Immunomodulatory treatments for persistent and chronic immune thrombocytopenic purpura: A PRISMA-compliant systematic review and meta-analysis of 28 studies.

BACKGROUND: Corticosteroid sparing is required in 15% to 40% of adults with persistent or chronic primary immune thrombocytopenic purpura (ITP). Herein, the efficacy of immunomodulatory drugs (dapsone, interferon alpha, danazol, and hydroxychloroquine as second-third-line therapies in ITP is investigated.

METHODS: MEDLINE was searched for studies that included patients with persistent or chronic primary ITP and published before the end of December 2014. Two investigators independently extracted data regarding study design, patient characteristics, dosage schedule, time to response, and occurrence of adverse events. The pooled overall response rate (ORR; platelet count >30 × 10 L) and the complete response rate (CRR; platelet count >100 × 10 L) were evaluated to determine drug efficacy by calculating weighted mean proportion using a fixed or random-effects model according to heterogeneity (I > 50%). The study was performed following the MOOSE and PRISMA guidelines.

RESULTS: A total of 28 studies (415 patients) were included (dapsone: k = 7 studies, n = 80; danazol: k = 12, n = 224; interferon alpha: k = 8, n = 83; hydroxychloroquine: k = 1, n = 28). The mean patient age was 50 years (female sex 70%, splenectomy 47%). The ORR and CRR were 55% (95% CI: 44%-66%, I = 0%) and 21% (95% CI: 13%-31%, I = 0%), respectively, for dapsone; 42% (95% CI: 22%-65%, I = 63%) and 18% (95% CI: 10%-29%, I = 9%), respectively, for interferon alpha; and 58% (95% CI: 42%-72%, I = 67%) and 29% (95% CI: 19%-42%, I = 63%), respectively, for danazol. The ORR was 50% (95% CI: 32%-67%) for hydroxychloroquine (data not available for CRR). Meta-regression analysis found a correlation between the ORR for interferon alpha and the splenectomized status of the patient (P = .02) and between the CRR for danazol and disease duration (P < .001). In total, 73%, 51%, 30%, and 0% of patients who received danazol, dapsone, interferon alpha, and hydroxychloroquine experienced side effects, respectively.

CONCLUSION: The ORR was equivalent for hydroxychloroquine, danazol, and dapsone in ITP. Regarding their low CRR, patients at high risk of infection or at low risk of bleeding should benefit from these treatments. Thanks to their best efficacy and safety profiles, dapsone and hydroxychloroquine in patients with antinuclear antibodies should be preferred over danazol and interferon alpha.