Adapted Treatment of Epstein–Barr Virus Infection to Prevent Posttransplant Lymphoproliferative Disorder After Heart Transplantation



Up to 35% of posttransplant lymphoproliferative disorder (PTLD) cases occur within 1 year of transplantation, and over 50% are associated with Epstein–Barr virus (EBV). EBV primary infection and reactivation are PTLD predictive factors, but there is no consensus for their treatment. We conducted a prospective single-center study on 299 consecutive heart-transplant patients treated with the same immunosuppressive regimen and monitored by repetitive EBV viral-load measurements and endomyocardial biopsies to detect graft rejection. Immunosuppression was tapered on EBV reactivation with EBV viral loads >105 copies/mL or primary infection. In the absence of response at 1 month or a viral load >106 copies/mL, patients received one rituximab infusion (375 mg/m2). All patients responded to treatment without increased graft rejection. One primary infection case developed a possible PTLD, which completely responded to diminution of immunosuppression, and one patient, whose EBV load was unevaluable, died of respiratory complications secondary to PTLD. Compared with a historical cohort of 820 patients, PTLD incidence was decreased (p = 0.033) by a per-protocol analysis. This is the largest study on EBV primary infection/reactivation treatment, the first using rituximab following solid organ transplantation to prevent PTLD and the first to demonstrate an acceptable tolerability profile in this setting.


cyclophosphamide, doxorubicin, vincristine and prednisone




computerized tomography


diminution of immunosuppression


Epstein–Barr virus


hematopoietic stem-cell transplantation


mycophenolate mofetil


polymerase chain reaction


positron emission tomography


primary infection


posttransplant lymphoproliferative disorder




standard incidence ratio


World Health Organization


Posttransplant lymphoproliferative disorder (PTLD) comprises a spectrum of lymphoid conditions associated with the use of potent immunosuppressive drugs after solid organ transplantation or hematopoietic stem-cell transplantation (HSCT) [1, 2]. PTLD is the second most frequent cause of neoplasm after cutaneous carcinoma in patients who have received a heart transplant, representing 10% of all posttransplant cancers [3]. PTLD is usually associated with poor prognosis, with a 5-year overall survival rate of 20% [3]; but recent data show that median overall survival can exceed 6 years when immunochemotherapy is used [4]. Most PTLDs are B cell neoplasms, and up to 35% occur within the first year following transplantation (early PTLD), with more than 50% of cases associated with Epstein–Barr virus (EBV) [4]. Use of immunosuppressive drugs to reduce the risk of graft rejection down-regulates physiologic control of EBV-infected lymphocytes by T cells and can lead to uninhibited B cell proliferation [5, 6]. EBV primary infection and EBV reactivation are well-defined predictive factors for PTLD development [7-10]. All of the currently used assays for predicting risk of developing PTLD are based on estimation of EBV viral load using polymerase chain reaction (PCR) amplification of EBV DNA isolated from peripheral blood. It has already been shown that transplant recipients with PTLD frequently have a significantly higher EBV viral load compared with transplant recipients without PTLD, that a high EBV viral load is associated with PTLD development [8, 11] and that early PTLD is almost always preceded by elevated EBV titers [7, 9, 10]. It has been suggested that diminution of immunosuppression (DIS) guided by EBV viral load may lead to a lower PTLD risk as a result of restoration of T cell control [10, 12].

Currently, there is no known modality for reliably preventing EBV infection. Development of antiviral drugs, such as aciclovir and ganciclovir, raised hopes that these drugs would be effective for preventing PTLD, but despite encouraging results in small studies [13, 14], there are no data from studies with larger numbers of patients.

Immunotherapy-based approaches for EBV therapy based on rituximab and cytotoxic T cells have been proposed. Rituximab is a chimeric monoclonal anti-CD20 antibody [15] that has been shown to deplete circulating B cells, including those infected with EBV. Published studies demonstrate rituximab use for treatment of EBV primary infection or reactivation after stem-cell transplantation but not, to our knowledge, after solid organ transplantation [16]. Infusion of EBV-specific cytotoxic T cells improves clinical symptoms and normalizes EBV immune serology [17]. To date, there is no published large prospective analysis on treatment of EBV reactivation for prevention of PTLD that includes data on safety and graft-rejection risk. We report the first large prospective study of adapted treatment of EBV primary infection and reactivation by DIS with or without rituximab and PTLD prevention in adult heart-transplant patients conducted in a single center in France.


Patients and immunosuppression

From January 2004 to December 2009, 299 patients who were alive 1 month after receiving a heart transplant in the Pitié Salpêtrière Hospital, Paris, France, were prospectively followed. Cytomegalovirus (CMV) and EBV viral loads were measured at each hospital follow-up visit, 15–20 times the first year. Patients received immunosuppressive treatment consisting of anti-lymphocyte globulin (Thymoglobulin®, [Genzyme Polyclonals, Lyon, France]; 1.5 mg/kg/day from January 2004 to June 2007, and 1–1.25 mg/kg/day thereafter) for 3–5 postoperative days, cyclosporin (adapted for serum level of 250 ng/mL), mycophenolate mofetil (MMF) (290/299 patients, 1500 mg/day) and prednisone (20 mg/day). This research was approved by the relevant institutional review boards or ethics committees and all human participants gave written informed consent.

EBV and CMV DNA loads

EBV and CMV DNA loads were measured using real-time quantitative PCR assays. DNA was isolated from 400 µL of EDTA-treated whole blood, using the MagNA Pure Compact extractor (Roche Diagnostics, Meylan, France) with the MagNA Pure Compact Nucleic Acid Isolation kit I (Roche Diagnostics), according to the manufacturer's instructions, and eluted in 100 µL of elution buffer. EBV load was assessed by real-time PCR using primers for the EBV thymidine kinase gene, BXLF1, as previously described [18, 19] using the LightCycler 1.5 (Roche Diagnostics). Reference DNA samples were obtained from Raji cells and from serial dilutions of a recombinant plasmid containing one copy of the target DNA fragment. Raji cell DNA had been previously quantified by PCR. This method had been validated according to the French EBV Working Group of the Agence Nationale de Recherche sur le Sida.

CMV real-time PCR was performed using an ABI Prism 7500 instrument (Applied Biosystems, Courtaboeuf, France), using TaqMan® technology (Invitrogen, Carlsbad, CA), as previously described [20]. The sensitivity threshold for both EBV and CMV reactions was 25 DNA copies/mL of whole blood.

EBV and CMV primary infection were defined as a positive viral load in a serologically negative patient. Reactivations were defined as a CMV viral load greater than 104 copies/mL of blood in CMV-positive patients and an EBV viral load greater than 105 copies/mL of blood in EBV-positive recipients [21].

Graft-rejection survey and treatment

Endomyocardial biopsies were performed weekly for the first 2 months following transplantation, then every 2 weeks for Months 3 and 4, every 3 weeks for Months 5 and 6, once monthly for Months 7–12, then every 4 months thereafter. A total of 20 biopsies per patient were taken over the first 12 months after transplantation.

Graft rejections were treated according to International Society for Heart and Lung Transplantation guidelines [22]. Acute cellular rejections (grade 1R) were treated by optimizing immunosuppression (increasing the target serum level of cyclosporin) with or without corticosteroid bolus (oral prednisone 100 mg/day for 5 days or intravenous (IV) methylprednisolone, 1 g/day for 3 days) depending on the hemodynamic and clinical symptoms. Grade 2R and 3R rejections were treated using anti-lymphocyte serum, depending on the hemodynamic symptoms.

Treatment for CMV primary infection or prophylaxis of reactivation

Patients who developed CMV primary infection or reactivation were treated by stopping MMF and initiating oral valganciclovir or IV ganciclovir, maintained until CMV viral load returned to baseline. In seronegative patients who received a CMV-positive transplant, patients received systemic prophylactic valganciclovir treatment.

Treatment for EBV primary infection or reactivation

Patients who were diagnosed with an EBV primary infection or reactivation received a mandatory second confirmatory assessment 1 week later to assess their infection kinetics. Patients with confirmed primary infection or reactivation were assessed for PTLD by computerized tomography (CT) or positron emission tomography (PET) scan. Patients who were PTLD-free were managed by DIS (by terminating MMF or decreasing cyclosporin dose with or without increasing corticosteroid dose) and were assessed weekly for viral load. During follow-up, if the EBV load was greater than 106 copies/mL or the viral load was stable at 1 month despite immunosuppression modification, a single IV infusion of rituximab (375 mg/m2) was administered. A second infusion of rituximab was administered if the EBV viral load had not markedly decreased 1 week later. A flow diagram showing the algorithm for EBV reactivation or primary infection treatment is shown in Figure 1. This study was approved by the institutional review board at La Pitié Salpêtrière Hospital.

Figure 1.

Epstein–Barr virus (EBV) reactivation or primary infection preemptive treatment algorithm. EVL ≥ 105 but < 106 indicates Epstein–Barr virus viral load (copies/mL). Rituximab administered as one infusion of 375 mg/m2.

PTLD diagnosis and prevention

Patients with EBV primary infection or reactivation and/or clinically relevant symptoms, with suspected PTLD detected by CT or PET scan, were biopsied for further assessment according to World Health Organization (WHO) classification guidelines [23]. In order to detect a potential preventative effect on PTLD incidence, we compared our results with a historical cohort of 820 heart-transplant patients from the same heart-transplant center treated between January 1987 and December 2003, in this cohort, the immunosuppressive regimen consisted of an induction treatment with anti-lymphocyte globulin (thymoglobulin 1.5 mg/kg/day) for 5 postoperative days, and an immunosuppressive prophylactic treatment associating cyclosporin (adapted for serum level of 250 ng/mL), azathioprine (25–50 mg/day) and prednisone (20 mg/day).


Descriptive statistics used median and range for quantitative variables, and numbers and percentages for qualitative variables. Most of the statistical analyses are based upon survival analysis methodology. Censored events are EBV infection (primary infection or reactivation), CMV infection (primary infection or reactivation), first rejection episode and death. Relationships between the censored event times and the covariables were assessed using log-rank tests for qualitative covariables and univariate Cox model for quantitative covariables. EBV infection, CMV reactivation and rituximab administration were also considered as time-dependent covariables for the analysis of the risk factors of the first rejection episode. Their effects were tested using the univariate Cox model. All the tests were two-sided, with a p-value of. 0.05 considered as significant. Data were analyzed using the SAS V9 statistical software (SAS Institute, Inc., Cary, NC).


Between January 2004 and December 2009, 299 patients were assessed: 226 males and 73 females, with a mean age of 50 years (range: 16–72 years); the mean number of viral load measured the first year was 17. Patient characteristics are listed in Table 1. Six patients who were seronegative at the time of transplantation and had an EBV-positive donor were considered to have a very high risk of EBV primary infection. A total of 115 patients were CMV-seronegative at the time of transplantation, and of these, 65 had a CMV-positive donor. Patient treatment is summarized in Figure 2.

Table 1. Patient characteristics
  1. R, serology status of transplant recipient; D, serology status of transplant donor. CMV, cytomegalovirus; EBV, Epstein–Barr virus.
Total, N299
Median age at transplantation, years50 (range: 16–72)
Gender, n
Type of transplant, n
Heart: first transplant288
Heart: second transplant3
Heart and lung3
Heart and kidney3
Heart and liver2
R+ CMV serotype at transplantation, n (%)184 (61)
R+ D+92 (30)
R+ D−91 (30)
R+ D unknown1
R− CMV serotype at transplantation, n (%)115 (39)
R− D+65 (22)
R− D−50 (17)
R+ EBV serotype at transplantation, n (%)291 (97)
R+ D+277 (93)
R+ D−11 (3)
R+ D unknown3 (1)
R− EBV serotype at transplantation, n (%)6 (2)
R− D+6 (2)
R− D−0
R unknown EBV serotype at transplantation, n (%)2 (1)
R unknown D+2 (1)
R unknown D−0
Figure 2.

Flow chart of Epstein–Barr virus reactivation or primary infection preemptive treatment. D+, Epstein–Barr virus-positive donor; DIS, diminution of immunosuppression; ICU, intensive care unit; NVL, normalization of EBV viral load; R-, Epstein–Barr virus-negative recipient.

CMV reactivation

A total of 158 patients developed a CMV infection (53%), representing 105 reactivations (35%) and 53 primary infections (18%); no patient had any clinical symptoms of CMV disease. All were treated using oral valganciclovir or IV ganciclovir. CMV reactivations and primary infections are detailed in Table 2.

Table 2. Characteristics of CMV and EBV reactivations and primary infections
 Patients, nInfections, n (%)
  1. Abbreviations are explained in Table 1.
  2. ReA, reactivation; PI, primary infection.
Total, N299105 (35)53 (18)31 (10)6 (2)
R+ CMV serotype at transplantation184    
R+ D+9251 (55)8 (9)0
R+ D−9154 (59)10 (11)2 (2)
R+ D unknown1000
R− CMV serotype at transplantation115    
R− D+6547 (72)6 (9)2 (3)
R− D−506 (12)7 (14)2 (4)
R+ EBV serotype at transplantation291    
R+ D+27797 (35)46 (17)31 (11)
R+ D−114 (36)4 (36)0
R+ D unknown32 (67)00
R− EBV serotype at transplantation6    
R− D+62 (33)2 (33)6 (100)
R− D−0
R unknown EBV serotype at transplantation2    
R unknown D+201 (50)00
R unknown D−0

EBV infection

A total of 37 patients (12%) developed an EBV infection, distributed in 31 EBV reactivations and 6 EBV primary infections. Neither CMV reactivation nor CMV primary infection was a predictive factor for EBV reactivation (p = 0.58 and p = 0.37, respectively). Similarly, EBV reactivation was not predictive for CMV reactivation (p = 0.51). However, of the six patients with EBV primary infection, four had a CMV infection occurring within 6 weeks: Two had CMV primary infection, one diagnosed the same day as EBV primary infection and one 13 days before EBV primary infection; and two had CMV reactivations, 10 and 42 days before EBV primary infection.

Treatment of EBV primary infections

All six high-risk patients (i.e. EBV-negative recipient/EBV-positive donor) developed a primary infection after a median of 71 days (range 26–223 days). DIS was effective in treating all six cases. One of these patients developed two other reactivations, respectively, at 212 and 308 days posttransplant. During the second reactivation, DIS was ineffective, and a single rituximab infusion was administered with success. The last reactivation was managed by DIS (cf. Figure 2).

EBV reactivation treatment

Reduction in EBV viral load to baseline levels was achieved by DIS in 22/31 patients. Eight patients received rituximab, either after failure of DIS (n = 3) or as a result of very-high initial viral loads (n = 5). In all cases, EBV viral load decreased below the pathologic threshold following treatment. One patient, who had received prior first-line rituximab, relapsed after 58 days; at this time, despite DIS, the patient's EBV viral load increased, necessitating further rituximab treatment, with success (cf. Figure 2).

Treatment safety

The use of DIS may theoretically increase the risk of graft rejection. We therefore assessed factors that may influence rejection, including EBV and CMV status of donors and recipients, gender, human leukocyte antigen mismatch of cardiac grafts, date of transplant, the nature of CMV infections and EBV management strategies used. Neither DIS during EBV infection nor rituximab treatment increased the incidence of graft rejection (Table 3). Rituximab-associated delayed neutropenia, herpetic infections and multi-focal leukoencephalitis were not observed in rituximab-treated patients in the current study. Thirty engrafted patients who did not receive rituximab developed severe neutropenia (<500 cells/mm3) during follow-up, mainly due to anti-CMV treatment.

Table 3. Graft rejection and patient characteristics
Factorsn (%)p (log-rank)
  1. Abbreviations are explained in Table 1.
  2. HLA, human leukocyte antigen.
EBV status
Donor positive285 (95)0.28
Recipient positive291 (97)0.72
CMV status
Donor positive157 (52)0.56
Recipient positive184 (61)0.69
Gender, male/female226/73 (76/24)0.58
HLA mismatch (1–6) 0.59
047 (16)
11 (0.003)
23 (0.01)
313 (4)
454 (18)
593 (31)
685 (28)
Unknown3 (0.01)
Transplant before/after June 2007174/125 (58/42)0.09
CMV infection
Primary infection0.24
EBV infection management
Diminution of immunosuppression37 (12)0.69
Rituximab infusion9 (3)0.39

Posttransplant lymphoproliferative disorder

One of the six patients with an EBV primary infection was diagnosed with several hepatic nodules, indicating a possible PTLD. Biopsies were not possible because of the small lesion sizes. All lesions disappeared after 1 month of DIS, and the patient was still in complete remission at last follow-up (5 years).

Of the patients with normal EBV infection risk, only one developed a PTLD, which was confirmed by lung biopsy 6 months after transplantation. This patient was EBV-positive, but EBV viral load could not be assessed after heart transplantation, due to rapid transfer to an intensive care unit for respiratory insufficiency. Despite rapid treatment with rituximab, the patient died within a week. With a mean follow-up of 760 days, no other patient developed a PTLD.

Influence of EBV prophylaxis on incidence of PTLD

In order to detect a protective effect of preemptive EBV treatment, we compared PTLD incidence in our study with historical data from before the introduction of systematic EBV viral load assessments from the same heart-transplant center.

In the patient cohort that received transplants between 2004 and 2008, one patient was diagnosed with an early EBV-positive PTLD (day 182) and one patient had a suspected PTLD and primary EBV infection (day 16). In the historical cohort, 24 patients were diagnosed with PTLD (mean 1.8 per year), 13 of whom had early EBV-positive PTLD (mean 137 days). A comparison of incidences of early EBV-positive PTLD for the intent-to-treat populations showed no significant difference (p = 0.111); however, a per-protocol analysis that excluded the patient who did not have EBV viral load follow-up found a significant decrease in PTLD incidence (p = 0.033).


PTLD represents a classical complication of organ transplantation. Analyzing 175 732 transplanted patients (58.4% kidney, 21.6% liver, 10% heart, 4% lung) in the US Scientific Registry of Transplant Recipients, Engels et al [24] found non-Hodgkin's lymphoma to be the most common malignancy, with an incidence of 194/100 000 per year and a standard incidence ratio (SIR) of 7.54 compared with the general population, an incidence even higher than lung cancer (173/100 000 per year, SIR 1.97). In this publication, heart transplant represents an intermediate risk of PTLD (SIR 7.79) compared with lung (SIR 18.73), liver (SIR 7.77) and kidney transplants (SIR 6.05), justifying the relevance of our study.

The prognosis and management of PTLD have evolved over the years. Traditional cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP) chemotherapy for PTLD management results in a complete response rate of 50%, but an early death rate of 33% [25]; patients treated with rituximab alone achieved a complete response rate of 28–53% and an overall response rate of 44–70% with only four 1× weekly infusions [26, 27], which can be further increased to 66% by administering eight infusions of rituximab [28]. However, relapse rate can be as high as 57% at 1 year after rituximab alone [29]. The last published prospective study of PTLD treatment demonstrated an overall response rate of 90% and a median overall survival of 79 months following treatment with four 1× weekly infusions of rituximab, then a 4-week drug holiday followed by four 3-week cycles of CHOP [4]. Patients treated with alternative treatments, such as allogeneic anti-EBV cytotoxic T lymphocytes, have achieved overall survival longer than 2 years; however, more data are needed [30]. Despite these advances, PTLD remains a severe complication and an important cause of death after transplantation, and preventative measures are important.

EBV plays a critical role in the physiopathology of PTLD and represents an attractive target. EBV is implicated in approximately 50% of PTLDs [4] and nearly 100% of PTLDs occurring in the first year after transplantation [4]. Presentation of EBV infection differs between primary infection and reactivation. Seronegative patients are at very high risk of EBV primary infection, with more than 90% of patients developing a primary infection [31], usually in the first year [32]. Up to 33% of these primary infections develop into a true PTLD [31, 32], often with a diagnosis of primary infection and PTLD at the same time, making prevention of PTLD in the patients with primary infection particularly challenging.

There are two main approaches to address EBV infection in the context of PTLD: avoiding EBV primary infection/reactivation and treating EBV primary infection/reactivation. A classical way to decrease the risk of EBV primary infection/reactivation is to improve T cell immunity. The use of immunosuppressive therapies such as OKT3 and azathioprine has been abandoned because of increased incidence of PTLD with a relative risk of 43.2 in patients treated with OKT3 [33], and an increased risk of PTLD and graft rejection with azathioprine compared with MMF [34].

Other approaches to decrease the incidence of EBV primary infection/reactivation include antiviral drugs (ganciclovir and aciclovir) [13] and commercial anti-CMV immunoglobulins (also rich in anti-EBV immunoglobulins) [35]. The link between CMV infection and EBV reactivation was not observed in our study. In a retrospective analysis of 40 patients who had received liver transplants and were seronegative for EBV, Mañez et al [31] showed that 35/37 evaluable patients developed EBV primary infection, 13/40 patients developed a PTLD and cases of PTLD were significantly associated with CMV infections. In our study, four of six seronegative patients with heart transplants had a CMV infection occurring within a 6-week time frame of EBV primary infection. Despite a low number of EBV primary infections, our data agree with those of Mañez et al, prompting us to monitor EBV and CMV viral loads in EBV-seronegative patients.

Despite the range of potential treatment strategies, EBV primary infection and reactivation remain a challenge in patients receiving heart transplants. Specific issues that need to be addressed as part of clinical studies assessing EBV primary infection/reactivation treatment are: identification of an EBV viral load threshold at which to begin treatment for EBV reactivation, verifying the efficacy and safety profiles for rituximab administered following solid organ transplantation, determining when treatment with DIS or rituximab is more appropriate and comparing rates of graft rejection and/or PTLD occurrence following management using DIS and/or rituximab.

With the introduction of regular monitoring of EBV viral load, EBV primary infection can be diagnosed very quickly, before detection of clinical symptoms and usually prior to occurrence of PTLD. In the present study, all EBV-seronegative patients had a primary infection during the first year after transplantation, and one had a concomitant diagnosis of possible PTLD. All responded to DIS, without an increase in graft-rejection rate, and PTLD resolved, without any new occurrences of PTLD during the follow-up (longer than 5 years for this patient), showing the importance of close EBV viral-load monitoring in seronegative patients and early intervention with primary infection therapy.

EBV reactivation represents the majority of EBV infections in transplanted adults. The incidence of EBV reactivation in the current era, using a combination of EBV viral-load monitoring and modern immunosuppressive regimens, has not been widely described. Bakker et al [12] reported EBV reactivation rates in patients who received a lung transplant between 1990 and 2001. A total of 35% of patients experienced an EBV reactivation when followed up between 2001 and 2006. DIS was safe and feasible, with no patients having acute rejection, but patients were treated using a range of immunosuppressive treatments, making it difficult to draw conclusions related to specific therapies. The incidence of EBV reactivation reported clearly depends on the viral-load threshold value used, and, as shown by Table 4, there are a very wide range of results [7, 11, 12, 16, 36-44]. Our threshold (≥105 copies/mL) is relatively high, which might explain the low rate of EBV reactivation compared with other studies. However, we did not see any complications or PTLD in patients with a viral-load <105 copies/mL, which justifies our choice. We did not normalize our viral load to WHO proposed standard as this study has been done before the WHO publication [45].

Table 4. Preventative and curative treatment of Epstein–Barr virus reactivation or primary infection
Refs.ThresholdOrgannPI/ReA (%)PTLDPreventativeCurative
  • Abbreviations are explained in Tables 1 and 2.
  • SOT, solid organ transplantation; Gan, ganciclovir; Aci, aciclovir; PBMC, peripheral blood mononuclear cell; DLI, donor leukocyte infusion; CTL, cytotoxic T lymphocytes; Ig, immunoglobulin; Val, valganciclovir.
  • 1Intent to treat.
  • 2Per protocol.
Current study>105 copies/mLSOT299120.3%1, 0%2DIS ± rituximab
McDiarmid et al [36]≥10 copies/µg DNASOT (pediatric)4064–832/40Gan and AciDIS
Rooney et al [27]>2000 copies/106 PBMCHSCT (pediatric)390Anti-EBV DLI
Baldanti et al [38]>1000 copies/0.5 µg DNAHSCT3231.20DIS
Gustafsson et al [39]≥4 log copies/106 PBMCHSCT9561Donor CTL
Stevens et al [7]>2000 copies/mLSOT (lung)146
Comoli et al [40]≥1000 copies/105 PBMCSOT7Autologous CTL
van Esser et al [16]>1000 copies/mLHSCT4931Rituximab
Lee et al [11]≥4000 copies/µg DNASOT (liver-pediatric)73262%DIS
Humar et al [41]≥1000 copies/106 PBMCSOT1656.38.8%Gan
1433.3 Gan + Ig
Green et al [42]≥2000 copies/106 PBMCSOT (liver-pediatric)432916%Placebo
39219%IgA CMV
Savoldo et al [43]≥1000 copies/µg DNASOT1215.50Autologous CTL
Bakker et al [12]>10 000 copies/mLSOT (lung)75251.5%DIS + Val
Worth et al [44]>40 000 copies/mLHSCT7028.61.4%Rituximab

Treatment of EBV primary infection/reactivation is not standardized. Antiviral drugs are not yet validated, and the use of anti-EBV cytotoxic T lymphocytes after solid organ transplantation is still considered experimental. However, several studies have demonstrated positive data in a nonsolid organ transplantation setting, showing the safety of DIS [11, 12, 36] and depletion of the B cell EBV reservoir using rituximab following HSCT [16, 44]. As a result of a lack of published information on B cell depletion to protect against PTLD, we preferred to use rituximab only in patients with viral loads ≥106 copies/mL, and to use the validated DIS in all patients. Compared with systematic use of rituximab, our algorithm is less expensive and limits any possible side effects of the antibody. Using this approach, we have obtained a 100% response rate. Two patients who relapsed during follow-up were retreated using the same approach and achieved a second complete response. As DIS and rituximab can impact organ engraftment, we assessed the potential risk of graft rejection. For this purpose, we systematically took serial myocardial biopsies, the gold standard for diagnosis of graft rejection. Using this rigorous protocol, we did not detect any increase in rate of graft rejection in patients treated with either DIS or rituximab (Table 3). No treatment-related neutropenia or infections were detected in patients treated with rituximab.

The primary goal of our study was prevention of PTLD. Since the incidence of PTLD is very low, prospective randomized studies are very difficult to design and recruit for; therefore, comparisons with historical data are appropriate to address this question. In our study, we minimized the bias seen in the literature by analyzing only patients who received heart transplants and who were treated using the same approaches to immunosuppression according to the same local practices, providing the most homogenous comparison possible. We also increased the statistical power of the analyses by including a large cohort of prospectively followed patients and a larger historical cohort. Due to the low incidence of PTLD, comparisons using the intent-to-treat population were not statistically significant. The per-protocol analysis, which excluded the patient who was lost to follow-up, showed a significant decrease in PTLD incidence in patients who received EBV primary infection/reactivation treatment. However, immunosuppressive regimens differ in the two groups, with use of azathioprine until the end of 2002 and MMF thereafter. Therefore, our data confirm that the change of azathioprine for MMF and the use of rituximab and DIS appear to decrease PTLD incidence in heart-transplant recipients. The individual impact of each change cannot be assessed in the present study. Similar effects have been suggested in a small cohort of patients receiving HSCT [44], although our findings are the first positive results in patients receiving solid organ transplants.

Our study includes several sources of potential bias. It concerns only heart transplantation, so we must proceed carefully if we extrapolate to other organs. A single-center study can lead to a “center effect,” which may not correlate with outcomes seen in smaller transplant units. In addition, EBV primary infection/reactivation incidences and immunosuppressive regimens can vary between transplant units. Furthermore, responses may vary depending on the organ transplanted; thus, our data cannot be extrapolated to other organs. The most important limitation of this study is that our algorithm is relevant only to EBV-positive PTLD, while EBV-negative PTLD represents half of all PTLD cases, especially in late diagnosis.

In conclusion, we have demonstrated that a simple algorithm, using a high viral-load threshold and simple treatments, available in all transplant units, can manage EBV primary infection/reactivation safely and effectively. This represents the largest study published so far in this field.


Support for third-party medical writing assistance for this manuscript was provided by F. Hoffmann-La Roche Ltd. All authors contributed to the study design, data collection and analysis and writing and review of the manuscript and had access to study data.

Author Contributions

SC performed the research, managed the rituximab treatment, analyzed the clinical and virological results, interpreted the data and wrote the manuscript. SV managed the posttransplantation survey, the immunosuppression decrease and graft-rejection treatment. CD performed the virological survey. JLG performed the statistical analysis. VL helped for the study edification and manuscript redaction.


The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. SC has acted as a consultant for F. Hoffmann-La Roche Ltd. and has received research funding from F. Hoffmann-La Roche Ltd. and Chugai Pharmaceutical Co., Ltd. VL has received honoraria from Janssen-Cilag and Mundipharma, travel support from Roche, and acted as an advisor to Janssen-Cilag and Pharmacyclics. CD, JLG and SV have no competing financial interests to declare.