Adult idiopathic autoimmune thrombocytopenic purpura is a chronic and sometimes refractory disease. Life-threatening haemorrages are rare but unpredictable, particularly in elderly patients. Corticosteroids are usually proposed as first-line therapy, but a long-term complete response (platelet count >100  ×  109/l) is obtained in only about 20% of patients, and splenectomy is the treatment of choice for non-responders, because it cures 60–80% of patients. However, the best therapy for refractory patients is a difficult challenge because the efficacy of the large number of drugs tried so far is unpredictable and often transient, and their side-effects are sometimes severe.

We describe our experience of treating eight patients (five females) with dapsone. All of the patients gave their informed consent. The results of laboratory tests, including full blood counts, renal and liver function, LDH and G6PDH determinations, showed that all of the patients were platelet autoantibody positive and none had G6PDH deficiency.

The mean age at the time of the start of dapsone therapy was 63 years (range 50–77). A complete response (CR) was defined as a platelet count of > 100  ×  109/l, a partial response (PR) as a platelet count of 50–100  ×  109/l, and no response (NR) as a platelet count of < 50  ×  109/l.

The clinical data and platelet counts of our patients are shown in 1Table I. Six patients had been splenectomized.

Table 1. Table I. Patient characteristics during dapsone therapy. DC, deltacortene; DNZ, danazol; VBL, vinblastine; VDN, vindesine; Ig, immunoglobulin; MTX, methotrexate; CTX, cyclophosphamide; AZT, azathioprine; S, splenectomy; IFN, interferon.Thumbnail image of

The mean disease duration before treatment was 147 months (range 12–300 months). The duration of therapy ranged from to 2 to 5 months, with dapsone being given orally at a dose of 75–100 mg/d. In two patients affected by severe thrombocytopenia and with platelet counts that were constantly about 5  ×  109/l, dapsone was combined with low-dose steroid treatment.

CR was observed in one patient (no. 5) and PR in two (nos. 1 and 4), but dapsone had to be withdrawn because of side-effects. We found haemolysis in patient 1 (haemoglobin values 12.8–10.5 g/dl) and a three-fold increase in transaminases in patient 4; both proved to be rapidly reversible after dapsone was stopped.

No response was observed in five patients, but its administration permitted the withdrawal of steroid therapy in two (nos. 6 and 8).

The mechanism of action of dapsone is unknown, although it has been suggested that the reticuloendothelial system may be blocked as a result of excessive red blood cell destruction.

In comparison with other series reported in the literature (Durand et al, 1991; Godeau et al, 1993, 1997; Hernandez et al, 1995), the mean age of our patients was higher (63 years), the starting platelet count lower (mean 17  ×  109/l) and the mean disease duration longer (147 months). Furthermore, a larger number had unsuccessfully undergone splenectomy (6/8 patients).

Our data also confirmed that patients with severe and long-lasting ITP do not respond to dapsone, but its good tolerability and limited cost mean that it can be proposed as an alternative to conventional second-line treatments.

It could be used in patients with contraindications to splenectomy or to the prolonged use of low-dose steroids, and in patients with a platelet count of >20  ×  109/l and/or a relatively recent diagnosis of ITP.