Search for a command to run...
Abstract Background: Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is a rare and clinically aggressive hematological malignancy, with an estimated incidence of 0.04 per 100,000 in the US (SEER 17 registry 2024). BPDCN was not classified as a distinct myeloid neoplasm until 2016 by the WHO, with reclassification under dendritic cell neoplasms in the 2022 WHO 5th edition (Arber et al., 2016; Khoury et al., 2022). Standardized treatment protocols have yet to be established, and therapeutic options are limited. Currently, tagraxofusp (Tag) is the only FDA-approved therapy for BPDCN (2018), although combination treatment with other antineoplastic agents occurs in clinical practice (Pemmaraju et al., 2023). Real-world (RW) studies of BPDCN are critical to understand the current treatment landscape and to identify areas of unmet need. Here we present patient characteristics, treatment patterns, and safety profiles of current therapeutic regimens using a US-based RW dataset. Method: A retrospective cohort of patients diagnosed with BPDCN (ICD-10 code: C86.4) between 10/1/2015 and 12/31/2024 was created using Optum Market Clarity®, which contains electronic health records (EHRs) linked to closed medical/prescription claims from over 115.4 million patients in the US. Eligible patients had ≥365 days of continuous health plan enrollment or EHR activity prior to BPDCN diagnosis (dx). Pre-defined pharmacologic therapies (PT) of interest included Tag, steroids (topical or systemic), and other antineoplastic agents used in clinical practice for BPDCN. First line of therapy (LoT1) was defined as the earliest initiation of PT of interest (index drug) following BPDCN dx, with combination therapy defined as initiation of any additional therapies ≤30 days of the index drug. Second line of therapy (LoT2) was defined by the initiation of a new agent after the 30-day period or a gap of ≥90 days before reinitiating the same regimen. Adverse events (AEs) were assessed by ICD-10 codes from treatment initiation to occurrence of an event or to earliest censoring event (death or end of follow-up). Results: The study included795 newly diagnosed patients with BPDCN, with a median follow up of 30 months. Patients were predominantly male (59%) and Caucasian (77%) with a median age at dx of 67 years (IQR: 55–76). Overall, 148 (19%) had no prior history of any malignancies, while 53% had prior or concomitant hematologic malignancies (PCHM) (most common: diffuse large B-cell lymphoma, acute myeloid leukemia, myelodysplastic syndrome, and chronic myelomonocytic leukemia) and 29% had prior solid tumor diagnoses (most common: neoplasm of uncertain behavior of skin, prostate, breast, and lung cancers). At baseline, 27% had a National Cancer Institute (NCI) Comorbidity Index score >3, with COPD, diabetes, renal, cardiovascular, and liver disease representing the most frequent individual NCI comorbidities. Median time from BPDCN dx to PT initiation was 85 days (IQR: 23–368). Among LoT1 patients (n=551, 69%), 61% were treated with steroid monotherapy, 4% with Tag (with or without steroids), and 34% with other agents (eg, venetoclax + hypomethylating agents; cyclophosphamide, hydroxydaunorubicin, oncovin, and prednisone [CHOP]; etc.). Median duration of LoT1 was 69 days (IQR: 13–219); of these, 363 (66%) patients proceeded to LoT2 with a median time to next treatment (TTNT) of 205 days. Among LoT2 patients, steroid monotherapy remained the most frequent regimen. Median duration of LoT2 was 77 days (IQR: 14–273) and 221 (61%) patients proceeded from LoT2 to the next LoT with a median TTNT of 219 days. A total of 85 (11%) patients underwent stem cell transplant during follow up. Among pts treated with Tag at any time during follow up (n=49), 18 (37%) developed capillary leak syndrome (CLS) with 14/18 (78%) hospitalized. Median time to CLS onset from BPDCN dx was 38 days (IQR: 19–86). Anemia, thrombocytopenia, neutropenia, and edema were common AEs across LoT1 and LoT2. Conclusions: This RW study highlights the limited treatment options and durability of currently available PTs for BPCDN. Of note, these RW data also show a high percentage of PCHM and solid tumors with BPDCN and should be explored further. While dx codes in claims may misclassify clinical features and AE incidence, application of medication and procedure codes define LoT more accurately than EHR data alone. More novel therapies and standardized treatment approaches remain an urgent priority.