Early hepatitis B and delta virus kinetics in patients undergoing liver transplantation Authors: Sara Battistella, Anna Pocurull, Thais Leonel, Sergio Rodríguez-Tajes, Yiliam Fundora, Louis Shekhtman, Leeor Hershkovich, Scott J Cotler, Harel Dahari, Tiffany Fortney, Mark Anderson, Gavin Cloherty, Juan Carlos Hurtado, Maria Saez-Palma, Sabela Lens, Sofía Pérez-del-Pulgar, Xavier Forns Published: 2026 ABSTRACT AbstractHBV recurrence after liver transplantation (LT) can occur despite removal of the infected liver and antiviral prophylaxis. Timing and mechanisms of HBV and HDV recurrence remain unclear. We evaluated post-transplant viral kinetics to address this gap. Twelve HBV patients undergoing LT, including five with HDV-coinfection, were prospectively enrolled. Serial serum and liver samples were rigorously collected before, during, and post-LT to characterize viral kinetics. HBV-DNA, HDV-RNA, HBsAg, and HDAg were detected in all explanted livers, whereas cccDNA was found only in HBV-monoinfected patients. Serum HBV-DNA, detectable in five recipients before LT, became undetectable within hours from LT. In HBV-monoinfected patients, baseline HBsAg levels were low (< 3logIU/mL) and cleared rapidly after LT, with faster disappearance in those receiving HBIG (5 days vs. 6 months). Among HDV-recipients, all treated with HBIG, HBsAg clearance was rapid when baseline HBsAg levels were <3log10IU/mL (12 h) and delayed at higher levels (6 months). HDV-RNA kinetics paralleled HBsAg decline. Post-transplant liver biopsies (reperfusion, 3- and 12-months) tested negative for viral markers, except for isolated intrahepatic HDV detection at 3-months in one patient. No intrahepatic HBV reinfection was observed post-LT. Occasional intrahepatic HDV detection supports the need for improved prophylactic strategies in HDV-recipients.Keywords: HBV, HDV, Liver transplant, HBsAg, Viral clearance, Antiviral prophylaxis, Graft reinfection, HBIG, Intrahepatic markers, cccDNASubject terms: Diseases, Gastroenterology, Medical research, Microbiology Introduction Hepatitis B virus (HBV) remains a common indication for liver transplantation (LT)1,2. The use of hepatitis B immune globulin (HBIG) and nucleos(t)ide analogues (NUCs) has drastically reduced HBV recurrence after LT, leading to patient and graft survival rates comparable to other indications for LT3. For patients undergoing LT due to HBV- or HBV/hepatitis delta virus (HDV)-related liver diseases, the combination of HBIG and high-barrier NUCs is considered the current gold standard of care4–6​. It is intriguing how HBV infection of the hepatic graft occurs despite the removal of the source of HBV (the liver explant) in patients that usually reach transplantation with undetectable HBV-DNA7. Moreover, the administration of high doses of HBIG during the anhepatic phase should neutralize circulating HBV particles. There are only a few studies analysing the presence of HBV markers in liver grafts of transplant recipients8–10. The results of these studies show significantly different prevalences of total intrahepatic HBV-DNA (iHBV-DNA) and covalently closed circular DNA (cccDNA) in liver grafts after LT. The latter may be explained by different patient characteristics (serum HBV-DNA levels at liver transplant, HBeAg status), heterogeneity in antiviral treatments, variations in sample preservation techniques, and the use of non-standardized in-house methods for the detection of intrahepatic viral markers with varying sensitivities. Hence, our study aimed to prospectively analyse HBV and HDV kinetics through serum and tissue sampling both during and immediately after LT. Materials and methods Study population Between 2020 and 2023, patients undergoing LT for HBV-related liver disease, with or without HDV coinfection were prospectively enrolled. Serum samples were prospectively obtained at LT (just before surgery), every 15 min during surgery (both during anhepatic and reperfusion phases), at days 1, 3, 5 and 7, 14, and at months 1, 3, 6 and 12 after LT. Fluid balances during surgery and immediately after LT were recorded. Frozen and formalin fixed paraffin embedded (FFPE) liver samples (explant, reperfusion and at 3- and 12-month post-LT) were collected for the assessment of intrahepatic viral replication markers (iHBV-DNA, cccDNA, intrahepatic HDV-RNA [iHDV-RNA]) and viral antigens (HBsAg, HDAg). HBV prophylaxis after LT included either combined prophylaxis (HBIG+NUCs) or NUCs monotherapy for patients with undetectable HBV-DNA and absence of HDV coinfection. In the combined prophylaxis, HBIG was administered as follows: intravenous 5,000 IU doses during the anhepatic phase and in the first day post-LT, followed by daily 1,000 IU doses from the second to the seventh day after transplantation. Long-term maintenance doses were administered as needed to sustain anti-HBs levels above 150–200 IU/L using intramuscular HBIG. All patients received entecavir (ETV), tenofovir disoproxil (TDF), or tenofovir alafenamide (TAF), according to pre-transplant therapy regimen or indications (Table 1). Table 1Characteristics of patients at the time of liver transplantation.Clinical characteristicsAll patients(n = 12)HBV patients(n = 7)HDV patients(n = 5)Age, years60 (54–65)63 (56–66)57 (56–58)Sex, M9 (75)6 (86)3 (60)Ethnicity Caucasian2 (16.7)2 (28.5)0 African8 (66.7)3 (42.9)5 (100) Asian1 (8.3)1 (14.3)0 South American1 (8.3)1 (14.3)0Indication for LT HCC8 (66.7)6 (85.7)2 (40) Decompensated cirrhosis4 (33.3)1 (14.3)3 (60)MELD score17 (9–31)18 (12–26)16 (11–26)Waiting list time, months1.1 (0.5–2.9)0.6 (0.5–1.6)2.9 (0.5–2.9)Virological status HBeAg positive000 HBcrAg positive6 (50)2 (28.5)4 (80) HBV-DNA detectable5 (58)3* (42.9)2^ (40) HBsAg, IU/mL312 (75–2.450)88 (27–269)7.650 (716–8.512) HDV-RNA, log10IU/mL--5.1 (3.6–5.4) HIV coinfection2 (17)-2 (40)HBV therapy before LT ETV6 (50)6 (85,7)0 TDF4 (33.3)1 (14,3)3 (60) TAF2 (16.7)02 (40)Duration of antiviral therapy before LT, months44.5 (23.9–90.5)40.8 (19.6–77.8)47.4 (45.3–84.9)*1 patient with detectable HBV-DNA but below LoQ (< 10 IU/mL), 2 patients with HBV-DNA of 16 and 23 IU/mL.^Both patients with HBV-DNA detectable but below LoQ (< 10 IU/ml). Characteristics of patients at the time of liver transplantation. *1 patient with detectable HBV-DNA but below LoQ (< 10 IU/mL), 2 patients with HBV-DNA of 16 and 23 IU/mL. ^Both patients with HBV-DNA detectable but below LoQ (< 10 IU/ml). Immunosuppressive (IS) therapy was managed in accordance with the Institutional protocol. Following LT, patients were initiated on either double or triple immunosuppressive therapy based on tacrolimus and prednisone, with the optional addition of mycophenolate mofetil. Corticosteroids were generally withdrawn within six months after transplantation. In selected patients, induction therapy with basiliximab (anti-IL2 receptor antibody) was used to postpone tacrolimus initiation. HBV and HDV viral markers Serum HBsAg was quantified using Elecsys Roche assay (Roche Diagnostics, Mannheim, Germany), with a limit of quantification (LoQ) of 0.05 IU/mL. To ensure we did not miss HBsAg in samples with very low concentrations or in the presence of HBIG, samples were retested using the ultrasensitive ARCHITECT HBsAg Next Qualitative assay (Abbott Diagnostics, Sligo, Ireland), which has an analytical sensitivity at the cut-off of 0.005 IU/mL. Determinations of HBeAg status and anti-HBe were performed by immunoassay using the Advia Centaur® System (Siemens, Erlangen, Germany). HBcrAg levels were assessed by chemiluminescent enzyme immunoassay using LUMIPULSE® G1200 Analyzer (Fujirebio Europe, Gent, Belgium; LoQ < 2.8logU/ml) according to the manufacturer’s instructions. Anti-HBs were quantified using Elecsys Anti-HBs II Roche assay (Roche Diagnostics, Mannheim, Germany), with a quantification range between 2 and 1000 UI/L. Serum HBV-DNA was measured using the Cobas® 6800 system (Roche Diagnostics, Manheim, Germany; LoQ < 10 IU/ml). iHBV-DNA and cccDNA were determined by real-time qPCR, as previously described11. In addition, iHBV-DNA and cccDNA were also tested by digital PCR using the QIAcuity Digital PCR System (Qiagen, Hilden, Germany) following the manufacturer’s instructions. All experiments were performed with appropriate positive (liver explant from a HBeAg positive patient) and negative (explant from a non-HBV infected cirrhotic) controls. Serum and intrahepatic HDV-RNA was measured by one-step real-time qPCR12,13, including serial dilutions of an HDV-RNA International Standard in each assay. Immunodetection of viral antigens in liver tissue To investigate the presence of HBsAg and HDAg in the hepatic parenchyma, we performed dual chromogenic immunohistochemistry (IHC) on FFPE liver samples using the ChromoPlex 1 Dual Detection for BOND and the BOND RX system (Leica Biosystems, Nussloch, Germany). HBsAg and HDAg were detected using mouse anti-HBsAg, (clone A10F1, BioLegend, San Diego, CA) and rabbit anti-HDAg (clone FD3A7, Absolute Antibody, Cleveland, UK), with DAB-brown and AP-pink as chromogens, respectively. Kinetics assessment The anhepatic phase was defined as the time elapsed between explant removal and portal vein anastomosis. Reperfusion phase was defined as the first 6 h after graft perfusion. The post-reperfusion phase was defined as the period from 6 h after graft reperfusion until the last follow up. HBsAg and HDV-RNA clearance curves during the different phases were analysed using linear regression and were fit only up to the first point that reached LoQ or to-not-detectable (TND) as previously done14,15. Since statistical tests showed no association between blood loss and patient kinetics, we did not incorporate potential declines in viral load due to fluids in the analysis. For some patients, a delay before the decline was observed, and the reported half-life (t1/2) corresponds to the period of decline after that delay. Similarly, for some patients there was a sharp decline followed by a flat response, in which case again only the period of decline was considered for the slopes and half-lives. Statistical analyses Categorical data were described using frequency and percentage. Continuous data were described using median with 25% and 75% quartile ranges (IQR). For continuous data, comparison between independent groups was performed using the Mann-Whitney U-test. Spearman correlation analysis was used to assess the association between HBsAg and HDV RNA levels after LT and baseline HBsAg and clearance time. Statistical significance was set at p ≤ 0.05. All analyses were completed using SPSS version 29.0, Python 3.11, and scipy 1.11. Ethics The study was approved by the Hospital Clinic institutional Ethics Committee (HCB/2019/0710) and was conducted in compliance with the principles of the Declaration of Helsinki, Good Clinical Practice guidelines, and local regulatory requirements. All patients provided written informed consent. Results Patients’ characteristics Twelve patients who underwent LT for HBV- (n = 7) or HBV/HDV-related liver disease (n = 5) were prospectively enrolled. Nine patients were male (75%), and the median age at LT was 60 (54–65) years. At LT, all patients were HBeAg negative and were receiving NUCs before LT for 44.5 (23.9–90.6) months, 5 patients (41.7%) had detectable HBV-DNA, and 6 (50%) had positive HBcrAg (Table 1). After LT, 8 patients (including all HDV coinfected) received a combination of HBIG and NUCs therapy, while 4 HBV monoinfected patients underwent NUCs monotherapy. All grafts were anti-HBc negative. Kinetics of serum viral markers in HBV monoinfected patients (n = 7) HBV-DNA became undetectable within few hours post-LT in the three patients who had residual quantifiable viral load and remained undetectable in all patients up to 12 months post-LT. Baseline HBsAg levels were low (< 3 log10IU/mL) in all patients, with a median of 88 (27–269) IU/mL. During the anhepatic phase, patients receiving HBIG (P02, P03, P10) exhibited significant declines in HBsAg (range 0.7–1.9 log10IU/mL), with very short HBsAg t1/2. Despite these results, the correlation between baseline HBsAg levels and clearance timings was not statistically significant (r = 0.599, p = 0.117 in HBIG-treated patients) due to the small sample size. In contrast, patients not treated with HBIG (P11, P12, P13, P14) exhibited either minimal (P12) or no decline in HBsAg levels (Fig. 1A; Table 2). Viral markers kinetics during and after LT. A HBsAg kinetics during and after LT in HBV monoinfected patients. B HBsAg and HDV-RNA kinetics during and after LT in HDV coinfected patients. In all patients who received HBIG, long-term maintenance doses were administered as needed after the first week from LT to sustain anti-HBs levels above 150–200 IU/L. The shaded region represents the anhepatic phase. ​ Open circles, squares, and diamonds represent detectable values; filled symbols correspond to non-detectable values. The anhepatic phase (AP) was defined as the time elapsed between explant removal and portal vein anastomosis. Reperfusion phase was defined as the first 6 h after portal vein anastomosis. The post-reperfusion phase was defined as the period from 6 h after graft reperfusion until the last follow up. Table 2HBsAg kinetics during anhepatic, reperfusion and post-reperfusion phases.Pt.HDVHBIGAnhepatic phaseReperfusion phasePost-Reperfusion phaseHBsAg at start of AP (log10IU/mL)HBsAg at end of AP (log10IU/mL)Decline (log10IU/mL)Slope (log10/hr)HBsAg half-life (hr)Blood loss (mL)HBsAg at start of RP (log10IU/mL)HBsAg at end of RP (log10IU/mL)Decline (log10IU/mL)Slope (log10/hr)HBsAg half-life (hr)HBsAg days reached TNDSlope until TND (log10/hr)HBsAg half-life (hr)HBsAg end FU (log10IU/mL)P02NoYes2.60.71.9− 0.58b1.224000.7− 0.71.4− 0.272.65− 0.001NETNDP03NoYes2.41.70.7− 0.22b3.120001.71.10.5− 0.1*6.9*2− 0.31NETNDP10NoYes2.51.31.2− 5.3†0.1†16001.10.50.6− 0.22*3.2*5− 0.010NETNDP11NoNo1.31.300.01Plateau18181.21.10.1− 0.14#Plateau90− 0.002NETNDP12NoNo1.11.00.1− 0.0417.326001.00.60.4− 0.04Plateau7− 0.027*25.6TNDP13NoNo1.92.0− 0.10.09Plateau12001.91.60.3− 0.06*11.6*180− 0.006NETNDP14NoNo0.91.0− 0.10.07Plateau28000.90.80.10.0Plateau14− 0.019*36.5TNDP07YesYes2.60.71.9− 1.750.4*25000.5− 0.51.0− 0.125.80.5− 0.065.0TNDP09YesYes1.9− 0.12.0− 2.010.3*3000− 0.6TND1.4− 0.088.60.250‡NETNDP01YesYes4.23.70.5− 0.15b4.63503.73.7a00.0Plateau90− 1.79NETNDP04YesYes3.93.40.5− 0.14b4.930803.43.30.1− 0.01Plateau180− 0.46NETNDP08YesYes3.83.60.2− 0.07Plateau16803.53.50− 0.01Plateau30− 2.83NETNDFor patients P01, P02, P03 and 04, sampling during the anhepatic phase was limited to 2 samples.†Only two data points are available therefore calculating a p-value was not possible;# Only fit through 1 h post reperfusion due to measurements missing.‡Patient was TND before the end of reperfusion phase;*p < 0.05, indicates a slope different than 0 (or plateau);a Only up to 2 h post reperfusion;b Slope is based on measurement prior to and immediately after AP phase since no measurements were taken during the AP phase. HBsAg kinetics during anhepatic, reperfusion and post-reperfusion phases. For patients P01, P02, P03 and 04, sampling during the anhepatic phase was limited to 2 samples. †Only two data points are available therefore calculating a p-value was not possible; # Only fit through 1 h post reperfusion due to measurements missing. ‡Patient was TND before the end of reperfusion phase; *p < 0.05, indicates a slope different than 0 (or plateau); a Only up to 2 h post reperfusion; b Slope is based on measurement prior to and immediately after AP phase since no measurements were taken during the AP phase. During the reperfusion phase, HBsAg levels continued to decline in patients who received HBIG, while HBsAg levels plateaued in 3/4 cases who did not undergo HBIG (Table 2; Fig. 1A). While all patients ultimately achieved undetectable HBsAg post-LT, the interval to clearance was shorter in those receiving HBIG (range 2–5 days) compared to those who did not (7 days – 6 months), although not statistically significant (p = 0.057). Results using the HBsAg Next Qualitative assay (Abbott Diagnostics, Sligo, Ireland), were concordant to that of the Elecsys Roche assay (Roche Diagnostics, Mannheim, Germany), except for two early time points in which the HBsAg Next Qualitative assay detected residual amounts of HBsAg that the Roche assay did not (data not shown). Anti-HBs levels were measured at baseline, reperfusion, days 1, 7, and 14 post-LT, and at 1, 3–6, and 12 months after LT. All HBV monoinfected patients who received HBIG exhibited high anti-HBs titres (> 150 IU/L) during the first month post-LT, with levels proportional to the administered doses thereafter. As expected, among patients who did not receive HBIG, anti-HBs levels remained < 10 IU/L until one month post-LT, when 3 of 4 patients developed low anti-HBs levels (range 22.1–240 IU/L). However, these levels were not sustained at 3–6 and 12 months of follow-up. Kinetics of serum viral markers in HDV coinfected patients (n = 5) HBV-DNA became undetectable within the first hours post-LT in 2 patients with residual HBV-DNA at time of transplantation and remained undetectable in all HDV coinfected patients up to the final measurement at 12 months post-LT. Among HDV coinfected patients, all of whom received HBIG, baseline HBsAg levels were significantly higher compared to HBV monoinfected patients (median 7,650 [716–8,612] IU/mL). During the anhepatic and reperfusion phases, the rate of HBsAg clearance was associated with baseline HBsAg levels. Patients with baseline HBsAg < 3 log10IU/mL (P07 and P09) exhibited sharp declines during the anhepatic phase (1.9 and 2 log10IU/mL), which continued during the reperfusion phase (1 and 1.4 log10IU/ml) (Fig. 1B). In contrast, in patients with baseline HBsAg > 3 log10IU/mL (P01, P04, P08), HBsAg plateaued, with minimal declines during the anhepatic and reperfusion phases (Fig. 1B; Table 2). In the post-reperfusion phase, all HDV coinfected patients eventually achieved undetectable HBsAg levels. Those with lower baseline HBsAg (< 3 log10IU/mL) reached undetectable HBsAg within 6–12 h, while those with higher baseline levels required a significantly longer period of time (1–6 months) (Table 2). Results using the HBsAg Next Qualitative assay were concordant to those provided by the Roche assay (data not shown). HDV-RNA kinetics mimicked that of HBsAg. Baseline HDV-RNA levels ranged from 2.61 to 6.4 log10IU/mL. In patients with HBsAg < 3 log10IU/mL (P07 and P09), HDV-RNA declined rapidly during both the anhepatic (1.2 and 2.0 log10IU/mL) and reperfusion phases (4.5 and 4.3 log10IU/mL) (Fig. 1B). In contrast, patients with HBsAg > 3 log10IU/mL (P01, P04, P08) showed slower HDV-RNA reductions during the anhepatic (max decline: 0.4 log10IU/mL) and reperfusion phase (max decline: 1.9 log10IU/mL), irrespective of baseline HDV-RNA levels (Fig. 1B; Table 3). This was confirmed by a strong and statistically significant correlation (p < 0.001) between HBsAg and HDV RNA levels after LT, with Spearman’s r > 0.9 in all patients. Table 3HDV-RNA kinetics during anhepatic, reperfusion and post-reperfusion phases.Pt.HBIGAnhepatic phaseReperfusion phasePost-reperfusion phaseHDV at start of AP (log10IU/mL)HDV at end of AP (log10IU/mL)Decline (log10IU/mL)HDV half-life (hr)HDV at start of RP (log10IU/mL)HDV at end of RP (log10IU/mL)Decline (log10IU/mL)Slope (log10/hr)HDV half-life (hr)HDV days reached TNDP07Yes4.062.91.20.6*2.5TND4.5− 18.1$0.040.03P09Yes4.112.22.00.5a2.3TND4.3− 17.1$0.040.02P01Yes6.46.40.0169‡6.44.51.9− 0.39Plateau90P04Yes2.612.30.32.61‡2.32.10.2− 0.04Plateau3P08Yes6.045.70.4Plateau5.65.20.30.04Plateau7‡ Half-life and decline are calculated from baseline to first point of reperfusion phase as no measurements were taken in the anhepatic phase.* p < 0.05, indicates a slope different than 0 (or plateau).a Though HDV declined 2 logs, due to fluctuations p = 0.07.$ Minimal estimate. HDV-RNA kinetics during anhepatic, reperfusion and post-reperfusion phases. ‡ Half-life and decline are calculated from baseline to first point of reperfusion phase as no measurements were taken in the anhepatic phase. * p < 0.05, indicates a slope different than 0 (or plateau). a Though HDV declined 2 logs, due to fluctuations p = 0.07. $ Minimal estimate. All HDV coinfected patients eventually achieved undetectable HDV-RNA levels post-LT, but clearance was faster in those with HBsAg < 3 log10IU/mL (30–45 min from reperfusion), compared to those with HBsAg > 3 log10IU/mL (3 days–3 months) (Table 3). Blood loss during surgery did not appear to influence HBsAg nor HBV-DNA and HDV-RNA kinetics (data not shown). Although all HDV coinfected patients received the same HBIG regimen, as expected, those with high baseline HBsAg levels (> 3 log₁₀IU/mL) exhibited considerably lower anti-HBs levels compared to those with low baseline HBsAg levels (< 3 log₁₀IU/mL) during and after LT. Longitudinal assessment of HBcrAg after LT HBcrAg levels were assessed at 7 days, 3–6 months, and 12 months following LT. Six patients (50%) were HBcrAg-negative at the time of LT and remained negative throughout the entire follow-up. Among HBV monoinfected patients, only 2/7 (29%) were HBcrAg-positive at LT. One of them became negative shortly after LT and remained so, while the other showed a decline in HBcrAg levels (from 5.9 to 4.5 IU/mL) at day 7, and tested negative at both 3 and 12 months. Four out of five (80%) HDV coinfected patients were HBcrAg-positive at LT, but they became immediately negative after LT and remained so at all follow-up points. Intrahepatic biomarkers of viral replication Using qPCR, iHBV-DNA was detected in liver explants from both HBV and HDV coinfected patients, while cccDNA was only detectable in patients with HBV monoinfection (Fig. 2). In contrast, neither iHBV-DNA nor cccDNA was detected in reperfusion or follow-up biopsies collected at 3 months (6 HBV and 2 HDV patients) and 12 months (3 HBV and 1 HDV patients) after LT, including those who did not receive HBIG. For HDV coinfected patients, iHDV-RNA was detected in all liver explants, while reperfusion biopsies were negative. At 3 months post-LT, only one HDV-coinfected patient (P01) had detectable iHDV-RNA (17 IU/µRNA tot), despite undetectable serum markers at the same time point. This patient had the highest baseline HDV-RNA levels (6.4 logIU/mL). Intrahepatic nucleic acids quantification in explants from HBV monoinfected and HDV coinfected LT recipients. HBV monoinfected patients are represented by blue circles; HDV coinfected patients are represented by red circles. Open circles represent patients with undetectable serum HBV DNA at the time of transplantation, whereas filled circles represent patients with detectable serum HBV DNA at the time of transplantation. Results obtained with ddPCR were consistent with those obtained with qPCR. At explant, iHBV-DNA was detectable in all patients by ddPCR but at substantially lower levels than those measured by qPCR. With respect to cccDNA, all HDV-coinfected patients were negative except one individual with a very low load (0.0002 cp/cell) who tested negative with qPCR, whereas among HBV monoinfected patients, all but one were positive, again at levels markedly lower than those detected by qPCR. Consistent with qPCR findings, all samples obtained after LT, including reperfusion, 3- and 12-month biopsies, were negative for both iHBV-DNA and cccDNA. Samples from liver explants and reperfusion biopsies were available for IHC analysis in all patients, while 3-month and 12-month biopsies for IHC viral antigen assessment were available for 10 patients (7 HBV and 3 HDV) and 3 patients (1 HBV and 2 HDV), respectively. HBsAg was detected in all liver explants and HDAg was detected in all 5 explants from HDV-coinfected patients. In contrast, liver samples obtained after LT were negative for both HBsAg and HDAg at all time points, except for P01 who exhibited mild expression of HDAg at 3 months post-LT (Fig. 3). This patient also had low levels of iHDV-RNA but undetectable serum markers at the same time point. HBsAg and HDAg detection in explanted and post-transplant liver biopsies from HBV monoinfected and HDV coinfected LT recipients. The left upper panel shows HBsAg and HDAg in an explant sample from an HDV coinfected patient (P01). Right upper panel shows a few hepatocytes positive for HDAg at the 3-month post-LT biopsy. The left lower panel shows HBsAg in an explant sample from an HBV monoinfected patient (P03). The patient tested negative for HBsAg in all samples collected after LT (right lower panel, biopsy taken at 12 months after LT). Image magnification: 10x. Discussion The precise mechanisms leading to HBV graft infection are not completely known5. One possible explanation is the persistence of residual circulating infectious virions that are not fully neutralized by HBIG and remain below the detection threshold of current diagnostic assays. Even at very low levels, these particles may retain infectivity and contribute to intermittent or sustained viral persistence. Alternative (and somewhat controversial) hypotheses, include the presence of extrahepatic reservoirs, particularly peripheral blood mononuclear cells, which have been shown to harbour replication-competent HBV and may serve as a source of viral reactivation or reseeding of hepatocytes16. Additionally, extracellular vesicles have emerged as potential mediators of viral persistence: these membrane-bound structures can encapsulate viral nucleic acids and proteins, shielding them from immune recognition and antibody-mediated neutralization, including HBIG. By facilitating intercellular transfer of viral components in a protected form, extracellular vesicles might represent a mechanism by which HBV evades host defences and maintains low-level infection despite prophylactic or therapeutic interventions17. Thus, we designed a prospective study meticulously assessing viral markers in both serum and liver samples during and after LT. At the time of transplant, all our patients had been on NUCs for a median time longer than 3 years and most had undetectable or unquantifiable serum HBV-DNA; in those with positive HBV-DNA, it became undetectable rapidly after graft reperfusion. Regarding HBIG administration, we observed a clear impact on HBsAg kinetics in patients with low baseline HBsAg levels, with a very rapid serum clearance. In contrast, among patients who did not receive HBIG (all of whom were HBV monoinfected with undetectable HBV-DNA at LT), HBsAg persisted in circulation for several weeks despite low baseline levels, although ultimately became undetectable. Importantly, HBIG administration did not appear to have a significant impact on HBsAg kinetics if baseline HBsAg levels were higher. These data were obtained from HDV coinfected patients, since all of them received HBIG. Although this finding remains exploratory and requires validation in larger cohorts, it suggests that in patients with low HBsAg levels at LT, HBIG administration might be limited to the immediate post-transplant period. In contrast, in patients with higher baseline HBsAg levels, the administered HBIG doses were probably insufficient to fully neutralize circulating HBsAg, resulting in its prolonged persistence18,19. In these cases, the kinetics of HBsAg elimination appear to resemble that observed in patients with low baseline HBsAg levels who do not receive HBIG, aligning with the reported natural half-life of HBsAg20. To further understand HBV and HDV kinetics, we also analysed liver samples obtained during (liver explants and reperfusion biopsies) and after transplantation. As expected, all liver explants tested positive for iHBV-DNA, albeit at very low levels. The detection of iHBV-DNA in patients undergoing long-term NUCs therapy is not surprising, as it may originate from integrated HBV-DNA, which is not targeted by antiviral therapy21. Conversely, cccDNA was detected in patients with HBV monoinfection but not in those coinfected with HDV. This is in line with other studies demonstrating that HDV reduces HBV replication both in vitro22 and in vivo23, although the exact underlying mechanisms have not been fully elucidated24. After LT, we were unable to detect any HBV markers (iHBV-DNA, cccDNA) in biopsies taken during the reperfusion phase or later (3 months and 1 year after LT), even in patients with slower HBsAg clearance. The detection of HBV intrahepatic markers after transplantation remains a topic of ongoing debate8,9. Villeret et al.10, recently reported the presence of total iHBV-DNA, cccDNA, and 3.5-kb RNA in a significant proportion of graft samples obtained less than one hour after reperfusion and at 1-year of follow-up, despite the absence of detectable serum viral markers and the administration of HBIG. More in line with our results, Lenci et al.9, analysed post-transplant liver samples from 44 patients and, among those without serological HBV recurrence, detected total iHBV-DNA in only 3 patients (6.8%) and cccDNA in only one patient (2.3%). The discrepancies observed among studies may be partially attributed to several factors, including differences in patient ethnicity, higher viral load in those with detectable viremia, shorter duration of pre-transplant antiviral therapy, and a higher prevalence of detectable pre-transplant HBcrAg in the French cohort compared to ours. Additionally, different sample preservation methods (RNA Later-stored samples vs. liquid nitrogen frozen samples) may have contributed. Although we cannot entirely exclude that differences between the studies are attributable to different sensitivities of PCR across studies, we double checked all the liver samples using both qPCR and ddPCR, obtaining consistent results. From a pathophysiological perspective, it is difficult to explain how cccDNA (and iHBV-DNA) can be found in the liver graft just minutes after reperfusion in patients with undetectable HBV-DNA in the blood. In fact, a recent study from Shekhtman et al. ​18​ in primary human hepatocytes estimated the time from HBV inoculation to the onset of productive HBV infection to be 22.1±0.3 h. On the contrary, since HDV-RNA remained detectable for hours or days following LT, we were not surprised to detect low levels of both iHDV-RNA and HDAg in one coinfected patient 3 months after LT. The latter confirms that at some point after LT, HDV was able to infect hepatocytes. However, the lack of circulating HDV markers indicates that HDV could not complete its life cycle due to the absence of HBsAg in those few hepatocytes bearing HDAg. Thus, in the absence of HBsAg, hepatocytes expressing HDV-Ag are of limited immediate clinical relevance. However, this condition may not be entirely benign. In fact, in patients not maintained on long-term HBIG prophylaxis, the risk of HBV reactivation persists, and renewed HBsAg expression could restore the capacity for HDV assembly and propagation. Bulevirtide, a recently approved entry inhibitor25, may represent a valuable option to prevent HDV entry into hepatocytes, particularly in patients with high HBsAg levels at LT in whom complete neutralization may not be attainable; however, its role in this setting requires further investigation in dedicated studies. This study has some limitations. First, our cohort is small. To mitigate this limitation, we conducted a detailed and thorough analysis of viral biomarker kinetics in serum and liver samples during and after LT. Second, the use of HBIG during and after LT may have interfered with serum HBsAg detection. To address this, we used a novel qualitative technique to improve HBsAg detection even in the presence of HBIG. Finally, current methodologies for the detection of intrahepatic HBV-DNA, and particularly cccDNA, remain poorly standardized due to its low abundance and technical challenges associated with its measurement, limiting direct comparison across studies. Therefore, although we sought to mitigate this limitation by employing two different techniques (qPCR and ddPCR), we cannot definitively exclude its presence in our samples. In conclusion, despite variable HBsAg clearance kinetics following liver transplantation, we were unable to detect HBV markers (including cccDNA) in serum and liver grafts at different time points after transplantation, even in patients receiving HBIG-free prophylaxis after LT. On the contrary, HDV could be sporadically detected in the liver grafts. Further studies evaluating additional prophylactic strategies in HDV coinfected patients awaiting LT are needed. Acknowledgements The authors extend their appreciation to the Instituto de Salud Carlos III (ISCIII) and European Union through the projects PI22/00013, PI19/00036, PI20/00609, FI23/00089, CM21/00081 and CM17/00015; to the Secretaria d’Universitats i Recerca del Departament d’Economia i Coneixement (grant 2021_SGR_01322) and the CERCA Programme/Generalitat de Catalunya; to the Binational Science Foundation (grant 207745), and US NIH (grants R01AI144112, R01AI078881 and R01AI146917) for funding this work. Author contributions Contributed to the study concept and design: SPP, XF. Patient selection/inclusion: AP, SRT, YF, SL, XF. Acquired data: SB, AP, TL, SRT, YF, JCH, SL, SPP, XF. Contributed to analysis and interpretation of data: SB, LS, LH, SC, HD, SL, SPP, XF. Drafted the manuscript: SB, SL, SPP, XF. Contributed to critical revisions and approved the final manuscript: all authors. Funding This study was sponsored in part by the Instituto de Salud Carlos III (ISCIII) through projects PI22/00013 (SPP and XF), PI19/00036 (SPP) and PI20/00609 (SL), co-funded by the European Union. XF has received support from the Secretaria d’Universitats i Recerca del Departament d’Economia i Coneixement (grant 2021_SGR_01322) and the CERCA Programme/Generalitat de Catalunya. MSP is supported by the PFIS program (FI23/00089) and AP and SRT by the Río Hortega program (CM21/00081 and CM17/00015, respectively) from the ISCIII, both co-funded by the European Social Fund (ESF). The study was also supported by the Binational Science Foundation grant number 207745, and US NIH grants R01AI144112, R01AI078881 and R01AI146917. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Data availability Data are available from the corresponding author upon reasonable request. Declarations Conflict of interest HD serves as advisor for VIR. Ethical approval The study was approved by the Hospital Clinic institutional Ethics Committee (HCB/2019/0710) and was conducted in compliance with the principles of the Declaration of Helsinki, Good Clinical Practice guidelines, and local regulatory requirements. Informed consent All patients provided written informed consent. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Sara Battistella and Anna Pocurull have contributed equally. Contributor Information Sofía Pérez-del-Pulgar, Email: sofiapp@clinic.cat. Xavier Forns, Email: xforns@clinic.cat. References Associated Data Data Availability Statement Data are available from the corresponding author upon reasonable request.