A local-saturation-and-delay MRI method for evaluation of red blood cells aggregation in vivo for tumor-bearing or drug-used rats

Abstract

Hyperviscosity syndrome (HVS) is a combination of clinical signs and symptoms related to increased blood viscosity. HVS can increase the thrombotic risk by causing a major disturbance to the blood flow, which is usually found in the advanced stages of the tumor. Moreover, some of the drugs used in chemotherapy, such as 5-fluorouracil and erythropoietin, are also capable of causing HVS through their respective pathways. Clinically, the viscosity of a patient's blood sample is measured by a rotary rheometer to estimate the risk of hyperviscosity syndrome. However, the measurement of blood viscosity in vitro is easily affected by storage time, storage environment, and anticoagulants. In addition, the fluid conditions in the rheometer are quite different from those in natural blood vessels, making this method inappropriate for evaluating blood viscosity and its effects in vivo under physiological condition. Herein, we presented a novel magnetic resonance imaging method called local-saturation-and-delay imaging (LSDI). The radial distributions of flow velocity measured by LSDI are consistent with the Ultrasonic (US) method (Spearman correlation coefficient r = 0.990). But the result of LSDI is more stable than US (p < 0.0001). With the LSDI method, we can directly measure the radial distribution of diastolic flow velocity, and further use these data to calculate the whole blood relative viscosity (WBRV) and erythrocyte aggregation trend. It was a strong correlation between the results measured by LSDI and rotary rheometer in the group of rats given erythropoietin. Furthermore, experimental results in glioma rats indicate that LSDI is equivalent to a rheometer as a method for predicting the risk of hyperviscosity syndrome. Therefore, LSDI, as a non-invasive method, can effectively follow the changes in WBRV in rats and avoid the effect of blood sampling during the experiment on the results. In conclusion, LSDI is expected to become a novel method for real-time in vivo recognition of the cancer progression and the influence of drugs on blood viscosity and RBC aggregation.

Keywords: hyperviscosity syndrome; local-saturation-and-delay imaging; magnetic resonance imaging; sprague-dawley rats; whole blood relative viscosity.

FIGURE 1

Schematic diagram of LSDI applied to the left common carotid artery of SD rats. The essence of the LSDI method is that the partially overlap of TOF-MRA and T₁-weight FLASH in the field of view, and the path of blood flow can be visualized by adjusting the delay time. This method performs a T₁-weight FLASH scan outside the saturated slice to observe the vascular geometry upstream and downstream of the region of interest and ensures that the saturated slice is perpendicular to the blood vessel. At the same time, a TOF scan with a specified delay time is performed in the saturated slice to observe the displacement of the blood flow at different radial positions and eventually obtain the flow velocity distribution.

FIGURE 2

Workflow for extracting WBRV from LSDI images.

FIGURE 3

The timeline of the experiments (A). Normalized blood flow velocity profiles (measured by LSDI) of control group (B) and EPO group (C). The profiles of the EPO group showed obvious passivation on day 11. The WBRV (measured by LSDI) of NC and EPO group of animals (D). The WBRV in EPO group increased significantly (**: p < 0.01). The correlation between the WBRV of the two group measured by LSDI and rheometer on day 11 (E). There is a strong correlation between the results measured by two methods. (Spearman r = 0.685, p < 0.05).

FIGURE 4

Normalized blood flow velocity profiles measured by LSDI and ultrasonic (A). Ultrasonic imaging of blood flow in CCA scanned by color Doppler flow imaging mode (B). LSDI imaging of blood flow in CCA (C). The correlation between the two groups of datapoints measured by LSDI and US (D). The velocity profiles obtained by the two methods are consistent (Spearman r = 0.990, p < 0.0001). Standard deviation distribution of datapoints in the two methods (E), ****: p < 0.0001).

FIGURE 5

T₂WI for the brain in C6 rat (A) and NC rat (B). The rats’ WBRV (measured in the left common carotid artery by LSDI) of NC and C6 groups (C), ^*: p < 0.05).