True reactivation versus transient viremia: a retrospective analysis of HBV reactivation in resolved infection Authors: Tadashi Inuzuka, Atsushi Takai, Masayuki Ueno, Masako Mishima, Haruhiko Takeda, Ken Takahashi, Takahisa Kayahara, Hiroyuki Takabatake, Hirokazu Mouri, Hiroshi Seno Published: 2026 ABSTRACT Hepatitis B virus (HBV) reactivation is a major complication of immunosuppressive therapy or chemotherapy in patients with resolved HBV infection, yet the clinical significance of low-level HBV DNA reappearance remains unclear. We retrospectively analyzed 69 patients with resolved HBV infection who developed HBV DNA reappearance under immunosuppression. Among them, 38 patients were initially managed without immediate nucleos(t)ide analogue therapy, allowing assessment of HBV DNA kinetics. True reactivation was defined as HBV DNA ≥ 3.3 log IU/mL after initial detection. Of the 38 monitored patients, 7 (18%) progressed to true reactivation, 18 (47%) achieved spontaneous viral clearance, and 13 initiated antiviral therapy before meeting either definition. Distinct risk patterns emerged: all patients with an initial HBV DNA level ≥ 2.0 log IU/mL progressed to true reactivation, whereas only 2 of 20 with lower levels did so. Hematologic malignancy was also strongly associated with clinically significant reactivation, with true reactivation occurring in 6 of 12 patients compared with 1 of 13 patients with other diseases. These findings suggest that low-level HBV DNA reappearance under immunosuppression represents transient viremia rather than progressive reactivation. Risk stratification based on initial viral load and underlying disease may enable individualized management, reducing unnecessary antiviral therapy while maintaining patient safety.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-56251-9. Introduction An estimated one-third of the global population has been exposed to hepatitis B virus (HBV)1,2. HBV is a partially double-stranded DNA virus that replicates in hepatocytes via covalently closed circular DNA (cccDNA) formed after infection, enabling lifelong viral persistence, and remains a major global health concern. Clinically, HBV reactivation is a well-recognized complication in patients undergoing chemotherapy or immunosuppressive therapy, potentially causing severe or even fatal liver injury3,4. Particularly among individuals with resolved HBV infection—defined as hepatitis B surface antigen (HBsAg)-negative but hepatitis B core antibody (HBcAb)-positive and/or hepatitis B surface antibody (HBsAb)-positive—many have little awareness of prior HBV infection; however, the virus can persist in a latent form by harboring cccDNA in hepatocytes and may reactivate under immunosuppression5,6. HBV reactivation is generally defined as an increase in viral replication compared with baseline, though the precise criteria differ slightly among major guidelines. Major guidelines for HBsAg-positive carriers use relative or absolute increases in HBV DNA, whereas for resolved infection they generally define reactivation as new detectability of HBV DNA or HBsAg seroreversion7–9. Over the last decade, American Gastroenterological Association (AGA), American Association for the Study of Liver Diseases (AASLD), European Association for the Study of the Liver (EASL), and Asian Pacific Association for the Study of the Liver (APASL) have emphasized risk-based prevention with nucleos(t)ide analogues (NAs) or close monitoring7–10. In contrast, the Japan Society of Hepatology (JSH) adopts a monitoring-centered strategy, prioritizing serial HBV DNA testing and deferring antiviral therapy until virological reactivation is documented. Specifically, NA therapy is recommended once HBV DNA exceeds 1.3 log IU/mL (≥ 20 IU/mL) in patients with resolved infection11. However, whether this low-level detectability reliably marks clinically meaningful reactivation has not been rigorously verified. This divergence in management strategies has important implications. In settings where prophylactic antiviral therapy is widely implemented, opportunities to observe the natural course of untreated low-level HBV DNA detectability are inherently limited. In contrast, the Japanese approach—supported by widespread access to frequent HBV DNA monitoring—creates a unique clinical environment in which low-level viral reappearance can be longitudinally assessed without immediate antiviral intervention. This context provides a rare opportunity to examine whether all detectable HBV DNA represents clinically meaningful reactivation or whether a subset reflects transient, self-limited viremia. A fundamental unresolved issue is therefore whether current definitions of HBV reactivation, which treat any detectable HBV DNA as clinically significant, adequately discriminate progressive viral amplification requiring therapy from transient viral reappearance that does not warrant clinical intervention. Clarifying this distinction is essential to avoid unnecessary antiviral therapy while ensuring timely treatment of true reactivation. To address this gap, we investigated the longitudinal dynamics of HBV DNA in patients with resolved HBV infection undergoing immunosuppressive or chemotherapeutic therapy within a monitoring-based clinical framework. To contextualize these findings, we also examined HBV DNA variability in untreated HBsAg-positive patients in whom HBV DNA had once fallen below the detection limit, as a reference cohort. Our objectives were to (i) characterize the natural course of low-level HBV DNA detectability, (ii) identify clinical factors associated with progressive reactivation, and (iii) explore an operational approach to distinguish clinically significant reactivation from transient low-level viremia. Results Longitudinal HBV DNA variability in inactive HBsAg-positive carriers (reference cohort) We analyzed 71 HBsAg-positive, HBeAg-negative patients with chronic HBV infection who became HBV DNA–undetectable at least once during follow-up. The baseline clinical and virological characteristics of these 71 patients are summarized in Supplementary Table S1. The median age at the time when HBV DNA first became undetectable was 62 years (range: 15–88), with 30 males and 41 females. The median serum ALT level was 17 U/L (range: 6–108), and the median platelet count was 22.3 × 10⁴/µL (range: 9.5–38.0). All 71 patients were persistently HBeAg-negative, and all were HBsAg-positive at baseline. During follow-up, however, HBsAg loss was documented in 13 of the 71 patients (18.3%), reflecting the possibility of eventual seroclearance in HBsAg-positive patients whose HBV DNA had once fallen below the detection limit. Based on the longitudinal HBV DNA profiles after the first achievement of undetectable HBV DNA during follow-up, patients were classified into two groups: (i)those who persistently maintained undetectable HBV DNA throughout follow-up (hereafter referred to as the “suppression” group) and.(ii)those who experienced at least one episode of HBV DNA reappearance after an undetectable state (hereafter referred to as the “intermittent viremia” group). (i)those who persistently maintained undetectable HBV DNA throughout follow-up (hereafter referred to as the “suppression” group) and. those who persistently maintained undetectable HBV DNA throughout follow-up (hereafter referred to as the “suppression” group) and. (ii)those who experienced at least one episode of HBV DNA reappearance after an undetectable state (hereafter referred to as the “intermittent viremia” group). those who experienced at least one episode of HBV DNA reappearance after an undetectable state (hereafter referred to as the “intermittent viremia” group). According to this classification, 19 cases (26.8%) were assigned to the suppression group, whereas 52 cases (73.2%) were assigned to the intermittent viremia group, indicating that even after HBV DNA becomes undetectable, re-detection on subsequent measurements is common (Figure S1A–C). Baseline clinical and virological characteristics of patients in the suppression group and the intermittent viremia group are summarized in Supplementary Table S1. There were no significant differences between the two groups in age, sex distribution, ALT levels, or platelet counts. However, HBsAg loss during follow-up was observed significantly more frequently in the suppression group than in the intermittent viremia group (42.1% vs. 9.6%, p < 0.05). Next, among the 52 cases classified as ‘intermittent viremia’, HBV DNA reappearance at ≥ 1.0 log IU/mL was observed on a total of 94 occasions. The median number of reappearance episodes per patient was 1 (range: 1–6), and 23 of 52 patients (44.2%) experienced multiple (≥ 2) episodes of HBV DNA reappearance. This indicates that, as illustrated in the representative case in Figure S1B, many patients experienced multiple episodes of HBV DNA disappearance followed by re-detection. Analysis of the viral load trends during these events (Figure S1D) shows that nearly all remained below 3.3 log IU/mL, corresponding to viral loads typically observed during the inactive HBV carrier phase. Only one event (1.1%) exceeded this threshold, but the increase was transient and not sustained, with HBV DNA levels promptly returning below 3.3 log IU/mL. When focusing on reappearance events that occurred within 6 months (182 days) from the undetectable status, 63 of the 94 events (67%) were documented, with a median time to reappearance of 175 days (range: 81–182 days) and a median HBV DNA level of 1.5 log IU/mL (range: 1.0–2.6 log IU/mL) (Figure S1E). Notably, none of these short-term reappearance events exceeded 3.3 log IU/mL. These findings support the concept that in the setting of chronic HBV infection without immunosuppression, once HBV DNA becomes undetectable, subsequent fluctuations generally do not result in marked viral rebounds, with reappearance levels rarely exceeding 3.3 log IU/mL, the threshold consistent with the inactive phase of chronic infection. Clinical characteristics of reactivation cases monitored without NA therapy Turning to the immunosuppressed population, among the 69 patients with resolved HBV infection who developed HBV reactivation and constituted the final study population (Fig. 1), 29 patients were started on NA therapy immediately at the time of HBV DNA detection, whereas 38 patients were initially managed without antiviral treatment; the remaining 2 patients died before follow-up assessment due to progression of their underlying diseases and were excluded from subsequent analyses. Fig. 1Flowchart of patient selection. A total of 77 patients with resolved HBV infection who experienced HBV reactivation between January 2014 and June 2025 were initially identified at Kyoto University Hospital and Kurashiki Central Hospital. After excluding 8 cases with HBV reactivation following living donor liver transplantation, 69 patients were included in the final analysis. Among them, 29 patients received NA therapy immediately after HBV reactivation, 2 patients died before follow-up assessment, and 38 patients were monitored without NA therapy, allowing observation of the natural course of HBV DNA levels. NA: nucleos(t)ide analogue. Flowchart of patient selection. A total of 77 patients with resolved HBV infection who experienced HBV reactivation between January 2014 and June 2025 were initially identified at Kyoto University Hospital and Kurashiki Central Hospital. After excluding 8 cases with HBV reactivation following living donor liver transplantation, 69 patients were included in the final analysis. Among them, 29 patients received NA therapy immediately after HBV reactivation, 2 patients died before follow-up assessment, and 38 patients were monitored without NA therapy, allowing observation of the natural course of HBV DNA levels. NA: nucleos(t)ide analogue. The latter group of 38 untreated patients (hereafter referred to as the observation group) provided a unique opportunity to characterize the natural course of HBV reactivation under immunosuppression without immediate antiviral intervention. Patients who received NA therapy at reactivation were analyzed as the immediate NA group. In the observation group (n = 38), the median age at HBV reactivation was 74 years (range, 46–86), compared with 71 years (56–88) in the immediate NA group (p = 0.20) (Table 1). Males accounted for 52.6% and 48.3% of patients in the observation and immediate NA groups, respectively (p = 0.81). The distribution of underlying diseases differed significantly between the two groups (p < 0.05). Hematologic malignancies were more frequent in the immediate NA group (44.7% vs. 69.0%), whereas rheumatic diseases were more frequent in the observation group (28.9% vs. 6.9%). Table 1Baseline characteristics and outcomes of HBV reactivation in patients with resolved infection, stratified by initial management after HBV DNA detection (observation vs. immediate NA).CharacteristicObservation group(no immediate NA, n = 38)Immediate NA group (n = 29)p valueAge, years74 [46–86]71 [56–88]0.20Sex0.81Male20 (52.6%)14 (48.3%)Female18 (48.4%)15 (51.7%)Underlying disease< 0.05*Hematologic malignancy17 (44.7%)20 (69.0%)Solid tumor8 (21.1%)3 (10.3%)Rheumatic disease11 (28.9%)2 (6.9%)Rheumatic disease with solid tumor1 (2.6%)0 (0.0%)Other disease1 (2.6%)4 (13.8%)Chemotherapy/immunosuppressive treatment0.095Chemotherapy with B-cell depletion therapy9 (23.7%)13 (44.8%)Chemotherapy without B-cell depletion therapy9 (23.7%)6 (20.7%)HSCT6 (15.8%)5 (17.2%)DMARDs9 (23.7%)0 (0.0%)Non-biologic immunomodulators4 (10.5%)4 (13.8%)Prednisolone alone1 (2.6%)1 (3.4%)Use of corticosteroids12 (31.6%)12 (41.4%)0.45Laboratory data at initiation of chemotherapy/immunosuppressive therapyALT, U/L16 [7–422]17 [7–383]0.69Anti-HBs antibody positive27 (71.1%)22 (75.9%)0.69Anti-HBc antibody positive34 (89.5%)27 (93.1%)0.79Anti-HBe antibody positive2 (18.2%)0 (0.0%)0.48HBV DNA, Log IU/mL1.4 [1.0–6.0]1.6 [1.3–8.1]< 0.01*HBV DNA < 1.3 log IU/mL, n (%)16 (42.1%)0 (0.0%)< 0.001*HBV DNA monitoring interval ≤ 91 days before reactivation25/38 (65.8%)23/29 (79.3%)0.35Outcomes related to HBV reactivationHepatitis due to HBV reactivation2 (5.3%)1 (3.4%)1.00Death due to HBV reactivation0 (0.0%)0 (0.0%)1.00Data are expressed as number (%) or median [range]. * Indicates a statistically significant difference.ALT, alanine aminotransferase; DMARDs, disease-modifying antirheumatic drugs; HBc, hepatitis B core; HBe, hepatitis B e; HBs, hepatitis B surface; HSCT, hematopoietic stem cell transplantation; NA, nucleos(t)ide analogue. Table 1Baseline characteristics and outcomes of HBV reactivation in patients with resolved infection, stratified by initial management after HBV DNA detection (observation vs. immediate NA).CharacteristicObservation group(no immediate NA, n = 38)Immediate NA group (n = 29)p valueAge, years74 [46–86]71 [56–88]0.20Sex0.81Male20 (52.6%)14 (48.3%)Female18 (48.4%)15 (51.7%)Underlying disease< 0.05*Hematologic malignancy17 (44.7%)20 (69.0%)Solid tumor8 (21.1%)3 (10.3%)Rheumatic disease11 (28.9%)2 (6.9%)Rheumatic disease with solid tumor1 (2.6%)0 (0.0%)Other disease1 (2.6%)4 (13.8%)Chemotherapy/immunosuppressive treatment0.095Chemotherapy with B-cell depletion therapy9 (23.7%)13 (44.8%)Chemotherapy without B-cell depletion therapy9 (23.7%)6 (20.7%)HSCT6 (15.8%)5 (17.2%)DMARDs9 (23.7%)0 (0.0%)Non-biologic immunomodulators4 (10.5%)4 (13.8%)Prednisolone alone1 (2.6%)1 (3.4%)Use of corticosteroids12 (31.6%)12 (41.4%)0.45Laboratory data at initiation of chemotherapy/immunosuppressive therapyALT, U/L16 [7–422]17 [7–383]0.69Anti-HBs antibody positive27 (71.1%)22 (75.9%)0.69Anti-HBc antibody positive34 (89.5%)27 (93.1%)0.79Anti-HBe antibody positive2 (18.2%)0 (0.0%)0.48HBV DNA, Log IU/mL1.4 [1.0–6.0]1.6 [1.3–8.1]< 0.01*HBV DNA < 1.3 log IU/mL, n (%)16 (42.1%)0 (0.0%)< 0.001*HBV DNA monitoring interval ≤ 91 days before reactivation25/38 (65.8%)23/29 (79.3%)0.35Outcomes related to HBV reactivationHepatitis due to HBV reactivation2 (5.3%)1 (3.4%)1.00Death due to HBV reactivation0 (0.0%)0 (0.0%)1.00Data are expressed as number (%) or median [range]. * Indicates a statistically significant difference.ALT, alanine aminotransferase; DMARDs, disease-modifying antirheumatic drugs; HBc, hepatitis B core; HBe, hepatitis B e; HBs, hepatitis B surface; HSCT, hematopoietic stem cell transplantation; NA, nucleos(t)ide analogue. Baseline characteristics and outcomes of HBV reactivation in patients with resolved infection, stratified by initial management after HBV DNA detection (observation vs. immediate NA). Data are expressed as number (%) or median [range]. * Indicates a statistically significant difference. ALT, alanine aminotransferase; DMARDs, disease-modifying antirheumatic drugs; HBc, hepatitis B core; HBe, hepatitis B e; HBs, hepatitis B surface; HSCT, hematopoietic stem cell transplantation; NA, nucleos(t)ide analogue. The distribution of chemotherapy and immunosuppressive regimens did not differ significantly between groups (p = 0.095). However, patients in the immediate NA group more frequently received B-cell depletion therapy (44.8% vs. 23.7% in the observation group), while disease-modifying antirheumatic drugs (DMARDs) were used exclusively in the observation group (23.7% vs. 0%). Liver enzyme levels at initiation of chemotherapy/immunosuppressive therapy were generally within normal to mildly elevated ranges; the median ALT was 16 U/L (range, 7–422) in the observation group and 17 U/L (7–383) in the immediate NA group (p = 0.69). Only one patient in the observation group exhibited marked ALT elevation, which was attributable to graft-versus-host disease following hematopoietic stem cell transplantation; all remaining patients had ALT levels ≤ 32 U/L. The proportion of patients who underwent HBV DNA monitoring at intervals of ≤ 91 days prior to the first detection of reactivation was similar between groups (65.8% vs. 79.3%, p = 0.35). With respect to virological profiles, all patients were HBsAg-negative at the time of initiation of immunosuppressive or chemotherapeutic therapy. Anti-HBs positivity (71.1% vs. 75.9%, p = 0.79) and anti-HBc positivity (89.5% vs. 93.1%, p = 0.68) were comparable between the observation and immediate NA groups. Notably, the median HBV DNA level at initial detection was significantly lower in the observation group than in the immediate NA group (1.4 [1.0–6.0] vs. 1.6 [1.3–8.1] log IU/mL, p < 0.01). Furthermore, a substantial proportion of patients in the observation group (42.1%) exhibited very low-level viremia (< 1.3 log IU/mL), a level corresponding to the guideline-recommended threshold for NA initiation in Japan, whereas no patients in the immediate NA group fell below this threshold (p < 0.001). During follow-up, clinically overt hepatitis attributable to HBV reactivation (ALT ≥ 100 U/L) occurred in 2 of 38 patients (5.3%) in the observation group and 1 of 29 patients (3.4%) in the immediate NA group (p = 1.00). No HBV-related deaths occurred in either group (p = 1.00). Overall, patients in the immediate NA group were characterized by a higher prevalence of hematologic malignancies, more frequent use of B-cell depletion therapy, and higher HBV DNA levels at reactivation, whereas the observation group more often included patients with rheumatic diseases and low-level HBV DNA reappearance. These findings suggest that, in routine clinical practice, immediate antiviral therapy tended to be selected for patients perceived to be at higher virological or immunosuppressive risk. Natural history of HBV reactivation under immunosuppression In this observation cohort (n = 38), we next examined the longitudinal dynamics of HBV DNA (Fig. 2). The reasons for continuing follow-up without initiating NA therapy included the clinician’s decision to opt for observation due to low initial HBV DNA levels and cases in which HBV DNA PCR results were not available on the day of testing, preventing immediate recognition of reactivation. The median interval between the first and second HBV DNA measurements after reactivation was 42 days (range: 6–125 days). In 18 of 38 cases, the virus became undetectable during follow-up after the initial reactivation. The median time to HBV DNA undetectability was 65.5 days (range, 8–196 days). Of these 18 cases, 13 were undetectable at the first follow-up HBV DNA measurement, whereas the remaining 5 showed a transient increase at the first follow-up measurement (median + 0.4 log IU/mL [+ 0.2–+1.0 log IU/mL]) but became undetectable at the subsequent measurement. Notably, approximately half of the patients demonstrated spontaneous disappearance of HBV DNA during follow-up. These findings suggest that at least some cases previously diagnosed as HBV reactivation may represent transient viremia rather than true reactivation induced by chemotherapy or immunosuppressive therapy. Consistent with this notion, we also observed a case of transient HBV DNA reappearance in an individual with resolved HBV infection who had not received immunosuppressive therapy (Figure S2). Fig. 2Natural course of HBV DNA levels in untreated patients with resolved HBV infection after HBV reactivation (n = 38). Longitudinal HBV DNA measurements were plotted over time until HBV DNA became undetectable, NA therapy was initiated, or follow-up was ended. The horizontal dashed line indicates the lower limit of HBV DNA detectability (1.0 log IU/mL). NA: nucleos(t)ide analogue. Natural course of HBV DNA levels in untreated patients with resolved HBV infection after HBV reactivation (n = 38). Longitudinal HBV DNA measurements were plotted over time until HBV DNA became undetectable, NA therapy was initiated, or follow-up was ended. The horizontal dashed line indicates the lower limit of HBV DNA detectability (1.0 log IU/mL). NA: nucleos(t)ide analogue. Risk stratification and definition of True Reactivation Based on these observations, “True Reactivation” in the present study was defined as HBV DNA ≥ 3.3 log IU/mL, either at the initial detection of reactivation or during subsequent monitoring in patients with resolved HBV infection. Among the 38 monitored patients followed without NA treatment after HBV reactivation, 7 fulfilled this definition (hereafter referred to as the True Reactivation group), whereas 18 patients achieved spontaneous HBV DNA clearance (hereafter referred to as the Spontaneous Clearance group). The remaining 13 cases were not assigned to either group because NA therapy was initiated before achieving spontaneous remission and before reaching 3.3 log IU/mL. When analysis was restricted to the 25 patients whose natural course could be clearly classified without early NA intervention, 18 patients (72%) experienced spontaneous HBV DNA clearance, whereas 7 patients (28%) progressed to true reactivation. The viral load trends of the True Reactivation group (n = 7) and the Spontaneous Clearance group (n = 18) are presented in Fig. 3. At the time of initial detection of HBV reactivation, the median HBV DNA level was significantly higher in the True Reactivation group than in the Spontaneous Clearance group (2.9 log IU/mL [range, 1.4–6.0] vs 1.2 log IU/mL [range, 1.0–1.9]) (p < 0.01) (Table 2). The interval between initiation of chemotherapy or immunosuppressive therapy and the initial detection of HBV DNA did not differ significantly between the True Reactivation group and the Spontaneous Clearance group (median 20 weeks [range, 6–112] vs 7.5 weeks [range, 2–111], respectively; Mann–Whitney U test, p = 0.28), although the median interval tended to be longer in the True Reactivation group. At the subsequent HBV DNA measurement, performed after a median interval of 62 days (range, 29–125) in the True Reactivation group and 42 days (range, 8–119) in the Spontaneous Clearance group, the median HBV DNA levels were 6.6 log IU/mL (range, 4.0–8.3) and 0 log IU/mL (range, 0–2.1), respectively. Correspondingly, the median change in HBV DNA per 30 days was + 1.2 log IU/mL (range, + 0.62 to + 1.9) in the True Reactivation group and − 0.74 log IU/mL (range, − 5.3 to + 0.86) in the Spontaneous Clearance group. A characteristic feature of the 7 cases in the True Reactivation group is that all showed a monotonous increase in viral load prior to the initiation of NA therapy. Detailed patient-level information on underlying diseases and immunosuppressive or chemotherapeutic regimens, together with guideline-based risk classifications according to the 2025 AGA Clinical Practice Guideline, in the True Reactivation and Spontaneous Clearance groups is provided in Supplementary Table S29. When comparing the clinical characteristics between the two groups, hematologic malignancies were present in 6 of 7 patients (85.7%) in the True Reactivation group, compared with 6 of 18 patients (33.3%) in the Spontaneous Clearance group (p < 0.05) (Table 2). Furthermore, patients with true reactivation were significantly younger (66 [46–78] vs. 76 [51–86] years, p < 0.05), and a higher proportion of these patients had undergone hematopoietic stem cell transplantation (HSCT). Fig. 3Differential viral load dynamics in patients with resolved HBV infection: True Reactivation vs. Spontaneous Clearance. The figure illustrates representative HBV DNA trends in patients classified as the True Reactivation group (n = 7) and the Spontaneous Clearance group (n = 18). True Reactivation was defined as HBV DNA ≥ 3.3 log IU/mL at the initial detection or thereafter in patients who had not received NA therapy. All seven patients in this group exhibited a progressive, monotonic increase in viral load prior to the initiation of antiviral treatment. In contrast, patients in the Spontaneous Clearance group demonstrated spontaneous resolution of HBV DNA without reaching the threshold of 3.3 log IU/mL. Dashed lines represent individual cases in the True Reactivation group, whereas solid lines represent cases in the Spontaneous Clearance group. The horizontal dashed line indicates the lower limit of HBV DNA detectability (1.0 log IU/mL), and the red solid line represents the viral load threshold for True Reactivation (3.3 log IU/mL). NA: nucleos(t)ide analogue. Differential viral load dynamics in patients with resolved HBV infection: True Reactivation vs. Spontaneous Clearance. The figure illustrates representative HBV DNA trends in patients classified as the True Reactivation group (n = 7) and the Spontaneous Clearance group (n = 18). True Reactivation was defined as HBV DNA ≥ 3.3 log IU/mL at the initial detection or thereafter in patients who had not received NA therapy. All seven patients in this group exhibited a progressive, monotonic increase in viral load prior to the initiation of antiviral treatment. In contrast, patients in the Spontaneous Clearance group demonstrated spontaneous resolution of HBV DNA without reaching the threshold of 3.3 log IU/mL. Dashed lines represent individual cases in the True Reactivation group, whereas solid lines represent cases in the Spontaneous Clearance group. The horizontal dashed line indicates the lower limit of HBV DNA detectability (1.0 log IU/mL), and the red solid line represents the viral load threshold for True Reactivation (3.3 log IU/mL). NA: nucleos(t)ide analogue. Table 2Comparison of baseline characteristics between True Reactivation group and Spontaneous Clearance group.CharacteristicTrue Reactivation(n = 7)Spontaneous Clearance(n = 18)p valueAge, years66 [46–78]76 [51–86]< 0.05*Sex0.41Male5 (71.4%)9 (50.0%)Female2 (28.6%)9 (50.0%)Underlying diseaseHematologic malignancy6 (85.7%)6 (33.3%)< 0.05*Solid tumor0 (0%)6 (33.3%)0.14Rheumatic disease1 (14.3%)7 (38.9%)0.36Chemotherapy/immunosuppressive treatment< 0.05*Chemotherapy with rituximab2 (28.6%)4 (22.2%)Chemotherapy without rituximab0 (0%)6 (33.3%)HSCT4 (57.1%)1 (5.6%)DMARDs1 (14.3%)6 (33.3%)Non-biologic immunomodulators0 (0%)1 (5.6%)Use of corticosteroids2 (28.6%)6 (33.3%)1.00Laboratory data at initiation detection of HBV-DNAALT, U/L16 [9–28]14 [7–30]0.75Anti-HBs antibody positive5 (71.4%)13 (72.2%)1.00Anti-HBc antibody positive6 (85.7%)17 (94.4%)0.49Anti-HBe antibody positive0 (0%)1 (33.3%)1.00HBV-DNA, Log IU/mL2.9 [1.4–6.0]1.2 [1.0–1.9]< 0.01*Data are expressed as number (%) or median [range]. * Indicates a statistically significant difference.ALT, alanine aminotransferase; DMARDs, disease-modifying antirheumatic drugs; HBc, hepatitis B core; HBe, hepatitis B e; HBs, hepatitis B surface; HSCT, hematopoietic stem cell transplantation. Table 2Comparison of baseline characteristics between True Reactivation group and Spontaneous Clearance group.CharacteristicTrue Reactivation(n = 7)Spontaneous Clearance(n = 18)p valueAge, years66 [46–78]76 [51–86]< 0.05*Sex0.41Male5 (71.4%)9 (50.0%)Female2 (28.6%)9 (50.0%)Underlying diseaseHematologic malignancy6 (85.7%)6 (33.3%)< 0.05*Solid tumor0 (0%)6 (33.3%)0.14Rheumatic disease1 (14.3%)7 (38.9%)0.36Chemotherapy/immunosuppressive treatment< 0.05*Chemotherapy with rituximab2 (28.6%)4 (22.2%)Chemotherapy without rituximab0 (0%)6 (33.3%)HSCT4 (57.1%)1 (5.6%)DMARDs1 (14.3%)6 (33.3%)Non-biologic immunomodulators0 (0%)1 (5.6%)Use of corticosteroids2 (28.6%)6 (33.3%)1.00Laboratory data at initiation detection of HBV-DNAALT, U/L16 [9–28]14 [7–30]0.75Anti-HBs antibody positive5 (71.4%)13 (72.2%)1.00Anti-HBc antibody positive6 (85.7%)17 (94.4%)0.49Anti-HBe antibody positive0 (0%)1 (33.3%)1.00HBV-DNA, Log IU/mL2.9 [1.4–6.0]1.2 [1.0–1.9]< 0.01*Data are expressed as number (%) or median [range]. * Indicates a statistically significant difference.ALT, alanine aminotransferase; DMARDs, disease-modifying antirheumatic drugs; HBc, hepatitis B core; HBe, hepatitis B e; HBs, hepatitis B surface; HSCT, hematopoietic stem cell transplantation. Comparison of baseline characteristics between True Reactivation group and Spontaneous Clearance group. Data are expressed as number (%) or median [range]. * Indicates a statistically significant difference. ALT, alanine aminotransferase; DMARDs, disease-modifying antirheumatic drugs; HBc, hepatitis B core; HBe, hepatitis B e; HBs, hepatitis B surface; HSCT, hematopoietic stem cell transplantation. Importantly, the patterns associated with true reactivation summarized in Table 3 can be described as follows: (i) when the initial HBV DNA level was ≥ 2.0 log IU/mL, all patients in this cohort progressed to true reactivation, and (ii) in patients with underlying hematologic malignancies, true reactivation could occur even when the initial HBV DNA level was < 2.0 log IU/mL. Indeed, when stratified by underlying disease, 6 of 12 patients (50.0%) with hematologic malignancies experienced true reactivation, compared to only 1 of 13 patients (7.7%) with other diseases. Moreover, initial HBV DNA levels were strongly predictive of risk: while only 2 of 20 patients (10.0%) with viral loads < 2.0 log IU/mL developed true reactivation, all 5 patients (100%) with levels ≥ 2.0 log IU/mL progressed to true reactivation. These results suggest that in cases of HBV reactivation due to resolved HBV infection, in the absence of underlying hematologic malignancy and with initial HBV DNA levels below 2.0 log IU/mL, detectable viremia often resolved spontaneously without progression to true reactivation in this cohort. Under the current JSH guidelines, HBV DNA monitoring would be recommended in 16 of 38 patients (42.1%) whose HBV DNA levels remained below 1.3 log IU/mL. In contrast, when applying our proposed criteria (initial HBV DNA level and hematologic malignancy), monitoring would extend to an additional nine cases, encompassing 25 of 38 patients (65.8%). This broader application highlights the potential to manage a larger proportion of ‘reactivation episodes’ that ultimately resolve spontaneously without requiring antiviral therapy. These findings suggest that, in selected patients, careful monitoring rather than immediate initiation of antiviral therapy may allow many cases of HBV DNA reappearance to resolve spontaneously without unnecessary treatment. Table 3Risk patterns of true reactivation stratified by underlying disease and initial HBV DNA level at reactivation. Underlying disease Hematologic malignancy Other diseases Total Initial viral load<2.0 log IU/mL 2/8 (25.0%) 0/12 (0%) 2/20 (10%)≥2.0 log IU/mL 4/4 (100%) 1/1 (100%) 5/5 (100%) Total 6/12 (50.0%) 1/13 (7.7%) 7/25 (28%) Table 3Risk patterns of true reactivation stratified by underlying disease and initial HBV DNA level at reactivation. Underlying disease Hematologic malignancy Other diseases Total Initial viral load<2.0 log IU/mL 2/8 (25.0%) 0/12 (0%) 2/20 (10%)≥2.0 log IU/mL 4/4 (100%) 1/1 (100%) 5/5 (100%) Total 6/12 (50.0%) 1/13 (7.7%) 7/25 (28%) Risk patterns of true reactivation stratified by underlying disease and initial HBV DNA level at reactivation. Discussion Despite advances in vaccines and antiviral therapies, HBV remains a global health challenge. As immunomodulatory therapies are increasingly used, HBV reactivation continues to pose a clinical concern, especially among individuals previously exposed to the virus2. Within this context, our findings challenge current management strategies, questioning both international guidelines—which recommend prophylactic antiviral therapy for high- to moderate-risk patients even without detectable HBV DNA—and the Japanese guideline, which initiates therapy at an HBV DNA level ≥ 1.3 log IU/mL7–11. In our cohort, only 7 of 25 patients (28%) progressed to “true reactivation” (≥ 3.3 log IU/mL), while 18 of 25 patients (72%) achieved spontaneous clearance without treatment, suggesting that many low-level DNA detections reflect transient viral activity rather than progressive reactivation. A key strength of the present study lies in the uniquely Japanese, monitoring-centered clinical framework in which it was conducted. In contrast to many international guideline settings, where prophylactic NA therapy is often initiated before HBV DNA becomes detectable, the Japanese guideline permits careful longitudinal monitoring of HBV DNA kinetics below predefined treatment thresholds. This strategy allowed us to directly observe the natural course of untreated low-level viremia under immunosuppression and to distinguish transient viral reappearance from clinically significant HBV reactivation. Such an evaluation would be difficult in settings where antiviral therapy is routinely initiated at or before the first detection of HBV DNA. If confirmed in larger cohorts, these data imply that a portion of clinically defined “HBV reactivation” may represent transient viremia. This redefinition has important implications: (i) the operational criteria for HBV reactivation, which treat any detectable HBV DNA as clinically significant, should be reconsidered; and (ii) while numerous drugs have been implicated as triggers for HBV reactivation, the true reactivation risk attributable to individual drugs may require reassessment. This is because some events previously classified as reactivation may have been self-limiting viral fluctuations rather than true reactivation. Low-level HBV DNA may result from intermittent cccDNA transcription, immune variation, or assay sensitivity rather than active replication requiring therapy. Thus, strict adherence to current thresholds— such as initiating treatment upon any HBV DNA detection or when levels reach 1.3 log IU/mL, as recommended by several international guidelines—may lead to unnecessary therapy in patients who would otherwise clear the virus spontaneously. In the inactive carrier phase, HBV DNA levels typically remain below approximately 3.3 log IU/mL and ALT levels are persistently normal, reflecting minimal ongoing hepatic inflammation7. Furthermore, ALT elevation typically occurs after viral amplification during HBV reactivation, indicating that HBV DNA kinetics represent an earlier marker of progressive viral activity12. Together, these observations support the concept that HBV DNA levels below approximately 3.3 log IU/mL may reflect transient low-level viral fluctuation in the setting of inactive chronic HBV infection, whereas higher levels may be more suggestive of progressive viral amplification. Our stratified analysis identified two factors associated with clinically relevant reactivation: hematologic malignancy and initial HBV DNA ≥ 2.0 log IU/mL. Half of patients with hematologic malignancy developed true reactivation, compared with only 7.7% of others; all patients with an initial HBV DNA level ≥ 2.0 log IU/mL progressed, while only 10% with lower levels did. Although these parameters may help discriminate patients who require early NA therapy from those suitable for monitoring, these findings should be interpreted cautiously given the limited sample size and the lack of additional virological markers such as quantitative HBsAg, ALT kinetics, or HBeAg status. Further validation in larger prospective cohorts is warranted. A possible explanation for why true reactivation cases are typically detected at relatively high viral loads is as follows. Transient viremia tends to remain below 2.0 log IU/mL, whereas true reactivation involves viral amplification, rapidly exceeding this threshold from initially undetectable levels. Therefore, the detection window for true reactivation at < 2.0 log IU/mL is narrow, and capture at low levels is likely less frequent compared to detection in transient viremia. The higher frequency of true reactivation among patients with hematologic malignancies may also be explained by the nature of immunosuppressive therapies used in this population. As shown in Table S2, many patients with hematologic malignancies received B-cell-depleting agents or underwent HSCT, both of which are recognized as high-risk factors for HBV reactivation. Therefore, the predominance of hematologic malignancies in the True Reactivation group is biologically plausible and consistent with previous reports2,9,12,13. Current JSH guideline recommends NA once HBV DNA exceeds 1.3 log IU/mL regardless of disease type11. However, our data indicate that patients with non-hematologic diseases and initial HBV DNA < 2.0 log IU/mL rarely progress and can be safely observed. Consistent with this, Fukuda et al. reported that reactivation in rheumatoid arthritis patients was uncommon (1.65/100 person-years) and often self-limited14. Conversely, patients with hematologic malignancies or higher viral loads remain high-risk and warrant early prophylaxis, aligning with prior evidence linking B-cell depletion and HSCT to elevated reactivation risk12,13,15. Beyond clinical outcomes, overtreatment may increase the risk of treatment-related adverse events, including renal dysfunction and bone mineral loss (particularly with tenofovir disoproxil fumarate), as well as treatment burden and healthcare costs16,17. Although antiviral resistance is uncommon with currently recommended first-line agents such as entecavir and tenofovir, it is not completely eliminated, particularly during long-term therapy. The Japanese monitoring-centered strategy has effectively prevented severe reactivation while minimizing such risks11,18–20. Fujita et al. also demonstrated that risk-tailored management for B-cell lymphoma reduced healthcare costs, reinforcing the value of HBV DNA monitoring over universal prophylaxis21. Incorporating both viral load and underlying disease type requiring immunosuppressive therapy or chemotherapy may further optimize this cost-effective framework. Limitations include the retrospective design of this study, the small number of true reactivation cases, and variable testing intervals determined by clinicians. In addition, because initiation of NA therapy was determined at the discretion of the treating physicians, patients considered to be at higher clinical risk—such as those with higher HBV DNA levels or hematologic malignancies—were more likely to receive immediate treatment. Consequently, the observation group represents a selected subset of patients with relatively lower-risk features and may not fully reflect the entire spectrum of HBV reactivation under immunosuppression. This selection process may also have resulted in underestimation of the true incidence of progressive reactivation in the overall population. Furthermore, because multivariable adjustment was not feasible in this small cohort, the independent contribution of age to reactivation risk could not be determined. In addition, as this study included patients with resolved HBV infection defined serologically by HBsAg negativity with anti-HBc positivity or anti-HBs positivity, the exact duration since viral resolution and the distinction between prior acute self-limited infection and resolved chronic infection could not be determined. Quantitative HBsAg levels were also not consistently available in this retrospective cohort and therefore could not be incorporated into the present analysis. However, previous studies have demonstrated that increases in HBV DNA typically precede both HBsAg seroreversion and ALT elevation during HBV reactivation, supporting the use of HBV DNA kinetics as an early indicator of progressive viral activity12. Finally, in the inactive HBsAg-positive reference cohort, HBV DNA measurements were obtained at discrete intervals during routine clinical follow-up; therefore, the precise timing of viral reappearance could not be determined, and the apparent clustering of reappearance events around approximately 6 months—particularly in the analysis shown in Figure S1E—may partly reflect the monitoring schedule rather than the exact biological timing of viral re-detection. Prospective studies with predefined treatment thresholds and standardized monitoring schedules will be necessary to further validate the generalizability of these findings and refine treatment criteria. In conclusion, combining the initial viral load at the timing of reactivation with the underlying disease type may help support a more individualized approach to HBV reactivation management. In our cohort, patients with non-hematologic conditions and an initial load below 2.0 log IU/mL were unlikely to progress and could be considered for careful monitoring, whereas those with hematologic malignancies or ≥ 2.0 log IU/mL at reactivation appeared more likely to require early antiviral therapy. Although these findings require validation in larger prospective studies, this stratified approach may help reduce overtreatment, limit adverse effects and healthcare costs, and contribute to a more personalized, evidence-based framework for HBV reactivation management in the field of Oncohepatology. Methods Study design and overall population This was a retrospective cohort study conducted at Kyoto University Hospital and Kurashiki Central Hospital. The primary objective was to investigate the clinical course and risk factors of HBV reactivation in patients with resolved HBV infection undergoing immunosuppressive or chemotherapeutic therapy. In addition, HBsAg-positive reference cohort was analyzed to assess longitudinal HBV DNA variability in the absence of antiviral treatment and immunosuppression. Reference cohort of untreated HBsAg-positive patients To contextualize the HBV DNA kinetics observed in the reactivation cohort, we additionally analyzed a reference cohort of untreated HBsAg-positive patients in whom HBV DNA had once fallen below the detection limit. This analysis was intended to provide contextual information on the magnitude of HBV DNA fluctuation in the absence of immunosuppression and was not designed as a direct comparator to the resolved infection cohort. To characterize longitudinal HBV DNA variability in the natural course, we retrospectively reviewed HBsAg-positive patients with chronic HBV infection followed at Kyoto University Hospital over the past 10 years. The cohort included 161 patients with no history of antiviral or immunosuppressive therapy. The median observation period for these 161 patients was 4.5 years (range, 0.7–13.3 years). Because the aim of this reference cohort analysis was to evaluate HBV DNA reappearance after an undetectable state, we restricted the analysis to patients who achieved undetectable HBV DNA at least once during follow-up. Accordingly, 71 patients met this criterion and were included in the fluctuation analysis, whereas the remaining 90 patients did not have a documented undetectable HBV DNA result during follow-up and were not included in this specific analysis. The median observation period for the 71 patients was 4.8 years (range, 2.4–13.3 years). HBV DNA measurements in this cohort were performed as part of routine clinical follow-up at approximately 6-month intervals. At the time when HBV DNA first became undetectable during the observation period, we recorded patient age, serum alanine aminotransferase (ALT) levels, and platelet counts to characterize the clinical background. In addition, HBe antigen status and HBsAg positivity were also assessed to further define the virological profile of the cohort. Serial HBV DNA levels were extracted, and patterns of undetectable status and reappearance were classified as “suppression” (sustained undetectable status) and “intermittent viremia” (reappearance after undetectable status). HBV DNA reappearance events were defined as transitions from a previously documented undetectable measurement to detectable levels (≥ 1.0 log IU/mL), with the undetectable status confirmed based on the immediately preceding measurement. The frequency and magnitude of HBV DNA reappearance were analyzed, focusing on episodes reaching ≥ 1.0–1.3 log IU/mL (corresponding to the lower limit of detection of the assays used during the study period) or ≥ 3.3 log IU/mL (2,000 IU/mL), the latter commonly regarded as the threshold defining the inactive phase of chronic HBV infection. HBV reactivation cohort (main analysis) For this study, HBV reactivation from resolved infection was defined as the reappearance of HBV DNA at levels ≥ 1.0 log IU/mL in patients with previously undetectable HBV DNA, or the seroreversion of HBsAg. At Kyoto University Hospital, we reviewed 77,712 HBV DNA measurements obtained between January 2014 and June 2025. Cases were screened for transitions from undetectable HBV DNA to ≥ 1.0 log IU/mL. Eligible cases were defined as patients with resolved HBV infection who had received chemotherapy and/or immunosuppressive therapy, based on medical records. Following this process, 65 cases of HBV reactivation were identified. At Kurashiki Central Hospital, a total of 56,468 HBV DNA measurements were performed between January 2017 and March 2025 as part of routine clinical monitoring across all tested patients. During the same period, 9,482 patients received chemotherapy or immunosuppressive therapy. Among them, 145 patients had detectable HBV DNA during those treatments. Of these, 32 had been negative for HBV DNA at chemotherapy initiation, and 12 cases ultimately fulfilled the definition of HBV reactivation and were included in the analysis. In total, 77 cases of HBV reactivation were identified across the two institutions (65 from Kyoto University Hospital and 12 from Kurashiki Central Hospital). After excluding 8 cases following living donor liver transplantation, 69 patients were included in the present study. Clinical data including demographics, underlying diseases, type of regimen, laboratory parameters, and treatment courses were collected from medical records. HBV DNA measurement At both Kyoto University Hospital and Kurashiki Central Hospital, HBV DNA levels were measured using the COBAS TaqMan 48 test (Roche Diagnostics, Tokyo, Japan) from January 2014 to March 2020, with a lower limit of detection (LOD) of 20 IU/mL (1.3 log IU/mL). From March 2020 onward, HBV DNA was measured using the COBAS 8800 system (Roche Diagnostics), with an LOD of 10 IU/mL (1.0 log IU/mL). Until 2017, HBV DNA values were reported in log copies/mL, whereas from 2018 onward they were expressed in log IU/mL. For measurements originally reported in log copies/mL, values were converted to log IU/mL using the formula: log IU/mL = log copies/mL − 0.72. as described in previous studies22, to ensure comparability across the entire dataset. Management after reactivation At the time of HBV reactivation, the decision to initiate NA therapy (entecavir, tenofovir disoproxil fumarate, or tenofovir alafenamide) was made at the discretion of the treating physicians. Patients who were not started on NA therapy underwent serial HBV DNA measurements at intervals determined by clinical practice, typically every 4–8 weeks. For patients treated with NA therapy, the timing and choice of drug were also extracted from the records. Follow-up was censored at the time when HBV DNA became undetectable or when NA therapy was initiated. Analysis of the natural course in untreated patients Among the 69 patients diagnosed with HBV reactivation, 29 received immediate NA therapy and 2 died before follow-up assessment, leaving 38 patients who were monitored without NA therapy (Fig. 1). These 38 patients underwent at least one additional HBV DNA measurement after reactivation. For this subset, serial HBV DNA levels were obtained until HBV DNA became undetectable, NA therapy was initiated, or the patient was lost to follow-up. The intervals from the time of reactivation to subsequent HBV DNA measurements, as well as the median interval between the first and second measurements, were calculated. Statistical analysis Categorical variables were compared using the chi-square or Fisher’s exact test, as appropriate. Continuous variables were expressed as median (range) and compared using the Mann–Whitney U test. A p-value < 0.05 was considered statistically significant. Statistical analyses were performed using PRISM 10.2.2. (GraphPad, MA, USA). Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1