- Email: [email protected]
Selective elimination of alloreactivity from immunotherapeutic T cells by photodynamic cell purging and memory T-cell sorting
Cytotherapy (2005) Vol. 7, No. 2, 126 /133
Selective elimination of alloreactivity from immunotherapeutic T cells by photodynamic cell purging and memory T-cell sorting NT Le, BJ Chen and NJ Chao Divisions of Cellular Therapy/Bone Marrow Transplantation and Medical Oncology, Departments of Medicine and Immunology, Duke University Medical Center, Durham, North Carolina, USA
Allogeneic stem cell transplantation (alloSCT), especially in the mismatched setting, carries a high risk of life-threatening GvHD because of activation of donor T cells by Ag present on host cells. Removal of mature donor T cells can prevent GvHD but leads to delayed immune reconstitution, and an increased incidence of opportunistic infections and disease relapse. These findings demonstrate the vital role of donor T cells in providing graft-versus-tumor (GvT) and anti-pathogen effects as well as facilitating immune reconstitution. It has been well documented that GvHD can be separated from GvT effects, making it possible potentially to eliminate GvHD while preserving the immunotherapeutic benefits of donor T cells. Over the past decade, major attempts have been made to reduce
GvHD incidence without loss of GvT effect, especially in the haploidentical setting. Novel techniques to deplete host-reactive donor T cells selectively have been explored. This review focuses on the use of the photodynamic cell purging (PDP) process and of sorting memory T cells for the selective elimination of alloreactivity. Minimizing the threat of GvHD while maximizing the beneficial GvT effect would broaden the scope and effectiveness of alloSCT.
Introduction
can prevent GvHD but leads to a delay in immune reconstitution, leaving the recipient at great risk of potentially fatal opportunistic infections [5/7]. Moreover, total T-cell depleted (TCD) stem cell grafts are associated with higher rates of graft rejection and disease relapse [2], demonstrating the important role of donor T cells in GvT effects. Strategies in which T cells causing GvHD are selectively depleted, while T cells capable of mounting responses to viral and tumor Ag are preserved, are thus highly desirable for clinical use. Over the past decade, novel techniques to eliminate alloreactive T cells selectively, either by their removal or by destruction, ex vivo or in vivo, have been explored [8]. As part of this effort, our group has investigated two approaches for the selective depletion of alloreactive donor T cells: by PDP and by sorting memory T cells. Here we review these two
Allogeneic stem cell transplantation (SCT) is now recognized as a curative treatment for various hematologic malignancies, and has been used in the treatment of solid tumors and inherited diseases [1]. Allografts exert their anti-tumor activity by a graft-versus-tumor (GvT) effect, mediated mainly by donor cytotoxic T cells, which recognize either leukemia-specific Ag or alloantigens present on normal and malignant host cells [2]. Lifethreatening GvHD develops when donor T cells that accompany the stem cell graft are activated by Ag expressed on normal host cells, causing severe damage to target tissues. While conventional GvHD prophylaxis with immunosuppressive agents can be effective in genotypically HLA-identical sibling transplants, prophylaxis does not always prevent severe GvHD, especially in mismatched SCT [3,4]. Removal of mature donor T cells
Keywords Allodepletion, photodynamic cell purging, naive T cells, memory T cells, stem cell transplantation.
Correspondence to: Nelson J Chao, MD, Division of Cellular Therapy/BMT, Duke University Medical Center, 2400 Pratt St., Suite 1100, Box 3961, Durham, NC 27705, USA. – 2005 ISCT
DOI: 10.1080/14653240510018163
Selective depletion of alloreactive T cells using PDP and memory T cells
techniques. Their potential benefits and disadvantages will be discussed in an attempt to identify the most suitable clinical approach.
(a)
B H2N
PDP treatment prevents GvHD but preserves the GvT effect in animal models We first tested the effects of PDP treatment in mice; the results of these studies were published in 2002 [15]. We found that treatment with PDP inhibited anti-host responses as measured by cytotoxic T lymphocyte (CTL) assay and IFN-g production. In contrast, anti-tumor and
B NH2Cl
O
Ex vivo PDP treatment to remove alloreactive T cells selectively Photodynamic therapy (PDT) has been used to treat malignant disease since the turn of the century. The technique involves the in vivo administration of a photosensitizing agent, innocuous in its native state, followed by activation of the agent by light of appropriate wavelength and energy level. The activation of the photosensitizing agent is thought to result in the generation of toxic oxygen derivatives and reactive oxygen species, which leads to cellular damage and eventually cell death. Rhodamine dyes, with their low toxicity and rapid clearance, are potentially useful photosensitizers for PDT [9]. A photoactivable rhodamine derivative, 4,5-dibromorhodamine-methyl ester (TH9402) (Figure 1a), has been used to purge chronic myelogenous leukemia [10,11], NHL [12] and CLL [12] in autologous BM grafts. TH9402 is taken up by all cells, but is actively extruded from the cytoplasm by the active multi-drug transporter P-glycoprotein 170 (P-gp 170) [13]. However, P-gp 170 becomes inactivated upon T-cell activation, leading to the retention of the photosensitizer TH9402 in the mitochondria of activated T cells [13,14]. Following exposure to visible light (514 nm), the dye becomes cytotoxic, resulting in cell death [13,14]. Thus, when the mixture of resting and activated T lymphocytes is treated with TH9402 and then exposed to visible light, only the activated T cells are eliminated. Our group has investigated the potential of this procedure, PDP, for the selective elimination of donor alloantigen-specific T cells. Briefly, donor T cells are activated by host APC in a one-way MLC. Following activation, these cells are treated with TH9402 ex vivo and subsequently exposed to visible light for elimination of activated alloresponsive T cells (Figure 1b).
127
COOCH3
4,5-dibromorhodamine 123 (TH9402)
(b)
one-way MLC
TH9402 40 min
Incubation 90 min
Visible light
GVH Necrosis GVT & others Intact
Figure 1. (a) Chemical structure of 4,5-dibromorhodamine-methyl ester (TH9402). (b) A schematic diagram of PDP treatment. Donor T cells are activated by host APC in a one-way MLC. After in vitro activation, cells are treated ex vivo with TH9402 for 40 min at 378C. Cells are then washed and kept at 378C for 90 min to allow extrusion of TH9402 from resting, non-activated T cells. Immediately following the incubation, cells are exposed to visible light energy (514 nm) delivered by the scanning lamp device Theralux (Theratechnologies, Montreal, Canada). As TH9402 is selectively retained in the mitochondria of activated cells, PDP treatment results in the specific elimination of activated, host-responsive T cells.
anti-third party responses were essentially conserved. We further demonstrated, in a murine model, that while all tumor-bearing mice transplanted with only TCD BM cells had leukemic cells in peripheral blood and more than half died within 40 days after transplantation, the addition of PDP-treated cells allowed 90% of the recipients to survive for more than 100 days, without detectable tumor cells. Moreover, these mice were healthy, free of GvHD and had full phenotypic recovery of immune cells. These data suggest that PDP can selectively eliminate host alloantigen-specific T cells responsible for GvHD while preserving the GvT effect.
Problems facing human pre-clinical studies Encouraged by the results obtained from animal studies, we have undertaken a pre-clinical study to assess the feasibility of using the ex vivo TH9402 PDP process to
128
NT Le et al .
purge GvHD-inducing T cells before transfusing donor T cells into patients. We have encountered several problems Finding the optimal duration for the primary MLC Studies using CFSE-labeled cells have revealed with great precision the kinetics and degree of alloreactive T-cell clonal expansion following allogeneic stimulation in vivo [16]. During the first 30 h post-stimulation, cell division is highly synchronous. However, over the next 24 h individual responding T cells proliferate asynchronously, some daughter cells achieving only one cell division while others undergoing five rounds of cell division. The diversity of proliferative behavior is maintained throughout the entire phase of clonal expansion, with the average responding precursor giving rise to 64 daughter cells through six rounds of cell division within 72 h of stimulation. The majority of the methods employed for the depletion of GvHD-inducing T cells identify these T cells based on their activation status after ex vivo stimulation in an allogeneic MLC. In the published studies, the duration of the primary MLC has ranged from 2 /7 days. It is currently unclear what can be judged as the optimal duration for primary MLC. Longer culture times would probably result in the activation of more alloresponsive T cells but at the cost of lower cell viability. Furthermore, if we can extrapolate the results of in vivo CFSE-labeling studies to ex vivo MLC, longer culture would lead to greater numbers of daughter cells arising from a single alloreactive precursor that would need to be eradicated. In the PDP approach, this would require more reagent and a higher efficiency of depletion. We are therefore using a 3day primary MLC, although the peak of proliferative response for human PBMC is on days 4 /5, based on 3Hthymidine incorporation assay. Selecting the most efficient stimulator cells A major limitation of current approaches for allodepletion is the source of recipient stimulator cells used in alloactivation. The data accumulating thus far suggest that crude PBMC preparations are inefficient at Ag presentation, presumably because of their heterogeneity. To circumvent this problem, some investigators have used DC as the stimulators. One study reported that HLAmismatched PBMC stimulators consistently resulted in 90/95% depletion of alloreactivity, whereas purified matured DC could produce a 100-fold (99%) reduction of response to alloantigens [17]. In practice, however,
purified DC are difficult to prepare in adequate number for clinical trials. Furthermore, they require expensive cytokines for culture and Ab for cell sorting. Another concern is that, in leukemic patients, PBMC may be contaminated with leukemic blasts, or alternatively DC may have processed leukemic Ag, resulting in a loss of anti-tumor activity after allodepletion. To address this problem, one study used EBV-transformed lymphoblastoid cell lines (LCL) as stimulator cells [18]. These investigators reported a more consistent depletion of alloreactivity when LCL were used for stimulation instead of PBMC, although the residual response remained significant. LCL have several advantages as stimulator cells: they are better at Ag presentation than PBMC, are relatively inexpensive to prepare, can be easily expanded in vitro, and do not express tumor Ag that may serve as targets for GvT effects. In our hands, LCL irradiated at 90 Gy still retained a residual proliferative response, as assessed by 3H-thymidine incorporation assay The development of EBVinduced lymphoproliferative disorders after SCT is, therefore, a major concern. Because of the limitations of LCL, and the possibility that when using PBMC the residual alloproliferative response could lead to GvHD, we are considering the generation of purified mature DC from patients for use in clinical trials. Finding reliable and practically applicable read-out assays A major problem that we have encountered is finding a reliable human in vitro model or assay to predict GvHD. We found that primed PDP-treated T cells retain residual proliferative responses to alloantigens in an MLC but do not elicit cytotoxicity towards the same allogeneic targets (Figure 2). One possible explanation for these findings is that the residual alloantigen-responsive T cells, as assessed in a proliferation assay, may not be true alloreactive T cells. Thus, 3H-thymidine incorporation, although inexpensive and easy to perform, may not be an appropriate assay for evaluating the residual alloantigen response in primed PDP-treated T cells. CTL and cytotoxic T-lymphocyte precursor (CTLp) assays have been viewed as the ‘gold standard’ in assessing T-cell alloreactivity. However, these assays are laborious and require prolonged culture, as well as use of radioactive materials. In order to assess the efficiency of allodepletion after PDP treatment, we have used the recently developed granzyme B (GrB) enzyme-linked immunospot
Selective depletion of alloreactive T cells using PDP and memory T cells
80000
unprimed, untreated 0 uM 20 uM
unprimed, untreated 0 uM 20 uM 3rd-party primed
(a)
70000 60000
20 % s pec i f ic r eleas e
cpm
50000 40000 30000 20000 10000 0 0.313
0.625
1.25
2.5
responder (x10E5)
129
(b)
15 10 5 0 -5
6.25 to 1
12.5 to 1
25 to 1
E:T ratio
Figure 2. PDP-treated T cells retain some residual proliferative response to alloantigens (a) but do not elicit cytotoxicity towards same-party targets (b). PBMC from a healthy donor (responder cells) were stimulated with irradiated (30 Gy) PBMC obtained from another healthy donor (stimulator cells) in an MLC. Following 3 days of stimulation, cells were PDP-treated at various concentrations of TH9402. Cells were then rested for 3 days before being used in subsequent assays. The proliferation assay was performed using graded numbers (0.313 /2.5 /105) of responder cells and 2.5 /105 irradiated same-party stimulator cells per well in a 96-well flat-bottomed plate. Cytotoxicity was measured by the standard 4-h 51 Cr-release assay. Responder cells that were neither stimulated in primary MLC nor PDP-treated were used as a control. In addition, responder cells primed with a different (designated third-party) stimulator were used as a control for the specificity of the CTL assay. Cells in both assays were collected after a 4-day MLC.
(ELISPOT) assay in addition to a CTLp assay. When the GrB ELISPOT was directly compared with the IFN-g ELISPOT and 51Cr-release assays, excellent cross-correlation between all three assays was shown [19]. Moreover, titration studies demonstrated a strong correlation between the number of effector cells and GrB spots per well, thus providing an estimation of cytotoxic effector cell frequency [19]. Using our PDP protocol, PDP treatment leads to a TH9402 dose-dependent reduction of alloreactive cytotoxic T-cell frequency and, at 20 mm of TH9402, an average 2-log depletion is obtained (Figure 3). Given the high sensitivity and specificity of the GrB ELISPOT assay, it might be useful for clinical trials where there are limited numbers of patient cells available for immunologic monitoring. In addition to these assays, the skin explant model has been developed and used quite extensively elsewhere to predict the occurrence of GvHD. This assay involves incubating the recipient skin biopsies with donor T cells
previously activated by the recipient’s APC in a one-way MLC. The GvH reaction is then assessed by examining the histopathologic changes of the skin explant. Although several studies have reported statistically significant correlations between skin explant results and the in vivo development of GvHD, individual variations have been observed [20]. More recent reports suggest that a combination of cytokine gene polymorphisms and skin explant results might be more predictive [21]. We are in the process of using the skin explant model to test the ability of primed PDP-treated donor T cells to induce GvHDlike histopathologic changes in the recipient’s skin biopsies. Of interest, it would be of great benefit to develop a human immune system in an animal model that could be used to evaluate GvHD. Traggiai et al . [22] recently showed that intrahepatic injection of human cord blood stem cells into a murine model lacking B, T and NK cells can lead to the de novo development of human B, T and DC as well as the production of functional immune responses.
130
NT Le et al .
unprimed, untreated
100000
0 uM
80000 cpm
900 (b) 800
(a)
GrB ELISPOTs
120000
10 uM
60000
20 uM
40000
700 600 500
unprimed, untreated 0 uM 10 uM 20 uM
400 300 200
2. 5
1.2 5
0.6 3
0.3 2
0.1 6
0
0.0 8
0 0.0 4
100
0.0 2
20000
responder (x10E5)
0.02 0.04 0.08 0.16 0.32 0.63 1.25 2.5 responder (x10E5)
Figure 3. PDP treatment significantly reduces the frequency of alloreactive cytotoxic T cells. (a) Proliferative response to same-party stimulator cells. (b) Reduction of alloreactive cytotoxic T-cell frequency as assessed by GrB ELISPOT assay. Cells were stimulated and PDP-treated as described in Figure 2. The proliferation assay was performed using graded numbers (0.02 /2.5 /105) of responder cells and 1.25 /105 irradiated same-party stimulator cells per well in a 96-well flat-bottomed plate. The GrB ELISPOT assay was performed according to the manufacture’s protocol (R&D Systems, Minneapolis, MN, USA). Responder cells that were neither stimulated in primary MLC nor PDP-treated were used as a control. Cells in both assays were collected after a 18-h MLC.
This approach could be developed as a valuable model for predicting human clinical GvHD. The potential risks and benefits of adding-back PDP-treated donor T cells The main concern with adding-back primed, PDP-treated donor T cells is the potential exacerbation of GvHD. It has been well documented that memory T cells have a lower activation threshold than naive T cells and, once reactivated, they produce cytokines, proliferate and differentiate into effector T cells within hours. Thus, it is possible that the post-PDP residual alloreactive T cells might quickly become activated upon transferring into the recipient, causing acute GvHD. However, the results of our murine studies demonstrated that, although the PDPtreated T cells retained some residual alloreactivity, as determined by IFN-g ELISPOT assay, transferring these T cells into the animals did not cause clinical GvHD. Moreover, the addition of PDP-treated donor T cells to tumor-bearing TCD BM-transplanted animals improved survival, prevented relapse and allowed full immunophenotypic recovery. If we could extrapolate data from animal studies to human patients, these results suggest that PDPtreated donor T cells could be transferred to posttransplant patients to facilitate engraftment and immune reconstitution as well as to provide GvT effects without increasing GvHD.
Selective depletion of naive T cells Naive versus memory T cells Peripheral T cells can be broadly divided into two groups: T cells that have never been activated by Ag (naive T cells), and Ag-experienced T cells (memory T cells). In mice, the lymphocyte homing receptor CD62L (L-selectin) can be used to distinguish these two groups of T cells: while naive T cells are CD62L , memory T cells can be either CD62L or CD62L. In humans, CD45RA is differentially expressed in naive and memory T cells: while human naive T cells are CD45RA , human memory T cells can be either CD45RA or CD45RA .
CD62L memory T cells confer GvT effects but do not cause GvHD in animal models We and others [23 /25] recently reported a novel approach to prevent GvHD by selectively depleting naive T cells. This approach was based on the following hypothesis: naive T cells, which express CD62L, are T cells that have never encountered Ag specific for their T-cell Ag receptor (TCR). Memory/activated T cells, which can be either CD62L or CD62L, are T cells that have been exposed to their corresponding Ag. If a donor has never encountered the host alloantigens, potential GvHD-inducing alloreactive T cells should be contained within the donor naive T-cell compartment. Thus, CD62L T cells, which
Selective depletion of alloreactive T cells using PDP and memory T cells
Human unprimed CD45RA memory T cells do not possess a cytotoxic response to alloantigens It has been postulated that there is an immunologic crossreactivity between allogeneic MHC molecules and environmental Ag being presented by autologous APC. This leads to a legitimate concern for human application of this approach because in humans, as opposed to laboratory mice raised in sterile conditions, donor T cells might have encountered environmental Ag cross-reacting with alloantigens present in the recipient. Thus, donor T cells might
250000
(a)
c pm
200000 150000 100000
Bulk CD45RA+ CD45RA-
50000 0 4
5
6
7
Days in culture
8
contain some memory T cells that are host reactive. To investigate this possibility, we compared the ability of memory T cells to respond to alloantigens with those of naive and bulk T cells. Purified T cells were first obtained from peripheral blood of healthy donors, then separated into memory and naive T-cell subsets based on the expression of CD45RA (memory, CD45RA ; naive, CD45RA ). The abilities of these T-cell subsets to respond to alloantigens were tested using proliferation and cytotoxicity assays. We found that, in contrast to the data obtaining from the mice, human memory T cells proliferated equally as well as naive and bulk T cells in MLC (Figure 4a). However, these same memory T cells failed to kill the allogeneic targets despite the vigorous proliferative responses against the same alloantigens (Figure 4b). It is possible that human CD45RA memory T cells do not contain true alloantigen-specific T cells or that they do not possess the effector functions necessary for the initiation of GvHD. Thus, transferring donor CD45RA memory T cells into the recipient may not cause GvHD.
The potential risks and benefits of transferring human memory T cells Transferring donor memory T cells into patients posttransplant would have several advantages. First is the ease
% s pec ific r eleas e
are devoid of naive T cells and represent a subset of memory/activated T cells, should not be able to induce GvHD. We showed that, in unprimed mice, CD62L T cells failed to proliferate in response to alloantigens and were unable to induce GvHD in allogeneic hosts [23]. Furthermore, CD62L T cells contributed to T-cell reconstitution by peripheral expansion as well as by promoting T-cell regeneration from BM stem/progenitor cells. Importantly, CD62L T cells from the animals previously primed with a tumor cell inhibited the tumor growth in vivo but were unable to induce GvHD in the third-party recipients. These results suggest that it is possible to selectively prime CD62L T cells in normal donors to carry the memory T cells that we desire and transfer them to the recipients without causing GvHD.
131
40
*
Bulk CD45RA+ CD45RA-
(a)
30 20 10 0 12.5 to 1
25 to 1
50 to 1
100 to 1
E:T ratio
Figure 4. The human CD45RA memory T-cell compartment does not contain alloantigen-specific T cells if it has never been exposed to those alloantigens. (a) Human memory T cells proliferated as well as naive and bulk T cells did in MLC. (b) Human CD45RA memory T cells, however, did not display a cytotoxic response towards alloantigens. Purified T cells were first obtained from peripheral blood from healthy donors and then separated into memory and naive T-cell subsets based on the expression of CD45RA (memory, CD45RA ; naive, CD45RA ). Memory T cells were subsequently tested for their ability to respond to alloantigens using proliferation and cytotoxicity assays, in comparison with naive and bulk T cells. The proliferation assay was performed using 1.25 /105 responder cells and 5 /105 irradiated stimulator cells per well in a 96-well flat-bottomed plate. Cytotoxicity was measured by a standard 4-h 51Cr-release assay after a 5-day MLC. *The values of percentage specific release by the T cells activated by third-party alloantigens were 6.8%, 8.5%, 9.5% and 5.6% for E:T ratios of 12.5:1, 25:1, 50:1 and 100:1.
132
NT Le et al .
of selection and availability of GMP-graded fluorescencelinked anti-CD45RA Ab that would facilitate translating the research from the bench to a clinical trial. Second, unlike PDP treatment, donor CD45RA memory T cells are not primed in an allogeneic MLC. Furthermore, as revealed by our preliminary data, although the CD45RA memory T cells display vigorous proliferative responses against the same alloantigens, they do not elicit cytotoxicity to the same-party targets, suggesting that they may not cause GvHD upon in vivo transfer. Lastly, but most importantly, the finding that CD62L memory T cells from the animals previously primed with a tumor cell line were capable of inhibiting the tumor growth in vivo demonstrated that these memory T cells retain the GvT effect and perhaps other memory T-cell functions. It further suggests a strategy of avoiding GvHD by selective priming of T cells in normal donors to carry the desired memory T cells followed by their transfer to the recipient. It may prove difficult to prime the donor with recipient tumor cells in some cases. However, it has been shown that healthy donors can be immunized with myeloma immunoglobulin from the plasma of the recipient and developed a specific T-cell response [26].
Conclusion Donor T cells selectively depleted of GvHD-inducing ability have great clinical potential. Allodepletion applied in the clinical setting should significantly reduce morbidity and mortality associated with alloSCT, as well as safely extend the donor pool to include fully mismatched donors. However, researchers do not yet agree on the best way to eliminate alloreactive T cells. Moreover, the optimal number of T cells, including the clinically allowable number of residual alloreactive T cells, to include in the graft remains unknown and may in fact vary among donor /recipient pairs. Lastly, as not all data obtained in murine or other animal models can be extrapolated to the clinic, the clinical trials performed or currently undertaken are essential in order to define fully the best strategies for each situation.
References 1 2
Armitage JO. Bone marrow transplantation. N Engl J Med 1994;330:827 /38. Kernan NA. T-cell depletion for prevention of graft-versushost disease. In: Forman SJ, Blume KG, Thomas ED, eds
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Bone Marrow Transplantation . Boston, MA: Blackwell, 1994: 124 /35. Storb R, Deeg HJ, Whitehead J et al . Methotrexate and cyclosporine compare with cyclosporine alone for prophylaxis of acute graft-versus-host disease after marrow transplantation for leukemia. New Engl J Med 1986;314:729 /35. Beatty PG. Results of allogeneic bone marrow transplantation with unrelated or mismatched donors. Semin Oncol 1992;19: 13 /9. Champlin R. T-cell depletion to prevent graft-versus-host disease after bone marrow transplantation. Hematol Oncol Clin North Am 1990;4:687 /98. Gratama JW, van Esser JW, Lamers CH et al . Tetramer-based quantification of cytomegalovirus (CMV)-specific CD8 T lymphocytes in T-cell-depleted stem cell grafts and after transplantation may identify patients at risk for progressive CMV infection. Blood 2001;98:1358 /64. van Esser JW, van der Holt B, Meijer E et al . Epstein /Barr virus (EBV) reactivation is a frequent event after allogeneic stem cell transplantation (SCT) and quantitatively predicts EBV-lymphoproliferative disease following T-cell-depleted SCT. Blood 2001;98:972 /8. Zeiser R, Marks R, Bertz H et al . Immunopathogenesis of acute graft-versus-host disease: implications of novel preventive and therapeutic strategies. Ann Hemtol 2004;83:551 /65. Pal P, Zeng H, Durocher G et al . Phototoxicity of some bromine-substituted rhodamine dyes: synthesis, photophysical properties and application as photosensitizers. Photochem Photobiol 1996;63:161 /8. Villeneuve L. Ex vivo photodynamic purging in chronic myelogenous leukaemia and other neoplasias with rhodamine derivatives. Biotechnol Appl Biochem 1999;30:1 /17. Roy DC, Boileau J, Laplante J et al . Phase I study of autologous progenitor cell transplantation (PCT) purged with a photodynamic approach for patients with chronic myeloid leukemia (CML). Blood 2000;96(suppl 1):583a. Roy DC, Dallaire N, Moreau B et al . Photodynamic therapy (PDT) of B-lineage non-Hodgkin’s lymphoma (NHL) and chronic lymphocytic leukemia (CLL). Blood 2000;96(suppl 1):185a. Guimond M, Balassy A, Barrette M et al . P-glycoprotein targeting: a unique strategy to selectively eliminate immunoreactive T cells. Blood 2002;100:375 /82. Guimond M, Brochu S, Perreault C et al . Specific elimination of anti-host T cell alloreactivity using a photodynamic approach. FASEB J 2000;14:A1074. Chen BJ, Cui X, Liu C, Chao NJ. Prevention of graft-versus-host disease while preserving graft-versus-leukemia effect after selective depletion of host-reactive T cells by photodynamic cell purging process. Blood 2002;99:3083 /8. Suchin EJ, Langmuir PB, Palmer E et al . Quantifying the frequency of alloreactive T cells in vivo : new answers to an old question. J Immunol 2001;166:973 /81. Godfrey WR, Krampf MR, Taylor PA, Blazar BR. Ex vivo depletion of alloreactive cells based on CFSE dye dilution,
Selective depletion of alloreactive T cells using PDP and memory T cells
18
19
20
21
activation antigen selection, and dendritic cell stimulation. Blood 2004;103:1158 /65. Amrolia PJ, Muccioli-Casadei G, Yvon E et al . Selective depletion of donor alloreactive T cells without loss of antiviral or antileukemic responses. Blood 2003;102:2292 /9. Shafer-Weaver K, Sayers T, Strobl S et al . The granzyme B ELISPOT assay: an alternative to the 51Cr-release assay for monitoring cell-mediated cytotoxicity. J Translational Med 2003;1:14 /23. Sabin S, Akoglu TF, Gurbuz O et al . Reevaluation of skinexplant model in graft-versus-host disease prediction. Clin Transplant 1995;9:370 /4. Dickinson AM, Cavet J, Cullup H et al . GvHD risk assessment in hematopoietic stem cell transplantation: role of cytokine gene polymorphisms and an in vitro human skin explant model. Hum Immunol 2001;62:1266 /76.
22
23
24
25
26
133
Traggiai E, Chicha L, Mazzucchelli L et al . Development of a human adaptive immune system in cord blood cell-transplanted mice. Science 2004;304:104 /7. Chen BJ, Cui X, Sempowski GD et al . Transfer of allogeneic CD62L memory T cells without graft-versus-host disease. Blood 2004;103:1534 /41. Anderson BE, McNiff J, Yan J et al . Memory CD4 T cells do not induce graft-versus-host disease. J Clin Invest 2003;112:101 / 8. Foster AE, Marangolo M, Sartor MM et al . Human CD62L memory T cells are less responsive to alloantigen stimulation than CD62L naive T cells: potential for adoptive immunotherapy and allodepletion. Blood 2004;104:2403 /9. Kwak LW, Taub DD, Duffey PL et al . Transfer of myeloma idiotype-specific immunity from an actively immunised marrow donor. Lancet 1995;345:1016 /20.
Selective elimination of alloreactivity from immunotherapeutic T cells by photodynamic cell purging and memory T-cell sorting NT Le, BJ Chen and NJ Chao Divisions of Cellular Therapy/Bone Marrow Transplantation and Medical Oncology, Departments of Medicine and Immunology, Duke University Medical Center, Durham, North Carolina, USA
Allogeneic stem cell transplantation (alloSCT), especially in the mismatched setting, carries a high risk of life-threatening GvHD because of activation of donor T cells by Ag present on host cells. Removal of mature donor T cells can prevent GvHD but leads to delayed immune reconstitution, and an increased incidence of opportunistic infections and disease relapse. These findings demonstrate the vital role of donor T cells in providing graft-versus-tumor (GvT) and anti-pathogen effects as well as facilitating immune reconstitution. It has been well documented that GvHD can be separated from GvT effects, making it possible potentially to eliminate GvHD while preserving the immunotherapeutic benefits of donor T cells. Over the past decade, major attempts have been made to reduce
GvHD incidence without loss of GvT effect, especially in the haploidentical setting. Novel techniques to deplete host-reactive donor T cells selectively have been explored. This review focuses on the use of the photodynamic cell purging (PDP) process and of sorting memory T cells for the selective elimination of alloreactivity. Minimizing the threat of GvHD while maximizing the beneficial GvT effect would broaden the scope and effectiveness of alloSCT.
Introduction
can prevent GvHD but leads to a delay in immune reconstitution, leaving the recipient at great risk of potentially fatal opportunistic infections [5/7]. Moreover, total T-cell depleted (TCD) stem cell grafts are associated with higher rates of graft rejection and disease relapse [2], demonstrating the important role of donor T cells in GvT effects. Strategies in which T cells causing GvHD are selectively depleted, while T cells capable of mounting responses to viral and tumor Ag are preserved, are thus highly desirable for clinical use. Over the past decade, novel techniques to eliminate alloreactive T cells selectively, either by their removal or by destruction, ex vivo or in vivo, have been explored [8]. As part of this effort, our group has investigated two approaches for the selective depletion of alloreactive donor T cells: by PDP and by sorting memory T cells. Here we review these two
Allogeneic stem cell transplantation (SCT) is now recognized as a curative treatment for various hematologic malignancies, and has been used in the treatment of solid tumors and inherited diseases [1]. Allografts exert their anti-tumor activity by a graft-versus-tumor (GvT) effect, mediated mainly by donor cytotoxic T cells, which recognize either leukemia-specific Ag or alloantigens present on normal and malignant host cells [2]. Lifethreatening GvHD develops when donor T cells that accompany the stem cell graft are activated by Ag expressed on normal host cells, causing severe damage to target tissues. While conventional GvHD prophylaxis with immunosuppressive agents can be effective in genotypically HLA-identical sibling transplants, prophylaxis does not always prevent severe GvHD, especially in mismatched SCT [3,4]. Removal of mature donor T cells
Keywords Allodepletion, photodynamic cell purging, naive T cells, memory T cells, stem cell transplantation.
Correspondence to: Nelson J Chao, MD, Division of Cellular Therapy/BMT, Duke University Medical Center, 2400 Pratt St., Suite 1100, Box 3961, Durham, NC 27705, USA. – 2005 ISCT
DOI: 10.1080/14653240510018163
Selective depletion of alloreactive T cells using PDP and memory T cells
techniques. Their potential benefits and disadvantages will be discussed in an attempt to identify the most suitable clinical approach.
(a)
B H2N
PDP treatment prevents GvHD but preserves the GvT effect in animal models We first tested the effects of PDP treatment in mice; the results of these studies were published in 2002 [15]. We found that treatment with PDP inhibited anti-host responses as measured by cytotoxic T lymphocyte (CTL) assay and IFN-g production. In contrast, anti-tumor and
B NH2Cl
O
Ex vivo PDP treatment to remove alloreactive T cells selectively Photodynamic therapy (PDT) has been used to treat malignant disease since the turn of the century. The technique involves the in vivo administration of a photosensitizing agent, innocuous in its native state, followed by activation of the agent by light of appropriate wavelength and energy level. The activation of the photosensitizing agent is thought to result in the generation of toxic oxygen derivatives and reactive oxygen species, which leads to cellular damage and eventually cell death. Rhodamine dyes, with their low toxicity and rapid clearance, are potentially useful photosensitizers for PDT [9]. A photoactivable rhodamine derivative, 4,5-dibromorhodamine-methyl ester (TH9402) (Figure 1a), has been used to purge chronic myelogenous leukemia [10,11], NHL [12] and CLL [12] in autologous BM grafts. TH9402 is taken up by all cells, but is actively extruded from the cytoplasm by the active multi-drug transporter P-glycoprotein 170 (P-gp 170) [13]. However, P-gp 170 becomes inactivated upon T-cell activation, leading to the retention of the photosensitizer TH9402 in the mitochondria of activated T cells [13,14]. Following exposure to visible light (514 nm), the dye becomes cytotoxic, resulting in cell death [13,14]. Thus, when the mixture of resting and activated T lymphocytes is treated with TH9402 and then exposed to visible light, only the activated T cells are eliminated. Our group has investigated the potential of this procedure, PDP, for the selective elimination of donor alloantigen-specific T cells. Briefly, donor T cells are activated by host APC in a one-way MLC. Following activation, these cells are treated with TH9402 ex vivo and subsequently exposed to visible light for elimination of activated alloresponsive T cells (Figure 1b).
127
COOCH3
4,5-dibromorhodamine 123 (TH9402)
(b)
one-way MLC
TH9402 40 min
Incubation 90 min
Visible light
GVH Necrosis GVT & others Intact
Figure 1. (a) Chemical structure of 4,5-dibromorhodamine-methyl ester (TH9402). (b) A schematic diagram of PDP treatment. Donor T cells are activated by host APC in a one-way MLC. After in vitro activation, cells are treated ex vivo with TH9402 for 40 min at 378C. Cells are then washed and kept at 378C for 90 min to allow extrusion of TH9402 from resting, non-activated T cells. Immediately following the incubation, cells are exposed to visible light energy (514 nm) delivered by the scanning lamp device Theralux (Theratechnologies, Montreal, Canada). As TH9402 is selectively retained in the mitochondria of activated cells, PDP treatment results in the specific elimination of activated, host-responsive T cells.
anti-third party responses were essentially conserved. We further demonstrated, in a murine model, that while all tumor-bearing mice transplanted with only TCD BM cells had leukemic cells in peripheral blood and more than half died within 40 days after transplantation, the addition of PDP-treated cells allowed 90% of the recipients to survive for more than 100 days, without detectable tumor cells. Moreover, these mice were healthy, free of GvHD and had full phenotypic recovery of immune cells. These data suggest that PDP can selectively eliminate host alloantigen-specific T cells responsible for GvHD while preserving the GvT effect.
Problems facing human pre-clinical studies Encouraged by the results obtained from animal studies, we have undertaken a pre-clinical study to assess the feasibility of using the ex vivo TH9402 PDP process to
128
NT Le et al .
purge GvHD-inducing T cells before transfusing donor T cells into patients. We have encountered several problems Finding the optimal duration for the primary MLC Studies using CFSE-labeled cells have revealed with great precision the kinetics and degree of alloreactive T-cell clonal expansion following allogeneic stimulation in vivo [16]. During the first 30 h post-stimulation, cell division is highly synchronous. However, over the next 24 h individual responding T cells proliferate asynchronously, some daughter cells achieving only one cell division while others undergoing five rounds of cell division. The diversity of proliferative behavior is maintained throughout the entire phase of clonal expansion, with the average responding precursor giving rise to 64 daughter cells through six rounds of cell division within 72 h of stimulation. The majority of the methods employed for the depletion of GvHD-inducing T cells identify these T cells based on their activation status after ex vivo stimulation in an allogeneic MLC. In the published studies, the duration of the primary MLC has ranged from 2 /7 days. It is currently unclear what can be judged as the optimal duration for primary MLC. Longer culture times would probably result in the activation of more alloresponsive T cells but at the cost of lower cell viability. Furthermore, if we can extrapolate the results of in vivo CFSE-labeling studies to ex vivo MLC, longer culture would lead to greater numbers of daughter cells arising from a single alloreactive precursor that would need to be eradicated. In the PDP approach, this would require more reagent and a higher efficiency of depletion. We are therefore using a 3day primary MLC, although the peak of proliferative response for human PBMC is on days 4 /5, based on 3Hthymidine incorporation assay. Selecting the most efficient stimulator cells A major limitation of current approaches for allodepletion is the source of recipient stimulator cells used in alloactivation. The data accumulating thus far suggest that crude PBMC preparations are inefficient at Ag presentation, presumably because of their heterogeneity. To circumvent this problem, some investigators have used DC as the stimulators. One study reported that HLAmismatched PBMC stimulators consistently resulted in 90/95% depletion of alloreactivity, whereas purified matured DC could produce a 100-fold (99%) reduction of response to alloantigens [17]. In practice, however,
purified DC are difficult to prepare in adequate number for clinical trials. Furthermore, they require expensive cytokines for culture and Ab for cell sorting. Another concern is that, in leukemic patients, PBMC may be contaminated with leukemic blasts, or alternatively DC may have processed leukemic Ag, resulting in a loss of anti-tumor activity after allodepletion. To address this problem, one study used EBV-transformed lymphoblastoid cell lines (LCL) as stimulator cells [18]. These investigators reported a more consistent depletion of alloreactivity when LCL were used for stimulation instead of PBMC, although the residual response remained significant. LCL have several advantages as stimulator cells: they are better at Ag presentation than PBMC, are relatively inexpensive to prepare, can be easily expanded in vitro, and do not express tumor Ag that may serve as targets for GvT effects. In our hands, LCL irradiated at 90 Gy still retained a residual proliferative response, as assessed by 3H-thymidine incorporation assay The development of EBVinduced lymphoproliferative disorders after SCT is, therefore, a major concern. Because of the limitations of LCL, and the possibility that when using PBMC the residual alloproliferative response could lead to GvHD, we are considering the generation of purified mature DC from patients for use in clinical trials. Finding reliable and practically applicable read-out assays A major problem that we have encountered is finding a reliable human in vitro model or assay to predict GvHD. We found that primed PDP-treated T cells retain residual proliferative responses to alloantigens in an MLC but do not elicit cytotoxicity towards the same allogeneic targets (Figure 2). One possible explanation for these findings is that the residual alloantigen-responsive T cells, as assessed in a proliferation assay, may not be true alloreactive T cells. Thus, 3H-thymidine incorporation, although inexpensive and easy to perform, may not be an appropriate assay for evaluating the residual alloantigen response in primed PDP-treated T cells. CTL and cytotoxic T-lymphocyte precursor (CTLp) assays have been viewed as the ‘gold standard’ in assessing T-cell alloreactivity. However, these assays are laborious and require prolonged culture, as well as use of radioactive materials. In order to assess the efficiency of allodepletion after PDP treatment, we have used the recently developed granzyme B (GrB) enzyme-linked immunospot
Selective depletion of alloreactive T cells using PDP and memory T cells
80000
unprimed, untreated 0 uM 20 uM
unprimed, untreated 0 uM 20 uM 3rd-party primed
(a)
70000 60000
20 % s pec i f ic r eleas e
cpm
50000 40000 30000 20000 10000 0 0.313
0.625
1.25
2.5
responder (x10E5)
129
(b)
15 10 5 0 -5
6.25 to 1
12.5 to 1
25 to 1
E:T ratio
Figure 2. PDP-treated T cells retain some residual proliferative response to alloantigens (a) but do not elicit cytotoxicity towards same-party targets (b). PBMC from a healthy donor (responder cells) were stimulated with irradiated (30 Gy) PBMC obtained from another healthy donor (stimulator cells) in an MLC. Following 3 days of stimulation, cells were PDP-treated at various concentrations of TH9402. Cells were then rested for 3 days before being used in subsequent assays. The proliferation assay was performed using graded numbers (0.313 /2.5 /105) of responder cells and 2.5 /105 irradiated same-party stimulator cells per well in a 96-well flat-bottomed plate. Cytotoxicity was measured by the standard 4-h 51 Cr-release assay. Responder cells that were neither stimulated in primary MLC nor PDP-treated were used as a control. In addition, responder cells primed with a different (designated third-party) stimulator were used as a control for the specificity of the CTL assay. Cells in both assays were collected after a 4-day MLC.
(ELISPOT) assay in addition to a CTLp assay. When the GrB ELISPOT was directly compared with the IFN-g ELISPOT and 51Cr-release assays, excellent cross-correlation between all three assays was shown [19]. Moreover, titration studies demonstrated a strong correlation between the number of effector cells and GrB spots per well, thus providing an estimation of cytotoxic effector cell frequency [19]. Using our PDP protocol, PDP treatment leads to a TH9402 dose-dependent reduction of alloreactive cytotoxic T-cell frequency and, at 20 mm of TH9402, an average 2-log depletion is obtained (Figure 3). Given the high sensitivity and specificity of the GrB ELISPOT assay, it might be useful for clinical trials where there are limited numbers of patient cells available for immunologic monitoring. In addition to these assays, the skin explant model has been developed and used quite extensively elsewhere to predict the occurrence of GvHD. This assay involves incubating the recipient skin biopsies with donor T cells
previously activated by the recipient’s APC in a one-way MLC. The GvH reaction is then assessed by examining the histopathologic changes of the skin explant. Although several studies have reported statistically significant correlations between skin explant results and the in vivo development of GvHD, individual variations have been observed [20]. More recent reports suggest that a combination of cytokine gene polymorphisms and skin explant results might be more predictive [21]. We are in the process of using the skin explant model to test the ability of primed PDP-treated donor T cells to induce GvHDlike histopathologic changes in the recipient’s skin biopsies. Of interest, it would be of great benefit to develop a human immune system in an animal model that could be used to evaluate GvHD. Traggiai et al . [22] recently showed that intrahepatic injection of human cord blood stem cells into a murine model lacking B, T and NK cells can lead to the de novo development of human B, T and DC as well as the production of functional immune responses.
130
NT Le et al .
unprimed, untreated
100000
0 uM
80000 cpm
900 (b) 800
(a)
GrB ELISPOTs
120000
10 uM
60000
20 uM
40000
700 600 500
unprimed, untreated 0 uM 10 uM 20 uM
400 300 200
2. 5
1.2 5
0.6 3
0.3 2
0.1 6
0
0.0 8
0 0.0 4
100
0.0 2
20000
responder (x10E5)
0.02 0.04 0.08 0.16 0.32 0.63 1.25 2.5 responder (x10E5)
Figure 3. PDP treatment significantly reduces the frequency of alloreactive cytotoxic T cells. (a) Proliferative response to same-party stimulator cells. (b) Reduction of alloreactive cytotoxic T-cell frequency as assessed by GrB ELISPOT assay. Cells were stimulated and PDP-treated as described in Figure 2. The proliferation assay was performed using graded numbers (0.02 /2.5 /105) of responder cells and 1.25 /105 irradiated same-party stimulator cells per well in a 96-well flat-bottomed plate. The GrB ELISPOT assay was performed according to the manufacture’s protocol (R&D Systems, Minneapolis, MN, USA). Responder cells that were neither stimulated in primary MLC nor PDP-treated were used as a control. Cells in both assays were collected after a 18-h MLC.
This approach could be developed as a valuable model for predicting human clinical GvHD. The potential risks and benefits of adding-back PDP-treated donor T cells The main concern with adding-back primed, PDP-treated donor T cells is the potential exacerbation of GvHD. It has been well documented that memory T cells have a lower activation threshold than naive T cells and, once reactivated, they produce cytokines, proliferate and differentiate into effector T cells within hours. Thus, it is possible that the post-PDP residual alloreactive T cells might quickly become activated upon transferring into the recipient, causing acute GvHD. However, the results of our murine studies demonstrated that, although the PDPtreated T cells retained some residual alloreactivity, as determined by IFN-g ELISPOT assay, transferring these T cells into the animals did not cause clinical GvHD. Moreover, the addition of PDP-treated donor T cells to tumor-bearing TCD BM-transplanted animals improved survival, prevented relapse and allowed full immunophenotypic recovery. If we could extrapolate data from animal studies to human patients, these results suggest that PDPtreated donor T cells could be transferred to posttransplant patients to facilitate engraftment and immune reconstitution as well as to provide GvT effects without increasing GvHD.
Selective depletion of naive T cells Naive versus memory T cells Peripheral T cells can be broadly divided into two groups: T cells that have never been activated by Ag (naive T cells), and Ag-experienced T cells (memory T cells). In mice, the lymphocyte homing receptor CD62L (L-selectin) can be used to distinguish these two groups of T cells: while naive T cells are CD62L , memory T cells can be either CD62L or CD62L. In humans, CD45RA is differentially expressed in naive and memory T cells: while human naive T cells are CD45RA , human memory T cells can be either CD45RA or CD45RA .
CD62L memory T cells confer GvT effects but do not cause GvHD in animal models We and others [23 /25] recently reported a novel approach to prevent GvHD by selectively depleting naive T cells. This approach was based on the following hypothesis: naive T cells, which express CD62L, are T cells that have never encountered Ag specific for their T-cell Ag receptor (TCR). Memory/activated T cells, which can be either CD62L or CD62L, are T cells that have been exposed to their corresponding Ag. If a donor has never encountered the host alloantigens, potential GvHD-inducing alloreactive T cells should be contained within the donor naive T-cell compartment. Thus, CD62L T cells, which
Selective depletion of alloreactive T cells using PDP and memory T cells
Human unprimed CD45RA memory T cells do not possess a cytotoxic response to alloantigens It has been postulated that there is an immunologic crossreactivity between allogeneic MHC molecules and environmental Ag being presented by autologous APC. This leads to a legitimate concern for human application of this approach because in humans, as opposed to laboratory mice raised in sterile conditions, donor T cells might have encountered environmental Ag cross-reacting with alloantigens present in the recipient. Thus, donor T cells might
250000
(a)
c pm
200000 150000 100000
Bulk CD45RA+ CD45RA-
50000 0 4
5
6
7
Days in culture
8
contain some memory T cells that are host reactive. To investigate this possibility, we compared the ability of memory T cells to respond to alloantigens with those of naive and bulk T cells. Purified T cells were first obtained from peripheral blood of healthy donors, then separated into memory and naive T-cell subsets based on the expression of CD45RA (memory, CD45RA ; naive, CD45RA ). The abilities of these T-cell subsets to respond to alloantigens were tested using proliferation and cytotoxicity assays. We found that, in contrast to the data obtaining from the mice, human memory T cells proliferated equally as well as naive and bulk T cells in MLC (Figure 4a). However, these same memory T cells failed to kill the allogeneic targets despite the vigorous proliferative responses against the same alloantigens (Figure 4b). It is possible that human CD45RA memory T cells do not contain true alloantigen-specific T cells or that they do not possess the effector functions necessary for the initiation of GvHD. Thus, transferring donor CD45RA memory T cells into the recipient may not cause GvHD.
The potential risks and benefits of transferring human memory T cells Transferring donor memory T cells into patients posttransplant would have several advantages. First is the ease
% s pec ific r eleas e
are devoid of naive T cells and represent a subset of memory/activated T cells, should not be able to induce GvHD. We showed that, in unprimed mice, CD62L T cells failed to proliferate in response to alloantigens and were unable to induce GvHD in allogeneic hosts [23]. Furthermore, CD62L T cells contributed to T-cell reconstitution by peripheral expansion as well as by promoting T-cell regeneration from BM stem/progenitor cells. Importantly, CD62L T cells from the animals previously primed with a tumor cell inhibited the tumor growth in vivo but were unable to induce GvHD in the third-party recipients. These results suggest that it is possible to selectively prime CD62L T cells in normal donors to carry the memory T cells that we desire and transfer them to the recipients without causing GvHD.
131
40
*
Bulk CD45RA+ CD45RA-
(a)
30 20 10 0 12.5 to 1
25 to 1
50 to 1
100 to 1
E:T ratio
Figure 4. The human CD45RA memory T-cell compartment does not contain alloantigen-specific T cells if it has never been exposed to those alloantigens. (a) Human memory T cells proliferated as well as naive and bulk T cells did in MLC. (b) Human CD45RA memory T cells, however, did not display a cytotoxic response towards alloantigens. Purified T cells were first obtained from peripheral blood from healthy donors and then separated into memory and naive T-cell subsets based on the expression of CD45RA (memory, CD45RA ; naive, CD45RA ). Memory T cells were subsequently tested for their ability to respond to alloantigens using proliferation and cytotoxicity assays, in comparison with naive and bulk T cells. The proliferation assay was performed using 1.25 /105 responder cells and 5 /105 irradiated stimulator cells per well in a 96-well flat-bottomed plate. Cytotoxicity was measured by a standard 4-h 51Cr-release assay after a 5-day MLC. *The values of percentage specific release by the T cells activated by third-party alloantigens were 6.8%, 8.5%, 9.5% and 5.6% for E:T ratios of 12.5:1, 25:1, 50:1 and 100:1.
132
NT Le et al .
of selection and availability of GMP-graded fluorescencelinked anti-CD45RA Ab that would facilitate translating the research from the bench to a clinical trial. Second, unlike PDP treatment, donor CD45RA memory T cells are not primed in an allogeneic MLC. Furthermore, as revealed by our preliminary data, although the CD45RA memory T cells display vigorous proliferative responses against the same alloantigens, they do not elicit cytotoxicity to the same-party targets, suggesting that they may not cause GvHD upon in vivo transfer. Lastly, but most importantly, the finding that CD62L memory T cells from the animals previously primed with a tumor cell line were capable of inhibiting the tumor growth in vivo demonstrated that these memory T cells retain the GvT effect and perhaps other memory T-cell functions. It further suggests a strategy of avoiding GvHD by selective priming of T cells in normal donors to carry the desired memory T cells followed by their transfer to the recipient. It may prove difficult to prime the donor with recipient tumor cells in some cases. However, it has been shown that healthy donors can be immunized with myeloma immunoglobulin from the plasma of the recipient and developed a specific T-cell response [26].
Conclusion Donor T cells selectively depleted of GvHD-inducing ability have great clinical potential. Allodepletion applied in the clinical setting should significantly reduce morbidity and mortality associated with alloSCT, as well as safely extend the donor pool to include fully mismatched donors. However, researchers do not yet agree on the best way to eliminate alloreactive T cells. Moreover, the optimal number of T cells, including the clinically allowable number of residual alloreactive T cells, to include in the graft remains unknown and may in fact vary among donor /recipient pairs. Lastly, as not all data obtained in murine or other animal models can be extrapolated to the clinic, the clinical trials performed or currently undertaken are essential in order to define fully the best strategies for each situation.
References 1 2
Armitage JO. Bone marrow transplantation. N Engl J Med 1994;330:827 /38. Kernan NA. T-cell depletion for prevention of graft-versushost disease. In: Forman SJ, Blume KG, Thomas ED, eds
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Bone Marrow Transplantation . Boston, MA: Blackwell, 1994: 124 /35. Storb R, Deeg HJ, Whitehead J et al . Methotrexate and cyclosporine compare with cyclosporine alone for prophylaxis of acute graft-versus-host disease after marrow transplantation for leukemia. New Engl J Med 1986;314:729 /35. Beatty PG. Results of allogeneic bone marrow transplantation with unrelated or mismatched donors. Semin Oncol 1992;19: 13 /9. Champlin R. T-cell depletion to prevent graft-versus-host disease after bone marrow transplantation. Hematol Oncol Clin North Am 1990;4:687 /98. Gratama JW, van Esser JW, Lamers CH et al . Tetramer-based quantification of cytomegalovirus (CMV)-specific CD8 T lymphocytes in T-cell-depleted stem cell grafts and after transplantation may identify patients at risk for progressive CMV infection. Blood 2001;98:1358 /64. van Esser JW, van der Holt B, Meijer E et al . Epstein /Barr virus (EBV) reactivation is a frequent event after allogeneic stem cell transplantation (SCT) and quantitatively predicts EBV-lymphoproliferative disease following T-cell-depleted SCT. Blood 2001;98:972 /8. Zeiser R, Marks R, Bertz H et al . Immunopathogenesis of acute graft-versus-host disease: implications of novel preventive and therapeutic strategies. Ann Hemtol 2004;83:551 /65. Pal P, Zeng H, Durocher G et al . Phototoxicity of some bromine-substituted rhodamine dyes: synthesis, photophysical properties and application as photosensitizers. Photochem Photobiol 1996;63:161 /8. Villeneuve L. Ex vivo photodynamic purging in chronic myelogenous leukaemia and other neoplasias with rhodamine derivatives. Biotechnol Appl Biochem 1999;30:1 /17. Roy DC, Boileau J, Laplante J et al . Phase I study of autologous progenitor cell transplantation (PCT) purged with a photodynamic approach for patients with chronic myeloid leukemia (CML). Blood 2000;96(suppl 1):583a. Roy DC, Dallaire N, Moreau B et al . Photodynamic therapy (PDT) of B-lineage non-Hodgkin’s lymphoma (NHL) and chronic lymphocytic leukemia (CLL). Blood 2000;96(suppl 1):185a. Guimond M, Balassy A, Barrette M et al . P-glycoprotein targeting: a unique strategy to selectively eliminate immunoreactive T cells. Blood 2002;100:375 /82. Guimond M, Brochu S, Perreault C et al . Specific elimination of anti-host T cell alloreactivity using a photodynamic approach. FASEB J 2000;14:A1074. Chen BJ, Cui X, Liu C, Chao NJ. Prevention of graft-versus-host disease while preserving graft-versus-leukemia effect after selective depletion of host-reactive T cells by photodynamic cell purging process. Blood 2002;99:3083 /8. Suchin EJ, Langmuir PB, Palmer E et al . Quantifying the frequency of alloreactive T cells in vivo : new answers to an old question. J Immunol 2001;166:973 /81. Godfrey WR, Krampf MR, Taylor PA, Blazar BR. Ex vivo depletion of alloreactive cells based on CFSE dye dilution,
Selective depletion of alloreactive T cells using PDP and memory T cells
18
19
20
21
activation antigen selection, and dendritic cell stimulation. Blood 2004;103:1158 /65. Amrolia PJ, Muccioli-Casadei G, Yvon E et al . Selective depletion of donor alloreactive T cells without loss of antiviral or antileukemic responses. Blood 2003;102:2292 /9. Shafer-Weaver K, Sayers T, Strobl S et al . The granzyme B ELISPOT assay: an alternative to the 51Cr-release assay for monitoring cell-mediated cytotoxicity. J Translational Med 2003;1:14 /23. Sabin S, Akoglu TF, Gurbuz O et al . Reevaluation of skinexplant model in graft-versus-host disease prediction. Clin Transplant 1995;9:370 /4. Dickinson AM, Cavet J, Cullup H et al . GvHD risk assessment in hematopoietic stem cell transplantation: role of cytokine gene polymorphisms and an in vitro human skin explant model. Hum Immunol 2001;62:1266 /76.
22
23
24
25
26
133
Traggiai E, Chicha L, Mazzucchelli L et al . Development of a human adaptive immune system in cord blood cell-transplanted mice. Science 2004;304:104 /7. Chen BJ, Cui X, Sempowski GD et al . Transfer of allogeneic CD62L memory T cells without graft-versus-host disease. Blood 2004;103:1534 /41. Anderson BE, McNiff J, Yan J et al . Memory CD4 T cells do not induce graft-versus-host disease. J Clin Invest 2003;112:101 / 8. Foster AE, Marangolo M, Sartor MM et al . Human CD62L memory T cells are less responsive to alloantigen stimulation than CD62L naive T cells: potential for adoptive immunotherapy and allodepletion. Blood 2004;104:2403 /9. Kwak LW, Taub DD, Duffey PL et al . Transfer of myeloma idiotype-specific immunity from an actively immunised marrow donor. Lancet 1995;345:1016 /20.