Abstract

Hyperleukocytosis, is generally defined as circulating white blood cell count (WBC) count exceeding 50 to100 × 109/L and mostly occurs in acute leukemias. High WBC count is a clear adverse prognostic factor for overall survival in both pediatric and adult ALL and its presence is associated with adverse survival outcome in AML as well. Furthermore hyperleukocytosis was associated with increased early mortality in a large group of pediatric and adult AML patients. The reason probably reside in the occurrence of blast cell aggregates and WBC thrombi in the capillary beds of the lung, brain, and heart (demonstrated at autopsy) resulting in clinically observed signs and symptoms. Novotny et al. (

Eur J Haematol
2005
:
74
:
501
–510
) developed a clinical score to grade the probability of leukostasis. It is based on the simple and fast clinical evaluation of overall severity of symptoms (slight, marked or severe limitation), the presence of pulmonary symptoms (from no limitation to dispnoea at rest) and the presence of neurologic symptoms (tinnitus, visual disturbances, confusion, somnolence) with assignment of a score: 0: no leukostasis; 1: possible leukostasis; 2: probable leukostasis; 3 highly probable leukostasis. We retrospectively applied the leukostasis grading score (LGS) to patients admitted to our Institution from 1995 to 2008 with a new diagnosis of acute leukemia presenting with hyperleukocytosis (>100 × 109/L) in order to identify patients at high risk of early death (within a week). Every patient was assigned a LGS by 4 blinded independent observer physicians and combined together to obtain a median LGS for each patient. We analyzed relationship between LGS and clinical or laboratory characteristics of patients. Statistical analysis was performed using “GraphPad Prism” GraphPad Software Inc. Five hundred fifty eight patients were admitted to our institution in the above mentioned timeframe with diagnosis of acute leukemia and 33 (6%) presented hyperleukocytosis (median WBC 209 × 109/L, range 121–427 × 109/L), 18 female and 15 male with a median age of 57 years (range 8 – 90). According to the French–American–British (FAB) Group, cases were classified as 4 M1, 7 M2, 6 M4, 7 M5, 4 with unclassified AML and 5 with ALL. On Institutional basis, all patients were submitted to leukapheretic procedure taking in account clinical conditions and laboratory testing (WBC and blasts count). All patients received standard induction chemotherapy immediately after the first leukapheresis procedure was completed. Early death occurred in 6 patients (20%), 2 female and 4 male with a median age of 70 years (range 57–90) affected by AML, 3 with monocyte lineage involvement. LGS was retrospectively applied to 31 out of 33 patients resulting in a median score of 2 (range 0–3). Three patients (9.7%) were assigned a LGS of 0, 4 patients (12.9 %) were assigned a LGS of 1, 14 patients (45.2 %) were assigned a LGS of 2 and 10 patients (32.2 %) were assigned a LGS of 3. Statistical analysis showed a statistical significant positive correlation between age and LGS (r+0.42, p=0.0206) and a lack of correlation between LGS and WBC, %Blasts, LDH at diagnosis. We found a statistically significant difference in LGS according to phenotype when we compare ALL to AML (LGS 1 vs 2, p=0.0136). Sorting out ALL, AML with and AML without monocyte lineage involvement, LGS was statistically higher in AML with monocyte lineage involvement (LGS 1 vs 2, vs 2: p=0.0248). In patients with early death we observed a significantly higher LGS (3 vs 2, p=0.0094). Finally, among clinical and laboratory variables, only age, LDH and LGS were factors statistically associated to the occurrence of early death. In our series >75% of hyperleukocytic patients were assigned a LGS >=2 and almost 50% an LGS of 3. Higher LGS was observed in patients with myeloid phenotype and, particularly with monocyte involvement. The correlation between LGS and immunophenotype together with the lack of correlation with WBC count or %Blasts or LDH reflects the role of the adhesion molecules rather than the tumor mass in the pathogenesis of leukostasis. The score is rapid, simple, clinical-based waiting for immunophenotype, cytogenetic and molecular characterization and is able to identify patients at higher risk of early death immediately requiring more aggressive treatment (leukapheresis every 12 hours, adhesion molecule antagonists).

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author