• Disease-free survival one year after NK cell infusion [ Designated as safety issue: No ]
• Achievement of complete or partial remission at one year following NK infusion [ Designated as safety issue: No ]

RATIONALE: Giving chemotherapy drugs, such as cytarabine and cyclophosphamide, before a donor natural killer cell infusion helps stop the growth of cancer cells. It also helps stop the patient's immune system from rejecting the donor's stem cells. When the natural killer cells from a donor are infused into the patient they may help kill more cancer cells.

PURPOSE: This phase II trial is studying how well giving cytarabine or cyclophosphamide followed by a donor natural killer cell infusion works in treating patients with relapsed or persistent acute myeloid leukemia, acute lymphoblastic leukemia, chronic myelogenous leukemia, or myelodysplastic syndrome after a hematopoietic stem cell transplant.

OBJECTIVES:

Primary

• Determine the anti-leukemic efficacy of allogeneic HLA-haploidentical related natural killer (NK) cell infusion following a cytoreductive regimen with either cytarabine or cyclophosphamide in patients with acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), chronic myelogenous leukemia (CML), or myelodysplastic syndromes (MDS) who have relapsed following allogeneic hematopoietic stem cell transplant, where efficacy is defined as the achievement of complete or partial remission at one year following NK cell infusion.

Secondary

• To assess treatment efficacy, as defined by achievement of complete or partial remission, at 3 and 6 months following HLA-haploidentical related NK cell infusion.
• To assess the effects of an HLA-haploidentical related NK cell infusion on the sustained engraftment and recovery of an HLA-matched stem cell allograft.
• To assess the risk of inducing graft-vs-host disease (GVHD) or altering its severity.
• To provide preliminary evidence that specific donor KIR-recipient HLA ligand combinations relating to missing self-MHC class I ligand or missing class I ligand are associated with higher NK alloreactivity and improved outcome.
• To monitor the extent and duration of NK cell donor chimerism.
• To monitor NK cell reconstitution through NK receptor cell surface phenotyping (CD94/NKG2A, ILT-2, KIR expression) and function (intracellular IFN-γ, cytotoxicity) on day 15, 30, 60, 100, and 200 following the NK infusion and to correlate with outcome.
• To correlate the magnitude of NK effect with disease (myeloid versus lymphoid) and known survival risk factors (time from allogeneic HSCT to relapse; < 6 months vs > 6 months).

OUTLINE: Patients receive either cytarabine IV on days -6 to -2 or cyclophosphamide IV over 1 hour on days -3 and -2. Patients then receive an infusion of HLA-haploidentical allogeneic natural killer cells on day 0.

After completion of study therapy, patients are followed at days 15, 30, 60, 100, and 200 and at 1 year.

• Leukemia
• Myelodysplastic Syndromes

• Biological: natural killer cell therapy
• Drug: cyclophosphamide
• Drug: cytarabine

DISEASE CHARACTERISTICS:

• Pathologically confirmed diagnosis of leukemia (acute myeloid leukemia [AML], acute lymphoblastic leukemia [ALL], blastic phase chronic myelogenous leukemia [CML]), or myelodysplastic syndromes (MDS)

  • Evidence of disease with > 5% bone marrow involvement detected by morphology or abnormal cytogenetics (by karyotype or FISH)
  • Patients with molecular detection of markers characteristic of the patient's disease from two consecutive bone marrow biopsies are also eligible

• Relapsed or persistent disease after prior hematopoietic stem cell transplantation (HSCT)

  • Patient has undergone prior related or unrelated HLA-compatible (9/10 or 10/10) HSCT within the past year

    • Patients with documented graft versus host disease (GVHD) are not excluded from this trial, but must not have used systemic immunosuppression (e.g., systemic steroids, calcineurin inhibitors, MTOR-inhibitors, budesonide, anti-thymocyte globulin) for at least one month
    • Patients with grade I/II acute GVHD or limited chronic GVHD and receiving localized GVHD therapy (e.g., topical steroids) are not excluded from this trial
• Deemed ineligible for second HSCT after consideration of adequacy of their physical function, extent of disease, and prior treatment-related toxicities
• CML patients should have already received and failed bulk donor leukocyte infusion or donor lymphocyte infusion (DLI) (if the stem cell donor is available for DLI) for relapsed/persistent disease
• Patients with extramedullary relapse are eligible except for those with CNS involvement
• Stem cell boost from the related or unrelated HSCT donor available
• Related HLA-haploidentical NK cell donor available

PATIENT CHARACTERISTICS:

• Patient must not be pregnant and must be using an adequate form of birth control
• Karnofsky (adult) or Lansky (pediatric) performance status 70-100%
• LVEF at rest must be > 50%
• ALT < 2 x upper limit of normal (ULN) and total bilirubin < 1.5 mg/dL, unless liver is involved with the disease or there is congenital benign hyperbilirubinemia
• Serum creatinine < 1.5 mg/dL or creatinine clearance > 50 mL/min
• Not oxygen-dependent
• No active infection

PRIOR CONCURRENT THERAPY:

• See Disease Characteristics
• Patients who have received previous adoptive cellular therapy such as donor lymphocyte infusion (DLI) are not excluded from this trial as long as their disease has been documented to progress within two months of receiving DLI or if the patient has not received DLI within the past two months

• At least 1 month since prior and no concurrent systemic immunosuppressive agents including, but are not limited to, systemic steroids, calcineurin inhibitors, MTOR-inhibitors, budesonide, or anti-thymocyte globulin

  • Localized therapy (e.g., topical steroids) for GVHD allowed