral prophylaxis (Table 2). No patients harboured drug-resistant strains.
At HBV-R, the median (IQR) HBV-DNA was 44 (27–40,509) IU/mL, and ALT was above the normal level (>UNL) in 44% of patients (median [IQR] ALT: 81 (49‐541) U/L), indicating a diagnosis of biochemical HBV-R in less than half of the patients (Table 2). These data are indicative of an early HBV-R diagnosis as a result of strict patient monitoring (every 3 months), followed in this study protocol. Median anti-HBc titer was 22 (3–46) COI.
Predictive role of serological HBV markers in the diagnosis of HBV-R
An anti-HBc >3COI combined with an anti-HBs persistent or declining to <50 mIU/ml during patient monitoring was significantly correlated with a higher risk of developing HBV-R (OR [95% CI ]: 9.1[2.7–30.2]; P<0.001). Indeed, 63% of patients with the combination of anti-HBc >3COI plus anti-HBs <50 mIU/ml experienced HBV-R compared to 15% of patients without this combination (P<0.001) (Fig. 2A). Notably, this result was also confirmed in the subset of patients experiencing HBV-R after completing anti-HBV prophylaxis (OR [95% CI ]: 8.8[2.0–38]; P=0.005). In this setting, 55% of patients with the combination of anti-HBc >3COI plus anti-HBs <50 mIU/ml experienced HBV-R versus 14% without this combination (P=0.005) (Fig. 2B).
A further step of this study was to evaluate the predictive role of virological markers in HBV-R diagnosis. The on-monitoring analysis revealed that the detection of HS_HBs and/or of serum HBV-DNA (target detected below LLOQ), confirmed at least 2 time points, was a predictive factor for HBV-R (OR [95% CI ]: 13.8 [3.6–52.6]; P<0.001). Indeed, 50% of patients positive to HS-HBsAg and/or to serum HBV-DNA versus 7% never positive to these markers experienced HBV-R (P<0.001) (Fig. 2C). This result was also confirmed in the subset of patients experiencing HBV-R after completing anti-HBV prophylaxis (P<0.001) (Fig. 2D), suggesting that monitoring these biomarkers could be useful to identify patients requiring prolonged or enhanced prophylaxis.
Multivariate analysis confirmed that either the combination of anti-HBc>3COI+ anti-HBs<50 mIU/ml or positivity for HS-HBsAg and/or serum HBV-DNA (target detected below LLOQ) were independent predictive factors for HBV-R (OR [95% CI ]: 7.2 [1.4–39.2], P=0.020 and 5.3 [1.0–27.8], P=0.049, respectively (Table 3).
Outcome of patients experiencing HBV-R
Among the 17 patients experiencing HBV-R, 16 received antiviral therapy with entecavir and/or tenofovir disoproxil fumarate/tenofovir alafenamide and were monitored for a median (IQR) time of 19 [15‐34] months, while one patient was lost to follow-up.
By analysing treatment outcome, ALT normalization was achieved by all patients after a median (IQR) time of 3 (3–9) months, and 81.3% (13/16) achieved virological suppression within 6 months of antiviral therapy.
Overall, 25% of patients (4/16) died from the progression of the underlying oncohaematological disease, while no death due to hepatic failure occurred.
In a subset of 6 (35%) out of 17 patients, serum HBV RNA was also quantified at HBV-R and during antiviral treatment (median [IQR] time of follow-up: 12 [11–24] months). At HBV-R, serum HBV-RNA ranged from <LLOQ to 6.6 log IU/ml. In 5 out of 6 patients, both serum HBV-DNA and HBV-RNA became undetectable within 6 months of treatment and remained undetectable in the subsequent follow-up, suggesting rapid and durable silencing of cccDNA transcriptional activity. A different scenario was observed in the remaining patient, characterized at HBV-R by serum HBV-DNA and HBV-RNA of 6.0 log IU/ml and 6.6 log IU/ml, respectively. Notably, after 9 months of treatment, a significant decay (up to 2.8 log IU/ml) was observed exclusively for serum HBV-DNA, while serum HBV-RNA remained stable at approximately 6 log IU/ml, indicating elevated cccDNA transcriptional activity. To date, the patient is still in entecavir treatment, and the last (June 2021) HBV DNA value was undetectable below 20 UI/mL. He is continuing regular follow up'
The added value of HS HBsAg quantification: a case report
Here, a clinical case is reported, highlighting the importance of HS-HBsAg quantification in unravelling minimal HBV replicative activity.
The patient was a 58-year-old Italian male diagnosed with Hodgkin lymphoma. The patient was anti-HBc isolated at baseline screening and thus received proper antiviral prophylaxis before starting chemotherapy, containing adriamycin, bleomycin, vinblastyn, and dacabarzin. During 24 months of patient monitoring, serum HBV-DNA was persistently undetectable, and HBsAg was persistently negative by classical assays, while transaminases fell within the normal values. Antiviral prophylaxis was suspended 18 months after completing chemotherapy. Unexpectedly, 1 month after prophylactic suspension, the patient experienced HBV-R with HBV-DNA and HBsAg reappearance in serum and ALT levels rising to 95 IU/ml (Fig. S1). The highly sensitive quantification of HBsAg revealed that HBsAg was already detected at each time point analysed during antiviral prophylaxis, supporting the intrahepatic activity of the HBV reservoir that can explain HBV-R occurrence immediately after the suspension of antiviral prophylaxis (Figure S1).