# Impact of Leukapheresis and Biological Risk Markers on Early Mortality in Patients with Hyperleukocytic Acute Myeloid Leukemia

## TL;DR

This study examines whether leukapheresis improves survival in patients with high white blood cell counts in acute myeloid leukemia and finds it does not significantly reduce early mortality.

## Contribution

The study provides new insights into the limited impact of leukapheresis on early mortality and identifies specific biological markers that predict outcomes in hyperleukocytic AML.

## Key findings

- Leukapheresis did not significantly reduce 15-, 30-, or 90-day mortality in hyperleukocytic AML patients.
- Elevated LDH, TLS, DIC, and ECOG PS ≥2 were strong predictors of early death across all patients.
- CD25 positivity and PB blast percentage were independent predictors of poorer overall survival.

## Abstract

Background and Objectives: Hyperleukocytosis in acute myeloid leukemia (AML) is life-threatening, often complicated by leukostasis, tumor lysis syndrome (TLS), and disseminated intravascular coagulation (DIC), with very high early mortality. Leukapheresis (LA) can rapidly reduce circulating blast burden, but its effect on survival and prognostic relevance of disease markers remains unclear. Materials and Methods: We retrospectively analyzed 74 adult AML patients with WBC > 100 × 109/L treated at the University Clinical Center of Serbia between 2014 and 2024: 28 received LA plus cytoreduction (LA group), and 46 received cytoreduction alone (non-LA group). We evaluated 15-, 30-, and 90-day mortality and overall survival (OS), and assessed clinical, laboratory, and immunophenotypic predictors using Cox regression, with separate subgroup analyses. Results: Patients in the LA group had significantly higher baseline leukocyte counts and LDH (p = 0.18 and p = 0.024, respectively). Although LA resulted in a median 34% reduction in WBC, there was no statistically significant difference in early mortality: 15-day survival was 68% vs. 76% (HR 0.70, p = 0.423), 30-day survival 50% vs. 65% (HR 0.62, p = 0.197), and 90-day survival 39.3% vs. 41.3% (HR 0.85, p = 0.604). Median OS was similarly poor, about 1 month in the LA group compared to 2 months in the non-LA (HR 0.73). Across all patients, ECOG PS ≥2, elevated LDH, TLS, and DIC were the strongest indicators of early death. In the LA group, elevated LDH and increased peripheral blood (PB) monocyte count predicted 15-day mortality (p = 0.021 and p = 0.031, respectively), but lost significance by day 90. In non-LA patients, CD25 positivity (p = 0.034) and DIC (p = 0.045) predicted 15-day death. By day 90, CD25 expression (p = 0.048) remained prognostic, while PB blast percentage (p = 0.045) and PB monocyte count (p = 0.017) emerged as additional adverse prognostic predictors in the non-LA group. In multivariate analysis, higher PB blast percentage, CD25 positivity, and ECOG PS ≥ 2 independently predicted poorer OS. Conclusions: Although LA did not reduce early mortality in the entire cohort, the loss of prognostic significance of elevated LDH, high PB blast percentage, PB monocyte burden, and CD25 expression in the LA group may suggest that the intervention can attenuate the impact of biologically aggressive disease.