The goal of this clinical research study is to find the safety of decitabine in patients with acute lymphocytic leukemia. Upon agreement of the patient, additional blood and bone marrow samples to be used to evaluate the effect of the treatment on leukemic cells. Also, with agreement of the patient, any leftover blood and bone marrow samples that are collected at the start of the study and during the regularly scheduled evaluations to be sent for research studies. The research studies will examine changes in the blood and bone marrow cells that might help explain the causes of leukemia.

Decitabine is a potent hypomethylating agent with clinical activity in myelodysplastic syndromes (MDS), and acute and chronic myelogenous leukemia (CML). In vitro, decitabine induces loss of cell viability and apoptosis in ALL derived cell lines with known DNA methylation alterations. Exposure of these cell lines to decitabine results in hypomethylation and reactivation of putative tumor suppressor genes, an effect that is thought to have a role in the antineoplastic activity of decitabine.

Aberrant DNA methylation of multiple promoter CpG islands is frequently observed in patients with ALL both at initial presentation and at the time of relapse. Indeed these methylation marks are stable in over 70% of patients with ALL at the time of relapse. Importantly, methylation of specific molecular pathways has been associated with an extremely poor prognosis in patients with ALL. For instance, data from our laboratory has identified methylation, and silencing, of a cell cycle pathway composed of p73 and the cyclin dependent kinase inhibitors p57KIP2 and p15, as a marker of poor prognosis in patients with Philadelphia chromosome (Ph) negative disease. These results have been corroborated at the protein level: expression of p57KIP2 and or p15/p73 has been associated with a better prognosis. Finally, although the global methylation patterns observed in children with ALL, that overall have an excellent prognosis, do not seem to differ with those of older patients with the same genetic characteristics, methylation of prognostically significant pathways, such as P73/P15/P57KIP2 are remarkably lower in the younger patients. Finally, introduction of p57KIP2 in methylated/silenced ALL cell lines results in cell cycle arrest and induction of apoptosis.

All these data indicates that aberrant methylation has a role in the clinical behavior of patients with ALL and that its reversal may result in clinical benefit.

Inclusion Criteria:

1. Patients with refractory or relapsed acute lymphocytic leukemia (ALL).
2. Signed informed consent indicating that patients are aware of the investigational nature of this study in keeping with the policies of University of Texas MD Anderson Cancer Center (UTMDACC).
3. Patients of any age are eligible.
4. Patients must have been off chemotherapy for 1 week prior to entering this study and recovered from the toxic effects (< grade 2) of that therapy, unless there is evidence of rapidly progressive disease. Use of high dose steroids with dexamethasone is allowed during the first 2 courses of therapy. Imatinib mesylate (Gleevec) must be stopped 1 week prior to entering this study.
5. Adequate liver function (bilirubin of < 3 mg/dL, serum glutamate pyruvate transaminase (SGPT) < 5 x ULN) and renal function (creatinine < 3mg/dL) unless proven to be related to disease infiltration.
6. Women of childbearing potential must practice contraception. Child bearing potential defined as not post-menopausal for 12 months or no previous surgical sterilization. Men and women must continue birth control for the duration of the trial.

Exclusion Criteria:

1) Nursing and pregnant females are excluded.