Analysis of Cerebral Vasospasm in a Murine Model of Subarachnoid Hemorrhage with High Frequency Transcranial Duplex Ultrasound

To the best of our knowledge, this study is the first to present a protocol for monitoring of cerebral vasospasm in a murine model of SAH with high frequency transcranial color-coded Duplex ultrasound. We show that this method can measure an increase in intracranial blood flow velocities after SAH induction in mice. In human medicine this phenomenon is well known^3,15. Several clinical studies have shown that elevated blood flow velocities of the large intracranial arteries and an elevated quotient of intra- and extracranial blood flow velocities are a functional consequence of vessel narrowing and correlate with angiographic vasospasm (reviewed in¹⁵). In clinical practice, it is therefore common to use TCD or TCCD for non-invasive bedside monitoring of cerebral vasospasm after SAH^3,15.

DCI is a significant factor influencing neurological outcome after non-traumatic SAH^2,3. As the pathophysiology of DCI is still unclear and effective strategies to prevent and treat DCI are lacking, it is in the focus of clinical and experimental research. Because vasospasm of the cerebral arteries contributes to DCI, many studies evaluate cerebral vasospasm as an endpoint^{5,6,7,8,9,11,12,20}. While formerly large animals were frequently used in experimental studies on SAH, there has been a shift towards small animal models in the last years, particularly to murine models²¹. However, a problem is that imaging methods for cerebral vasospasm used in human medicine cannot be directly transferred to mice and other small animals. Clinical sonography equipment does not yield a sufficient resolution to monitor cerebral vasospasm in mice. There is the possibility of small animal MRI or CT scanning²². However, these methods are cost-intensive and time-consuming. Furthermore, they induce distress in the animals because of the duration of the imaging protocols and contrast application. Moreover, a precise measurement of diameters or volumes of intracranial vessel segments is also limited with these in vivo methods. In SAH studies using mice, it is therefore common to determine the degree of cerebral vasospasm ex vivo^{5,6,7,8,9,11,12,20}. The method presented here is fast, reducing the anesthesia time for the exam to less than 10 minutes, and therefore presumably induces only little distress in the animals. The examination is noninvasive and exhibits a sufficient resolution to visualize and determine the blood flow velocities of large intracranial vessels (ICA and middle cerebral artery). It would therefore be well suited for functional monitoring of cerebral vasospasm in longitudinal studies, examining the same animals at different time points. In studies not requiring histology or other tissue examinations together with the examinations on vasospasm, a longitudinal study design could be used to reduce animal numbers. For future studies focusing on modulation of vasospasm after SAH, determination of blood gases should be performed at the time of the ultrasonographic determinations of cerebral blood flow velocities.

The method shown here contains several critical steps, which should be reviewed in case of methodological problems. It is critical that the body temperature of the animal is kept constant during the whole procedure. Mice quickly develop hypothermia after induction of anesthesia if they are not warmed (e.g., with a heating plate). Hypothermia may alter the results of the measurements. Because of this the ultrasound gel should also be warmed to 37 °C in a water bath before application. Secondly, in order to standardize the measurements, it is necessary that the angle in which the ultrasound probe is applied is constant between the exams. It is therefore necessary to position the animal carefully. The ultrasound probe should not be used free hand but be mounted on a holder with a micromanipulator to allow insonation at a defined position and angle. Furthermore, it is critical to use constant technical settings of the ultrasound device within an experimental series to reduce technical variations. Thirdly, it should be noted that the Duplex examination is not feasible in the time immediately after SAH induction. During this period, an elevated intracranial pressure leads to cerebral hypoperfusion, which limits the application of transcranial Duplex sonography. The Duplex examination of the extracranial carotid artery exposed during the operation for SAH induction may furthermore be impaired by surgical artifacts.

Finally, we want to discuss limitations and future directions of the method presented here. Similar to TCD or TCCD in clinical practice, we cannot directly measure the vessel diameter. An acceleration of the blood flow velocities of cerebral arteries could therefore also be caused by cerebral hyperperfusion. However, clinical studies showed a correlation between an accelerated blood flow velocity and angiographic vasospasm¹⁵. Furthermore, we did not observe cerebral cortical hyperperfusion after SAH induction in the murine model used here¹⁹, and the increase of intracranial blood flow velocities was accompanied by an increase of the quotients of intra- and extracranial blood flow velocities of the ICA, which was reported to indicate vasospasm in a clinical study²³. We therefore assume that the accelerated blood flow velocities also indicate vasospasm in the SAH mouse model, although, as in the clinical application of Doppler ultrasonography, it is not possible to distinguish between vasospasm and cerebral hyperperfusion with hyperdynamic flow. Secondly, functional monitoring of cerebral blood flow velocities only allows conclusions on cerebral vasospasm. Direct imaging and quantification of cerebral perfusion in the context of DCI is not possible. Nevertheless, it should be noted that determination of cerebral perfusion with ultrasonography has been reported in a clinical application²⁴. We therefore speculate that ultrasonographic quantification of cerebral perfusion in mice will become available in the future. A modification of the method in this respect would then allow conclusions not only on vasospasm of the large vessels, but also on microcirculatory disturbances. Thirdly, clinical studies have reported a high investigator dependence of bedside transcranial ultrasonography studies^17,25. However, this is presumably not the case for the experimental application shown here, because of the highly standardized and controlled settings in experimental studies, and because in mice the imaging resolution allowed a clear identification of the vessel segments to be analyzed. Lastly, it is a disadvantage that vasospasm is determined at defined anatomical positions. Vasospasm of neighboring segments could therefore escape evaluation. It should be noted, however, that this problem also arises with other methods determining vasospasm. A measure to reduce errors from this source in future experimental studies would be to determine cerebral blood flow velocities of several intracranial vessel segments.