Redefining Hyperviscosity in Acute Leukemia: Implications for Red Cell Transfusions

Hyperleukocytosis, most commonly defined as a white blood cell (WBC) count greater than 100,000/μL, is an emergency in acute leukemia, possibly resulting in life-threatening microvascular obstruction, or leukostasis, leading to neurologic (CNS hemorrhage, thrombosis) or pulmonary (respiratory distress, hypoxia) complications. The underlying mechanisms remain poorly understood and are canonically attributed to blood hyperviscosity secondary to high WBC count and abnormal biophysical properties of leukemia cells themselves (leukemia immunophenotype, increased cell size, adhesion, and stiffness). Leukapheresis is a commonly-used therapy for rapid cytoreduction in symptomatic patients, but the procedure carries risk and existing guidelines are supported by scant evidence. Interestingly, despite hematocrit(Hct)/hemoglobin(Hgb) levels being major drivers of blood viscosity due to the high prevalence of circulating red cells (RBCs), how Hct/Hgb mediates hyperviscosity in acute leukemia is unknown. This is clinically important as Hct/Hgb often decrease as leukemic cell counts rise, and acute leukemia patients with anemia are often transfused. While sickle cell disease guidelines advise using a target post transfusion Hct of 30% to minimize iatrogenic hyperviscosity and its morbid complications, no guidelines have been established for acute leukemia. As such, can RBC transfusion actually increase leukostasis risk in acute leukemia?

To explore this question requires new biophysical tools as the complexity of blood viscosity increases substantially at the microvascular level as the physical properties of the cells themselves become the major determinants of resistance to blood flow. To that end, we developed "microvasculature-on-a-chip" devices that recapitulate microvascular biophysical and hemodynamic conditions to investigate how the differing presentations of acute leukemia and transfusion support affect the effective blood viscosity at the microvascular level to cause "in vitro leukostasis."

A multiple-vessel "multiplex" microfluidic device that operates at the appropriate size scale and mimics the microvascular geometry was designed to enable assessing accurate biophysical measurements of blood hyperviscosity. The devices were microfabricated using standard polydimethylsiloxane-based photolithography (Figure 1). Acute B-cell lymphoblastic (B-ALL, 697), acute T-cell lymphoblastic (T-ALL, Jurkat) and acute myelocytic (AML, HL60) leukemia cell lines were maintained via standard cell culture conditions. Patient samples were obtained through our institution's IRB. RBCs from healthy donors were isolated and mixed with leukemia cells to achieve target Hct/Hgb and WBC levels. Various physiologic leukemia "mixtures" were then perfused under physiologic microcirculatory flow conditions through the microvascular device and microchannels occlusion was tracked via videomicroscopy (Figure 2).

Using a standard least squares multivariable linear regression with first and second order effects, microchannel size, Hct/Hgb, WBC count and leukemia cell type all showed statically significant effect on in vitro leukostasis, or microchannel occlusion over time, (all p values < 0.03) (Figure 3). Overall, severe anemia appears to be protective against in vitro leukostasis and there appears to be Hct/Hgb thresholds above which in vitro leukostasis becomes more prevalent, though this is different for B-ALL versus T-ALL. This is in contrast to AML, where severe anemia does not appear to offer protection against in vitro leukostasis as occlusion was appreciated at all Hct/Hgb levels.

These data suggest when determining risk for leukostasis, WBC count and leukemia immunophenotype should not be the sole determinants. Here we show Hct/Hgb levels affect microvascular blood viscosity and risk for microvascular occlusion. These results may impact decisions regarding RBC transfusions and possibly initiation of leukapheresis in asymptomatic patients. Having a model to assess risk associated with RBC transfusions and informing clinicians when a patient might become at risk for leukostasis can have a significant impact on their clinical outcome, morbidity and mortality. Ongoing studies incorporating patient lymphoid and myeloid leukemia cells are needed to support this cell line data.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal