Persistent detection of Epstein-Barr virus DNA after pediatric liver transplantation: Unclear risks and uncertain responses

Abstract

Epstein-Barr virus (EBV) is second to cytomegalovirus (CMV) as the most frequent and important viral pathogen affecting pediatric recipients of liver transplantation. Children undergoing liver transplantation frequently develop a primary EBV infection, which places them at marked risk of developing EBV disease, including posttransplant lymphoproliferative disease (PTLD).1, 2 Clinically important EBV infections and PTLD occur 5 to 7 times more frequently in patients who are seronegative prior to transplantation in comparison with patients who are seropositive at the time of liver transplantation.3, 4 Although many children who seroconvert will not have clinically recognizable disease, high rates of PTLD have been historically reported after pediatric liver transplantation under both cyclosporine-based and tacrolimus-based immune suppression.4, 5 CMV, cytomegalovirus; EBV, Epstein-Barr virus; PCR, polymerase chain reaction; PTLD, posttransplant lymphoproliferative disease. The high prevalence of EBV disease and its associated morbidity and mortality have resulted in efforts to identify children at risk for this complication. Over the last 10 years, many centers have incorporated the longitudinal measurement of EBV load in the peripheral blood with DNA amplification techniques as a strategy to identify both patients at risk for EBV disease and the presence of EBV disease.2, 6 The ability to identify patients prior to the development of clinical disease has led to interest in developing preemptive strategies to prevent EBV-related complications. Potential preemptive interventions in response to elevated EBV loads have included reduction of immune suppression, antiviral therapy, or a combination of the two.5, 7 Although there is consensus agreement on the importance of reduction of immunosuppression for such strategies, the therapeutic value of preemptive use of antiviral therapies (for example, ganciclovir or acyclovir) remains unproven. In this issue of Liver Transplantation, Hierro and colleagues8 explore the potential value of oral valganciclovir in reducing the risk of developing EBV-related PTLD in pediatric liver transplant recipients. In their retrospective study, these investigators report on their experience with 47 children who underwent liver transplantation at their center and in whose blood EBV DNA was detected by polymerase chain reaction (PCR). Forty-two of the 47 children were treated with valganciclovir in response to these results, beginning their treatment a median of 35 months after transplantation. Because the level of EBV DNA was not specified, it appears that these results were determined with a qualitative assay, which did not differentiate between those who may have had low or high EBV loads. Given the fact that the valganciclovir was started nearly 3 years post-transplant, it is not surprising that EBV infection had been previously recognized in most of these children, including some with a proven prior history of PTLD. This may explain at least in part the fact that the vast majority of these children had previously been treated with antiviral therapy earlier in their posttransplant course. Thirty-four of the patients were considered asymptomatic at the time of initiation of valganciclovir, whereas 13 were felt to be symptomatic, including 2 with PTLD. Treatment was initially planned to be given for 1 month. However, the protocol was eventually changed to provide treatment until EBV DNA was undetectable in the blood. The authors point out that immune suppression was not decreased in these patients unless they showed evidence of PTLD or systemic symptoms attributable to EBV. Virologic: What happened to the EBV load? Clinical: Did the patient develop symptoms of EBV? Safety: Were there side effects attributable to the long-term treatment with valganciclovir in this population? From the virologic perspective, the authors note that 20 of the 42 treated patients (48%) achieved undetectable EBV DNA after a median of 8 months of therapy. From the clinical perspective, they report no new cases of PTLD. However, 1 of the children diagnosed with PTLD at the time of starting valganciclovir experienced worsening of symptoms. The authors felt that the safety profile of the prolonged course of valganciclovir was excellent. Although the authors felt that their results must be interpreted with caution, they concluded that the results of their trial demonstrated that the use of valganciclovir was “safe, and that the treated patients did not develop PTLD warranting future trials to confirm a positive effect of valganciclovir in the management of EBV infection.” The article by Hierro and colleagues8 raises several important questions associated with the monitoring of EBV loads in children undergoing liver transplantation. As the authors point out, the presence of EBV DNA detected by measurement with PCR “creates anxiety in families and physicians due to the risk of PTLD.” In considering whether or not the use of prolonged therapy with valganciclovir will be of benefit to such patients, one must first determine if these patients are truly at risk of a complication. We and others have reported on the relatively frequent finding of persistent EBV loads in children undergoing liver transplantation.9-11 Much of this previous work has focused on determining the risk associated with persistent detection of EBV. In contrast to the current report, we have used quantitative PCR assays to allow us to attempt to differentiate the relative risk of development of EBV disease and PTLD on the basis of the height of the persistent load. The use of the quantitative result is an attempt to add specificity to EBV DNA measurement, as detection of at least some level of EBV load is exceedingly common in pediatric transplant recipients, particularly those who have experienced primary infection after transplant. Although we believe that there is reason for concern for those patients with persistently high EBV loads,12 our experience does not suggest that there is significant risk of disease progression in children with persistent detection of low EBV loads. Accordingly, it is not clear what proportion, if any, of the 42 children treated with valganciclovir were truly at risk of developing EBV disease. It is also important to discuss the appropriate timing of EBV surveillance. As noted earlier, the development of primary EBV infection after liver transplantation is associated with the highest risk of progression to EBV disease and PTLD. EBV seroconversion typically occurs in the first 3 to 6 months following transplantation. Accordingly, it is not surprising that the bulk of EBV disease and PTLD is seen within the first year following liver transplantation.13 Because of this, we use intensive EBV load surveillance early after transplant in an effort to identify primary EBV infection. The frequency of EBV monitoring is diminished over time as the risk of primary EBV infection lessens. We typically obtain EBV surveillance beyond the first year after liver transplantation relatively infrequently, often in response to clinical events or in follow-up to previous positive results. This practice is supported by the fact that PTLD occurs quite infrequently beyond the first year after transplant.13 Accordingly, the relative cost benefit of random screening is diminished in this later time period when disease is uncommon, and the interpretation of the result may be more confusing. Although not all of the children in this series with EBV DNA identified in their blood were necessarily at risk of developing EBV disease or PTLD, it is possible that some of them were. This raises a third important question related to this report: what is the appropriate response to persistently elevated EBV load detection? As Hierro et al.8 point out, commonly considered interventions include reduction of immune suppression and/or antiviral therapy. Although antiviral therapy is a perfect preemptive treatment for detection of CMV DNA, the pathogenesis of EBV infection in transplant recipients differs from that of the other herpesviridae, and this makes the use of these agents less certain. Although CMV-related disease is clearly attributable to lytic viral replication (a process amenable to inhibition with ganciclovir), EBV-related disease is more complicated. EBV-related pathogenesis in the transplant recipient is driven more by uncontrolled expansion of EBV-transformed B cells and less by lytic viral replication.14 The expansion occurs through human DNA polymerases driven by EBV as opposed to viral DNA polymerase. The human DNA polymerases are not affected by acyclovir, ganciclovir, or their related prodrugs valacyclovir and valganciclovir. Accordingly, the expansion and accumulation of EBV-transformed B cells (the source of PTLD) are not clearly limited by these antiviral therapies. The lack of impact of antiviral therapies on this expansion is consistent with the apparent lack of a clear temporal relationship between the initiation of valganciclovir and the resolution of viral DNA detection in the current study. In contrast to CMV, for which we would expect to see complete clearance of viral DNA within a month or two of ganciclovir being started, more than half of the patients in this study still had detectable EBV DNA despite the use of valganciclovir for an average of 7 months. This mirrors our own experience in pediatric intestinal transplant recipients, in whom the use of prolonged IV ganciclovir therapy for persistent elevations in EBV load has not reliably resulted in resolution of this condition. Similarly, in the original report by McDiarmid and colleagues5 describing preemptive treatment of rising EBV loads early after pediatric liver transplantation, the majority of patients developed their primary EBV infection and experienced their highest loads while on continuous IV ganciclovir during the first 100 days after transplant.5 Because antiviral therapy is of uncertain value, we recommend reduction of immune suppression to provide a safe and effective response to pediatric liver transplant recipients with persistently high EBV loads who appear to truly be at risk for the development of EBV disease. Evidence in support of the efficacy of the preemptive reduction of immune suppression in response to elevated EBV loads early after transplant (and likely nearer the time of primary EBV infection) is provided by the work of Lee and colleagues.7 These investigators noted a decline in the incidence of PTLD from 16% to 2% in a group of pediatric liver transplant recipients in comparison with historical controls associated with reducing tacrolimus dosing to achieve trough levels of 4 to 6 ng/mL. In contrast to the prior study by McDiarmid et al.5 (who combined reduction of immune suppression with the use of ganciclovir), Lee et al. did not use antiviral therapy in patients with an elevated EBV load. The benefits of this strategy were achieved with minimal risks of rejection. We have employed a strategy similar to that of Lee and colleagues since 2001 in our pediatric liver transplant population with similar outstanding results in the prevention of early EBV disease, including PTLD. Although rapid reduction of immune suppression appears to be an effective preemptive intervention for patients who develop primary EBV infection early after transplant, the management of patients with persistently elevated EBV loads beyond the first 3 to 6 months after transplant may require a different strategy. We recently reported our experience with persistently high EBV load carriers in our pediatric liver transplant population.10 We identified 36 pediatric liver transplant recipients who maintained EBV viral loads of >16,000 genome copies/mL of whole blood (or >200 genome copies/105 peripheral blood lymphocytes) in at least 50% of load measurements over a minimum 6-month period of monitoring. Although some of these patients had previously experienced symptomatic EBV disease (including a prior history of PTLD), they had maintained their persistent load state in the absence of clinical symptoms for 6 months to meet inclusion criteria for this study. Antiviral therapy was not used. Only 1 of the 36 high load carriers developed PTLD. This child appeared to have a unique predisposition to EBV disease as he experienced multiple episodes of PTLD during the 41 months of maintaining chronic high EBV loads. We believe that the minimal rate of development of PTLD in these children was a result of our management of these patients. This management strategy has evolved as we have gained experience with EBV load monitoring in our pediatric liver transplant recipients. Early in our experience, the presence of an elevated EBV load prompted aggressive reductions (and in some cases even withdrawal) of immunosuppression. However, the development of severe rejection in at least 1 of these patients prompted a more careful approach to the management of the asymptomatic EBV high load carrier. We have developed a relatively standardized approach to the presence of the high load carrier state that aims to gradually reduce immune suppression while carefully monitoring for any evidence of breakthrough rejection. Our practice has been to maintain low levels of immune suppression for prolonged periods. Many of our pediatric liver transplant recipients who have been identified as high load carrier states are currently maintained with trough tacrolimus levels < 5 ng/mL as their targeted level of immune suppression. In general, the approach of gradually reducing immunosuppression to achieve this goal has avoided the development of intercurrent rejection. Although the use of this approach does not appear to result in the timely resolution of the carrier state in most children, we believe that the presence of lower levels of immune suppression likely protects against progression to EBV disease despite the persistence of the carrier state. In contrast to the experience in our pediatric liver transplant recipients, 45% of pediatric heart transplant recipients at our institution meeting the same definition of persistently high EBV load carriage subsequently developed PTLD.12 We believe that the difference in outcomes between these 2 patient populations is likely attributable to a willingness to gradually and progressively reduce immune suppression over time in our liver transplant recipients to a greater extent than is possible in our pediatric heart transplant population. This difference in management reflects the inadequacy of current methods of rejection surveillance for heart transplant recipients and the fact that although rejection is generally asymptomatic in the liver transplant population, acute rejection in heart transplant recipients may present with dysrhythmia, cardiogenic shock, and sudden death. The growing number of centers that routinely perform surveillance EBV load measurement has led to increasing recognition that EBV loads are frequently detectable for prolonged periods of time in these patients. Recognition of this fact requires clinicians caring for these patients to understand the potential risks associated with the phenomenon and to develop strategies to minimize these complications. Although the study of Hierro et al.8 attempts to address the latter, the limitations of this study (including the use of a qualitative EBV PCR, the absence of any control group, and its retrospective nature) limit the conclusions that can be drawn from this work. Given the relatively short follow-up of patients in their study and the fact that less than half of the patients treated with valganciclovir in their cohort experienced clearance of EBV DNA, we agree with these investigators that additional trials of valganciclovir are needed to establish a beneficial effect of this drug.