The retrospective study by Chang et al 1 describes the characteristics of a cohort of 375 ITP patients. The first conclusion is that corticosteroids are efficient as a first-line option, which is not really a surprise. The paper then focuses on two second-line treatments, splenectomy, and azathioprine. Whereas the former has long been the reference for chronic/refractory ITP, 2 the latter is used much more rarely and is not even mentioned in a timely review of "old" ITP medications. 3 However, there are some reports about its efficiency in the literature. 4,5 The authors obtain a surprisingly high rate of success and explain the discrepancy with former studies with "difference in case enrollment and definition of response". 1The outcome of a high number of ITP cases has improved since the addition of rituximab and thrombopoietin receptor agonists to the therapeutic arsenal. Nevertheless, there will always be patients not responding to these treatments or presenting contraindications, so that alternatives remain necessary. One of such valuable alternatives is danazol, which Chang et al 1 do not mention, but which was effective in almost all of the studies having included sufficient patient numbers. The series of Ahn et al 6 reports about 96 patients of whom 61.4% achieved a platelet count of ≥50 x 10 9 /L. Among the 26 patients described by Fermand et al, 7 51.4% responded and entered remission. In 2004, Maloisel et al 8 published a study of 57 patients with an overall response rate (defined as a platelet count above 50 x 10 9 /L) of 67% sustained for at least 2 months. More recently, Liu et al 9 saw an overall response rate of 65% in a cohort of 294 patients (103 treated with danazol alone and 191 with danazol and corticosteroids). In a metaanalysis of the effect of immunomodulatory medications for chronic ITP, danazol reached an overall response rate of 58%, the definition of response being in this case a platelet count >30 x 10 9 /L. 10 Finally, Feng et al tested the combination of danazol with all-trans retinoic acid compared to danazol alone in a multicentre, randomized, open-label, phase 2 trial and observed a response rate at 12 months of 62% with the combination therapy vs 28% in the danazol only arm. 11Overall, danazol appears as a good choice for the second-line treatment of ITP (especially in older patients) and is clearly at least as efficient as most of the other possibilities. A dosing of 200-400 mg daily is generally well tolerated. 9,11 However, platelet counts do usually not increase immediately after the introduction of danazol. An additional medication might be initially required. Danazol is less expensive than rituximab.Nevertheless, caution is necessary because of the contraindications (particularly pregnancy but also liver diseases, current or past thromboembolism, hormone-dependent cancers) and potential side effects of this androgen, especially in female patients, like hirsutism, weight gain, or acne. Further adverse reactions include pseudotumor cerebri, stroke, hepatotoxicity, but their i...