Recirculatory Fibrinolytic-Assisted External Ventricular Drainage (EVD) for Intraventricular Hemorrhage

Background Elevated intracranial pressure and acute obstructive hydrocephalus secondary to intraventricular hemorrhage (IVH) can be treated by external ventricular drainage (EVD). The treatment time and the risk of EVD-related complications can be reduced with brinolytic agents’ instillation via an EVD catheter, but previous clinical trial results did not reveal a signicant improvement in terms of long-term functional outcomes. A recirculatory brinolytic-assisted EVD system was designed. The clot dissolution effectiveness of the system under different drug dosages and uid ow rates was tested in an ex vivo model. Results The results showed that the mean clot mass was quickly reduced in an initial brinolytic agent dose-independent stage, followed by a dose-dependent stage. Elevating brinolytic agent dosages beyond a certain threshold did not contribute to shorter dissolution times. Optimal treatment parameters for such a system were determined. A recirculatory ow rate of 10–18 ml/min with a low-dose of 30 000–60 000 IU of uPA resulted in an 80% clot mass reduction within four hours.


Background
Intracerebral hemorrhage (ICH) has an annual incidence of 10 to 30 per 100 000 people and accounts for 2 million, or 13%, of all newly diagnosed stroke cases worldwide [1][2][3]. The mortality rate is 30% and can result in signi cant disability [4,2]. 40% of hemorrhagic stroke patients have concomitant intraventricular hemorrhage (IVH), which is an independent predictor for 30-day mortality and poor six-month functional outcomes [5-8, 4, 9-12]. The normalization of intracranial pressure (ICP) and the resumption of normal cerebrospinal uid (CSF) circulation are dependent on e cient IVH clearance, which increases the likelihood of a good recovery.
External ventricular drainage (EVD) is a common procedure performed in daily neurosurgical practice for the treatment of IVH. Blood products and CSF are passively drained along a pressure gradient via an intraventricular catheter, and life-threatening obstructive hydrocephalus can be relieved [13][14][15]7]. However, such catheters are at risk of occlusion, migration, and inducing central nervous system (CNS) infections such as bacterial meningitis or ventriculitis. Intraventricular brinolytic agent administration via the catheter can reduce the risk of catheter occlusion and improve treatment outcomes. After instilling recombinant tissue plasminogen activator (rt-PA) [16][17][18] or urokinase plasminogen activator (uPA) [19][20][21][22]12], the catheter is typically temporarily clamped for 60 minutes to allow time for clot dissolution. The catheter is subsequently reopened for the siphoning of blood products and CSF from the ventricular system. The process is generally repeated every 8 or 12 hours until IVH clearance is observed or when obstructive hydrocephalus is relieved on serial computed tomography (CT) scans. The treatment duration generally lasts 2 to 10 days. However, potential complications associated with intraventricular brinolytic drug instillation exist, including higher rates of neuro-in ammation, recurrent hemorrhage, and infection [23,24,7].
The causes for these procedure-related complications are believed to be associated with the mode of drug delivery and ineffectiveness of passive drainage. The brinolytic agent's movement into the clot is governed by its dosage concentration gradient at the clot surface. However, studies showed that increasing dosages failed to shorten the dissolution time beyond a threshold [25,26]. Instead of increasing the dosage, which also increases the risk of neuro-in ammation or recurrent hemorrhage, the application of a ow/pressure gradient could assist drug penetration. Active drainage of the dissolved products could also contribute to faster clot removal. A closed-loop uid circulation prototype system designed to drive brinolytic agent penetration into the clot was developed. The effects of the uid ow gradient and the brinolytic agent concentration gradient on clot dissolution time were investigated in this study.

Clot Preparation
Clot volumes ranged from 10 ml -50 ml [9,5,12,4,11]. In this study, 1.6 ml of 0.25 M calcium chloride was added to 20 ml of blood drawn from healthy donors (aged between 20 to 30 years) to reverse the intrinsic anticoagulation cascade [27]. The blood was then transferred into a warming receptacle at 37°C for 90 minutes.

Active EVD System Development
The Clot Lysis: Evaluating Accelerated Resolution of Intraventricular Hemorrhage (CLEAR III) is the most recent randomized placebo-controlled trial utilizing alteplase, a rt-PA, as the thrombolytic agent. However, rt-PA has been shown to elicit signi cantly more severe in ammatory processes when exposed to nervous tissue compared to uPA [28][29][30][31]. In addition, since uPA is more widely available in China and Southeast Asia, the agent was adopted for this study [21,5,20]. In clinical practice, uPA dosages prescribed for intraventricular instillation varies from 10 000 to 100 000 International Units (IU) constituted in 2 ml -10 ml of saline solution. The drug is infused through the EVD catheter into the ventricular system every 8-12 hours [32,33,28,[20][21][22]12]. As the entire treatment course generally lasts 2-10 days, the total uPA dosage can range from 40 000-3 000 000 IU. In this experiment, uPA (Urokinase -Green Cross Injection 60 000 IU, China Chemical & Pharmaceutical Co. Ltd; Taiwan Green Cross Co. Ltd) of sequentially escalating doses of 6 000, 12 000, 30 000, and 60 000 IU was diluted in 50 ml of saline solution and transferred into the investigated active EVD system. Normal saline solution was used as the control infusion. An EVD catheter with an outer diameter (OD) of 3.1 mm and 20 inlet holes located within 27 mm from its tip (catalog number, 82-1735, Codman Neuro™, Depuy Synthes, Raynham [MA], USA) was used to deliver uPA. A co-axial microcatheter with an OD of 1.3 mm and inner diameter (ID) of 1.1 mm was used to drain CSF and lysed blood products. As the ID of most commercial EVD catheters range from 1.3 mm to 1.5 mm, the 1.1 mm ID of the microcatheter was not considered a potential risk factor for occlusion since continuous uid pressure will be applied through the system. The drainage catheter and the uPA delivery catheter were connected to a container with 50 ml of uPA solution and a lter to separate blood products from CSF and uPA solution. Recirculatory ow was controlled by two peristaltic pumps.

Experiment Setup
The mass of the tested blood clots was rst determined. They were then transferred into a polyethylene bag submerged in a 37°C water bath. The co-axial catheters were then inserted into the clot (Fig. 1). The ow settings were adjusted from 10-11 ml/min, 12-13 ml/min, 16-18 ml/min, to 20-24 ml/min, with drug delivery at a lower rate and uid drainage at a higher rate.
The CLEAR III trial results concluded that poor functional outcome within the alteplase patient group was attributed to insu cient clot clearance. For every milliliter of clot remaining, the trial investigators observed a patient mortality hazard ratio of 1.03 and an mRS≤3 odds ratio of 0.96. A signi cant association was also identi ed between the attainment of > 80% of intraventricular clot clearance with both good functional outcome and lower mortality [34,35]. In this study, clot mass was measured every 30 minutes on a precision balance until it fell below 20% of its initial mass and this was de ned as the dissolution time. For the control tests, the experiment was halted after 8 hours of treatment. Each test was repeated three times.

Results
The mean initial clot mass was 21.7±0.9 g. The effects of ow rates on the percentage of clot mass change regarding uPA dosage are shown in Fig. 2 and Table 1. Clot mass reduction was rapid, independent of ow rates and uPA dosage in the rst 30 minutes for all tests. Dependence on uPA dosage became apparent after 30 minutes, while the clot mass reduction in the control group saline tests were signi cantly slower. The remained clot after 2-hour treatment with 6 000 IU was 28.34%±10.85%, compared with 45.61%±7.16% for saline infusion (t-score = 9.04, p-value < 0.00001). In contrast, increasing ow rates only marginally increased the clot mass reduction in the control saline group tests.
The remained clot after 2-hour treatment with 11 − 10 ml/min was 46.87%±3.05%, compared with 36.55% ±4.27% for 24 − 20 ml/min (t-score = 0.98, p-value = 0.43). The saline group results showed that infusion ow alone was responsible for approximately 60% of clot mass reduction.  Fig. 3(a). Clot dissolution times were shortened with increased dosage for all ow conditions. For uPA dosages > 30 000 IU, the shortest clot dissolution times achieved were from 2-4 hours. Clot dissolution times were also shortened with increasing ow rates ( Fig. 3(b)) with a minimum threshold rate of 16ml/min. These results suggest that 80% of clot mass reductions can be accomplished with a lower uPA dosages of 30 000 to 60 000 IU at an infusion ow rate > 16ml/min within 4 hours.

Discussion
Intraventricular hemorrhage is a commonly seen neurological emergency in 40% of patients with ICH and 50% of patients with subarachnoid hemorrhage (SAH) [36,37]. Studies have shown that IVH is a poor independent prognostic factor for both types of hemorrhagic stroke [38]. The mortality rate of IVH complicating ICH is as high as 50%, with the volume of intraventricular blood being proportional to the risk of this outcome [37,34]. Even among survivors of IVH, only 20% of patients were observed to have good functional outcomes [37].
The detrimental effects of IVH can be broadly classi ed into two causes: A) primary brain insult due to intracranial hypertension, either as a result of the hematoma volume per se or obstructive hydrocephalus; and B) secondary insult due to hematoma-induced cytotoxicity, oxidative stress, and neuroin ammation.
Both immediate and delayed hydrocephalus can occur as a consequence of IVH. The commonest immediate life-threatening cause is acute obstructive hydrocephalus from the blockage of normal CSF ow within the ventricular system. Studies also demonstrated that both the initial total hematoma volume and the duration of intraventricular hematoma presence before its dissolution were independent determinants of communicating hydrocephalus after the acute obstructive phase [39]. The CLEAR III trial showed that 44% of patients required permanent CSF shunting when ICP exceeded 30 mmHg [34]. In a population-based study, shunting was indicated in 18% of IVH patients, and shunt-dependence was identi ed as an independent predictor for long-term morbidity [40,41]. Secondary brain injury due to IVH can occur due to disruption of the blood-CSF barrier, neuroin ammation, or cytotoxicity arising from blood components. Swine model studies have shown that obstructive hydrocephalus resulted in hypoperfusion changes of periventricular structures that persisted beyond hematoma resolution [42]. The ependymal surface, the thin single-layer epithelioid glial membrane that lines the ventricular system, are often damaged during IVH [42]. This results in dysregulation of the normal transfer of extracellular uid, ions, and small molecules between the brain parenchyma and CSF [43,44]. Iron, a degradation product of hemoglobin, is cytotoxic and can exacerbate oxidative injury [45,46]. Normally, ependymal cells prevent excess iron deposition in the brain, but with their disruption following IVH, its accumulation can cause persistent hydrocephalus [47]. In ammation, mediated by transforming growth factors (TGF-β1 and TGF-β2), has also been observed to cause obliterative arachnoiditis or dysfunction of the arachnoid granulations resulting in post-hemorrhagic communicating hydrocephalus [48,49]. In animal studies, intraventricular injection of lysed red blood cells and iron led to a signi cant increase in ventricular volume and elevated CSF in ammatory markers [50]. These ndings were supported by another study where aseptic CSF in ammation was detected in post-IVH brain tissue [44]. Intraventricular blood volumes were also positively correlated with more extensive neuroin ammatory responses re ected by elevated CSF white blood cell (WBC) counts [44].
Despite greater understanding of the multiple pathophysiological sequelae of IVH-induced secondary brain injury, current clinical management remains directed at ICP control. For patients with obstructive hydrocephalus, EVD functions both as therapeutic relief of intracranial hypertension and pressure monitoring. The e cacy of EVD for selected patients with IVH is widely accepted. The CLEAR III trial observed a proportional increase in mortality and poorer functional outcomes for every milliliter of clot remaining in the ventricular system, with 80% clot clearance being the determining threshold for improved outcomes [34]. However, drainage alone does not effectively assist clot dissolution. The mechanism of intraventricular blood clot lysis, similar to elsewhere in the body, is highly dependent on circulatory brinolytic activity. Unlike in the systemic vasculature, normal CSF does not contain brinolytic enzymes.
Studies revealed that plasminogen and tissue plasminogen activator enzyme systems in IVH become saturated by 24-48 hours, and after that, further clot dissolution reaches a steady constant rate [51,52]. This could explain why clot clearance with EVD alone is often frequently slow, lasting several days to weeks. In addition, ventricular catheter blockage by blood clots or by the choroid plexus often necessitates ushing or even repeated catheter revisions. The number of drainage system manipulations and duration of EVD are risk factors for bacterial ventriculitis and catheter-associated intraparenchymal tract hematomas [53].
The intraventricular administration of brinolytic agents through the ventricular catheter was introduced to facilitate clot dissolution and drainage in the 1990s [22]. Not only were catheter obstruction rates signi cantly reduced, but the duration required for clot clearance was also considerably shortened compared to EVD alone [38]. Meta-analyses showed that intraventricular brinolysis reduced IVH mortality by half and improved patient functional outcomes [54,37,55]. A recent systemic review also demonstrated a reduction in shunt dependence in patients that received such treatment [54].
Preclinical studies have observed that intraventricular rtPA could be neuro-toxic and pro-in ammatory [56][57][58]. A randomized-controlled trial noted that CSF cytokine concentrations, particularly tumor necrosis factor-α, interferon-γ, interleukin (IL)-1, IL-4, IL-6, and WBC counts were signi cantly elevated compared to placebo-treated patients [59]. There is also the theoretical risk of inducing further IVH with brinolytic agent instillation, although this has not been con rmed with RCTs [34]. Therefore, the duration and dosage of such drugs should be carefully titrated to balance clot lysis effects against its potential adverse effects.
Current intraventricular brinolytic agent administration involves temporarily clamping the catheter to permit CSF ow stasis after their introduction. As the blood clot is a crosslinked solid with micropores, drug permeation is primarily driven by a concentration gradient in a closed system. However, the rate and effectiveness of drug diffusion are di cult to monitor and studies demonstrated that merely increasing dosages did not shorten clot dissolution times [25,26].
To increase clot dissolution, a prototype ow-and pressure-gradient actuated system with uPA was designed to accelerate drug permeation into the micropores of the clot. In this proof-of-concept study, we observed that this recirculatory brinolytic drug-assisted EVD system shortened clot clearance durations and required signi cantly lower overall uPA dosages. No catheter occlusions were observed during the treatment process. These ndings may have clinical implications including reducing in-dwelling catheter durations along with its attendant related complications, decreasing the risk of neurotoxic blood degradation product exposure, and the minimizing the need for CSF shunting. The potential bene ts of this novel system could translate to shortened hospitalization durations and improved functional outcomes.
Endoscopic-assisted microsurgical evacuation is an emerging technique for intraventricular clot removal. Clinical trials revealed higher clot removal rate, fewer complications, and better postoperative outcomes [60-62]. However, the operative time is signi cantly longer than EVD catheter placement. There is a steeper skill learning curve to attain procedure pro ciency and often an experienced assistant is required to hold the neuro-endoscope. With such direct instrumental intraventricular clot manipulation close attention must be paid to avoid foramen, thalamic and thalamostriate vein injury that could result in catastrophic neurological consequences. Clots, especially during the acute phase, may be highly adherent to the ventricular wall unamenable to simple suction. The narrow working eld of vision of a rigid neuroendoscope's working channel also limits its capability to reach deep intraventricular locations, in particular the fourth ventricle. For these reasons, endoscopic-assisted microsurgical clot evacuations of IVH are not commonly performed and the e cacy and safety of this technique compared to EVD requires further investigation [63, 61, 64]. In view of these facts, it is believed that this novel recirculatory brinolytic drug assisted EVD system addresses the drawbacks of both these clot evacuation therapies.
There are several limitations to this study. First, the rate of clot dissolution is multifactorial. Apart from IVH volume, clearance is also affected by the clot's anatomical location and the relative position of the catheter. IVH within the third ventricle and the frontal horns of the lateral ventricles were subject to more pronounced hematoma thrombolysis than elsewhere [65]. The morphology of the experimental receptacle containing the clot in our experiment was not a patient-driven anatomical replica of the ventricular system. Clinical studies showed that catheter placement on the side predominantly involved by IVH resulted in more rapid clot clearance [66]. Future studies investigating an active recirculatory drainage system of this nature should adopt a stereolithographic model of the ventricular system. Inter-individual variations in baseline serum plasminogen and platelet count levels in uenced IVH resolution durations [52]. These factors were not controlled in the blood samples collected from healthy individuals in our study. Our model also did not consider ICP elevations that would require emergent relieving measures. Therefore, maintaining a closed EVD system without disruption of the thrombolytic uid ow gradient is a design challenge that needs to be overcome before it can be clinically applicable. Finally, the possibility of rebleeding, natural history complications commonly encountered in clinical practice, could not be studied in our ex-vivo experimental model.
To conclude, this novel recirculatory brinolytic agent assisted EVD system demonstrated rapid hematoma clearance at signi cantly lower doses. A window of optimal ow and uPA dosage rates was determined in this study. Further, in vivo animal studies are required to address the model's limitations and to realize the potential clinical bene ts for this novel IVH therapy.

Conclusions
EVD assisted by quasi-static brinolytic drug immersion can shorten blood clot mass reduction time.
Replacing quasi-static drug immersion with drug recirculation signi cantly reduced treatment durations. A recirculatory brinolytic-assisted EVD system could clear 80% of the clot within four hours using a low uPA doses of less than 60 000IU. Patients with IVH may bene t from this novel clot clearance system.

Declarations
Ethics approval and consent to participate This article does not contain any studies with human participants or animals performed by any of the authors. Ethics approval is not required as the blood samples were donated by the research team members voluntarily.

Consent for publication
Not applicable.
Availability of data and material The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.