Biology of Blood and Marrow Transplantation
Volume 10, Issue 11 , Pages 743-747, November 2004

Temporal discordance between graft-versus-leukemia and graft-versus-host responses: A strategy for the separation of graft-versus-leukemia/graft-versus-host reactivity?

  • Parameswaran Hari

      Affiliations

    • Bone Marrow Transplant Program, Medical College of Wisconsin, Milwaukee, Wisconsin, USA
    • Department of Medicine, Medical College of Wisconsin, Milwaukee, Wisconsin, USA
    • ,
  • Brent Logan

      Affiliations

    • Biostatistics, Medical College of Wisconsin, Milwaukee, Wisconsin, USA
    • ,
  • William R. Drobyski

      Affiliations

    • Bone Marrow Transplant Program, Medical College of Wisconsin, Milwaukee, Wisconsin, USA
    • Department of Medicine, Medical College of Wisconsin, Milwaukee, Wisconsin, USA
    • Corresponding Author InformationCorrespondence and reprint requests: William R. Drobyski, MD, Bone Marrow Transplant Program, 9200 W. Wisconsin Ave., Milwaukee, WI 53226

Received 22 May 2004; accepted 25 July 2004.

Article Outline

Abstract 

The graft-versus-leukemia (GVL) effect is often coexpressed with graft-versus-host disease (GVHD), although the temporal kinetics of these responses have not been critically examined. To evaluate this question in the absence of the confounding effects of the conditioning regimen, 23 patients who received donor lymphocyte infusions from HLA-identical siblings and subsequently developed GVHD and/or a GVL response were studied to determine whether these were temporally synchronous events. The GVL effect occurred significantly earlier than GVHD, being that 19 of 23 patients had a sustained GVL response that antedated the onset of clinical GVHD. The median difference between time to GVL and graft-versus-host (GVH) reactivity in the entire cohort was 14 days. There was no correlation between total T-cell dose and the relative onset of GVL versus GVH reactivity, indicating that temporal dissociation of GVL and GVH responses was not a function of the absolute number of infused donor T cells. These data support existing murine bone marrow transplantation studies indicating that GVL and GVH responses are not temporally synchronous events and raise the possibility that targeted elimination of alloreactive donor T cells after bone marrow transplantation may be an effective strategy for the separation of GVL/GVH reactivity.

Key words:  Graft-versus-leukemia effect , Graft-versus-host disease , Donor lymphocyte infusion , Allogeneic bone marrow transplantation

 

Back to Article Outline

Introduction 

Allogeneic bone marrow transplantation (BMT) is curative in part because of a graft-versus-leukemia (GVL) effect mediated primarily by immunocompetent donor T cells transferred in the marrow graft [1]. Unfortunately, in most instances, GVL reactivity is also coexpressed with graft-versus-host disease (GVHD), which is the major complication of allogeneic marrow transplantation [2, 3]. The precise relationship between the T cells that mediate the GVL effect and those responsible for the initiation of GVHD has not been fully resolved. The preponderance of experimental and clinical data support the existence of GVL effector populations that have specificity either for antigens expressed exclusively on leukemia cells or antigens that are shared on both leukemia cells and host tissues [3, 4, 5, 6, 7]. An underlying assumption is that GVL and graft-versus-host (GVH) responses are synchronous events, and attempts to dissociate these 2 events have been predicated on the selection of donor T cells that have preferential specificity for leukemia as opposed to host cell antigens [8, 9, 10, 11, 12]. Studies in animal models, however, have indicated that GVL and GVH reactivity are not necessarily temporally concordant but can occur with differential kinetics after allogeneic BMT [13, 14]. The relevance of these findings to clinical BMT, however, has not been studied. To examine whether this supposition is valid in humans, we selected a population of patients who had received donor lymphocyte infusions (DLI) for the treatment of relapsed leukemia after allogeneic marrow transplantation, in whom the temporal kinetics of GVL and GVH reactivity could be examined in the absence of the confounding effects of the conditioning regimen.

Back to Article Outline

Materials and methods 

Patient Population 

Between 1992 and 2002, 46 adult patients received DLI from HLA-identical sibling donors at the Medical College of Wisconsin for treatment of relapsed hematologic malignancies after allogeneic BMT. A retrospective analysis of these patients and their clinical course after DLI was approved by the Internal Review Board of the Medical College of Wisconsin. Patients were excluded from this analysis if they received DLI for posttransplantation lymphoproliferative disorder (n = 1) or received DLI with thymidine kinase-transduced activated T cells [15] (n = 2). Patients were also not included who were treated with chemotherapeutic agents immediately before or after DLI before a response could be evaluated (n = 10); with rituxan (n = 2); or with prolonged (>2 weeks) post-DLI therapy with immunomodulating agents such as interferon-alfa that were capable of inducing a direct antileukemic effect (n = 4). Patients with chronic myelogenous leukemia (CML) who were treated with hydroxyurea to control increased white blood cell counts immediately after DLI were included because hydroxyurea has not been shown to induce cytogenetic responses in CML. Three patients who underwent a second transplantation immediately after DLI, were infused with granulocyte colony-stimulating factor-mobilized DLI, or had an unassessable GVL response were excluded.

Definitions of GVHD and GVL reactivity 

The onset of GVHD was defined as the first day on which there was clinical or histologic evidence of GVHD. The onset of an antileukemic response was defined as the first day on which there was evidence of a cytogenetic response for patients with CML, whereas a bone marrow examination demonstrating morphologic remission was required for patients with myelodysplastic syndrome. Evidence of a sustained decrease in serum paraprotein levels or computed tomographic scan evidence of nodal regression was required for response in multiple myeloma or non-Hodgkin’s lymphoma, respectively. One patient with CML who was treated in molecular relapse was deemed to have responded after having a molecular response as determined with a nested polymerase chain reaction assay [16]. For a patient to be classified as having had a GVL response, the morphologic, serologic, cytogenetic, or molecular response had to be sustained on at least 2 consecutive determinations.

Back to Article Outline

Results 

Of the 24 patients who were evaluable for GVH and GVL reactivity on the basis of the criteria noted previously, 1 patient had no evidence of a GVH or GVL response and was excluded from the analysis. Three patients had only GVHD, 2 had a GVL response only, and 18 patients had evidence of both GVL and GVH reactivity. The demographics and the temporal kinetics of the GVL and GVH responses in these 23 patients are presented in Table 1. In 13 patients the GVL response occurred before GVHD, in 6 patients GVL and GVH responses were documented concurrently, and in 4 patients GVHD occurred before or in the absence of a GVL response (Figure 1A). Statistical analysis with the sign test demonstrated that the GVL response occurred significantly earlier than the clinical expression of GVHD in these 23 patients (P = .049). The temporal kinetics of GVL and GVH responses in the patients who developed 1 or both are shown in Figure 1B. The difference between time to GVL and GVH reactivity was computed for each patient, and the median difference was determined to be 14 days. The magnitude of the GVL response at the time GVHD was documented was then evaluated in 18 patients who had evidence of both GVL and GVH reactivity. Ten patients had a complete cytogenetic or hematologic response that occurred either before or concurrently with the onset of GVHD, whereas 7 patients had a partial cytogenetic/molecular response (for CML patients) or a hematologic response before GVHD became clinically evident (Table 1). In only 1 patient did GVHD present before any antileukemic response was observed.

Table 1. Characteristics of Patients with GVL and/or GVH Responses
Patient No.Age (y)SexDisease Status at DLIDLI T-Cell Dose (× 108)Onset of GVL after DLI (days)Onset of GVHD after DLI (days)Disease Response to DLI at GVHD OnsetMagnitude of GVL Response at GVHD Onset*Current Clinical Status†
22535FCML, AP4.06789Cy Rem90%→0% Ph+Dead, GVHD, day 2120
30643MCML, AP2.37537NoneNACCR, day 4080
33536MCML, Cy Rel3.42734Cy Resp100%→90% Ph+CCR, day 3163
35024FCML, CP2.62063Cy Rem80%→0% Ph+CCR, day 3550
35240FCML, CP1.0NA36NoneNADead, relapse, day 697
35748FCML, CP2.74949Cy Resp95%→80% Ph+Dead, sepsis, day 1477
38652MMDS3.0NA50NoneNADead, sepsis, day 57
48756MCML, CP0.13232Cy Rem100%→0% Ph+Dead, GVHD, day 71
50625MCML, AP2.882124Cy Resp100%→90% Ph+CCR, day 3226
56342MCML, CP2.62794Cy Resp95%→3% Ph+CCR, day 3228
59330MCML, BC3.0NA31NoneNADead, GVHD, day 56
60235MCML, CP2.959157Cy Rem100%→0% Ph+Dead, relapse, day 375
73250FCML, CP0.127148Cy Rem5%→0% Ph+CCR, day 2562
75545FCML, Cy Rel0.15367Cy Resp70%→40% Ph+Dead, sepsis, day 1024
80148MMM1.03636Hem Resp↓IgGκ paraproteinCR, day 2480
88134MCML, CP2.69797Cy Rem80%→0% Ph+Dead, relapse, day 456
88235MCML, CP2.52648Cy Rem90%→0% Ph+CCR, day 1182
88428MCML, Cy Rel3.05050Cy Rem30%→0% Ph+CCR, day 3030
92742MCML, CP0.98296Cy Rem100%→0% Ph+CCR, day 1629
104742FNHL, Rel0.6146NACRNodal regression by CTCR, day 825
114844MCML, Mol Rel1.15555Mol Resp1%→2% round PCR+MR, day 676
126752MMDS0.185133CR27%→0% blastsDead, relapse, day 679
132128FCML, Cy Rel0.581NACy Rem30%→0% Ph+CCR, day 571

Cy Rel indicates cytogenetic relapse; Mol Rel, molecular relapse; NHL, non-Hodgkin lymphoma; AP, accelerated phase; CP, chronic phase; Cy Resp, cytogenetic response; Cy Rem, cytogenetic remission; Mol Resp, molecular response; CCR, complete cytogenetic response; MR, molecular remission; NA, not applicable; MDS, myelodysplastic syndrome; MM, multiple myeloma; Ig, immunoglobulin; CT, computed tomography; PCR, polymerase chain reaction; BC, blast crisis; Rel, relapsed; Hem Resp, hematological response.

Response criteria were defined as follows: cytogenetic response, a decrease in the percentage of Philadelphia chromosome-positive (Ph+) metaphases on marrow examination; cytogenetic remission, the absence of Ph+ metaphases; molecular response, a decrease in the bcr/abl signal with a semiquantitative nested PCR approach [16]; complete remission (CR), normal blood counts and <5% marrow blasts (for MDS) or complete regression of adenopathy (for NHL).

* The magnitude of the GVL response was assessed by determining the extent of antileukemic response at the time GVHD was first documented. This was assessed by analysis of the most recent marrow examination that antedated GVHD. This was done to avoid biasing the results in favor of a more complete GVL response at the time of GVHD. In patient 884, a marrow examination was performed 5 days after the onset of GVHD and showed cytogenetic remission. Because of the proximity of this examination to the onset of GVHD and the extent of the response, this patient was deemed to have had concurrent onset of GVL and GVH reactivity.

† Patients 357 and 755, who died from sepsis, were in cytogenetic remission at the time of their demise.

  • View full-size image.
  • Figure 1. 

    Temporal kinetics of GVL and GVH responses. A, Number of patients in whom GVL and GVH reactivity occurred concurrently (white bar), in whom a GVL response antedated GVHD (black bar), or in whom a GVL effect was documented after the onset of GVHD (gray bar). B, Percentage of patients who had either a GVL (closed squares) or a GVH (open squares) response after the administration of DLI. C, Scatterplot of total T-cell dose versus the interval between the relative onset of GVL and GVH reactivity in 18 patients who had both GVL and GVH responses. Data points represent individual patients.

Back to Article Outline

Discussion 

These data indicate that GVL and GVH reactivity are not temporally synchronous events but can occur with differential kinetics after infusion of alloreactive donor T cells. In the vast majority of patients, the GVL response antedated clinical GVHD. Moreover, the magnitude of the GVL response in many of these patients was such that they were in remission before or at the time GVHD was documented, demonstrating that a quantitatively robust GVL response can occur before GVHD initiation. There was no correlation between the total T-cell dose and the relative onset of GVL versus GVH reactivity (Figure 1C; P = .67,r = −0.01 by Pearson correlation coefficient), indicating that temporal dissociation of GVL and GVH responses was not a function of the absolute number of infused donor T cells. Although most patients in this study had CML, there is evidence from animal models that this observation may be valid in other hematologic diseases as well. Boranic and Tonkovic [13] examined the tempo of myeloid and lymphoid leukemia eradication in mice that underwent transplantation with allogeneic marrow grafts by adoptively transferring cells from various organs into secondary recipients. They demonstrated that by 7 days after BMT, when mice had yet to develop clinical GVHD, leukemia could no longer be transferred into secondary animals. In more recent studies, we examined the kinetics of leukemia eradication by using a murine acute lymphoblastic leukemia transplant model in which donor T cells were genetically engineered to express a thymidine kinase suicide gene [14]. These data demonstrated that, under conditions in which untreated mice could not be cured of leukemia without dying from GVHD, the timely administration of ganciclovir was able to result in the elimination of leukemia without lethal GVHD, indicating that GVL and GVH reactivity were temporally discordant events. In both studies, mice were administered pretransplantation total body irradiation; this indicates that this observation is also valid in conditioned recipients. Findings by Michalek et al. [11] further support this premise in human acute myelogenous leukemia. Using T-cell receptor-β CDR3 sequences to define putative GVH- and GVL-reactive T-cell clones, they were able to track the appearance of these clones in the peripheral blood of a patient after transplantation. These studies showed that there was an earlier and quantitatively greater increase in the putative GVL-reactive T-cell clone in the peripheral blood in comparison to the GVH-reactive clone, suggesting that the temporal kinetics of these 2 responses were discordant. Collectively, these data provide support for the premise that GVL and GVH responses can occur with differential kinetics after allogeneic stem cell transplantation. Why leukemia cells would be preferentially targeted by donor T cells is an unresolved but interesting question. Possible explanations include, but are not limited to, a more restricted expression of antigens on the leukemia for which donor T cells have a greater precursor frequency or affinity, and the relative accessibility of leukemia cells in marrow and nodal sites that confer enhanced susceptibility to immune-mediated attack.

The findings of this study may have important clinical implications given the technology that now exists to regulate donor T-cell survival in vivo in humans. The incorporation of a suicide gene through retroviral transduction [17, 18] or the triggering of suicide genes via chemical dimerizing agents [19, 20] are examples of existing and emerging technologies that allow for the selective elimination of alloreactive donor T cells. When coupled with sensitive cytogenetic and molecular tools for the monitoring of residual disease, the timed elimination of alloreactive donor T cells in patients deemed to be in remission may be an effective way to separate GVL and GVH reactivity after BMT.

Back to Article Outline

References 

  1. Truitt RL , Johnson BD . Principles of graft versus leukemia reactivity . Biol Blood Marrow Transplant . 1995;1:61–68
  2. Sullivan KM , Weiden PL , Storb R , et al.   Influence of acute and chronic graft versus host disease on relapse and survival after bone marrow transplantation from HLA-identical siblings as treatment of acute and chronic leukemia . Blood . 1989;73:1720–1728
  3. Horowitz MM , Gale RP , Sondel PM , et al.   Graft-versus-leukemia reactions after bone marrow transplantation . Blood . 1990;75:555–562
  4. Truitt TR , Shih CY , LeFever AV , Tempelis LD , Andeani M , Bortin MM . Characterization of alloimmunization-induced T lymphocytes reactive against AKR leukemia in vitro and correlation with graft versus leukemia activity in vitro . J Immunol . 1983;131:2050–2058
  5. Falkenburg JH , Wafelman AR , Joosten P , et al.   Complete remission of accelerated phase chronic myeloid leukemia by treatment with leukemia-reactive cytotoxic T lymphocytes . Blood . 1999;94:1201–1208
  6. Rezvani K , Grube M , Brenchley JM , et al.   Functional leukemia-associated antigen-specific memory CD8+ T cells exist in healthy individuals and in patients with chronic myelogenous leukemia before and after stem cell transplantation . Blood . 2003;102:2892–2900
  7. Kloosterboer FM , Van Luxemburg-Heijs SA , Soest RV , et al.   Direct cloning of leukemia-reactive T cells from patients treated with donor lymphocyte infusion shows a relative dominance of hematopoiesis-restricted minor histocompatibility antigen HA-1 and HA-2 specific T cells . Leukemia . 2004;18:798–808
  8. Mutis T , Verdijk R , Schrama E , Esendam B , Brand A , Goulmy E . Feasibility of immunotherapy of relapsed leukemia with ex vivo generated cytotoxic T lymphocytes specific for hematopoietic system-restricted minor histocompatibility antigens . Blood . 1999;93:2336–2341
  9. Bonnet D , Warren EH , Greenberg PD , Dick JE , Riddell SR . CD8 (+) minor histocompatibility antigen-specific cytotoxic T lymphocyte clones eliminate human acute myeloid leukemia stem cells . Proc Natl Acad Sci U S A . 1999;96:8639–8644
  10. Marjit WA , Heemskerk MH , Kloosterboer FM , et al.   Hematopoiesis-restricted minor histocompatibility antigens HA-1 or HA-2-specific T cells can induce complete remissions of relapsed leukemia . Proc Natl Acad Sci U S A . 2003;100:2742–2747
  11. Michalek J , Collins RH , Durrani HP , et al.   Definitive separation of graft versus leukemia and graft versus host specific CD4+ T cells by virtue of their receptor beta loci sequences . Proc Natl Acad Sci U S A . 2003;100:1180–1184
  12. Crough T , Nieda M , Morton J , Bashford J , Durrant S , Nicol AJ . Donor-derived b2a2-specific T cells for immunotherapy of patients with chronic myeloid leukemia . J Immunother . 2002;25:469–475
  13. Boranic M , Tonkovic I . Time pattern of the antileukemic effect of graft-versus-host reaction in mice . Cancer Res . 1971;31:1140–1147
  14. Drobyski WR , Gendelman M , Vodanovic-Jankovic  , Gorski J . Elimination of leukemia in the absence of lethal graft versus host disease . J Immunol . 2003;170:3046–3053
  15. Burt RK , Drobyski W , Seregina T , et al.   Herpes simplex thymidine kinase gene transduced donor lymphocyte infusions for relapsed hematologic malignancies following allogeneic hematopoietic stem cell transplantation . Exp Hematol . 2003;31:903–910
  16. Hessner MJ , Roth MS , Drobyski WR , Baxter-Lowe LA . Development of a sensitive, highly controlled assay for molecular detection of the Philadelphia chromosome in patients with chronic myelogenous leukemia . Genet Anal Tech Appl . 1994;11:90–94
  17. Link CJ , Burt RK , Traynor AE , Drobyski WR . Adoptive immunotherapy for leukemia (donor lymphocytes transduced with the herpes thymidine kinase gene for remission induction) . Hum Gene Ther . 1998;9:115–134
  18. Bonini C , Ferrari G , Verzeletti S , et al.   HSV-TK gene transfer into donor lymphocytes for control of allogeneic graft-versus-leukemia . Science . 1997;276:1719–1724
  19. Berger C , Blau CA , Huang ML , et al.   Pharmacologically regulated fas-mediated death of adoptively transferred T cells in a nonhuman primate model . Blood . 2003;103:1261–1269
  20. Thomis DC , Marktel S , Bonini C , et al.   A fas-based suicide switch in human T cells for the treatment of graft versus host disease . Blood . 2001;97:1249–1257

PII: S1083-8791(04)00397-0

doi:10.1016/j.bbmt.2004.07.006

Biology of Blood and Marrow Transplantation
Volume 10, Issue 11 , Pages 743-747, November 2004

Article Tools

* Image Usage & Resolution