A critical appraisal of low-dose cytosine arabinoside in patients with acute non-lymphocytic leukemia and myelodysplastic syndromes.

We reviewed 53 publications reporting 751 patients with hematologic malignancies treated with low doses (5 to 20 mg/m2/d) of cytosine arabinoside (LoDAC). Of 507 patients evaluable for response, complete remission (CR) rates varied from 32% for patients with primary acute non-lymphoblastic leukemia (1 degree ANLL) to 16% for patients with hematologic malignancies secondary to previous chemotherapy or following a myelodysplastic syndrome (MDS) (2 degrees ANLL), and 16% for MDS. Median duration of CR was 9.5 months for 1 degree ANLL, and 10.5 months for both 2 degrees ANLL and MDS. Based on limited available survival data, overall median survival for these groups was 9 months, 3 months, and 15 months, respectively. Only three CRs were reported of 31 evaluable patients treated for a variety of other hematologic malignancies. CR rates for patients with 1 degree ANLL greater than or equal to 50 years old was 56%, compared with 29% less than 50 years old (P = .10). While prior chemotherapy was more common in 1 degree ANLL patients less than 50 years of age (86% v 21%; P less than .001), it did not influence response rates in those greater than 50 years of age, suggesting other biases. Hematologic toxicity was mentioned in only 33 of 53 publications, affecting 254 of 289 patients (88%), with at least 15% treatment-related deaths. LoDAC hypothetically acts as a differentiating agent; however, correlative laboratory studies were rarely performed. Cytogenetic data were available for only 15%, and in vitro studies for 10% of all patients with marked discrepancies in the interpretation of results. LoDAC is clearly cytotoxic for both malignant and normal hematopoietic cells. While large numbers of patients have been reported, the lack of well-designed clinical trials prohibits definitive conclusions as to its role as a differentiating agent. Future LoDAC studies should determine optimal dose and schedule, with clinical laboratory correlates to assess differentiation. Trials in ANLL comparing LoDAC with conventional chemotherapy, and in MDS with supportive care alone, may help identify the role of LoDAC. Until appropriate indications can be identified, LoDAC should not be routinely used in clinical practice.