DONOR HELPER T-CELL FREQUENCIES AS PREDICTORS OF ACUTE... : Transplantation (original) (raw)
Allogeneic bone marrow transplantation (BMT*) is an accepted treatment for numerous hematological disorders (1-2). The success rate of BMT has steadily increased in recent years, but graft-versus-host disease (GVHD) is still the major cause of posttransplantation morbidity and mortality (3-4). Improved matcching of donors and recipients for HLA has limited the incidence and severity of acute GVHD after BMT. Despite the current level of sophistication of molecular typing for class I and class II alleles, some disparities still cannot be defined by molecular techniques. A significant proportion (20-40%) of recipients of HLA-identical sibling marrow develop severe, acute GVHD (5). It has been proposed that minor histocompatibility (mH) antigens (6) or as yet undefined HLA subtypes may be responsible for the development of this posttransplantation complication. To search for the best available donor, molecular matching and possibly functional matching aimed at detecting undefined antigens would seem to be necessary to reduce the risk of GVHD. Previously, the mixed lymphocyte reaction (MLR) was the functional assay of choice. This assay has failed to correlate with graft outcome (7), therefore, other functional methods should be investigated.
The active involvement of T lymphocytes in the development of GVHD has been the focus of many studies. The role of these cells was summarized by Vogelsang and Hess in 1994 (4). Ferrara and Burakoff (8) viewed GVHD as a “cytokine storm.” According to these authors, cytokines are initially released in response to tissue damage induced by radiotherapy and chemotherapy, which in turn up-regulates the expression of MHC antigens. With increased MHC expression there is increased recognition of donor/recipient differences by alloreactive T cells in the donor graft. These reactive T cells proliferate and secrete cytokines, in particular, interleukin (IL)-2. The cytokines released by the T cells activate more donor T cells and other mononuclear cells, which in turn induce the secretion of IL-1 and tumor necrosis factor-α. The resulting cycle of cytokine release manifests itself clinically as GVHD. By analyzing donor T cells, their specific subsets, and the cytokines they produce in response to recipient antigenic stimulation, it might be possible to predict those recipients of HLA-identical sibling marrow that develop GVHD after transplantation. A quantitative analysis of T lymphocytes with a defined specificity is possible when a limiting dilution analysis is used in donor assessment. Limiting dilution analysis allows for the measurement of functionally defined lymphocyte subsets (9). Several authors have studied the role of T-cell precursor frequencies and their involvement in cellular immunity (10-15). Helper T lymphocytes have been shown to secrete lymphokines such as IL-2, colony-stimulating factor, or interferon-gamma in response to antigenic stimuli (16). Subsequently, these lymphokines may affect the function of the immune system by influencing the development of antigen-presenting cells, lymphocyte progenitors, and other T cells.
Using a colorimetric bioassay, Theobald et al., Nierle et al., and Bunjes et al. (10-12) established frequencies of IL-2-secreting T-helper cells in the donor, directed against the recipient. In studies using HLA-identical sibling transplants, frequencies determined before and after transplantation were found to correlate with development of both acute and chronic GVHD. Their methodology is somewhat cumbersome and time consuming, involving the production and maintenance of patient Epstein Barr virus-transformed cell lines. Orosz et al. (13) developed a 3-day method, which they tested using HLA disparate donor/recipient combinations. Schwarer et al. (14), using a similar methodology, demonstrated a correlation between helper T lymphocyte precursor (HTLp) frequency and development of acute GVHD in recipients of HLA-identical sibling bone marrow. All three methodologies used murine cytotoxic T lymphocyte lines (CTLL), which have been shown to proliferate only in the presence of human or murine IL-2 or murine IL-4 (17).
Our study, which was based on this methodology, involves an important modification that renders this method more reliable (L.E. Weston, A.F. Geczy, and C. Farrell, manuscript in preparation). Repeated testing of our assay method showed it to be reproducible and sensitive. In this study we examined the correlation between frequency of HTLp in the donor and the development of acute GVHD in 42 recipients of HLA-matched sibling bone marrow.
MATERIALS AND METHODS
Transplant recipient population. Forty-two recipients received bone marrow transplants from HLA-identical sibling donors as treatment for their hematological disorder. Peripheral blood mononuclear cells (PBMC) for both recipient and donor were cryopreserved and stored in liquid nitrogen. Recipients used in this study all received bone marrow transplants from January 1990 to December 1996 and received methotrexate and cyclosporine as prophylaxis for GVHD.
HLA matching. Recipients and donors were selected on the basis of identity using serological typing for HLA-A and HLA-B alleles, whereas HLA-DRB and HLA-DQB alleles were determined by restriction fragment length polymorphism and sequence-specific oligonucleotide (results not shown). Table 1 lists some of the clinical features of the HLA-identical sibling pairs who were investigated in our study. Of the 30 male recipients, 12 had female donors and 18 had male donors. Seven female recipients received marrow from male sibling donors and five received marrow from female donors. Recipients' ages ranged from 7 to 58 years with a mean age of 34.9 years, and donors' ages ranged from 4 to 61 years with a mean age of 35.5 years.
Materials. All cells were grown in culture medium consisting of RPMI 1640 (Cytosystems, Sydney, Australia) buffered with sodium hydrogen carbonate and HEPES supplemented with 20 IU/ml penicillin, 20 μg/ml streptomycin, 20 mM L-glutamine, and 10-5 M 2-mercaptoethanol. The culture medium was supplemented by 7.5% heat-inactivated pooled human serum collected from nontransfused male donors. The serum was screened for its ability to support cell proliferation and for the absence of HLA antibodies.
Peripheral blood was collected from both the recipient and donor before the BMT. After separation by density gradient centrifugation on Ficoll Paque (Pharmacia Biotech, Uppsala, Sweden), PBMC were washed twice and resuspended at 4°C in culture medium (see above) containing 20% dimethyl sulfoxide and 15% human serum, and then frozen in a controlled rate freezer. Cells were stored in liquid nitrogen until required for testing.
CTLL-2 maintenance. The IL-2-dependent murine cell line CTLL-2 was grown in culture medium supplemented with 10 IU human recombinant (r)IL-2/ml (Amersham, Buckinghamshire, UK) and 15% heat-inactivated fetal calf serum. The CTLL-2 cells were cultured in 75 cm2 flasks (Corning Glassworks, Corning, NY) and subcultured every 3 days with fresh culture media. CTLL-2 cells were cryopreserved on the morning of the third day after subculturing, without IL-2. These cryopreserved cells were assessed in doseresponse assays, and their sensitivity was judged acceptable if the cells were capable of detecting 0.01 U of rIL-2. CTLL-2 cells for use in assays were thawed using 20% fetal calf serum in RPMI, then washed, counted, and adjusted to a concentration of 103 cells per milliliter in media without IL-2.
Measurement of the T helper frequencies using limiting dilution analysis. Cryopreserved PBMC, described above, were thawed using 20% pooled normal human serum in culture medium. Cells were washed, counted, and diluted to a concentration of 106 cells/ml in 7.5% heat-inactivated pooled normal human serum diluted in culture medium.
All assays were performed in the graft-versus-host direction with the donor cells as the responders and the recipient cells as the stimulators. Seven serial twofold dilutions were made with 24 replicates at each dilution. Donor cells (50 μl) were added to each well in U bottom 96-well plates (Nunc, Roskilde, Denmark) starting at 50,000 and diluting down to 390 cells per well. Recipient PBMC were gamma-irradiated (25 grays) and 50,000 cells in a volume of 50 μl were added to all wells. Based on previous reports (10, 13, 14, 18), which examined the acceptable dose of radiation on stimulator cells (from 20 to 50 grays), and in our experience with varying levels of irradiation, it was not considered necessary to increase the level of irradiation beyond 25 grays. As these are HLA-identical siblings, an increase in the level of irradiation to 100-grays would significantly reduce the capacity of the recipient cells to function effectively as allostimulators (19). The IL-2 produced by the alloreactive donor T cells was assessed by the addition of CTLL-2.
HTLp controls. PBMC from a third-party individual (HLA-A, -B, and -DR mismatched) were included as controls with each assay. When used as responders, the third-party cells demonstrated the ability of the recipient's cells to stimulate a response. When used as stimulators, they demonstrated the donor cell's ability to respond to alloantigen. The CTLL-2 cells were placed in background media alone to ensure that there was no endogenous stimulation due to the media. To establish the cut-off point for each assay, 24 background control wells containing irradiated recipient cells were set up.
Incubation. All trays were incubated at 37°C in an humidified 5% CO2 atmosphere. After 36 hr of incubation the trays were gamma-irradiated (25 grays) to prevent further proliferation of responder cells, and 103 CTLL-2 cells, prepared as outlined above, were added to each well in a 10 μl volume. After 6 hr all wells were labeled with 1 μCi of tritiated thymidine (Dupont NEN, Boston, MA) and proliferation of the CTLL-2 cells was assessed by the incorporation of radioactivity at the end of a further 16-hr incubation. All trays were harvested using a Skatron (Lier, Norway) harvester onto glass fiber mats. These mats were dried and sealed in a bag with 10 ml of Betaplate Scint (Wallac, Milton Keynes, UK), before counting by liquid scintillation spectrophotometry (LKB 1205 beta plate; Wallac, Turku, Finland).
Statistical analysis. A well was scored positive if the counts per minute were greater than the mean plus 3 SD of the control (irradiated recipient cells only). The frequency of donor IL-2-producing helper cells was determined using maximum likelihood estimation, and the variance was determined by the use of 95% confidence limits (20). Chi-square analysis was used to show that these data were in accordance with single-hit kinetics. Frequencies were only considered on assays where the P value and the chi-square were >0.05 and <10, respectively. Correlation of the helper frequencies and development of acute GVHD was assessed using the Mann-Whitney test and Fisher's exact test.
RESULTS
The results in Figure 1 show dose-response curves obtained from four different batches, each grown on separate occasions. The ability of the cryopreserved CTLL-2 cells to detect human rIL-2 is demonstrated, and there is remarkable similarity in the performance and sensitivity of three of the batches (13/12, 21/8, and 5/96). Having met with our criteria of sensitivity (as outlined in CTLL-2 maintenance), these batches were suitable for use in the assay. The fourth batch, grown on the 8/12, was discarded as it failed to meet our criteria.
To establish the reproducibility of the assay, the HTLp frequency of two control responder/stimulator pairs were determined on a number of occasions. The following individuals were used as responders: P.W.B., HLA-A1, -; B8, -; DR3, -; DP1, 4; and C.H., HLA-A1, 30; B7, 35; DR12, 15. On each occasion the stimulator was N.E.J., HLA-A1, -; B8, -; DR3, -; DP4, -. The cryopreserved PBMC used for these controls were obtained from the individuals on different occasions. The reproducibility of the assay is shown in Table 2. The HTLp frequency remained relatively consistent when tested on six separate occasions.
The HTLp frequency, the grade of acute GVHD, and the MLR relative response (%RR) for each donor/recipient pair are shown in Table 3. HTLp frequencies detected for the study group ranged from 1:1,517 to 1:2,384,263 (with three donors having undetectable frequencies). The grade of acute GVHD was assessed by the Seattle Criteria (21). Six of the donor/recipient pairs were excluded from statistical analysis as they were nonevaluable (the recipient failed to survive longer than 100 days or achieve a grade II or higher GVHD). The MLR%RR was calculated in the direction of GVHD and showed no correlation with the development of acute GVHD.
Figure 2 is a histogram of the HTLp frequencies calculated for this study group. Statistical descriptives for the HTLp frequencies are outlined in Table 4. The data were classified in one of two categories: those with mild acute GVHD (grades 0-I) and those with clinically significant acute GVHD (grades II-IV). A Mann-Whitney test was used to statistically assess the data. The two-sided P value was <0.0001 and was considered extremely significant.
Donor/recipient pairs were divided into high (>1:100,000) and low (<1:100,000) HTLp frequency groups as defined in previous studies (10, 14). Of the 16 sibling donors with a high HTLp frequency, 14 recipients developed clinically significant acute GVHD (grades II-IV). In contrast, of the 20 donors with low (<1:100,000) HTLp frequencies all recipients experienced nil to mild acute GVHD (grades 0-I). Using a 2×2 contingency table and the Fisher's exact test the two-sided P value was <0.0001 and was considered extremely significant (Table 5). These results indicate a strong association between development of acute GVHD by the recipient and the frequency of HTLp detected in the donor's circulation before transplantation.
DISCUSSION
The determination of HTLp frequencies used in this study was a modification of the 3-day method developed by Orosz et al. (13) and Schwarer et al. (14). A crucial factor in the success of this assay was the ability to reproducibly detect the production of IL-2 by the patient-reactive donor T cells. Previously, investigators have relied upon the use of freshly cultured CTLL-2 cells, the sensitivity of which was unknown at the time of testing. We found that the use of fresh CTLL-2 cells was difficult to standardize and control. We overcame this problem with a modification of the methodology by preparing a large batch of reactive “indicator” cells and cryopreserving them. In this way, a ready supply of sensitive cells allowed both intra- and interassay consistency and reliability.
Dose-response curves showed that the CTLL-2 batches are capable of detecting IL-2 at levels at least as low as 0.01 U IL-2/ml. Vie and Miller (22) demonstrated that one IL-2 secreting effector cell, in culture for 3 days, produced on average 1 pg (3 U) of IL-2. We have ensured that the CTLL-2 cells used in our study are capable of detecting levels of IL-2 that are less than 0.01U IL-2/ml. Any batch cryopreserved and subsequently tested that failed to meet this criterion was discarded.
In this study we have been able to demonstrate a highly significant correlation between circulating donor HTLp frequency directed against recipient antigen and the development of clinically significant acute GVHD in the recipient. The determination of alloreactive precursor frequencies by limiting dilution analysis has been shown to be a powerful tool for the prediction of GVHD. This confirms the results found in smaller studies (10, 14), demonstrating that by increasing the sample population size the correlation becomes extremely significant. An assessment of the relationship between HTLp frequency, chronic GVHD, and survival is at present not possible because the number of recipients who have survived for a period of time sufficient to permit analysis is too small.
The HTLp frequency detected in this assay is produced in response to stimulation by alloantigen that has so far not been detected by molecular methods. Between genotypically HLA-identical sibling donor-recipient pairs it is possible that patient-reactive donor T cells are directed against mH antigens, presented in the context of molecules encoded by the MHC (6). Van Els et al. (15) correlated the involvement of T-helper cells responding to mH antigens and the development of acute GVHD in patients after HLA-identical BMT. They found that the mH antigens are not stimulatory in primary responses. This conclusion was based on the observation that 11 of 12 donors tested before transplantation failed to generate an in vitro host-specific T-helper cell response. However, using a primed lymphocyte test, their study showed a significant correlation between mH antigens and the development of acute GVHD. In this study we have also failed to demonstrate a primary response (lack of MLR reactivity as seen in Table 3), however, by using CTLL-2 we were able to detect IL-2 production by donor T cells in response to stimulation by recipient cells. Further investigation is required to establish the nature of the antigenic determinants that are responsible for the HTLp frequencies detected in this study.
In a small preliminary study (our unpublished data) that investigated HTLp frequency in donors of volunteer unrelated donor (VUD) bone marrow transplants, we found no correlation between HTLp frequency and the development of GVHD in the patient after BMT. Therefore, the predictive value of the HTLp for VUD transplants does not seem promising. This is possibly due to the high level of HLA-DPB1 disparity seen in VUD. Petersdorf et al. (23) looked at HLA-DPB1 incompatibility in 129 VUD transplants and found that 78% were disparate. Potolicchio et al. (18) reported an association between high HTLp frequencies and an HLA-DPB1 mismatch at position 69 in VUD transplant pairs. In this laboratory high precursor frequencies have always been observed in donor/recipient pairs with HLA-DP disparity. These results suggest that HLA-DP disparity may mask another more subtle disparity as seen in the matched siblings. The assay system described in this study may thereby fail to provide an appropriate degree of discrimination necessary for the VUD donor selection.
In 1995 DeBruyne et al. (24) published the results of a study of cardiac transplant patients and found that increases in donor-reactive HTLp frequencies preceded acute cardiac allograft rejection episodes. The implication, that the frequency of HTLp may also correlate with solid organ rejection, places this assay in a unique position of being an extremely useful diagnostic tool in the prediction of transplantation outcome.
Relatively easy to perform, rapid, reliable, and cost effective, this assay seems to have a role in the selection of a sibling donor and the monitoring of GVHD prophylaxis necessary for each sibling bone marrow transplant. Additionally, the HTLp assay is not only a predictor of the possible development of acute GVHD after BMT, but possibly also of organ rejection after solid organ transplantation. Additional studies of solid organ transplantation are currently in progress in this laboratory.
The HTLp methodology described in this study is a relevant functional assay that should be performed in conjunction with molecular typing, to ensure the selection of the most suitable sibling donor for BMT.
Acknowledgments. The authors thank the Bone Marrow Transplant Units at The New Children's Hospital, Royal Prince Alfred Hospital, St. Vincents Hospital, Sydney Children's Hospital, and Westmead Hospital for supplying samples and clinical data on their bone marrow transplant recipients. The authors also acknowledge the expert statistical advice of Dr. Abie Ekangaki of Macquarie University. We thank Dr. J. Sullivan for critical reading of the manuscript.
The ability of the cryopreserved CTLL-2 cells to detect human rIL-2, measured as counts per minute of tritiated thymidine incorporation by the proliferating cells. Four batches grown independently of each other are represented as follows: the batch frozen on 5/96 (▴) or on 21/8 (•); and the batch frozen on 13/12 (♦) or on 8/12 (□). 103 CTLL-2 cells were used at each dilution. Batch 8/12 failed to detect levels of rIL-2 at and below 0.01 U, and having failed to meet our criteria was discarded.
Histogram of the HTLp frequencies calculated for the study group. The x axis shows the grade of acute GVHD for each recipient. The y axis is the calculated HTLp frequency for each recipient/donor pair.
Footnotes
Abbreviations: BMT, bone marrow transplantation; CTLL, cytotoxic T lymphocyte line; GVHD, graft-versus-host disease; HTLp, helper T lymphocyte precursor; IL, interleukin; mH, minor histocompatibility; MLR, mixed lymphocyte reaction; PBMC, peripheral blood mononuclear cells; r, recombinant; VUD, volunteer unrelated donor.
REFERENCES
1. O'Reilly RJ. Allogeneic bone marrow transplantation: current status and future directions. Blood 1983; 62: 941.
2. Martin PJ, Hansen JA, Storb R, Thomas ED. Human marrow transplantation: an immunological perspective. Adv Immunol 1987; 40: 379.
3. Ferrara JLM, Deeg HJ. Graft-versus-host-disease. N Engl J Med 1991; 324:667.
4. Vogelsang GB, Hess AD. Graft-versus-host disease: new directions for a persistent problem. Blood 1994; 84: 2061.
5. Bortin MM, Horowitz MM, Mrsic M, Rimm AA, Sobocinski KA. Progress in bone marrow transplantation for leukemia: a preliminary report from the Advisory Committee of the International Bone Marrow Transplantation Registry. Transplant Proc 1991; 23: 61.
6. Goulmy E. Minor histocompatibility antigens in man and their role in transplantation. Transplant Rev 1988; 2: 29.
7. DeGast G, Mickelson E, Beatty P, et al. Mixed leucocyte culture reactivity and graft-versus-host disease in HLA-identical marrow transplantation for leukemia. Bone Marrow Transplant 1992; 9: 87.
8. Ferrara J, Burakoff S. The pathophysiology of acute graft-versus-host disease in a murine transplant model. In: Burakoff S, Deeg HJ, Ferrara J, Atkinson K, eds. Graft-versus-host disease: immunology, pathophysiology and treatment. New York: Dekker, 1990: 9.
9. Sharrock C, Kaminski E, Man S. Limiting dilution analysis of human T-cells: a useful clinical tool. Immunol Today 1990; 11: 281.
10. Theobald M, Nierle T, Bunjes D, et al. Host-specific interleukin-2-secreting donor T-cell precursors as predictors of acute graft-versus-host disease in bone marrow transplantation between HLA-identical siblings. N Engl J Med 1992; 327: 1613.
11. Nierle T, Bunjes D, Arnold R, et al. Quantitative assessment of posttransplant host-specific interleukin-2-secreting T-helper cell precursors in patients with and without acute graft-versus-host disease after allogeneic HLA-identical sibling bone marrow transplantation. Blood 1993; 81: 841.
12. Bunjes D, Theobald M, Nierle T, et al. Presence of host-specific interleukin-2 -secreting T helper cell precursors correlates closely with active primary and secondary chronic graft-versus-host disease. Bone Marrow Transplant 1995; 15: 727.
13. Orosz C, Adams P, Ferguson R. Frequency of human alloantigen-reactive T lymphocytes. Transplantation 1987; 43: 718.
14. Schwarer A, Jiang Y, Brookes P, et al. Frequency of anti-recipient alloreactive helper T-cell precursors in donor blood and graft-versus-host disease after HLA-identical sibling bone-marrow transplantation. Lancet 1993; 341: 203.
15. van Els C, Bakker A, Zwinderman A, et al. Effector mechanisms in graft-versus-host disease in response to minor histocompatibility antigens. II. Evidence of a possible involvement of proliferative T cells. Transplantation 1990; 50: 67.
16. Prystowsky M, Ely J, Beller D, et al. Alloreactive cloned T cell lines. VI. Multiple lymphokine activities secreted by helper and cytolytic cloned T lymphocytes. J Immunol 1982; 129: 2337.
17. Gillis S, Smith K. Long term culture of tumour-specific cytotoxic T cell lines. Nature 1977; 268: 154.
18. Potolicchio I, Brookes P, Madrigal A, et al. HLA-DPB1 mismatch at position 69 is associated with high helper T lymphocyte precursor frequencies in unrelated bone marrow transplant pairs. Transplantation 1996; 62: 1347.
19. Reisaeter A, Thorsby E, Brinchmann J. Analysis of alloreactive helper T lymphocyte precursor frequencies: influence of interleukin-2 produced by the stimulating cells. J Immunol Methods 1996; 189: 65.
20. Fazekas de St. Groth S. The evaluation of limiting dilution assays. J Immunol Methods 1982; 49: R11.
21. Glucksberg H, Storb R, Fefer A, et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HLA-matched sibling donors. Transplantation 1974; 18: 295.
22. Vie H, Miller R. Estimation by limiting dilution analysis of human IL-2-secreting T cells: detection of IL-2 produced by single lymphokine-secreting T cells. J Immunol 1986; 136: 3292.
23. Petersdorf EW, Smith AG, Mickelson EM, et al. The role of HLA-DPB1 disparity in the development of acute graft-versus-host disease following unrelated donor marrow transplantation. Blood 1993; 81: 1923.
24. DeBruyne L, Renlund D, Bishop D. Evidence that human cardiac allograft acceptance is associated with a decrease in donor-reactive helper T lymphocytes. Transplantation 1995; 59: 778.
© Williams & Wilkins 1997. All Rights Reserved.