An “ATRA-ctive” new treatment of ITP?

In this issue of Blood, Wu and colleagues¹ present encouraging results from one of the accumulating studies evaluating use of all-trans retinoic acid (ATRA) in adult patients with immune thrombocytopenia (ITP).

The most widely recognized use of oral ATRA is for treatment of acute promyelocytic leukemia, in which it has achieved the status of standard of care worldwide in combination with arsenic trioxide. Recently, data suggesting beneficial immunomodulatory effects and rescue of impaired thrombopoiesis in models of ITP^2-4 have supported further investigation into a potential therapeutic role for ATRA in ITP. Because of the heterogeneity of ITP pathogenesis and the potential for changes over time, the availability of agents with diverse mechanisms of actions for use in various combinations is needed for refractory ITP.⁵ 

Regarding ATRA, the few reports currently available explore combinations of ATRA with prednisone (10 mg twice daily)⁴ or danazol² in patients with corticosteroid-resistant ITP. These studies reported response rates of ∼55% to 60% with side effects of dry skin, dry mouth, and headaches. A recently published multicenter randomized controlled phase 2 trial of adults with newly diagnosed ITP in China demonstrated a higher sustained response rate (68%) in patients assigned to high-dose dexamethasone plus ATRA compared with high-dose dexamethasone alone (41%).⁶ 

Wu and colleagues had previously administered ATRA with danazol with the goal of using only oral agents and combining different mechanisms of action.⁴ As addressed by the authors in the discussion, issues with liver disease, hirsutism, and uncertainty about long-term effects made danazol less than ideal. In this report, Wu and colleagues compare the combination of ATRA 20 mg/m² for 12 weeks plus low-dose rituximab (100 mg weekly for 6 weeks) with low-dose rituximab alone in adult patients with corticosteroid-resistant or relapsed ITP. The goal was to achieve “cure” (eg, 12-month sustained response), as well as faster onset of platelet increase with less toxicity. Notable findings included a higher overall response rate in the combination treatment arm (80%, 90/112) compared with monotherapy with low-dose rituximab (59%, 33/56), and a higher sustained response rate, 61% (68/112) vs 41% (23/56), respectively. Both groups had a median time to response of 28 days. Overall toxicity was low, mainly grade 1 to 2 adverse effects, of which dry skin, headache, and dizziness were reported in the ATRA group.

Several limitations to the study deserve mention, including inability to attain the planned sample size (188); 168 patients were enrolled and randomized, which could lead to underpowered analyses, potential impact of concomitant medications (14% to 18% in each arm), and limited follow-up. It is unclear if the promising results with a combination of ATRA with low-dose rituximab translate to long-term response. Per supplemental Figure 2, it appears that the platelet counts decline steadily toward the end of the study. It is also of interest that approximately three-fourths of the patients were entered at 1 study site.

Other issues, too, deserve consideration. First, both arms shared a time to response of 28 days. This is a relatively long time, particularly for patients with low platelet counts, and indeed, there was 1 case of intracranial hemorrhage (on day 27) in the combination arm. The authors address this and suggest utilization of a third agent in bleeding patients. Second, the investigators used low-dose rituximab, albeit for 6 weekly infusions instead of the usual 4 infusions. Higher-dose rituximab might have yielded faster and more durable results in both arms. Third, some of the platelet endpoints were not standard in that they have almost only been used for initial response⁷; this especially applies to sustained response where counts >50 000 per µL are more often used. Finally, the median duration of ITP in enrolled patients was 20 months. This combination may be better applicable to earlier-stage patients rather than ones with ongoing low platelet counts after >1.5 years from diagnosis. Given the current emphasis on not continuing steroids for >6 to 12 weeks, it seems like earlier might be a more appropriate time to explore the treatment combination described here.

Overall, this study supports the hypothesis that “two drugs are better than one”; however, which combination is optimal for use remains an open question. Consideration needs to be given to the mechanisms of effect of each treatment; cost of the agents; potential for overlapping toxicities; and the route of administration. Overall, the intriguing findings reported by Wu and colleagues further the possibilities of ATRA as a novel treatment of ITP and advance its utilization especially in combination perhaps with dexamethasone as recently reported⁶ as well as with rituximab.