Engraftment of Allogeneic Hematopoietic Progenitor Cells With Purine Analog-Containing Chemotherapy: Harnessing Graft-Versus-Leukemia Without Myeloablative Therapy

HIGH-DOSE CHEMO-radiotherapy with allogeneic bone marrow transplantation (BMT) has been extensively used to treat patients with hematologic malignancies.1,2 This procedure has generally been limited to younger patients in good general medical condition due to the increased risk of regimen related toxicities and graft-versus-host disease (GVHD) that occurs with age and poor performance status.3-8 Improvements in supportive care and GVHD prophylaxis have now enabled many centers to treat older patients, but few consider patients greater than 55 years of age.4 6 However, acute myeloid leukemia (AML) is most common in older adults. These patients have a poor prognosis and novel therapeutic options need to be explored.

The curative potential of allogeneic BMT is mediated in part by an immune mediated graft-versus-leukemia effect.9-14 Evidence supporting the graft-versus-leukemia effect includes a lower relapse rate in patients with GVHD and a higher risk of relapse after syngeneic or T-cell–depleted transplants.9-14 The most striking evidence for this phenomenon is the observation that infusions of donor lymphocytes can reinduce remissions in many patients.15 Patients with chronic myelogenous leukemia (CML) are most likely to respond, but selected patients with acute leukemia, chronic lymphocytic leukemia (CLL), myeloma, and lymphoma relapsing after allogeneic BMT have also responded.15-17 This finding suggests an alternative strategy of inducing graft-versus-leukemia after standard-dose nonablative chemotherapy as primary treatment for susceptible malignancies in patients ineligible for high-dose chemotherapy or total body irradiation.

The purine analogs fludarabine and 2-chlorodeoxyadenosine (2-CDA) have been shown to be active against a variety of hematologic malignancies.18 These compounds are also immunosuppressive, effectively inhibiting the mixed lymphocyte reaction in vitro.19 20 

We performed a pilot trial to determine whether purine analog-containing nonmyeloablative chemotherapy could be sufficiently immunosuppressive to enable engraftment of allogeneic hematopoietic progenitor cells in patients considered ineligible for myeloablative therapy either because of age or medical condition.

Eligibility criteria.Patients between 55 and 70 years of age with acute leukemia or myelodysplasia beyond first complete remission were considered eligible for this study. Patients less than 55 years of age who were not eligible for a conventional myeloablative transplant protocol because of a concurrent medical condition were also eligible. Patients required an HLA-identical or one-antigen–mismatched related donor, a serum bilirubin level of less than 3.0 mg/dL, a serum creatinine level of less than 2.0 mg/dL, a cardiac ejection fraction of 40% or greater, and a Zubrod performance status of ≤2. Patients and donors signed written informed consents and the treatment was approved by the institutional review board.

Stem cell collection.Donors received filgrastim at 6 μg/kg subcutaneously twice daily starting 4 to 5 days before the first collection. Leukaphereses were performed daily using conventional techniques for blood progenitor cell collection until greater than 3 × 10⁶ CD34⁺ cells/kg recipient were collected.21 Donor cells were cryopreserved using standard techniques. If insufficient number of cells were collected or if peripheral blood stem cell collection was not feasible, donor bone marrow was harvested.

Chemotherapy.Fifteen patients were treated. Seven patients without prior exposure to fludarabine received fludarabine at 30 mg/m²/d intravenously at the same time over 30 minutes for 4 days with either ara-C at 2 g/m² administered intravenously over 4 hours starting 4 hours after the beginning of the fludarabine infusion and idarubicin at 12 mg/m² intravenously for 3 days.22 One patient received the same fludarabine schedule with melphalan at 140 mg/m² administered once at the end of chemotherapy. Seven patients with prior fludarabine exposure received ara-C at 1.0 g/m²/d over 2 hours intravenously for 5 days and 2-CDA at a dose of 12 mg/m²/d by continuous intravenous infusion for 5 days beginning 4 hours after the first dose of ara-C.23 Cells were infused 2 days after the last dose of chemotherapy.

Supportive care.All patients were treated as inpatients in private rooms. Patients received antibacterial and antifungal prophylaxis with intravenous vancomycin at 1 g daily or oral penicillin at 250 mg four times daily in addition to norfloxacin at 400 mg orally twice daily and fluconazole at 200 mg/d. Broad spectrum antibiotics were begun for temperature greater than 38.3°C or clinical signs of infection.

Patients received filgrastim at 5 μg/kg/d from the day of the transplantation (day 0) until achievement of an absolute neutrophil count (ANC) of greater than 1.5 × 10⁹/L. Hemoglobin was maintained at a level of ≥8 g/dL and the platelet count was maintained at ≥20 × 10⁹/L with filtered and irradiated blood products.

GVHD prophylaxis consisted of cyclosporine at a dose of 3 mg/kg/d by continuous intravenous infusion changed to oral as soon as tolerated and methylprednisolone at 1.0 mg/kg/d beginning on day +5 tapered over the following 8 weeks. Patients without signs of GVHD on day 50 had their cyclosporine tapered by 10% to 20% weekly. Patients developing GVHD received methylprednisolone at 2 mg/kg/d intravenously and tapered upon response as tolerated.

Study endpoints.The primary objectives of this study were to evaluate the engraftment potential and antileukemic effects of allogeneic hematopoietic progenitor cells in patients with acute leukemia or myelodysplastic syndromes (MDS) receiving intensive, but nonmyeloablative purine analog-containing chemotherapy. Chimerism and evidence of minimal residual disease were determined by conventional cytogenetics and restriction fragment length polymorphisms (RFLPs) using published techniques.24,25 Additional endpoints included toxicity as defined by the criteria of Bearman et al26 and hematologic recovery defined as the first of 3 consecutive days with an ANC greater than 0.5 × 10⁹/L. Acute and chronic GVHD were scored according to standard criteria.27,28 Complete remission was defined as less than 5% blasts in bone marrow with normal maturation, an ANC greater than 1.5 × 10⁹/L, and platelet transfusion independence. Survival was calculated as of August 8, 1996 from the day of transplantation according to the methods of Kaplan and Meier.29 

Patient and disease characteristics.From July 1995 to July 1996, 15 patients with acute leukemia or MDS were treated. Patient and disease characteristics are summarized in Table 1. In brief, the median age was 59 years (range, 27 to 71 years). Thirteen patients had AML (3 preceded by a prior hematologic disorder), 1 patient had chronic myelomonocytic leukemia (CMML), and 1 patient had refractory anemia with excess blasts in transformation (RAEB-T). The median time to transplantation from diagnosis was 391 days (range, 50 to 1,315 days). Nine patients were refractory to salvage chemotherapy, 3 patients were in untreated relapse (first, second, and third relapse), 1 patient was in a third remission, 1 patient had an untreated secondary AML 4 years after an autologous BMT, and the patient with CMML had received hydroxyurea chemotherapy. The median number of prior therapies was 2 (range, 0 to 4).

The median donor age was 57 years (range, 25 to 68 years). All donors were either fully HLA-compatible siblings (n = 13) or one-antigen–mismatched siblings (n = 2). Six donors were women and 9 were men; 6 donor-recipient pairs were sex-mismatched. Donors received a median of 5 days of treatment with filgrastim. Eight underwent one apheresis procedure and 5 required two procedures. The median CD34⁺ cell yield expressed as number of CD34⁺ cells per liter of blood processed during the first collection was 21.3 × 10⁶ (range, 6.2 to 45.5 × 10⁶). The median number of CD34⁺ cells infused was 4.5 × 10⁶/kg (range, 1.7 to 9.9 × 10⁶/kg).

Toxicity.The chemotherapy was well tolerated in 13 of the 15 patients. One patient (unique patient no. [UPN] 1100), who had secondary AML and progressive pneumonia, died from multiorgan failure after the second dose of fludarabine, idarubicin, and ara-C. She developed progressive respiratory failure with renal and hepatic deterioration. Further chemotherapy was discontinued, and she died 48 hours later without receiving her donor cells. The only other instance of grade 3 toxicity was a patient with extensive anthracycline exposure and an ejection fraction of 54% before therapy (UPN 6108) who developed congestive heart failure that responded to diuretics and inotropic support.

Engraftment.Thirteen patients achieved an ANC of greater than 0.5 × 10⁹/L a median of 10 days posttransplantation (range, 8 to 17 days). One patient never cleared his peripheral blood blasts and failed to recover normal hematopoiesis. Ten patients achieved a platelet count of 20.0 × 10⁹/L or greater without transfusions a median of 13 days postinfusion (range, 7 to 78 days).

GVHD.Acute GVHD occurred in 3 patients, 1 limited to skin alone (grade 1) and the other 2 involving skin and gut (grade 2). GVHD responded to steroid therapy alone in 2 patients and required antithymocyte globulin in another. Chronic GVHD has not occurred, but only 5 patients are evaluable for this complication.

Response and chimerism.Thirteen of 15 patients cleared their peripheral blood blasts. Eight patients achieved remission criteria, with 6 having ≥90% donor cells at that time (between days 14 and 30). One patient achieved remission criteria without evidence of donor cell engraftment by cytogenetics or RFLP, and 1 patient could not be assessed. Five patients relapsed between 43 and 127 days posttransplantation (median, 65 days). Responses and chimerism for each patient are shown in Tables 2 and 3.

Survival.Six patients remain alive between 34 and 175 days posttransplantation (median, 100 days). Two patients remain in remission 34+ and 170+ days posttransplantation; the other 4 have active disease. None of the relapsing patients has responded to cyclosporine withdrawal or filgrastim therapy as treatment of relapse.30 The median survival was 78 days (range, 0 to 175+ days). The causes of death were leukemia (n = 5), infection after subsequent therapy (n = 2), multiorgan failure (n = 1), and aspergillosis (n = 1). The actuarial 100 day survival was 66% ± 19% for patients achieving a complete remission and 21% ± 17% for patients not achieving this response (P = .16 log rank).

Despite modern combination chemotherapy, most patients with acute leukemia relapse within the first 2 years of achieving complete remission. Durable responses to salvage therapy are rare, with 24% of patients expiring before achieving a response and approximately 33% achieving second remissions; the median survival is 18 weeks. The prognosis with conventional chemotherapy for second or third relapse is worse, with a median survival of 7 weeks and a complete remission rate of less than 10%.31 

Myeloablative chemoradiotherapy with allogeneic progenitor cells can result in long-term disease-free survival in selected patients with advanced acute leukemia. This therapy is associated with substantial toxicity and the treatment-related mortality rate approaches 40%. This limits the use of transplantation therapies to younger patients with a relatively good performance status.1 2 

The antileukemic effect seen with donor lymphocyte infusions in patients relapsing after allogeneic transplantation suggests that if engraftment of donor hematopoietic cells can be achieved after standard-dose chemotherapy, it may be possible to exploit the graft-versus-leukemia effect without the potential morbidity associated with myeloablative therapy.15 16 

The purine analogs fludarabine and 2-CDA produce lymphocytopenia and substantial immunosuppression and could therefore enhance engraftment of allogeneic hematopoietic progenitors.18-20 Fludarabine and 2-CDA combinations have been extensively studied in combination with ara-C as salvage therapy for patients with acute leukemia.22 23 The addition of allogeneic hematopoietic progenitor cells could potentially enhance hematopoietic recovery as well as provide a graft-versus-leukemia effect. The presence of stable chimerism would allow subsequent donor lymphocyte infusions in patients with minimal residual disease or a high likelihood of relapse.

This study examined the feasibility of this approach in poor prognosis patients with recurrent or resistent AML or MDS who were not eligible for myeloablative therapy because of advanced age or concomitant medical illness. Thirteen patients recovered granulocytes between 8 and 17 days postinfusion (15 to 24 days after the beginning of chemotherapy), with 6 patients in remission with predominantly donor cells. Moreover, 3 patients in remission 60 to 90 days postinfusion had ≥80% donor cells as assessed by cytogenetics or RFLP. We did not perform studies concerning lineage-specific engraftments and the extent of lymphoid engraftment is unknown. Longer follow-up and specific assays for chimerism in each lineage are required for evaluation of the extent and durability of engraftment. In this study, the durability of the graft was difficult to assess in our patients due to the fact that many of them had recurrent leukemia within a short period of time. However, the observation that 3 of 4 patients in remission had predominantly donor cells between 2 and 3 months posttransplantation without other signs of graft failure suggests the possibility of durable engraftment in some patients.

We were unable to demonstrate the feasibility or efficacy of donor lymphocyte infusions for treating relapse in this setting because many patients relapsed quickly after transplantation with rapidly progressive disease. However, this study suggests that a strategy of purine analog-containing nonmyeloablative chemotherapy followed by infusion of donor lymphocytes for patients with evidence of minimal residual disease or with a high risk of relapse could be feasible in patients with less advanced hematologic malignancies. Such a strategy may be particularly useful for patients with CML, multiple myeloma, or acute leukemia in first remission who have an HLA-compatible donor but are deemed ineligible for a conventional myeloablative preparative regimen due to their age or medical condition.

Many other potential applications exist for a nonmyeloablative regimen that can consistently achieve engraftment of allogeneic hematopoietic progenitor cells. These include treatment of nonmalignant hematologic disorders or of immune or metabolic diseases or induction of tolerance for solid organ transplantation. The results of this study suggest that purine analog-containing nonmyeloablative regimens should be explored in these settings.

In summary, this study shows that purine analog-containing nonmyeloablative chemotherapy followed by allogeneic hematopoietic progenitor cells is feasible in elderly or debilitated patients with advanced leukemia. This therapy can result in engraftment of donor cells with rapid hematologic recovery and little regimen-related toxicity. Further studies with a larger number of patients and longer follow-up will be necessary to determine if this approach will be associated with an acceptable risk of acute and chronic GVHD and if further infusions of donor lymphocytes for treatment of minimal residual disease or as relapse prevention will improve leukemia-free survival in these patients or in patients with less advanced disease. Further exploration of this strategy as alternative therapy for patients with hematologic malignancies who have an HLA-compatible donor but are ineligible for a conventional myeloablative therapy is warranted.

The authors acknowledge the contributions of the nurses, fellows, pharm D's, and laboratory technicians at the M.D. Anderson Cancer Center, whose invaluable assistance and dedication to excellent and compassionate patient care made this work possible.