Long-Term Overuse of Corticosteroids in Patients with Immune Thrombocytopenia: A Real-World Analysis of a US Claims Database

Background: Corticosteroids (CSs) are standard first-line (1L) therapy for immune thrombocytopenia (ITP); prolonged high-dose exposure is often associated with serious toxicity. Multiple second-line (2L) treatment options exist, but with little consensus on the preferred sequence.

Aim: To assess real-world ITP treatment patterns using the US IBM^® MarketScan^® database.

Methods: This retrospective, observational analysis used claims in the US IBM MarketScan database (2011–2017) for patients (pts) newly diagnosed with ITP, with commercial or private Medicare health coverage. Selection criteria: adults with ≥ 1 primary ITP diagnosis (ICD-9-CM 287.31 or ICD-10-CM D69.3); ≥ 12 months of continuous enrollment before the date of ITP diagnosis; ≥ 1 ITP treatment; ≥ 1 month of continuous enrollment following initiation of first ITP treatment (treatment index date). Pts with viral hepatitis and HIV were excluded. Line of therapy (LOT) was continuous until 1 of the following occurred: discontinuation of all treatments for ≥ 60 days (d), addition of a new therapy after discontinuing all or ≥ 28 d after initiation of a new regimen, disenrollment, death, or end of the observation period.

Results: Eligible ITP pts (7,837/50,428 [16%]) were included; the majority of pts were excluded due to not meeting criteria for newly diagnosed ITP and ≥ 1 month of continuous enrollment after first ITP treatment. For each LOT, > 50% of pts received monotherapy (Fig 1A). CSs were the most common treatment for each LOT, even at 7L; CS use did not decrease with subsequent LOT: 85% of pts received CS as part of 1L, and > 85% for each LOT thereafter, with 89% receiving CS as part of 7L. Prednisone was the most commonly used; prednisone re-initiation was the most common treatment sequence across 1L–3L. Prednisone was frequently used as monotherapy (Fig 1B); highest rate: 58% in 1L, lowest rate: 38% in 2L. For pts treated with CSs, ≥ 92% in each LOT were prescribed a prednisone dose of ≥ 5 mg/d or its equivalent. Pts were prescribed CSs for a median of 76 d during the first year of follow-up. Rituximab (highest in 2L, 17%), thrombopoietin receptor agonists ([TPO-RAs] 12% in 2L, 14% in 3/4L), and splenectomy (highest in 3L, 5%) were used much less frequently than CSs.

Conclusion: Long-term, higher-than-minimal-dose CS was the most commonly used ITP treatment per line (including repeated use of the same CS). This was true for up to 7L. Monotherapy was more common than combination therapy in all LOTs; TPO-RAs, rituximab, and splenectomy were used much less often than CSs, irrespective of LOT. Limitations of our study include: data variability; claims data only denote receipt of medication, not pt use; variable duration of follow-up after treatment index date, and pts with a shorter follow-up more likely to have fewer LOT variable time elapsed (< 60 d) between lines where pts may have been left untreated; requiring 1 year of enrollment pre‑ITP diagnosis and commercial or private Medicare supplemental health coverage may have altered the pt population; and the changing treatment landscape from 2011 to 2017. Despite these limitations, the results demonstrate that real-world CS use was not low dose or short term. Previous ITP pt surveys found a significant association between the duration of CS use and mean number and severity score of pt-reported side effects. Further education and dissemination of information about other 2L ITP treatments and their outcomes is needed to enable hematologists to follow ITP expert recommendations and make informed treatment decisions to reduce CS overuse in clinical practice.

This abstract has been previously presented at the 25^th Congress of the European Hematology Association (EHA), Virtual Congress, June 11-21, 2020.