Published online Oct 27, 2026. doi: 10.4254/wjh.124709
Revised: July 27, 2026
Accepted: September 23, 2026
Published online: October 27, 2026
Processing time: 120 Days and 10.3 Hours
Liver transplantation is the gold standard for end-stage liver disease, but the shortage of donor organs results in extended waiting lists and considerable mor
To assess the impact of hepatitis delta infection on waiting-list outcomes through a competing-risks analysis.
We conducted an ambispective national cohort study of adults listed for liver transplantation (1 January 2013-31 December 2025). The retrospective phase covered 2013-2017 and the prospective phase 2018-2025. Two separate etiologic analyses were performed, comparing HDV infection first with hepatitis B virus (HBV) monoinfection and then with non-delta viral infection, comprising HBV monoinfection, HBV/hepatitis C virus (HCV) coinfection, and HCV monoinfection. Death on the waiting list and transplantation were treated as competing events. Cumulative incidence functions (CIF) were estimated using the Aalen-Johansen method. Multivariable cause-specific Cox and Fine-Gray subdistribution hazard models were fitted for death and for transplantation.
Of 313 adults, 174 (55.6%) had delta infection. The delta group had a more advanced baseline profile. The 5-year CIF of waiting-list death did not differ between the delta and non-delta viral groups (45.9% vs 42.9%; Gray’s test, P = 0.696), whereas transplantation was more frequent in the delta group (43.1% vs 21.7%; P = 0.001). In cause-specific models for the viral cohort, HDV approximately doubled the hazard of both waiting-list death [cause-specific hazard ratio (csHR) = 2.12, 95%CI: 1.38-3.24, P < 0.01] and transplantation [csHR = 2.71, 95%CI: 1.58-4.64, P < 0.01]. On the subdistribution scale for the viral cohort, HDV was not independently associated with waiting-list death [subdistribution hazard ratio (sHR) = 1.19, 95%CI: 0.84-1.68; P = 0.22] but was associated with transplanta
HDV increased the cause-specific hazard of waiting-list death, but this effect was attenuated by more frequent transplantation. HDV status may aid risk stratification in endemic settings, although current evidence is insu
Core Tip: This national competing-risks study shows that hepatitis D virus infection approximately doubled the cause-specific hazards of both waiting-list death and liver transplantation. However, the cumulative incidence of waiting-list death was not increased because patients with hepatitis D underwent transplantation more frequently. This divergence between cause-specific and subdistribution effects highlights how transplantation modifies the observed mortality risk. Hepatitis D status may therefore improve clinical risk stratification in endemic settings, although the current evidence does not support a formal model for end-stage liver disease exception.
- Citation: Peltec A, Buga D, Hotineanu A, Wedemeyer H. Impact of hepatitis delta virus infection on waiting list outcomes: A competing-risks analysis. World J Hepatol 2026; 18(10): 124709
- URL: https://www.wjgnet.com/1948-5182/full/v18/i10/124709.htm
- DOI: https://dx.doi.org/10.4254/wjh.124709
Liver transplantation is the gold standard treatment for end-stage liver disease. However, demand for donor organs has long exceeded supply, resulting in prolonged waiting times and considerable mortality before a graft becomes available[1,2]. Once decompensation occurs, median survival without transplantation falls below two years[3]. Allocation policy changed in 2002, when the model for end-stage liver disease (MELD) was adopted to estimate short-term mortality[4,5]. The sodium-adjusted MELD-Na and, more recently, MELD 3.0 have refined this score by incorporating hyponatremia, sex and albumin[6,7].
MELD-based allocation has reduced waiting-list mortality across many national programs, but its predictive accuracy is not uniform. Etiology, comorbidity and frailty modulate the relationship between score and outcome[8], leaving residual etiology-specific risk that the score does not capture. This is particularly relevant in hepatitis D virus (HDV)-related advanced chronic liver disease, in which rapid clinical deterioration and portal-hypertensive complications may occur despite comparable MELD values.
Hepatitis delta virus infection promotes faster progression of fibrosis, leads to earlier decompensation, and increases the risk of hepatocellular carcinoma about threefold compared with hepatitis B virus (HBV) monoinfection[9-12]. HDV prevalence among HBsAg-positive individuals is highest in Eastern Europe, the Mediterranean region, Central Asia and sub-Saharan Africa[13,14]. The Republic of Moldova has one of the highest age-standardized burdens of cirrhosis worldwide[15], and more than half of patients on the national liver transplant waiting list have HDV-related end-stage liver disease[16,17]. Whether HDV status provides independent prognostic information beyond MELD in predicting waiting-list outcomes remains insufficiently investigated.
On the waiting list, death and transplantation are competing events, as transplantation removes a patient from the risk of subsequent waitlist death. Transplantation is also an informative event, because transplanted patients tend to be the most severely ill[18]. In this setting, the Kaplan–Meier method may overestimate waiting-list mortality[19]. Therefore, competing-risks analysis is more appropriate, with cumulative incidence functions (CIF) compared using Gray’s test[20] and covariate effects assessed using cause-specific Cox and Fine-Gray subdistribution hazard models[18,21]. We applied this methodology to the complete national cohort of adults listed for liver transplantation between 2013 and 2025, to assess the prognostic role of HDV infection with respect to both waiting-list mortality and access to transplantation.
This was an ambispective national cohort study of adult patients with advanced chronic liver disease listed for liver transplantation between January 1, 2013 and December 31, 2025. Clinical assessment was performed at the Hepatology Department and Outpatient Clinic of the Republican Clinical Hospital “Timofei Mosneaga”, Chisinau, Republic of Moldova. The study was approved by the Research Ethics Committee of the Nicolae Testemitanu State University of Medicine and Pharmacy (No. 40). Written informed consent was obtained from prospectively enrolled patients, while the retrospective phase used anonymized archival data under institutional approval.
Eligible patients were aged 18 years or older at listing, had a diagnosis of liver cirrhosis based on standard clinical, biochemical and imaging criteria, were registered on the national waiting list between 1 January 2013 and 31 December 2025, and had complete data for calculation of MELD-Na at listing. Patients with documented human immunodeficiency virus co-infection, missing listing date or no follow-up were excluded. After exclusion of five pediatric records from the 318 initially identified records, the final analytical cohort comprised 313 adult patients (Figure 1). Patients were stratified by era of inclusion into an early retrospective period (2013–2017, n = 177) and a later prospective period (2018–2025, n = 136). Among the 266 patients with viral liver disease, two comparisons were performed. Patients with non-viral liver disease were excluded from these analyses to reduce etiological heterogeneity and to allow a more direct assessment of the effect of HDV infection within a viral-disease population. First, patients with HDV infection (n = 174) were compared with those with HBV monoinfection (n = 30) to provide an etiologically specific assessment. Because the HBV monoinfection group was relatively small, an additional broader comparison was performed between patients with HDV infection (n = 174) and those with non-delta viral infection (n = 92), including HBV monoinfection (n = 30), HBV/HCV coinfection (n = 10), and HCV monoinfection (n = 52).
Baseline variables collected at listing included demographics (age, sex, residence), body mass index (BMI), hematology, standard liver biochemistry (alanine and aspartate aminotransferases, gamma-glutamyltransferase, total bilirubin, albumin, international normalized ratio, sodium, creatinine), virological and immunological markers (HBsAg, anti-HDV, HDV RNA, HBV DNA, anti-HCV), prognostic scores (MELD, MELD-Na, MELD 3.0), and markers of decompensation (ascites, hepatic encephalopathy, bleeding from esophageal varices) and complications. All clinical, laboratory, and virological data were collected at the time of listing. The primary outcomes were waiting-list death and liver transplan
Continuous variables were summarized as mean ± SD or median (interquartile range) according to distribution (Shapiro-Wilk test) and compared using the student t test or the Mann-Whitney U test; categorical variables were compared using the χ2 or Fisher exact test. Baseline differences between groups were additionally assessed using standardized mean differences (SMDs) to quantify the magnitude of between-group differences independently of sample size and to comple
Waiting-list outcomes were analyzed within a competing-risks framework. Death and transplantation were the events of interest, each competing for the other. Patients who remained active on the waiting list without experiencing the event of interest were followed until their last available follow-up. In the cause-specific Cox analysis, occurrence of a competing event ended follow-up for the outcome being analyzed. Thus, transplantation ended follow-up when modeling waiting-list death, whereas death ended follow-up when modeling transplantation. Patients who remained alive and active on the waiting list at the end of the study contributed follow-up through December 31, 2025, without being classified as having experienced death or transplantation.
CIFs were estimated using the Aalen-Johansen estimator and compared between groups using Gray’s test. Group-specific estimates with 95%CIs were reported at 1 and 5 years. Because cause-specific and subdistribution hazard models address complementary questions, both approaches were applied to each outcome. Cause-specific Cox models assessed the association of HDV with the instantaneous rate of each event among patients remaining at risk, whereas Fine-Gray models assessed its association with cumulative incidence in the presence of competing events. Both models were adjusted for HDV status, age, sex, MELD-Na, and era. Bilirubin, international normalized ratio, creatinine, and sodium were not entered separately because they are components of the MELD-Na score.
The proportional-hazards assumption was assessed using scaled Schoenfeld residuals for the cause-specific Cox models and an HDV-by-time interaction for the Fine-Gray models. Robustness of the findings was evaluated in prespecified sensitivity analyses, including incremental covariate adjustment, restriction of the comparator group to HBV monoinfection or to non-delta viral disease, replacement of MELD-Na with MELD 3.0, and additional adjustment for hepatocellular carcinoma and antiviral therapy. Baseline data were nearly complete, with ≤ 0.6% missingness for any variable. Therefore, complete-case analysis was used. No patients were lost to follow-up. Statistical tests were two-sided, with P < 0.05 considered statistically significant. Analyses were performed using R version 4.5.2 and SPSS version 27.0.
Of the 313 adults, 174 (55.6%) had HDV infection, 92 (29.4%) had non-delta viral infection, of whom 30 (9.6%) had HBV monoinfection and the remaining 47 (15.0%) had non-viral cirrhosis. Compared with patients with HBV monoinfection, those with HDV infection had more pronounced manifestations of portal hypertension and hepatocellular injury at baseline. They were younger (48 years vs 52 years, P = 0.037), included a higher proportion of women (45% vs 20%, P = 0.019), and had a lower BMI (24.8 kg/m2 vs 27.1 kg/m2, P = 0.004). The HDV group also had lower platelet counts (median, 72 × 109/L vs 98 × 109/L, P = 0.005) and higher alanine aminotransferase (ALT) levels (58.1 U/L vs 43.3 U/L, P = 0.009) (Table 1).
| Characteristic | Delta (n = 174) | HBV monoinfection (n = 30) | SMD | P value |
| Age (years) | 48.0 (42.0-54.0) | 52.0 (47.0-57.8) | 0.42 | 0.037 |
| Female sex | 78 (45) | 6 (20) | 0.55 | 0.019 |
| Rural residence | 107 (61) | 14 (47) | 0.30 | 0.185 |
| BMI (kg/m2) | 24.8 (22.8-27.1) | 27.1 (24.7-30.4) | 0.54 | 0.004 |
| ALT (U/L) | 58.1 (39.3-93.1) | 43.3 (26.9-64.2) | 0.27 | 0.009 |
| AST (U/L) | 76.8 (55.5-116.6) | 61.1 (44.2-90.5) | 0.27 | 0.067 |
| Platelets (× 109/L) | 72.0 (49.0-103.5) | 98.0 (68.5-161.3) | 0.44 | 0.005 |
| Albumin (g/L) | 28.0 (24.9-31.9) | 29.4 (24.0-35.6) | 0.07 | 0.643 |
| Total bilirubin (μmol/L) | 40.6 (23.5-60.5) | 45.0 (21.8-79.1) | 0.23 | 0.477 |
| INR | 1.5 (1.3-1.7) | 1.4 (1.2-1.7) | 0.14 | 0.252 |
| Creatinine (μmol/L) | 66.8 (57.7-75.0) | 71.2 (57.1-79.2) | 0.21 | 0.327 |
| Sodium (mmol/L) | 138.0 (135.0-140.1) | 139.0 (136.0-140.8) | 0.09 | 0.466 |
| Ascites | 110 (63) | 15 (50) | 0.27 | 0.242 |
| Hepatic encephalopathy | 68 (39) | 10 (33) | 0.12 | 0.676 |
| Esophageal variceal hemorrhage | 35 (20) | 5 (17) | 0.09 | 0.660 |
| Hepatocellular carcinoma | 31 (18) | 8 (27) | 0.21 | 0.375 |
| MELD | 15.0 (12.0-17.0) | 15.5 (11.0-17.0) | 0.01 | 0.992 |
| MELD-Na | 15.0 (12.0-18.8) | 15.5 (11.0-18.0) | 0.07 | 0.510 |
| MELD 3.0 | 16.5 (13.0-20.0) | 16.5 (12.0-19.0) | 0.10 | 0.652 |
| Waiting-list time (days) | 247.5 (102.5-631) | 281 (135.2-762.2) | 0.13 | 0.593 |
| Waiting-list death | 78 (44.8) | 19 (63.3) | 0.38 | 0.094 |
| Transplantation | 72 (41.4) | 6 (20.0) | 0.48 | 0.043 |
| Delisted | 1 (0.6) | 0 (0.0) | 0.11 | 1.000 |
| Active in waiting list | 23 (13.2) | 5 (16.7) | 0.10 | 0.826 |
Given the small size of the HBV monoinfection group (n = 30), patients with HDV infection (n = 174) were additionally compared with the broader non-delta viral infection group (n = 92) (Table 2). A similar baseline pattern was observed. Patients with HDV infection were younger [48.0 (42.0–54.0) years vs 53.0 (48.0–58.0) years, P < 0.001], had a lower BMI [24.8 (22.8–27.1) kg/m2 vs 26.4 (24.5–28.1) kg/m2, P < 0.001], and were more frequently from rural areas [107/174 (61%) vs 38/92 (41%), P = 0.003]. They also had lower platelet counts [72.0 (49.0–103.5) × 109/L vs 97.8 (71.5–124.2) × 109/L, P < 0.001] and albumin levels [28.0 (24.9–31.9) g/L vs 30.8 (25.7–36.0) g/L, P = 0.004], together with higher ALT [58.1 (39.3–93.1) U/L vs 48.3 (27.0–77.6) U/L, P = 0.005] and aspartate aminotransferase [76.8 (55.5–116.6) U/L vs 65.2 (44.0–93.0) U/L, P = 0.034] levels. The frequencies of ascites, hepatic encephalopathy, esophageal variceal hemorrhage, and hepatocellular carcinoma were also comparable (Table 2). Overall, these findings indicate a less favorable baseline clinical and biochemical profile in patients with HDV infection, and crude outcome estimates were therefore interpreted alongside multivariable models adjusted for relevant clinical and demographic covariates.
| Characteristic | Delta group (n = 174) | Non-delta viral group (n = 92) | SMD | P value |
| Age (years) | 48.0 (42.0-54.0) | 53.0 (48.0-58.0) | 0.58 | < 0.001 |
| Female sex | 78 (45) | 30 (33) | 0.25 | 0.072 |
| Rural residence | 107 (61) | 38 (41) | 0.41 | 0.003 |
| BMI (kg/m2) | 24.8 (22.8-27.1) | 26.4 (24.5-28.1) | 0.40 | < 0.001 |
| ALT (U/L) | 58.1 (39.3-93.1) | 48.3 (27.0-77.6) | 0.19 | 0.005 |
| AST (U/L) | 76.8 (55.5-116.6) | 65.2 (44.0-93.0) | 0.23 | 0.034 |
| Platelets (×109/L) | 72.0 (49.0-103.5) | 97.8 (71.5-124.2) | 0.30 | < 0.001 |
| Albumin (g/L) | 28.0 (24.9-31.9) | 30.8 (25.7-36.0) | 0.34 | 0.004 |
| Total bilirubin (μmol/L) | 40.6 (23.5-60.5) | 34.2 (20.6-56.5) | 0.07 | 0.198 |
| INR | 1.5 (1.3-1.7) | 1.4 (1.2-1.6) | 0.18 | 0.063 |
| Creatinine (μmol/L) | 66.8 (57.7-75.0) | 69.3 (60.3-80.9) | 0.29 | 0.070 |
| Sodium (mmol/L) | 138.0 (135.0-140.1) | 139.1 (136.0-141.0) | 0.22 | 0.065 |
| Ascites | 110 (63) | 46 (50) | 0.27 | 0.051 |
| Hepatic encephalopathy | 68 (39) | 25 (27) | 0.26 | 0.067 |
| Esophageal variceal hemorrhage | 35 (20) | 25 (27) | 0.17 | 0.190 |
| Hepatocellular carcinoma | 31 (18) | 18 (20) | 0.04 | 0.854 |
| MELD | 15.0 (12.0-17.0) | 14.0 (10.0-17.0) | 0.15 | 0.249 |
| MELD-Na | 15.0 (12.0-18.8) | 15.0 (10.0-17.2) | 0.22 | 0.074 |
| MELD 3.0 | 16.5 (13.0-20.0) | 15.0 (11.0-18.2) | 0.29 | 0.029 |
| Waiting-list time (days) | 247.5 (102.5-631) | 601.5 (173.5-1936.5) | 0.60 | < 0.001 |
| Waiting-list death | 78 (44.8) | 42 (45.7) | 0.02 | 1.00 |
| Transplantation | 72 (41.4) | 20 (21.7) | 0.43 | 0.002 |
| Delisted | 1 (0.6) | 5 (5.4) | 0.29 | 0.035 |
| Active in waiting list | 23 (13.2) | 25 (27.2) | 0.35 | 0.008 |
Baseline MELD and MELD-Na did not differ significantly between the delta and HBV monoinfection groups or between the delta and non-delta viral groups. MELD 3.0 at listing was higher in the delta group than in the non-delta viral group (16.5 vs 15.0, P = 0.029).
Median waiting-list time did not differ significantly between patients with HDV infection and those with HBV mo
When patients with HDV infection were compared with those with non-delta viral infection, median waiting-list time was significantly shorter among patients with HDV infection than among those with non-delta viral infection [247.5 (102.5-631.0) days vs 601.5 (173.5-1936.5) days, P < 0.001]. Waiting-list mortality was comparable in the two groups (44.8% vs 45.7%, P = 1.000). Transplantation was more frequent among patients with HDV infection (41.4% vs 21.7%, P = 0.002), while a higher proportion of patients with non-delta viral infection remained active on the waiting list at the end of follow-up (27.2% vs 13.2%, P = 0.008) and delisting was more frequent in that group (5.4% vs 0.6%, P = 0.035) (Table 2).
In the overall cohort, the CIF for waiting-list death reached 19.5% at 1 year and 42.6% (95%CI: 37.3-48.0) at 5 years. For transplantation, the corresponding values were 25.4% at 1 year and 34.4% (95%CI: 29.1-39.8) at 5 years. During the first 12 months the two cumulative incidences were broadly comparable. From the second year onward, however, mortality progressively exceeded transplantation. This divergence may reflect a mismatch between the rate of clinical deterioration and the availability of timely transplantation during longer follow-up. These estimates are consistent with the 35%-45% 5-year waiting-list mortality reported by the European Liver Transplant Registry among patients with viral cirrhosis[22], supporting the external relevance of the present national cohort.
The 5-year cumulative incidence of waiting-list death was 45.9% (95%CI: 38.1-53.8) in the delta group and 42.9% (95%CI: 32.1-53.7) in the non-delta viral group, without a significant difference (Gray’s test, P = 0.696), whereas the cumulative incidence of transplantation was 43.1% (95%CI: 35.4-50.9) vs 21.7% (95%CI: 13.0-30.4) (P = 0.001) (Figure 2A and Table 3). In the comparison with HBV monoinfection group, the 5-year cumulative incidence of death was nume
| Group | Death 1 year | Death 5 years | Gray P | Transplantation 1 year | Transplantation 5 years | Gray P value |
| Delta group vs non-delta viral group | ||||||
| Delta group | 23.7 (17.1-30.2) | 45.9 (38.1-53.8) | 0.696 | 31.7 (24.6-38.8) | 43.1 (35.4-50.9) | 0.001 |
| Non-delta viral group | 15.7 (8.1-23.3) | 42.9 (32.1-53.7) | 19.2 (10.9-27.5) | 21.7 (13.0-30.4) | ||
| Delta group vs HBV monoinfection group | ||||||
| Delta group | 23.7 (17.1-30.2) | 45.9 (38.1-53.8) | 0.065 | 31.7 (24.6-38.8) | 43.1 (35.4-50.9) | 0.030 |
| HBV monoinfection group | 27.7 (10.9-44.5) | 66.4 (47.3-85.5) | 20.7 (5.5-35.9) | 20.7 (5.5-35.9) | ||
In the multivariable cause-specific Cox model for waiting-list death, HDV infection was independently associated with mortality [cause-specific hazard ratio (csHR) = 2.01, 95%CI: 1.40-2.89; P < 0.01], indicating a twofold higher cause-specific hazard of death after adjustment for age, sex, MELD-Na score, and era of inclusion (Figure 3 and Table 4). Female sex was associated with a lower cause-specific hazard of death (csHR = 0.53, 95%CI: 0.37-0.77), and MELD-Na predicted mortality (csHR = 1.07 per point, 95%CI: 1.03-1.11). In the corresponding Fine-Gray model, the subdistribution hazard of death associated with HDV was not significantly increased [subdistribution hazard ratio (sHR) = 1.20, 95%CI: 0.89-1.62], whereas MELD-Na remained associated with the cumulative incidence of death (sHR = 1.04 per point, 95%CI: 1.01-1.07) (Table 5). Although the contrast between the two hazard scales may appear paradoxical, it is consistent with the structure of competing-risks analysis. Because transplantation removes patients from the risk set for death, an elevated cause-specific hazard can coexist with an unchanged cumulative incidence when the competing event occurs more frequently in the same group. Age and era of inclusion were not independently associated with waiting-list death on either scale.
In the multivariable models for transplantation, HDV infection was independently associated with greater access to transplantation on both scales (csHR = 2.41, 95%CI: 1.56-3.73; sHR = 1.89, 95%CI: 1.25-2.86), indicating that HDV candidates reached transplantation both earlier and more frequently than non-delta candidates (Figure 3, Tables 4 and 5). MELD-Na predicted higher transplantation probability (csHR = 1.06 per point, 95%CI: 1.02-1.11), which is the expected behavior of a score designed around the principle of allocating organs to the sickest first. Female sex was not inde
When the comparator was restricted, the delta group remained more severely affected on granular markers of disease severity despite similar MELD scores (Table 6). The association between HDV infection and cause-specific waiting-list mortality was unchanged when the comparator was restricted to non-delta viral infection (csHR = 2.12, 95%CI: 1.38-3.24) and when MELD-Na was replaced by MELD 3.0 (csHR = 2.03, 95%CI: 1.41-2.92); the comparison with HBV monoinfection was limited by the small size of that subgroup (n = 30) and remained exploratory (Figure 4 and Table 6). The greater access to transplantation associated with HDV infection was likewise reproduced with the restricted comparators: vs viral non-delta cirrhosis, csHR = 2.71 (95%CI: 1.58-4.64) and sHR = 2.09 (95%CI: 1.20-3.64); vs HBV monoinfection, csHR = 2.18 (95%CI: 0.95-5.02) and sHR = 2.38 (95%CI: 1.02-5.56), the latter estimated with a reduced model because of the small subgroup size (Table 7). Estimates were stable across incremental adjustment (csHR: 1.88-2.11; sHR: 1.16-1.20; Table 8). In an extended model additionally including hepatocellular carcinoma, antiviral therapy, albumin and ascites, hepatocellular carcinoma was an independent predictor of waiting-list death (csHR = 2.38, 95%CI: 1.48-3.84) and the HDV effect persisted (csHR = 1.90, 95%CI: 1.30-2.77; Table 9).
| Model | Covariates added | sHR (95%CI) | csHR (95%CI) |
| M0 | HDV | 1.17 (0.87-1.57) | 1.88 (1.32-2.66)b |
| M1 | + age | 1.17 (0.87-1.58) | 1.93 (1.35-2.74)b |
| M2 | + sex | 1.20 (0.89-1.61) | 2.11 (1.48-3.03)b |
| M3 | + MELD-Na | 1.17 (0.87-1.57) | 2.01 (1.40-2.88)b |
| M4 | + era | 1.20 (0.89-1.62) | 2.01 (1.40-2.89)b |
| M5 | + albumin + ascites | 1.16 (0.86-1.57) | 1.99 (1.37-2.87)b |
| Covariate | csHR (95%CI) | sHR (95%CI) |
| HDV infection | 1.90 (1.30-2.77)b | 1.14 (0.84-1.55) |
| Age (year) | 1.01 (0.99-1.03) | 1.00 (0.99-1.02) |
| Female sex | 0.60 (0.41-0.88)b | 0.63 (0.46-0.86)b |
| MELD-Na per point | 1.07 (1.03-1.11)b | 1.03 (1.00-1.07) |
| Era II vs I | 0.86 (0.57-1.31) | 0.75 (0.53-1.08) |
| Albumin (per g/L) | 1.01 (0.97-1.04) | 0.99 (0.97-1.02) |
| Ascites | 1.50 (1.02-2.21)a | 1.31 (0.95-1.80) |
| Hepatocellular carcinoma | 2.38 (1.48-3.84)b | 1.33 (0.91-1.93) |
| Antiviral therapy | 0.92 (0.48-1.76) | 0.88 (0.49-1.57) |
The proportional-hazards assumption showed no evidence of violation for either cause-specific model (global scaled Schoenfeld residual test, P = 0.82 for waiting-list death and P = 0.41 for transplantation; HDV-specific P = 0.48 and P = 0.06, respectively; Figure 5 and Table 10). Thus, there was no statistically significant evidence that the effect of HDV on the instantaneous cause-specific hazard of either waiting-list death or transplantation changed over time. Although the HDV-specific P value for transplantation (P = 0.06) was close to the conventional significance threshold, it did not indicate a formal violation of the proportional-hazards assumption.
| Covariate | Waiting-list death (P value) | Transplantation (P value) |
| HDV | 0.479 | 0.061 |
| Age (year) | 0.537 | 0.334 |
| Female sex | 0.894 | 0.377 |
| MELD-Na (per point) | 0.938 | 0.556 |
| Era 2018-2025 | 0.230 | 0.971 |
| Global | 0.819 | 0.408 |
For the Fine-Gray models, the HDV-by-time interaction was non-significant for transplantation (P = 0.55) but signifi
| Model | Hepatitis delta virus × time coefficient | P value |
| Waiting-list death | -0.22 | 0.036 |
| Transplantation | -0.07 | 0.550 |
Splitting follow-up at one year showed that the adverse association between HDV infection and waiting-list death persisted throughout follow-up. HDV was associated with a higher cause-specific hazard of death both during the first year (csHR = 1.81, 95%CI: 1.05-3.14) and thereafter (csHR = 2.16, 95%CI: 1.35–3.44), with no evidence of attenuation over time. In contrast, although HDV was also associated with a higher cause-specific hazard of transplantation during both periods, this association was substantially stronger after the first year, increasing from a csHR of 1.81 (95%CI: 1.11-2.97) during the first year to 4.95 (95%CI: 2.13-11.50) thereafter (Table 12). The marked increase in transplantation among pa
| Follow-up interval | Waiting-list death, csHR (95%CI) | Transplantation, csHR (95%CI) |
| < 1 year | 1.81 (1.05-3.14) | 1.81 (1.11-2.97) |
| ≥ 1 year | 2.16 (1.35-3.44) | 4.95 (2.13-11.50) |
The convergence of the cumulative incidence curves for waiting-list death may therefore reflect the progressively stronger competing effect of transplantation rather than attenuation of the underlying mortality hazard associated with HDV. As patients with HDV were increasingly transplanted during follow-up, they were removed from the population remaining at risk of waiting-list death, thereby reducing the extent to which their persistently higher cause-specific mor
Using competing-risks methodology in the complete national liver transplant cohort, we assessed whether HDV infection provided prognostic information beyond standard severity markers. In the model for waiting-list death, HDV infection was independently associated with a twofold higher cause-specific hazard of death after adjustment for age, sex, MELD-Na score, and era of inclusion (csHR = 2.01, 95%CI: 1.40-2.89), whereas the corresponding subdistribution hazard was not significantly increased (sHR = 1.20, 95%CI: 0.89-1.62). In the unadjusted cumulative incidence analysis, the 5-year cumulative incidence of waiting-list mortality was comparable between the HDV and non-delta viral groups (45.9% vs 42.9%; Gray’s test, P = 0.696). However, this crude comparison was influenced by differences in baseline characteristics and, importantly, by the competing event of transplantation. Patients with HDV underwent transplantation more fre
In contrast, HDV infection was independently associated with greater access to transplantation (csHR = 2.41; sHR = 1.89). Female sex was independently associated with a lower cause-specific hazard of waiting-list death (csHR = 0.53), corresponding to a higher mortality risk in men, without a significant difference in access to transplantation.
The excess cause-specific mortality of HDV candidates is compatible with the biological aggressiveness of the infection rather than with a systematic disadvantage imposed by the allocation system, which in fact granted HDV candidates more frequent access to transplantation. These observations may inform future consideration of HDV status within frameworks analogous to existing MELD exception policies, such as those for hepatocellular carcinoma, portopulmonary hypertension, and hepatopulmonary syndrome. In each of these conditions, exception points compensate for a survival risk that standard MELD-Na fails to capture. The excess cause-specific waiting-list mortality observed in HDV candidates suggests that a similar rationale may apply; however, a formal exception policy would require external validation in inde
The attenuation of the MELD-Na association on the subdistribution scale (sHR = 1.04 per point) relative to the cause-specific scale (csHR = 1.07 per point) likely reflects the competing effect of transplantation, as higher-priority candidates are removed from the waiting list earlier, rather than a lack of prognostic value of MELD-Na. Conditional on remaining listed, their cause-specific hazard of death is certainly higher, but the subdistribution hazard is attenuated because trans
The lower cause-specific hazard of waiting-list death among women (csHR = 0.53), corresponding to a higher risk in men, should be interpreted with caution, although the association persisted after adjustment for MELD-Na score, age, etiology, and era. Similar sex-related differences in waiting-list outcomes have been reported in other cohorts and contributed to the development of the sex adjustment incorporated into MELD = 3.0[7]. Potential explanations include differences in body size and donor-recipient size matching, timing of referral, and factors related to allocation pathways that are not fully captured by MELD-based prioritization. These findings support the need for further evaluation of sex-specific access to transplantation within the national allocation system.
The main practical implication is that these data may inform future consideration of HDV status as a potential modifier of waiting-list prioritization, within a framework analogous to existing MELD exception policies. A complementary prognostic tool combining MELD-Na with HDV status could further improve risk stratification in HDV-endemic settings. This possibility is being explored separately through the derivation of a Delta Prognostic Score.
The national allocation system uses the descriptive categories “urgent” and “super-urgent”, but during the study period these did not correspond to a formally codified allocation status applied according to explicit clinical criteria, and they were therefore not modeled as a covariate. Because early transplantation could nevertheless reflect urgent prioritization, transplantations performed within 30 and 90 days of listing were excluded in a sensitivity analysis. The number of such procedures was small (13 and 33, respectively), and the association between HDV infection and transplantation persisted, indicating that the greater access to transplantation observed in HDV candidates was not driven by urgent cases but occurred within the elective allocation pathway.
Although the study is based on a complete national liver transplant cohort, its single-center design limits the generalizability of the findings, and external validation in other HDV-endemic populations remains necessary. The national transplant system is distinctive in that it combines national and international transplant pathways. Therefore, findings about allocation in Moldova should be interpreted carefully before applying them to larger transplant systems. Residual confounding cannot be excluded, in particular because the primary comparator group was etiologically heterogeneous; this was mitigated, but not eliminated, by the restricted-comparator analyses. The retrospective phase relied on archival records, and although the ambispective design allowed harmonization of variables, some misclassification of secondary outcomes, particularly delisting, cannot be excluded. MELD-Na was modeled as a baseline exposure at listing, without incorporating dynamic changes during the waiting period. Finally, the subdistribution hazard ratio does not carry the same clinical interpretation as a cause-specific hazard, a distinction that is relevant when the results are used to inform allocation policy.
In this national cohort, HDV infection was associated with an approximately twofold higher cause-specific hazard of waiting-list death. This excess did not translate into higher absolute waiting-list mortality, because delta candidates underwent transplantation earlier and more frequently. HDV status may be used as an additional factor in transplant risk stratification, especially in HDV-endemic settings. A formal MELD exception is not supported, given the similar absolute mortality and the lack of external validation.
The authors thank the National Agency of Transplantation of the Republic of Moldova and the team of the Hepatobiliary Surgery and Hepatology Department, Republican Clinical Hospital Timofei Mosneaga, Chisinau, for support with data collection and clinical follow-up.
| 1. | World Health Organization. Global hepatitis report 2024: action for access in low- and middle-income countries. [cited 7 August 2026]. Available from: https://www.who.int/publications/i/item/9789240091672. |
| 2. | GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396:1204-1222. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 14906] [Cited by in RCA: 13337] [Article Influence: 2222.8] [Reference Citation Analysis (40)] |
| 3. | D’Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J Hepatol. 2006;44:217-231. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 2533] [Cited by in RCA: 2280] [Article Influence: 114.0] [Reference Citation Analysis (6)] |
| 4. | Kamath PS, Wiesner RH, Malinchoc M, Kremers W, Therneau TM, Kosberg CL, D’Amico G, Dickson ER, Kim WR. A model to predict survival in patients with end-stage liver disease. Hepatology. 2001;33:464-470. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 4208] [Cited by in RCA: 3818] [Article Influence: 152.7] [Reference Citation Analysis (11)] |
| 5. | Wiesner R, Edwards E, Freeman R, Harper A, Kim R, Kamath P, Kremers W, Lake J, Howard T, Merion RM, Wolfe RA, Krom R; United Network for Organ Sharing Liver Disease Severity Score Committee. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology. 2003;124:91-96. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 2097] [Cited by in RCA: 1895] [Article Influence: 82.4] [Reference Citation Analysis (8)] |
| 6. | Kim WR, Biggins SW, Kremers WK, Wiesner RH, Kamath PS, Benson JT, Edwards E, Therneau TM. Hyponatremia and mortality among patients on the liver-transplant waiting list. N Engl J Med. 2008;359:1018-1026. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1236] [Cited by in RCA: 1141] [Article Influence: 63.4] [Reference Citation Analysis (5)] |
| 7. | Kim WR, Mannalithara A, Heimbach JK, Kamath PS, Asrani SK, Biggins SW, Wood NL, Gentry SE, Kwong AJ. MELD 3.0: The Model for End-Stage Liver Disease Updated for the Modern Era. Gastroenterology. 2021;161:1887-1895.e4. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 534] [Cited by in RCA: 528] [Article Influence: 105.6] [Reference Citation Analysis (3)] |
| 8. | Lai JC, Tandon P, Bernal W, Tapper EB, Ekong U, Dasarathy S, Carey EJ. Malnutrition, Frailty, and Sarcopenia in Patients With Cirrhosis: 2021 Practice Guidance by the American Association for the Study of Liver Diseases. Hepatology. 2021;74:1611-1644. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 575] [Cited by in RCA: 594] [Article Influence: 118.8] [Reference Citation Analysis (2)] |
| 9. | Wranke A, Heidrich B, Deterding K, Hupa-Breier KL, Kirschner J, Bremer B, Cornberg M, Wedemeyer H. Clinical long-term outcome of hepatitis D compared to hepatitis B monoinfection. Hepatol Int. 2023;17:1359-1367. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 36] [Cited by in RCA: 34] [Article Influence: 11.3] [Reference Citation Analysis (0)] |
| 10. | Gish RG, Wong RJ, Di Tanna GL, Kaushik A, Kim C, Smith NJ, Kennedy PTF. Association of hepatitis delta virus with liver morbidity and mortality: A systematic literature review and meta-analysis. Hepatology. 2024;79:1129-1140. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 30] [Cited by in RCA: 60] [Article Influence: 30.0] [Reference Citation Analysis (0)] |
| 11. | Turcanu A, Pitel E, Dumbrava VT, Tcaciuc E, Donscaia A, Peltec A, Pineau P. Profile of hepatocellular carcinoma in the Republic of Moldova: first-hand information on the presentation, distribution and etiologies. Rom J Intern Med. 2019;57:37-46. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 3] [Cited by in RCA: 4] [Article Influence: 0.6] [Reference Citation Analysis (1)] |
| 12. | Alfaiate D, Clément S, Gomes D, Goossens N, Negro F. Chronic hepatitis D and hepatocellular carcinoma: A systematic review and meta-analysis of observational studies. J Hepatol. 2020;73:533-539. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 159] [Cited by in RCA: 160] [Article Influence: 26.7] [Reference Citation Analysis (1)] |
| 13. | Stockdale AJ, Kreuels B, Henrion MYR, Giorgi E, Kyomuhangi I, de Martel C, Hutin Y, Geretti AM. The global prevalence of hepatitis D virus infection: Systematic review and meta-analysis. J Hepatol. 2020;73:523-532. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 583] [Cited by in RCA: 551] [Article Influence: 91.8] [Reference Citation Analysis (19)] |
| 14. | Miao Z, Zhang S, Ou X, Li S, Ma Z, Wang W, Peppelenbosch MP, Liu J, Pan Q. Estimating the Global Prevalence, Disease Progression, and Clinical Outcome of Hepatitis Delta Virus Infection. J Infect Dis. 2020;221:1677-1687. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 301] [Cited by in RCA: 297] [Article Influence: 49.5] [Reference Citation Analysis (4)] |
| 15. | Wang Y, Wang M, Liu C, Hao M, Wang W, Li Y, Shi J, Jia X, Zhang X, Dang S. Global burden of liver cirrhosis 1990-2019 and 20 years forecast: results from the global burden of disease study 2019. Ann Med. 2024;56:2328521. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 13] [Cited by in RCA: 54] [Article Influence: 27.0] [Reference Citation Analysis (2)] |
| 16. | Buga D, Hotineanu A, Taran N, Peltec A. Characteristics of candidates from the waiting list for liver transplantation in the Republic of Moldova. Public Health Econ Manage Med. 2023;4:117-121. |
| 17. | Buga D, Hotineanu A, Tcaciuc E, Taran N, Peltec A. Liver cirrhosis induced by delta virus: clinical and evolutive pecularities in patients on the waiting list for liver transplantation in Republic of Moldova. [cited 7 August 2026]. Available from: https://ibn.idsi.md/sites/default/files/imag_file/143_17.pdf. |
| 18. | Fine JP, Gray RJ. A Proportional Hazards Model for the Subdistribution of a Competing Risk. J Am Stat Assoc. 1999;94:496-509. [RCA] [DOI] [Full Text] [Cited by in Crossref: 8085] [Cited by in RCA: 10237] [Article Influence: 379.1] [Reference Citation Analysis (0)] |
| 19. | Kim WR, Therneau TM, Benson JT, Kremers WK, Rosen CB, Gores GJ, Dickson ER. Deaths on the liver transplant waiting list: an analysis of competing risks. Hepatology. 2006;43:345-351. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 110] [Cited by in RCA: 123] [Article Influence: 6.2] [Reference Citation Analysis (0)] |
| 20. | Gray RJ. A Class of $K$-Sample Tests for Comparing the Cumulative Incidence of a Competing Risk. Ann Statist. 1988;16. [DOI] [Full Text] |
| 21. | Austin PC, Lee DS, Fine JP. Introduction to the Analysis of Survival Data in the Presence of Competing Risks. Circulation. 2016;133:601-609. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 1040] [Cited by in RCA: 2092] [Article Influence: 209.2] [Reference Citation Analysis (0)] |
| 22. | Adam R, Karam V, Cailliez V, O Grady JG, Mirza D, Cherqui D, Klempnauer J, Salizzoni M, Pratschke J, Jamieson N, Hidalgo E, Paul A, Andujar RL, Lerut J, Fisher L, Boudjema K, Fondevila C, Soubrane O, Bachellier P, Pinna AD, Berlakovich G, Bennet W, Pinzani M, Schemmer P, Zieniewicz K, Romero CJ, De Simone P, Ericzon BG, Schneeberger S, Wigmore SJ, Prous JF, Colledan M, Porte RJ, Yilmaz S, Azoulay D, Pirenne J, Line PD, Trunecka P, Navarro F, Lopez AV, De Carlis L, Pena SR, Kochs E, Duvoux C; all the other 126 contributing centers (www. eltr.org) and the European Liver and Intestine Transplant Association (ELITA). 2018 Annual Report of the European Liver Transplant Registry (ELTR) - 50-year evolution of liver transplantation. Transpl Int. 2018;31:1293-1317. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 424] [Cited by in RCA: 377] [Article Influence: 47.1] [Reference Citation Analysis (6)] |