Continuous intravertebral injection of fasudil hydrochloride in the treatment of cerebral vasospasm
Background and Objective : Cerebral vasospasm is a serious complication of subarachnoid hemorrhage (SAH) and is associated with clinical deterioration and mortality. The objective of this study is to investigate the effect of continuous intravertebral artery (cIVA) injection of fasudil hydrochloride on delayed cerebral vasospasm (CVS). Material and Methods : Forty white rabbits were alloted into groups: (i) seven-day (cIVA injection of fasudil hydrochloride for seven days after injection of blood) group, (ii) five-day (cIVA injection of fasudil hydrochloride for five days from the third day after injection of blood) group, (iii) intravenous treatment (intravenous infusion of fasudil hydrochloride after the first blood injection twice a day) group, and (iv) control group. All the rabbits in all the four groups underwent selective vertebrobasilar angiography. The pathological changes in the basal artery were observed by light and electron microscopy. Fasudil hydrochloride injection (2ml:30mg) was provided by Tianjin Chase Sun Pharmaceutical Company Limited. Results : Severe CVS occurred after subarachnoid hemorrhage (SAH) in the control group, whereas, it was significantly lower after intravertebral artery and intravenous injection of fasudil hydrochloride. The difference between the intravenous and intravertebral artery groups was statistically significant on the seventh day. Conclusions : The effect of cIVA injection of fasudil hydrochloride in treating delayed CVS due to SAH at different time points was better than intravenous administration.
Keywords: Cerebral vasospasm, continuous intravertebral artery injection, fasudil hydrochloride, subarachnoid hemorrhage
Cerebral vasospasm (CVS) is a serious complication of subarachnoid hemorrhage (SAH) and is associated with in clinical deterioration and mortality. ,, Though administration of vasodilators, such as nimodipine and fasudil, have been used to prevent CVS,  the efficacy of intravenous delivery is questionable. ,, This study in the experimental animal models of SAH is to assess and compare the efficacy of continuous intravertebral artery (cIVA) injection of fasudil hydrochloride versus intravenous administration in reversing SAH-induced delayed CVS at different time points.
Establishment of animal model of CVS
The animal model of CVS was established by injecting blood into the cisterna magna using the two-injection method. , All animals underwent the first such injection 30 minutes after selective vertebrobasilar artery angiography, on the first day; the rate of injection was 0.5 ml/min and the total amount was 0.3 ml/kg. After 48 hours, the second blood injection was repeated; the blood was taken from the central artery of the rabbits' ears and no anticoagulant was added. After injection, the rabbits were put in a 30-degree, head-down position for 10 minutes, so that the blood get distributed throughout the subarachnoid space. Whether the non-anticoagulant blood spread throughout the subarachnoid space depended on the cerebrospinal fluid outflow from the needle casing during puncture; thus, the results were reliable. 
Experimental animals and grouping
This study was approved by the Ethics Committee of Wuhan General Hospital, Guangzhou Command. A total of 40 healthy Japanese white rabbits (provided by the Experimental Animal Center of the Wuhan General Hospital of Guangzhou Military Command), weighing between 2.5 to 3.0 kg, were randomly divided into four groups: (i) seven-day group -- cIVA injection of fasudil hydrochloride was conducted continually after the first blood injection for seven days (1 mg/kg/24 hours); (ii) five-day group -- cIVA injection of fasudil hydrochloride (1 mg/kg/24 hours) was conducted on the thrid day after the first blood injection and continued for five days (one day after the second blood injection); (iii) intravenous treatment (iv) group -- fasudil hydrochloride given intraveously after the first blood injection (0.5 mg/kg, bid); and (iv) control group -- CIV injection of 5% glucose solution was performed after the first blood injection. Fasudil Hydrochloride Injection (2ml:30mg) was provided by the Tianjin Chase Sun Pharmaceutical Company Limited.
Selective vertebrobasilar artery angiography and catheter technology
Selective vertebrobasilar artery angiography was performed using the femoral approach in the all animals before the first injection of blood on the fifth day and seventh day.  After the rabbits were anesthetized in the supine position under sterile conditions, an oblique incision, 3 - 4 cm long, was made in the right groin, and 2 cm of the right femoral artery was exposed and separated. The distal end was ligated with silk ligatures, and blood flow was blocked temporarily with an artery clamp at the proximal end. Ophthalmic scissors were used to make a small oblique cut in the femoral artery at the distal end of the artery clamp, and an ultra-smooth guide wire (Terumo, Japan), connected to a 4F catheter sheath, was led into the femoral artery. Subsequently, the artery clamp was released, while the 4F catheter sheath was delivered into the femoral artery. With the help of a 'seeker', an Echelon-10 microcatheter, supported by a Silverspeed-10 micro-guide wire (Micro Theraprutics, USA), was led into the 4F sheath. By X-ray fluoroscopy, the microcatheter was introduced into the left subclavian artery through the iliac artery, abdominal aorta, descending aorta, and the aortic arch, and 0.2 ml of nonionic contrast media was injected to determine the course of the left vertebral artery. The microcatheter was inserted into the opening of the left vertebral artery, and 0.2 ml of contrast media was injected to produce a road map (tracing chart). Then, the microcatheter cooperated with the micro-guide wire under the guidance of the road map. The micro-guide wire was selectively inserted into the left vertebral artery of the extra vertebrae (V1) segment. Next, the microcatheter was slid into the V1 segment along with the micro-guide wire, and together, the microcatheter and the micro-guide wire slid into the left vertebral artery of the intervertebral foramen (V2 segment). Later, the micro-guide wire was pulled out and 0.5 ml of heparin saline (25 u/ml of heparin) was injected into the microcatheter, and 1.0 ml of contrast media was injected using a 1 ml syringe at low pressure. Finally, vertebrobasilar artery angiography was performed at three frames per second, lasting for eight seconds. 
In addition to the animals in the intravenous treatment group, indwelling microcatheters were placed in the left vertebral artery of the animals in the other three groups after angiography, to prepare for the connection of a micro-injection pump, to inject the drugs. The specific procedures were as follows: Under X-ray fluoroscopy, the microcatheter in the vertebral artery was kept in a fixed position and the 4F catheter sheath was withdrawn. The femoral artery was ligated with silk ligation on a microcatheter surface and the incision was sutured in layers. The microcatheter was fixed in the animal thigh with skin sutures, covered with dressing, and fixed. A three-way tee switch connector was connected to the end of the microcatheter for drug delivery.
The rabbits' heart rate, blood pressure, respiration, and electrocardiogram (ECG) were monitored using a physiological recorder and an ECG instrument, during cerebral angiography. After blood injection, the rabbits were fed in a special well-ventilated cage frame and given pellet feed and tap water; room temperature was kept at 18 - 20C. The neurological function was checked daily before and after blood injection and classified into four levels, according to Endo standards. 
Measurement of basilar artery diameter
The double-blind method was used to measure the basilar artery (BA) diameter before blood injection on the fifth and seventh days, during selective vertebrobasilar artery angiography. In addition, HPIAS-1000 (Qian Ping Imaging Technology Co., Ltd., Wuhan) was used to measure the vessel diameter of the upper, middle, and lower segments of the BA - averaged as the BA diameter. The BA diameter before blood injection was taken as the baseline value and the degree of vasospasm was expressed as the (basic value - spastic value) / basic value Χ 100%. 
Color Doppler ultrasound detection of BA blood flow spectrum and velocity
The SEQUOZA-512 color ultrasonic Doppler (German SIEMENS) was used; the spectrum scale was centimeter per second (cm/s). A color ultrasound probe at 2 mHz was placed in the foramen magnum of the rabbits. The spectrum shape of the BAs in each group before blood injection and on the fifth and seventh days after blood injection was obtained and the peak blood flow velocity in each group was measured and recorded.
On the seventh day after angiography, the rabbits were sacrificed by infusion. After rapid craniotomy the BA was exposed, and sections were obtained after washing it thrice with normal saline. The BAs were fixed in 2.5% glutaraldehyde. Light and electron microscopy was performed to determine the structure of the BA wall.
The data were analyzed with the statistical software SPSS 13.0. Continuous variables were expressed as mean ± standard deviation. Variance analysis was used for multi-group comparisons. Statistical significance was set at P < 0.05.
Mortality and neurological status
Most rabbits showed symptoms immediately after blood injection, which included opisthotonos, neck stiffness, horizontal nystagmus, and rapid breathing, which eased gradually after two-to-three hours. One rabbit each from five-day, seven-day, and control groups and two rabbits from the intravenous group died within 24 hours of the first injection. Two rabbits each from five-day, seven-day, and intravenous groups, and one rabbit from the control group died within 12 hours after the second blood injection. The total mortality rate was 30%. After the first blood injection, the rabbits presented with apathetic mood, reduced locomotor activity and sleepiness, decreased feeding, and drinking, and occasional neck stiffness. These symptoms aggravated significantly after the second blood injection.
A total of 84 cerebral angiographies were performed on the 28 rabbits that survived. Angiography and measurements of percentage change in diameter are shown in [Table 1]. No significant difference was found in the baseline vessel diameter between the three groups. Significant BA vasospasm occurred in the control group compared to the five-day (27.06 ± 3.34%) and seven-day groups (24.98 ± 4.19%). In all groups, both arterial and venous injection of fasudil hydrochloride had a significant effect in preventing BA spasm induced by SAH. On the fifth day after injection, BA vasospasm in the intravenous treatment group (15.43 ± 4.82%) did not differ significantly when compared with the five-day group (10.37 ± 6.92%), but was significantly different from the seven-day group (2.02 ± 6.21%, n = 7) [Figure 1]. On the seventh day after injection, no such difference was seen between the five-day (2.60 ± 6.42%) and seven-day groups (0.44 ± 5.12%), but both groups had a significant difference when compared with the intravenous group (8.40 ± 4.28%) [Figure 2] and [Figure 3].
Color Doppler ultrasound detection
[Table 2] shows the changes in peak blood flow velocity (Vm) in the BA of each group before blood injection and on the fifth and seventh days after blood injection. No significant difference was seen in the baseline peak blood flow velocity between the three groups before blood injection. The peak blood flow velocity (Vm) (cm/s) in the control group increased significantly after blood injection (five-day group: 52.20 ± 1.09; seven-day group: 51.14 ± 1.40). In all the groups, both arterial and venous injection of fasudil hydrochloride had significant effects in preventing the basilar artery spasm induced by SAH. The difference in Vm on the fivth day was not significant between the intravenous (48.46 ± 3.49) and five-day groups (46.93 ± 3.10), but was significant between the intravenous and seven-day groups (40.70 ± 4.22). On the seventh day, there was no difference in Vm between the five-day and seven-day groups (41.31 ± 4.3; 38.53 ± 4.94, respectively), but both the groups differed significantly compared to the intravenous group (46.74 ± 2.83) [Figure 2] and [Figure 3].
Grossly, the specimens in each group generally showed blood clots, distributed primarily under the brain, covering the surface of the basilar artery. Light microscope examination of the brain tissue of the animals in the study group showed an absence of bleeding, nerve cell degeneration, and necrosis. The brain tissue of the control group showed basilar artery spasm and wall thickening. Swelling of the endothelial cells and shedding and shrinkage of the elastic intima were also observed. The intermittent, basement membrane became thickened and the endothelial cell separated. Cytoplasm in the smooth muscle cells expanded, leading to vacuolation. Partial necrosis, membrane hyperplasia, and inflammatory cell infiltration were seen. The changes in the intravenous group were between the changes those of the arterial treatment and control groups [Figure 4].
Under a transmission electron microscope, the brain cell structures of the animals in the arterial and IV groups were intact. No swelling or destruction was observed in the mitochondria or other organelles, and the lamellar structure of the myelin was intact. In the control group, the elastic intima and layers in the BA had shrunk. Degeneration and necrosis of smooth muscle cells between the layers was observed. The endothelial cells degenerated and the mitochondria within the cytoplasm were swollen. Pyknosis of cell nuclei, obvious changes in the media close to the luminal side, smooth muscle cell degeneration, and partial necrosis were observed. The ultrastructures changed slightly in each arterial treatment group. The changes in the intravenous group fell between those of the arterial treatment and control groups [Figure 5].
Cerebral vasospasm is one of the main causes of disability and mortality in patients with aneurysmal SAH. ,, Nimodipine, a calcium channel blocker, has been shown to improve the clinical outcome in patients with SAH by preventing vasospasm. However, a quarter of the SAH patients develop severe symptoms of delayed cerebral ischemia even after treatment with nimodipine.  Nicardipine, another calcium channel blocker, can also reduce the frequency of symptomatic CVS. In a study of 365 patients with aneurysms who were randomized into two different doses of nicardipine versus placebo, delayed cerebral ischemia occurred in 31% of the patients in both the control group and the treatment group. 
In the absence of a clear benefit from intravenous drug delivery, balloon angioplasty and intra-arterial drug delivery provide alternative treatment modalities to treat CVS. ,, Because of the risk of arterial rupture and the need for special techniques, balloon angioplasty can only be used in larger arteries in patients with localized CVS. ,, Application of intra-arterial papaverine has been shown to have some beneficial effects; however, recent reports indicate a rather poor long-term efficacy. , Apart from that, intra-arterial delivery of papaverine may lead to complications, such as, pupillary changes, epilepsy, and respiratory depression. ,
Intravenous administration of fasudil hydrochloride, a protein kinase inhibitor, can effectively prevent vasospasm after SAH. , Fasudil hydrochloride and its active metabolite hydroxyfasudil act by inhibiting Rho kinase. Rho kinase is involved in myosin light chain phosphorylation during the final stage of smooth muscle contraction.  Thus, inhibition of Rho kinase by fasudil hydrochloride results in vasodilatation. Fasudil hydrochloride is usually injected intravenously (30 mg, Q8h).
In 1999, Tachibana et al. reported for the first time that a symptomatic vasospasm can be treated successfully with an intra-arterial injection of fasudil hydrochloride.  In a recent study of 23 patients with symptomatic vasospasm, who were treated with an intra-arterial injection of fasudil hydrochloride, improvement in vasospasm was observed in 100% of patients, as shown by angiography, and good recovery was observed in 65.3% of patients at three months. Furthermore, an intra-arterial injection of fasudil hydrochloride was not associated with complications such as pupillary changes, epilepsy or respiratory depression.  Intra-arterial injection of fasudil hydrochloride was associated with significant dilatation of the spastic cerebral blood vessels as well as improvement in the clinical symptoms. Post-treatment angiography failed to show any vessel complications. These results indicate that intra-arterial injection of fasudil hydrochloride may be a new, safe, and effective method for the treatment of symptomatic CVS.
In our study using animal models, remarkable results were achieved by continuous injection of fasudil hydrochloride, using indwelling microcatheters, in the vertebral arteries. Arterial as well as venous injection of fasudil hydrochloride had significant effects in preventing SAH-induced basilar artery spasm. On the fifth day after the injection, a significant difference in vasospasms was not observed between the intravenous group and the five-day group. However, the difference was significant between the IV and seven-day groups (P < 0.05). This was probably related to the short-term administration of the drug in the five-day group resulting in failure to achieve a full effect of the drug. On the seventh day after injection, no significant difference was seen between the five-day group and the seven-day group, but there was a significant difference between both intra-arterial groups when compared with the intravenous groups (both P < 0.05). The peak blood flow velocity of the BA, monitored using the color Doppler, was consistent with the above-mentioned results. This indicated that that there was a definite advantage of intra-arterial administration over intravenous administration, which became apparent with a longer period of administration.
We used indwelling microcatheters in the vertebral artery and continuous injection of fasudil hydrochloride was achieved with a micropump. This would prevent direct contact with the spasmic part of the blood vessel and prevent catheter thrombosis. Apart from that, low-molecular-weight heparin was given subcutaneously, to prevent thrombosis outside the catheter. During administration, a specific frame was used for rabbits, to restrict their movement and prevent microcatheter prolapse. This method of drug delivery was associated with the following advantages: (1) drugs could be delivered directly to the spastic site without going through the pulmonary circulation; (2) delivery was direct and the effect was rapid; (3) the amount of drug needed was small; and (4) there was little impact on the peripheral blood vessels.
In conclusion, our experiments show that CVS following SAH can be effectively treated with intra-arterial administration of fausidil hydrochloride. This is a new method that may have potential benefits in clinical practice. Some of the benefits of this method include rapid, but short-term action of the medications that provide sustained prevention of CVS. Further studies are needed to assess the feasibility and clinical effect of this method in humans.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2]