- Email: [email protected]
Dendritic cells improve the generation of Epstein-Barr virus–specific cytotoxic T lymphocytes for the treatment of posttransplantation lymphoma
Dendritic cells improve the generation of Epstein-Barr virus–specific cytotoxic T lymphocytes for the treatment of posttransplantation lymphoma Grayson H. Wheatley III, MD, Karen P. McKinnon, PhD, Michelle Iacobucci, BS, Sarah Mahon, BS, Cohava Gelber, PhD, and H. Kim Lyerly, MD, Durham, NC
Background. The use of immunosuppressive therapies after solid organ transplantation has been shown to increase a patient’s risk for Epstein-Barr virus (EBV)–associated lymphoma. A potential therapy for this disorder is the adoptive transfer of EBV-specific cytotoxic T lymphocytes (CTLs). We proposed that dendritic cells (DCs) could be loaded with EBV antigens and be used to improve the in vitro generation of EBVspecific CTLs. Methods. Autologous EBV-transformed B-lymphoblastoid cell lines (BLCLs) were generated from normal donors, and CTLs were initiated by culturing peripheral blood mononuclear cells with DCs alone, disrupted BLCLs alone, intact, irradiated BLCLs alone, and DCs loaded with disrupted BLCLs. Lytic activities were determined with a 4-hour chromium-release assay against autologous BLCLs, and statistical calculations were performed by a Student t test assuming equal variance. Results. The lytic activity of CTLs generated with DCs loaded with disrupted BLCLs reached 78% and was statistically significant (P < .01) at all effector/target ratios compared with CTLs generated with DCs alone, disrupted BLCLs alone, or intact BLCLs alone. Total numbers of CTLs were also greater than those of control groups for DCs loaded with disrupted BLCLs. Conclusions. DCs improved the in vitro generation of EBV-specific CTLs as evidenced by this group’s significantly increased lytic activity over that of the control group. The improved lytic activity of DC-generated EBV-CTLs suggests that adoptive transfer of these cells could lead to a more effective immunotherapeutic response against posttransplantation EBV-associated lymphoma. (Surgery 1998;124:171-6.) From the Departments of Surgery, Pathology, and Immunology, Duke University Medical Center, Durham, NC
SOLID ORGAN TRANSPLANTATION HAS become the accepted standard of therapy for a number of acquired and inherited diseases. The number of patients undergoing solid organ transplantation has been increasing steadily, and 19,017 patients in the United States underwent solid organ transplantation in 1994.1 The implementation of increasingly potent immunosuppressive therapies has allowed many allograft recipients to achieve successful outcomes. However, a potentially serious complication associated with prolonged immunosuppression in allograft recipients is the development of posttransplantation lymphoproliferative disorder (PTLD). PTLD is a frequent tumor arising in patients after transplantation and ranges from lymphoid hyperPresented at the Fifty-ninth Annual Meeting of the Society of University Surgeons, Milwaukee, Wis, Feb 12-14, 1998. Reprint requests: H. Kim Lyerly, MD, Box 2606, DUMC, Durham, NC 27710. Copyright © 1998 by Mosby, Inc. 0039-6060/98/$5.00 + 0
11/6/91179
plasia to malignant non-Hodgkin’s lymphoma (NHL).2 The incidence of PTLD in allograft recipients has paralleled the increased use of immunosuppressants cyclosporine, anti-CD3 monoclonal OKT3, and FK506.3,4 PTLD is now one of the most frequent tumors arising in patients after transplantation and totals 15% to 25% of neoplasms, compared with a 5% incidence in the general population.4 PTLD in allograft recipients is primarily of B-cell origin, and host infection with Epstein-Barr virus (EBV) is an important etiologic factor.5 A 200-kilobase double-stranded DNA virus belonging to the human herpesvirus family, EBV has been implicated in the pathogenesis of many lymphoproliferative diseases, including PTLD.6 EBV, either type 1 or 2, infects primarily resting B cells through a receptor identical to the C3d complement receptor.7,8 Once intracellular, the virus is maintained as extrachromosomal episomes. A persistent life-long infection results, and the balance between latency and immortalized B-cell proliferation rests solely on the host immune SURGERY 171
172 Wheatley et al.
Surgery August 1998
used to activate EBV-specific CTLs (EBV-CTLs) in vitro. In addition, we proposed that EBV-CTLs generated with DCs pulsed with disrupted tumor cells have increased lytic activity compared with EBV-CTLs generated with tumor cells alone.
Fig. 1. Flow cytometric analysis of DCs derived from day 8 of culture of PBMCs. A, Isotypic control; B, CD14FITC/HLA-DR-PE; C, CD18-FITC/CD80-PE; D, CD86FITC/CD83-PE.
system. When the host immune system is disrupted, as in allograft recipients receiving exogenous immunosuppression, latently infected B cells can undergo rapid polyclonal or monoclonal expansion. Although some allograft recipients with PTLD respond to first-line chemotherapy, this treatment modality is limited by excessive toxicity and frequent tumor resistance and recurrence. A potential anticancer therapy for EBV-associated NHL in patients after transplantation is immunotherapy.9 Tumor-specific CD8+ cytotoxic T lymphocytes (CTLs) are an important effector arm in antitumor immunity.10 These lymphocytes recognize short peptide epitopes derived from tumor-rejection antigens, presented by class I major histocompatability complex (MHC) molecules on the tumor cell surface. Although recognition of peptide class I complexes is sufficient to trigger lysis, priming of CTL responses requires the presentation of the relevant tumor-rejection antigen by professional antigen-presenting cells capable of providing co-stimulation. Dendritic cells (DCs) are considered to be the most potent antigen-presenting cells for mediating a variety of antitumor immune responses. These include stimulation of helper T cell–dependent antibody responses, as well as the generation of primary and secondary cytotoxic T cell responses to tumor- associated antigens.11,12 As a consequence, DCs loaded with tumor antigens may represent a potentially powerful method of inducing antigen-specific CTL expansion and resulting antitumor immunity. The objective of this study was to demonstrate that DCs could be loaded with EBV antigens and be
MATERIAL AND METHODS Generation of B-lymphoblastoid cell lines (BLCLs). After obtaining informed, written consent, healthy donors underwent leukapheresis. Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Hypaque density-gradient centrifugation, washed, and suspended at 5 × 106 cells/mL in AIM V lymphocyte activation medium (Gibco, Grand Island, NY) containing 2 µg/mL cyclosporine (Sandoz Pharmaceutical Co, East Hanover, NJ) and 2 mL viral supernatant harvested from an EBV-expressing cell line (B95.8). The solution was incubated at 37° C with 5% carbon dioxide and humidified room air for 18 hours. The PBMCs were resuspended at 2 × 106 cells/mL in AIM V medium, and the cultures were inspected daily for evidence of clonal expansion. Culture medium was added as needed, and cultures were expanded by splitting 1:2 with fresh AIM V medium. When adequate numbers of cells were available, the presence of B-cell markers CD19/CD20 and the absence of T and natural killer markers were confirmed by flow cytometry. Generation of DCs from PBMCs. After leukapheresis, fresh PBMCs were suspended in AIM V culture medium, placed in tissue culture flasks, and incubated at 37° C with 5% carbon dioxide and humidified room air for 2 hours. The nonadherent population was then removed and cryopreserved until stimulated with mature DCs. The adherent PBMCs were cultured in AIM V supplemented with 10% autologous serum and 500 units/mL human interleukin (IL)-4 (Schering-Plough) and 800 units/mL human granulocyte-macrophage colonystimulating factor (GM-CSF) (Schering-Plough). Cytokines were replaced every 48 hours. On day 8 the DCs were removed from culture and a portion was cryopreserved. Flow cytometric analysis. Phenotypic analysis of harvested, cultured DCs was determined by flow cytometry. Briefly, for dual-color immunofluorescence analyses, cells were typed with either fluorescein isothiocyanate–conjugated (FITC) mononuclear antibodies against human CD14, CD18, and CD86 and phycoerythrin-conjugated (PE) monoclonal antibodies against human HLA-DR, CD80, and CD83. The antibodies conjugated with both FITC and PE were used at saturating concentrations. Conjugated monoclonal antibodies with specificity
Wheatley et al. 173
Number of Cells (x106)
Surgery Volume 124, Number 2
Fig. 2. Total number of lymphocytes in culture. Proliferation of EBV-CTLs for each experimental condition after initial stimulation was determined by counting total number of lymphocytes beginning on day 0 and continuing through every other day of culture. EBV-CTL cultures were initiated on day 0 by culturing PBMCs with DCs alone, disrupted autologous BLCLs alone, intact irradiated autologous BLCLs alone, or DCs loaded with disrupted autologous BLCLs. Cultures were counted and restimulated with original experimental conditions on day 14.
for human immunoglobulin (Ig)G1 and IgG2 were used as isotypic controls. Cells were analyzed with a FACScan cytometric flow cytometer, and the large cell population was gated to exclude contaminating lymphocytes. Likewise, phenotypic analysis of EBVCTLs was determined with the same isotypic controls, FITC mononuclear antibodies against human CD3 and CD20, and PE mononuclear antibodies against human CD20 and CD56. Initiation and expansion of autologous EBV-CTLs. Four experimental conditions for initiating and expanding EBV-CTLs were studied (DCs alone, disrupted BLCLs alone, intact, irradiated BLCLs alone, and DCs loaded with disrupted BLCLs). Responder/stimulator ratios of 100:1 were used to initiate all conditions. Autologous BLCLs were disrupted by suspending washed BLCLs at a concentration of 10 × 106 cells/mL, subjected to three cycles of freezing to –70° C and thawing to +40° C, and irradiated with 5000 rads by a cesium source. Intact, irradiated BLCLs were washed, resuspended at 10 × 106 cells/mL, and irradiated with 5000 rads by a cesium source. DCs were loaded with disrupted BLCLs by coculturing DCs with disrupted BLCL solution (1 DC/1 BLCL equivalent) in siliconized glass tubes in 0.5 mL AIM V for 16 hours at 37° C with 5% carbon dioxide and humidified room air. All EBV-CTL cultures were initiated in AIM V medium containing AIM V medium with 10% autologous serum in the presence
of 100 units/mL human IL-2 (Proleukin) and 10 ng/mL human IL-7 (Gibco). Cytokines and fresh medium were added every 48 hours beginning on day 4. Cultures were restimulated according to identical experimental conditions on day 14. Cytotoxicity assay. Cytolytic activities of EBVCTLs were determined with a standard 4-hour 51Crrelease assay. Briefly, target cells, consisting of autologous BLCLs, were labeled with 51Cr for 90 minutes. The targets were washed and added to a 96-well round-bottomed tissue culture plate as 5000 targets/well in 100 µL RPMI (Gibco) with 10% autologous serum. EBV-CTLs for each experimental condition were washed and added to target wells at effector/target (E/T) ratios of 10:1, 20:1, and 40:1. All assays were performed in triplicate. After 4 to 5 hours of incubation, the supernatant fluids were collected and counted in a Wallac gamma counter. Percent specific lysis was calculated according to the following formula: Percent specific lysis = ([Observed release – Spontaneous target release]/[Maximal target release – Spontaneous target release]) × 100. Maximal target release was obtained by incubating targets with 100 µL 10% Triton-X solution, and spontaneous target release was obtained by incubating targets with 100 µL culture medium for the full 4-hour period. Statistical analysis. The lytic activities of EBVCTLs for each experimental group were compared
174 Wheatley et al.
Surgery August 1998
Fig. 3. Cytotoxicity assays. Lytic activities of each experimental group of EBV-specific CTLs were determined on day 24 of culture against autologous BLCLs with standard 4-hour chromium-release assay. EBV-CTL cultures were initiated on day 0 by culturing PBMCs with DCs alone, disrupted autologous BLCLs alone, intact irradiated autologous BLCLs alone, or DCs loaded with disrupted autologous BLCLs. Cultures were restimulated with original experimental conditions on day 14, and lytic activities were determined at E/T ratios of 10:1, 20:1, and 40:1.
at E/T ratios of 10:1, 20:1, and 40:1 with a Student t test, assuming equal variance. RESULTS Dendritic cells. PBMCs were obtained from healthy donors. The adherent PBMC fraction was cultured for 1 week in GM-CSF and IL-4 containing culture medium. Fig. 1 shows a representative cytometrogram of the phenotype of cultured DCs harvested on day 8. The mature DCs were harvested on day 8 and analyzed for the presence of surface markers MHC class I, MHC class II, and CD86 by flow cytometry. DCs generated under the same growth conditions from many preparations of other PBMCs reproducibly expressed high levels of HLA-DR and CD86 but were negative for CD14. B-lymphoblastoid cell lines. After establishment of an autologous BLCL, the resulting cells were phenotyped for determination of B-cell markers. The cells had very high levels of CD19/CD20, phenotypic markers for B cells, and an absence of CD3 and natural killer phenotypic markers on flow cytometric analysis. In vitro expansion of EBV-CTLs. EBV-CTL cultures were initiated by stimulating the nonadherent PBMC fraction with DCs alone, disrupted BLCLs alone, intact, irradiated BLCLs alone, or DCs loaded with disrupted BLCLs. Fig. 2 shows the numeric expansion of EBV-CTLs for each experimental condition. The only condition that had a net expansion of EBV-CTLs on day 24 of culture,
after a restimulation on day 14, were those initiated with DCs loaded with disrupted BLCLs. There was a 50% increase in total number of cells on day 24 compared with the number of PBMCs on day 0, whereas all other conditions had a net decline. Cytotoxicity assays. A 51Cr cytotoxicity-release assay was performed on day 24 after a restimulation on day 15. Fig. 3 shows the percent specific lysis of each category of EBV-CTLs (effector) against autologous BLCLs (targets) at E/T ratios of 10:1, 20:1, and 40:1. Percent specific lysis by EBV-CTLs stimulated with DCs loaded with disrupted tumor cells was greater than each of the other categories at all E/T ratios (P < .01, Student t test). The background natural killer activity was minimal for all groups examined. The lytic activity of EBV-CTLs stimulated with DCs loaded with disrupted BLCLs against K562 cells was less than 5% at each E/T ratio. DISCUSSION Although the EBV genome contains more than 100 genes and encodes more than 80 proteins, B cell transformation is mediated specifically by 6 nuclear antigens (EBNAs 1, 2, 3A, 3B, 3C, and LP) and 2 membrane-derived proteins (LMP1 and LMP2).13 CD8+ CTLs are the immune effector cells responsible for eliminating EBV-infected B cells and recognize EBNA 3A, 3B, 3C, and LMP antigens expressed on the cell surface in the context of MHC class I molecules.14-16 These antigens are processed individually within the cytoplasm of antigenpresenting cells into smaller epitope fragments and
Wheatley et al. 175
Surgery Volume 124, Number 2
are then presented on the cell surface, in the context of MHC class I, to CD8+ T lymphocytes. Ex vivo expansion and adoptive transfer of EBVspecific CTLs offer the possibility of enhancing the cellular immune response against previously transformed B cells. EBV-associated NHL in patients after transplantation results from intentional exogenous suppression of the cellular immune system and manifests physiologically as a deficit in immune surveillance. The ongoing balance between the number of functioning EBV-specific CTLs and the rate of EBV-transformed B-cell proliferation is critical to preventing malignancy. Selectively restoring functional aspects of the immune system in immunosuppressed patients through adoptive transfer of EBV-specific CTLs could augment the depleted immune system and potentially target established B-cell lymphomas.17 DCs are present in small numbers in most tissues including skin, lung, liver, spleen, blood, lymphoid organs, and bone marrow.18 Morphologically, they are large cells with elongated, stellate processes called dendrites. They lack cell surface markers typical for B, T, natural killer, and monocytemacrophage lineages but express high levels of MHC classes I and II, as well as co-stimulatory molecules. These cells have been shown to internalize, process, and present soluble antigen as peptides in conjunction with MHC classes I and II.19,20 In addition, DCs have the unique ability to cluster naive T cells and respond to antigen by rapid up-regulation of the expression of MHC and co-stimulatory molecules, the production of cytokines, and migration toward lymphatic organs.20,21 The method of presentation of EBV antigens for in vitro CTL expansion is critical to generating highly specific CTLs that efficiently target and lyse EBV-expressing tumor cells. Intact, whole B lymphocytes can act as antigen-presenting cells and possess the requisite MHC class I and co-stimulatory molecules to activate CD8+ T lymphocytes. In addition, the in vitro pattern of EBV antigen expression by EBV-transformed BLCLs mimics that expressed by B cell neoplasms in vivo. Although intact BLCLs possess both EBV antigens and MHC class I molecules, they are not the optimal method for presenting EBV antigens to CTLs. Our results indicate that DCs improve the activation and expansion of EBV-CTLs in vitro. DCs pulsed with disrupted BLCLs appear to generate EBV-CTLs with increased EBV-specific lytic activities at all E/T ratios examined. Most likely, this is related to the improved antigen-presenting capabilities of DCs compared with BLCLs. The additional co-stimulatory molecules present on DCs, combined with
the enhanced antigen-processing capabilities of DCs, indicate that DCs can be used for the generation of EBV-CTLs in vitro. Although the subpopulations of EBV-specific CTLs contained in the EBVCTL lymphocyte cultures were not specifically determined in this study, the optimal strategy for adoptive immunotherapy is to use a bulk culture approach to the expansion of EBV-CTLs. It is possible that the EBV-CTLs did undergo some expansion in the bulk cultures used in this study, as demonstrated by the increased EBV-specific lytic activity of CTLs against EBV-expressing BLCLs. Future studies will focus on improving the loading of DCs with disrupted BLCLs. Potential improvements include use of cationic liposome preparations, in combination with disrupted tumor cells, to augment loading of DCs and optimize the co-culture conditions between PBMCs and pulsed DCs. The improved lytic activity of DC-generated EBV-CTLs suggests that adoptive transfer of these cells could lead to a more effective immunotherapeutic response in vivo against posttransplantation EBV-associated lymphoma. REFERENCES 1. United Network for Organ Sharing, Richmond, Va, 1994. 2. Basgoz N, Preiksaitis JK. Post-transplant lymphoproliferative disorder. Infect Dis Clin North Am 1995;9:901-23. 3. Lucas KG, Pollok KE, Emanuel DJ. Post-transplant EBV induced lymphoproliferative disorders. Leuk Lymphoma 1997;25:1-8. 4. Boubenider S, Hiesse C, Goupy C, Kriaa F, Marchand S, Charpentier B. Incidence and consequences of post-transplantation lymphoproliferative disorders. J Nephrol 1997;10:136-45. 5. Mcknight JL, Cen H, Riddler SA, Breinig MC, Williams PA, Ho M, et al. EBV gene expression, EBNA antibody responses and EBV+ peripheral blood lymphocytes in posttransplant lymphoproliferative disease. Leuk Lymphoma 1994;15:9-16. 6. Young L, Alfieri C, Hennessy K, Evans H, O’Hara C, Anderson KC, et al. Expression of Epstein-Barr virus transformation–associated genes in tissues of patients with EBV lymphoproliferative disease. N Engl J Med 1989;321:1080-5. 7. Nemerow GR, Mold C, Schwend VK, Tollefson V, Cooper NR. Identification of gp350 as the viral glycoprotein mediating attachment of Epstein-Barr virus (EBV) to the EBV/C3d receptor of B cells: sequence homology of gp350 and C3 compliment fragment C3d. J Virol 1987;61:1461-70. 8. Thorley-Lawson DA, Geilinger K. Monoclonal antibodies against the major glycoprotein (gp350/220) of Epstein-Barr virus neutralize infectivity. Proc Natl Acad Sci USA 1980;77:5307-11. 9. Heslop HE, Rooney CM. Adoptive cellular immunotherapy for EBV lymphoproliferative disease. Immunol Rev 1997;157:217-22. 10. Townsend A, Bodmer H. Antigen recognition by class I–restricted T-lymphocytes. Ann Rev Immunol 1989;7:601-24. 11. Steinman RM. The dendritic cell system and its role in immunogenicity. Annu Rev Immunol 1991;9:271-96.
176 Wheatley et al.
12. Sornasse T, Flamand V, De Becker G, Bazin H, Tielemans F, Tielemans K, et al. Antigen- pulsed dendritic cells can efficiently induce an antibody response in vivo. J Exp Med 1992;175:15-21. 13. Young L, Alfieri C, Hennessy K, Evans H, O’Hara C, Anderson KC, et al. Expression of Epstein-Barr virus transformation–associated genes in tissues of patients with EBV lymphoproliferative disease. N Engl J Med 1989;321:1080-5. 14. Murray RJ, Kurilla MG, Brooks JM, Thomas WA, Rowe M, Kieff E, et al. Identification of target antigens for the human cytotoxic T cell response to Epstein-Barr virus (EBV): implications for the immune control of EBV-positive malignancies. J Exp Med 1992;176:157-68. 15. Murray RJ, Wang D, Young LS, Wang F, Rowe M, Kieff E, et al. Epstein-Barr virus- specific cytotoxic T-cell recognition of transfectants expressing the virus-coded latent membrane protein LMP. J Virol 1988;62:3747-55. 16. Burrows SR, Sculley TB, Misko IS, Schmidt C, Moss DJ. An Epstein- Barr virus–specific cytotoxic T-cell epitope in EBNA 3. J Exp Med 1990;171:345-50. 17. Rickinson AB, Murray RJ, Brooks J. T-cell recognition of Epstein-Barr virus–associated lymphomas. In: Franks LM, editor. Cancer surveys, a new look at tumor immunology, vol 13. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 1992. p 53-80. 18. Inaba K, Metlay J, Crowley M, Witmer-Pack M, Steinman RM. Dendritic cells as antigen presenting cells in vivo. Int Rev Immunol 1990;6:197-206. 19. Crowley M, Inaba K, Steinman R. Dendritic cells are the principal cells in mouse spleen bearing immunogenic fragments of foreign proteins. J Exp Med 1990;172:383-90. 20. Steinman R, Witmer-Pack M, Inaba K. Dendritic cells: antigen presentation, assessory function and clinical relevance. Adv Exp Med Biol 1993;329:1-9. 21. Inaba K, Romani N, Steinman R. An antigen-independent contact mechanism as an early step in T cell proliferative responses to dendritic cells. J Exp Med 1989;170:527-42.
DISCUSSION Dr Michael Abecassis (Chicago, Ill). Stan Rodell has shown that adoptive immunotherapy can be effective in cytomegalovirus disease. Are these populations of CTLs specifically directed against an epitope? In other words, there are different viral antigens that are responsible for different phases of the lytic infection. Have you looked into the populations of CTLs to see which particular antigens they might be directed against? Dr Wheatley. In terms of these cultures, our end point was the lytic activity against the BLCLs. We did not evaluate the particular subpopulations of CTLs in this study. Prior studies in our laboratory and others have used specific peptide epitopes to help define better areas of lytic activity. I think it will be important to break down the subpopulations involved in determining the total specific lysis. Dr Francesco M. Marincola (Bethesda, Md). To continue with the question raised by Dr Abecassis, I would like to ask whether you are planning to analyze further the specificity of the CTL reactivity. Also, I am
Surgery August 1998 surprised that you did not get any CTL reactivity with EBV lines, which are wonderful antigen-presenting cells. Dr Wheatley. In response to your first question, for this particular study we were not interested in epitope mapping per se. We are looking at different and newly defined epitopes involved in the immune response for EBV cancers. In terms of why we did not get increased lytic activity against BLCLs, I think this is an important question but not one to pursue in the context of our current collaboration. BLCLs are characteristically good antigen-presenting cells, and I think in this particular study the DCs were superior to the BLCLs alone. Dr Ronald J. Weigel (Stanford, Calif). When normal people are infected with EBV, they get this polyclonal expansion that can be suppressed. However, something is different about the polyclonal lymphomas that occur. If you stop the immunosuppression in those people, they still have aggressive tumors. So something is different about the lymphomas once they occur. Even though, as your model confirms, you can generate an immune response to normal immortalized EBV cells, clearly something is different about the polyclonal lymphomas. Could you comment on how this is going to be different when you are showing that you can lyse polyclonal immortalized cells versus a true polyclonal aggressive lymphoma that occurs from EBV? Dr Wheatley. Our laboratory studied a similar situation in patients with acquired immune deficiency syndrome who are immunosuppressed through human immunodeficiency virus. We have looked at the EBVspecific, non-Hodgkin’s lymphomas in those patients and we have actually been able to demonstrate some lysis against a monoclonal malignant EBV-expressing tumor isolated from a patient. I think additional studies are necessary to help elicit the exact nature of this immune response, but I think our studies have shown that there is the potential for addressing this in malignant lymphoma in patients. Dr Nancy L. Ascher (San Francisco, Calif). I have questions again about the specificity of your CTLs and wonder whether you have examined their activity against allogeneic cells. Clinically we frequently see EBV infections actually growing along the blood vessels of the organs that we transplant. There have been some reports in the literature that the EBV infection actually comes from donor origin rather than host origin. In the setting in which you might expect a CTL response against a graft, how specific are these cells and what effects do you anticipate looking at the allogeneic system? Dr Wheatley. I would like to reiterate that I think this study is useful as a starting point in generating EBV-specific CTLs that have good lytic activity against EBV-expressing tumor cells. I think it is important to look at these subpopulations to help define subpopulations with increased lytic activity. We hope that this study will provide a launching point for additional studies of that nature.
Background. The use of immunosuppressive therapies after solid organ transplantation has been shown to increase a patient’s risk for Epstein-Barr virus (EBV)–associated lymphoma. A potential therapy for this disorder is the adoptive transfer of EBV-specific cytotoxic T lymphocytes (CTLs). We proposed that dendritic cells (DCs) could be loaded with EBV antigens and be used to improve the in vitro generation of EBVspecific CTLs. Methods. Autologous EBV-transformed B-lymphoblastoid cell lines (BLCLs) were generated from normal donors, and CTLs were initiated by culturing peripheral blood mononuclear cells with DCs alone, disrupted BLCLs alone, intact, irradiated BLCLs alone, and DCs loaded with disrupted BLCLs. Lytic activities were determined with a 4-hour chromium-release assay against autologous BLCLs, and statistical calculations were performed by a Student t test assuming equal variance. Results. The lytic activity of CTLs generated with DCs loaded with disrupted BLCLs reached 78% and was statistically significant (P < .01) at all effector/target ratios compared with CTLs generated with DCs alone, disrupted BLCLs alone, or intact BLCLs alone. Total numbers of CTLs were also greater than those of control groups for DCs loaded with disrupted BLCLs. Conclusions. DCs improved the in vitro generation of EBV-specific CTLs as evidenced by this group’s significantly increased lytic activity over that of the control group. The improved lytic activity of DC-generated EBV-CTLs suggests that adoptive transfer of these cells could lead to a more effective immunotherapeutic response against posttransplantation EBV-associated lymphoma. (Surgery 1998;124:171-6.) From the Departments of Surgery, Pathology, and Immunology, Duke University Medical Center, Durham, NC
SOLID ORGAN TRANSPLANTATION HAS become the accepted standard of therapy for a number of acquired and inherited diseases. The number of patients undergoing solid organ transplantation has been increasing steadily, and 19,017 patients in the United States underwent solid organ transplantation in 1994.1 The implementation of increasingly potent immunosuppressive therapies has allowed many allograft recipients to achieve successful outcomes. However, a potentially serious complication associated with prolonged immunosuppression in allograft recipients is the development of posttransplantation lymphoproliferative disorder (PTLD). PTLD is a frequent tumor arising in patients after transplantation and ranges from lymphoid hyperPresented at the Fifty-ninth Annual Meeting of the Society of University Surgeons, Milwaukee, Wis, Feb 12-14, 1998. Reprint requests: H. Kim Lyerly, MD, Box 2606, DUMC, Durham, NC 27710. Copyright © 1998 by Mosby, Inc. 0039-6060/98/$5.00 + 0
11/6/91179
plasia to malignant non-Hodgkin’s lymphoma (NHL).2 The incidence of PTLD in allograft recipients has paralleled the increased use of immunosuppressants cyclosporine, anti-CD3 monoclonal OKT3, and FK506.3,4 PTLD is now one of the most frequent tumors arising in patients after transplantation and totals 15% to 25% of neoplasms, compared with a 5% incidence in the general population.4 PTLD in allograft recipients is primarily of B-cell origin, and host infection with Epstein-Barr virus (EBV) is an important etiologic factor.5 A 200-kilobase double-stranded DNA virus belonging to the human herpesvirus family, EBV has been implicated in the pathogenesis of many lymphoproliferative diseases, including PTLD.6 EBV, either type 1 or 2, infects primarily resting B cells through a receptor identical to the C3d complement receptor.7,8 Once intracellular, the virus is maintained as extrachromosomal episomes. A persistent life-long infection results, and the balance between latency and immortalized B-cell proliferation rests solely on the host immune SURGERY 171
172 Wheatley et al.
Surgery August 1998
used to activate EBV-specific CTLs (EBV-CTLs) in vitro. In addition, we proposed that EBV-CTLs generated with DCs pulsed with disrupted tumor cells have increased lytic activity compared with EBV-CTLs generated with tumor cells alone.
Fig. 1. Flow cytometric analysis of DCs derived from day 8 of culture of PBMCs. A, Isotypic control; B, CD14FITC/HLA-DR-PE; C, CD18-FITC/CD80-PE; D, CD86FITC/CD83-PE.
system. When the host immune system is disrupted, as in allograft recipients receiving exogenous immunosuppression, latently infected B cells can undergo rapid polyclonal or monoclonal expansion. Although some allograft recipients with PTLD respond to first-line chemotherapy, this treatment modality is limited by excessive toxicity and frequent tumor resistance and recurrence. A potential anticancer therapy for EBV-associated NHL in patients after transplantation is immunotherapy.9 Tumor-specific CD8+ cytotoxic T lymphocytes (CTLs) are an important effector arm in antitumor immunity.10 These lymphocytes recognize short peptide epitopes derived from tumor-rejection antigens, presented by class I major histocompatability complex (MHC) molecules on the tumor cell surface. Although recognition of peptide class I complexes is sufficient to trigger lysis, priming of CTL responses requires the presentation of the relevant tumor-rejection antigen by professional antigen-presenting cells capable of providing co-stimulation. Dendritic cells (DCs) are considered to be the most potent antigen-presenting cells for mediating a variety of antitumor immune responses. These include stimulation of helper T cell–dependent antibody responses, as well as the generation of primary and secondary cytotoxic T cell responses to tumor- associated antigens.11,12 As a consequence, DCs loaded with tumor antigens may represent a potentially powerful method of inducing antigen-specific CTL expansion and resulting antitumor immunity. The objective of this study was to demonstrate that DCs could be loaded with EBV antigens and be
MATERIAL AND METHODS Generation of B-lymphoblastoid cell lines (BLCLs). After obtaining informed, written consent, healthy donors underwent leukapheresis. Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Hypaque density-gradient centrifugation, washed, and suspended at 5 × 106 cells/mL in AIM V lymphocyte activation medium (Gibco, Grand Island, NY) containing 2 µg/mL cyclosporine (Sandoz Pharmaceutical Co, East Hanover, NJ) and 2 mL viral supernatant harvested from an EBV-expressing cell line (B95.8). The solution was incubated at 37° C with 5% carbon dioxide and humidified room air for 18 hours. The PBMCs were resuspended at 2 × 106 cells/mL in AIM V medium, and the cultures were inspected daily for evidence of clonal expansion. Culture medium was added as needed, and cultures were expanded by splitting 1:2 with fresh AIM V medium. When adequate numbers of cells were available, the presence of B-cell markers CD19/CD20 and the absence of T and natural killer markers were confirmed by flow cytometry. Generation of DCs from PBMCs. After leukapheresis, fresh PBMCs were suspended in AIM V culture medium, placed in tissue culture flasks, and incubated at 37° C with 5% carbon dioxide and humidified room air for 2 hours. The nonadherent population was then removed and cryopreserved until stimulated with mature DCs. The adherent PBMCs were cultured in AIM V supplemented with 10% autologous serum and 500 units/mL human interleukin (IL)-4 (Schering-Plough) and 800 units/mL human granulocyte-macrophage colonystimulating factor (GM-CSF) (Schering-Plough). Cytokines were replaced every 48 hours. On day 8 the DCs were removed from culture and a portion was cryopreserved. Flow cytometric analysis. Phenotypic analysis of harvested, cultured DCs was determined by flow cytometry. Briefly, for dual-color immunofluorescence analyses, cells were typed with either fluorescein isothiocyanate–conjugated (FITC) mononuclear antibodies against human CD14, CD18, and CD86 and phycoerythrin-conjugated (PE) monoclonal antibodies against human HLA-DR, CD80, and CD83. The antibodies conjugated with both FITC and PE were used at saturating concentrations. Conjugated monoclonal antibodies with specificity
Wheatley et al. 173
Number of Cells (x106)
Surgery Volume 124, Number 2
Fig. 2. Total number of lymphocytes in culture. Proliferation of EBV-CTLs for each experimental condition after initial stimulation was determined by counting total number of lymphocytes beginning on day 0 and continuing through every other day of culture. EBV-CTL cultures were initiated on day 0 by culturing PBMCs with DCs alone, disrupted autologous BLCLs alone, intact irradiated autologous BLCLs alone, or DCs loaded with disrupted autologous BLCLs. Cultures were counted and restimulated with original experimental conditions on day 14.
for human immunoglobulin (Ig)G1 and IgG2 were used as isotypic controls. Cells were analyzed with a FACScan cytometric flow cytometer, and the large cell population was gated to exclude contaminating lymphocytes. Likewise, phenotypic analysis of EBVCTLs was determined with the same isotypic controls, FITC mononuclear antibodies against human CD3 and CD20, and PE mononuclear antibodies against human CD20 and CD56. Initiation and expansion of autologous EBV-CTLs. Four experimental conditions for initiating and expanding EBV-CTLs were studied (DCs alone, disrupted BLCLs alone, intact, irradiated BLCLs alone, and DCs loaded with disrupted BLCLs). Responder/stimulator ratios of 100:1 were used to initiate all conditions. Autologous BLCLs were disrupted by suspending washed BLCLs at a concentration of 10 × 106 cells/mL, subjected to three cycles of freezing to –70° C and thawing to +40° C, and irradiated with 5000 rads by a cesium source. Intact, irradiated BLCLs were washed, resuspended at 10 × 106 cells/mL, and irradiated with 5000 rads by a cesium source. DCs were loaded with disrupted BLCLs by coculturing DCs with disrupted BLCL solution (1 DC/1 BLCL equivalent) in siliconized glass tubes in 0.5 mL AIM V for 16 hours at 37° C with 5% carbon dioxide and humidified room air. All EBV-CTL cultures were initiated in AIM V medium containing AIM V medium with 10% autologous serum in the presence
of 100 units/mL human IL-2 (Proleukin) and 10 ng/mL human IL-7 (Gibco). Cytokines and fresh medium were added every 48 hours beginning on day 4. Cultures were restimulated according to identical experimental conditions on day 14. Cytotoxicity assay. Cytolytic activities of EBVCTLs were determined with a standard 4-hour 51Crrelease assay. Briefly, target cells, consisting of autologous BLCLs, were labeled with 51Cr for 90 minutes. The targets were washed and added to a 96-well round-bottomed tissue culture plate as 5000 targets/well in 100 µL RPMI (Gibco) with 10% autologous serum. EBV-CTLs for each experimental condition were washed and added to target wells at effector/target (E/T) ratios of 10:1, 20:1, and 40:1. All assays were performed in triplicate. After 4 to 5 hours of incubation, the supernatant fluids were collected and counted in a Wallac gamma counter. Percent specific lysis was calculated according to the following formula: Percent specific lysis = ([Observed release – Spontaneous target release]/[Maximal target release – Spontaneous target release]) × 100. Maximal target release was obtained by incubating targets with 100 µL 10% Triton-X solution, and spontaneous target release was obtained by incubating targets with 100 µL culture medium for the full 4-hour period. Statistical analysis. The lytic activities of EBVCTLs for each experimental group were compared
174 Wheatley et al.
Surgery August 1998
Fig. 3. Cytotoxicity assays. Lytic activities of each experimental group of EBV-specific CTLs were determined on day 24 of culture against autologous BLCLs with standard 4-hour chromium-release assay. EBV-CTL cultures were initiated on day 0 by culturing PBMCs with DCs alone, disrupted autologous BLCLs alone, intact irradiated autologous BLCLs alone, or DCs loaded with disrupted autologous BLCLs. Cultures were restimulated with original experimental conditions on day 14, and lytic activities were determined at E/T ratios of 10:1, 20:1, and 40:1.
at E/T ratios of 10:1, 20:1, and 40:1 with a Student t test, assuming equal variance. RESULTS Dendritic cells. PBMCs were obtained from healthy donors. The adherent PBMC fraction was cultured for 1 week in GM-CSF and IL-4 containing culture medium. Fig. 1 shows a representative cytometrogram of the phenotype of cultured DCs harvested on day 8. The mature DCs were harvested on day 8 and analyzed for the presence of surface markers MHC class I, MHC class II, and CD86 by flow cytometry. DCs generated under the same growth conditions from many preparations of other PBMCs reproducibly expressed high levels of HLA-DR and CD86 but were negative for CD14. B-lymphoblastoid cell lines. After establishment of an autologous BLCL, the resulting cells were phenotyped for determination of B-cell markers. The cells had very high levels of CD19/CD20, phenotypic markers for B cells, and an absence of CD3 and natural killer phenotypic markers on flow cytometric analysis. In vitro expansion of EBV-CTLs. EBV-CTL cultures were initiated by stimulating the nonadherent PBMC fraction with DCs alone, disrupted BLCLs alone, intact, irradiated BLCLs alone, or DCs loaded with disrupted BLCLs. Fig. 2 shows the numeric expansion of EBV-CTLs for each experimental condition. The only condition that had a net expansion of EBV-CTLs on day 24 of culture,
after a restimulation on day 14, were those initiated with DCs loaded with disrupted BLCLs. There was a 50% increase in total number of cells on day 24 compared with the number of PBMCs on day 0, whereas all other conditions had a net decline. Cytotoxicity assays. A 51Cr cytotoxicity-release assay was performed on day 24 after a restimulation on day 15. Fig. 3 shows the percent specific lysis of each category of EBV-CTLs (effector) against autologous BLCLs (targets) at E/T ratios of 10:1, 20:1, and 40:1. Percent specific lysis by EBV-CTLs stimulated with DCs loaded with disrupted tumor cells was greater than each of the other categories at all E/T ratios (P < .01, Student t test). The background natural killer activity was minimal for all groups examined. The lytic activity of EBV-CTLs stimulated with DCs loaded with disrupted BLCLs against K562 cells was less than 5% at each E/T ratio. DISCUSSION Although the EBV genome contains more than 100 genes and encodes more than 80 proteins, B cell transformation is mediated specifically by 6 nuclear antigens (EBNAs 1, 2, 3A, 3B, 3C, and LP) and 2 membrane-derived proteins (LMP1 and LMP2).13 CD8+ CTLs are the immune effector cells responsible for eliminating EBV-infected B cells and recognize EBNA 3A, 3B, 3C, and LMP antigens expressed on the cell surface in the context of MHC class I molecules.14-16 These antigens are processed individually within the cytoplasm of antigenpresenting cells into smaller epitope fragments and
Wheatley et al. 175
Surgery Volume 124, Number 2
are then presented on the cell surface, in the context of MHC class I, to CD8+ T lymphocytes. Ex vivo expansion and adoptive transfer of EBVspecific CTLs offer the possibility of enhancing the cellular immune response against previously transformed B cells. EBV-associated NHL in patients after transplantation results from intentional exogenous suppression of the cellular immune system and manifests physiologically as a deficit in immune surveillance. The ongoing balance between the number of functioning EBV-specific CTLs and the rate of EBV-transformed B-cell proliferation is critical to preventing malignancy. Selectively restoring functional aspects of the immune system in immunosuppressed patients through adoptive transfer of EBV-specific CTLs could augment the depleted immune system and potentially target established B-cell lymphomas.17 DCs are present in small numbers in most tissues including skin, lung, liver, spleen, blood, lymphoid organs, and bone marrow.18 Morphologically, they are large cells with elongated, stellate processes called dendrites. They lack cell surface markers typical for B, T, natural killer, and monocytemacrophage lineages but express high levels of MHC classes I and II, as well as co-stimulatory molecules. These cells have been shown to internalize, process, and present soluble antigen as peptides in conjunction with MHC classes I and II.19,20 In addition, DCs have the unique ability to cluster naive T cells and respond to antigen by rapid up-regulation of the expression of MHC and co-stimulatory molecules, the production of cytokines, and migration toward lymphatic organs.20,21 The method of presentation of EBV antigens for in vitro CTL expansion is critical to generating highly specific CTLs that efficiently target and lyse EBV-expressing tumor cells. Intact, whole B lymphocytes can act as antigen-presenting cells and possess the requisite MHC class I and co-stimulatory molecules to activate CD8+ T lymphocytes. In addition, the in vitro pattern of EBV antigen expression by EBV-transformed BLCLs mimics that expressed by B cell neoplasms in vivo. Although intact BLCLs possess both EBV antigens and MHC class I molecules, they are not the optimal method for presenting EBV antigens to CTLs. Our results indicate that DCs improve the activation and expansion of EBV-CTLs in vitro. DCs pulsed with disrupted BLCLs appear to generate EBV-CTLs with increased EBV-specific lytic activities at all E/T ratios examined. Most likely, this is related to the improved antigen-presenting capabilities of DCs compared with BLCLs. The additional co-stimulatory molecules present on DCs, combined with
the enhanced antigen-processing capabilities of DCs, indicate that DCs can be used for the generation of EBV-CTLs in vitro. Although the subpopulations of EBV-specific CTLs contained in the EBVCTL lymphocyte cultures were not specifically determined in this study, the optimal strategy for adoptive immunotherapy is to use a bulk culture approach to the expansion of EBV-CTLs. It is possible that the EBV-CTLs did undergo some expansion in the bulk cultures used in this study, as demonstrated by the increased EBV-specific lytic activity of CTLs against EBV-expressing BLCLs. Future studies will focus on improving the loading of DCs with disrupted BLCLs. Potential improvements include use of cationic liposome preparations, in combination with disrupted tumor cells, to augment loading of DCs and optimize the co-culture conditions between PBMCs and pulsed DCs. The improved lytic activity of DC-generated EBV-CTLs suggests that adoptive transfer of these cells could lead to a more effective immunotherapeutic response in vivo against posttransplantation EBV-associated lymphoma. REFERENCES 1. United Network for Organ Sharing, Richmond, Va, 1994. 2. Basgoz N, Preiksaitis JK. Post-transplant lymphoproliferative disorder. Infect Dis Clin North Am 1995;9:901-23. 3. Lucas KG, Pollok KE, Emanuel DJ. Post-transplant EBV induced lymphoproliferative disorders. Leuk Lymphoma 1997;25:1-8. 4. Boubenider S, Hiesse C, Goupy C, Kriaa F, Marchand S, Charpentier B. Incidence and consequences of post-transplantation lymphoproliferative disorders. J Nephrol 1997;10:136-45. 5. Mcknight JL, Cen H, Riddler SA, Breinig MC, Williams PA, Ho M, et al. EBV gene expression, EBNA antibody responses and EBV+ peripheral blood lymphocytes in posttransplant lymphoproliferative disease. Leuk Lymphoma 1994;15:9-16. 6. Young L, Alfieri C, Hennessy K, Evans H, O’Hara C, Anderson KC, et al. Expression of Epstein-Barr virus transformation–associated genes in tissues of patients with EBV lymphoproliferative disease. N Engl J Med 1989;321:1080-5. 7. Nemerow GR, Mold C, Schwend VK, Tollefson V, Cooper NR. Identification of gp350 as the viral glycoprotein mediating attachment of Epstein-Barr virus (EBV) to the EBV/C3d receptor of B cells: sequence homology of gp350 and C3 compliment fragment C3d. J Virol 1987;61:1461-70. 8. Thorley-Lawson DA, Geilinger K. Monoclonal antibodies against the major glycoprotein (gp350/220) of Epstein-Barr virus neutralize infectivity. Proc Natl Acad Sci USA 1980;77:5307-11. 9. Heslop HE, Rooney CM. Adoptive cellular immunotherapy for EBV lymphoproliferative disease. Immunol Rev 1997;157:217-22. 10. Townsend A, Bodmer H. Antigen recognition by class I–restricted T-lymphocytes. Ann Rev Immunol 1989;7:601-24. 11. Steinman RM. The dendritic cell system and its role in immunogenicity. Annu Rev Immunol 1991;9:271-96.
176 Wheatley et al.
12. Sornasse T, Flamand V, De Becker G, Bazin H, Tielemans F, Tielemans K, et al. Antigen- pulsed dendritic cells can efficiently induce an antibody response in vivo. J Exp Med 1992;175:15-21. 13. Young L, Alfieri C, Hennessy K, Evans H, O’Hara C, Anderson KC, et al. Expression of Epstein-Barr virus transformation–associated genes in tissues of patients with EBV lymphoproliferative disease. N Engl J Med 1989;321:1080-5. 14. Murray RJ, Kurilla MG, Brooks JM, Thomas WA, Rowe M, Kieff E, et al. Identification of target antigens for the human cytotoxic T cell response to Epstein-Barr virus (EBV): implications for the immune control of EBV-positive malignancies. J Exp Med 1992;176:157-68. 15. Murray RJ, Wang D, Young LS, Wang F, Rowe M, Kieff E, et al. Epstein-Barr virus- specific cytotoxic T-cell recognition of transfectants expressing the virus-coded latent membrane protein LMP. J Virol 1988;62:3747-55. 16. Burrows SR, Sculley TB, Misko IS, Schmidt C, Moss DJ. An Epstein- Barr virus–specific cytotoxic T-cell epitope in EBNA 3. J Exp Med 1990;171:345-50. 17. Rickinson AB, Murray RJ, Brooks J. T-cell recognition of Epstein-Barr virus–associated lymphomas. In: Franks LM, editor. Cancer surveys, a new look at tumor immunology, vol 13. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 1992. p 53-80. 18. Inaba K, Metlay J, Crowley M, Witmer-Pack M, Steinman RM. Dendritic cells as antigen presenting cells in vivo. Int Rev Immunol 1990;6:197-206. 19. Crowley M, Inaba K, Steinman R. Dendritic cells are the principal cells in mouse spleen bearing immunogenic fragments of foreign proteins. J Exp Med 1990;172:383-90. 20. Steinman R, Witmer-Pack M, Inaba K. Dendritic cells: antigen presentation, assessory function and clinical relevance. Adv Exp Med Biol 1993;329:1-9. 21. Inaba K, Romani N, Steinman R. An antigen-independent contact mechanism as an early step in T cell proliferative responses to dendritic cells. J Exp Med 1989;170:527-42.
DISCUSSION Dr Michael Abecassis (Chicago, Ill). Stan Rodell has shown that adoptive immunotherapy can be effective in cytomegalovirus disease. Are these populations of CTLs specifically directed against an epitope? In other words, there are different viral antigens that are responsible for different phases of the lytic infection. Have you looked into the populations of CTLs to see which particular antigens they might be directed against? Dr Wheatley. In terms of these cultures, our end point was the lytic activity against the BLCLs. We did not evaluate the particular subpopulations of CTLs in this study. Prior studies in our laboratory and others have used specific peptide epitopes to help define better areas of lytic activity. I think it will be important to break down the subpopulations involved in determining the total specific lysis. Dr Francesco M. Marincola (Bethesda, Md). To continue with the question raised by Dr Abecassis, I would like to ask whether you are planning to analyze further the specificity of the CTL reactivity. Also, I am
Surgery August 1998 surprised that you did not get any CTL reactivity with EBV lines, which are wonderful antigen-presenting cells. Dr Wheatley. In response to your first question, for this particular study we were not interested in epitope mapping per se. We are looking at different and newly defined epitopes involved in the immune response for EBV cancers. In terms of why we did not get increased lytic activity against BLCLs, I think this is an important question but not one to pursue in the context of our current collaboration. BLCLs are characteristically good antigen-presenting cells, and I think in this particular study the DCs were superior to the BLCLs alone. Dr Ronald J. Weigel (Stanford, Calif). When normal people are infected with EBV, they get this polyclonal expansion that can be suppressed. However, something is different about the polyclonal lymphomas that occur. If you stop the immunosuppression in those people, they still have aggressive tumors. So something is different about the lymphomas once they occur. Even though, as your model confirms, you can generate an immune response to normal immortalized EBV cells, clearly something is different about the polyclonal lymphomas. Could you comment on how this is going to be different when you are showing that you can lyse polyclonal immortalized cells versus a true polyclonal aggressive lymphoma that occurs from EBV? Dr Wheatley. Our laboratory studied a similar situation in patients with acquired immune deficiency syndrome who are immunosuppressed through human immunodeficiency virus. We have looked at the EBVspecific, non-Hodgkin’s lymphomas in those patients and we have actually been able to demonstrate some lysis against a monoclonal malignant EBV-expressing tumor isolated from a patient. I think additional studies are necessary to help elicit the exact nature of this immune response, but I think our studies have shown that there is the potential for addressing this in malignant lymphoma in patients. Dr Nancy L. Ascher (San Francisco, Calif). I have questions again about the specificity of your CTLs and wonder whether you have examined their activity against allogeneic cells. Clinically we frequently see EBV infections actually growing along the blood vessels of the organs that we transplant. There have been some reports in the literature that the EBV infection actually comes from donor origin rather than host origin. In the setting in which you might expect a CTL response against a graft, how specific are these cells and what effects do you anticipate looking at the allogeneic system? Dr Wheatley. I would like to reiterate that I think this study is useful as a starting point in generating EBV-specific CTLs that have good lytic activity against EBV-expressing tumor cells. I think it is important to look at these subpopulations to help define subpopulations with increased lytic activity. We hope that this study will provide a launching point for additional studies of that nature.