The treatment of EBV-associated post-transplant lymphoproliferative disease (PTLD) poses a considerable challenge. Efforts have been made to define regimens based on combination of the available therapeutic agents, chosen and tailored on a patient-by-patient basis, with the aim of augmenting event-free patient and graft survival. Recently, autologous EBV-specific cytotoxic T-lymphocytes (CTL) have proved effective in enhancing EBV-specific immune responses and reducing viral load in organ transplant recipients with active infection. We investigated the use of a tailored combined approach including autologous EBV-specific CTL for the treatment of EBV-related PTLD developing after pediatric kidney transplantation.
Five patients with disseminated monoclonal (n = 3) or localized polyclonal (n = 2) PTLD unresponsive to reduction of immunosuppression were enrolled. The patients with disseminated PTLD received 4–5 courses of reduced-dosage polychemotherapy, accompanied by rituximab on the first day of each course, while localized disease was removed surgically. At treatment completion, autologous EBV-specific CTL were infused. All patients showed a complete response to treatment, without therapy-related toxicity or rejection, and persist in remission with good renal function at a median follow-up of 31 months. These preliminary results suggest that a combined chemoimmunotherapy regimen including virus-specific T-cells is well tolerated and potentially effective as first-line treatment of EBV-related PTLD.
Post-transplant lymphoproliferative disorders (PTLD) are potentially lethal complications of immunosuppression in solid organ transplant recipients (1,2), mostly associated with the proliferation of EBV-infected B cells, whose expansion in immunocompetent individuals is controlled by cytotoxic T lymphocytes (CTLs) (3). The reported incidence of PTLD after kidney transplantation is estimated to be between 1% and 2% (4,5), with higher rates observed in the presence of intense immunosuppression and primary EBV infection (6,7).
The outcome of PTLD has been correlated to various parameters including clinical presentation, histology and clonality (8,9). Reduction of immune suppression may be sufficient to control the disease in early-state, low-risk patients (10,11), by allowing the recovery of host immune surveillance against the virus, and it is therefore considered the initial treatment of choice in the setting of SOT. Patients who do not tolerate or do not respond to immunosuppression reduction require more aggressive therapy and have a poorer prognosis, with high mortality rates. Given the lack of controlled trials, and heterogeneity of reported case series, there is as yet no consensus on a standard approach to the treatment of established PTLD unresponsive to immunosuppression reduction (12). Anatomically localized PTLD can be eradicated by surgical resection or radiation (2). Regression of lymphoproliferation has been anecdotally reported following treatment with antiviral drugs (1,11) and interferon-α (13); however, some other form of treatment was almost invariably associated, and therefore their therapeutic role remains unclear.
Early results of cytotoxic chemotherapy indicated that aggressive regimens were associated with high response rates, but also with severe treatment-related toxicity and increased susceptibility to infections, with disease-free survival <50% (14). Encouraging results in terms of stable, complete remission rates have been more recently described (15), and a low-dose regimen has proved well tolerated and effective for PTLD in children who failed reduction of immune suppression (16). Since successful treatment of PTLD requires control of EBV-transformed B-cells, a strategy based on anti-B-cell monoclonal antibodies (mAbs) has been employed to reduce the B-cell compartment in patients with PTLD (17). A humanized murine anti-CD20 monoclonal antibody (rituximab) has produced complete remissions in both hematopoietic stem cell transplant and SOT recipients (18,19). In the latter cohort, the reported incidence of complete responses has varied widely, with the overall response rate observed in the largest serie being 46% (20). Recently, it has been suggested that efficacy may augment when rituximab is used as first-line treatment, and with more prolonged treatment duration (21).
Since EBV-related PTLD derive from a disruption of the balance between the virus lytic and latent life cycles and host immune control, an attractive alternative or implementation to a treatment strategy based on abrogation of the B-cell compartment is restoration of EBV-specific immunocompetence by infusion of T cells with EBV-specific activity (19,22). In previous studies, we and others found that infusion of autologous EBV-specific CTLs obtained from patient peripheral blood mononuclear cells (PBMC) recovered at the time of viral reactivation augmented virus-specific immune responses and reduced viral and tumor load in SOT recipients (23,24). Therefore, we added autologous EBV-specific CTL transfer to a treatment regimen tailored according to disease extension and histological subtype, with the aim of improving disease control without additional toxicity.
Patients and Methods
Between July 1998 and November 2002, all pediatric kidney allograft recipients diagnosed with EBV-related PTLD at the Pediatric Nephrology Unit of the G. Gaslini Institute-Genova received autologous EBV-specific CTL infusions in addition to a tailored, multiple treatment approach.
The clinical characteristics of the patients are summarized in (Table 1). All patients received a cadaveric kidney transplant. During the post-transplant course, patient 2 experienced prolonged delayed graft function due to acute tubular necrosis that required the use of anti-thymocyte globulin; patients 1, 2 and 4 developed acute rejection episodes requiring treatment with pulse steroids.
Work-up for suspected PTLD included peripheral blood EBV DNA analysis and computerized tomodensitometry (CT scan) of neck, chest, abdomen and pelvis. Diagnosis was confirmed by histological evaluation.
Patients gave written informed consent at the time of enrolment; the study was conducted according to the institutional guidelines.
EBV-specific CTL reactivation
Autologous EBV-specific CTLs were reactivated and expanded in vitro according to a method previously reported, following GLP standard procedures (24). In detail, EBV-specific CTLs were prepared from fresh or frozen PBMC obtained at the time of diagnosis, plated in 2 mL X-VIVO 20 medium (BioWhittaker, Walkersville, MD) with 2% autologous plasma, at 2 × 106 cells per well and stimulated with irradiated autologous B-lymphoblastoid cell line (LCL) at a responder-to-stimulator (R:S) ratio of 40:1. After 10 days, cultures were restimulated with irradiated autologous B-LCL at an R:S ratio of 4:1. Starting on day 14, 20 U/mL recombinant IL-2 (rIL-2) (Hoffman-La Roche, Basel, Switzerland) were added to the wells, and the cultures were subsequently restimulated weekly with irradiated autologous B-LCL in the presence of rIL-2. Before cryopreservation, T cells were examined for EBV specificity in a standard 51Cr-release assay against a panel of targets including recipient B-LCL and PHA blasts, HLA-mismatched allogeneic B-LCL and PHA blasts, and the LAK-permissive Daudi cell line. CTL lines were also evaluated for immunophenotype, and for sterility. Samples with satisfactory test results were thawed and administered intravenously to patients.
The allograft recipients with localized poli/oligoclonal disease had surgical removal of the tumor. The patients with disseminated monoclonal PTLD received polychemotherapy, accompanied by 1 dose (375 mg/m2) of rituximab on the first day of each course. Chemotherapy consisted of reduced-dosage CHOP (two-third of the conventional dosage) courses, alternated to courses of a BFM-derived block (dexamethasone 3 mg/m2, days 1–6; cyclophosphamide 200 mg/m2, days 1–4; vincristine 1.5 mg/m2, day 1; etoposide 100 mg/m2, days 3–4; cytarabine 150 mg/m2, days 3–4) (25) in two patients.
All allograft recipients received two monthly protocol infusions of autologous EBV-specific CTLs (2 × 107 cells/m2/dose) at completion of treatment regimen. Additional EBV CTL infusions were administered upon evidence of increased EBV DNA blood levels, in the presence of low or undetectable frequency of virus-specific T cells.
Details on treatment modalities for each patient are reported in (Table 2)).
Table 2. PTLD treatment, patient response and outcome
EBV DNA baseline*
EBV DNA at treatment completion*
Patient status (time since beginning of treatment)
*EBV DNA levels in peripheral blood mononuclear cells.
2 × (2×107) + 3 × (2×107)
CsA: 4.1 mg/kg
PDN: 0.08 mg/kg
2 × rCHOP 2 × mBFM
2 × (2×107) + 2 × (2×107)
Sirolimus: 2.8 mg/m2
MMF: 200 mg/m2
PDN: 0.06 mg/kg
2 × (2×107)
CsA: 2.5 mg/kg
PDN: 0.05 mg/kg
5 × rCHOP
2 × (2×107)
CsA: 1.2 mg/kg
PDN: 0.09 mg/kg
3 × rCHOP 2 × mBFM
2 × (2×107)
CsA: 3 mg/kg
PDN: 0.1 mg/kg
Patients were evaluated after two and four courses with CT scan, and at the completion of therapy with positron emission tomography (PET) and/or CT scan. Therapy was discontinued upon evidence of full remission. Thereafter, follow-up was carried out every 3–4 months for the first year, and subsequently progressively extended. Allograft function was tightly monitored with the aim of rejection surveillance in the context of initial reduced immunosuppression and CTL infusion.
EBV DNA levels were monitored by PCR (24) on peripheral blood mononuclear cells or plasma, as appropriate, at monthly intervals in the first year, and thereafter at 3-month intervals.
To evaluate the effects of CTL infusion on the frequency of EBV-specific T cell immunity, IFN-γ secreting lymphocytes were measured by ELISPOT assay on peripheral blood samples collected at baseline, 2–4 weeks after infusion, and subsequently at 3-month intervals, following a method previously described (26). For ELISPOT assay, 96-well multiscreen filter plates (MAIPS 4510, Millipore, Bedford, MA) were coated with 100 μL of primary antibody (IFN-γ, Mabtech, Nacka, Sweden) at 2.5 μg/mL, and incubated overnight at 4°C. PBMCs were thawed and cultured overnight in RPMI-FCS medium before use in the assay, and were then seeded in the absence or in the presence of EBV-LCL or 2μg/mL of a peptide mix containing 15-mer peptides spanning the EBV LMP-2 protein (Jerini, Berlin, Germany). After incubation for 24 h at 37°C, 100 μL of biotinylated secondary antibody (Mabtech, 0.5μg/mL) was added, and plates were then processed according to standard procedure. IFN-γ-producing spots were counted using an ELISPOT reader (Bioline, Torino, Italy). The number of spots per well was calculated after subtraction of assay background, quantitated as average of 24 wells containing only sterile complete medium, and specific background, quantitated as the sum of cytokine spots associated with responders alone or responders plated with DMSO solvent control, as appropriate.
Patients and PTLD characteristics
Between July 1998 and November 2002, five patients were enrolled in our tailored combined treatment protocol. The five patients developed PTLD 5–155 months after transplantation. Four out of the five patients had received additional immunosuppression to their baseline regimen.
Clinical and pathological presentation of PTLD in these patients are detailed in (Table 1). Two patients had disease localized in the tonsils or as skin nodules, while the other three kidney recipients presented with disseminated disease involving Waldeyer's ring, abdominal and retroperitoneal nodes, gastric mucosa, liver and spleen. Allograft localization was present in two patients (cases 2 and 5). Pathological evaluation of tumor cells demonstrated that three PTLD were monoclonal, one had oligoclonal cell populations, while the fifth was polyclonal. All tumors were EBV+ by EBER, and 4/5 were CD20+. PTLD morphology indicated that two patients had polymorphic disease, whereas three tumors had monomorphic histology.
Generation and characteristics of autologous EBV-specific CTL lines
Between 1997 and November 2002, autologous EBV-specific CTL lines were generated at the Laboratory of Transplant Immunology of Policlinico S Matteo-Pavia for solid organ transplant recipients who were diagnosed with EBV-related PTLD or identified as being at risk of developing EBV-related PTLD through the finding of elevated numbers of EBV-DNA genome copies on two or more consecutive samples. We succeeded in generating EBV LCLs for 64 of the 69 graft recipients referred to the laboratory, and proceeded to reactivate autologous EBV-specific CTLs in 41 of the 64 patients whose LCLs had expanded. CTL lines were successfully generated in all 41 patients. The median time for LCL generation was 30 days (range 25–65), while CTL expansion sufficient for infusion requirements was obtained at a median time of 28 days (range 23–45).
In the case of CTL lines reactivated from the five patients included in this study, median expansion at 4 weeks was 10-fold (range 4–16). Phenotypic analysis indicated that the majority of cells were CD8+ (60 ± 17%), with 29 ± 21% CD4+, and variable, albeit low, numbers of CD3+/CD8+/CD56+ and CD56+/CD3− lymphocytes. EBV specificity of the CTL lines was indicated by the fact that all CTL lines showed strong lysis of the autologous EBV-LCL (52 ± 5 mean% lysis at effector:target ratio of 5:1), while little or no reactivity was observed against HLA-mismatched LCLs (8 ± 3 mean% lysis), and allogeneic HLA-mismatched PHA blasts (mean lysis < 1%).
PTLD treatment and response
The patients were initially treated with reduction of immunosuppression, consisting of reduction/discontinuation of the third drug (azathioprine or micophenolate mofetil) and reduction of calcineurin inhibitor, according to the severity of disease. All patients either failed to regress or showed a progression of their initial tumor after 2–4 weeks of observation, which prompted the introduction of additional treatment. Specifically, the two poly/oligoclonal, localized PTLD were removed by surgical excision, whereas the patients with monoclonal disseminated disease were treated with courses of polychemotherapy associated to anti-CD20 monoclonal antibody (immunochemotherapy, ICT) in the case of tumor positivity by histology. In order to reduce chemotherapy-related toxicity, the number and composition of cycles were tailored for each patient on the basis of histology and response to treatment. In an attempt to increase treatment efficacy, without addition of further toxicity, all patients received two doses of autologous EBV-specific CTLs at the completion of treatment, to consolidate remission.
The overall response rate was 100%. All five patients achieved complete remission, and are currently without evidence of disease at a median follow-up of 31 months (range 19–71 months) (Table 2). The three patients with disseminated monoclonal disease achieved CR after a median time of 18 weeks (range 17–24 weeks). Two of the latter patients received protocol CTL infusions while in remission. Patient 5, who presented with multiple PTLD nodules at the graft, liver and spleen, had an excellent response to the five cycles of ICT, with disappearance of kidney and spleen disease, but persisted with a single residual lesion, reduced in diameter compared with baseline, in the liver (Figure 1A, B). To avoid further chemotherapy load, we decided to interrupt ICT and treat the residual disease with the two protocol doses of autologous EBV-specific CTL alone. T-cell therapy was able to clear the tumor, and the patient persists in complete remission 19 months after beginning treatment (Figure 1C).
Therapy was well tolerated by all patients. None of the three patients receiving chemotherapy required erythrocyte or platelet transfusion; patient 4 required hospitalization for an episode of febrile neutropenia. All three patients showed decreased IgG and IgM immunoglobulin levels after treatment with rituximab, and were therefore given replacement therapy with i.v. γ-globulins. No adverse events ascribable to autologous EBV CTL infusion were recorded.
Clinical and immunological follow-up
In the course of treatment, immunosuppression was decreased or discontinued in patients with localized or disseminated disease, respectively. In detail, patients 1 and 3 were maintained on lower levels of CsA and steroids, while no immunosuppression was administered to the patients who were receiving chemotherapy. At completion of PTLD treatment, the former two patients continued with unaltered levels of lowered immunosuppression, while immunosuppressive drugs were gradually reintroduced in the other three allograft recipients starting at day +45, +60 and +1 from treatment completion, respectively. Decisions on time of restart and composition/dosage of IS regimens were dictated by time interval between PTLD onset and transplantation and by clinical considerations regarding the risk of rejection. Details on the IS regimen at the last follow-up are reported in (Table 2)). Good function of the graft was maintained throughout treatment, and none of the patients developed rejection during or after ICT-CTL therapy.
PTLD follow-up in our patients included, in addition to clinical and radiological evaluation, serial monitoring of EBV DNA load and EBV-specific T cell frequency in peripheral blood. The two patients with rebound elevations of viral load in the presence of low or undetectable levels of virus-specific T cells received additional infusions of EBV-CTL. Upon T-cell transfer, the patients showed an increase in the frequency of specific IFN-γ-producing T cells, with consequent marked reduction or clearance of viral load (Figure 2). Since circulating EBV-infected resting memory B cells, responsible for the augment of EBV DNA levels in the peripheral blood, carry only a restricted pattern of EBV antigens, we tested the presence of LMP2-specific lymphocytes both in the original CTL lines and in PBMC before and after the additional CTL infusions. In patient 2, whose CTL line contained a high percentage of LMP2-specific T cells, the number of circulating LMP-2 specific lymphocytes increased considerably (Figure 2A). In the case of patient 1, EBV-specific T cell lines included only a small fraction of LMP-2 specific cells, accounting for 7% of the total EBV-specific population, and the infusion of CTLs did not result in the augment of LMP2-specific activity (Figure 2B). Therefore, cells directed towards other antigens, such as EBV lytic-cycle-specific proteins, may have been responsible for viral load reduction.
Treatment management of PTLD after SOT is a complex issue, since decision making is conditioned by the heterogeneity of pathologies included under the heading of PTLD, PTLD clonality, genetic features and clinical presentation, as well as by patient- and transplant-specific characteristics. Restoration of specific immune competence by reduction of immune suppression is currently considered the front-line treatment for SOT recipients with PTLD. In a recent study, the association of immunosuppression reduction and the antiviral drug acyclovir has allowed to obtain an 82% long-term continuous complete remission in a cohort of 11 renal graft recipients, albeit with a 45% incidence of graft failures (11). Likewise, though excellent disease-free organ and patient survival data were reported in small single-center pilot studies (21), it appears from studies on larger patient cohorts that single-agent treatment with either chemotherapy or anti-B cell mAbs does not result in a satisfying event-free patient and graft survival (15,20), likely due to selection of escape mutants or to excess toxicity. Therefore, efforts are currently being made to define new treatment protocols based on combination of the different available therapeutic agents, chosen and tailored on a patient-by-patient basis according to clinico-pathological classification, disease location and patient/transplant characteristics, with the aim of augmenting efficacy through the synergistic activity of multiple agents, while reducing overall toxicity and treatment-related adverse events. The preliminary results obtained in a pilot trial of chemoimmunotherapy including rituximab and cyclophosphamide/prednisone, with 83% disease-free survival at a median follow-up of 14 months are encouraging, though they need to be confirmed in a controlled trial (27).
Since EBV-specific T-cell deficiency is a prerequisite for the development of PTLD, the use of virus-specific CTL to recover antiviral immune surveillance is an attractive, low-toxicity option to prevent or treat EBV-related PTLD in the post-transplant setting. Thus, there is a rationale for adding this therapeutic agent to the other treatment options. Moreover, our feasibility data indicate that EBV-specific CTLs can be reactivated and expanded from the vast majority of patients after SOT.
Adoptive cellular immunotherapy has proved effective in preventing PTLD and treating a number of patients with established disease after hematopoietic stem cell transplantation (19). The experience with EBV-specific CTL treatment in PTLD after SOT is far more limited (19,22–24,28,29). Analysis of the available data suggest that, notwithstanding the encouraging preliminary results indicating how EBV-specific CTL reactivated from previously seronegative patients who developed a primary infection under immunosuppression could mediate an efficient antiviral response both in vitro and in vivo (23,24), the use of CTL as monotherapy, at least in patients with extended, monoclonal PTLD who failed conventional treatment, may have the same limitations described for rituximab and chemotherapy, namely selection of neoplastic clones with poor expression of EBV antigens, and thus refractory to CTL killing (23,29). However, earlier administration and the simultaneous association of CTL therapy to other forms of treatment may reduce considerably the risk of tumor selection. In our pilot study, T-cell therapy with autologous EBV-specific lymphocytes was given as part of the primary treatment, in combination with rituximab and tailored chemotherapy. The regimen was well tolerated, and associated with minimal toxicity. None of the patients experienced toxic effects, or developed graft rejection following EBV CTL infusion. Moreover, a 100% overall complete response rate was obtained, that resulted in a sustained disease-free and event-free survival ranging between 19 and 71 months. These data are obtained in a small cohort, and are not the result of a controlled study. Therefore, they cannot demonstrate an advantage of this regimen over other approaches, or a specific benefit derived from the use of EBV CTL. However, they compare favorably with the 60% overall survival and 20% graft survival in our historical population (7), and with recent published series (11,20,21,27). In addition, we could demonstrate a direct anti-tumor effect of autologous EBV-specific CTL in patient 5, who presented with a disseminated monoclonal immunoblastic lymphoma that could not be totally eradicated after five chemoimmunotherapy courses, and was rescued by CTL infusion. The reason for the long-lasting complete response in this patient, compared to previous observations of PTLD relapse after a primary complete response following cell therapy (23, personal unpublished observations), may be ascribable to the fact that CTL were infused in the presence of minimal immunosuppression, and could therefore expand sufficiently in vivo. In this view, the combination of low-intensity chemotherapy to a regimen of anti-CD20 and EBV-CTL may be crucial to allow a temporary discontinuation of immunosuppression, thus avoiding the detrimental effects of in vivo immunosuppression on the function and/or persistence of CTL.
Follow-up management of our patients included serial monitoring of EBV DNA load in blood, though its role in patients who have recovered from PTLD is unclear, and the frequent rebounds observed after immunosuppression reintroduction are mostly associated with increased levels of resting B cells carrying the virus rather than of EBV-transformed B lymphoblasts, which are the cells directly involved in PTLD development (29). In order to increase the predictive value of this test, we associated the measurement of virus-specific T lymphocyte frequency. Thus, whenever a patient experienced an increase in viral load in the absence of measurable levels of virus-specific T cells, 'recall' doses of EBV-specific CTL were infused to ensure protection against a possible PTLD relapse. The observation of an increase in circulating virus-specific T cells was considered as a protective response also in the absence of a concomitant decrease in peripheral blood viral load, since residual EBV loads may indicate not cell therapy failure, but rather the persistence of circulating EBV-infected resting memory B cells or lytically infected B cells, usually poorly recognized and killed by conventional EBV-specific CTL lines (24,29). This said, in both patients requiring additional EBV CTL infusions we could observe a complete decline of EBV DNA levels, likely due to the presence of LMP2-specific and/or lytic cycle-specific cells in the infused CTL line.
In our small cohort of SOT recipients, the use of a tailored regimen including virus-specific T cell therapy has proved feasible, safe and effective in treating EBV-related PTLD, while maintaining good graft function. The addition of T cell therapy to other treatment strategies may have resulted in a significant improvement of patient outcome, although the small sample size and use of multiple interventions simultaneously makes the contribution of any single therapeutic agent difficult to assess. In order to confirm the beneficial effect of autologous EBV CTL therapy, a randomized study on a larger cohort of patients is warranted. Additionally, studies focusing on the development of strategies able to activate autologous virus-specific CTL from virus-seronegative patients, and on the identification of novel surrogate markers of PTLD progression, may consent to further optimize T cell infusion protocols.
The authors wish to acknowledge the patients who contributed, the Kidney Transplant Unit in Genova and the Nord Italia Transplant program for their cooperation, and M. Paulli, J.L. Ravetti, C. Gambini for assistance in the characterization of patients.
This is supported in part by grants from the Associazione Italiana Ricerca sul Cancro (AIRC) to PC, FL and RM; grants RFM/02, RFM/03, RFM/04 to PC, FL, RM and FB; grant FP6-Allostem to FL; grant RCR/02 80541 from Policlinico S. Matteo to FB; grant RCR/2003 from Istituto G. Gaslini, grant from Fondo Malattie Renali del Bambino and grant from Fondazione ‘Istituto di Ricerca Virologica O. Bartolomei Corsi’ to FG, FP and GB. SB is the recipient of a grant from Associazione 'Morgan di Gianvittorio'.