Volume 12, Issue 4 , Pages 414-421, April 2006
Graft-versus-Host Reactions and the Effectiveness of Donor Lymphocyte Infusions
Article Outline
Abstract
We retrospectively analyzed 83 consecutive recipients of donor lymphocyte infusions (DLI) after allogeneic transplantation for factors associated with disease response and graft-versus-host disease (GVHD). DLI was highly effective in relapsed chronic phase chronic myeloid leukemia (CML), with 71% of patients achieving durable complete remissions (CR). In relapsed acute myeloid leukemia, DLI led to durable CRs in 31% of patients; the rate was <20% in all other diseases. Achieving full donor chimerism and GVHD were predictive of CR. Grade II or higher acute or chronic GVHD occurred in 36 (43%) patients and contributed to death in 13 (16%). Even more patients, 33 (40%), died of their underlying malignancy, including 10 who developed active GVHD. In relapsed CML, most durable CRs occurred without clinically apparent GVHD, yet all responders achieved full donor chimerism, including 6 with coincident normal host hematopoiesis at the time of DLI. Thus, in CML, potent lymphohematopoietic graft-versus-host reactions occurred even in the absence of clinically apparent GVHD; this confirms the ability to dissociate these processes and argues against a leukemia-specific immunologic effect. DLI clearly has efficacy in the treatment of relapsed disease after allogeneic transplantation. However, with the exception of CML, most patients die of their underlying disease because of insufficient antitumor activity even with active GVHD.
Key words: Graft-versus-host reactions , Donor lymphocyte infusions , Allogeneic transplantation , Disease response
Introduction
Allogeneic blood or marrow transplantation (alloBMT) has the potential to produce long-term disease-free survival and even cures in some patients with hematologic malignancies. A major component of the antitumor activity of alloBMT is immunologic and is mediated by donor T cells reacting against host antigens [1, 2, 3]. Perhaps the clearest evidence of immunologic antitumor activity is the demonstration that donor lymphocyte infusions (DLIs) can induce remissions in patients who relapse after alloBMT [4, 5, 6, 7, 8]. This is further supported by the finding that patients who relapse after alloBMT can achieve remissions by withdrawal of immunosuppression [9, 10, 11].
The efficacy of DLI is a function of the underlying disease and disease status at the time of DLI. The best responses are seen in patients with chronic myeloid leukemia (CML) in cytogenetic or chronic phase (CP) relapse; durable remissions are achieved in 70% to 80% of these patients, often at T-cell doses (≤1 × 107 CD3+ T cells per kilogram) that do not produce clinically significant graft-versus-host disease (GVHD) [6, 12, 13]. DLI is less successful in other hematologic malignancies, with response rates ranging from 10% in acute lymphocytic leukemia (ALL) to as high as 40% in multiple myeloma (MM), myelodysplasia (MDS), chronic lymphocytic leukemia, and indolent non-Hodgkin lymphoma (NHL); however, many of the responses are not durable [6, 14, 15, 16, 17, 18, 19, 20, 21, 22].
We report a single-institution experience of DLI in 83 consecutive patients with relapsed hematologic malignancies over a 7-year period. We demonstrate that the percentage of donor chimerism at the time of DLI is a powerful predictor of response, as is the development of GVHD. Furthermore, patients who achieved full donor chimerism after DLI were 22-fold more likely to achieve a complete remission (CR) than patients with residual host cells present (P < .01). Although essential for achieving a CR, except in CML, full donor chimerism was insufficient to prevent relapse.
Methods and materials
All patients aged 18 years and older who received DLI(s) from an HLA-identical sibling after a myeloablative allogeneic transplantation for a hematologic malignancy at Johns Hopkins Hospital were included. The treatment period extended from October 1995 through January 2003. DLI was given for relapsed disease after alloBMT. Patients were eligible to receive DLI if they had evidence of donor chimerism (≥5%) and were not being actively treated for GVHD. All patients gave informed consent for DLI and follow-up as approved by the Institutional Review Board of the Johns Hopkins Hospital and University. Follow-up on disease status and survival was completed through October 2004. Complete follow-up information was available on 77 patients. Patients with severe GVHD and overwhelming infections were considered deaths from GVHD.
Response Criteria and Measurement of Donor/Host Chimerism
Responses were assessed by using standard response criteria for each of the diseases being treated. Donor-host chimerism was determined by 1 of 3 methods on peripheral blood and bone marrow samples: the use of restriction fragment linked polymorphisms [23, 24], polymerase chain reaction analysis of variable nucleotide tandem repeats [25, 26], or by fluorescence in situ hybridization using X and Y chromosome probes, if informative [27]. Pre-DLI chimerism was measured within 4 weeks before the receipt of DLI. Post-DLI chimerism was measured 2 to 3 months after the administration of DLI.
Graft-versus-Host Disease
Patients were evaluated using the Keystone staging system for acute GVHD and the International Bone Marrow Transplantation Registry criteria for chronic GVHD [28, 29].
Chemotherapy/Immune Adjuvants
Twenty-five patients received chemotherapy within 4 weeks before their DLI as part of a planned combination approach. These patients all had aggressive disease and were in need of cytoreduction to allow time for development of an immunologic antitumor effect from DLI. In each case, DLI was given without assessing response to chemotherapy. Two patients received chemotherapy within the first 4 weeks after DLI because of rapid disease progression. Two patients with CML received interferon alfa concomitant with their first DLI, whereas interferon alfa or interleukin 2 was given to 8 patients as an immune adjuvant in the setting of DLI dose escalation.
Donor Lymphocyte Infusions
Doses of DLI ranged from 0.1 to 5 × 108 CD3+ T cells per kilogram. Patients with CML in CP or cytogenetic relapse received 0.1 × 108 CD3+ T cells per kilogram as their initial dose. The initial T-cell doses for other patients ranged from 0.1 to 1.5 × 108 CD3+ T cells per kilogram. Patients with aggressive diseases, relapsed acute leukemia, or aggressive lymphomas (Hodgkin’s disease [HD] and diffuse large cell [DLC]) were given an initial T-cell dose of 1 × 108 CD3+ T cells per kilogram. Patients with myeloma or MDS/myeloproliferative disorder (MP) received 0.1 to 0.5 × 108 CD3+ T cells per kilogram according to the protocol they were enrolled on. Dose escalation was used in patients who were alive and free of GVHD and who did not achieve a CR after the initial dose of DLI or whose disease had clearly progressed at least 3 months after the initial DLI. One patient with ALL was given escalated DLI at 2 months after the initial dose for evidence of disease progression. All other patients who received dose escalated DLI received it between 3 and 8 months after the initial DLI.
Statistical Methods
Statistical analysis was performed by using SPSS 11.0 (SPSS Inc., Chicago, IL), Stata (StataCorp, College Station, TX), and SAS (SAS Institute, Cary, NC). Progression-free survival was calculated for patients who achieved a remission, from the date of DLI to the date of disease progression or death from any cause. Overall survival was calculated from the date of DLI to the date of death. Patients without evidence of progression were censored at the time of last contact. Follow-up on disease status and survival was completed through October 2004. Complete follow-up information was available on 77 patients. The Fisher’s exact test was used to compare binomial outcomes between groups. Kaplan-Meier methods were used to calculate survival probability, and log-rank tests were used to test for differences in survival curves between diagnostic groups. Cox proportional hazards models and χ2 tests were used to determine which variables were associated with the onset and development of GVHD. These analyses were performed by using only the first DLI that each patient received so as not confound the interpretation by including the same patient twice. Logistic regression models were also used to describe which characteristics were associated with a CR.
Results
Patient Characteristics
DLIs were given to 83 patients (Table 1) with relapsed hematologic malignancies between October 1995 and January 2003. Of these 83 patients, 64 received 1 infusion, and 19 received ≥2 infusions. The total number of DLIs given was 112. The median age was 45 years (range, 21-67 years); 49 patients were men, and 34 were women.
Table 1. Patient Characteristics
| Variable | Data |
|---|---|
| Median age, y (range) | 45 (21-67) |
| Type of allogeneic transplantation | |
| T-cell depleted | 73 |
| T-cell replete | 7 |
| Second allogeneic | 2 |
| Allogeneic after autologous | 1 |
| Reason for DLI | |
| Relapsed disease | 83 |
| CML in CP/cytogenetic relapse | 17 |
| CML in accelerated phase/blast crisis | 5 |
| Multiple myeloma | 20 |
| AML | 13 |
| MDS/MP | 12 |
| HD | 8 |
| Aggressive NHL | 4 |
| ALL | 3 |
| Follicular NHL | 1 |
| Median interval (d) between BMT and DLI (range) | |
| CML in CP (n = 17) | 916(232-5339) |
| All other diseases (n = 66) | 427(69-3702) |
| Median follow-up, d (range) | |
| CML in CP (n = 17) | 857(487-2657) |
| All other diseases (n = 66) | 383(10-3031) |
| T-cell dose (112 infusions) (T cells per kilogram) | |
| 1 × 107 CD3+ | 29 |
| 3-5 × 107 CD3+ | 27 |
| 0.7-1.5 × 108 CD3+ | 53 |
| 2-5 × 108 CD3+ | 3 |
| Number of T-cell infusions per patient (n = 83) | |
| One | 64 |
| Two or more | 19 |
DLI was given to 73 recipients of T cell–depleted myeloablative allogeneic transplants and 7 recipients of unmanipulated myeloablative allogeneic transplants. Three patients received 2 myeloablative transplants before DLI: 2 had a second alloBMT (1 T-cell depleted [by elutriation] and 1 unmanipulated), and 1 received alloBMT for disease relapse after an autologous transplantation. DLI was given to treat the following relapsed diseases: CML in CP or cytogenetic relapse (n = 17), CML in accelerated phase/blast crisis (AP/BC; n = 5), MM (n = 20), acute myeloid leukemia (AML; n = 13), MDS/MP (n = 12), HD (n = 8), aggressive NHL (n = 4), ALL (n = 3), and follicular NHL (n = 1; Table 1).
The median interval between transplantation and first DLI was 492 days (range, 69-5339 days). For patients with CML in CP or cytogenetic relapse, the median time to first DLI was 916 days (range, 232-5339 days), which is significantly longer than the median time to DLI for all other indications combined: 427 days (range, 69-3702 days; Table 1).
DLI for Relapsed Disease
Salvage chemotherapy was given to 25 patients with relapsed disease, from 3 to 28 days before DLI. The diseases treated included AML (n = 7), HD (n = 3), MM (n = 5), MP (n = 2), ALL (n = 2), CML in AP/BC (n = 2), MDS (n = 2), aggressive NHL (n = 1), and indolent NHL (n = 1). The regimens varied and included cyclophosphamide, etoposide, topotecan, cytarabine, and doxorubicin. In each case, the decision to give chemotherapy was based on the aggressiveness of the disease and the need to achieve cytoreduction to allow time for development of an immunologic antitumor effect. Response to chemotherapy was not assessed before the administration of DLI because it was a planned combination approach. Of the 25 patients, 6 achieved CR, all after developing GVHD with DLI. Only 2 of these 6 patients are alive in continuous CR; 4 patients subsequently died (2 of disease and 2 of GVHD in CR). Three patients achieved partial remissions (PR), all with GVHD; 2 subsequently relapsed and died of their disease, and the other died of GVHD with an ongoing response.
As has been found in most series, patients with CML in CP or cytogenetic relapse had the highest CR rate, at 77% (Table 2), with a similar rate for both CP and cytogenetic relapse (Table 3). DLI also had significant activity in most other relapsed diseases, with CR rates of 50% in aggressive NHL, 33% in ALL, 46% in AML, 30% in MM, 20% in CML in AP/BC, 15% in MDS/MP, and 13% in HD (Table 2). With the exception of CML in CP or cytogenetic relapse, for which all responding patients except 1 (who died of GVHD) remain in CR, with a median follow-up of 857 days (range, 487-2657 days), relapses were common in other diseases. Thus, of the 66 patients with diseases other than CML in CP or cytogenetic relapse, only 14% (9/66) are alive in CR, with a median follow-up of 383 days (range, 10-3031 days) and a median survival of 321 days (range, 10-3031 days; Table 3). Although 31% of relapsed AML patients remain in remission after DLI (median follow-up of 1526 days; range, 1495-1551 days), the durable remission rate is ≤20% in all other diseases.
Table 2. DLI Outcomes
| Variable | Initial T-Cell Dose | Dose Escalation | Alive in Continuous CR |
|---|---|---|---|
| Response | |||
| CML in CP/cytogenetic relapse (n = 17) | 10/17(59%) | 13/17(77%) | 12/17(71%) |
| AML—relapsed (n = 13) | 6/13(46%) | 6/13(46%) | 4/13(31%) |
| All other relapsed diseases (n = 53) | 18/53(34%) | 19/53(36%) | 5/53(9%) |
| Deaths (n = 55) | |||
| Underlying disease | 33 | ||
| GVHD | 13 | ||
| Infection | 6 | ||
| Other (MI/interstitial pneumonitis/second malignancy) | 3 |
Table 3. Disease-Specific Responses (n = 83)
| Diagnosis | n | CR to Initial Dose | CR/PR with Initial Dose | CR to DLI Including Dose Escalation | CR/PR to DLI Including Dose Escalation | Relapse after DLI | Alive in Continuous CR | Death (Disease, GVHD, Other) | Median OS, d (range) |
|---|---|---|---|---|---|---|---|---|---|
| CML | 17 | 10(59%) | 10(59%) | 13(77%) | 13(77%) | 0/13 | 12(71%) | 2(1,1,0) | NR(449-2657) |
| Cyto | 12 | 8 | 8 | 9 | 9 | 8(67%) | |||
| CP | 5 | 2 | 2 | 4 | 4 | 4(80%) | |||
| CML (AP/BC) | 5 | 1(20%) | 1(20%) | 1(20%) | 1(20%) | 0/1 | 1(20%) | 4(2,0,2) | 1252(10-2392) |
| AML relapsed | 13 | 6(46%) | 6(46%) | 6(46%) | 6(46%) | 2/6 | 4(31%) | 9(4,2,3) | 127(15-1551) |
| MM | 20 | 6(30%) | 8(40%) | 6(30%) | 8(40%) | 2/8 | 3(15%) | 15(11,4,0) | 614(25-3031) |
| MDS/MP | 12 | 2(15%) | 2(15%) | 2(15%) | 2(15%) | 0/2 | 0 | 10(5,2,3) | 210(49-1027) |
| NHL aggressive relapsed | 4 | 2(50%) | 3(75%) | 2(50%) | 3(75%) | 3/3 | 0 | 4(2,2,0) | 91(91-1312) |
| HD | 8 | 1(13%) | 3(38%) | 1(13%) | 3(38%) | 2/3 | 1(13%) | 7(6,1,0) | 512(475-1297) |
| ALL relapsed | 3 | 0(0%) | 0(0%) | 1(33%) | 1(33%) | 1/1 | 0 | 3(1,1,1) | 34(22-383) |
| NHL follicular | 1 | 0(0%) | 0(0%) | 0(0%) | 0(0%) | 0/0 | 0 | 1(1,0,0) | N/A |
Additional dose-escalated DLI was given to 19 patients who did not achieve CR or had progression within 3 months after the initial T-cell infusion. The diseases treated included CML-CP (n = 6), MDS/MP (n = 5), MM (n = 2), CML in AP (n = 2), small lymphocytic lymphoma (n = 1), AML (n = 1), ALL (n = 1), and HD (n = 1). Of the 6 CML patients in CP or cytogenetic relapse who received dose escalation, 3 (50%) achieved CR. Of the 13 patients with diseases other than CML in CP or cytogenetic relapse, only 1 (7%) achieved CR with dose escalation, and the CR was not durable despite the development of GVHD.
Graft-versus-Host Disease
The overall incidence of acute (grade II or higher) or chronic GVHD was 36% after the first dose of DLI and was 43% when taking dose escalation into consideration (Table 4). The overall incidence of acute GVHD (grade II or higher) was 35%. Chronic GVHD was seen in 33% of the patients treated. The incidence of acute and chronic GVHD did not differ significantly with the T-cell dose given; it was 32%, 42%, 36%, and 33%, respectively, at CD3+ T-cell doses of 1 × 107, 3 to 5 × 107, 0.7 to 1.5 × 108, and 5 × 108 cells per kilogram. The median time to the development of acute GVHD was 35 days (range, 12-159 days) and to chronic GVHD was 85 days (range, 57-108 days) after DLI. Fatal GVHD was seen in 16% (13/83) of the patients treated in this series.
Table 4. GVHD According to T-Cell Infusion
| Variable | Initial T-Cell Dose (n = 83) | All Doses Including Dose Escalation (n = 112 doses) |
|---|---|---|
| Acute GVHD | 26/83 (31%) | 29/83 (35%) |
| Stage II | 10 | 12 |
| Stage III | 5 | 5 |
| Stage IV | 11 | 12 |
| Chronic GVHD | 21/83 (25%) | 27/83 (33%) |
| Limited | 5 | 4⁎ |
| Extensive | 16 | 23 |
| Antecedent acute | 17 | 20 |
| Acute or chronic GVHD | 30/83 (36%) | 36/83 (43%) |
⁎ One patient with limited-stage chronic GVHD developed extensive chronic GVHD with dose escalation. |
Causes of Death
To date, 55 (66%) of 83 patients have died. By far, the most common cause of death was relapsed disease (60%; 33/55). GVHD caused or contributed to death in 24% (13/55) of patients who died (5 in CR, 2 in PR, and 2 without measurable disease). Six patients died of infection (including 1 in CR and 1 in PR); 3 patients died of other causes (myocardial infarction, interstitial pneumonitis, and second malignancy). A Kaplan-Meier survival curve stratifying survival for patients with CML in CP or cytogenetic relapse versus all other diseases is shown in Figure 1, thus confirming earlier findings that demonstrated superior survival for patients with CML in CP or cytogenetic relapse as compared with all other diseases (P < .001).
Figure 1.
Kaplan-Meier plot—overall survival after DLI (dashed lines indicate 95% confidence intervals). This Kaplan-Meier survival curve compares overall survival for patients with CML in CP or cytogenetic relapse with all other indications. The Fisher exact test for the relationship between death and CML group was P < .001.
Factors Predictive of GVHD and Response
By using a χ2 test, patients with relapsed disease who developed grade II or higher acute or chronic GVHD were significantly more likely to achieve a CR than patients who did not (P < .001). When the analysis was repeated by looking at the CML in CP subgroup, the development of GVHD did not correlate with achieving CR (P = .10), whereas the effect persisted in patients who received DLI for all other relapsed diseases (P < .01). Univariable logistic regression models were fit to identify variables associated with a response to DLI (Table 5). In these models, the only variable that was predictive of a response was the reason for DLI; patients who received DLI for CML in CP (cytogenetic relapse included with CP for this analysis) were more likely to respond than patients who received DLI for relapsed disease other than CML in CP. Univariable and multivariable logistic regression models were fit to describe the relationships among chimerism, GVHD, and response. The percentage of donor chimerism before DLI had a significant effect on the development of GVHD and subsequent responses. Patients with >50% donor chimerism before DLI were 4.5 times (95% confidence interval, 1.7-12.5 times; P < .01) more likely to achieve a CR than patients whose donor chimerism was <50% at the time of DLI (Table 5). Furthermore, according to the Fisher t test, patients who achieved full donor chimerism were 21-fold more likely to achieve a CR than patients with residual host cells present (P < .001). No other variable, including age, patient/donor sex, time to DLI, or T-cell dose, reached statistical significance in these models when the group was analyzed as a whole. When the subgroup of patients with AML was analyzed with the Wilcoxon rank test, the time between BMT and DLI was statistically significant: the median time was 544 days (range, 203-1631 days) in the responders, compared with 152 days (range, 69-420 days) in the nonresponders (P = .035).
Table 5. Univariate Logistic Regression Models Describing Variables Associated with CR Taking Only Initial DLI into Consideration (n = 83)
| Variable | Odds Ratio | P Value | 95% CI |
|---|---|---|---|
| Age (y) | |||
| <35 | 1.4 | .58 | 0.41-4.8 |
| 36-50 | 0.7 | 0.26-2.1 | |
| >50 | — | — | |
| Sex | 0.95 | .91 | 0.37-2.4 |
| T-cell dose | |||
| 1 × 107 | 2.7 | .17 | 0.94-8.1 |
| 3-5 × 107 | 1.4 | 0.41-4.7 | |
| >5 × 107 | — | — | |
| Time from BMT to DLI (d) | |||
| 0-492 | 1.2 | .70 | 0.48-3.0 |
| >492 | — | — | |
| Reason for DLI | |||
| CML in CP | 3.0 | .05 | 1.00-9.0 |
| Relapsed disease | — | — | |
| Donor chimerism before DLI | |||
| 0%-50% | — | — | — |
| 51%-100% | 4.5 | <.01 | 1.7-12.5 |
| Donor chimerism after DLI | |||
| 0%-50% | — | — | — |
| 51%-100% | 16.5 | <.01 | 5.3-51.2 |
Cox proportional hazards models were used to look at factors associated with the development of GVHD (Table 6). Only the percentage of donor chimerism before DLI had a significant effect on the development of GVHD. Patients with >50% donor chimerism before DLI were 3.4 times (hazard ratio, 3.4; 95% confidence interval, 1.5-7.6; P < .01) more likely to develop GVHD at any point in time than patients whose donor chimerism was <50% at the time of DLI (Table 6).
Table 6. Univariate Cox Proportional Hazards Models Were Fit Correlating These Variables with the Onset and Development of GVHD, Taking Only Initial DLI into Consideration (n = 83)
| Variable | Hazard Ratio | P Value | 95% CI |
|---|---|---|---|
| Age (y) | |||
| <35 | 0.42 | .22 | 0.14-1.3 |
| 36-50 | 0.63 | 0.29-1.4 | |
| >50 | — | — | |
| Sex | |||
| Male to male | 0.56 | .50 | 0.23-1.4 |
| Male to female | 0.50 | 0.18-1.4 | |
| Female to male | 0.58 | 0.18-1.9 | |
| Female to female | — | — | |
| T-cell dose | |||
| 1 × 107 | 0.42 | .11 | 0.17-1.1 |
| 3-5 × 107 | 1.0 | 0.44-2.3 | |
| >5 × 107 | — | — | |
| Time from BMT to DLI (d) | |||
| 0-492 | 0.65 | .25 | 0.32-1.4 |
| >492 | — | ||
| Donor chimerism before DLI | |||
| 0%-50% | — | — | |
| 51%-100% | 3.4 | <.01 | 1.5-7.6 |
| Donor chimerism after DLI | |||
| 0%-50% | — | — | |
| 51%-100% | 1.9 | .09 | 0.90-3.8 |
| Reason for DLI | |||
| CML in CP | 0.64 | .25 | 0.24-1.7 |
| Relapsed disease | — | — | |
| Time from BMT to DLI | 1.00 | 0.99-1.0 |
Mechanisms Responsible for the Activity of DLI in CML
CML is unique in its sensitivity to DLI in that responses are frequently seen in the absence of clinically evident GVHD. The mechanisms responsible for the responsiveness of CML to DLI, however, remain unclear. Although most (17/19) of the CRs in diseases other than CML were associated with clinically apparent GVHD, 7 of the 13 CML patients (CP and cytogenetic) who achieved a CR to DLI exhibited no evidence of GVHD. Of the 17 patients who received DLI for CML (in CP or cytogenetic relapse), 10 were part of a trial that used T-cell depletion followed by granulocyte-macrophage colony-stimulating factor for 2 months after BMT in an attempt to prevent relapse [30]. Granulocyte-macrophage colony-stimulating factor has been shown to preferentially terminally differentiate CML progenitors and was thus added after transplantation in an effort to eradicate residual CML cells that survived T cell–depleted alloBMT [31, 32, 33]. Of the 23 patients who underwent transplantation on this trial, 8 exhibited stable mixed chimerism (>10% host hematopoiesis; median, 18%; range, 11%-45%) without evidence of relapse (absence of Philadelphia chromosome by fluorescent in situ hybridization and cytogenetics; unpublished data). Although 2 of these 8 patients remain in remission with stable mixed chimerism at 2 and 3 years after BMT, 6 of 8 have subsequently relapsed and gone on to receive DLI. All 6 cases achieved a complete cytogenetic and molecular remission with DLI. All 6 also converted to persistent full donor chimerism, although only 3 of the 6 developed any evidence of GVHD. Because normal and CML host hematopoiesis were both eliminated, the DLI responses in CML were associated with effective graft-versus-host reactions, even when clinical GVHD was not apparent.
Discussion
Most disease responses after DLI occur in the setting of GVHD [6, 15, 34]. However, durable responses are often seen in CP CML with T-cell doses (1 × 107 cells per kilogram) that are associated with a low incidence of clinically apparent GVHD [7, 8, 35]. We found that residual normal host hematopoiesis was also eradicated in all CML patients who achieved CRs with DLI. Thus, the CML patients who responded to DLI exhibited potent lymphohematopoietic graft-versus-host reactions even though GVHD was not clinically apparent in many patients. The similar sensitivity of CML and normal hematopoiesis to DLI argues against a leukemia-specific immunologic reaction and probably explains CML’s high response to DLI. With malignant progression to AP/BC, CML shows the same limited responsiveness to DLI as other hematologic malignancies.
Patients with >50% donor chimerism before DLI were more likely to achieve a CR. Furthermore, those who achieved full donor chimerism after DLI were 22 times more likely to achieve a CR than patients who did not. The finding that the degree of donor chimerism correlates with disease response to DLI could be explained by donor chimerism serving as a marker of minimal residual disease; that is, patients with high levels of donor chimerism before DLI have the lowest disease burdens at the time of DLI and thus might be expected to benefit the most from DLI. However, the degree of donor chimerism was also highly predictive of the development of GVHD, and this finding should not be explained by donor chimerism being a marker of disease burden. Because the development of GVHD after DLI was the strongest predictor of disease response, the correlation between disease response to DLI and donor chimerism is likely a reflection of donor chimerism being a marker for graft-versus-host reactions rather than disease burden.
Although necessary for the immunologic antitumor activity, GVHD produced significant morbidity and mortality. GVHD was the cause of or contributed to death in 24% of the patients who died. Recurrent disease, however, was an even larger problem and led to 60% of the deaths in this series. Moreover, only 2 of 25 patients who received chemotherapy for relapsed disease before DLI are alive in continuous remission. Clearly, outside of CML, novel approaches that enhance the antitumor activity of DLI while minimizing GVHD are needed. One approach that has shown promise is the use of autologous tumor vaccines in combination with DLI in a murine model [36, 37]. Thus, it may be possible to incorporate tumor vaccines with DLI in an endeavor to educate donor T cells and enhance antitumor activity, perhaps by using lower doses of donor T cells that may reduce GVHD [38]. Even if this is effective, autologous tumor will not be available for all patients, and other approaches will be needed to improve the antitumor activity of adoptive immunotherapy.
Dose escalation of DLI was used in 19 patients who did not respond or who relapsed after an initial response to DLI and did not have active GVHD. Although dose escalation has been reported to induce remissions in patients who do not respond [7, 8], CML was the only disease in which we saw durable responses with dose escalation. Of the 6 CML CP patients who received dose-escalated DLI, 3 achieved a durable CR. As other groups have reported [6, 13, 15], we found no correlation between the dose of T cells administered and the risk of developing GVHD.
DLIs clearly have the potential to induce remissions in patients who relapse after allogeneic transplantation. Their efficacy seems greatest in patients with CML in CP or cytogenetic relapse, those with >50% donor chimerism at the time of DLI, and in those who develop GVHD and full donor chimerism after DLI. Unfortunately, except in CML, most patients continue to die of their underlying disease because of insufficient antitumor activity even in the face of active GVHD.
Acknowledgments
Supported by National Institutes of Health grant nos. P01 CA15396 (RJJ) and K23 CA09657 (CAH).
References
- Antileukemic effect of graft-versus-host disease in human recipients of allogeneic-marrow grafts . N Engl J Med . 1979;300:1068–1073
- . Antileukemic effect of chronic graft-versus-host disease . N Engl J Med . 1981;304:1529–1533
- Graft-versus-leukemia reactions after bone marrow transplantation . Blood . 1990;75:555–562
- Donor leukocyte transfusions for treatment of recurrent chronic myelogenous leukemia in marrow transplant patients . Blood . 1990;76:2462–2465
- . Allogeneic cell therapy for relapsed leukemia after bone marrow transplantation with donor peripheral blood lymphocytes . Exp Hematol . 1995;23:1553–1562
- Donor leukocyte infusions in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation . J Clin Oncol . 1997;15:433–444
- Comparison of single-dose and escalating-dose regimens of donor lymphocyte infusion for relapse after allografting for chronic myeloid leukemia . Blood . 2000;95:67–71
- . Adoptive immunotherapy using donor leukocytes following bone marrow transplantation for chronic myeloid leukemia (is T cell dose important in determining biological response?) . Bone Marrow Transplant . 1995;15:591–594
- . Hematologic relapse of chronic myelogenous leukemia following allogeneic bone marrow transplantation (apparent graft-versus-leukemia effect following abrupt discontinuation of immunosuppression) . Bone Marrow Transplant . 1992;10:391–395
- . Induction of graft-versus-host disease as immunotherapy of leukemia relapsing after allogeneic transplantation (single-center experience of 32 adult patients) . Bone Marrow Transplant . 1997;20:129–135
- . Induction of a graft-versus-leukemia reaction by cyclosporin A withdrawal as immunotherapy for leukemia relapsing after allogeneic bone marrow transplantation . Bone Marrow Transplant . 1999;23:771–777
- Graft-versus-leukemia effect of donor lymphocyte transfusions in marrow grafted patients. European Group for Blood and Marrow Transplantation Working Party Chronic Leukemia . Blood . 1995;86:2041–2050
- Treatment of relapsed leukemia after unrelated donor marrow transplantation with unrelated donor leukocyte infusions . Blood . 2000;95:1214–1221
- Donor leukocyte infusions in acute lymphocytic leukemia . Bone Marrow Transplant . 2000;26:511–516
- . Donor leukocyte transfusions for treatment of leukemic relapse after bone marrow transplantation. EBMT Immunology and Chronic Leukemia Working Parties . Vox Sang . 1998;74(suppl 2):321–329
- Long-term follow-up of patients who achieved complete remission after donor leukocyte infusions . Biol Blood Marrow Transplant . 1999;5:253–261
- . Donor lymphocyte infusions to treat hematologic malignancies in relapse after allogeneic blood or marrow transplantation . Cancer Control . 2002;9:123–137
- T-cell-depleted allogeneic bone marrow transplantation followed by donor lymphocyte infusion in patients with multiple myeloma (induction of graft-versus-myeloma effect) . Blood . 2001;98:934–939
- Donor leukocyte infusions for multiple myeloma . Bone Marrow Transplant . 2000;26:1179–1184
- High response rate in refractory and poor-risk multiple myeloma after allotransplantation using a nonmyeloablative conditioning regimen and donor lymphocyte infusions . Blood . 2001;97:2574–2579
- . Donor leukocyte infusions are effective in relapsed multiple myeloma after allogeneic bone marrow transplantation . Blood . 1997;90:4206–4211
- . Adoptive immunotherapy with donor mononuclear cell infusions to treat relapse of acute leukemia or myelodysplasia after allogeneic bone marrow transplantation . Bone Marrow Transplant . 1996;18:975–980
- Sequential monitoring of chimerism and detection of minimal residual disease after allogeneic blood stem cell transplantation (BSCT) using multiplex PCR amplification of short tandem repeat-markers . Leukemia . 2001;15:293–302
- . Strategies and clinical implications of chimerism diagnostics after allogeneic hematopoietic stem cell transplantation . Acta Haematol . 2004;112:16–23
- . Bone marrow engraftment analysis after allogeneic bone marrow transplantation . Clin Lab Med . 2000;20:197–225
- The use of amplified variable number of tandem repeats (VNTR) in the detection of chimerism following bone marrow transplantation. A comparison with restriction fragment length polymorphism (RFLP) by Southern blotting . Am J Clin Pathol . 1997;107:292–298
- Interphase FISH for Y chromosome, VNTR polymorphisms, and RT-PCR for BCR-ABL in the monitoring of HLA-matched and mismatched transplants . Cancer Genet Cytogenet . 2000;120:25–29
- 1994 Consensus Conference on Acute GVHD Grading . Bone Marrow Transplant . 1995;15:825–828
- . Consensus among bone marrow transplanters for diagnosis, grading and treatment of chronic graft-versus-host disease. Committee of the International Bone Marrow Transplant Registry . Bone Marrow Transplant . 1989;4:247–254
- GM-CSF-mediated differentiation therapy following bone marrow transplantation (BMT) for chronic myeloid leukemia (CML) . [abstract] Blood . 1997;90:235a
- Growth factor-mediated terminal differentiation of chronic myeloid leukemia . Cancer Res . 1994;54:5535–5538
- . Selective expansion of primitive normal hematopoietic cells in cytokine-supplemented cultures of purified cells from patients with chronic myeloid leukemia . Blood . 1997;90:64–69
- . Chronic myelogenous leukemia primitive hematopoietic progenitors demonstrate increased sensitivity to growth factor-induced proliferation and maturation . Exp Hematol . 2000;28:1401–1412
- The occurrence of graft-versus-host disease is the major predictive factor for response to donor lymphocyte infusions in multiple myeloma . Blood . 2004;103:4362–4364
- Durability of responses following donor lymphocyte infusions for patients who relapse after allogeneic stem cell transplantation for chronic myeloid leukemia . Blood . 2000;96:2712–2716
- Successful therapy of metastatic cancer using tumor vaccines in mixed allogeneic bone marrow chimeras . Blood . 2003;101:1645–1652
- In vivo priming of donor lymphocyte infusions by tumor vaccines after allogeneic bone marrow transplantation induces potent tumor immunity without lethal GVHD . Blood . 2000;96:581a
- Comparative outcomes of T-cell-depleted and non-T-cell-depleted allogeneic bone marrow transplantation for chronic myelogenous leukemia (impact of donor lymphocyte infusion) . J Clin Oncol . 1999;17:561–568
PII: S1083-8791(05)00813-X
doi:10.1016/j.bbmt.2005.11.520
© 2006 American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.
Volume 12, Issue 4 , Pages 414-421, April 2006
