University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma 73190, USA.
BACKGROUND: Treatment of chronic refractory idiopathic thrombocytopenic purpura is a dilemma because many patients have minimal symptoms, response to treatment is uncertain, and treatments may have serious adverse effects. PURPOSE: To determine the effectiveness of treatments for adult patients with idiopathic thrombocytopenic purpura who have not responded to splenectomy. DATA SOURCES: English-language reports from 1966 through 2003 that were retrieved from MEDLINE and Reference Update and bibliographies of retrieved articles. STUDY SELECTION: Articles reporting 5 or more total patients were reviewed to select eligible patients. Patients were eligible for inclusion if they were more than 16 years of age, had idiopathic thrombocytopenic purpura for more than 3 months, had a previous splenectomy, and had a platelet count less than 50 x 10(9) cells/L. DATA EXTRACTION: Patients were assessed for platelet count response, bleeding complications, duration of follow-up, and death. Complete remission was defined as a normal platelet count with no treatment for more than 3 months and for the duration of follow-up. DATA SYNTHESIS: 90 articles with 656 patients treated with 22 therapies met selection criteria. Azathioprine, cyclophosphamide, and rituximab had the most reported complete responses, but they were reported in only 41 to 109 patients. Reported complete response rates ranged from 17% to 27%, but 36% to 42% of patients had no response with these 3 treatments. Most reports described only platelet count responses; bleeding outcomes were reported in only 63 patients (10%). Only 111 (17%) of the 656 eligible patients had pretreatment platelet counts of less than 10 x 10(9) cells/L. No treatment method was reported in more than 20 patients. CONCLUSIONS: Evidence for the effectiveness of any treatment for patients with idiopathic thrombocytopenic purpura and persistent severe thrombocytopenia after splenectomy is minimal. Potentially effective treatments must be evaluated by randomized, controlled trials to determine both benefit and safety.