Prevention and treatment of epstein–barr virus‐associated lymphoproliferative disorders in recipients of bone marrow and solid organ transplants

Abstract

Reactivations of the Epstein–Barr virus (EBV), which may progress to EBV-associated lymphoproliferative disorders (EBV-LPD), are a major threat in recipients of allogeneic bone marrow transplants (BMT) and solid organ transplants (SOT). An overview is given of the monitoring and pre-emptive treatment of EBV reactivations and the incidence, prevention and therapy of EBV-LPD. In high-risk BMT recipients, monitoring of EBV viral load in preferably cell-free plasma should be performed once a week until 6 months post transplant. No strict guidelines for frequency and duration of monitoring in SOT recipients can be given, largely due to the variable time period in which post-transplant EBV-LPDs can occur in this patient group. When EBV reactivation is diagnosed, pre-emptive therapy with anti-B-cell monoclonal antibodies (mAbs) is advised. First-line treatment of EBV-LPD consists of withdrawing or decreasing immunosuppression together with the infusion of anti-B-cell mAbs. Second-line treatment options are also reviewed. The link between severe immunosuppression of transplant recipients and increased incidence of lymphoma has long been apparent and the association with the Epstein–Barr virus (EBV) is widely recognized (Nalesnik, 1998). EBV is the prototype of the gamma subfamily of potentially oncogenic herpesviruses. Taxonomists have renamed EBV as human herpesvirus 4 (HHV4). Two EBV types (type 1 and 2) circulate in most populations, of which type 1 is far more common in most populations (Kieff & Rickinson 2001). There are various isolates of type 1 and 2 EBV. Persistent infection with more than one EBV isolate is not unusual, particularly for immunocompromised patients (Gratama et al, 1992). In vitro, efficient EBV infection of cells is restricted to mature human B lymphocytes. This results in a latent infection in 10% of cells which subsequently proliferate as immortalized lymphoblastoid cell lines (LCLs). Latently infected B cells can be induced to become permissive for lytic viral replication, while viral replication occurs spontaneously in some (Kieff & Rickinson, 2001). The presence of latent virus in an infected cell can be readily detected using antibodies to any of the eight different virus proteins that are characteristically expressed in LCLs. These viral proteins include six nuclear proteins (EBNAs) and two integral membrane proteins (LMPs). Two small non-polyadenylated RNAs (EBERs) and highly spliced BAMH1 A rightward frame (BARF) transcripts are also expressed in LCLs. This type of latency is termed latency III (Rowe et al, 1998; Kieff & Rickinson, 2001). EBV lymphoproliferative lesions are considered to result from proliferating latently infected B cells, expressing latency type III genes, in the absence of EBV specific cytotoxic CD8+ T-cell surveillance (Rowe et al, 1998; Cohen, 2000). However, also more restricted patterns of EBV latency are observed in EBV lymphoproliferative disorders (EBV-LPD) and cases with latency type I (only expressing EBNA1, BARF transcripts and EBERs) have been described (Rea et al, 1994). In addition to the expression of latent EBV genes, viral gene products associated with replicative or lytic infection have been detected in LPDs (Rea et al, 1994; Montone et al, 1996; Rowe et al, 1998). At the moment it is unclear whether EBV replication or lytic infection is of significance in the pathogenesis of EBV-LPD. However, the detection of cell-free EBV-DNA and the high sensitivity and specificity of this test in diagnosing EBV-LPD suggests that lytic EBV infection might be more than a bystander in EBV-LPD. An excellent review of the four different histo-pathological classifications of post-transplant LPDs was published by Nalesnik (1998). Most LPDs are of B-cell origin, although T-cell LPDs sometimes (12%) occur in recipients following solid organ transplantation (SOT) (Leblond et al, 1998). van Gorp et al (1994) reported on three SOT patients with EBV-negative T-cell LPDs. A literature search performed by these authors resulted in 22 transplant (SOT: n = 19) recipients diagnosed with T-cell LPD. A summary of these patients was given and in only five of them an association with EBV was established. Most of these T-cell LPDs were occurring late (> 1 year post transplant), while prognosis was variable. Leblond et al (1998) diagnosed 34 cases with LPD after SOT. Four of 34 LPDs were of T-cell origin and three of these four were EBV negative. The other 30 cases were of B-cell origin and eight of them were EBV negative. EBV-negative LPDs more often occurred late after transplantation (> 2 years), while survival time after diagnosis of LPD was significantly shorter compared with patients with EBV-associated LPDs. All EBV-negative B-cell LPDs were monomorphic, meeting the criteria of diffuse large B-cell lymphoma according to the Revised European–American Lymphoma classification. The findings of Leblond were largely supported by other reports (Dotti et al, 2000; Nelson et al, 2000). No data are available for T-cell LPDs or EBV-negative LPDs after bone marrow transplantation (BMT). Some studies were undertaken to analyse whether EBV-LPD is derived from donor or host lymphoid tissue. In both BMT and SOT recipients post-transplant lymphomas of recipient origin as well as donor origin were found (BMT: Shapiro et al, 1988; Zutter et al, 1988; Schafer et al, 2001; SOT: Armes et al, 1994; Larson et al, 1996; Cheung et al, 1998). The origin of the EBV strain infecting these lymphoma (B) cells is unknown. Gratama et al (1988) showed that the conditioning regimen pre-BMT and/or graft-versus-host disease (GVHD) were able to eliminate the EBV strain of the host. In this study, one patient became infected with a strain indistinguishable from the virus isolated from her husband and another with the donor strain. In two EBV-seronegative and two EBV-seropositive SOT recipients with EBV-LPD, donor strains and non-donor strains, respectively, were identified (Haque et al, 1996). At this time it is unknown how often EBV strains of donor origin cause re-infection in BMT recipients or primo-infections in seronegative SOT recipients. Oral transmission or transmission through transfusion of blood products might be other possibilities, although after BMT irradiated blood products are used that may prevent transfusion-related transmission. EBV reactivations or EBV primo-infections in severely immunocompromised patients may result in the development of EBV-LPD, which is associated with a mortality of 80% (Armitage et al, 1991; Gross et al, 1999; van Esser et al, 2001a).The reported incidence of EBV-LPD varies, but is generally higher in recipients of solid organ transplants (2–8%; Armitage et al, 1991; Leblond et al, 1995; Swinnen et al, 1995; Rogers et al, 1998; Le Meur et al, 1998; Ben-Ari et al, 1999) compared with BMT recipients. In SOT recipients EBV-LPDs usually develop during the first year post transplant; however, they continue to occur thereafter (Armitage et al, 1991; Riddler et al, 1994; Walker et al, 1995; Fontan et al, 1998; Nalesnik, 1998; Ben-Ari et al, 1999). Primary EBV infection in EBV-seronegative SOT recipients, the type of transplanted allograft, cytomegalovirus (CMV) serostatus mismatch (R–/D+), CMV disease and the type (T-cell antibodies) and intensity of immunosuppression are important risk factors for the development of EBV-LPD in SOT recipients (Swinnen et al, 1990; Walker et al, 1995; Manez et al, 1997; Preitsakis & Keay, 2001). Buda et al (2000) showed that in heart transplant recipients hepatitis C virus infection is probably also a risk factor. Primary EBV infection and EBV-LPD is of greater concern in paediatric SOT recipients: in children with small-intestine transplants the incidence of EBV-LPD was 32% (Finn et al, 1998). After BMT the overall cumulative incidence is 1% in 10 years, with most EBV-LPDs occurring within the first 6 months post transplant (Curtis et al, 1999). Four major risk factors for early EBV-LPD (< 1 year) after BMT have been identified (Curtis et al, 1999): (1) T-cell depletion (TCD) using monoclonal antibodies (mAbs) directed at T cells or T and natural killer (NK) cells or TCD using E-rosetting; (2) use of unrelated or ≥ 2 human leucocyte antigen (HLA) mismatched related donors; (3) use of anti-thymocyte globulin (ATG) for prophylaxis or treatment of acute GVHD (aGVHD); and (4) treatment of aGVHD with anti-CD3 mAbs. In patients with three or more risk factors the incidence of EBV-LPD was 22%. Other studies reported EBV-LPD incidences in recipients of matched unrelated donor (MUD) grafts from 4·3% to 24% (Shapiro et al, 1988; Gerritsen et al, 1996; Micallef et al, 1998), in recipients of matched related donor (MRD) grafts from 0 to 0·7% (Shapiro et al, 1988; Zutter et al, 1988; Gerritsen et al, 1996; Micallef et al, 1998) and in recipients of unrelated umbilical cord transplants of 2% (Barker et al, 2001). Since 1994 many studies have been performed to analyse the value of EBV-DNA detection in diagnosing EBV-LPD or other EBV-associated diseases. EBV-DNA detection by polymerase chain reaction (PCR) techniques can be performed in peripheral blood mononuclear cells (PBMC), whole blood or cell-free plasma. Results of studies performed among SOT and BMT recipients are summarized in Tables I and II. In 5 of 13 studies measuring viral load (VL) in PBMC the sensitivity for diagnosing EBV-LPD was less than 100% (Riddler et al, 1994; Rooney et al, 1995a; George et al, 1998; Lucas et al, 1998a,b; Kimura et al, 1999; Baldanti et al, 2000; Hoshino et al, 2001; Stevens et al, 2001; Wagner et al, 2001a,b; Gartner et al, 2002; Sirvent-von Bueltzingsloewen et al, 2002). VL detection in cell-free plasma seems to be more accurate: in five out of six studies 100% sensitivity was obtained (Fontan et al, 1998; Beck et al, 1999; Limaye et al, 1999; Ohga et al, 2001; van Esser et al, 2001a; Wagner et al, 2001a). Limaye et al (1999) report the only patient with a negative PCR result in cell-free plasma who was diagnosed with EVB-LPD. However, the only manifestation of EBV-LPD in this patient was a skin nodule. Specificity varied from 73% to 100% and was generally higher when VL detection was performed in cell-free plasma. Wagner et al (2001a) performed real-time quantitative EBV-DNA detection in PBMC as well as cell-free plasma in recipients of renal transplants and healthy volunteers. Sensitivity of both methods was 100%, while specificity of EBV-DNA detection in PBMC was 89% and in cell-free plasma 100%. When remission of EBV-LPD was accomplished, VL was more effectively cleared in plasma compared with PBMC, which might suggest that cell-free EBV DNA better reflects response to therapy. In contradiction, Stevens et al (2001) were not able to detect EBV-DNA in serum of recipients of lung transplants with EBV-LPD, while 68% of all samples from these six patients tested positive in the PBMC fraction. However, as this is very different from all other reports, these results might be doubted. It has to be stressed that studies summarized in Tables I and II are hard to compare as different viral load detection techniques were used. Furthermore, most were retrospectively performed in selected patients with and without EBV-LPD, while some were prospectively undertaken. Despite this drawback, in the majority of cases EBV viral load was increased in patients with EBV-LPD. Overall, according to sensitivity and specificity, cell-free EBV-DNA detection seems to be the most accurate technique to predict the presence or development of EBV-LPD. An increase in EBV VL often preceded the development of EBV-LPD in BMT recipients by several weeks (Rooney et al, 1995a; Lucas et al, 1998a; Beck et al, 1999; Sirvent-von Bueltzingsloewen et al, 2002). In SOT recipients the time period for EBV-DNA detection prior to the development of EBV-LPD was more variable and ranged from 0 to > 10 months (Riddler et al, 1994; Stevens et al, 2001). van Esser et al (2001b) monitored plasma VL in 14 BMT recipients with EBV-LPD. In patients with response to treatment the VL decreased at least 50% within 72 h after treatment was started. Patients with progressive disease showed an increase in viral load. VL measurements might therefore also be used to monitor response to therapy. EBV-DNA detection by PCR techniques has also proven to be useful in some other EBV-related diseases such as infectious mononucleosis (Yamamoto et al, 1995; Fontan et al, 1998; Kimura et al, 1999; Zingg et al, 1999; Meerbach et al, 2001; Wagner et al, 2001a), chronic active EBV disorder (CAEBV) (Kanegan et al, 1999; Kimura et al, 1999), nasopharyngeal carcinoma (Lo et al, 1999, 2000; Lin et al, 2001) and human immunodeficiency virus (HIV)-associated central nervous system lymphomas (de Luca et al, 1995; Antorini et al, 1999; Portolani et al, 1999). van Esser et al (2002) performed a prospective study in recipients of TCD-BMT. EBV-DNA in cell-free plasma was monitored weekly in 49 patients. Pre-emptive therapy, consisting of a single infusion of rituximab (anti-CD20 monoclonal antibody) was given to patients with a VL ≥ 1000 copies/ml. Seventeen patients showed EBV reactivation of which 15 received pre-emptive therapy. Only one progressed to EBV-LPD, responding completely after two infusions of rituximab and donor lymphocyte infusion (DLI). In two patients EBV-LPD and EBV VL ≥ 1000 copies/ml was diagnosed on the same day. These patients achieved complete remission (CR) after two rituximab infusions. In a historical control group of 85 recipients of TCD-BMT 26 patients showed EBV reactivations, of which 10 developed EBV-LPD (38%). In the prospective study, 3 of 17 patients with VL ≥ 1000 copies/ml developed EBV-LPD (18%). Mortality in the historical group was 80% compared with 0% in the prospective study. This study highlighted the importance of monitoring high-risk patients and the effectiveness of pre-emptive therapy with anti-CD20 therapy. Currently no other study has been published that prospectively analyses the value of pre-emptive therapy to prevent EBV-LPD, apart from two small studies in which three and five patients were treated pre-emptively with rituximab and EBV-specific cytotoxic T cells (EBVs-CTL), respectively, for rising EBV-VL. One of five and none of three patients progressed to EBV-LPD (Gustafsson et al, 2000; Gruhn et al, 2001). Development of EBV-LPD is strongly associated with T-cell depletion of donor marrow. The risk for EBV-LPD varied according to the techniques used for T-cell depletion, being lowest (< 2%) when the Campath-1 or counterflow elutriation methods were used which, in contrast to T cell-specific mAbs, removed both T and B cells (Gross et al, 1998; Hale & Waldmann, 1998; Curtis et al, 1999). Cavazzana et al (1998) observed that none of 19 patients receiving transplants from a partially matched related donor (PMRD) developed EBV-LPD when ex vivo T- and B-cell depletion was performed, whereas 7 out of 19 historical controls developed EBV-LPD when only T-cell depletion was performed. One other study showed that B-cell depletion might be of benefit for decreasing the incidence of EBV-LPD (Liu et al, 1999). When, in our institute, grafts from matched unrelated donors (MUDs) were depleted of both T and B cells, 4 out of 31 patients (13%) developed EBV-LPD. Without B-cell depletion five out of seven patients (71%) receiving BMT from a MUD developed EBV-LPD (Meijer et al, 2002). In summary, B-cell depletion of grafts is efficacious in preventing EBV-LPD in recipients of T cell-depleted grafts from MUDs. The degree of B-cell depletion needed is still uncertain but is clearly closely related to the degree of TCD (Meijer et al, 2002). A mechanism explaining the importance of B-cell depletion might be a reduction of the EBV viral load transmitted by the marrow graft. This is probably more important than a reduction of the amount of B cells itself, as EBV-LPDs do not always consist of donor lymphoid tissue (see Background). Most studies using antiviral drugs have been performed with acyclovir and ganciclovir, which are both nucleoside analogues. The nucleosides first have to be converted to monophosphate by a viral enzyme [which is thymidine kinase (TK) in the case of EBV]. Second and third phosphorylations are performed by cellular kinases. Acyclovir or ganciclovir triphosphate is then preferentially incorporated in DNA by viral DNA polymerase and acts as an obligate chain terminator (Ljungman, 2001). The effectiveness of newer agents such as cidofovir and foscarnet for prevention or treatment of EBV-LPD has not been studied. Cidofovir is a nucleotide analogue of deoxycytidine monophosphate, while foscarnet is a pyrophosphate analogue forming a complex with the pyrophosphate binding site of viral DNA polymerase. Similar to acyclovir and ganciclovir, both drugs are dependent on viral DNA polymerase expression to be functional. TK and viral DNA polymerase are enzymes expressed only during lytic infection, while EBV-LPD is considered to result from latently infected proliferating B cells. Therefore, theoretically, no effect of these drugs can be expected with respect to prevention and treatment of EBV-LPD. However, as is described in the background, some results suggest lytic infection might have a role in the pathogenesis of EBV-LPD (Rea et al, 1994; Montone et al, 1996; Rowe et al, 1998). Prevention. Several studies have shown that treatment with acyclovir results in transient inhibition of EBV shedding in the oropharynx in patients with acute infectious mononucleosis and also in long-term carriers. However, the frequency of circulating EBV-infected B cells remained completely unchanged (Ernberg & Andersson, 1986; Yao et al, 1989; Luxton et al, 1993; Tynell et al, 1996). EBV is also able to transform human lymphocytes despite the presence of 500 μmol/l acyclovir (Sixbey & Pagano, 1985), while a 2-week exposure of B-lymphoblastoid cell lines to 100 μmol/l acyclovir did not prevent release of infectious EBV virus after irradiation to 75 Gray (Keever-Taylor et al, 2001). Many non-randomized studies have been published describing a decrease in incidence of EBV-LPD among SOT and BMT recipients treated prophylactically with acyclovir or ganciclovir; however, an equal amount of studies observed no effect at all of antiviral prophylaxis (Davis, 1998). In paediatric liver transplant recipients (Green et al, 1997), prophylaxis with a short course (2 weeks) of ganciclovir (intravenously) followed by long-term oral high-dose acyclovir resulted in EBV disease in 33% of the recipients compared with 21% in recipients receiving the short course ganciclovir alone. In other randomized trials among SOT recipients using acyclovir or ganciclovir prophylaxis, just a trend towards a lower incidence of EBV-LPD was seen (Davis, 1998). McDiarmid et al (1998) treated high-risk (EBV serostatus recipient/donor: –/+) paediatric liver transplant recipients prophylactically with intravenously administered ganciclovir for at least 100 d. In low-risk patients ganciclovir was replaced by oral acyclovir at discharge. Semiquantitative EBV-DNA monitoring was performed and immunosuppression was decreased when VL increased. The overall incidence of EBV-LPD decreased from 10% (historical) to 5%. This study, however, did not yield any evidence for effectiveness of ganciclovir. The decreased incidence of EBV-LPD might very well be attributed to the EBV-DNA-based reduction of immunosuppressive therapy. Therapy. According to Cohen (1991), acyclovir therapy generally has not been effective for SOT and BMT patients with EBV-LPD. The reduction in immunosuppression that often accompanied acyclovir therapy made it difficult to assess the real effectiveness of acyclovir. Nevertheless, as toxicity of acyclovir therapy is low, treatment with acyclovir is often instituted when EBV-LPD has been diagnosed. Two case reports described the achievement of CR of EBV disease after treatment with ganciclovir or foscarnet (Gruhn et al, 1999; Schneider et al, 2000). Little information is available on the effectiveness of newer antiviral agents regarding prevention or treatment of EBV-LPD. Withdrawal of or decreasing immunosuppressive therapy has proven to be effective in solid organ transplant recipients and is often undertaken as initial strategy (Penn, 1998). However, this is associated with a risk of graft rejection, which can be supported better in renal transplant recipients compared with other SOT recipients. BMT recipients have a far more pronounced immune suppression, which makes withdrawal of immunosuppressive therapy alone usually not sufficient for treating EBV-LPD (Aguilar et al, 1999). Surgical removal or radiotherapy has been effective in patients with localized disease. Survival in SOT recipients with localized EBV-LPD treated with surgical resection alone was 74% compared with 31% in all transplant recipients with EBV-LPD (Cohen, 1991). Chemotherapy (CT) is generally considered to be a treatment option when other therapies have failed (Cohen, 1991, 2000), although several case reports/small studies are available demonstrating the effectiveness of chemotherapy (Garrett et al, 1993; Raymond et al, 1995; Balfour et al, 1999; Smets et al, 2000; Watts et al, 2001). Cohen (1991) did not detect any survival advantage for patients treated with chemotherapy. Results of other larger studies are summarized in Table III. The only study in which treatment with chemotherapy resulted in a favourable outcome is that of Fohrer et al (2001). Twenty-seven recipients of SOT with EBV-LPD were treated with chemotherapy consisting of adriamycin, cyclophosphamide, vincristine, bleomycin and steroids. Granulocyte colony-stimulating factor was given and a total of six cycles were scheduled every 2–3 weeks. In 19 patients a CR was observed (70%), of which seven showed an early relapse (within a median time of 3 months). Actuarial survival at 3, 5 and 10 years was 72%, 66% and 49% respectively. Leblond et al (2001) and Dotti et al (2001) found, after univariate analysis, that treatment of EBV-LPD with CT was an adverse risk factor for overall survival in SOT recipients. Several case reports have been published showing the effectiveness of interferon-alpha (IFN) in the treatment of patients with EBV-LPD after SOT and BMT (summarized by Faro, 1998). In total, 14 SOT recipients and 4 BMT recipients with EBV-LPD received IFN, of which 12 obtained CR. Davis et al (1998) showed that 8 of 14 recipients of SOT with EBV-LPD obtained CR after treatment with IFN. Patients were treated daily (3 × 106 U/m2) for at least 3 weeks and treatment was continued for 6–9 months in responders. Gross et al (1999) describe 26 BMT recipients with EBV-LPD. Thirteen patients received therapy for EBV-LPD of whom only two patients responded. Both these patients were treated with IFN. It should be noted that all patients described in the varying studies received additional therapies. Therefore, it remains unclear whether IFN might be an effective treatment approach for EBV-LPD. Results of anti-cytokine (anti-interleukin 6) therapy in 12 SOT recipients were promising showing CR in 5 of 12 patients with EBV-LPD, partial remission (PR) in 3 of 12 and stable disease in one. Data were preliminary and larger studies have to be performed to confirm these results (Durandy, 2001). T-cell immunotherapy is able to control EBV-LPD in recipients of BMT (O'Reilly et al, 1997). Data on 18 patients with EBV-LPD who were treated with non-specific donor T-lymphocyte infusions (DLI) have been reported (O'Reilly et al, 1997). In 16 of 18 patients eradication of EBV-LPD was accomplished. Ten of 18 patients survived in sustained CR, while three died from GVHD and one from progressive EBV-LPD. This response rate is rather favourable to the data from Lucas et al (1998a), who observed complete response in 4 of 13 patients, while a similar proportion experienced GVHD and only 2 of 13 patients survived. A major side-effect of DLI is GVHD. Bordignon et al (1995) and Bonini et al (1997) treated eight patients with relapse or EBV-LPD with donor T lymphocytes, which were transduced with the herpes simplex virus thymidine kinase (HSV-tk) suicide gene. Three patients developed GVHD that was successfully treated with ganciclovir (CR in two, PR in one). This approach, however, is still experimental. A strategy to limit the risk of GVHD is the administration of EBV-specific cytotoxic T lymphocytes (EBVs-CTL). Rooney et al (1995b) treated 10 BMT recipients of MUD/PMRD grafts with EBVs-CTLs, three of whom had evidence of uncontrolled EBV reactivation. In all, VL fell to normal levels and symptoms disappeared. In a subsequent study (Rooney et al, 1998), 39 BMT recipients of MUD/PMRD grafts received prophylactic EBVs-CTLs. None developed EBV-LPD in contrast to 7 of 61 controls not receiving prophylactic therapy. Acute GVHD did not develop in any patient receiving EBVs-CTLs. Gustafsson et al (2000) described nine BMT recipients, of whom five showed a rapidly rising EBV VL. These patients were pre-emptively treated with EBVs-CTLs, only one progressed to fatal EBV-LPD. This patient received CTLs lacking an EBV-specific component. Altogether, the use of donor-derived (HLA-matched) EBV-specific CTLs seems to be very effective; however, it is limited by the long time periods required for the generation of these cells. Furthermore, generating these CTLs for every transplant recipient prior to the development of EBV-LPD is very expensive. Therefore, this technique will not be available in every transplantation centre. Another drawback was highlighted by Gottschalk et al (2001), whereby an EBV deletion mutant was associated with fatal lymphoproliferative disease unresponsive to therapy with EBVs-CTLs. Donor-derived CTLs are generally not used for prevention or treatment of EBV-LPD in SOT recipients, as the donor is not usually available and donor and recipient generally are not HLA-matched (Haque & Crawford, 1999). However, two case reports have been published in which SOT recipients with EBV-LPD were treated successfully with DLI from an HLA-identical sibling donor (Emanuel et al, 1997) and with EBVs-CTLs from a partially HLA-matched unrelated blood donor (Haque et al, 2001). Several studies described the development and effectiveness of autologous EBVs-CTLs (Nalesnik et al, 1997; Haque et al, 1998; Khanna et al, 1999; Savoldo et al, 2001; Comoli et al, 2002). As is the case in BMT, autologous EBVs-CTLs have to be prepared for all SOT recipients prior to the development of EBV-LPD, which (again) is time-consuming and expensive. Therefore, the use of EBVs-CTLs from partially HLA-matched unrelated blood donors creates new possibilities (Haque et al, 2001). Several case reports have been published demonstrating the effectiveness of anti-B-cell therapy for treatment of EBV-LPD in BMT and SOT recipients (Lazarovits et al, 1994; Cook et al, 1999; McGuirk et al, 1999; Kuehnle et al, 2000; Zompi et al, 2000; Wagner et al, 2001b; Zilz et al, 2001). Fischer et al (1991) and Benkerrou et al (1998) (see Table IV) treated 58 SOT and BMT recipients with EBV-LPD with anti-CD21 plus CD24 antibodies. CR was seen in 61%, while 5-year overall survival (OS) was 46% compared with 29% in historical controls. However, in recipients of BMT, the 5-year OS was only 35%. Milpied et al (2000) treated 32 SOT and BMT recipients with EBV-LPD with rituximab. After a median follow-up of only 8 months, 1-year OS was 73%. Response (CR and PR) was 65% for SOT recipients and 83% for BMT recipients. Rituximab was also given to 12 paediatric BMT recipients (Faye et al, 2001) and to seven adult SOT recipients (Kentos et al, 2001) with EBV-LPD showing CR rates of 66% and 71% respectively. Thus, anti-B-cell therapy seems to be very promising, especially when it is started pre-emptively in high-risk patients with increasing EBV VL (van Esser et al, 2002). It should be noted that EBV-LPD in the central nervous system (CNS) generally does not respond to anti-B-cell mAbs because of lack of penetration in the CNS (Benkerrou et al, 1998). Recently, one patient with EBV-LPD and CNS localization was treated with rituximab and cidofovir, which resulted in a CR. Plasma and cerebrospinal fluid EBV VL became negative during treatment (Hanel et al, 2001). Marshall et al (2000) used HLA class I tetramers complexed with multiple latent and lytic EBV peptides to characterize the dynamics of EBVs T cells in BMT recipients. In recipients of unmanipulated allogeneic BMT from related donors it was demonstrated that expansion of EBVs T-cell populations occurred even in the presence of immunosuppressive therapy. The amount of EBVs T cells correlated with EBV VL in PBMC. In contrast, after TCD or unrelated cord blood transplantation, EBVs T cells were undetectable, even in the presence of EBV viraemia. Curtis et al (1999) already showed that TCD and the use of unrelated donor grafts were risk factors for EBV-LPD. Nevertheless, the use of these tetramers might enable us to detect transplant recipients without circulating EBVs T cells. These patients have a high risk of developing EBV-LPD and should be monitored intensively to institute pre-emptive therapy (anti-CD20 mAbs, EBVs-CTLs) when EBV VL is rising. In high-risk BMT recipients monitoring of EBV VL in preferably cell-free plasma should be performed once a week until 6 months post transplant. No strict guidelines for frequency and duration of monitoring in SOT recipients can be given, largely due to the variable time period in which post-transplant EBV-LPDs can occur in this patient group. However, in high-risk SOT recipients monitoring may be performed fortnightly or at every outpatient visit until 1 year post transplant. When EBV reactivation is diagnosed, pre-emptive therapy with anti-B-cell mAbs is advised in BMT and SOT recipients. In BMT recipients receiving T cell-depleted grafts from unrelated donors, additional B-cell depletion can reduce the incidence of EBV-LPD dramatically. Treatment of EBV-LPD should start with the withdrawal or decrease of immunosuppression together with anti-B-cell mAbs. Donor lymphocyte infusion should be reserved for BMT recipients not responding to anti-B-cell therapy or with CNS localization. SOT recipients with CNS localization might receive additional radiotherapy and/or chemotherapy as well. The efficacy of antiviral therapy in preventing or treating EBV-LPD, if there is any, is very low. Chemotherapy or IFN might be given to SOT recipients when other treatment options have failed or are not available. Localized disease in this patient group can be cured with surgery or radiotherapy. When available, EBVs-CTLs from HLA-identical donors or autologous EBVs-CTLs can be used as (pre-emptive) treatment of EBV-LPD in BMT or SOT recipients respectively. Further studies will be necessary to evaluate the safety and effectiveness of EBVs-CTLs obtained from (partially) HLA-matched related and unrelated blood donors in both BMT and SOT recipients, which will make this approach more accessible.