Predicting Vaso-Occlusion in Sickle Cell Disease By Microfluidics and Modeling

Sickle cell disease (SCD) affects approximately 100,000 Americans and occurs in 1 out of every 365 African-American births. One of the most debilitating symptoms of SCD is pain crisis caused by sickled red blood cells clotting microvasculature, which is termed vaso-occlusion. There are currently no methods to monitor or predict pain crises. This study aims to design a method by applying a microfluidic device and modeling system to predict vaso-occlusion events in SCD.

The study was designed to incorporate experimental and modeling approaches to predict the clotting of the microvasculature. The microfluidic device monitors and records the passage of healthy and sickled red blood cells through pores that mimic the smallest human microvasculature. Blood samples were obtained from the sickle cell patients. Using this approach, the migration of normal and sickled red blood cells through the pores of the microfluidic device was monitored, and the passage time was recorded.

The recorded data was integrated with a mathematical model of the passage of sickle cells through microvasculature to predict the clotting under various conditions. Specifically, the images of blood smear from the blood samples were used to calculate the statistical distributions of different types of sickle cells and construct computational model. The cells were modeled using a multiscale cell model based on the finite element method coupled with a boundary integral model of the surrounding flows. Pressures at the physiological conditions in various microvasculatures were applied to push the cells to pass through the pores. The altered cell shape and mechanical properties of different types of sickle cells were incorporated into the multiscale model to distinct them from normal red blood cells. After calibrating the passage time against measurements in the microfluidic experiments, the probability of vaso-occlusions was predicted based on the measured statistical distributions of sickle cells and dimensions of microvasculatures.

We found that sickled red blood cells could pass through the pores in the microfluidics. However, the dynamic and quantity of the migration were significantly compromised. This is closely related to the percentages of different types of sickled red cells, reflecting the severity of patients' disease. Through the changes in the size of the microvasculatures, the possibility of clotting of various vasculature in the body can be predicted using this model with patient-specific blood smear images. In this respect, we analyzed 72 blood smears from 24 patients with sickle cell disease. The percentage of the sickle cells out of total red blood cells were quantified in each smear using a computational approach. Indeed, the frequency of sickled red cells is positively correlated with the frequency of vaso-occlusion events in these patients. We conclude that this novel design of the vaso-occlusion prediction system could be useful in predicting the clotting and pain crisis of SCD patients.

No relevant conflicts of interest to declare.