Real-world use of venetoclax and hypomethylating agents (HMA) in blastic plasmacytoid dendritic cell neoplasm (BPDCN): A multi-institutional case series

Abstract

Abstract Introduction: Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is a rare and aggressive hematologic malignancy originating from plasmacytoid dendritic cell precursors, historically associated with poor patient outcomes. Due to the rarity of BPDCN, there is limited data to guide therapeutic strategies. Tagraxofusp-erzs (formerly SL-401) is the only FDA-approved therapy for BPDCN and is frequently used as frontline treatment. Venetoclax combined with hypomethylating agents (Ven+HMA) is approved for frontline treatment of acute myeloid leukemia (AML) in individuals >75 years of age and is employed off-label for BPDCN, particularly in older or transplant-ineligible patients. Despite some case reports in the literature, the clinical outcomes and optimal use of Ven+HMA in BPDCN remain poorly characterized. Methods: We performed a retrospective, multi-institutional case series of BPDCN patients treated with Ven+HMA at Dana-Farber Cancer Institute and University of Miami Sylvester Comprehensive Cancer Center between 2009–2024. Clinical, genomic, and treatment data were extracted from medical records. Outcomes included response rates, duration of response (DOR), relapse patterns, and overall survival (OS). Descriptive statistics were used. All patients were treated off-protocol, outside of a clinical trial setting. Results: Fourteen patients with BPDCN received Ven+HMA in any line of therapy. The median age at diagnosis was 64 years (range, 14–84) and 93% were male. Most patients (86%) had isolated BPDCN, while 14% had concurrent myelodysplastic syndrome. ECOG performance status was 0 in 71%, 1 in 21%, and 2 in 7%. Median hemoglobin was 14.1 g/dL, platelet count 106 ×10⁹/L, and WBC 4.06 ×10⁹/L (max 44.3 ×10⁹/L). Peripheral blasts were generally absent. Bone marrow involvement was variable, with 45% (5/11) having ≥5% (range 0-95). Among 10 patients with sequencing data at diagnosis, the most frequent mutations were TET2 (60%), ASXL1 (40%), JAK2 and NRAS (30% each), and DNMT3A, IKZF1, and ZRSR2 (20% each). Venetoclax was administered as first-line (1L) therapy in 4 patients (29%) and in the relapsed/refractory (R/R) setting in 10 patients (71%). Specifically, 2L in 4, 3L in 2, and in subsequent lines (4L–7L) in one patient each. In the 1L group, 75% achieved complete response (CR) and 25% had progressive disease (PD). In 2L, 50% had CR, 25% had partial response (PR), and 25% had PD. Among 3L patients, one achieved CR and one had PD. In later lines, one each achieved CR (5L and 6L), PR (4L), and PD (7L). To assess venetoclax efficacy by line of therapy, we performed a Fisher's exact test comparing CR rates between 1L and R/R groups (3/4 vs. 5/10, p = 0.58), suggesting venetoclax retains activity in later lines and may be deferred to the R/R setting in select patients. The overall response rate (ORR) in the R/R setting was 70%. Overall median time to response was 59 days (range, 10–159), and median DOR for patients who responded was 263 days (range, 126–760). For venetoclax 1L, the median DOR was 224 days (n=3, range, 209-295). For patients treated 2L+, the median DOR was 363 days (n=7, range, 126-760). Allogeneic hematopoietic stem cell transplantation (allo-HSCT) was performed in 50% of patients (n=7); however, only two patients received allo-HSCT following Ven+HMA. Relapse occurred in 93% (n=13), frequently with central nervous system (CNS, 4) or extramedullary disease (EMD, 4). The median OS was 27 months (range, 3–129), with 1, 2, and 5-year OS rates of 77%, 50%, and 14%, respectively. For patients treated with Ven+HMA in R/R setting, the median OS was 10.5 months. Conclusions: Ven+HMA shows encouraging activity in BPDCN, including in R/R cases, with a high ORR similar to Tagraxofusp in the R/R setting. However, relapses involving CNS or EMD were common, highlighting the need for improved CNS surveillance and prophylaxis. Although limited by its retrospective design and small sample size, these real-world findings support continued evaluation of Ven+HMA, particularly in frontline and combination regimens. Notably, CR rates were comparable between 1L and 2L+ settings (Fisher's exact p = 0.58), suggesting that Ven+HMA may retain meaningful efficacy in the R/R context. Prospective, multi-center trials are needed to confirm its efficacy compared to established therapies and determine the optimal sequencing of Ven+HMA with Tagraxofusp and allo-HSCT.