TEL.: +91-11-29871246; FAX: +91-11-41611010
Cytomegalovirus infection: Its incidence and management in cytomegalovirus-seropositive living related liver transplant recipients: A single-center experience
Article first published online: 7 DEC 2012
Copyright © 2012 American Association for the Study of Liver Diseases
Volume 18, Issue 12, pages 1448–1455, December 2012
How to Cite
Wadhawan, M., Gupta, S., Goyal, N., Vasudevan, K. R., Makki, K., Dawar, R., Sardana, R., Lal, N. and Kumar, A. (2012), Cytomegalovirus infection: Its incidence and management in cytomegalovirus-seropositive living related liver transplant recipients: A single-center experience. Liver Transpl, 18: 1448–1455. doi: 10.1002/lt.23540
- Issue published online: 7 DEC 2012
- Article first published online: 7 DEC 2012
- Accepted manuscript online: 18 AUG 2012 03:35AM EST
- Manuscript Accepted: 12 AUG 2012
- Manuscript Received: 17 MAR 2012
It is believed that antiviral prophylaxis decreases the incidence of cytomegalovirus (CMV) reactivation and disease. There are few data regarding weekly assays for CMV DNA after transplantation and the subsequent management of CMV. Here we report a cohort of living related liver transplantation (LRLT) patients who were treated for invasive CMV disease or for CMV infections if they were receiving steroids for rejection. Patients who underwent liver transplantation at our center between September 2006 and August 2010 and were recipient-positive/donor-positive (R+/D+) were prospectively included. Patients were tested for CMV DNA 3 weeks after transplantation. CMV DNA–positive patients underwent weekly DNA monitoring until there were 2 consecutive negative reports. Those who developed CMV disease or had rising DNA titers while they were on treatment for rejection were treated. A Cox regression analysis was performed for factors predicting survival. Two hundred sixty-six of the 306 R+/D+ patients were CMV DNA–negative 3 weeks after transplantation, and 40 had detectable DNA. One of the DNA-negative patients developed CMV disease after treatment for rejection with methylprednisolone. Thirty patients had <500 copies/mL, and 10 had ≥500 copies/mL. Two of the 30 patients with DNA levels < 500 copies/mL developed CMV disease. Six of the 10 patients with DNA levels ≥500 copies/mL developed disease. CMV disease occurred in 9 of the 306 patients (2.9%). One patient received treatment for a rise in DNA titers while he was receiving steroids. There was a significant correlation between steroid administration for acute cellular rejection (ACR) and CMV reactivation (P = 0.003) and disease (P = 0.002). A multivariate analysis showed that CMV reactivation/disease did not predict survival. There was no difference in survival between CMV DNA–positive patients and CMV DNA–negative patients (P = 0.68). In conclusion, CMV reactivation is common after LRLT (13%), but the disease is rare (2.9%) without prophylaxis in CMV immunoglobulin G–positive recipients. The administration of steroids for ACR strongly correlates with CMV reactivation and disease. CMV reactivation and disease did not affect survival in our patient cohort. Liver Transpl, 2012. © 2012 AASLD.
Cytomegalovirus (CMV) infection is regarded as a major problem after liver transplantation. Therefore, patients are routinely tested for CMV serology and CMV viremia before and after transplantation, respectively. It is believed that antiviral prophylaxis significantly decreases the incidence of CMV infection/reactivation and disease.1-4 Many transplant centers routinely administer antiviral prophylaxis to all patients at risk for CMV disease after organ transplantation. The 2 basic approaches to preventing CMV disease are universal prophylaxis and preemptive therapy. With the prophylactic approach, antiviral therapy is administered to all patients for a defined period after transplantation (usually approximately 3 months). With the preemptive approach, clinicians assess the risk of CMV infection by monitoring patients with CMV DNA every 1 to 2 weeks, and they treat the patients if the results are positive. The risk of CMV infection and disease can be categorized by the pretransplant CMV immunoglobulin G (IgG) status of both the recipient and the donor.5, 6 The incidence of infection and disease is greatest in recipient-negative/donor-positive (R−/D+) patients. The reported incidence of CMV reactivation in recipient-positive/donor-positive (R+/D+) patients is 15% to 50%.6-8 However, it is not very clear how many patients who suffer reactivation clear the virus spontaneously and how many develop CMV disease. Furthermore, there is a controversy about the duration of prophylaxis and treatment for CMV disease. The incidence of CMV IgG positivity in the Indian subcontinent is >95%,9 and this places this population at intermediate risk for CMV disease.
Here we report a cohort of living related liver transplantation (LRLT) patients who were systematically monitored for CMV reactivation and disease and were treated either for established CMV disease or for CMV infections if they were receiving high-dose steroids for rejection. We also analyze the effects of CMV reactivation and disease on overall patient survival.
PATIENTS AND METHODS
All CMV IgG–positive patients who underwent liver transplantation at our center from September 2006 to August 2010 were prospectively included in this study. Patients who were negative for CMV IgG before transplantation were excluded from the final analysis. All patients were followed until July 2011 or death. Patients who died within 1 month of transplantation from causes unrelated to a CMV infection were excluded from the final analysis. The recorded data included the following: baseline demographics, pretransplant Child-Pugh Score and Model for End-Stage Liver Disease (MELD) scores, CMV antibody seropositivity, incidence of rejection after transplantation, administration of high-dose steroids for acute cellular rejection (ACR), biliary complications, sepsis (defined as clinical evidence of an infection with positive bacterial or fungal cultures in blood, sputum, drains, or urine), mortality, and follow-up. The baseline demographic data for these patients are presented in Table 1.
|Median age*||48 years (6 months to 76 years)|
|Hepatocellular carcinoma: all etiologies (n)||67|
|Median MELD score*||19 (6-50)|
|Child-Pugh score (n)|
|Median follow-up (months)*||28 (1-58)|
Informed consent was obtained from all patients. The ethics committee of the hospital approved the study protocol.
CMV infection and disease were defined in accordance with previously published definitions.10 A CMV infection was defined as the presence of CMV nucleic acid (eg, DNA polymerase chain reaction) in any body fluid or tissue specimen. All CMV infections were assumed to be reactivations because all patients were positive for CMV serology before transplantation. Patients who were positive for CMV DNA and had no clinical signs or symptoms were considered to be experiencing an asymptomatic CMV reactivation. Patients who were positive for CMV DNA and had clinical signs or symptoms (fever, neutropenia, and/or thrombocytopenia not attributable to any other cause) but had no evidence of end-organ disease were considered to have CMV syndrome. Finally, patients who were positive for CMV DNA and had evidence of end-organ disease (typically based on organ-specific symptoms and histological evidence of invasive CMV) were considered to have tissue-invasive CMV disease.
Immunosuppression and Rejection Treatment
All patients initially received triple immunosuppression (a calcineurin inhibitor, mycophenolate, and steroids). Tacrolimus was used in 90% of the recipients, and trough levels were maintained between 5 and 8 ng/mL. Cyclosporine was used in children less than 3 years old and when tacrolimus was not tolerated. The mycophenolate dosage was 500 mg twice a day for recipients who weighed 60 kg or less and 750 mg twice a day for recipients who weighed more than 60 kg. Steroids were tapered off by the end of 3 months for most patients. Rejection episodes were treated with high-dose methylprednisolone (500 mg/day) for 3 days and then a tapering dose of prednisolone (from 100 to 20 mg/day over the course of 5 days). No patient followed an antibody induction regimen with T cell–depleting agents.
CMV antibodies were detected with the enzyme-linked immunosorbent assay technique (Vidas, BioMerieux, Marcy-l'Étoile, France). The quantitative estimation of CMV DNA was performed with a real-time polymerase chain reaction technique (Sacace Biotechnologies, Como, Italy). CMV DNA titers were reported as negative, as positive but <500 copies/mL, or as a quantitative value if a level ≥ 500 copies/mL was detected. Tissue-invasive CMV was diagnosed on the basis of characteristic viral inclusions on biopsy specimens.
CMV Prophylaxis, Treatment, and Follow-Up
All recipients and donors underwent preoperative CMV IgG testing. Recipients negative for CMV IgG before transplantation were excluded from the study. Recipients positive for CMV IgG before transplantation received no prophylaxis for CMV.
A predefined protocol was followed for patients undergoing liver transplantation (Fig. 1). All recipients underwent CMV DNA estimation 3 weeks after transplantation. No CMV prophylaxis was given to any CMV IgG–positive recipients.
Any patients who were positive for CMV DNA 3 weeks after transplantation underwent weekly CMV DNA assays until 2 consecutive samples were negative. In patients with a rise in titers (versus 3-week values) and with unexplained febrile illness, graft dysfunction, cytopenia, or diarrhea, treatment with valganciclovir at a dose of 450 mg twice a day was administered after other causes of the symptoms were ruled out (the dose was adjusted according to creatinine clearance measured by 24-hour urinary collection: 450 mg daily for 25-50 mL/minute and 450 mg on alternate days for <25 mL/minute).
Patients who were negative for CMV DNA 3 weeks after transplantation were prospectively followed for the following symptoms: unexplained febrile illness, graft dysfunction, cytopenia, and diarrhea. Patients who developed any of these symptoms underwent repeat CMV DNA testing and were treated with valganciclovir at a dose of 450 mg twice a day if the CMV DNA findings were positive (the dose was adjusted according to creatinine clearance as described previously).
Patients who were treated for ACR with high-dose methylprednisolone did not receive any prophylactic valganciclovir. However, they were monitored for CMV DNA weekly for 4 weeks while they were receiving steroids and afterward. Any rise in CMV DNA titers was treated with valganciclovir.
Patients treated with valganciclovir were monitored for CMV DNA for 2 weeks. Therapy was stopped once 2 consecutive CMV DNA samples were negative.
The statistical analysis was performed with SPSS 20 (SPSS, Inc., Chicago, IL). Categorical variables were compared with the χ2 test or Fisher's exact test. The correlation between CMV reactivation and the administration of a steroid pulse was assessed with the χ2 test. A multivariate analysis was performed with a Cox regression analysis for factors predicting survival; ACR requiring high-dose steroids and CMV DNA positivity were used as time-dependent covariates. The actuarial probability of survival for CMV DNA–positive patients versus CMV DNA–negative patients was calculated with the Kaplan-Meier method, and comparisons were made with a log-rank test. A P value less than 0.05 was considered statistically significant.
Between September 2006 and August 2010, we performed 338 liver transplants. The baseline demographic data for these patients are presented in Table 1. Four recipients who were CMV IgG–negative were excluded (2 patients were R−/D+ and received CMV prophylaxis with valganciclovir for 3 months; 2 patients were recipient-negative/donor-negative and received no prophylaxis). Twenty-eight recipients died within 1 month of transplantation and hence were excluded from the final analysis. None of these patients died of CMV disease. Two of these 28 patients were positive for CMV DNA just before death. The causes of death are outlined in Table 2. Three hundred six patients who were R+/D+ were included; these patients were treated according to Fig. 1. The 1-year actuarial survival rate for our series was 85% (standard error = 2.2%). Three hundred thirty-six of the 338 donors were CMV IgG–positive.
|Cause of Death||Patients (n)||CMV DNA|
|Hepatic artery thrombosis||2||Positive in 1 patient just before death|
|Hepatic vein block||1||Negative|
|Cardiac (myocardial infarction, left ventricular failure)||4||Negative in all|
|Neurological (intracranial bleeding, infarct, raised intracranial pressure)||4||Negative in all|
|Necrotizing fasciitis||2||Negative in all|
|Small-for-size syndrome||2||Negative in all|
|Renal failure||2||Negative in all|
|Primary nonfunction||2||Negative in all|
|Antibody-mediated rejection||1||Positive just before death|
|Bile leak: sepsis||2||Negative in all|
|Fungal septicemia||3||Negative in all|
|Bacterial sepsis||3||Negative in all|
CMV Reactivation and Disease
Two-hundred sixty-six of the 306 R+/D+ patients were CMV DNA–negative 3 weeks after transplantation. Forty patients (13%) had detectable CMV DNA 3 weeks after transplantation. Eight patients developed CMV syndrome, and 1 patient developed CMV tissue-invasive disease involving the gastrointestinal tract (immunohistochemistry for a colonic biopsy sample showed CMV positivity). Overall, CMV syndrome or disease occurred in 9 of the 306 patients (2.9%).
Thirty patients had detectable CMV DNA at 3 weeks, but there were fewer than 500 copies/mL. Two of these 30 patients developed CMV syndrome (both developed bone marrow depression with fever) and were found to be CMV DNA–positive on follow-up. Ten patients had CMV DNA levels ≥ 500 copies/mL 3 weeks after transplantation (median = 21,650 copies/mL, range = 1040-62,750 copies/mL; Table 3). Six of the 10 patients with CMV DNA levels ≥ 500 copies/mL developed CMV syndrome/disease and received treatment (5 for cytopenia and 1 for diarrhea). The remaining 4 patients showed spontaneous clearance of CMV on follow-up weekly screening. In patients who developed CMV disease, the median time to disease after transplantation was 17 weeks (range = 5-27 weeks); the CMV DNA level at the time of disease varied from 45,000 to 496,000 copies/mL (Table 4). One patient was preemptively started on valganciclovir because he had a rise in CMV DNA titers while he was receiving high-dose methylprednisolone for rejection. One of the 266 patients who were CMV DNA–negative at 3 weeks developed CMV syndrome [she was CMV DNA–positive (496,000 copies/mL) with leukopenia after she was treated for rejection with methylprednisolone].
|Patient Number||CMV DNA Level at 3 Weeks (Copies/mL)||Symptom||Evidence of Disease||Treatment||Duration (Days)|
|1||25,322||Cytopenia||Other causes excluded||Valganciclovir||30|
|2||32,019||Cytopenia||Other causes excluded||Valganciclovir||30|
|3||14,500||Cytopenia||Inclusion bodies in bone marrow||Valganciclovir||30|
|4||19,300||Cytopenia||Other causes excluded||Valganciclovir||45|
|8||62,750||Diarrhea||Inclusion bodies on rectal biopsy||Valganciclovir||60|
|9||42,450||Cytopenia||Inclusion bodies in bone marrow||Valganciclovir||45|
|Patient Number*||CMV DNA Level at 3 Weeks (Copies/mL)||CMV DNA Level at Time of Disease (Copies/mL)||Time to Disease After Transplantation (Weeks)||Symptom|
|10||24,000||Not applicable||Not applicable||Preemptive treatment (steroid bolus for rejection)|
Rejection and CMV Reactivation
All patients received triple immunosuppression (a calcineurin inhibitor, mycophenolate, and steroids). Fifty-two of the 306 patients (17%) were treated with high-dose methylprednisolone. Five of the 10 patients with CMV DNA levels ≥ 500 copies/mL received methylprednisolone for ACR. One patient received preemptive treatment because he had a rise in CMV DNA titers while he was receiving methylprednisolone for ACR. Nine of the 30 patients who were CMV DNA–positive but had levels < 500 copies/mL received methylprednisolone for ACR. Only 1 of these patients developed CMV syndrome with high CMV DNA levels. Thirty-eight of the 266 CMV DNA–negative patients received methylprednisolone; after receiving methylprednisolone for ACR, 1 patient developed CMV viremia and tissue-invasive disease. There was a significant correlation between high-dose methylprednisolone administration and CMV reactivation (P = 0.003, χ2 test) as well as CMV syndrome/disease (P = 0.002).
Overall, 10 patients received treatment for CMV disease (9 for actual CMV disease and 1 as preemptive treatment). For all patients, CMV DNA was cleared, and the signs and symptoms of CMV syndrome/disease were resolved. All patients who received valganciclovir were monitored for CMV DNA for 2 weeks. Therapy was stopped once 2 consecutive CMV DNA results were negative. The median duration of therapy for the treated group was 30 days (range = 30-60 days). The median follow-up for these patients was 28 months (range = 1-58 months). Our cohort experienced no treatment failures with valganciclovir, and no patient required treatment with foscarnet or other antivirals.
Overall, 61 of the 338 patients died (survival = 82%) during follow-up (median = 28 months). Twenty-eight patients died within 1 month. Two of these 28 patients were CMV DNA–positive; the other 26 patients were CMV DNA–negative. Among those surviving more than 1 month (n = 306), only 1 CMV DNA–positive patient died during follow-up (recurrence of hepatocellular carcinoma). There was no mortality due to CMV in our cohort of patients.
A multivariate analysis was performed with a Cox regression analysis to predict the factors affecting survival. The following were included in the analysis: age, sex, baseline MELD and Child-Pugh scores, CMV DNA status, CMV disease, high-dose steroids for ACR, biliary complications, and sepsis. The administration of high-dose steroids (before or after 3 weeks after transplantation) and the CMV DNA status (positivity at or after 3 weeks) were used as time-dependent variables in the Cox modeling. Only sepsis (P < 0.001) and the preoperative MELD score (P = 0.01) were shown to independently predict survival (Table 5). There was no statistically significant difference in survival between CMV DNA–positive patients and CMV DNA–negative patients (P = 0.68, log-rank test; Fig. 2).
|Factor||Patients (n)||Odds Ratio (95% Confidence Interval)||P Value|
|MELD score||—||1.05 (1.01-1.10)||0.01|
|ACR (before 3 weeks)||34||0.62 (0.26-1.42)||0.26|
|ACR (at or after 3 weeks)||18||1.04 (0.22-4.74)||0.96|
|Biliary complications||55||0.89 (0.38-2.05)||0.96|
|CMV infection at 3 weeks||40||0.54 (0.21-1.37)||0.20|
|CMV infection after 3 weeks||15||5.37 (0.59-48.59)||0.13|
The incidence of CMV reactivation is lower for R+/D+ liver transplant recipients versus R−/D+ liver transplant recipients. Available data are sparse for the proportions of CMV DNA–positive patients developing disease or spontaneously clearing their infections. Another controversy is the duration of prophylaxis/treatment; policies vary from center to center, with most programs administering treatment for 3 months.11 Also, data suggest that CMV disease may occur in these patients later after the discontinuation of CMV prophylaxis.11, 12 A meta-analysis of published randomized controlled trials on the prevention of CMV disease by Kalil et al.7 demonstrated that both universal prophylaxis and preemptive strategies reduced the incidence of CMV disease and allograft rejection. However, this analysis included all possible recipient/donor combinations. The importance of universal prophylaxis versus preemptive strategies was shown for patients with the highest risk of disease (R−/D+ patients and patients using induction antibodies). We have reported prospective data for a wait-and-watch strategy used in a large cohort of R+/D+ patients who underwent liver transplantation.
In our study, patients who had CMV DNA levels ≥ 500 copies/mL at 3 weeks underwent weekly CMV DNA surveillance until the findings were negative or the disease developed. Patients who were CMV DNA–negative or had <500 copies/mL at 3 weeks did not undergo CMV DNA surveillance. Instead, they were monitored for symptoms and tested for CMV DNA if they were symptomatic. Patients received treatment only if they were CMV DNA–positive and their symptoms were attributed to CMV disease (all other causes were ruled out).
The overall rate of CMV DNA positivity (ie, CMV reactivation) in our R+/D+ patients was 13% (40/306) at a median follow-up of 28 months. Previously published trials have reported prevalence rates varying from 15% to 50%.6-8 Our cohort had a prevalence of CMV reactivation lower than that previously reported. This can be explained by the select population used for our cohort (majority of recipients and donors were CMV IgG–positive). In our population, an induction regimen with T cell–depleting agents was not used; the patients were given lower dose immunosuppression because in general they were malnourished and their general condition was poor before transplantation. This may also have contributed to the lower rates of reactivation. Nine of the 40 CMV DNA–positive patients developed CMV syndrome (n = 8) or tissue-invasive disease (n = 1). One patient was treated preemptively because of a rise in CMV DNA titers after treatment for ACR. The incidence of asymptomatic CMV reactivation in this cohort was 10% (31/306) during the median follow-up period of 28 months. CMV syndrome/disease was seen in only 2.9% of the patients (9/306). Those who developed CMV disease responded to valganciclovir treatment. The median duration of treatment was only 30 days because most of the patients were CMV DNA–negative within 15 days of treatment. These patients were followed for a median of 28 months (range = 1-58 months). Thirty-one of the 40 patients who had CMV viremia 3 weeks after transplantation did not develop the disease during a long follow-up period. Importantly, none of the patients died because of CMV reactivation/disease. There was only 1 death among the patients who developed CMV disease, and this was due to the recurrence of hepatocellular carcinoma.
Fifty-two patients received high-dose methylprednisolone for ACR. Thirty-eight of these patients were CMV DNA–negative at 3 weeks; only 1 of these 38 patients developed CMV disease. The other 14 patients who received methylprednisolone were positive for CMV DNA at 3 weeks. Five of these 14 developed CMV syndrome/disease. There was a significant correlation between the use of methylprednisolone for ACR and CMV positivity as well as the development of CMV disease. Although the rate of rejection seemed to be higher for patients with CMV viremia (14 of the 40 CMV DNA–positive patients versus 38 of the 266 CMV DNA–negative patients), the treatment of rejection with methylprednisolone may have been the cause of CMV viremia because steroid administration predated the development of CMV DNA positivity. We systematically monitored the patients who were CMV DNA–positive until negative DNA findings were achieved or the patients developed disease. In the final analysis, which used both high-dose steroid treatment for ACR and CMV DNA positivity as time-dependent variables, neither predicted survival. In most cases, the administration of steroids predated the development of CMV disease. Although we cannot definitely conclude that steroids for ACR led to CMV disease in our cohort, the data point toward that conclusion. This issue could be better addressed in a subsequent prospective trial. Therefore, the use of high-dose methylprednisolone was a risk factor for the development of CMV reactivation as well as disease in our cohort.
CMV infection at or after 3 weeks was not predictive of survival in univariate and multivariate analyses (P = 0.20, 0.13 respectively. Table 5). The only factors predictive of survival were evidence of sepsis (a positive bacterial or fungal culture from blood, urine, or drain fluids in the clinical setting of infection) and a higher preoperative MELD score.
Late-onset CMV disease has been reported to be a problem after the use of valganciclovir for CMV prophylaxis.13, 14 None of our patients developed late-onset CMV disease (we did not monitor for asymptomatic CMV reactivation after treatment). CMV prophylaxis and an R−/D+ status have been identified as risk factors for late-onset CMV disease.12, 14, 15 Because our population was R+/D+ and did not receive CMV prophylaxis, it may have been protected against late-onset disease. The incidence of tissue-invasive disease in our cohort was very low; only 1 patient developed diarrhea (CMV according to histopathology).
Another trial in an R+/D+ population by Kim et al.8 from Korea showed somewhat similar results. Although they reported a high CMV infection/reactivation rate (55.7%), the risk of CMV disease was only 5.5%. In their study, CMV infections did not affect survival, but CMV disease was a negative predictor of survival. In contrast, CMV reactivation/disease did not influence survival in our study cohort with the described approach of regular CMV DNA monitoring and treatment for positive patients.
This trial had several limitations. First, we did not monitor the CMV DNA titers in patients who were CMV DNA–negative at 3 weeks unless they developed symptoms. However, this was not reflected in the survival of these patients during long-term follow-up. Second, because we were dealing with almost universal CMV seropositivity in our population (both donors and recipients), our results cannot be generalized to all populations. Nevertheless, the results of this study should stimulate a re-evaluation of CMV prophylaxis strategies in the R+/D+ subgroup of liver transplant recipients. The incidence of CMV reactivation in our cohort was slightly lower than previously reported rates for R+/D+ patients, and during follow-up, only a minority developed disease. The duration of treatment is 60 days as treatment was stopped after 2 consecutive CMV DNA negative at 2 weeks interval. Hence, maximum duration to achieve DNA negative was 45 days.
In conclusion, the following can be stated:
- 1CMV reactivation is common after LRLT (13%), but the disease is rare (2.9%) without prophylaxis in CMV IgG–positive recipients.
- 2CMV IgG–positive recipients may not require ganciclovir/valganciclovir prophylaxis if they are systematically monitored (as our cohort was).
- 3The administration of high-dose methylprednisolone strongly correlates with CMV reactivation as well as disease.
- 4CMV DNA levels should be determined 3 weeks after LRLT; when the level is greater than 500 copies/mL, prophylactic treatment may be indicated if the patient is being treated with high-dose methylprednisolone for rejection.
- 5CMV reactivation and disease did not affect survival in our study cohort (positive recipients) with our center's approach to treatment.
- 13for Valganciclovir Solid Organ Transplant Study Group. Efficacy and safety of valganciclovir vs. oral ganciclovir for prevention of cytomegalovirus disease in solid organ transplant recipients. Am J Transplant 2004; 4: 611-620., , , , , , et al.;