Paper Details

ABSTRACT

Introduction

Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is an orphan, ultra-rare aggressive hematologic malignancy derived from plasmacytoid dendritic cell precursors. Therapeutic options include one of two FDA-approved CD123-directed therapies (tagraxofusp and pivekimab sunirine), ALL, AML or lymphoma chemotherapy (chemo) regimens. While the reported median survival is approximately 18 months, outcomes are heterogenous. Small studies have identified prognostic features of BPDCN, but there are no validated prognostic models. Previously, we generated a prognostic scoring system in a single center cohort based on age and disease features: BPDCN blasts ≥5% in bone marrow and presence of signaling pathway mutations (NRAS, KRAS, or FLT3) (Shimony et al. ASH 2024). Here, we validate this scoring system in an international cohort and evaluate its prognostic value in various treatment contexts.

Methods

Data from BPDCN patients were collected from over 15 academic centers in the US (DFCI, MDACC) France, and the international PETHEMA group. Based on our previous work, a risk score was assigned to each patient: favorable (age < 43 years), intermediate (age ≥ 43 years and no overt disease; or overt disease with absence of signaling mutations), and adverse (all others). Subsequently, the association of the risk score with overall (OS) and event free survival (EFS), upfront treatment and allogenic hematopoietic stem cell transplantation (HCT) was evaluated. Comparisons were made by Fisher’s exact (categorical variables) and Wilcoxon rank-sum (continuous variables) tests. Time to event analyses (OS and EFS, both presented as point estimates at 2 years) were calculated using the Kaplan-Meier method and comparisons made by log-rank test. A C-statistic was calculated by Harrell’s/Concordance index tests. Univariable and multivariable Cox regression models were fitted for OS, with the final multivariable model chosen by backward selection by AIC. HCT was included as a time-varying covariate.

Results

The entire cohort included 241 patients with newly diagnosed BPDCN (DFCI derivation cohort n=66, validation cohort of all other patients n=175) with a median age of 65 years (IQR 54-74). Most (90%) had ECOG of 0-1. Overt bone marrow disease was common (n=172, 71%) and 30% (n=48) had signaling pathway mutations. Initial therapy was tagraxofusp (n=83, 34%), ALL chemo (n=61, 25%), AML chemo (n=57, 24%) or other therapies (n=40, 17%). Almost half (n=108, 45%) were consolidated with HCT.

In the DFCI cohort (previously presented), OS differed significantly between the 3 risk groups (p<0.001): 2-year OS favorable 100% (95% CI 100-100%); intermediate 47% (95% CI 28-66%); and adverse 5% (95% CI 0-16%). In the validation cohort, we also saw significant differences in OS by risk group (p<0.001): 2-year OS favorable 70% (95% CI 52-87%); intermediate 38% (95% CI 26-51%); and adverse 25% (95% CI 14-37%). Finally, in the combined cohort, OS was different by risk group (p<0.001): 2-year OS favorable 77% (95% CI 63-91%); intermediate 41% (95% CI 31-52%); and adverse 20% (95% CI 11-29%). The C-statistic of the model (combined cohort) was 0.63. Although the model was based on OS, it also differentiated EFS by risk group (p<0.001): 2-year EFS favorable 63% (95% CI 47-79%); intermediate 29% (95% CI 20-39%), and adverse 13% (95% CI 6-21%).

After validating our prognostic model, now referred to as the BPDCN International Prognostic Scoring System (BIPSS), we next determined its prognostic significance for OS when adjusting for front-line treatment and HCT. In univariable analysis, BIPSS, ECOG, initial treatment and HCT were significant. However, in multivariable analysis only BIPSS (intermediate=reference, favorable HR 0.25 [95% CI 0.12-0.50], p<0.001; adverse HR 1.56 [95% CI 1.08-2.26], p=0.018), ECOG (2-3 vs. 0-1 HR 2.83 [95% CI 1.70-4.71], p<0.001) and HCT (HR 0.57 [95% CI 0.37-0.87], p=0.0085) retained prognostic significance for OS, while treatment was not retained.

Conclusions

This study validates a new prognostic model – BIPSS – in an international cohort of patients with BPDCN receiving heterogenous upfront therapy. In a multivariable analysis of OS, BIPSS and HCT were independently prognostic while initial therapy was not. In clinical practice BIPSS is applicable to guide clinical discussions. The BIPSS can be integrated into future BPDCN studies to facilitate cross-cohort analyses and to test risk-adapted treatment strategies.