Gemtuzumab ozogamicin (GO) was approved for marketing in 2000 by the United States Food and Drug Administration (FDA) for older patients with relapsed acute myeloid leukemia (AML). Four months later, 14 phase II clinical trial participants who received novel GO-containing combination chemotherapy regimens developed an unexpected hepatic toxicity termed sinusoidal obstructive syndrome (SOS) or hepatic veno-occlusive disease (VOD). Investigators associated with the Research on Adverse Drug Events and Reports (RADAR) project reviewed safety reports for GO included in reports of clinical trials and observational studies, interim reports from an FDA mandated Prospective Observational Registry, and the Food and Drug Administration's Adverse Event Reporting System. Medline searches provided incidence estimates of GO-associated SOS and comparative rates of SOS without GO. SOS is characterized by hyperbilirubinemia, painful hepatomegaly, ascites, and sudden weight gain developing at a median of 10 days following GO administration for patients who did not undergo an allogeneic SCT procedure and 13 days following an allogeneic SCT for patients who had previously received GO. Among adult AML patients who received GO in clinical trials, SOS incidence was 3% at doses < or =6 mg/m(2) if administered as monotherapy or in combination with non-hepatotoxic agents versus 28% if administered with thioguanine and 15% when administered as monotherapy at a dose of 9 mg/m(2). Observational studies identified SOS rates between 15% and 40% if an SCT is performed within 3 months of GO administration. The FDA mandated Prospective Observational Registry of patients who receive care at 60 medical centers has identified GO-associated SOS rates of 14% if an SCT is performed and 9% otherwise. Caution is advised when administering GO in routine clinical practice, particularly if administered with other hepatotoxic agents, at doses and schedules more intensive than those approved by the FDA, or within 3 months of a SCT procedure.