Rituximab Yielded a Significantly Longer Response Compared to Placebo in Steroid Free Population - a Post Hoc Analysis of the Ritp Study

Immune thrombocytopenia (ITP) is an autoimmune disease characterized by immune mediated platelet destruction causing thrombocytopenia and increase risk of bleeding.

Treatment is indicated in patients who present with bleeding or are at risk of bleeding. Corticosteroids (CS) are the main first-line treatment in adult patients, but since not all patients achieve adequate response to CS, most patients require a second line therapy. Rituximab is one of the most widely used second line therapies; its advantages include the ability to induce relatively long lasting remissions after 2-4 infusions and its favorable safety profile.

The RITP study was a randomized placebo-controlled trial in which ITP patients who failed to achieve adequate response to CS were randomized to receive rituximab or placebo. The study outcomes were treatment failure (splenectomy or meeting criteria for splenectomy), response rates and duration of response during an 18-month follow-up period. Apart from a longer duration of response in the rituximab arm, the study showed no significant differences in the other outcomes (Ghanima et al. Lancet 2015; 385: 1653-61). The use of stable dose of CS was allowed during the study.

This report represents a post-hoc analysis of the RITP trail, restricting the population to those who did not receive any CS during the study, until relapse or treatment failure in the two arms. By this we aimed to eliminate the possible interference resulting from concomitant use of CS on the effect and safety of rituximab and placebo. For effect, we estimated the duration of response, splenectomy rates, platelet counts, and time to bleeding and for safety we estimated time to infection episodes in both arms.

We included patients who did not receive CS before treatment failure or relapse or until the end the of the study, if none of the endpoints was achieved. The duration of response was estimated from time of achieving response to relapse (platelet count <30x10⁹/L). Kaplan Meier plots were constructed to depict time to relapse, bleeding and infections using log-rank test to determine statistical significance between the two study groups.

Of the 109 included patients, 45 (41%) patients did not receive CS in the two arms. Of these 45 patients, 27 achieved a response during the study; 14 in the rituximab arm and 13 in the placebo arm.

Median duration of response was significantly longer in the rituximab arm (median 308 days; IQR: 168- not reached) vs placebo (41 days; IQR: 36-42) arm (p=0.0019), figure 1. Splenectomy rates did not differ between the two arms; 4 in the rituximab arm vs 3 in the placebo. A trend towards lower bleeding episodes was observed in the rituximab arm (p=0.12), figure 2. Mean platelet count was higher during the study in the rituximab arm compared to placebo as shown in figure 3. There was no difference in time to infection in both arms.

One of the limitations of the RITP study was that CSs (prednisolone) were allowed in an unstandardized fashion. Thus, use of corticosteroids could have reduced the difference in effect between the two treatment arms masking some of the benefit of rituximab. Interestingly, no difference was found in the rate of response in the 2 arms; however, almost all those who were classified as responders in the placebo arm, relapsed within the first few weeks, whereas a much longer time to relapse was observed after rituximab. The unexpectedly high response rates in the placebo arm may be explained by the use of pre-medications, with CS received prior to infusion. Although the duration of response was also significantly longer in the original study, a greater difference was observed when restricting the population to those who did not use CS. Nevertheless, splenectomy rates were similar.

In conclusion, rituximab yielded significantly longer duration of response and although response to rituximab was transient, half of the patients, displayed a treatment free response of more than 10 months, as well as it resulted in higher platelet count and less bleeding episodes without causing more infections.

View largeDownload slide

View largeDownload slide

Close modal