Organ Involvement in BPDCN Is Associated with Sun Exposure History, TET2 and RAS Mutations, and Survival

Introduction

Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is a rare, aggressive hematologic neoplasm that can involve skin, bone marrow (BM), central nervous system (CNS), and non-CNS extramedullary disease (EMD) sites. Recent findings show that ultraviolet light (UV) exposure may contribute to the development of BPDCN (Griffin, Nature 2023). Conflicting data exist as to whether anatomic involvement is clinically relevant, and genetics have not been incorporated into disease site prognostic models. In addition, the role of non-overt (<5%, microscopic) BM involvement remains unclear as trials have only regarded ≥5% as evidence of BPDCN.

Methods

We assessed associations between organ involvement, UV exposure, molecular characteristics, and outcome in patients with biopsy-confirmed BPDCN seen at Dana-Farber Cancer Institute between 2006-2022. UV exposure was classified as high vs. low by an independent reviewer blinded to outcomes, based on medical, social, and occupational history. We classified 3 groups at diagnosis: systemic without skin (either BM with ≥5% BPDCN cells or EMD and no skin involvement), skin only (with no overt BM [<5% BPDCN cells] and no EMD), or skin and systemic; groups were classified irrespective of CNS involvement at diagnosis. Next generation sequencing was performed on blood/BM at diagnosis. Comparisons were performed by Fisher’s exact test. Survival was calculated from treatment initiation and comparisons made by the log-rank test. Uni- and multivariable Cox regression models were fit for association with overall survival (OS), with SCT as a time varying covariate.

Results

Overall, we included 66 consecutive patients with a median age at diagnosis of 68 years (interquartile range [IQR] 61-76). The majority (n= 57, 86%) of patients were male. Disease distribution at diagnosis involved skin (n=52, 79%), BM (n=50, 76% [39 (59%) with ≥5% BPDCN cells]) and EMD (n=30, 45%), (Table, Figure).

The median time from first symptom to diagnosis was 61 days (IQR 35-134) and from diagnosis to treatment was 30 days (IQR 20-46). The first therapeutic regimen was tagraxofusp (n=30, 45%), AML (n=12, 18%) or ALL (n=12, 18%) chemotherapy, or other treatment (n=12, 18%). Complete remission (CR) after first treatment was documented in 35 (57%) with 20/35 consolidated in CR1 with SCT (57%); however, after a median of 6.1 months from response assessment (IQR 2.4-8.1 months), 22/35 who achieved CR relapsed, with skin (n=13, 54%), BM (n=12, 50%), EMD (n=5, 21%), or CNS (n=3, 13%) as relapse sites. The median number of therapies was 2 (0-6), and 35 (53%) had SCT at any time. The pattern of post-SCT relapse included disease involvement in skin (8/14, 57%), BM (5/14, 36%), EMD (3/14, 21%) and CNS (3/14, 21%).

Patients with skin only at diagnosis (n=19) vs. any systemic involvement (n=47) were more frequently ≥75 years (9/19 [47%] vs. 9/47 [9%], p=0.032), had lower rates of complex karyotype (0 vs. 12 [32%], p=0.022) and activated signaling mutations (N/KRAS, FLT3, CBL, CKIT, SH2B3; 1 [8%] vs. 14 [40%], p=0.04), mainly driven by lower NRAS (0 in skin only vs. 10 [29%] with any systemic, p=0.044, Figure). In contrast, those with any skin involvement had higher rates of UV exposure (27 [59%] vs. 3 [23%], p=0.03) and TET2 mutations (28 [72%] vs. 3 [33%], p=0.051) compared to those with no skin involvement, consistent with the recently aforementioned link between UV exposure and selection for TET2 mutated plasmacytoid dendritic cells.

The median OS of the entire group was 18.2 months (95% CI 13-24) and did not statistically differ between presentation groups (skin only: 23.5 [95% CI 13-NR]; systemic + skin 20.4 [95% CI 11-26]; systemic only 17.5 [95% CI 9-22], p=0.14). In a multivariable analysis, age <60 (HR 0.35 [95% CI 0.13-0.94], p=0.038) was associated with improved OS, whereas BM involvement ≥5% (but not microscopic involvement) was independently associated with worse OS (HR 2.65 [95% CI 1.1-6.36]).

Conclusions

Specific organ involvement in BPDCN is associated with UV exposure, age, cytogenetic and molecular characteristics. Skin is the most common relapse site post CR or SCT. Survival is better in younger patients (age <60 years), and worse in those with ≥5% BM involvement after adjustment for age and SCT. These data establish additional factors that, in combination, have prognostic value in BPDCN, and provide independent evidence that sun exposure influences BPDCN genetics and clinical presentation.