Background. Epstein-Barr virus (EBV) reactivation is a frequent event after allogeneic stem cell transplantation and may progress to life-threatening lymphoproliferative disease (EBV-LPD) in the absence of adequate EBV-specific T cell immunity. Quantification of EBV DNA load in asymptomatic individuals who are at risk is a useful (although not entirely predictive) indicator of progression to EBV-LPD and guide for preemptive treatment with CD20 antibodies.
Methods. With the aim of improving the identification of patients at risk, we retrospectively analyzed, within a cohort of 25 consecutive allogeneic stem cell transplant recipients at risk for EBV-LPD, the pattern of T cell reconstitution during EBV reactivation in all preemptively treated patients (8 patients).
Results. In 6 of 8 cases, a significant T cell reconstitution (i.e., a CD3+ T cell count of >300 cells/µL) was documented during EBV reactivation, which included an expansion of EBV-specific memory T cells, as shown by human leukocyte antigen class I tetramer analysis. Additional evidence for the antiviral potential of this T cell reconstitution was obtained prospectively from a cohort of 14 consecutive allogeneic stem cell transplant recipients at risk for EBV-LPD. EBV reactivation occurred in 3 patients. Preemptive treatment was successfully withheld for 2 of these patients in light of concurrent (EBV-specific) T cell recovery.
Conclusion. We conclude that analysis of the level of (EBV-specific) T cell reconstitution during EBV reactivation is an important second parameter, in addition to quantification of EBV DNA load, that will be instrumental in a more accurate definition of patients at risk for EBV-LPD who, given their immunoincompetence, will be most certainly dependent on preemptive interventions.
Epstein-Barr virus (EBV) reactivation is a frequent event following receipt of an allogeneic stem cell transplant (alloSCT) that may progress to life-threatening EBV-related lymphoproliferative disease (EBV-LPD). Risk factors for EBV reactivation and subsequent EBV-LPD include the use of unrelated or mismatched family donors, T cell depletion in vitro, antithymocyte globulin, and nonmyeloablative stem cell transplants [1]. Clinical symptoms are frequently lacking in the early stages of EBV reactivation and are often only recognized in the later stages if they coincide with progressive EBV-LPD. The introduction of real-time PCR has provided a powerful tool to monitor EBV reactivation in alloSCT recipients who are still asymptomatic and to predict an increased risk of developing EBV-LPD [2, 3]. Both prophylactic and therapeutic strategies have been reported to either prevent or treat EBV-LPD, including B cell depletion of the graft [4] and restoration of T cellular immunity by means of donor lymphocyte infusion or administration of EBV-specific cytotoxic T lymphocytes [5, 6]. Recently, evidence has been provided that EBV DNA load—guided preemptive treatment with B cell—depleting CD20 monoclonal antibodies (rituximab) is effective in preventing EBV-LPD in alloSCT recipients at high risk [2].
Although effective in preventing EBV-LPD, preemptive treatment on the basis of EBV DNA load as a single parameter definitely results in unnecessary treatment in a significant number of patients [2]. Because EBV-LPD only occurs in the absence of adequate T cell immunity, we retrospectively analyzed (EBV-specific) T cell reconstitution during EBV reactivation in pediatric alloSCT recipients who had received preemptive treatment with CD20 antibodies (rituximab). On the basis of these results and supported by the outcome in a prospectively monitored cohort, we propose that simultaneous and analysis of both EBV DNA load and T cell recovery will improve the identification of patients at high risk for EBV-LPD.
Patients and transplantation. The retrospective study cohort consisted of 50 consecutive children with a hematological malignancy or genetic disease who received an alloSCT in the Leiden University Medical Center (Leiden, The Netherlands) from January 2002 through September 2003. Pretreatment of the patients was according to the disease-specific protocols of the relevant working parties of the European Group for Blood and Marrow Transplantation. Twenty-five children were regarded as at risk for the development of EBV-LPD, because they received a graft from an EBV-seropositive, unrelated (n = 19) or mismatched family (n = 6) donor and antithymocyte globulin during the conditioning (Imtix [Sangstat]; total dose,10 mg/kg). In 11 of 25 cases, in vitro manipulation of the graft was performed by CD34 selection (CliniMACS; Miltenyi) or T/B cell depletion by immunorosetting. The prospective cohort consisted of 52 consecutive pediatric alloSCT recipients who received their transplants in the Leiden University Medical Center from October 2003 through May 2005. Of these patients, 14 were regarded as at risk for EBV-LPD, because they received a graft from an EBV-seropositive, unrelated (n = 12) or mismatched family (n = 2) donor and antithymocyte globulin during the conditioning. In 3 of 14 cases, in vitro manipulation of the graft was performed by CD34 selection (CliniMACS) or T/B cell depletion by immunorosetting. Graft-versus-host disease prophylaxis consisted of cyclosporin A and a short course of methotrexate for T cell—repleted alloSCT recipients. No antiviral prophylaxis was given, and all patients received intravenous immunoglobulin supplementation during 3–12 months after receipt of transplant, depending on the level of immune recovery.
Monitoring and treatment of EBV reactivation. The real-time quantitative PCR for detection of EBV DNA in plasma was performed as described elsewhere [7] and was performed on an Icycler IQ Multi-Color Real Time PCR Detection system (BioRad). EBV DNA load was prospectively monitored weekly during the first 12 weeks after stem cell transplantation and subsequently every other week until 5–6 months after stem cell transplantation or longer, until EBV DNA became undetectable. EBV reactivation was defined as occurring in patients with a PCR positive for EBV DNA in plasma. Patients with an EBV DNA load >1000 copies/mL at 2 consecutive time points were considered to be at high risk to proceed to EBV-LPD [2, 3] and received a single infusion of rituximab (375 mg/m2). A second infusion was administered when the reduction in viral load was <1 order of magnitude (i.e., <1 log10 copies/mL) within the first week. EBV DNA load was monitored twice a week until at least 2 negative test results were obtained.
Immune reconstitution and tetramer analysis. Analysis of immune reconstitution, including tetramer staining, was performed by multicolor flow cytometry, as described elsewhere [8, 9]. The staining of lymphocytes was performed by incubating the cells with a pre-titrated concentration of tetramer at 37°C for 15 min. The cells were then stained for surface markers by incubation on ice for 30 min with allophycocyanin-conjugated anti-human CD8 antibody (Immunotech) and fluorescein isothiocyanate—conjugated anti-human CD4 antibody (BDBiosciences). For further phenotypic analysis, tetramer-stained cells were incubated with a fluorescein isothiocyanate—labeled CD45RO (DAKO) or an unlabeled anti-human CCR7 monoclonal antibody (BD Pharmingen) followed by detection with a goat anti-mouse IgM fluorescein isothiocyanate—labeled secondary antibody (Southern Biotechnology Associates). Flow cytometry was performed on a Facs Calibur (Becton Dickinson Immunocytometry Systems) using CellQuest software (Becton Dickinson). The EBV tetramers used in this study are shown in table 1 and were generated by the Institute for Cancer Studies (Birmingham, United Kingdom) and Sanguin Research (Amsterdam, The Netherlands). Written informed consent was obtained in all cases for immunological monitoring and cryopreservation of blood mononuclear cells. Approval was obtained from the local institutional review board.
Epstein-Barr virus (EBV)—specific T cell reconstitution during EBV reactivation for patients who received an allogeneic stem cell transplant (alloSCT). The lower limit for reliable quantification of the EBV DNA load by real-time PCR is 2.7 log10 copies/mL. PBMCs were stained with phycoerythrin-conjugated HLA-A2.1 tetramers containing either the GLC, YVL, or TLD peptide, HLA-B8 tetramers containing the RAK or FLR peptide, HLA-A11 tetramer containing the AVF peptide, HLA-B7 tetramer containing the RPP peptide, or HLA-A3 tetramer containing the RLR peptide. The cells were also stained with anti-CD8 and anti-CD4 antibodies. The numbers of EBV tetramer—positive T cells are shown as percentage of the total CD8+ T cells. Patients 1, 3, 5, and 6 received preemptive treatment with rituximab (see Patients and Methods), whereas treatment was withheld for the prospectively monitored patients 9 and 10 because of concurrent (EBV-specific) T cell reconstitution. AVF, epitope sequence AVFDRKSDAK; FLR, epitope sequence FLRGRAYGL; GLC, epitope sequence GLCTLVAML; RAK, epitope sequence RAKFKQLL; RLR, epitope sequence RLRAEAQVK; RPP, epitope sequence RPPIFIRRL; TLD, epitope sequence TLDYKPLSV; YVL, epitope sequence YVLDHLIVV.
General and EBV-specific T cell immune reconstitution during EBV reactivation. In a cohort of 50 consecutive alloSCT recipients, 25 of these patients were regarded as at risk for EBV reactivation according to the criteria defined in Patients and Methods. EBV reactivation was documented in 8 (32%) of these patients (patients 1–8; table 2). Median time to the first EBV reactivation was 59 days (range, 26–111 days) after stem cell transplantation, and preemptive therapy with rituximab was started at a median time of 67 days (range, 28–121 days) after receipt of an alloSCT (table 2). In 2 patients (patients 3 and 5) a second infusion was administered because of a lack of response after the first infusion. Retrospective analysis of general and EBV-specific T cell reconstitution was performed. In 2 of 8 cases (patients 1 and 2), clearance of EBV DNA was seen in the absence of significant T cell recovery (i.e., CD3+ T cell count <20 cells/µL; figure 1 and data not shown). In the remaining 6 patients, an increase in CD3+ T cell count to at least 300 cells/µL was documented during the initial 2–3 weeks of EBV reactivation, which was associated with a sustained decrease in EBV DNA load (figure 1 and data not shown). This rapid T cell reconstitution included both CD4+ cells and CD8+ T cells, with a predominance of CD8+ T cells. The magnitude of CD8+ T cell reconstitution varied between 7-fold and 50-fold during the period of EBV reactivation (figure 1 and data not shown). The large majority (>95%) of the CD8+ T cells were CD45RO+/CCR7- compatible, with an effector memory phenotype (data not shown). A significant and rapid increase in EBV-specific CD8+ T cell count was demonstrated during EBV reactivation in all patients in whom the HLA class I genotype and the availability of cryopreserved lymphocytes allowed us to perform an analysis with HLA class I tetramers (4 of 6 patients; figure 1 and data not shown). The peak value of the individual peptide-specific CD8+ T cell populations represented 0.1%–12% of the total CD8+ T cell pool for these patients and included responses to both latent and lytic EBV epitopes. Apart from biological variation, the variable number of available HLA class I tetramers per individual HLA genotype (0–3) most probably explains the quantitative differences between patients in our study. Because of these technical limitations, it remains difficult to determine the relative contribution of T cell responses against single lytic and latent epitopes in individual patients. Taking into consideration that the tetramer-based results are an underestimation of the overall EBV-specific CD8+ T cell response, the cumulative EBV-specific CD8+ T cell repertoire expanded from <0.5 cells/µL to 4–450 cells/µL during the EBV reactivation. The significant interindividual and peptide-dependent differences are similar to what has been reported for patients with infectious mononucleosis [11, 17].
Characteristics of patients who underwent allogeneic stem cell transplantation (alloSCT) with Epstein-Barr virus (EBV) reactivation.
The kinetics of these virus-induced memory T cell responses strongly suggest that these T cells might have been able to control EBV reactivation and prevent progression to EBV-LPD without rituximab. This is consistent with the observation by van Esser et al. [18] that, within a similar population of adult alloSCT recipients at risk and without preemptive treatment, only a minority of the patients with EBV reactivation finally progress to EBV-LPD. The observation regarding patient 7, for whom EBV DNA load had decreased prior to infusion of rituximab but in the presence of T cell reconstitution, further supports the functional relevance of T cell recovery.
Simultaneous analysis of viral load and T cell reconstitution. Further evidence that assessment of the level of T cell reconstitution at the time of EBV reactivation is of relevance for clinical decision making was obtained prospectively. In a cohort of 52 consecutive alloSCT recipients, 14 of these patients were regarded as at risk for EBV-LPD, according to the criteria defined in Patients and Methods. EBV reactivation was documented for 3 (21%) of these patients (patients 9–11; table 2). On the basis of a significant and rapid (EBV-specific) T cell expansion during the initial phase of EBV reactivation, rituximab was withheld for 2 of these 3 patients (patients 9–10). EBV reactivation was controlled and cleared in a fashion similar to that observed in the aforementioned patients who received additional treatment with rituximab (figure 1). Because T cell recovery was absent (CD3+ T cell count <100 cells/µL during the period of EBV reactivation) for the third patient (patient 11), rituximab was administered to this patient as described for the first 8 patients and resulted in clearance of the EBV reactivation. Altogether, our results strongly suggest that the need for preemptive intervention seems to be limited to those patients who lack an expansion of (EBV-specific) memory T cells during the initial phase of EBV reactivation. In our experience, although this threshold is arbitrary, an increase in CD3+ T cell count to at least 300 cells/µL during the initial phase of EBV reactivation appears to be a practically useful threshold for determining the need for preemptive intervention. In all patients except for patients 4 and 8, CD3+ T cell reconstitution was found to be exclusively of donor origin during the period of EBV reactivation; for patients 4 and 8, CD3+ T cells were of mixed and autologous origin, respectively (data not shown). Notably, and in contrast to the other patients, the latter 2 patients received transplants following a nonmyeloablative conditioning regimen.
Abrogation of cyclosporin A treatment has been suggested as an effective preemptive intervention for alloSCT recipients to control EBV reactivation [19]. Unfortunately, no information was provided in this study on the characteristics of (EBV-specific) T cell reconstitution before and after this intervention. In some of our patients, T cell reconstitution was documented while receiving continued cyclosporin A treatment, whereas in other patients, cyclosporin A was tapered. Therefore, in the presence of significant memory T cell reconstitution, the additional requirement of abrogating cyclosporin A to control EBV reactivation may be variable and needs to be weighed against the risk of eliciting graft-versus-host disease. The additional contribution of acyclovir to the control of EBV reactivation is probably limited and is difficult to assess, because it will not be administered as a single therapeutic intervention. Recently, Wagner et al. [20] provided evidence that prompt treatment on the basis of the combined analysis of viral load and clinical appearance of symptoms may be preferred to preemptive treatment on the basis of viral load only. However, although EBV-LPD was successfully controlled in all cases, 3 of 8 patients required treatment with EBV cytotoxic T lymphocytes. Therefore, in a significant amount of patients, the general feasibility of this approach of management of EBV reactivation may rely on the local availability of clinical-grade EBV cytotoxic T lymphocytess.
The role of virus-specific T cell responses in the control of cytomegalovirus infection in alloSCT recipients is rather similar to what we have reported for EBV infection, and assessment of these responses can be used in clinical management of cytomegalovirus infection. Cytomegalovirus-specific T cells rapidly expand during viremia and are able to prevent the recurrence or persistence of infection [21]. Similar to what we have demonstrated during EBV infection, the peak of cytomegalovirus-specific CD8+ T cell responses parallels the level of viraemia. The absence of such response defines patients at risk, who may benefit from adoptive transfer of virus-specific T cells [22].
Previous studies have provided evidence that the risk for EBV-LPD is correlated with both the general and EBV-specific immune reconstitution after stem cell transplantation [23–26]. Our data indicate that frequent analysis of T cell immune reconstitution at the time of EBV reactivation is, in combination with EBV DNA load, an important second parameter for recognition of alloSCT recipients who are at high risk for EBV-LPD and will thus further improve the identification of patients who will benefit most from preemptive interventions.
Financial support. Dutch Cancer Society (UL 2005–3657 to N.E.A.).
Potential conflicts of interest. All authors: no conflicts.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy