Real-world treatment patterns & outcomes among patients with immune thrombocytopenia (ITP) who switched treatment from eltrombopag or romiplostim to avatrombopag in the United States: Results from the real-AVA 3.5 study

Abstract In the management of immune thrombocytopenia (ITP), an autoimmune disorder characterized by low platelet counts (PCs) and an increased risk of bleeding, thrombopoietin receptor agonist (TPO-RA) therapies are frequently used as second-line treatments following the failure of steroids and/or immunoglobulin therapy. While all three approved TPO-RAs (romiplostim [ROMI], eltrombopag [ELT], avatrombopag [AVA]) target the thrombopoietin receptor, they differ in their binding site, route of administration, hepatotoxicity, and dietary restrictions. Real-world (RW) studies have shown strong effectiveness in patients with chronic ITP who switched to AVA from a prior TPO-RA (i.e., ELT or ROMI), including those with an inadequate response. However, available data is limited by small sample sizes, the unavailability of comprehensive PC data, and short duration of follow-up. This study aims to describe treatment patterns and outcomes in adult patients with primary ITP who switched treatment from ELT or ROMI to AVA in a predominantly community-based setting. This retrospective chart review study used the Cardinal Health Oncology Provider Extended Network (OPEN). Eligible patients were adults with primary ITP who initiated ELT or ROMI on or after July 1, 2019, switched to AVA within 30 days of discontinuing ELT or ROMI, and had ≥6 months follow-up after AVA initiation. Patients who died within 6 months of AVA initiation were eligible for the study, and patients were not required to remain on AVA throughout the study period. Patients were followed from AVA initiation until date of last contact, death, or study end (21-Mar-2025), whichever occurred first. Treatment response was defined as achieving ≥1 PC above the PC response thresholds (i.e., ≥30x109/L, ≥50x109/L, or ≥100x109/L) at any time during AVA treatment in the absence of rescue therapy. Durability of response was defined as the percentage of time on AVA with PC above the response threshold without the use of rescue therapy. Medical charts for 201 patients who switched to AVA (100 ELT-to-AVA; 101 ROMI-to-AVA) were abstracted. The cohort was majority White (59.2%) and female (60.7%). The mean age at AVA initiation was 56.7 years (SD: 12.4) and median time from primary ITP diagnosis to AVA initiation was 14.6 months (IQR: 8.4-25.2). The mean follow-up duration from AVA initiation was 13.0 months (SD: 10.7). Most common reasons for discontinuing the prior TPO-RA were lack of efficacy (59.2%), patient preference (26.9%), and inconvenient administration (22.9%). Among patients with baseline PCs <30x109/L (n=79; 39.3%), 98.7%, 94.9%, and 49.4% achieved responses ≥30x109/L, ≥50x109/L, and ≥100x109/L on AVA, respectively. Most patients with baseline PCs <50x109/L (95.6%; 108/113) achieved a response ≥50x109/L while on AVA. Mean durability of response was 90.8% (SD: 10.4%), 80.5% (SD: 22.1%), and 61.0% (SD: 29.9%) for PC thresholds ≥30x109/L, ≥50x109/L, and ≥100x109/L, respectively. Both patients receiving concomitant steroids at AVA initiation discontinued while on AVA. Rescue therapy was administered to 10.4% (21/201) of patients while on AVA. At last follow-up, 81.1% (163/201) of patients were still receiving AVA, with a median AVA treatment duration of 8.4 months (IQR: 6.7-10.8). Among patients who discontinued AVA, 47.4% (18/38) discontinued due to achieving target PC levels or remission, 10.5% (4/38) due to adverse events, and 5.3% (2/38) due to death. At last follow-up, 66.7% (24/36) of patients who discontinued AVA for reasons other than death had not received any subsequent ITP treatment. None of the patients who discontinued AVA due to achieving target PCs or remission (n=18) had received subsequent ITP treatment, with a median follow-up duration after AVA discontinuation of 10.0 months (IQR: 4.6-13.2). Response and durability of response with AVA were generally similar for patients switching from ELT or ROMI. In this RW study evaluating the effectiveness of AVA treatment following a switch from ELT or ROMI, most patients achieved or maintained a clinically meaningful and durable platelet response throughout the study observation period. The need for rescue therapy was rare and all patients on concomitant steroids were able to discontinue them, further supporting the effectiveness of AVA in patients with primary ITP who switched from ELT or ROMI.