Novel Rat Model of Severe Cerebral Venous Sinus Thrombosis Established by Combination of Semi-Ligation, Ferric Chloride, and Thrombin

Cerebral venous sinus thrombosis (CVST) is a distinct form of cerebrovascular disorder that differs from arterial stroke. It occurs more frequently in younger individuals^1,2, particularly in pregnant women³, and its incidence has been rising steadily in recent years. An epidemiological study conducted in Australia indicates that the annual incidence of CVST ranges from 13.0 to 15.7 cases per million individuals⁴. In the United States, the incidence of CVST exhibits an upward trend with age. Nevertheless, the peak age of onset for women is lower compared to that for men⁵. Between 2021 and 2023, 43 per million patients who visited emergency departments in the United States were diagnosed with CVST⁶. In clinical settings, multiple sinus thrombosis is the most commonly observed form, accounting for approximately 57.14% of all cases. This is followed by involvement of the superior sagittal sinus (SSS; 16.19%), the transverse sinus (11.43%), and the sigmoid sinus (8.57%)⁴. The main clinical manifestations include headache, which may be accompanied by epilepsy, disorders of consciousness, and focal neurological symptoms^7,8,9. However, these manifestations are not specific. Approximately 60% of patients with CVST develop cerebral venous infarction or hemorrhage, which is categorized as severe CVST^10,11. Although the treatment of CVST includes anticoagulation therapy and surgical intervention, the mortality rate of severe CVST remains as high as 34.2%⁹. This high mortality may be attributed to the fact that the underlying pathophysiological mechanisms of severe CVST are not yet fully understood. An appropriate animal model of severe CVST is therefore an essential experimental tool for investigating its pathogenesis, pathophysiological progression, and potential therapeutic strategies¹².

Currently, the methods for inducing CVST animal models can generally be categorized into the following types: permanent ligation^13,14, chemical induction^15,16,17, interventional approaches^18,19, implantation of self-made grafts^20,21,22, and bipolar electrocoagulation^23,24. The thrombus induced by previous animal models had a short duration and was unable to simultaneously induce multiple venous sinus thrombi and large-area venous cerebral infarction. Therefore, the development of a severe CVST model that involves multiple venous sinuses simultaneously has become an urgent priority. The method of semi-ligation combined with ferric chloride and thrombin can establish a novel rat model of severe CVST. Furthermore, the novel severe CVST model more accurately simulates the human pathophysiological process of severe CVST through three key aspects: venous sinus thrombus burden, venous cerebral infarction, and disruption of the blood-brain barrier. This model is applicable to research on pathophysiological mechanisms and treatment strategies for severe CVST. This novel rat model can be utilized in the future for investigating the pathophysiological mechanism of severe CVST.

The experimental protocol was approved by the Animal Experiments and Experimental Animal Welfare Committee of Capital Medical University (AEEI-2020- 119) in Beijing, China, and was conducted in compliance with the institution's Animal Care and Use Committee regulations.

1. Animal preparation

1. Randomly assign healthy adult male Sprague-Dawley (SD) rats, with an average body weight of 300 ± 10 g, to either the experimental group or the sham operation group. All the animals were maintained on an alternating 12 h light/dark cycle with access to food and water ad libitum.

2. Application of PSI

1. Use PSI to continuously monitor alterations in cerebral blood flow (CBF) within the SSS of rats throughout the modeling process. Select a fixed circular area within the SSS as the region of interest (ROI). By monitoring the changes in CBF of the ROI during surgery, determine the completion of semi-ligation and the success of thrombus induction.

3. Establishment of severe CVST rat model

1. Anesthetize rats using 5% enflurane, with a gas mixture consisting of 70% nitrous oxide and 30% oxygen delivered via an anesthesia mask. Maintain the concentration of enflurane within the range of 1%-2% throughout the surgical procedure. Place the anesthetized rats in the prone position with their heads fixed at the midline using a stereotaxic frame. Maintain the body temperature at 37 ± 0.5 °C throughout the surgical procedure using a thermostatic heating pad.
2. Make a 1.5 cm longitudinal incision along the midline of the rat's cranial region. Dissect the subcutaneous tissue to expose the underlying skull.
3. Under microscopic visualization, make a longitudinal cranial window (10 x 4 mm) between the bregma and the lambda using a high-speed dental drill. Expose the SSS and bilateral cerebral cortex while preserving the integrity of the dura mater. During the cranial drilling process, cool the drill bit continuously with normal saline to prevent thermal injury to the dura mater and cortex.
4. Following exposure to SSS, assess the CBF of the ROI in the SSS using the PSI, and record the corresponding data. Semi-ligate the SSS rostrally and caudally using an 8-0 polyamide sutures under microscopic observation.
   NOTE: When ligating the SSS, the force should be moderate and not excessive. Otherwise, a semi-ligation effect cannot be achieved. Ligation should ideally be performed in a single continuous procedure. Repeated ligation attempts may result in perforation of the SSS, thereby compromising the success of model fabrication.
5. Utilize the PSI to assess CBF of the ROI located at the ligated segment of the SSS, confirming that the CBF had diminished to approximately 50% of that before semi-ligation. Record the results systematically.
6. Under light-protected conditions, dip a 7 mm length 3-0 silk thread in 40% ferric chloride and use it to cover the surface of the SSS ligation site for a duration of 5 min. Wash the field with normal saline after the silk thread is removed.
7. Inject 0.1 mL thrombin solution (500 U/mL) into the sinus cavity of the ligated segment using a microinjector within 1 min. Repeat the injection two additional times consecutively, for a total thrombin dose of 150 U. Thrombin reflux to the epidural space did not lead to poor thrombosis.
   1. To facilitate the injection of thrombin into the SSS, gently bend the microinjector needle prior to insertion. This adjustment allows for better control of the needle's angle during puncture and helps prevent unintended penetration through the SSS.
8. Re-evaluate the CBF of the ROI using the PSI after thrombin injection. Ensure that the results confirm a significant decrease in CBF compared to that following semi-ligation, indicating successful induction of thrombosis. Record data systematically for further analysis.
9. Seal the bone window with bone wax, followed by suturing of the incision. After the surgical procedure, administer ketoprofen (5 mg/kg) subcutaneously to each rat to alleviate postoperative pain.
10. In the sham operation group, perform the craniotomy only, without any additional surgical interventions.

4. Venous sinus thrombosis burden and cerebral infarction

1. Evaluate the burden of venous sinus thrombosis at three time points: 1 day, 2 days, and 7 days after model establishment. Sacrifice rats and perform cardiac perfusion with normal saline at different time points.
2. Conduct a craniotomy to visually assess the extent of venous sinus thrombosis involvement. Isolate the thrombus and subject to quantitative analysis. Excise the brain tissue and section it into 2 mm thick coronal slices.
3. Immerse the slices in a 2% TTC solution and incubate in a 37 °C water bath for 20 min to facilitate staining. Following this, visually assess cerebral venous infarction and quantify the infarct volume using ImageJ software for subsequent statistical analysis.
   infarct volume = area of infarct in square mm x thickness [2 mm]

5. Disruption of blood-brain barrier (BBB) permeability

1. On the second day following the surgery, randomly select rats from both the model group and the sham operation group and administer an intravenous injection of 2% Evans blue (EB) through the tail vein.
2. After 2 h of successful administration of the injection, euthanize the rats, and perfuse with phosphate-buffered saline (PBS) through the left ventricle. Harvest the brain tissues and section them into 2 mm thick slices to assess the extent of EB penetration.

6. Statistical analysis

1. Conduct statistical analysis using SPSS 17.0 software. Present measurement data as mean ± standard deviation. Use one-way analysis of variance (ANOVA) or an independent samples t-test for group comparisons.
2. For the analysis of continuous variables measured at multiple time points within the same group, use repeated measures ANOVA to assess intergroup differences across time points. A p-value of less than 0.05 was considered statistically significant.

Modeling process
Figure 1A depicts the SSS of the rat, as denoted by the arrow. The rostral and caudal of the SSS were semi-ligated (Figure 1B), followed by application of ferric chloride-soaked thread onto the surface of ligated regions (Figure 1C). Thrombin was subsequently injected into the ligated segments to induce thrombosis. The area denoted by the arrow, where thrombin was injected, has turned black (Figure 1D). The white circle in Figure 1E-H indicates the ROI. In Figure 1E, a strong red blood flow signal was observed in the ROI following exposure to the SSS. In Figure 1F following semi-ligation, the blood flow signal in this region was markedly reduced. In Figure 1G, upon application of the ferric chloride-soaked thread, the red blood flow signal further diminished. Notably, after thrombin injection, the red blood flow signal was completely replaced by a blue signal indicating successful thrombus formation (Figure 1H). The success of thrombus induction was assessed based on these CBF alterations.

Changes in CBF during the modeling process
Figure 2A,B illustrates the alterations in CBF during the modeling procedure in both the model group and the sham operation group. Following semi-ligation, CBF in the ROI decreased to approximately half of the pre-ligation level. Subsequent application of ferric chloride and thrombin resulted in a further reduction in CBF, which indicated successful induction of thrombosis. However, CBF in the ROI remained constant throughout the observation period in the sham operation group (Figure 2B). Figure 2C presents a comparative analysis of CBF within the model group across three time points: before semi-ligation, after semi-ligation, and after ferric chloride and thrombin application. CBF exhibited statistically significant differences in the model group when comparing measurements before and after semi-ligation, as well as following the administration of ferric chloride and thrombin (359.4 ± 72.61, 190.0 ± 31.12, 95.93 ± 28.31, F=151.9, p < 0.001).

Observation of thrombus formation at various time points during the modeling process
Figure 3A illustrates the condition of venous sinus thrombosis 1 day after model establishment in the model group. 2 days after model induction, thrombus formation in the SSS and the transverse sinus was observed (Figure 3B). At 7 days after model induction, the thrombus in the SSS remained visible (Figure 3C), whereas no thrombus was observed in the transverse sinus, suggesting partial recanalization. These findings indicated that venous sinus thrombosis could persist for at least 1 week. However, we did not find out thrombosis in the venous sinus at any time point subsequent to the operation in the sham operation group (Figure 3D).

TTC staining and EB staining in the model group and the sham operation group
As shown in Figure 4A, venous cerebral infarction adjacent to the SSS was observed 1 day after surgery in the model group. 2 days after the procedure in the model group, extensive venous cerebral infarction was observed (Figure 4B). A venous cerebral infarction adjacent to the SSS remained observable 7 days post-operation in the model group, indicating that venous infarction may persist for up to one week (Figure 4C). No venous cerebral infarction was observed at any time point following the operation in the sham operation group (Figure 4D). As shown in Figure 4E, Evans blue staining in the model group was clearly evident 2 days after surgery, indicating increased blood-brain barrier permeability. However, no EB staining was observed in the sham-operated group (Figure 4F).

Comparison of thrombus weight and venous cerebral infarction volumes at various time points in the model group
Figure 5A demonstrated that there was no substantial difference in the weight of the thrombus between the first day and the second day post-operation (3.330±0.5290, 3.082±1.004, n = 5, P = 0.8282). However, significant differences were observed in the thrombus weight on the 1st (3.330±0.5290, 0.2480±0.1977, n = 5, P < 0.0001) and 2nd days (3.082±1.004, 0.2480±0.1977, n = 5, P < 0.0001) post-operation, respectively, when compared with that on the 7th day post-operation. Figure 5B indicated that the volume of venous cerebral infarction on the 1st day post-operation exhibited a significant difference compared to that on the 2nd day (65.08±19.97, 422.3±292.6, n = 5, P = 0.0197). However, there was no significant difference when compared to that on the 7th day post-operation (65.08±19.97, 117.8±89.48, n = 5, P = 0.8861). Moreover, the volume of venous cerebral infarction on the second day post-operation showed a significant difference from that on the 7th day post-operation (422.3±292.6, 117.8±89.48, n = 5, P = 0.0456). It indicated that the infarct volume peaked on the 2nd day post-operation and subsequently decreased by the 7th day. Over time, the weight of the thrombus gradually decreased, reaching its minimum on the 7th day post-operation. These findings indicated that upon the recanalization of venous sinus thrombosis, venous cerebral infarction might experience partial improvement.

Epileptic seizures
Epileptic seizures were observed in the rats (5.17%, 3/58) of the model group (refer to Supplementary Video 1 for details). As could be observed from the video, a rat in the model group abruptly underwent generalized convulsions, tremors of both forelimbs, and chewing movements. Following the convulsions, the rat exhibited listlessness. The behavior of the other two rats remained normal.

The results collectively show that the protocol led to the establishment of an epileptic seizure model.

Figure 1: Modeling process. (A-D) The complete modeling process, and (E-H) the corresponding changes in CBF at each procedural step. Please click here to view a larger version of this figure.

Figure 2: Changes in CBF During the Modeling Process. (A) Changes in CBF during the modeling process in the model group. (B) Changes in CBF in the sham group. (C) Comparative analysis of CBF within the model group across three time points, n = 20. Please click here to view a larger version of this figure.

Figure 3: Observation of thrombus formation at various time points during the modeling process. (A) Condition of venous sinus thrombosis 1 day after model establishment. As shown, white arrows indicate SSS thrombosis, green arrows indicate transverse sinus thrombosis, and yellow arrows indicate cortical venous thrombosis. (B) Condition 2 days after model induction, thrombus formation in the SSS (indicated by the white arrow) and the transverse sinus (indicated by the green arrow). (C) At 7 days after model induction, the thrombus in the SSS, indicated by the white arrow, remained visible. (D) No thrombosis in the venous sinus at any time point in the sham operation group. Please click here to view a larger version of this figure.

Figure 4: TTC staining and EB staining were conducted in both the model group and the sham operation group. (A-C) TTC staining of brain tissue in the model group on days 1, 2, and 7 post-operation, arrow indicates venous cerebral infarction adjacent to the SSS. (D) TTC staining in the sham operation group. (E,F) Evans Blue staining brain tissue in the model group and the sham operation group on day 2 after surgery (indicated by the white arrow). Please click here to view a larger version of this figure.

Figure 5: Comparison of thrombus weight and venous cerebral infarction volumes in the model group. (A-B) The thrombus weight and venous cerebral infarction volumes in the model group were compared at 1, 2, and 7 days post-surgery. Please click here to view a larger version of this figure.

Supplementary Video 1: Epileptic seizures in rats. Please click here to download this File.

A novel rat model of severe CVST constructed by semi-ligation combined with ferric chloride and thrombin cannot only simulate blood stasis and endothelial injury, but also the hypercoagulable state of blood, that is, it simulates the three elements of thrombosis formation²⁵. What is even more exciting is that the rat model developed using this method is capable of exhibiting epileptic seizures. Epileptic seizures represent a notable clinical manifestation in patients with CVST, occurring in approximately 12% to 31.9% of cases. Notably, patients with severe CVST who present with cortical venous thrombosis and venous cerebral infarction are at increased risk for developing epileptic seizures²⁶. As such, it more closely aligns with the pathophysiological mechanisms underlying human CVST.

In the model involving ligation of the SSS combined with thrombin injection, the approach described by Li et al.¹⁴is considered highly representative. This method induces SSS thrombosis by permanently ligating both the anterior and posterior segments of the SSS, followed by intra-sinus thrombin injection and transient bilateral carotid artery occlusion. The model effectively replicated key pathophysiological features of CVST, including blood stasis and a hypercoagulable state, and was capable of producing cortical venous thrombosis and associated venous infarction. However, due to the permanent nature of the sinus ligation, its applicability in evaluating potential therapeutic interventions is limited. Compared with Li's model, this semi-ligation method induces blood stasis while preserving partial blood flow, thereby minimizing interference with the evaluation of therapeutic strategies. In the chemically induced model, Rottler et al.¹⁵ applied filter paper strips containing 40% ferric chloride to the surface of the SSS for a duration of 5 min. The corrosive action of ferric chloride induced endothelial cell injury, thereby triggering platelet activation and subsequent thrombus formation²⁷. Although this method was technically straightforward, the resulting thrombi were found to be unstable, with complete recanalization occurring within 1 week. Furthermore, the model failed to induce cortical venous thrombosis or venous cerebral infarction. Wei et al.¹⁷ improved upon the method by combining the application of ferric chloride-soaked filter paper strips on the SSS surface with thrombin injection. This modified approach produced a more stable thrombus compared to the previous technique and effectively induced ischemic brain injury. However, the use of ferric chloride-soaked filter paper strip was associated with significant damage to the surrounding brain tissue. Compared with these models, the current model incorporates a modified 3-0 silk thread immersed in ferric chloride, significantly reducing the contact area between ferric chloride and the dura mater, thereby minimizing the risk of brain tissue damage. Bourrienn et al.²¹ induced thrombus formation in the SSS of mice by injecting autologous thrombi generated in vitro, combined with simultaneous bilateral ligation of the external jugular veins. Although this model successfully replicated pathological features such as ischemia and hemorrhage and demonstrated no significant thrombus recanalization within 1 week, the in vitro-generated thrombi differed substantially from naturally formed in vivo thrombi. Recently, Mu et al.²² introduced water-swellable rubber into the SSS of rats to simulate thrombosis. However, this approach failed to fully replicate the entire pathophysiological process of CVST. The ideal CVST animal model should fulfill three essential criteria: first, it must accurately simulate the pathophysiological processes of human CVST; second, it should be capable of inducing key pathological changes, including venous sinus thrombosis, cortical venous thrombosis, venous cerebral infarction, and intracerebral hemorrhage; third, it should serve as a reliable research platform for the development and evaluation of novel therapeutic strategies²⁸.

Given that thrombosis persists over an extended period, the current model can be utilized for long-term observation and the development of novel treatment strategies. Previous research has consistently shown that inflammation is correlated with venous cerebral infarction²⁹. The model can be employed to investigate the relationship between inflammation and venous cerebral infarction, explore novel therapeutic targets, and conduct evaluations of therapeutic effects such as anticoagulation and anti-inflammation.

The current model presents several limitations. First, male rats were used in the study despite the higher prevalence of CVST in females, which may affect the generalizability of the findings. Second, the semi-ligation procedure for the SSS requires precise application of force; excessive or insufficient tension may compromise its effectiveness. Mastery of this technique necessitates repeated practice to achieve consistent and reliable results.

In conclusion, a novel rat model of severe CVST can be constructed through semi-ligation in combination with the application of ferric chloride and thrombin. This model is characterized by a substantial thrombotic burden, simultaneous involvement of multiple venous sinuses, and the ability to induce extensive venous cerebral infarction, thereby more accurately recapitulating the pathophysiological features of severe CVST in clinical settings.