Unique Abnormalities of CD4 and CD8 Central Memory Cells ...

Biology of Blood and Marrow Transplantation 12:22-30 (2006) 䊚 2006 American Society for Blood and Marrow Transplantation 1083-8791/06/1201-0123$32.00/0 doi:10.1016/j.bbmt.2005.11.004

Unique Abnormalities of CD4ⴙ and CD8ⴙ Central Memory Cells Associated with Chronic Graft-versus-Host Disease Improve after Extracorporeal Photopheresis Kouhei Yamashita,1 Mitchell E. Horwitz,2 Akua Kwatemaa,1 Effie Nomicos,1 Kathleen Castro,3 Robert Sokolic,3 Susan F. Foster,1 Mary Garofalo,1 Uimook Choi,1 Mark Ryherd,1 Margaret R. Brown,4 Susan F. Leitman,5 Alan S. Wayne,6 Daniel H. Fowler,3 Michael R. Bishop,3 Richard W. Childs,7 A. John Barrett,7 Steven Z. Pavletic,3 Harry L. Malech1 1

Laboratory of Host Defenses, National Institute of Allergy and Infectious Diseases; 2Bone Marrow Transplant Program, Duke University Medical Center, Durham, North Carolina; 3Experimental Transplantation and Immunology Branch, National Cancer Institute; Departments of 4Laboratory Medicine and 5Transfusion Medicine, Warren Grant Magnuson Clinical Center; 6Pediatric Oncology Branch, National Cancer Institute; and 7 Hematology Branch, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland Correspondence and reprint requests: Harry L. Malech, MD, Laboratory of Host Defenses, National Institute of Allergy and Infectious Diseases, National Institutes of Health, 10 Center Dr, MSC 1456, Bldg 10, Room 5-3750, Bethesda, MD 20892 (e-mail: [email protected]). Received October 21, 2005; accepted November 7, 2005

ABSTRACT Chronic graft-versus-host disease (cGVHD) remains a problematic complication of allogeneic hematopoietic stem cell transplantation. Laboratory parameters correlated with cGVHD have not been fully defined, although changes in CD4/CD8 ratios occur and a decrease in CD4ⴙ central memory T cells has been noted. Extracorporeal photopheresis (ECP) is an effective therapy for steroid-refractory cGVHD. We have noted changes in lymphocyte subsets after ECP. CD4ⴙ and CD8ⴙ T-cell central and effector memory populations were enumerated by flow cytometry in a cohort of 37 patients postallogeneic transplantation with symptomatic cGVHD. Of the patients with symptomatic cGVHD, 7 were treated with ECP over 6 months and prospectively assessed for changes in lymphocyte subsets. There was a highly significant correlation of an increase in CD8ⴙ central memory cells and a concomitant decrease in CD4ⴙ central memory cells in patients with symptomatic cGVHD. These changes were not detected in patients without cGVHD posttransplantation. In all, 7 patients with cGVHD followed up prospectively during ECP treatment showed a statistically significant normalization of the pattern of CD4ⴙ and a trend toward normalization of CD8ⴙ central memory T cells coincident with improvement of cGVHD. These data indicate a high correlation between disturbances in the balance of central and effector memory populations and cGVHD suggesting use in following up responses to therapy. The normalization of central and effector memory populations in response to ECP coincident with clinical improvement of cGVHD support a correlation between these laboratory parameters and cGVHD. Further studies are needed to demonstrate whether laboratory measurements of the magnitude of changes in central and effector memory populations are useful prognostically or can be used to guide response to therapy. The contrasting change in central memory cells (CD8ⴙ increased versus CD4ⴙ decreased) in cGVHD provide support for recent reports suggesting unique differences in the differentiation pathways for CD8ⴙ versus CD4ⴙ T cells. © 2006 American Society for Blood and Marrow Transplantation

KEY WORDS Chronic GVHD transplantation

22

● ●

Central memory T cell Alloimmune

●

Extracorporeal photopheresis

●

Bone marrow

Abnormal Central/Effector Memory T Cell Ratios

INTRODUCTION After allogeneic stem cell transplantation, chronic graft-versus-host disease (cGVHD) is the most common and life-threatening delayed transplantation-related complication [1]. Donor-derived, alloreactive T cells play an important role in the pathogenesis of cGVHD, but their mechanism of action is uncertain. The incidence of cGVHD of 30% to 50% is increasing because of an increase in allogeneic stem cell transplantations for older patients, HLA antigen–mismatched transplants, peripheral blood stem cell transplantations, and donor lymphocyte infusions for relapsed disease or unstable mixed chimerism [2,3]. Extracorporeal photopheresis (ECP) is an effective therapy for cGVHD in patients who are unresponsive to first-line treatment with corticosteroids and calcineurin inhibitors [4-7]. Lymphocytes exposed to 8-methylpsoralen and UVA light undergo apoptosis [8]; however, only 2% to 5% of the total circulating mononuclear pool is exposed to UVA with each treatment. Thus, ECP likely reduces manifestations of graft-versus-host diseases (GVHD) by changing the balance of alloimmune activity by mechanisms other than simply the removal of alloreactive lymphoid clones. Recently, it has been reported that ECP results in a normalization of the inverted CD4/CD8 ratio, increased CD3⫺CD56⫹ natural killer (NK) cells, decreased CD80⫹CD123⫹ circulating dendritic cells, with a shift in dendritic polarization from a predominantly dendritic cell 1 to a dendritic cell 2 pattern [9,10]. Concurrently there is a shift in CD4⫹ T-cell polarity from a predominantly TH1 pattern to a TH2 cytokine profile [10]. However, the precise mechanism of action of ECP has yet to be defined and a recent report from a study of 28 patients with cGVHD treated with ECP found no laboratory parameter that predicted a favorable response to ECP [7]. Human peripheral blood CD4⫹ and CD8⫹ T cells can be grouped into functionally distinct subpopulations by the pattern of expression of chemokine receptor, CCR7. CD4⫹ T cells consist of one naive (CD45RA⫹CCR7⫹) and two memory subsets, CD45RA⫺ CCR7⫹ (central memory) and CD45RA⫺CCR7⫺ (effector memory). Within CD8⫹ T cells, the same 3 subsets have been identified, with a fourth CD45RA⫹ CCR7⫺CD8⫹ T-cell (effector memory) subset [11]. In subsets of memory T cells it has been noted that the adhesion molecule, CD62L (L-selectin) is usually expressed on the same cells that express CCR7 [11]. Both molecules appear pivotal for migration into secondary lymphoid organs [12,13]. Supported by Japan Society for the Promotion of Science Research Fellowships for Japanese Biomedical and Behavioral Researchers at National Institutes of Health (Dr Yamashita) and Therakos Inc.

BB&MT

We have recently reported that severe cGVHD is characterized by a preponderance of CD4⫹ effector memory T cells relative to CD4⫹ central memory T cells, compared with healthy individuals or patients with stem cell transplantation without cGVHD, in whom central memory CD4⫹ T cells are predominant in the peripheral blood [14]. In this study, we further examined whether clinical amelioration of cGVHD in response to ECP normalizes the abnormal patterns of T-cell memory subpopulations. We also extend our previous observations on memory CD4⫹ T cells in GVHD to the CD8⫹ T-cell memory compartments. In contrast to CD4⫹ T cells where there is a significant deficiency of central memory cells with preservation of effector memory populations associated with cGVHD, the CD8⫹ T-cell compartment demonstrates a marked increase in the proportion of central memory cells with little change in the numbers of effector memory cells. There is also a modest decrease in the RA⫹ effector memory population unique to CD8⫹ cells associated with cGVHD. We report here for the first time that the unique changes of both CD4⫹ and CD8⫹ central memory cells associated with cGVHD normalize after ECP treatment. These findings have implications for modeling memory T-cell differentiation pathways and may provide a marker to monitor response to ECP.

METHODS Patients and ECP Treatment

All patients studied had been enrolled onto National Institutes of Health (NIH) allogeneic stem cell transplantation protocols, conditioned with either myeloablative or nonmyeloablative regimens, and were undergoing transplantation from 6/6 or 5/6 HLA antigen– matched related donors. All patients were enrolled on one of a number of NIH institutional review board– approved protocols and all signed informed consent allowing study of blood cell phenotypes. A single blood sample was obtained from 37 patients with cGVHD. Patients had at least one unequivocal manifestation of cGVHD at the time of evaluation affecting skin, mouth, eyes, lungs, joints, fascia, or gastrointestinal system that required corticosteroids for control of symptoms. Most patients also received a calcineurin inhibitor, and some were being treated with additional medications. From this group, 7 patients were referred to and enrolled in a protocol to study the effect of ECP and allowing observation of longitudinal changes in lymphocyte subsets in response to this therapy. All 7 patients referred for ECP therapy had active and extensive cGVHD refractory to corticosteroid and/or calcineurin inhibitor therapy. More detailed characteristics of these 7 patients enrolled in the ECP protocol are shown in Table 1. The 23

K. Yamashita et al.

Table 1. Extracorporeal Photopheresis Patient Characteristics

Patient

Age (y), Sex

1 2 3 4 5 6 7

19/F 37/F 43/M 32/F 19/M 40/M 53/F

Disease

Duration of cGVHD

Initiation of ECP after SCT

ALL NHL NHL CML HD NHL AA

2 y 2 y 10 y 9 mo 4 mo 5 mo 2 mo

2.5 y 3 y 10 y 18 mo 8 mo 11 mo 5 mo

Involvement Sites Skin, Skin, Skin, Skin, Skin, Skin, Skin,

mouth musculoskeletal mouth, lung eyes, musculoskeletal mouth, lung eyes, musculoskeletal mouth, liver musculoskeletal lung mouth, musculoskeletal mouth, eyes musculoskeletal

Immunosuppressive Treatment at Enrollment Pred/CSA Pred/FK506 Pred Pred/FK506 Pred/FK506/MMF Pred/FK506 Pred/FK506

AA indicates aplastic anemia; ALL, acute lymphoblastic leukemia; cGVHD, chronic graft-versus-host disease; CML, chronic myelogenous leukemia; CSA, cyclosporin A; ECP, extracorporeal photopheresis; F, female; FK506, tacrolimus; HD, Hodgkin’s disease; M, male; MMF, mycophenolate mofetil; NHL, non-Hodgkin’s lymphoma; Pred, prednisone; SCT, stem cell transplantation.

blood samples from normal healthy control subjects and from patients with non-GVHD autoimmune disorders on high-dose steroids were also obtained under informed consent under institutional review board– approved protocol ECP Treatment Protocol

ECP treatment was administered by trained nursing staff under medical supervision using the UVAR XTS system consisting of the extracorporeal blood cell separation and irradiation device, associated attachments, and UVADEX (methoxsalen), all manufactured by Therakos Inc, Exton, Pa, and used according to manufacturer’s instructions. Therakos Inc provided some support for the conduct of this study in the context of a cooperative research and development agreement with the National Institute of Allergy and Infectious Diseases. ECP treatment was administered in the Apheresis Section of the Division of Transfusion Medicine, Warren H. Magnuson Clinical Center, NIH. Each procedure typically consisted of 5 to 6 cycles of collection and separation of buffy coat leukocytes (consisting of about 2%-5% of circulating leukocytes), followed by exposure of the pooled leukocytes to 8-methylpsoralen and UVA, and return to the patient. Each patient received two treatments per week for the first 3 months, followed by two treatments every other week for up to an additional 3 months. Response to ECP was evaluated by several methods. The primary clinical objective of the ECP trial was to reduce the level of corticosteroid dosage required to control pain and/or maintain performance of activities of daily living. All 7 patients achieved a greater than 30% reduction in steroid dose while showing improvement of one or more performance parameters. Cell Preparation and Flow Cytometry

Peripheral blood mononuclear cells were isolated from anticoagulated venous blood using lymphocyte 24

separation medium according to manufacturer’s instructions (ICN Biomedicals Inc, Aurora, Ohio). Freshly isolated peripheral blood mononuclear cells were analyzed using a flow cytometer (FACSort, BD Biosciences, San Jose, Calif). CD4⫹ cells were stained with fluorescein isothiocyanate– conjugated antihuman CD4 monoclonal antibody (mAb), phycoerythrin-conjugated antihuman CD45RO mAb, allophycocyanin-conjugated antihuman CD62L mAb, and biotinylated antihuman CCR7 mAb in combination with cychrome-conjugated streptavidin (BD Biosciences) following the manufacturer’s protocol. CD8⫹ cells were stained with fluorescein isothiocyanate– conjugated antihuman CD8 mAb, phycoerythrin-conjugated antihuman CCR4 mAb, allophycocyanin-conjugated antihuman CD45RA mAb, and biotinylated antihuman CCR7 mAb in combination with cychrome-conjugated streptavidin (BD Biosciences) following the manufacturer’s protocol. Data acquisition and postacquisition analysis was performed with software (CellQuest, BD Biosciences). Statistical Analysis

The data are shown as means ⫾ SEM. Two-group comparison was done on the analysis of variance and Students t test. For the analysis of 2-dimensional dot plots, logarithmic regression was performed. P less than .05 was considered statistically significant. RESULTS Figure 1A shows representative dot plots of flow cytometry analysis of CD62L and CCR7 expression on CD4⫹ T cells in a healthy individual (left), and two types of abnormal patterns that we observed in patients with cGVHD (middle and right). The vertical line in the graph is set at the upper end of isotype antibody binding background, but the horizontal line is set just below the CD62Lhigh population seen in healthy individuals. Note that even in healthy individuals there are cells that fall into the left upper quad-

Abnormal Central/Effector Memory T Cell Ratios

Figure 1. A, Flow cytometry dot plot analyses of peripheral blood CD4⫹ T cells assessing surface expression of CD62L (allophycocyanin chromoflor) on vertical axis and CCR7 (CyChrome chromoflor) on horizontal axis. Left, Typical pattern seen in healthy individuals. Middle and right panels, Patterns seen in two patients with cGVHD with extreme change noted in about 10% (right) and more moderate change typical of majority where loss of expression of CCR7 is more profound than loss of CD62L expression (middle). Vertical line in each graph was set just above staining by isotype CyChrome control antibody, whereas horizontal line was set just below population of cells in healthy donors with high expression of CD62L. Numbers in each quadrant indicate percent of events (cells) falling into that quadrant. B, Scatter plot (vertical logarithmic scale) of mathematic transformation of data from flow cytometry generated dot plots identical to those shown in A, where ratio of CCR7⫹CD62Lhigh/CCR7⫺CD62Llow CD4⫹ T cells (percent in upper right quadrant divided by percent in lower left quadrant) in 52 healthy individuals (solid square, healthy), 21 patients postallogeneic transplantation without cGVHD ( , no GVHD), and 37 patients postallogeneic transplantation with symptomatic cGVHD (solid circle, cGVHD). Each dot represents one patient.

‘

rant that cannot be classified by the standard definition as either central memory or effector memory cells. These may represent cells in transition or some other unique population. The majority of patients (about two thirds) with cGVHD have a pattern closer to that seen in the middle panel. We empirically tested several mathematic transformations of the data to see which transformation appeared to provide the greatest distinction between a large sampling of healthy individuals and the symptomatic cGVHD group of patients. This transformation consisted of the ratio of the percent of cells in the upper right quadrant (central memory CD4 T cells) divided by the percent of cells in the lower left quadrant (effector memory CD4 T cells). This ratio is plotted for the entire group of 52 healthy control blood samples, for 21 patients postallogeneic transplantation without cGVHD, and for 37 patients with symptomatic cGVHD postallogeneic transplantation in Figure 1B.

BB&MT

Healthy individuals had an average ratio of 3.2 ⫾ 0.37 indicating a preponderance of central memory and naive CD4⫹ T cells. Patients postallogeneic transplantation without cGVHD had similar average ratios (2.8 ⫾ 0.69, not significantly different from control subjects). Patients with symptomatic cGVHD had lower average ratios (0.78 ⫾ 0.11; P ⬍ .0001 compared with normal, P ⬍ .0002 compared with the non-cGVHD posttransplantation group). It is of note that the lowest ratio value for any healthy individual was 0.72 and for any patient postallogeneic transplantation without cGVHD was 0.73, whereas 23 of the 37 patients with cGVHD (62%) had a ratio below 0.72. We have previously reported that corticosteroid administration does not appear to affect numbers of central memory CD4⫹ T cells [14]. When we applied the above mathematic transformation (ratio of central memory to effector memory CD4 cells) to these previously published data from patients with autoimmune 25

K. Yamashita et al.

disease on high-dose steroids the values were not significantly different from the normal range (2.5 ⫾ 0.50). This confirms that chronic steroid administration is not likely a primary cause of the abnormal ratio described for patients with symptomatic cGVHD. However, patients with cGVHD in general were also on other immunosuppressives (most commonly tacrolimus), which could also influence these ratios. In data not shown, we re-analyzed blood samples from some healthy individuals and patients to assess the same mathematic ratio when the dot plot analysis was restricted to only the memory (CD45RO⫹ CD4⫹ T-cell) populations (i.e., not including naive T cells). We found the ratio for 25 healthy individuals to be 2.2 ⫾ 0.37, for 12 patients postallogeneic transplantation without cGVHD to be 1.6 ⫾ 0.38, and for 31 patients with cGVHD to be 0.59 ⫾ 0.07. The lowest ratio for a healthy individual was 0.48, where one of 12 patients posttransplantation without GVHD was below that (at 0.42), but 16 of 31 (52%) were below 0.48. Thus, the general conclusions regarding the ratio calculations apply even when the analysis was restricted to the memory populations (statistical differences at P ⬍ .0001 for healthy compared with cGVHD; and P ⬍ .0002 for no cGVHD posttransplantation versus cGVHD). However, restriction to the memory populations did not appear to increase the specificity of the analysis to distinguish the cGVHD population from the other groups. Because a 3-color analysis by flow cytometry is significantly easier to perform than a 4-color analysis, we ignored the RA or RO status of the CD4⫹ T cells in the subsequent analyses that we show below.

Figure 2. Scatter plot showing data from flow cytometry analysis of subtypes of peripheral blood CD8⫹ T cells. Each filled symbol represents value for analysis of healthy individual whereas each open symbol represents value for analysis of patient with symptomatic cGVHD. Indicated are comparisons of populations of CD8⫹ T cells that are defined as naive (CD45RA⫹CCR7⫹), effector memory (Tem ⫽ CD45RA⫺CCR7⫺), RA⫹ effector memory (Temra ⫽ CD45RA⫹CCR7⫺) and central memory (Tcm ⫽ CD45RA⫺ CCR7⫹). Each dot represents one patient. 26

We next analyzed the CD8⫹ T-cell populations with respect to expression of CD45RA and CCR7 to enumerate the following populations: naive (CD45RA⫹ CCR7⫹), central memory (CD45RA⫺CCR7⫹), effector memory (CD45RA⫺CCR7⫺), and RA⫹ effector memory (CD45RA⫹CCR7⫺). The results of this analysis for 26 healthy individuals and 23 patients with cGVHD are shown in Figure 2. Of note is that the CD8⫹ T-cell central memory population stands out with regard to significance of the difference (P ⬍ .0001) between the healthy individuals (11.8 ⫾ 1.0%) and the patients with cGVHD (23.3 ⫾ 2.0%). Specifically, there is a significant increase in peripheral blood CD8⫹ central memory T cells in patients with symptomatic cGVHD. This is in contrast to the CD4⫹ T-cell analysis noted above where there appears to be a loss of central memory cells in these patients. We also see a trend (P ⬍ .05) toward a decrease in the RA⫹ effector memory cell population in the cGVHD group (25.9 ⫾ 2.7%) relative to healthy individuals (33.6 ⫾ 2.5%). We note that similar changes in these CD8⫹ T-cell populations have been reported for patients with rheumatoid arthritis, although not seen in a comparision group of patients in that study who had systemic lupus erythematosis [15]. The level of CCR4 expression in central effector memory CD8⫹ T cells may further distinguish between cytotoxic effector precursors (CCR4⫺) and Tc2 cells (CCR4⫹) [16]. Therefore, we examined CCR4 expression specifically by CD8⫹ central memory (CD45RA⫺CCR7⫹) cells. There was no statistical difference in the percent of CD8⫹ central memory T cells that expressed CCR4 relative to those that did not express CCR4 in healthy individuals (n ⫽ 14, 39.6 ⫾ 4.7% of all central memory cells that express CCR4) versus patients with cGVHD (n ⫽ 19, 38.7 ⫾ 4.6%, P ⫽ .90) (data plots not shown). In light of the contrasting changes in CD4⫹ and CD8⫹ cells, we combined the data in one plot as shown in Figure 3 where the ratio of CD4⫹ T cells that are CCR7⫹CD62Lhigh/CCR7⫺CD62Llow is plotted on the vertical axis and the percent of CD8⫹ CD45RA⫺CCR7⫹ (central memory) is plotted on the horizontal axis. The group of 20 healthy individuals that were so analyzed cluster at the upper left side of the graph whereas 20 patients with cGVHD plotted in this way are distributed across the lower part of the graph with a trend toward lower right side of the graph. The 20 patients and 20 control subjects having both CD4 and CD8 analysis represent a nonselected group representing all control subjects and patients studied after the point in time when we began to study not only the CD4, but also the CD8 populations. These plots demonstrate that when both CD4 and CD8 populations are taken into account, there is a significant separation of most of the cGVHD group from the control group, but there is some overlap. Seven patients with cGVHD were treated with ECP

Abnormal Central/Effector Memory T Cell Ratios

Figure 3. Plots combine data for peripheral blood CD4⫹ and CD8⫹ T-cell subpopulations, respectively, as shown from scatter plots in Figure 1B (vertical axis) and Figure 2 (horizontal axis) for subset of healthy donors (left) and patients with symptomatic cGVHD (right) for which both sets of data are available. Regular symmetric ellipse was drawn and placed to tightly include all values for healthy donors and then reproduced over data similarly plotted for patients with cGVHD. Lines were produced by best-fit algorithm of regression analysis as indicated in formulae in each graph.

as indicated earlier. All 7 patients demonstrated a response to ECP in that performance criteria stabilized or improved while corticosteroid dosage was decreased by at least 30% during the period of treatment (ECP treatments occurred twice weekly for 3 months and then twice every other week for the second 3 months). Tlymphocyte subset analysis was performed before the start of ECP and at the end of the 6-month period of ECP treatments. Shown in Figure 4A is the paired preand post-ECP analysis of the ratio of CD4 central memory to effector memory cells (same measurements as in Figure 1B, but plotted as paired bar graphs). In 6 of 7 patients there was a normalization (increase) in this ratio

in the post-ECP sample relative to the pre-ECP and this change was statistically significant (P ⬍ .046). When the percent of CD8 T cells that were central memory cells were plotted in Figure 4B there was a decrease toward normal values in 5 of the 7 patients and in many cases this decrease was quite large. However, this trend for the CD8 central memory cell values did not reach statistical significance for the group of 7 as a whole. DISCUSSION In this study, we demonstrate that there is a highly significant increase in central memory CD8⫹ T cells

Figure 4. Paired measurements of ratio of CCR7⫹CD62Lhigh/CCR7⫺CD62Llow CD4⫹ T cells (percent in upper right quadrant divided by percent in lower left quadrant) (A) and percent of CD8⫹ CD45RO⫹T-cell that are also CCR7⫹ (i.e., percent of central memory CD8⫹ T cells) (B) from each of 7 patients before (blue) and after 6 months of ECP treatments (red).

BB&MT

27

K. Yamashita et al.

together with a highly significant decrease in central memory CD4⫹ T cells associated with cGVHD. Furthermore, we show that this abnormality of the balance of central and effector memory cells in cGVHD was progressively normalized after ECP treatment in association with improvement of clinical symptoms. Because of the small number of patients studied with ECP it is not possible to say whether the clinical improvement achieved with ECP is a result of normalization of central and effector memory cell ratios, nor can we determine yet whether such laboratory changes are predictive of clinical response. The purpose of the current pilot study was to search for abnormalities of T-lymphocytes that appear to normalize after ECP treatment for cGVHD to identify candidate blood cell markers that should be examined in a larger series of patients undergoing ECP or other types of treatments for cGVHD. In all 7 patients participating in the program of ECP treatments there was clear evidence of a trend toward normalization of the central/effector memory cell patterns in both CD4 and CD8 cells, and this trend was statistically significant for the group of 7 as a whole for at least the CD4 central memory to effector memory ratio. Alloreactive donor-derived T cells are believed to play an important role in the pathogenesis of cGVHD; however, consistent statistically significant patterns of changes in T cells associated with cGVHD have been difficult to define. Many clinical manifestations of cGVHD overlap with features of autoimmune disorders, thus, there may be similar patterns of changes in lymphocyte phenotypes in these conditions. Notably, an increase in central memory CD8⫹ T cells has been reported in patients with rheumatoid arthritis, but not patients with systemic lupus erythematosus [15]. CD8⫹ central memory T cells have been reported to be heterogenous with respect to expression of CCR4, which may discriminate cells that are destined to be committed to the TC2 pathway (CCR4⫹) versus a cytotoxic TC1 pathway (CCR4⫺) [16]. Although we documented an increase in the central memory CD8⫹ T-cell group as a whole, when we further delineated the percentage of these that were CCR4⫹ versus those that were CCD4⫺, we did not see a difference between patients with cGVHD and healthy donors. We were interested in assessing this because Fowler and Gress [17] have reported that allospecific donor Tc2 cells reduce GVHD in murine transplantation models. Despite this complexity it is possible to differentiate central and effector memory T-cell roles in immune function. For example, when stimulated in vitro by T-cell receptors, central memory T cells produce predominantly interleukin (IL)-2, whereas effector memory T cells produce reduced levels of IL-2, but significant amounts of interferon-␥ and IL-12 (TH1/ TC1) or IL-4, IL-5, and IL-10 (TH2/TC2) [11]. 28

There is controversy regarding the differentiation pathway of these cells, including specific precursor/ progeny relationship among naive, central memory, and effector memory T cells [18,19]. Furthermore, there are likely differences between CD4⫹ and the CD8⫹ T-cell populations in this regard. Lanzavecchia and Sallusto [20] proposed a lineage model (applied primarily to CD4⫹ cells) and suggested that naive T cells differentiate first into central memory, and then into effector memory cells. Wherry et al. [21] have proposed an alternative pathway whereby naive CD8⫹ T cells differentiate into effector cells, then effector memory cells, and finally into central memory cells. Baron et al. [22] have proposed that CD8⫹ central and effector memory cells are independent subpopulations that do not interchange. To our knowledge, this is the first report that demonstrates that cGVHD is associated with an imbalance of central and effector memory cells that occurs in both the CD4⫹ and CD8⫹ T-cell populations. Furthermore, our initial findings suggest that ECP is associated with normalization of T-cell population pattern over time. The decrease in CD4⫹ central memory cells and increase in CD8⫹ central memory cells associated with cGVHD suggests different roles played by CD4⫹ and CD8⫹ central memory cells in response to the intensive and unremitting alloimmune antigenic challenge that can occur in the setting of transplantation. Alternatively it is possible that differences in differentiation pathways for the two types of T cells may explain the contrasting abnormal patterns of CD4⫹ versus CD8⫹ T cells in GVHD. As noted above there are a variety of models proposed for differentiation pathways. Recently, Seder and Ahmed [18] proposed a model for CD4⫹ T cells that is slightly different for cells destined for a TH1 type response versus those destined for a TH2 type response. The model proposes that some naive CD4⫹ T cells differentiate either directly into central memory cells or into effector cells, and that central memory cells then give rise to effector memory cells. For the TH1 group, it is proposed that effector cells producing interferon-␥ and related cytokines are end-stage cells progressing to apoptosis, whereas for TH2 cells the effector cells producing IL-4 and related cytokines may progress to either central memory or effector memory cells. In any case the predominant pathway for CD4⫹ T cells is naive to central memory to effector memory and then possibly to effector cells and finally to apoptosis. Seder and Ahmed [18] propose that for CD8⫹ T cells the pathway appears to be a direct conversion of naive to RA⫹ effector to effect memory to central memory and then possibly to effector cells and apoptosis. In attempting to fit our observations to these models of the differentiation pathways for CD4⫹ and CD8⫹ T cells, we propose in a schematic outline what

Abnormal Central/Effector Memory T Cell Ratios

CD4 cell

CD8 cell

healthy donor

healthy donor

naive

Tcm

Tem

naive

Tem Temra

Tcm

Tem Temra

Tcm

cGvHD patient

cGvHD patient naive

Tcm

thymic impairment

Tem

naive

after ECP

continuous allo-stimulation

thymic impairment

continuous allo-stimulation

after ECP

naive

Tcm

?

thymic impairment

Tem

Tem Temra

naive

?

continuous allo-stimulation

thymic impairment

Tcm

continuous allo-stimulation

Figure 5. Proposed scheme indicating differentiation pathways for T cells from healthy individuals, patients with cGVHD, and patients with cGVHD after response to ECP treatment. Arrows from left to right indicate direction of differentiation. Size of circles suggests relative size of that population. Naive, effector memory (Tem), RA⫹ effector memory (Temra), and central memory (Tcm) cells are indicated.

may be occurring in cGVHD (Figure 5). In healthy individuals there is continuous age-appropriate generation of naive CD4⫹ and CD8⫹ T cells. In the case of CD4⫹ T cells the naive cells initially progress to a central memory phenotype that accumulates as the predominant population. By contrast the normal differentiation pathway for CD8⫹ T cells leads from naive cells to RA⫹ effector memory or to effector memory cells that accumulate as the predominant population. It is known that allogeneic stem cell transplantation and cGVHD adversely affect thymic function likely limiting the rate of generation of naive T cells [23]. This would reduce the size of both CD4⫹ and CD8⫹ naive cells. Added to this effect strong allogeneic antigen stimulation would likely drive cells through the pathway further reducing the size of the CD4⫹ central memory population and the CD8⫹ effector memory populations while slightly increasing the CD4⫹ effector memory and CD8⫹ central memory populations. ECP might reverse this process by a number of mechanisms. Induction of apoptosis, particularly that of dendritic antigen-presenting cells, may trigger tolerance perhaps by increasing the numbers of lymphocytes that acquire regulatory function. Furthermore, the reduction in response to antigenic stimulation might both reduce ongoing thymic damage/dysfunction and reduce the antigen-stimulated progression of lymphocyte differentiation. Thus, we propose that changes in the balance of central memory and effector memory cells occur as a consequence of thymic dysfunction and unremitting antigen stimulation associated with cGVHD. Similarly, normalization of the balance of central memory and effector memory cells associated with ECP is proposed to be a

BB&MT

consequence of the induction of tolerance by mechanisms yet to be defined. Larger prospective studies are warranted to determine if the measurements we have defined are useful prognostic indicators and/or could be used as a measure of and to guide response to therapies.

ACKNOWLEDGEMENT We thank Karen Diggs and Julie Hopkins for outstanding apheresis expertise.

REFERENCES 1. Vogelsang GB. How I treat chronic graft-versus-host disease. Blood. 2001;97:1196-1201. 2. Flowers ME, Parker PM, Johnston LJ, et al. Comparison of chronic graft-versus-host disease after transplantation of peripheral blood stem cells versus bone marrow in allogeneic recipients: long-term follow-up of a randomized trial. Blood. 2002;100:415-419. 3. Mohty M, Kuentz M, Michallet M, et al. Chronic graft-versushost disease after allogeneic blood stem cell transplantation: long-term results of a randomized study. Blood. 2002;100:31283134. 4. Greinix HT, Volc-Platzer B, Knobler RM. Extracorporeal photochemotherapy in the treatment of severe graft-versushost disease. Leuk Lymphoma. 2000;36:425-434. 5. Apisarnthanarax N, Donato M, Korbling M, et al. Extracorporeal photopheresis therapy in the management of steroid-refractory or steroid-dependent cutaneous chronic graft-versushost disease after allogeneic stem cell transplantation: feasibility and results. Bone Marrow Transplant. 2003;31:459-465. 6. Messina C, Locatelli F, Lanino E, et al. Extracorporeal photochemotherapy for pediatric patients with graft-versus-host dis29

K. Yamashita et al.

7.

8.

9.

10.

11.

12.

13.

14.

15.

30

ease after hematopoietic stem cell transplantation. Br J Haematol. 2003;122:118-127. Seaton ED, Szydlo RM, Kanfer E, et al. Influence of extracorporeal photopheresis on clinical and laboratory parameters in chronic graft-versus-host disease and analysis of predictors of response. Blood. 2003;102:1217-1223. Tambur AR, Ortegel JW, Morales A, et al. Extracorporeal photopheresis induces lymphocyte but not monocyte apoptosis. Transplant Proc. 2000;32:747-748. Alcindor T, Gorgun G, Miller KB, et al. Immunomodulatory effects of extracorporeal photochemotherapy in patients with extensive chronic graft-versus-host disease. Blood. 2001;98: 1622-1625. Gorgun G, Miller KB, Foss FM. Immunologic mechanisms of extracorporeal photochemotherapy in chronic graft-versus-host disease. Blood. 2002;100:941-947. Sallusto F, Lenig D, Forster R, et al. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature. 1999;401:708-712. Tedder TF, Penta AC, Levine HB, et al. Expression of the human leukocyte adhesion molecule, LAM1. Identity with the TQ1 and Leu-8 differentiation antigens. J Immunol. 1990;144: 532-540. Yoshida R, Nagira M, Kitaura M, et al. Secondary lymphoidtissue chemokine is a functional ligand for the CC chemokine receptor CCR7. J Biol Chem. 1998;273:7118-7122. Yamashita K, Choi U, Woltz PC, et al. Severe chronic graftversus-host disease is characterized by a preponderance of CD4⫹ effector memory cells relative to central memory cells. Blood. 2004;5:5. Maldonado A, Mueller YM, Thomas P, et al. Decreased effec-

16.

17.

18.

19.

20.

21.

22.

23.

tor memory CD45RA⫹ CD62L- CD8⫹ T cells and increased central memory CD45RA- CD62L⫹ CD8⫹ T cells in peripheral blood of rheumatoid arthritis patients. Arthritis Res Ther. 2003;5:R91-R96. Geginat J, Lanzavecchia A, Sallusto F. Proliferation and differentiation potential of human CD8⫹ memory T-cell subsets in response to antigen or homeostatic cytokines. Blood. 2003;101: 4260-4266. Fowler DH, Gress RE. Th2 and Tc2 cells in the regulation of GVHD, GVL, and graft rejection: considerations for the allogeneic transplantation therapy of leukemia and lymphoma. Leuk Lymphoma. 2000;38:221-234. Seder RA, Ahmed R. Similarities and differences in CD4⫹ and CD8⫹ effector and memory T cell generation. Nat Immunol. 2003;4:835-842. Tough DF. Deciphering the relationship between central and effector memory CD8⫹ T cells. Trends Immunol. 2003;24:404407. Lanzavecchia A, Sallusto F. Dynamics of T lymphocyte responses: intermediates, effectors, and memory cells. Science. 2000;290:92-97. Wherry EJ, Teichgraber V, Becker TC, et al. Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat Immunol. 2003;4:225-234. Baron V, Bouneaud C, Cumano A, et al. The repertoires of circulating human CD8(⫹) central and effector memory T cell subsets are largely distinct. Immunity. 2003;18:193-204. Weinberg K, Blazar BR, Wagner JE, et al. Factors affecting thymic function after allogeneic hematopoietic stem cell transplantation. Blood. 2001;97:1458-1466.