These authors contributed equally to this work.
Epstein–Barr virus early-antigen antibodies before allogeneic haematopoietic stem cell transplantation as a marker of risk of post-transplant lymphoproliferative disorders
Article first published online: 23 NOV 2006
British Journal of Haematology
Volume 136, Issue 2, pages 305–308, January 2007
How to Cite
Agbalika, F., Larghero, J., Esperou, H., Marais, D., Robin, M., Foïs, E., De Latour, R. P., Gluckman, E., Rocha, V., Benbunan, M., Socié, G. and Marolleau, J.-P. (2007), Epstein–Barr virus early-antigen antibodies before allogeneic haematopoietic stem cell transplantation as a marker of risk of post-transplant lymphoproliferative disorders. British Journal of Haematology, 136: 305–308. doi: 10.1111/j.1365-2141.2006.06420.x
- Issue published online: 23 NOV 2006
- Article first published online: 23 NOV 2006
- Received 31 August 2006, accepted 11 October 2006
- Epstein–Barr virus;
- post-transplant lymphoproliferative disorder;
- allogeneic haematopoietic stem cell transplantation;
- early-antigen immunoglobulin G
The occurrence of post-transplant lymphoproliferative disorders (PTLDs) after allogeneic haematopoietic stem cell transplantation (allo-HSCT) represents a clinical problem. Pretransplant Epstein–Barr virus serological status and viral load was determined in 21 recipients and 28 control transplanted patients, with (+) and without (−) PTLD, respectively. Early-antigen immunoglobulin G (EA-IgG) was detected in 12/21 PTLD+ patients, but only 2/28 PTLD− patients (P = 0·00023, Odds ratio = 17·42). High viral load was detected in peripheral blood mononuclear cells at PTLD diagnosis, independently of deleted LMP1. Detection of EA-IgG in allo-HSCT recipients pretransplantation might be considered as risk factor for PTLD development.
It is now known that the deficit in Epstein–Barr virus (EBV)-directed immune surveillance, often encountered in allogeneic haematopoietic stem cell transplantation (allo-HSCT), results in EBV-induced post-transplant lymphoproliferative disorders (PTLDs). These PTLDs, which are clinically and morphologically heterogeneous, can develop early after transplant and represent the most common neoplastic diseases among patients in the first year post-transplant (Curtis et al, 1999). The importance of the PTLDs lie in their high mortality, particularly in the adult population, approaching a rate of 24% in patients receiving T-cell-depleted bone marrow (Shapiro et al, 1988). The use of quantitative polymerase chain reaction (PCR) as a diagnostic tool has dramatically changed the diagnostic criteria and prognosis of these disorders. PTLDs are frequently diagnosed as a result of the EBV viral load, although patients develop only an isolated fever, with or without tumour burden and monoclonal gamma-globulin. Thus, early or pre-emptive treatment with anti-B-cell monoclonal antibodies or donor-derived EBV-specific cytotoxic T lymphocytes can be initiated when high viral load is detected by quantitative PCR of EBV-DNA (Wagner et al, 2004). Several risk factors have been shown to increase the risk of PTLDs (Gottschalk et al, 2005), but their predictive role is yet undetermined in allo-HSCT. In this study, we showed that the presence of EBV early-antigen immunoglobulin G (EA-IgG) in recipients before transplantation is associated with higher incidence of developing early PTLD (within 1-year post HSCT), therefore identifying those patients at risk who could benefit from better surveillance and monitoring of EBV reactivations after allo-HSCT.
Materials and methods
This retrospective single-centre study included 21 patients diagnosed with PTLD after allo-HSCT between December 1999 and January 2003. Their characteristics are listed in Table I. Cells infused were collected from bone marrow (n = 16) or peripheral blood (n = 5 of which one was T-cell depleted). They were compared with 28 matched control allo-HSCT recipients without clinical PTLD during the same time period. All patients included in the study gave informed consent. PTLD was defined using well established criteria (Knowles et al, 1995).
|Patient||Donor EBV status||Sibling||Cell source||Diagnosis||Infused nucleated cells (109)||Infused CD3+ cells (107)||Infused CD19+ cells (107)||EBV copies per 1 × 106 cells||Lymphoma diagnosis (in months after graft)|
Graft cell content
Automated cell counts and cell subpopulation quantification were performed both before and after processing, but only the latter was taken into account in this study. A Sysmex instrument (Roche Diagnostics, Meylan, France) was used to determine total nucleated cells. CD3+ and CD19+ cell quantification was performed by fluorescence analysis. Briefly, one million cells were incubated for 10 min at room temperature with monoclonal antibodies or matched isotype controls (all purchased from Becton Dickinson, Le Pont de Claix, France). Immunofluorescence analysis was performed with a 5-parameter FACSscan (Becton Dickinson Immunocytometry Systems, San Jose, CA, USA).
Evaluation of EBV status
Epstein–Barr virus evaluation in recipients was based on serology, viral load measurement and identification of the wild-type (w-LMP1) and of the deleted variant (del-LMP1) of the LMP1 gene.
Serological EBV infection status. Sera were collected 2 weeks before transplant in the 49 patients. IgG and IgA antibodies against viral capsid antigen (VCA) and EA were detected using enzyme immunoassay (EIA) from a commercial Diasorin kit (DiaSorin Laboratories, Sallugia, Italy). A reactive response was defined as the optical density ratio between sample and negative control mean ≥ 1, i.e. 20 arbitrary units. Immunofluorescence assay (IFA), as gold standard, was also performed using specific-antigen-coated slides purchased from Gull laboratories (Zeus Scientific Inc., Raritan, NJ, USA). According to the manufacturers’ recommendation, the assay was reactive when fluorescent reactions were noted in samples diluted at 1:10. VCA-IgM antibodies were also evaluated using EIA and IFA. In all, no discordant results were obtained by these two methods.
Viral load measurement. Quantification of EBV-DNA extracted from cells infused to 49 recipients and from recipients’ peripheral blood mononuclear cells (PBMC) was performed using quantitative real time PCR as previously described (Clave et al, 2004). EBV genome copy number was determined in comparison with standard curves constructed by plotting the serial dilution of plasmid standards containing BALF5 gene and calculated as indicated by elsewhere (Kimura et al, 1999).
Identification of EBV wild-type or deleted LMP1 gene. All samples containing viral EBV-DNA were analysed to search for deletion of the LMP1 gene using a two-step procedure. First, the sequence within the C-terminal region of the LMP1 gene was amplified as previously described (Tao et al, 1998), to obtain products of either 260 bp (w-LMP1) or 230 bp (del-LMP1), respectively, analysed on agarose gel at 3% (w/v). Second, these results were confirmed by a dual qualitative PCR within the C-terminal sequence of LMP1 gene by using two primers sets and two probes designed on primer express software (PE Biosystems, Foster City, CA, USA) between nucleotides 168208 and 168318 of the EBV sequence. The first primers set was as follows: 5′-TCCACCGGAACCA-3′ and 5′-GGCCCGCCTTTGATGC-3′ with FAM-TAMRA-labelled probe (5′-CAAAAGCAGCGTAGGAAGGTGTGGATCA-3′). The second primers set, nested within the first set, was as follows: 5′-TCCACCGGAACCAGAAGAAC-3′ and 5′-GGCCCGCCTTTGATGAC-3′ with VIC-MGB-labelled probe (5′-CATGGCCGGAATC-3′) partially spanning the deleted sequence. The run was performed with the same quantity of DNA as previously described (Clave et al, 2004). The w-LMP1 gene was amplified with the two sets and the products were detected with both probes. When LMP1 was deleted, only the first set amplified the product detectable with 6-carboxy-fluorescein (FAM)-labelled probe.
Differences between the two groups were analysed using Fisher's exact test. Odds ratio (OR) was performed on EA-IgG data combined from PTLD and control patients. A value of P < 0·05 was used to define statistical significance.
Results and discussion
In this study, 21 recipients with PTLD (1–8 months, median 2 months) after allo-HSCT were analysed and compared with 28 matched control recipients without PTLD. The initial diseases in both groups were similar and included acute and chronic leukaemia, aplastic anaemia, Fanconi anaemia and multiple myeloma. The median age was 23 years (range, 5–50 years) in the patient group and 21 years (range, 6–45 years) in the control group (P = 0·4). The two groups were also matched for use of antithymocyte globulin before transplant and transplant outcomes, such as engraftment and relapse.
The 21 patients that developed PTLD received a median number of total nucleated, CD3+ and CD19+ cells of 15·62 × 109, 196·3 × 107 and 50·2 × 107 respectively (Table I). EBV-DNA quantification in the infused cells revealed that two of 21 grafts were positive in the PTLD+ recipient group (1·2–2·4 log/106 cells) and three of 28 in the PTLD− recipient group (0·3–0·5 log/106 cells), showing that infused cells were rarely infected with EBV. The 21 recipients that developed PTLDs were treated with Rituximab® and 15 (71·4%) responded to this treatment. The time to PTLD diagnosis was found to be independent of donor sex, cell source and infused cell dose.
In these recipients, before conditioning, we also evaluated the pretransplant serological responses to the three classical EBV antigens (VCA, EA and EBNA) using EIA and IFA to detect M, G and A immunoglobulins. Table II clearly indicates that the 49 recipients [patients (PTLD+) and controls (PTLD−)] had serological patterns of previous EBV infections. On the other hand, EA-IgG was more significantly detected in PTLD+ recipients (12/21) than in control group (2/28, P = 0·00023, OR = 17·42, 95% confidence interval (95% CI), range: 2·1–175·7). These data suggest that anti-EA antibodies, considered as the hallmark of EBV-reactivation before transplantation, in the presence of VCA- and EBNA-antibodies, could be a predictive marker for development of PTLD. These results contrast with data reported in recipients of solid organ transplants at the time of lymphoproliferative disease (Carpentier et al, 2003; Leruez-Ville et al, 2004). Thus, from a small patient cohort of paediatric recipients of solid organ transplants, it was stated that some EA-IgG positive recipients before transplantation did not develop PTLDs, even if a high EBV load was detected. In addition, the absence of EBV-EA serological response was considered as a negative predictive value at the time of PTLD. For that, samples from 10 out of 12 EA-IgG positive PTLD+ patients were tested for EBV serological antibodies at the time of PTLD, using the same commercial kit purchased from Diasorin. Nine of these 10 patients were negative for EA-IgG and one remained positive. These results underline the variability of EBV serological responses, as clearly demonstrated in cellular immunity linked to the efficiency of epitope presentation during the lytic viral cycle (Pudney et al, 2005). Thus, our pretransplant data suggest that EA-IgG detection before transplantation is a hallmark of a transiently unbalanced immunity that predisposes to early PTLD. The hypothesis of contamination of the newly reconstituting cells by EBV-infected epithelial cells through intimate contact in the reticulated crypt epithelium can be advanced to explain the emergence of the disease (Pegtel et al, 2004).
|Pretransplant serological patterns (seroreactivity)||EBV viral load and LMP1 deletion|
|Before graft||After PTLD diagnosis|
|Anti-VCA IgG||Anti-VCA IgM||Anti-VCA IgA||Anti-EBNA IgG||Anti-EA IgG*||Anti-EA IgA||EBV load (infused cells)||EBV load/106 PBMCs||Deleted C-terminus LMP1 (30 bp)|
|PTLDs patients (n = 21)|
|21/21 (100)||0/21 (0)||0/21 (0)||21/21 (100)||12/21 (57)*||0/21 (0)||2/21 (9·5) 1·2log-2·4log/106 cells||21/21 (100) 4·3log-6·8log/106 cells||3/21 (14·3)†|
|Control patients (n = 28)|
|28/28 (100)||0/28 (0)||0/28 (0)||28/28 (100)||2/28 (9·5)*||0/28 (0)||3/28 (10·7) 0·3log-0·5log/106 cells||0/28 (0)||0/28 (0)|
In contrast, pretransplant EBV genome number quantification by real time PCR in infused cells (2/21 and 3/28 with EBV-DNA detectable in PTLD+ and PTLD− respectively) and detection of C-terminal deleted LMP1 gene (only 3/21 vs. 0/28) were not predictive for early PTLD with high viral load (4·3–6·8 log/106 PBMC). Despite a large 95% CI as a result of the small subgroups of recipients included in this study and that the EA-IgG parameter is normally absent in healthy individuals, our results showed that the presence of pretransplant EA antibodies in recipients, concomitantly with VCA- and EBV nuclear antigen (EBNA)-IgG, might represent a new, reliable marker that is predictive for early PTLD after allogeneic haematopoietic stem cell transplant. Whether this serological pattern is a useful tool for screening and monitoring future candidates for transplantation remains to be confirmed in a larger cohort of patients. In order to confirm the results obtained in this retrospective study, we prospectively followed our allo-HSCT recipients for EA antibodies. During year 2005, seven recipients were EA-IgG positive before transplantation. Six of these displayed a dramatic increase in EBV viral load after transplant and were treated at the time with Rituximab®. On the contrary, none of the EA-IgG negative recipients showed increase in EBV viral load during the same period.
- 2003) Epstein-Barr virus (EBV) early-antigen serologic testing in conjunction with peripheral blood EBV DNA load as a marker for risk of posttransplantation lymphoproliferative disease. Journal of Infectious Diseases, 188, 1853–1864. , , , & (
- 2004) Epstein-Barr virus (EBV) reactivation in allogeneic stem-cell transplantation: relationship between viral load, EBV-specific T-cell reconstitution and rituximab therapy. Transplantation, 77, 76–84. , , , , , , , , , , , , , , , & (
- 1999) Risk of lymphoproliferative disorders after bone marrow transplantation: a multi-institutional study. Blood, 94, 2208–2216. , , , , , , , , , , , , , , , , , & (
- 2005) Post-transplant lymphoproliferative disorders. Annual Review of Medicine, 56, 29–44. , & (
- 1999) Quantitative analysis of Epstein-Barr virus load by using a real-time PCR assay. Journal of Clinical Microbiology, 37, 132–136. , , , , , , & (
- 1995) Correlative morphologic and molecular genetic analysis demonstrates three distinct categories of posttransplantation lymphoproliferative disorders. Blood, 85, 552–565. , , , , , & (
- 2004) Epstein-Barr virus (EBV) early-antigen antibodies and EBV DNA load in blood, in postransplantation lymphoproliferative disease. Journal of Infectious Diseases, 190, 1524–1525. , , , , & (
- 2004) Epstein-Barr virus infection in ex vivo tonsil epithelial cell cultures of asymptomatic carriers. Journal of Virology, 78, 12613–12624. , & (
- 2005) CD8+ immunodominance among Epstein-Barr virus lytic cycle antigens directly reflects the efficiency of antigen presentation in lytically infected cells. Journal of Experimental Medicine, 201, 349–360. , , & (
- 1988) Epstein-Barr virus associated B cell lymphoproliferative disorders following bone marrow transplantation. Blood, 71, 1234–1243. , , , , , , , , & (
- 1998) Epstein-Barr virus (EBV) variant with a 30-bp deletion at the carboxyl terminus (amino acids 346–355) of latent membrane protein-1 (LMP1) gene is able to transform cells and evade immune surveillance. American Journal of Pathology, 152, 1398–1399. , & (
- 2004) Prompt versus preemptive intervention for EBV lymphoproliferative disease. Blood, 103, 3979–3981. , , , , , , , & (