ABSTRACT
-
Background/Aims
The role of reactivation of human cytomegalovirus (HCMV) in determining outcomes of cirrhotic patients with acute decompensation (AD) is unknown. We aimed to investigate HCMV incidence and potential correlation with hepatic outcomes in AD patients.
-
Methods
Two prospective multicentre cohorts with AD patients were investigated. Patients in cohort 1 were recruited from 4 centres, while patients in cohort 2 were randomly selected from a second multicentre cohort. HCMV reactivation was established with quantitative real-time polymerase chain reaction assay in seropositive patients.
-
Results
HCMV reactivation was found in 35 patients from cohort 1 (n=722) and 14 from cohort 2 (n=291), with an incidence of 4.8% in both cohorts. Bacterial infection and liver failure were independently correlated with HCMV reactivation. HCMV reactivation was an independent predictor of 90-day mortality. Among bacterial infection populations in these two cohorts, patients with HCMV reactivation had worse prognosis compared to those without. Incidence of acute-on-chronic liver failure (ACLF) was higher in patients with HCMV reactivation compared to those without and was also independently correlated with development of ACLF. In a total of 49 HCMV reactivation cases, 8 patients were treated with ganciclovir, in whom a significantly lower 90-day mortality compared with those not treated was observed. All 3 patients who underwent liver transplantation with reactivation of HCMV died.
-
Conclusions
In AD patients, HCMV reactivation was common, especially in those with bacterial infection or liver failure, and they were more prone to having ACLF and 90‑day mortality. The data propose the need for active surveillance for HCMV infection in AD patients.
-
Keywords: Acute-on-chronic liver failure; Bacterial infection; Cytomegalovirus; End stage liver disease; Liver cirrhosis
Study Highlights
• HCMV reactivation frequently occurs in patients with acute decompensation during hospitalization, especially those with bacterial infection or liver failure.
• HCMV reactivation is an independent predictor of transplant-free 90-day mortality; some patients might benefit from ganciclovir therapy.
• HCMV reactivation may be a precipitating event for ACLF and is associated with respiratory failure.
Graphical Abstract
INTRODUCTION
Human cytomegalovirus (HCMV), as a member of the
Betaherpesvirinae, establishes latency and persists for the life of the individual. The global HCMV seroprevalence approximates to 60% in the general population and close to 100% in Chinese counterparts [
1,
2]. HCMV reactivation often emerges in immunocompromised patients with latent infection such as transplant recipients [
3,
4]. HCMV reactivation is related to an excess of graft rejection and is a major cause of morbidity and mortality during the first year after transplantation [
5,
6]. Both prophylactic [
7] and pre-emptive therapy [
8] with ganciclovir could reduce the risk of HCMV disease and death in organ transplantation populations. HCMV reactivation also occurs frequently in previously immunocompetent critically ill patients [
9], which has been reported to be associated with prolonged stay in the intensive care unit (ICU), prolonged need for mechanical ventilation and higher mortality [
10-
13].
Cirrhosis associated immune system dysfunction in patients with acute decompensation (AD) resembles that observed in sepsis patients, with the key components including systemic inflammation and immune deficiency [
14]. HCMV reactivation has been observed in populations with AD via microbial next generation sequencing [
15] or with liver failure via real-time quantitative polymerase chain reaction (qPCR) approach [
16] and its clinical relevance of adverse clinical outcomes has been suggested. However, the incidence of HCMV reactivation and its impact on the clinical course of patients with AD are not known. Therefore, the aims of this study were to prospectively screen for HCMV reactivation using commercially available real-time qPCR to evaluate the incidence of HCMV reactivation and its impact on the clinical course and short-term mortality of patients with liver cirrhosis presenting with AD.
MATERIALS AND METHODS
Study design
The study was performed with 2 prospective series of patients consecutively admitted for AD cirrhosis. Cohort 1 was used to explore the incidence, clinical characteristics of HCMV reactivation and relationships between HCMV reactivation and clinical outcomes. To validate the incidence of HCMV reactivation in AD patients, the clinical characteristics of patients experiencing HCMV reactivation and its impact on prognosis, an independent analytic cohort (cohort 2) extracted from the Chinese Acute-on-Chronic Liver Failure (CATCH-LIFE) study was used. Patients in the current analysis were followed up for 90 days.
Cohort 1 (NCT06039696)
This prospective multicentre cohort was conducted from November 2020 to December 2023 in four tertiary hospitals in China: Nanfang Hospital (Guangzhou, Southern China), Mengchao Hepatobiliary Hospital (Fuzhou, Eastern China), the First Affiliated Hospital of Xi’an Jiaotong University (Xi’an, Western China) and the Third Hospital of Hebei Medical University (Shijiazhuang, Northern China). Demographic, clinical and laboratory data were collected at admission from all patients screened for eligibility. Patients with AD who did not meet any exclusion criteria underwent HCMV serological screening (anti-HCMV IgG). HCMV seropositive participants were enrolled and HCMV DNA quantification was performed with plasma sample. Details for patients from each centre were shown in
Supplementary Table 1.
Cohort 2 (NCT02457637)
This is an independent analytic subgroup with AD patients selected from the investigation cohort of the CATCHLIFE study. The CATCH-LIFE cohort is a large cohort with 1,560 AD participants, and thus we used simple random sampling method to select the participants into the current study for reasons of cohort sustainability and economic considerations. The protocol and cohort profile of CATCHLIFE study have been described previously [
17]. Clinical and demographic characteristics of patients before and after random selection are shown in
Supplementary Table 2. Selected participants with available plasma samples underwent anti-HCMV IgG screening. HCMV DNA quantification tests were further performed in positive anti-HCMV IgG cases which were included in the current analysis.
The study protocols were approved by local ethics committee of the leading centre in cohort 1, Nanfang Hospital (NFEC-2020-255) and the leading centre in cohort 2, Renji Hospital (No. [2024] 148 k). Written informed consent was obtained from all study participants. Patients or the public were not involved in the design, conduct, reporting or dissemination plans of our research.
Study population
The inclusion criteria were: 1) able to give informed consent; 2) cirrhotic inpatients; 3) admission for AD (overt ascites, overt hepatic encephalopathy and acute variceal bleeding);18 4) HCMV seropositive. The exclusion criteria were: 1) age <18 or >80 years; 2) malignancy of liver or other organs (including leukaemia); 3) human immunodeficiency virus infection; 4) medical history of immune deficiency (organ transplantation, receipt of immunosuppressive medications for non-hepatic reasons, etc.); 5) insufficient volume of plasma sample for HCMV assays; 6) severe organ failure related to non-hepatic reasons.
Definition
Liver cirrhosis was diagnosed based on clinical, biochemical, and imaging or histological features. Acute-on-chronic liver failure (ACLF), division of ACLF grades and organ failure were defined according to clinical practice guidelines from the European Association for the Study of the Liver [
18]. Acute kidney injury was defined according to the current definition of the International Club of Ascites [
19]. Episodes of bacterial infection were defined as consistent with our previous study [
15]. Model for end-stage liver disease (MELD) score was calculated as recommended [
20]. HCMV reactivation was diagnosed if HCMV viremia was detectable by real-time qPCR.
Plasma samples were adopted for all HCMV assays. Briefly, blood samples collected at admission were centrifuged at 3,000 rpm for 10 minutes and the supernatant (plasma) was stored at –80℃ until analysis. A commercial enzyme immunoassay kit (Human Cytomegalovirus IgG ELISA Kit, Beijing Beier Bioengineering Co., Ltd., Beijing, China) was used for detection of IgG to HCMV indicating prior infection. DNA was extracted from 200 μl stored plasma using QIAamp MinElute Virus Spin Kit (57704, Qiagen) and HCMV DNA was quantified using a commercial real-time qPCR assay (S3014E, Sansure Biotech Inc., Changsha, China). The assay was performed and interpreted according to manufacturer recommendations with low limit detection level at 100 copies/mL.
Statistical analysis
Categorical data were reported as frequency (percentage), and continuous data were reported as median (interquartile range). Comparisons between variables were carried out using the chi-square test (or Fisher’s test) for categorical variables and the Mann–Whitney test for continuous variables. Survival function was described using the Kaplan–Meier method, and log-rank tests were used to compare the hazards of mortality between groups. All reported P-values are 2-sided, and P-value lower than 0.05 was considered significant.
Binary logistic regression model was used to evaluate the relationship between HCMV reactivation and baseline clinical characteristics. Due to the limited number of events (HCMV reactivation) in cohort 1 and cohort 2, the logistic regression analysis was conducted in the combined data from these two cohorts. Potential factors with P-value lower than 0.05 in the univariate analysis were considered for entry into multivariable models.
Factors associated with 90-day mortality were identified in a univariate Cox proportional hazards models. Transplanted patients were censored at the time of treatment. Risk factors with P-value lower than 0.05 and without collinearity were considered for entry into multivariable models in cohort 1. Due to the limit of the number of events (mortality) in cohort 2 (n=50), four confounding factors were included in the multivariable model in cohort 2. Point estimates and 95% confidence intervals (CIs) were depicted with forest plots.
Cumulative incidence of ACLF was estimated in a competing risk setting, where death and liver transplantation were considered as competing risks. The unadjusted and confounder-adjusted effects on the development of ACLF were analysed by using Fine and Gray competing risk regression models. Potential confounders were considered if a variable was significantly associated with ACLF development in univariate analysis at a level of 0.05 and without collinearity. The sub-distribution hazard ratio (sHR) with 95% CIs was used to describe the contribution of each variable to the risk of developing ACLF. Point estimates and 95% CIs were depicted with forest plots.
Cox proportional-hazards regression models were used to estimate the association between antiviral treatment for HCMV reactivation and 90-day death. In addition, to help account for the non-randomized treatment administration of antiviral treatment, propensity-score methods were used to reduce the confounding effects. The individual propensities for receipt of antiviral treatment were estimated with a multivariable logistic regression model that included the same covariates as the multivariate Cox regression model. Associations between antiviral treatment and 90-day mortality were then estimated by multivariable Cox regression models with the inverse probability weighting and the overlap weighting [
21-
23]. The statistical analyses were performed with the use of R version 4.4.1 (R Project for Statistical Computing), IBM SPSS Statistics 22.0 (IBM Co., Armonk, NY, USA) and PRISM Version 8.3 (GraphPad Software, San Diego, CA, USA).
RESULTS
Participants and HCMV serological profiles
Prevalence of latent HCMV infection was over 95% in individuals from all participating centres in various geographic locations of China (
Supplementary Fig. 1). Of the 729 prospectively enrolled patients from 4 centres, 722 patients (99.0%) positive anti-HCMV IgG were included in cohort 1 (
Fig. 1A). Demographic and clinical characteristics of participants are shown in
Table 1. The median age of participants was 53 (44–61) years, and 76.7% were male. The aetiology of liver cirrhosis was mainly hepatitis B virus (HBV) infection (62.3%), followed by alcohol consumption (19.4%). The median MELD-Na score at baseline was 19 (11–27). Among 139 patients (19.3%) with ACLF, 34 patients (4.7%) were classified as Grade 1, 79 (10.9%) as Grade 2 and 26 (3.6%) as Grade 3.
After HCMV serological assessment from 295 patients with plasma samples available, 291 patients (98.6%) with positive anti-HCMV IgG were included in cohort 2 (
Fig. 1B and
Table 1). HBV (60.1%) infection was also the leading aetiology of liver cirrhosis. The median MELD-Na score at baseline was 16 (10–24). Among 22 patients (7.6%) with ACLF at baseline, 2 (9.1%) were categorized into grade 1, 18 (81.8%) grade 2 and 2 (9.1%) grade 3.
There were 35 patients (35/722, 4.8%; 95% CI 3.5–6.7%) with detectable HCMV viremia at baseline in cohort 1 (
Supplementary Fig. 2A). HCMV reactivation was also noted in 4.8% patients (14/291; 95% CI 2.9–7.9%;
Supplementary Fig. 2A) in cohort 2. The median HCMV viral load was 2.3 log10 copies/mL (IQR 2.1–2.8) in cohort 1 and 2.7 log10 copies/mL (IQR 2.3–3.7) in cohort 2 (
Supplementary Fig. 2B). Details of their demographic and clinical characteristics were reported in
Supplementary Table 3.
Baseline characteristics of subjects with HCMV reactivation from cohort 1 and cohort 2 are shown in
Table 2. To further investigate the potential clinical characteristics correlated with HCMV reactivation, univariate and multivariate logistic regression analyses were performed in pooled population from two cohorts. In the univariate analysis, the presence of bacterial infection, presence of organ failure (including circulatory, liver and coagulation failure), ACLF and more severe liver disease (MELD-Na score) correlated with HCMV reactivation (
Fig. 2A). In multivariable models, presence of bacterial infection (OR 3.794; 95% CIs 1.621-8.881;
P=0.002) and liver failure defined as total bilirubin ≥12 mg/dL (OR 3.093; 95% CIs 1.637–5.843;
P=0.001) were independently associated with HCMV reactivation (
Fig. 2B).
The incidence of HCMV reactivation in patients with bacterial infection was higher than those without (7.8% vs. 1.2%,
P<0.001, cohort 1; 8.1% vs. 1.9%,
P=0.013, cohort 2;
Fig. 2C). Meanwhile, HCMV reactivation was more prevalent in patients with liver failure than those without (10.3% vs. 2.2%,
P<0.001, cohort 1; 10.7% vs. 2.8%,
P=0.011, cohort 2;
Fig. 2D).
90-day mortality
In cohort 1, 139 patients (19.3%) died and 27 (3.7%) received liver transplantation during a 90-day follow-up period. Among patients with HCMV reactivation in cohort 1, the 90-day cumulative mortality was 49.6% (95% CIs 32.7–66.5%), which was significantly higher than those without HCMV reactivation (18.1%; 95% CIs 15.2–21.0%;
P<0.001;
Fig. 3A). In cohort 2, 50 (17.2%) individuals died within 90-day follow-up. Patients with HCMV reactivation also had significantly higher 90-day mortality rate than those without (57.1% vs. 15.3%;
P<0.001,
Fig. 3B) in cohort 2.
Univariate analysis of 90-day survival in cohort 1 is shown in
Supplementary Fig. 3A. In the multivariable analysis, the presence of HCMV reactivation was independently associated with 90-day mortality after adjustment for other significant baseline variables (HR 1.821; 95% CIs 1.080–3.071;
P=0.025,
Fig. 3C). Univariate (
Supplementary Fig. 3B) and multivariable analysis showed that HCMV reactivation was also an independent predictor of 90-day mortality in cohort 2 (HR 3.156; 95% CIs 1.428–6.977;
P=0.005,
Fig. 3C).
In cohort 1, over 75% of patients with HCMV reactivation coexisted with bacterial infection (88.6% in cohort 1 and 78.6% in cohort 2). Bacterial infection patterns between patients with or without HCMV reactivation are shown in
Table 2. Pneumonia was the most common site of bacterial infection among both groups of patients in cohort 1, and the same phenomenon could be observed in cohort 2. Additionally, in cohort 1, patients with HCMV reactivation had higher incidence of pneumonia compared to those without (80.6% vs. 56.8%;
P=0.012). And the higher incidence of pneumonia could be observed among patients with HCMV reactivation in cohort 2 but without statistical difference (72.7% vs. 65.3%;
P=0.749). To further illustrate the contribution of HCMV reactivation to adverse outcomes in AD patients with bacterial infection, 90-day mortality in patients with HCMV reactivation was compared with those without HCMV reactivation. Patients with HCMV reactivation had higher 90-day mortality than those without HCMV reactivation (51.6% vs. 28.6%;
P=0.003,
Fig. 3D). This observation was confirmed in cohort 2 (72.7% vs. 28.0%;
P<0.001,
Fig. 3E).
Preclpltatlng factor for ACLF
In cohort 1, there were 139 patients (19.3%) diagnosed with ACLF and 583 patients (80.7%) with AD alone (
Table 1) at baseline. During follow-up of 28 days, 70 AD patients (70/583, 12.0%) developed ACLF. The incidence of ACLF in AD patients with HCMV reactivation was 42.9% (9/21), which was significantly higher than those without HCMV reactivation (10.9%, 61/562,
P<0.001,
Fig. 4A). In cohort 2, there were 22 patients (22/269, 8.2%) who developed ACLF within 28 days. In patients with HCMV reactivation, 30.0% patients (3/10) developed ACLF within 28 days of follow-up, which was significantly greater than in patients without HCMV reactivation (7.3%, 19/259,
P=0.039,
Fig. 4A).
AD patients without ACLF from two cohorts were pooled together to investigate the potential of whether HCMV reactivation was independently associated with the development of ACLF. Among pooled AD patients (n=852), the 28-day cumulative incidence of ACLF was 10.8% (95% CIs 8.8–13.0%). The estimated cumulative incidences of ACLF were 3.4% (95% CIs 2.0–5.4%) in patients without bacterial infection nor HCMV reactivation, 19.0% (95% CIs 13.4–21.0%) in those with bacterial infections alone, 33.3% (95% CIs 3.2–70.4%) in those only with HCMV reactivation and 40.0% (95% CIs 20.8–58.6%) in those with both bacterial infection and HCMV reactivation (
Fig. 4B).
Univariate analysis of factors associated with the development of ACLF in the pooled non-ACLF patients is shown in Figure 4C. Multivariate analysis showed that HCMV reactivation whether with or without bacterial coinfection (adjusted sHR 2.235 [P=0.010] and 1.956 [P<0.001], respectively) was independently associated with occurrence of ACLF within 28 days (
Fig. 4C). Biochemical parameters including total bilirubin, serum creatinine and INR were also independently associated with the development of ACLF during 28 days follow-up (
Fig. 4C). Patients with HCMV reactivation and bacterial infection were more likely to develop respiratory failure during 28-day follow-up compared to those with only bacterial infections and to those without HCMV infection (
Supplementary Fig. 4).
Liver transplantation
A total of 32 patients from cohort 1 and cohort 2 underwent liver transplantation during 90-day follow-up (
Supplementary Table 4). There were three patients with pre-transplant HCMV reactivation, who died (100.0%) soon after liver transplantation. Two developed extrahepatic bile duct necrosis with biliary fistula formation, complicated by severe infection, and ultimately died of septic shock. The third patient experienced graft failure post-transplantation, subsequently leading to multiorgan failures and death. Among patients without HCMV reactivation, four patients were lost to follow-up, and 6 out of the remaining patients (6/25, 24.0%) died after liver transplantation with septic shock.
Among a total of 49 patients with HCMV reactivation, nine from cohort 1 and none from cohort 2 received ganciclovir therapy (
Supplementary Table 3). The decision to treat or not was left to the judgement of the treating physician. All nine but one patient had ganciclovir treatment for over 72 hours. The clinical characteristics of patients with or without anti-HCMV treatment were shown in
Supplementary Table 5. The patient treated with ganciclovir for less than 72 hours was excluded from the following survival analysis. Patients with higher viral load (>1,000 copies/mL) who did not receive antiviral therapy had significantly worse 90-day prognosis than treated group of patients and low-level viral load group of untreated patients (
Fig. 5A). Further analysis was carried out to explore the impact of antiviral therapy on clinical outcomes in patients with HCMV reactivation. In the unadjusted analysis, there was no significant association between antiviral therapy and 90-day mortality (HR 0.323; 95% CI 0.076–1.377;
P=0.127). In the multivariable analysis, patients who had received ganciclovir were more likely to survive than patients not treated with ganciclovir (adjusted HR 0.237; 95% CI 0.053–1.052;
P=0.059,
Fig. 5B). Analyses with inverse probability weighting (HR 0.173; 95% CI 0.124–0.599;
P=0.001,
Fig. 5B) or overlap weighting (HR 0.211; 95% CI 0.076–0.584;
P=0.003,
Fig. 5B) both showed that patients with ganciclovir treatment had lower 90-day mortality.
DISCUSSION
The main observations of this multicentre and prospective study were that HCMV reactivation occurred in 4.8% patients with cirrhosis with AD, which was significantly higher in patients with bacterial infections or liver failure. This was independently correlated with 90-day mortality and might be a precipitating event for development of ACLF. Antiviral therapy with ganciclovir in the patients with evidence of HCMV reactivation was associated with better survival. Our data provide compelling evidence for surveillance for HCMV reactivation in seropositive cirrhosis patients with AD, particularly those with bacterial infection and liver failure.
HCMV is an important cause of opportunistic viral infections in immunosuppressed patients such as those with organ transplantation [
9,
24]. Cirrhosis patients in the ICU were found to be vulnerable to bacterial infection [
25], which may be due to cirrhosis associated immune dysfunction [
14]. Our previous study adopted metagenomic next-generation sequencing (mNGS) approach and found HCMV reactivation rate of 20.9% in a small AD population, which is a highly sensitive technique [
15]. However, its specificity is poor and mNGS is not available for routine clinical use. Therefore, a qPCR approach, which is available in routine clinical practice, was adopted in the current study. HCMV viremia was diagnosed in 4.8% patients with AD. Prevalence of HCMV viremia has been reported to be 5% and 16.7% in two sets of ACLF patients [
16,
26] with very different clinical consequences. Hu et al. [
26] found that HCMV infection was mild and asymptomatic in ACLF patients, while Yang et al. [
16] and our previous study [
15] observed a close relationship between HCMV infection and adverse hepatic outcomes in AD patients. These prior studies were limited by relatively small sample sizes, inclusion of patients from single centers, and bias due to the retrospective nature of these studies.
To address some of these limitations, this large, prospective, multicentre study was conducted. The data indicated that incidence of HCMV reactivation was relatively high, particularly in patients with bacterial infection or liver failure. Bacterial infection consists of an initial hyper-inflammatory phase followed by an immunosuppressive phase, both of which might potentially facilitate transcriptional activation of latent HCMV infection to reactivation [
27]. In our patients diagnosed with liver failure defined by total bilirubin ≥12 mg/dL, incidence of HCMV reactivation reached 10.3% in cohort 1 and 10.7% in cohort 2. The association between elevated bilirubin levels and HCMV reactivation may reflect severe liver dysfunction which is known to be associated with immune dysfunction or may be due to its direct immune-regulatory effects [
28]. Conjugated bilirubin has been shown to negatively affect the viral infection process by altering the immune microenvironment and immune cell proliferation [
29,
30].
Our study demonstrated that HCMV reactivation was independently correlated with 90-day mortality and may be an important precipitating event for ACLF. As reviewed by Imlay et al. [
27], at least 3 mechanisms may explain a potential causal relationship between HCMV reactivation and clinical outcomes; direct or indirect lung injury, amplification of systemic inflammation, and/or secondary immunosuppression that increases risk for secondary nosocomial infections. Our data showed that coexistence of HCMV reactivation and bacterial infection increased the incidence of lung failure compared to those with bacterial infection alone. Among our 42 HCMV patients with infection, over 75% of patients developed pneumonia. Seven patients (100%) who developed lung failure had pneumonia. These correlations were also observed with HCMV reactivation among transplant recipients [
9]. The high mortality of patients undergoing transplantation with active HCMV infection argues strongly for routine testing for reactivation in patients with ACLF and potentially exclusion of these patients from transplantation until the infection has been controlled. Our data also showed that HCMV reactivation correlated with the occurrence of ACLF in about 10% patients with no ACLF at presentation, which was mostly attributed to new onset of respiratory failure. Larger prospective studies that include tissue biopsy and laboratory investigations (innate and adaptive immune factors, and HCMV specific immune responses) will be necessary to define the mechanism(s) underlying the association of HCMV reactivation with adverse clinical outcomes in patients with critical illness.
Considering the benefit of prophylactic or pre-emptive therapy of HCMV reactivation in other scenarios, we further explored the potential benefit of anti-HCMV with ganciclovir in our AD patients. Due to observational design of the current study, screening and surveillance of HCMV reactivation, as well as the initiation of antiviral treatment was left to the decision of the clinician. Besides, screening CMV reactivation is not a routine practice among patients with AD in most of participating centres, this might be the reason why there was no difference in CMV viral load between treated and untreated patients. To help account for the non-randomized treatment administration of antiviral treatment, propensity score methods were used to reduce the effects of confounding. Our small sample size of patients receiving ganciclovir therapy showed that antiviral treatment over 3 days could improve the 90-day transplantation-free survival compared to those without antiviral therapy. Meanwhile, the results of this subgroup analysis also suggest that 1,000 copies/mL may serve as an actionable threshold for initiating antiviral treatment for HCMV reactivation. However, these data need to be confirmed in larger studies to better define whether ganciclovir should be used pre-emptively and confirm actionable thresholds to initiate treatment of AD patients with HCMV reactivation.
The data presented here should be interpreted taking into account its limitations. First, our patients were from China, where seropositivity for HCMV is extremely high, and thus HCMV reactivation in the current work was defined within the clinical context. HCMV reactivation could not be assured without more complicated molecular approaches like whole-genome sequencing, but de novo infection of HCMV might be relatively low in our patients based on previous studies in other population [
31-
34]. Second, HCMV reactivation was assessed only at baseline, which possibly underestimates the true incidence of HCMV reactivation rates. Third, the predominant end organ disease in our patients with HCMV reactivation was the lung. However, the lack of broncho-alveolar lavage samples limits conclusion on whether the HCMV was directly responsible for this or not. Finally, participants in cohort 2 were selected from the investigation cohort of the CATCH-LIFE study for economic and cohort sustainable considerations. The data and samples from 350 patients rather than the entire cohort were used in this study, and there are some differences in clinical characteristics between the selected subgroup and the entire cohort. Although this may lead to different results when estimating the incidence of viral reactivation, this external cohort of AD patients confirms that HCMV reactivation is relatively high in this population and is correlated with adverse outcomes such as higher mortality and greater incidence of ACLF.
In summary, with large sample size of AD patients from two multicentre cohorts, this study showed that HCMV reactivation was common, particularly in the patients with bacterial infection and liver failure. HCMV reactivation was independently associated with increased mortality in AD patients and development of ACLF. In particular, mortality was universal among patients who underwent liver transplantation with active HCMV reactivation providing a note of caution when listing these patients with ACLF for transplantation. Survival benefits from antiviral therapy were evidenced in a small number of patients with HCMV reactivation. Taken together, the data suggest that HCMV seropositive cirrhosis patients with AD should undergo surveillance for evidence of reactivation, particularly those with bacterial infection and liver failure.
FOOTNOTES
-
Authors’ contribution
C.H. collected clinical data and samples, performed experiments and analysis and drafted manuscript. Z.H., Y.H., and R.W. provided patients care, collected clinical data, and interpreted data in cohort 1. X.W., G.D., Y.G., Y.H., F.L., X.L., Z.M., and H.L. provided patients care, collected clinical data and samples and interpreted data in cohort 2. J.L., Y.L., B.L., Q.W., Y.Y., and W.L. collected clinical data and samples in cohort 1. X.L., Q.H., W.L., Q.L., L.Z., T.Q., Y.J., and M.L. provided patients care, participated in data acquisition and interpretation in two cohorts. Y.N., H.L., B.L., and R.J. made critical revision of the manuscript for important intellectual content. J.C. conceived the project, had full access to the data and controlled the decision to publish.
-
Acknowledgements
We thank the faculty and nurses from all the participating centres, and the patients who participated in the study and their families. The graphical abstract was created in Bio-Render.
This work was supported by the National Key Research and Development Program of China (2022YFC2304800, Jinjun Chen), National Natural Science Foundation of China (82370614, 82070650; Jinjun Chen), National Natural Science Foundation Youth Fund (82200688, Beiling Li), the Nat ional Science and Technology Major Project (2018ZX10723203, Jinjun Chen), Local Innovative and Research Teams Project of Guangdong Pearl River Talents Program (2017BT01S131, Jinjun Chen), Clinical Research Program of Nanfang Hospital, Southern Medical University (2020CR026, Jinjun Chen), Clinical Research Startup Program of Southern Medical University by High-level University Construction Funding of Guangdong Provincial Department of Education (LC2019ZD006, Jinjun Chen), Key-Area Research and Development Program of Guangdong Province (2019B020227004, Jinjun Chen), Guangzhou Basic Research Program (2025A04J5126, Beiling Li).
-
Conflicts of Interest
Rajiv Jalan is the Founder of Yaqrit Discovery, a spin out company from University College London, which has four subsidiaries, namely Amalive Ltd., Hepyx Ltd., CytoX Limited and Enterosorb Ltd. He is also a co-founder of Cyberliver Ltd. He had research collaborations with Yaqrit Discovery. Other authors declare that there is no conflict of interest.
SUPPLEMENTARY MATERIAL
Supplementary material is available at Clinical and Molecular Hepatology website (
http://www.e-cmh.org).
Supplementary Figure 1.
HCMV-seropositive rate in participating centres in different locations of China. Cohort 1: ⑥ Nanfang Hospital: 463/469, 98.7%; ⑦ Mengchao Hepatobiliary Hospital of Fujian Medical University: 127/127, 100%; ⑧ The First Affiliated Hospital of Xi’an Jiaotong University: 71/71, 100%; ⑨ The Third Hospital of Hebei Medical University: 61/62, 98.4%. Cohort 2: ① The First Hospital of Jilin University: 25/25, 100%; ② Beijing Ditan Hospital: 42/43, 97.7%; ③ Ren Ji Hospital: 73/73, 100%; ④ Southwest Hospital: 37/37, 100%; ⑤ Xiangya Hospital: 20/20, 100%; ⑥ Nanfang Hospital: 64/65, 98.5%. *Centres in Cohort 2 with no less than 20 participants are shown. HCMV, human cytomegalovirus.
Supplementary Figure 2.
Details for HCMV reactivation in both cohorts. (A) Incidence of HCMV reactivation at baseline in cohort 1 (n=722) and cohort 2 (n=291). (B) HCMV viral load among patients with HCMV reactivation in cohort 1 (n=35) and cohort 2 (n=14). HCMV, human cytomegalovirus.
Supplementary Figure 3.
Univariate analysis of HCMV reactivation and 90-day transplant-free survival. Univariable analysis of factors associated with 3-month mortality in cohort 1 (A) and cohort 2 (B). #A total of 32 patients received liver transplantations within 90-day follow-up: 27 in cohort 1 and 5 in cohort 2. ACLF, acute-on-chronic liver failure; CI, confidential interval; HBV, hepatitis B virus; HCMV, human cytomegalovirus; HR, hazard ratio; INR, international normalized ratio; MELD, model for end-stage liver disease.
Supplementary Figure 4.
Infection and six major organ failures. Proportion of development of six major organ failures, including lungs, circulation, brain, liver, kidney and coagulation in “Neither bacterial infection nor HCMV reactivation”, “Bacterial infection without HCMV reactivation” and “HCMV reactivation without bacterial infection” and “Both bacterial infection and HCMV reactivation” in pooled study population. *Presents P-value lower than 0.05; the absence of labelling indicates that there is no difference between the two groups. BI, bacterial infection; HCMV, human cytomegalovirus.
Figure 1.Flowchart of cohort 1 and cohort 2. Flowchart of the inclusion and exclusion process for the study population in cohort 1 (A) and cohort 2 (B). AD, acute decompensation; HCMV, human cytomegalovirus.
Figure 2.Clinical characteristics of patients with HCMV reactivation. Univariable (A) and multivariable (B) logistic regression analysis of HCMV reactivation in pooled study population from cohort 1 and cohort 2. Variables included in multivariate analysis were bacterial infection, circulatory failure, liver failure, coagulation failure, ACLF and MELD-Na score. Incidence of HCMV reactivation according to status of bacterial infection (C) and liver failure (D) in cohort 1 and cohort 2. Level of significance: P<0.001 in cohort 1 and P=0.013 in cohort 2 (chisquare test); P<0.001 in cohort 1 and P=0.011 in cohort 2 (Fisher’s test), respectively. ACLF, acute-on-chronic liver failure; CI, confidence interval; HBV, hepatitis B virus; HCMV, human cytomegalovirus; INR, international normalized ratio; MELD, model for end-stage liver disease; OR, odds ratio.
Figure 3.HCMV reactivation and 90-day mortality. Kaplan–Meier survival curves grouped in “HCMV reactivation” and “Without HCMV reactivation” in cohort 1 (A) and cohort 2 (B). Level of significance: P-value <0.001 (log-rank test). (C) Independent predictive factors associated with 3-month mortality in cohort 1 and cohort 2. Variables included in multivariate analysis in cohort 1 were leucocyte count, albumin, alanine transaminase, aspartate transaminase, MELD-Na scores, ACLF, bacterial infection and HCMV reactivation. Variables included in multivariate analysis in cohort 2 were MELD-Na scores, ACLF, bacterial infection and HCMV reactivation. Kaplan–Meier survival curves grouped in “HCMV reactivation” and “Without HCMV reactivation” among bacterial infection population from cohort 1 (D) and cohort 2 (E). Level of significance: P=0.003 in cohort 1 and P<0.001, respectively (log-rank test). ACLF, acute-on-chronic liver failure; CI, confidence interval; HBV, hepatitis B virus; HCMV, human cytomegalovirus; HR, hazard ratio; INR, international normalized ratio; MELD, model for end-stage liver disease.
Figure 4.HCMV reactivation and development of ACLF. (A) Proportion of development of ACLF in patients “without HCMV reactivation” and patients “with HCMV reactivation” in cohort 1 and cohort 2. Level of significance: P<0.001 (Fisher’s test) in cohort 1 and P=0.039 (Fisher’s test) in cohort 2. (B) Cumulative incidence curves of development of ACLF according to the status of infection. Level of significance: P<0.001 (Gray’s test). (C) Univariate and multivariable competing risk model analysis of ACLF development in pooled patients. Variables included in multivariate analysis were HBV, infection, leucocyte count, bilirubin, INR, and serum creatinine. ACLF, acute-on-chronic liver failure; BI, bacterial infection; CI, confidence interval; HBV, hepatitis B virus; HCMV, human cytomegalovirus; HR, hazard ratio; INR, international normalized ratio; MELD, model for end-stage liver disease.
Figure 5.Association between anti-HCMV treatment and 90-day transplant-free mortality in patients with HCMV reactivation. (A) Kaplan–Meier survival curves grouped in “Treated”, “Untreated with high-level viral load (>1,000 copies/mL)” and “Untreated with low-level viral load (≤1,000 copies/mL)” in patients with HCMV reactivation. The area between two dashed lines represents pointwise 95% confidence intervals. Level of significance: P=0.024 (log-rank test). (B) Associations between antiviral treatment and 90-day mortality in the multivariable analysis, and propensity-score analyses. #Variables included in multivariate analysis were bacterial infection, MELD-Na scores and ACLF. ‡Multivariable Cox proportional-hazards model with the same covariates with inverse probability weighting according to the propensity score. †Multivariable Cox proportional-hazards model with the same covariates with overlap weighting according to the propensity score. CI, confidence interval; IPW, inverse probability weighting; LT, liver transplantation; OW, overlap weighting.
Table 1.Baseline characteristics of the study population
Table 1.
|
Variable |
Cohort 1 (n=722) |
Cohort 2 (n=291) |
|
Age (yr) |
53 (44–61) |
52 (45–60) |
|
Male |
554 (76.7) |
211 (72.5) |
|
Etiology |
|
|
|
HBV |
406 (56.2) |
145 (49.8) |
|
HBV plus others |
44 (6.1) |
30 (10.3) |
|
Alcohol related |
140 (19.4) |
43 (14.8) |
|
Autoimmune liver diseases |
52 (7.2) |
33 (11.3) |
|
Others |
80 (11.1) |
40 (13.7) |
|
Presence of ascites |
641 (88.8) |
157 (54.0) |
|
Presence of HE |
115 (15.9) |
67 (23.0) |
|
Presence of acute bleeding |
117 (16.2) |
96 (33.0) |
|
Presence of infection |
397 (55.0) |
135 (46.4) |
|
Leucocyte count (×109/L) |
5.3 (3.6–7.8) |
4.9 (3.3–7.8) |
|
Bilirubin (mg/dL) |
4.3 (1.6–16.1) |
3.4 (1.3–12.2) |
|
INR |
1.5 (1.3–2.0) |
1.5 (1.2–1.8) |
|
Serum creatinine (mg/dL) |
0.8 (0.7–1.0) |
0.7 (0.6–0.9) |
|
Diagnosis of ACLF |
139 (19.3) |
22 (7.6) |
|
ACLF |
|
|
|
Grade 1 |
34 (4.7) |
2 (0.7) |
|
Grade 2 |
79 (10.9) |
18 (6.2) |
|
Grade 3 |
26 (3.6) |
2 (0.7) |
|
MELD-Na score |
19 (11–27) |
16 (10–24) |
Table 2.Baseline characteristics of the study population according to HCMV reactivation at baseline
Table 2.
|
Variable at baseline |
Cohort 1
|
Cohort 2
|
|
HCMV reactivation (n=35) |
No HCMV reactivation (n=687) |
P-value |
HCMV reactivation (n=14) |
No HCMV reactivation (n=277) |
P-value |
|
Age (yr) |
52 (44–61) |
53 (44–61) |
0.859 |
55 (46–65) |
52 (45–60) |
0.434 |
|
Male |
28 (80.0) |
526 (76.6) |
0.639 |
10 (71.4) |
201 (72.6) |
>0.999 |
|
Etiology |
|
|
0.947 |
|
|
0.815 |
|
HBV |
22 (62.9) |
428 (62.3) |
|
8 (57.1) |
167 (60.3) |
|
|
Non-HBV |
13 (37.1) |
259 (37.7) |
|
6 (42.9) |
110 (39.7) |
|
|
Ascites |
33 (94.3) |
608 (88.5) |
0.443 |
10 (71.4) |
147 (53.1) |
0.179 |
|
HE |
12 (34.3) |
103 (15.0) |
0.002 |
5 (35.7) |
62 (22.4) |
0.325 |
|
Acute bleeding |
6 (17.1) |
111 (16.2) |
0.877 |
3 (21.4) |
93 (33.6) |
0.561 |
|
Leucocyte count (×109/L) |
9.3 (6.1–12.6) |
5.2 (3.6–7.5) |
<0.001 |
11.7 (5.1–17.6) |
4.7 (3.2–7.5) |
0.006 |
|
Bilirubin (mg/dL) |
19.6 (11.2–24.6) |
4.0 (1.5–15.2) |
<0.001 |
19.6 (6.0–31.4) |
3.2 (1.2–11.7) |
0.001 |
|
Albumin (g/L) |
30.2 (28.1–34.1) |
30.2 (26.7–33.7) |
0.486 |
30.6 (28.8–33.3) |
29.5 (25.6–33.4) |
0.326 |
|
Alanine transaminase (U/L) |
60 (39–109) |
32 (20–70) |
0.002 |
51 (28–85) |
43 (26–103) |
0.786 |
|
Aspartate transaminase (U/L) |
116 (61–159) |
54 (34–105) |
<0.001 |
73 (47–125) |
59 (34–133) |
0.406 |
|
INR |
2.1 (1.5–2.7) |
1.5 (1.3–2.0) |
<0.001 |
1.8 (1.4–2.5) |
1.4 (1.2–1.8) |
0.032 |
|
Serum creatinine (mg/dL) |
0.8 (0.7–1.0) |
0.8 (0.7–1.0) |
0.851 |
0.8 (0.6–1.2) |
0.7 (0.6–0.9) |
0.564 |
|
Bacterial infection |
31 (88.6) |
366 (53.3) |
<0.001 |
11 (78.6) |
124 (44.8) |
0.013 |
|
Site of infection |
|
|
|
|
|
|
|
Pneumonia |
25 (80.6) |
208 (56.8) |
0.012 |
8 (72.7) |
81 (65.3) |
0.749 |
|
Bacteremia |
0 (0) |
13 (3.6) |
0.611 |
0 (0) |
2 (1.6) |
>0.999 |
|
Peritonitis |
2 (6.5) |
66 (18.0) |
0.135 |
2 (18.2) |
21 (16.7) |
>0.999 |
|
Urinary tract infection |
1 (3.2) |
20 (5.5) |
1.000 |
1 (9.1) |
4 (3.2) |
0.351 |
|
Other infection |
2 (6.5) |
11 (3.0) |
0.269 |
0 (0) |
8 (6.5) |
>0.999 |
|
Unproven infection |
4 (12.9) |
83 (22.7) |
0.262 |
2 (18.2) |
15 (12.1) |
0.630 |
|
Organ failure |
|
|
|
|
|
|
|
Respiratory failure |
0 (0) |
7 (1.0) |
1.000 |
0 (0) |
1 (0.4) |
>0.999 |
|
Circulatory failure |
3 (8.6) |
10 (1.5) |
0.021 |
0 (0) |
2 (0.7) |
>0.999 |
|
Brain failure |
3 (8.6) |
37 (5.4) |
0.435 |
1 (7.1) |
7 (2.5) |
0.329 |
|
Liver failure |
24 (68.6) |
209 (30.4) |
<0.001 |
8 (57.1) |
67 (24.2) |
0.011 |
|
Coagulation failure |
11 (31.4) |
85 (12.4) |
0.003 |
4 (28.6) |
23 (8.3) |
0.031 |
|
Kidney failure |
1 (2.9) |
42 (6.1) |
0.715 |
1 (7.1) |
5 (1.8) |
0.258 |
|
ACLF |
14 (40.0) |
125 (18.2) |
0.001 |
4 (28.6) |
18 (6.5) |
0.015 |
|
MELD-Na score |
27 (23–32) |
18 (11–26) |
<0.001 |
27 (20–33) |
15 (10–23) |
0.001 |
Abbreviations
acute-on-chronic liver failure
Chinese Acute-on-Chronic Liver Failure study
international normalized ratio of thrombin time
model for end-stage liver disease
polymerase chain reaction
sub-distribution hazard ratio
REFERENCES
- 1. Zuhair M, Smit GSA, Wallis G, Jabbar F, Smith C, Devleesschauwer B, et al. Estimation of the worldwide seroprevalence of cytomegalovirus: A systematic review and meta-analysis. Rev Med Virol 2019;29:e2034.
- 2. Wen L, Qiu Y, Cheng S, Jiang X, Ma YP, Fang W, et al. Serologic and viral genome prevalence of HSV, EBV, and HCMV among healthy adults in Wuhan, China. J Med Virol 2018;90:571-581.
- 3. Kotton CN, Kumar D, Caliendo AM, Huprikar S, Chou S, Danziger-Isakov L, et al. The third international consensus guidelines on the management of cytomegalovirus in solid-organ transplantation. Transplantation 2018;102:900-931.
- 4. Ljungman P, de la Camara R, Robin C, Crocchiolo R, Einsele H, Hill JA, et al. Guidelines for the management of cytomegalovirus infection in patients with haematological malignancies and after stem cell transplantation from the 2017 European Conference on Infections in Leukaemia (ECIL 7). Lancet Infect Dis 2019;19:e260-e272.
- 5. Rubin RH. The indirect effects of cytomegalovirus infection on the outcome of organ transplantation. Jama 1989;261:3607-3609.
- 6. Griffiths P. The direct and indirect consequences of cytomegalovirus infection and potential benefits of vaccination. Antiviral Res 2020;176:104732.
- 7. Vernooij RW, Michael M, Ladhani M, Webster AC, Strippoli GF, Craig JC, et al. Antiviral medications for preventing cytomegalovirus disease in solid organ transplant recipients. Cochrane Database Syst Rev 2024;5:Cd003774.
- 8. Owers DS, Webster AC, Strippoli GF, Kable K, Hodson EM. Pre-emptive treatment for cytomegalovirus viraemia to prevent cytomegalovirus disease in solid organ transplant recipients. Cochrane Database Syst Rev 2013;2013:Cd005133.
- 9. Ong DSY, Chong GM, Chemaly RF, Cremer OL. Comparative clinical manifestations and immune effects of cytomegalovirus infections following distinct types of immunosuppression. Clin Microbiol Infect 2022;28:1335-1344.
- 10. Heininger A, Jahn G, Engel C, Notheisen T, Unertl K, Hamprecht K. Human cytomegalovirus infections in nonimmunosuppressed critically ill patients. Crit Care Med 2001;29:541-547.
- 11. Jaber S, Chanques G, Borry J, Souche B, Verdier R, Perrigault PF, et al. Cytomegalovirus infection in critically ill patients: associated factors and consequences. Chest 2005;127:233-241.
- 12. Limaye AP, Kirby KA, Rubenfeld GD, Leisenring WM, Bulger EM, Neff MJ, et al. Cytomegalovirus reactivation in critically ill immunocompetent patients. Jama 2008;300:413-422.
- 13. Heininger A, Haeberle H, Fischer I, Beck R, Riessen R, Rohde F, et al. Cytomegalovirus reactivation and associated outcome of critically ill patients with severe sepsis. Crit Care 2011;15:R77.
- 14. Albillos A, Martin-Mateos R, Van der Merwe S, Wiest R, Jalan R, Álvarez-Mon M. Cirrhosis-associated immune dysfunction. Nat Rev Gastroenterol Hepatol 2022;19:112-134.
- 15. Li B, Hong C, Fan Z, Cai S, He Q, Lan X, et al. Prognostic and therapeutic significance of microbial cell-free DNA in plasma of people with acute decompensation of cirrhosis. J Hepatol 2023;78:322-332.
- 16. Yang Q, Zhou Z, Yang X, Chen Y, Liu A, Zhang B, et al. Latent cytomegalovirus reactivation in patients with liver failure: a 10-year retrospective case-control study, 2011-2020. Front Cell Infect Microbiol 2021;11:642500.
- 17. Gu WY, Xu BY, Zheng X, Chen J, Wang XB, Huang Y, et al. Acute-on-chronic liver failure in China: rationale for developing a patient registry and baseline characteristics. Am J Epidemiol 2018;187:1829-1839.
- 18. European Association for the Study of the Liver. EASL Clinical Practice Guidelines on acute-on-chronic liver failure. J Hepatol 2023;79:461-491.
- 19. Angeli P, Gines P, Wong F, Bernardi M, Boyer TD, Gerbes A, et al. Diagnosis and management of acute kidney injury in patients with cirrhosis: revised consensus recommendations of the International Club of Ascites. Gut 2015;64:531-537.
- 20. Kim WR, Mannalithara A, Heimbach JK, Kamath PS, Asrani SK, Biggins SW, et al. MELD 3.0: The model for end-stage liver disease updated for the modern era. Gastroenterology 2021;161:1887-1895.e1884.
- 21. Chesnaye NC, Stel VS, Tripepi G, Dekker FW, Fu EL, Zoccali C, et al. An introduction to inverse probability of treatment weighting in observational research. Clin Kidney J 2022;15:14-20.
- 22. Li F, Morgan KL, Zaslavsky AM. Balancing covariates via propensity score weighting. J Am Stat Assoc 2018;113:390-400.
- 23. Thomas LE, Li F, Pencina MJ. Overlap weighting: a propensity score method that mimics attributes of a randomized clinical trial. Jama 2020;323:2417-2418.
- 24. Ljungman P, Boeckh M, Hirsch HH, Josephson F, Lundgren J, Nichols G, et al. Definitions of cytomegalovirus infection and disease in transplant patients for use in clinical trials. Clin Infect Dis 2017;64:87-91.
- 25. Gustot T, Felleiter P, Pickkers P, Sakr Y, Rello J, Velissaris D, et al. Impact of infection on the prognosis of critically ill cirrhotic patients: results from a large worldwide study. Liver Int 2014;34:1496-1503.
- 26. Hu J, Zhao H, Lou D, Gao H, Yang M, Zhang X, et al. Human cytomegalovirus and Epstein-Barr virus infections, risk factors, and their influence on the liver function of patients with acute-on-chronic liver failure. BMC Infect Dis 2018;18:577.
- 27. Imlay H, Limaye AP. Current understanding of cytomegalovirus reactivation in critical illness. J Infect Dis 2020;221:S94-s102.
- 28. Vítek L, Tiribelli C. Bilirubin: The yellow hormone? J Hepatol 2021;75:1485-1490.
- 29. Castro-García FP, Corral-Jara KF, Escobedo-Melendez G, Sandoval-Hernandez MA, Rosenstein Y, Roman S, et al. Conjugated bilirubin affects cytokine profiles in hepatitis A virus infection by modulating function of signal transducer and activator of transcription factors. Immunology 2014;143:578-587.
- 30. Liu Y, Zhu P, Wang W, Tan X, Liu C, Chen Y, et al. Mucosal-associated invariant T cell dysregulation correlates with conjugated bilirubin level in chronic HBV infection. Hepatology 2021;73:1671-1687.
- 31. Bowden RA, Sayers M, Flournoy N, Newton B, Banaji M, Thomas ED, et al. Cytomegalovirus immune globulin and seronegative blood products to prevent primary cytomegalovirus infection after marrow transplantation. N Engl J Med 1986;314:1006-1010.
- 32. Bowden RA, Slichter SJ, Sayers M, Weisdorf D, Cays M, Schoch G, et al. A comparison of filtered leukocyte-reduced and cytomegalovirus (CMV) seronegative blood products for the prevention of transfusion-associated CMV infection after marrow transplant. Blood 1995;86:3598-3603.
- 33. Kotton CN, Kumar D, Manuel O, Chou S, Hayden RT, Danziger-Isakov L, et al. The fourth international consensus guidelines on the management of cytomegalovirus in solid organ transplantation. Transplantation 2025;109:1066-1110.
- 34. Kalil AC, Florescu DF. Prevalence and mortality associated with cytomegalovirus infection in nonimmunosuppressed patients in the intensive care unit. Crit Care Med 2009;37:2350-2358.
, Hai Li12
