Treatment of relapsed acute myeloid leukemia after ... - Nature

Leukemia (2004) 18, 1789–1797 & 2004 Nature Publishing Group All rights reserved 0887-6924/04 $30.00 www.nature.com/leu

Treatment of relapsed acute myeloid leukemia after allogeneic bone marrow transplantation with chemotherapy followed by G-CSF-primed donor leukocyte infusion: a high incidence of isolated extramedullary relapse S-J Choi1, J-H Lee1, J-H Lee1, S Kim1, M Seol1, Y-S Lee1, J-S Lee1, W-K Kim1, H-S Chi2 and K-H Lee1 1

Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea; and Department of Laboratory Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea

2

For patients with acute myeloid leukemia (AML) relapsed after allogeneic bone marrow transplantation (BMT), donor leukocyte infusion (DLI) as sole therapy has very limited efficacy. We tested the effects of cytoreductive chemotherapy, followed immediately by G-CSF-primed DLI (chemotherapy followed by DLI, Chemo-DLI), in 16 AML patients who relapsed after allogeneic BMT. In all, 10 of these patients achieved complete remission (CR), four of whom remain alive in CR at a median follow-up of 1488 days after DLI. The 2-year overall survival (OS) for the entire cohort was 31%. The 1-year OS for patients with post-BMT remission of 6 months or longer was 55%, compared with 0% for patients with post-BMT remission of less than 6 months, making post-BMT remission duration the only significant prognostic factor for OS (P ¼ 0.015). These findings suggest that Chemo-DLI could induce durable remissions in a proportion of relapsed AML patients with relatively long postBMT remission duration. All five patients who relapsed after achieving CR with Chemo-DLI relapsed at extramedullary sites in the presence of continuous bone marrow remission, suggesting uneven graft-versus-leukemia effects in different parts of the body. Although our data should be interpreted cautiously considering the limited number of patients, isolated extramedullary relapse seems to be common after Chemo-DLI. Leukemia (2004) 18, 1789–1797. doi:10.1038/sj.leu.2403523 Published online 23 September 2004 Keywords: acute myeloid leukemia; relapse; allogeneic BMT; donor leukocyte infusion; extramedullary relapse

Introduction For patients with acute myeloid leukemia (AML) receiving allogeneic bone marrow transplantation (BMT), relapse remains the major cause of treatment failure and is associated with a very poor prognosis.1–3 The optimal therapy for these relapsed patients has not been defined, since most therapies are of limited utility. While reinduction chemotherapy can induce remission in about 30–40% of patients, this remission is usually of short duration, and most of these patients eventually relapse and die.2,3 A second BMT results in long-term survival of about 10–35% of patients, but it is associated with extremely high treatment-related mortality (TRM), ranging from 40 to 50%.4–8 Another approach to the treatment of AML patients who relapse after allogeneic BMT is donor leukocyte infusion (DLI), which induces a graft-versus-leukemia (GVL) effect. DLI has been shown to be highly effective in patients with chronic myelogenous leukemia who relapse into chronic phase, achieving complete and durable remission in 70–80% of these patients.9,10 In contrast, DLI as sole therapy for relapsed AML has been much less effective, yielding complete remission (CR) Correspondence: Professor S-J Choi, Department of Internal Medicine, Asan Medical Center, 388-1 Poongnap-2-dong, Songpa-ku, Seoul 138736, Korea; Fax: þ 82 2 3010 6961; E-mail: [email protected] Received 13 May 2004; accepted 18 August 2004; Published online 23 September 2004

rates of 15–29% and rare long-term remission.9–12 Limited efficacy of DLI alone for relapsed AML may be due in part to the diverse immune escape mechanisms of leukemic cells from GVL. These mechanisms may include the downregulation of HLA molecules,13,14 defects in presentation of antigenic peptides,15 deficient expression of costimulatory molecules,16,17 secretion by leukemic cells of cytokines that inhibit activated lymphocytes,18 and FAS ligand expression by leukemic cells.19 Another factor limiting the response of DLI alone in AML may be too rapid proliferation rate of leukemic cells coupled with the high leukemic cell burden at the time of relapse for clinically evident GVL effect to develop. Since a GVL effect of DLI typically requires several weeks or months to become clinically apparent,10,20,21 in patients with rapidly advancing diseases such as AML, it would be reasonable to attempt cytoreductive chemotherapy before DLI in order to allow time for the development of the GVL effect. Several retrospective studies have reported that durable remission could be achieved in some patients with relapsed AML with chemotherapy followed by DLI implemented at the time of nadir blood counts after chemotherapy or as consolidation of a chemotherapy-induced remission.9,10,22,23 These findings, however, are limited by the retrospective nature of the studies, heterogeneous treatment schemes, and the small numbers of patients included. Likewise, there have been few prospective studies evaluating the effectiveness of Chemo followed by DLI (Chemo-DLI) in relapsed AML.24–26 Thus, the role of Chemo-DLI in the treatment of relapsed AML after allogeneic BMT is still largely unknown. We previously reported preliminary results of a prospective study evaluating the efficacy of Chemo-DLI for patients with relapsed acute leukemia, including eight patients with AML, after allogeneic BMT.25 Our previous study showed that a high CR rate and some durable remissions could be achieved with this approach in patients with relapsed acute leukemia following allogeneic BMT. The present study updates our previous findings and reports our experience with Chemo-DLI in a larger cohort of patients with relapsed AML following allogeneic BMT. Our objectives were to determine the overall outcome in these patients, to assess the major prognostic factors that affect the outcome, and to determine the circumstances in which it may be appropriate to offer this treatment option.

Patients and methods

Patient eligibility Patients were eligible for treatment if they had undergone allogeneic BMT for AML and subsequently relapsed. To be included, patients were required to have a sufficient cardiac function (ejection fraction X40%), a creatinine level of less than 2.0 mg/dl, a bilirubin level of less than 2.0 mg/dl, liver

Chemo-DLI for relapsed AML after allogeneic BMT S-J Choi et al

1790 transaminases level lower than three times the upper limit of the normal range, and Karnofsky performance status of equal to or over 70. The study protocol was approved by the Institutional Review Board of the Asan Medical Center. Written informed consent was obtained from each patient and donor. Patients were enrolled in this prospective study between August 1997 and February 2003. During the study period, 21 patients with AML relapsed following allogeneic BMT. Of these patients, five patients were excluded due to following reasons: two due to poor performance status, two due to the original unrelated marrow donor’s refusal to donate G-CSF-primed leukocyte, and one due to the patient’s refusal to participate in the study. Thus, remaining 16 patients were enrolled in the study.

Initial BMT procedure Detailed procedures for initial transplantation have been described previously.27 At the time of initial BMT, 13 patients were in CR, two had untreated first relapse, and one had refractory disease. Three patients who were not in CR at the time of BMT achieved CR after BMT. All patients received Bu-Cy (busulfan 4 mg/kg/day on days 7 to 4 and cyclophosphamide 60 mg/kg/day on days 3 to 2) as a preparative regimen. A total of 15 patients received bone marrow grafts from HLAidentical siblings, whereas one patient (unique patient number, UPN 165) received a bone marrow graft from an HLA-matched unrelated donor. None of the bone marrow grafts was T-cell depleted graft-versus-host-disease (GVHD) prophylaxis consisted of cyclosporine (1.5 mg/kg intravenously every 12 h starting on day –1, then switched to an oral dose when oral intake became feasible) plus a short course of methotrexate (15 mg/m2 intravenously on day 1, and 10 mg/m2 intravenously on days 3, 6, and 11). In the absence of GVHD, cyclosporine was tapered from day 60 until day 180. After initial BMT, one patient (UPN 67) developed acute GVHD (aGVHD) of grade I at post-BMT day 26, and three patients developed chronic GVHD (cGVHD), one (UPN 61) with limited disease at day 111 and two (UPN 23 and 238) with extensive disease at days 83 and 157, respectively. In three of these four patients with aGVHD or cGVHD, GVHD resolved thereafter and all immunosuppressive agents were off before relapse (UPN 23 at post-BMT day 195; UPN 61 at day 264; UPN 67 at day 154). In one patient (UPN 238) whose cGVHD resolved with the use of steroid plus cyclosporine and four other patients (UPN 62, 152, 158, and 224) who relapsed before post-BMT day 180, cyclosporine was being tapered at the time of relapse.

Patients given immunosuppressive agents for the prophylaxis or treatment of GVHD at the time of the relapse after initial BMT were discontinued from these immunosuppressive agents before induction chemotherapy. No post-DLI GVHD prophylaxis was administered. Acute and cGVHD were graded according to established criteria.28,29 Grade II–IV aGVHD or extensive cGVHD occurring after DLI were treated with corticosteroid plus other immunosuppressive agent(s) such as cyclosporine, tacrolimus, or mycophenolate. Limited cGVHD was treated with cyclosprorine and/or corticosteroid when symptoms or signs of GVHD were progressive during observation without treatment. Bone marrow examination of patients showing full hematologic recovery in peripheral blood was performed to evaluate their remission status after Chemo-DLI. Patients presenting with extramedullary leukemia at relapse after initial BMT were evaluated by physical examination and appropriate radiologic methods. CR was defined by cellular marrow with less than 5% blasts, no evidence of extramedullary leukemia, and peripheral neutrophil and platelet counts of at least 1500/ml and 100 000/ml, respectively.

Evaluation of hematopoietic chimerism Since December 1997, post-BMT and post-DLI hematopoietic chimerism have been prospectively evaluated on peripheral blood mononuclear cells of the patients using polymerase chain reaction for short tandem repeats or amelogenin loci, as described previously.30,31

Statistical analysis Overall survival (OS) was measured from the date of DLI to the date of death or last follow-up. Disease-free survival (DFS) for patients achieving CR was measured from the date of CR documentation to the date of relapse or last follow-up. OS, DFS, and cumulative incidence of GVHD after DLI were plotted using the Kaplan–Meier method, and differences were assessed by the log-rank test. The variables analyzed for their impact on CR, OS, DFS, and post-DLI GVHD included disease status at the time of initial BMT, age at DLI, presence of GVHD after initial BMT, time from initial BMT to relapse (post-BMT remission duration), presence of extramedullary leukemic relapse after initial BMT, and CD34 and CD3 cell dose for DLI. For the analysis of OS and DFS, presence of post-DLI GVHD was also considered as a potential prognostic variable. All statistical tests were two-sided, with Po0.05 indicating statistical significance.

Treatment of relapse after allogeneic BMT Results The first day of DLI was designated day 0. The induction chemotherapy regimen, administered from days 7 to 2, consisted of cytarabine (1 g/m2/day continuous intravenous infusion on days 7 to 2), idarubicin (12 mg/m2/day intravenous infusion on days 7 to 5), and etoposide (150 mg/m2/day intravenous infusion over 5 h on days 7 to 5). Each original bone marrow donor was administered G-CSF (10 mg/kg/day subcutaneously from days 3 to 0), and their peripheral blood mononuclear cells, including lymphocytes and CD34 þ cells, were collected on days 0 and 1 of DLI and infused into the recipient on the same days to induce a GVL effect and to minimize marrow aplasia after chemotherapy. Patients subsequently received G-CSF 450 mg/day intravenously from day 5 until their absolute neutrophil counts were greater than 3000/ml. Leukemia

Patient characteristics The clinical characteristics of the patients are shown in Table 1. Median age of the patients at the time of DLI was 30 years (range, 20–46 years). The median CD3 þ and CD34 þ cell dose infused for DLI were 4.5 108/kg (range, 1.6–8.3) and 3.7 106/kg (range, 0.4–80.5), respectively. All 16 patients finished a course of induction chemotherapy and received G-CSF-primed DLI at a median of 13 days (range, 9–157 days) after documentation of relapse. One patient (UPN 67), who relapsed initially at an extramedullary site (left ulnar) without bone marrow involvement, achieved CR after localized radiation therapy to the extramedullary relapse site. After 5 months of


Yes (10.1) 7.3

UPN, unique patient number; BMT, bone marrow transplantation; DLI, donor leukocyte infusion; Chemo-DLI, chemotherapy followed by DLI; aGVHD, acute GVHD; cGVHD, chronic GVHD; CR, complete remission; CR1, first CR; CR2, second CR; NE, not evaluable.

Yes (6.1) Yes (3.6) No Yes (3.4) No No 5.1 No medication 17.5 21.2 1.9

1.8 0.4 1.8

No Yes (22.4) No 10.8 17.5 27.9

Ext Ext Ext NE Lim Lim Ext Ext NE No NE Lim NE NE NE NE II III II NE II I IV III III No NE III NE IV III IV 12.9 34.5 6.0 17.0 23.7 5.0 16.9 6.5 3.7 8.6 2.4 8.6 4.9 3.4 12.3 7.4 36 17 36 20 24 35 19 29 43 26 46 23 42 41 20 28 20 23 43 59 61 62 67 83 152 157 158 159 165 224 228 238

CR1 CR1 CR1 CR1 CR1 CR1 CR2 CR1 CR1 CR1 Refractory Relapse Relapse CR1 CR1 CR1

CR CR CR Non-CR CR CR CR CR CR CR Non-CR CR Non-CR Non-CR Non-CR Non-CR

cGVHD aGVHD

GVHD after DLI Response after Chemo-DLI

Post-BMT remission duration (months) Disease status at BMT Age (years) UPN

Table 1

Baseline characteristics and clinical course of patients after chemotherapy and subsequent DLI

Last day of taking immunosuppressive agent for post-DLI GVHD (post-DLI, months)

Relapse after Chemo-DLI (postDLI, months)

Alive Died Alive Died Died Died Alive Died Died Alive Died Died Died Died Died Died

Current status (post-DLI, months)

and NED (58.2) with leukemia (28.1) and NED (67.8) with leukemia (0.1) with leukemia (10.9) with leukemia (7.2) and NED (41.0) with leukemia (21.2) of GVHD (1.9) and NED (25.7) of pneumonia (0.3) with leukemia (11.1) with leukemia (0.4) with leukemia (1.8) of GVHD (0.4) of GVHD (1.8)

1791 initial relapse, this patient experienced relapses in bone marrow and multiple skin sites, and received Chemo-DLI 157 days after the time of initial isolated extramedullary relapse.

Response and overall clinical course Data regarding the clinical course after Chemo-DLI are summarized in Table 1. In all, 10 patients (63%; 95% confidence interval (CI), 38–87%) achieved CR at a median of 30 days (range, 18–78 days) after DLI. As a result of low statistical power, none of the variables was found to be predictive of achieving CR with Chemo-DLI. Of the 10 patients who achieved CR, four remain in CR, at a median follow-up time of 1488 days (range, 771–2034 days) after DLI. Five of the 10 CR patients relapsed at a median of 182 days (range, 101– 671 days) after DLI and died at a median of 333 days (range, 215–843 days). The remaining one CR patient died of aGVHD 58 days after DLI. The six patients who failed to achieve CR died at a median of 11 days (range, 3–55 days) after DLI, three with persistent leukemia, and three due to treatment-related toxicity.

OS and DFS For all 16 patients, six and five patients were alive at post-DLI 1 and 2 year, respectively, making 1- and 2-year OS of 38% (95% CI, 14–62%) and 31% (95% CI, 8–54%), respectively (Figure 1a). The median survival duration of all patients was 215 days. For 10 patients who achieved CR, five and four patients were alive in CR at post-DLI 1 and 2 year, respectively, making 1- and 2-year DFS of 56% (95% CI, 23–89%) and 44% (95% CI, 11–77%), respectively (Figure 1b). Univariate analysis of risk factors influencing OS and DFS is presented in Table 2. Although small number of patients prevented proper analysis, post-BMT remission duration of less than 6 months was found to be predictive of lower OS. For 11 patients with post-BMT remission duration of 6 months or longer, eight patients achieved CR, and among these eight patients with CR, four patients died after relapse and another four patients remained alive without relapse. For five patients with post-BMT remission duration of less than 6 months, two patients achieved CR, and among these two patients with CR, one died in CR due to GVHD and the other died after relapse. The 1-year OS for the patients with post-BMT remission duration of 6 months or longer was 55% (95% CI, 26–84%), whereas the 1-year OS for the patients with post-BMT remission duration of less than 6 months was 0% (P ¼ 0.015) (Figure 2). In prognostic factor analysis for DFS, none of the clinical factors was identified to be predictive for DFS due to low statistical power, although patients with shorter post-BMT remission duration showed trend for lower DFS (P ¼ 0.094).

Treatment-related toxicities and death All patients experienced severe myelosuppression, with a lowest neutrophil count less than 500/ml and a lowest platelet count less than 20 000/ml. Omitting the four early deaths, the remaining 12 patients achieved neutrophil counts over 500/ml at a median of 12 days (range, 9–18 days) after DLI. A total of 10 patients achieved a platelet count over 20 000/ml without transfusion support at a median of 16 days (range, 11–27 days) after DLI. All 16 patients experienced febrile neutropenia necessitating treatment with antimicrobials. Within 2 months Leukemia


1792 treatment-related toxicities. As bone marrow status after ChemoDLI was not evaluated in these three patients, due to early death (UPN 158 and 228) or not achieving full hematologic recovery (UPN 238), we were unable to determine whether the major cause of death was treatment-related toxicity or resistant leukemia. However, there was no clinical evidence of leukemia in their peripheral blood or extramedullary site. The proximate causes of death in these three patients were GVHD in two (UPN 228 and 238) and pneumonia in one (UPN 158). On this assumption, there were four treatment-related deaths, giving a TRM of 25%.

Pattern of relapse after Chemo-DLI

Figure 1 OS and DFS. Kaplan–Meier product-limit estimate of OS among all 16 patients (a) and DFS among 10 patients who achieved CR with chemotherapy and subsequent DLI (b).

All five patients who relapsed after achieving CR with ChemoDLI had experienced post-DLI GVHD previously before relapse. In three patients (UPN 23, 61, and 159), GVHD was resolved with the use of immunosuppressive agents, and all immunosuppressive agents were discontinued 4.9, 1, and 2.8 months before relapse, respectively. In two patients (UPN 62, and 83), relapse occurred in the presence of cGVHD (Table 1). The five patients who relapsed all did so at extramedullary sites in the presence of continuous marrow remission (Table 3). Three of these five patients also had extramedullary leukemia at the time of relapse after initial BMT (bone in UPN 23; skin in UPN 62; breasts, skin, and soft tissue in UPN 159). The most frequently involved relapse sites were leptomeninges (three patients; UPN 23, 61, and 83) and skin (two patients; UPN 62 and 159). Evaluation of bone marrow status at the time of relapse in all these five patients showed continuous remission in bone marrow. Except for UPN 83, the other four patients subsequently relapsed at other extramedullary sites (vulva in UPN 23; skin and common bile duct in UPN 61; leptomeninges in UPN 62; and peritoneum in UPN 159). Of these five patients, only one (UPN 23) subsequently showed leukemic cells in peripheral blood, whereas the other four had no leukemic cells in the peripheral blood until death.

Hematopoietic chimerism after Chemo-DLI after DLI, five patients experienced pneumonia and five experienced bacteriologically documented septicemia. Of 13 assessable patients, 12 developed aGVHD at a median of 23 days after DLI (range, 6–41 days), one with grade I, three with grade II, five with grade III, and three with grade IV. Eight of nine assessable patients developed cGVHD at a median of 106 days after DLI (range, 96–139 days), two with limited and six with extensive disease. None of the clinical factors was found to be predictive for the occurrence of grade III–IV aGVHD or extensive cGVHD after DLI. Detailed data regarding post-DLI GVHD and its treatment are shown in Table 1. Except for three patients (UPN 152, 228, and 238) who died of aGVHD and two patients (UPN 83 and 224) who died of leukemia in the presence of GVHD, GVHD of all the other patients resolved completely and all immunosuppressive agents were discontinued in these patients. Seven patients died early within 2 moths after DLI, six without achieving CR, and one (UPN 152) who achieved CR as a result of aGVHD. Of the six patients not achieving CR, three (UPN 59, 165, and 224) died with persistent leukemia in their peripheral blood or extramedullary site (intestine in UPN 59). Three other patients (UPN 158, 228, and 238) died, at least in part, from Leukemia

Data regarding hematopoietic chimerism are summarized in Table 4. Hematopoietic chimerism status at the time of relapse after initial BMT was evaluated in 13 patients. Except for two patients (UPN 62 and 159) who relapsed as isolated extramedullary relapse, chimerism study at the time of relapse in all the other patients showed mixed chimerism with recipient cell fraction of 10–100%. Except for four patients who died early within 1 month of DLI, post-DLI hematopoietic chimerism was evaluated in remaining nine patients. Except for UPN 224 who still showed leukemic blasts in peripheral blood at post-DLI 1 month, the other eight patients showed complete donor chimerism at post-DLI 1 month. Hematopoietic chimerism status at the time of relapse after Chemo-DLI was evaluated in four patients, three of whom (UPN 61, 62, and 159) showed complete donor chimerism, whereas one (UPN 83) showed mixed chimerism with recipient cell faction of 35%. Mixed chimerism in UPN 83 spontaneously converted to complete donor chimerism 1 month later and complete donor chimerism was maintained thereafter until death. UPN 62 showed mixed chimerism just before the time of death, but leukemic cells were not observed in peripheral blood smear specimens. UPN 61 and 159 showed complete donor chimerism throughout the entire follow-up period after DLI.


1793 Table 2 Variable

Impact of the clinical factors on OS and DFS N

OS at 1 year after DLI (%)

P-value

N

DFS at 1 year after DLI (%)

P-value

0.92

5 5

60 50

0.88

0.816

8 2

57 50

0.765

0.605

3 7

67 50

0.904

0.015

2 8

0 63

0.094

0.811

4 6

50 60

0.513

38 38

0.978

6 4

40 75

0.476

25 50

0.205

5 5

25 80

0.124

60 38

0.133

5 5

60 50

0.44

33 83

0.348

3 6

33 67

0.751

Pre-DLI variables Age group (years) o30 8 38 X30 8 38 Disease status at initial BMT CR1 12 42 Beyond CR1 4 25 aGVHD or cGVHD after initial BMT Yes 4 50 No 12 33 Time from initial BMT to relapse o6 months 5 0 X6 months 11 55 Extramedullary leukemia at relapse after initial BMT Yes 6 33 No 10 40 Post-DLI variables CD 34 dose at DLI ( 106/kg) o3.7 8 X3.7 8 8 CD3 dose at DLI ( 10 /kg) o4.5 8 X4.5 8 aGVHD after DLI Grade 0–II 5 Grade III–IV 8 cGVHD after DLI None or limited 3 Extensive 6

OS, overall survival; DFS, disease-free survival; DLI, donor leukocyte infusion; BMT, bone marrow transplantation; CR1, first complete remission; GVHD, graft-versus-host disease.

Figure 2 OS according to post-BMT remission duration. The 1year OS for patients with post-BMT remission of 6 months or longer was 55%, compared with 0% for patients with post-BMT remission of less than 6 months, making post-BMT remission duration the only significant prognostic factor for OS (P ¼ 0.015).

Discussion We have shown here that Chemo-DLI could produce a relatively high CR rate and induce durable remissions in a considerable proportion of patients with relapsed AML after allogeneic BMT,

especially in patients with relatively long post-BMT remission duration. We also have observed that isolated extramedullary relapse were common after this treatment approach. Recently, Levine et al24 reported the results of a prospective study evaluating the efficacy of anthracycline-based chemotherapy followed by DLI in patients with advanced myeloid malignancy relapsed after allogeneic BMT. The treatment approach of this study was similar to that of our study, including the use of G-CSF-primed DLI, the absence of post-DLI GVHD prophylaxis, and early application of DLI after chemotherapy (7 and 10–14 days after chemotherapy in our trial and trial by Levine et al, respectively). Although the CR and 2-year OS rates in our study (63 and 31%, respectively) were slightly superior to those observed in the trial by Levine et al (42 and 19%, respectively), overall outcomes seem to be similar, inasmuch as the latter trial included more patients with short post-BMT remission duration (ie patients with poor prognosis). TRM was similar between the two trials (25 and 23%), but the incidence of overall and grades III–IV aGVHD was much higher in our trial (93 and 62%, respectively) than in the other (56 and 28%, respectively). Comparing the results from Chemo-DLI with historical data of DLI alone,9,10 CR rate was higher with Chemo-DLI than with DLI alone (42–63 vs 15–29%). However, it is not clear whether this higher CR rate after Chemo-DLI translates into improved survival. Long-term survival after DLI alone for relapsed AML has been reported to be approximately 10–15%.9,10 Although the long-term survival rates in our study and study by Levine et al (31 and 19%, respectively) were slightly superior to those after DLI alone, this small difference in long-term survival is difficult to interpret due to the inherent limitations of historical Leukemia


1794 Table 3 UPN

23 61 62 83 159

Pattern of relapse after DLI Extramedullary leukemia at the time of relapse after initial BMT

Initial relapse site after Chemo-DLI (post-DLI, months)

Bone

Leptomeninges, bone (22.4) Leptomeninges (6.1) Skin (3.6) Leptomeninges (3.4) Skin (10.1)

None Skin None Breast, skin, soft tissue

BM status at relapse after Chemo-DLI

Subsequent relapse site after Chemo-DLI (post-DLI, months)

Current status (postDLI, months)

NED

Vulva (25.4), PB (26.0) Skin (8.8), CBD (9.8) Leptomeninges (5.2) None Peritoneum (10.4)

Died (28.1)

NED NED NED NED

Died Died Died Died

(10.9) (7.2) (21.2) (11.1)

UPN, unique patient number; BMT, bone marrow transplantation; Chemo-DLI, chemotherapy followed by DLI; BM, bone marrow; NED, no evidence of disease; PB, peripheral blood; CBD, common bile duct.

Table 4

Hematopoietic chimerism at relapse and after DLI

UPN

At relapse

59 61 62 67 83 152 157 158 159 165 224 228 238

100 45 CC 50a 45 50 100 10 CC 40 85 65 55

1 mo

2 mo

3 mo

CC CC CC CC CC CC

CC CC CC CC CC CC

CC CC CC 35 CC

CC

CC

CC

4 mo

5 mo

CC

CC

CC

CC

6 mo

12 mo

15 mo

18 mo

21 mo

24 mo

CC CC

CC CC

CC CC

CC CC

CC

CC

CC

CC

CC

CC

CC

CC

CC

CC

CC

CC CC CC CC

7 mo

9 mo CC

60 CC

25 CC

UPN, unique patient number; CC, complete donor chimerism; DLI, donor leukocyte infusion. Number in bold represents the percentage of the recipient cell population showing mixed chimerism. a Denotes the percentage of the recipient cell population at the time of bone marrow relapse.

comparison. Thus, it is uncertain whether administration of chemotherapy prior to DLI improves long-term survival following DLI for treating patients with relapsed AML following allogeneic BMT. In many aspects, the approach taken in our trial is similar to a second allogeneic BMT. The critical differences between our approach and second BMT are the use of nonmyeloablative chemotherapy, and absence of prophylactic immunosuppression for GVHD. Considering chemotherapy regimen before DLI is not required for immunosuppressive because patients undergoing DLI retain donor-derived immunity, we included only cytotoxic agents (such as cytarabine, idarubicin, and etoposide) in chemotherapy regimen before DLI. Comparing the result of Chemo-DLI in our trial with those from second allogeneic BMT,4–8 TRM seems to be lower with our approach (25 vs 40–50%), although long-term survival was similar (31 vs 10–35%). However, it should be noted that all patients in our study received Chemo-DLI at the status of untreated relapse, while most of the patients who received second allogeneic BMT were in remission induced by salvage chemotherapy before second BMT.4–8 Moreover, it has been reported that only 7–20% of highly selected patients can reach the stage of a second BMT after relapse according to the performance and remission status after salvage chemotherapy.3,6 Considering this fact, ChemoDLI seems to compare favorably with second allogeneic BMT of myeloablative conditioning. Although Chemo-DLI induced a durable remission in some patients with relapsed AML after allogeneic BMT, the limited overall efficacy of this procedure emphasizes the importance of identifying the specific subgroup of patients who might benefit Leukemia

from this treatment approach. Although small number of patients prevented proper analysis, our study found that postBMT remission duration was significantly associated with OS. In patients with post-BMT remission duration of 6 months or longer, 1-year OS was 55%, indicating that Chemo-DLI is worth trying in this group of patients. In contrast, patients with postBMT remission duration of less than 6 months are unlikely to benefit from Chemo-DLI and should be considered for other investigational therapies in the context of well-designed clinical trials. One possible reason for these observations would be that patients with shorter post-BMT remission duration might experience more severe treatment-related toxicities. In these patients, tissue damage after initial BMT may not be fully healed at the time of Chemo-DLI, thus increasing the severity of treatment-related toxicities. Another possible reason for the poor prognosis of patients with short post-BMT remission duration may be that these patients have leukemic cells intrinsically resistant to GVL mechanisms and/or chemotherapy. In our study, the major problem after Chemo-DLI was the high incidence of GVHD and GVHD-related deaths. We did not use GVHD prophylaxis to avoid the possibility of increasing the risk of relapse due to a reduction in the GVL effect, and the high incidence of GVHD in our study may be at least partly due to the absence of post-DLI GVHD prophylaxis. Another factor possibly associated with high incidence of GVHD may be cumulative tissue damage caused by chemotherapy for ChemoDLI and conditioning for initial BMT. Any damaged tissues would release large amounts of inflammatory cytokines, which may trigger and/or increase the severity of GVHD,32,33 and this may have been further exacerbated by our early application of


1795 DLI after chemotherapy. Animal studies have shown that a given cell dose is more likely to induce GVHD if infused sooner rather than later after a preparative regimen.34,35 The other factor that may be associated with the high incidence of GVHD is our use of a higher median dose of infused T-lymphocytes (4.5 108/kg) than that used in a similar trial (1.0 108/kg).24 These results suggest that outcome might be negatively affected by performing DLI too soon after chemotherapy, by using a high-dose-intensity chemotherapy regimen before DLI, and by infusing too many Tlymphocytes. It should be remembered, however, that a balance must be struck between the risk of GVHD and the risk of relapse, and that treatment schemes, including the GVHD prophylaxis regimen, the chemotherapy regimen before DLI, the timing of DLI, and the dose of infused T-lymphocytes, should be optimized to decrease GVHD while maintaining the GVL effect. Little has been reported regarding the relapse pattern after DLI or after second allogeneic BMT for relapsed AML following allogeneic BMT. In the present study, we have observed the unusually high incidence of isolated extramedullary relapse without bone marrow involvement after Chemo-DLI. Of the 10 patients with CR, five relapsed, all initially at extramedullary sites in the presence of continuous marrow remission, making isolated extramedullary relapse the major reason for failure of long-term survival in patients who achieved CR with ChemoDLI. Interestingly, all five relapsers experienced post-DLI GVHD before relapse, and in two of them, relapse occurred even in the presence of cGVHD. Of these five patients, one subsequently showed leukemic cells in peripheral blood, whereas the other four did not show any clinical evidence of leukemic cells in the peripheral blood until the time of death. Furthermore, most patients subsequently relapsed at other extramedullary sites without the concomitant presence of leukemic cells in peripheral blood. Thus, as has been suggested in a postBMT36,37 or post-DLI38,39 setting, this finding suggests that the potency of the GVL effect may be uneven in different parts of the body, being stronger in the bone marrow than in extramedullary tissues. One possible explanation for the weaker GVL effect at extramedullary sites might be tissue-selective lymphocyte homing based on the expression of certain cell surface molecules. Lymphocyte activation occurring in the marrow or spleen most likely led to the expression of lymphocyte surface molecules directing selective homing only to those tissues.40 Alternatively, isolated extramedullary relapse might reflect biological characteristics of certain leukemic cells that are intrinsically resistant to chemotherapy and/or GVL mechanisms and meanwhile tend to dwell in extramedullary tissues through diverse immune escape mechanisms such as the downregulation of HLA molecules,13,14 defects in presentation of antigenic peptides,15 deficient expression of costimulatory molecules,16,17 secretion of cytokines that inhibit activated lymphocytes,18 and FAS ligand expression.19 Mechanisms for the deficient GVL effect at extramedullary sites require further investigation to establish strategies to prevent relapse at these sites. Relapse limited to extramedullary sites following allogeneic BMT has been reported to constitute 13–30% of all AML relapses.2,37,41,42 Although it should be interpreted cautiously, considering the limited number of patients in our study, our finding, that all the initial relapses after DLI occurred only at extramedullary sites, suggests that these isolated extramedullary relapses might occur more frequently after DLI than after BMT. Although the reason this uneven GVL effect between marrow and extramedullary sites tends to be stronger in the post-DLI than in the post-BMT setting remains unknown, it is possible that the patients enrolled in our study may represent a selected group. Specifically, three of our five patients with isolated

extramedullary relapse after Chemo-DLI had also experienced extramedullary relapse after their initial BMT, suggesting that these patients may have constituted a considerable fraction of those patients in whom a proper GVL effect would not be operative at extramedullary sites. On the basis of relapse pattern observed in our study, long-term DFS after Chemo-DLI seems to depend on proper GVL effect, especially in the immunologically privileged extramedullary sites, rather than the dose intensity of chemotherapy prior to DLI. One potential way to enhance GVL effect of DLI might be the use of cytokines, such as interleukin-2 (IL-2). IL-2 might be a potentially active agent to enhance GVL effect of DLI because it induces the activation and proliferation of both T and NK cells. Although experience with combination therapy of IL-2 plus DLI for treatment of leukemia relapse after allogeneic BMT is limited, several studies reported its feasibility and potential benefit in a variety of hematologic malignancies including AML.43–45 Considering the limited efficacy of DLI or Chemo-DLI for treatment of relapsed AML following allogeneic BMT, the effectiveness of combination therapy of DLI or ChemoDLI plus IL-2 needs to be evaluated in well-designed prospective trials. In addition, it might also be important to maintain adequate cytotoxic drug level of the chemotherapeutic agents in the extramedullary sites such as central nervous system. The high frequency of leptomeningeal relapse observed in our study suggests that a high-dose cytarabine-containing regimen might be employed as cytoreductive chemotherapy before DLI, reducing the likelihood of relapse at this site by increasing the level of drug in the cerebrospinal fluid.46 Alternatively, intrathecal chemotherapy using methotrexate or cytarabine could be considered for the same purpose. The effectiveness of these therapeutic approaches for reducing leptomeningeal relapse after Chemo-DLI should be evaluated in future studies. In summary, we have shown here that Chemo-DLI could induce durable remissions in a considerable proportion of patients with relapsed AML after allogeneic BMT, especially in patients with relatively long post-BMT remission duration. Although limited by the small number of patients, isolated extramedullary relapse seems to be common after Chemo-DLI. The high frequency of isolated extramedullary relapse requires particular attention in future studies, as well as in designing future treatment programs.

References 1 Copelan EA, Biggs JC, Thompson JM, Crilley P, Szer J, Klein JP et al. Treatment for acute myelocytic leukemia with allogeneic bone marrow transplantation following preparation with BuCy2. Blood 1991; 78: 838–843. 2 Mortimer J, Blinder MA, Schulman S, Appelbaum FR, Buckner CD, Clift RA et al. Relapse of acute leukemia after marrow transplantation: natural history and results of subsequent therapy. J Clin Oncol 1989; 7: 50–57. 3 Frassoni F, Barrett AJ, Granena A, Ernst P, Garthon G, Kolb HJ et al. Relapse after allogeneic bone marrow transplantation for acute leukaemia: a survey by the EBMT of 117 cases. Br J Haematol 1988; 70: 317–320. 4 Mrsic M, Horowitz MM, Atkinson K, Biggs JC, Champlin RE, Ehninger G et al. Second HLA-identical sibling transplants for leukemia recurrence. Bone Marrow Transplant 1992; 9: 269–275. 5 Radich JP, Sanders JE, Buckner CD, Martin PJ, Petersen FB, Bensinger W et al. Second allogeneic marrow transplantation for patients with recurrent leukemia after initial transplant with total-body irradiation-containing regimens. J Clin Oncol 1993; 11: 304–313. 6 Mehta J, Powles R, Treleaven J, Horton C, Meller S, Pinkerton CR et al. Outcome of acute leukemia relapsing after bone marrow transplantation: utility of second transplants and adoptive immunotherapy. Bone Marrow Transplant 1997; 19: 709–719. Leukemia


1796 7 Michallet M, Tanguy ML, Socie G, Thiebaut A, Belhabri A, Milpied N et al. Second allogeneic haematopoietic stem cell transplantation in relapsed acute and chronic leukaemias for patients who underwent a first allogeneic bone marrow transplantation: a survey of the Societe Francaise de Greffe de moelle (SFGM). Br J Haematol 2000; 108: 400–407. 8 Bosi A, Laszlo D, Labopin M, Reffeirs J, Michallet M, Gluckman E et al. Acute Leukemia Working Party of the European Blood and Marrow Transplant Group. Second allogeneic bone marrow transplantation in acute leukemia: results of a survey by the European Cooperative Group for Blood and Marrow Transplantation. J Clin Oncol 2001; 19: 3675–3684. 9 Kolb HJ, Schattenberg A, Goldman JM, Hertenstein B, Jacobsen N, Arcese W et al. 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. 10 Collins Jr RH, Shpilberg O, Drobyski WR, Porter DL, Giralt S, Champlin R et al. Donor leukocyte infusions in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation. J Clin Oncol 1997; 15: 433–444. 11 Kolb HJ. 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. 12 Porter DL, Collins Jr RH, Shpilberg O, Drobyski WR, Connors JM, Sproles A et al. Long-term follow-up of patients who achieved complete remission after donor leukocyte infusions. Biol Blood Marrow Transplant 1999; 5: 253–261. 13 Dermime S, Mavroudis D, Jiang YZ, Hensel N, Molldrem J, Barrett AJ. Immune escape from a graft-versus-leukemia effect may play a role in the relapse of myeloid leukemias following allogeneic bone marrow transplantation. Bone Marrow Transplant 1997; 19: 989–999. 14 Brouwer RE, van der Heiden P, Schreuder GM, Mulder A, Datema G, Anholts JD et al. Loss or downregulation of HLA class I expression at the allelic level in acute leukemia is infrequent but functionally relevant, and can be restored by interferon. Hum Immunol 2002; 63: 200–210. 15 Ossendorp F, Eggers M, Neisig A, Ruppert T, Groettrup M, Sijts A et al. A single residue exchange within a viral CTL epitope alters proteasome-mediated degradation resulting in lack of antigen presentation. Immunity 1996; 5: 115–124. 16 Hirano N, Takahashi T, Takahashi T, Ohtake S, Hirashima K, Emi N et al. Expression of costimulatory molecules in human leukemias. Leukemia 1996; 10: 1168–1176. 17 Brouwer RE, Hoefnagel J, Borger van Der Burg B, Jedema I, Zwinderman KH, Starrenburg IC et al. Expression of co-stimulatory and adhesion molecules and chemokine or apoptosis receptors on acute myeloid leukaemia: high CD40 and CD11a expression correlates with poor prognosis. Br J Haematol 2001; 115: 298–308. 18 Bergmann L, Schui DK, Brieger J, Weidmann E, Mitrou PS, Hoelzer D. The inhibition of lymphokine-activated killer cells in acute myeloblastic leukemia is mediated by transforming growth factor-beta 1. Exp Hematol 1995; 23: 1574–1580. 19 Buzyn A, Petit F, Ostankovitch M, Figueiredo S, Varet B, Guillet JG et al. Membrane-bound Fas (Apo-1/CD95) ligand on leukemic cells: a mechanism of tumor immune escape in leukemia patients. Blood 1999; 94: 3135–3140. 20 Drobyski WR, Keever CA, Roth MS, Koethe S, Hanson G, McFadden P et al. Salvage immunotherapy using donor leukocyte infusions as treatment for relapsed chronic myelogenous leukemia after allogeneic bone marrow transplantation: efficacy and toxicity of a defined T-cell dose. Blood 1993; 82: 2310–2318. 21 Dazzi F, Szydlo RM, Craddock C, Cross NC, Kaeda J, Chase A et al. 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. 22 Verdonck LF, Petersen EJ, Lokhorst HM, Nieuwenhuis HK, Dekker AW, Tilanus MG et al. Donor leukocyte infusions for recurrent hematologic malignancies after allogeneic bone marrow transplantation: impact of infused and residual donor T cells. Bone Marrow Transplant 1998; 22: 1057–1063. Leukemia

23 Keil F, Prinz E, Kalhs P, Lechner K, Moser K, Schwarzinger I et al. Treatment of leukemic relapse after allogeneic stem cell transplantation with cytotoreductive chemotherapy and/or immunotherapy or second transplants. Leukemia 2001; 15: 355–361. 24 Levine JE, Braun T, Penza SL, Beatty P, Cornetta K, Martino R et al. Prospective trial of chemotherapy and donor leukocyte infusions for relapse of advanced myeloid malignancies after allogeneic stem-cell transplantation. J Clin Oncol 2002; 20: 405–412. 25 Lee JH, Lee KH, Kim S, Seol M, Kim SH, Kim WK et al. Combination chemotherapy of intermediate-dose cytarabine, idarubicin, plus etoposide and subsequent mobilized donor leukocyte infusion for relapsed acute leukemia after allogeneic bone marrow transplantation. Leukemia Res 2001; 25: 305–312. 26 Alessandrino EP, Bernasconi P, Caldera D, Colombo A, Malcovati L, Martinelli G et al. Chemotherapy and donor peripheral blood progenitor cells for acute leukemia in early relapse after allogeneic bone marrow transplantation. Bone Marrow Transplant 1999; 23: 607–612. 27 Lee KH, Lee JH, Choi SJ, Kim S, Lee JS, Kim SH et al. Randomized comparison of two different schedules of granulocyte colonystimulating factor administration after allogeneic bone marrow transplantation. Bone Marrow Transplant 1999; 24: 591–599. 28 Przepiorka D, Weisdorf D, Martin P, Klingemann HG, Beatty P, Hows J et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant 1995; 15: 825–828. 29 Shulman HM, Sullivan KM, Weiden PL, McDonald GB, Striker GE, Sale GE et al. Chronic graft-versus-host syndrome in man. A longterm clinicopathologic study of 20 Seattle patients. Am J Med 1980; 69: 204–217. 30 Choi SJ, Lee KH, Lee JH, Kim S, Chung HJ, Lee JS et al. Prognostic value of hematopoietic chimerism in patients with acute leukemia after allogeneic bone marrow transplantation: a prospective study. Bone Marrow Transplant 2000; 26: 327–332. 31 Lee KH, Lee JH, Choi SJ, Lee JH, Kim S, Seol M et al. Monthly prospective analysis of hematopoietic chimerism after allogeneic hematopoietic cell transplantation. Bone Marrow Transplant 2003; 32: 423–431. 32 Xun CQ, Thompson JS, Jennings CD, Brown SA, Widmer MB. Effect of total body irradiation, busulfan-cyclophosphamide, or cyclophosphamide conditioning on inflammatory cytokine release and development of acute and chronic graft-versus-host disease in H-2-incompatible transplanted SCID mice. Blood 1994; 83: 2360–2367. 33 Hill GR, Crawford JM, Cooke KR, Brinson YS, Pan L, Ferrara JL. Total body irradiation and acute graft-versus-host disease: the role of gastrointestinal damage and inflammatory cytokines. Blood 1997; 90: 3204–3213. 34 Johnson BD, Drobyski WR, Truitt RL. Delayed infusion of normal donor cells after MHC-matched bone marrow transplantation provides an antileukemia reaction without graft-versus-host disease. Bone Marrow Transplant 1993; 11: 329–336. 35 Johnson BD, Truitt RL. Delayed infusion of immunocompetent donor cells after bone marrow transplantation breaks graft-host tolerance allows for persistent antileukemic reactivity without severe graft-versus-host disease. Blood 1995; 85: 3302–3312. 36 Lee KH, Lee JH, Kim S, Lee JS, Kim SH, Kim WK. High frequency of extramedullary relapse of acute leukemia after allogeneic bone marrow transplantation. Bone Marrow Transplant 2000; 26: 147–152. 37 Chong G, Byrnes G, Szer J, Grigg A. Extramedullary relapse after allogeneic bone marrow transplantation for haematological malignancy. Bone Marrow Transplant 2000; 26: 1011–1015. 38 Berthou C, Leglise MC, Herry A, Balcon D, Hardy E, Lessard M et al. Extramedullary relapse after favorable molecular response to donor leukocyte infusions for recurring acute leukemia. Leukemia 1998; 12: 1676–1681. 39 Singhal S, Powles R, Kulkarni S, Treleaven J, Saso R, Mehta J. Long-term follow-up of relapsed acute leukemia treated with immunotherapy after allogeneic transplantation: the inseparability of graft-versus-host disease and graft-versus-leukemia, and the problem of extramedullary relapse. Leukemia Lymphoma 1999; 32: 505–512.


1797 40 Butcher EC, Picker LJ. Lymphocyte homing and homeostasis. Science 1996; 272: 60–66. 41 Simpson DR, Nevill TJ, Shepherd JD, Fung HC, Horsman DE, Nantel SH et al. High incidence of extramedullary relapse of AML after busulfan/cyclophosphamide conditioning and allogeneic stem cell transplantation. Bone Marrow Transplant 1998; 22: 259–264. 42 Lee KH, Lee JH, Choi SJ, Lee JH, Kim S, Seol M et al. Bone marrow versus extramedullary relapse of acute leukemia after allogeneic hematopoietic cell transplantation: risk factors and clinical course. Bone Marrow Transplant 2003; 32: 835–842. 43 Slavin S, Naparstek E, Nagler A, Ackerstein A, Samuel S, Kapelushnik J et al. Allogeneic cell therapy with donor peripheral blood cells and recombinant human interleukin-2 to treat

leukemia relapse after allogeneic bone marrow transplantation. Blood 1996; 87: 2195–2204. 44 Mehta J, Powles R, Singhal S, Tait D, Swansbury J, Treleaven J. Cytokine-mediated immunotherapy with or without donor leukocytes for poor-risk acute myeloid leukemia relapsing after allogeneic bone marrow transplantation. Bone Marrow Transplant 1995; 16: 133–137. 45 Nadal E, Fowler A, Kanfer E, Apperley J, Goldman J, Dazzi F. Adjuvant interleukin-2 therapy for patients refractory to donor lymphocyte infusions. Exp Hematol 2004; 32: 218–223. 46 Frick J, Ritch PS, Hansen RM, Anderson T. Successful treatment of meningeal leukemia using systemic high-dose cytosine arabinoside. J Clin Oncol 1984; 2: 365–368.

Leukemia