Introduction

Intracerebral and particularly intraventricular hemorrhages (IVH) are severe types of stroke with interrelated etiologies and a complex pathophysiology representing a challenge for neurocritical care [1,2,3,4,5,6,7]. Combined neurosurgical and pharmacological treatment may be one option to treat these severely ill IVH patients (Fig. 1). Hence, clinical trials addressing intraventricular thrombolysis [8,9,10] including the most recent trial [11] and its sequential sub-analyses [12,13,14] were anticipated and discussed with great interest [15]. For example, CLEAR III was a randomized, multicenter, placebo-controlled trial that included 500 patients suffering from IVH, mostly originating from intracerebral hemorrhage (ICH) [11]. Patients were randomized to receive intraventricular alteplase (recombinant tissue plasminogen activator, rt-PA) as treatment or intraventricular saline as placebo. While the primary outcome measurement (moderate disability, modified Rankin scale, mRS ≤ 3) was slightly but not significantly higher in the rt-PA group (48%) compared to placebo (45%), the rt-PA group had a significantly lower mortality rate after 180 days. Nevertheless, some surviving patients had severe disabilities (mRS = 5) in the rt-PA group, which is not unusual for this critical cerebrovascular disease. The ambitious and well-designed trial was expected to improve knowledge on how to treat IVH patients. However, the pathophysiologic sequelae following IVH treatment with intraventricular alteplase raised new questions and may potentially implicate a therapeutic double-edged sword. But the question arises—why?

Fig. 1
figure 1

Cranial CT scans of two patients suffering from massive (a–d) or less severe (e–h) intraventricular hemorrhages (IVH). Both patients received intraventricular rt-PA via external ventricular drainage. a–d Axial images of a cranial CT scan of a male patient (82 years) suffering from a subarachnoid and intraparenchymal hemorrhage with a massive IVH. Most likely, the IVH is a result of an aneurysm rupture of the left distal anterior cerebral artery. The CT-based morphology and distribution of the intracranial hemorrhage and the presentation of the patient’s symptoms suggest the occurrence of an initial bleed and a later re-bleed. Note the edema formation particularly at the semi-oval center of the left hemisphere adjacent to the large blood clot, particularly visible in a and b. c, d The external ventricular drainage in the right lateral ventricle. e–h Axial and coronal CT images of a male patient (61 years). The images show a less severe intraventricular hemorrhage. The external ventricular drainage is seen in f and particularly in the coronal image g. The implantation accidentally caused some amounts of intracranial air (e, f). Blood clot-adjacent regional edema formation can be seen in g and h. Yellow arrows signifies blood or blood clots. Blue arrow heads signifies brain parenchyma with vasogenic edema formation and/or cerebrospinal fluid diapedesis from the ventricles. Red arrows signifies external ventricular drainage

Full size image

Please find the search criteria for this review online (Supplementary material: Method and Search criteria).

Classic and Hypothetical Routes of Cerebrospinal Fluid

To constructively debate this question, we should briefly recall the routes of the cerebrospinal fluid (CSF, Fig. 2a). The CSF is filtrated and actively transported (“produced”) at a rate of approximately 0.35 mL/min predominantly by the choroid plexuses, specialized endothelium and ependyma-like two-layered structures building the fairly tight blood-CSF barrier (BCSFB). Other parts of the CSF may directly leak from cerebral capillaries. However, the respective amounts are still under debate [16]. From the choroid plexuses, the CSF circulates through the entire ventricular system and reaches various cisterns via physiologic apertures. As a CSF-brain barrier hardly exists, CSF partially penetrates into the brain parenchyma, quasi-communicating with the interstitial fluid (ISF) [17], particularly when severe IVH occurs (compare Fig. 1a–d). Due to arterial pulsation and inspiration, the glymphatic system (see below) participates in driving the CSF/ISF flow (in IVH conditions together with blood-derived solutes) forward to the perivascular Virchow-Robin space, passing through the brain parenchyma and alongside cerebral vessels into the subarachnoid space [17, 18]. In the cortex, CSF is mainly resorbed by arachnoid granulations (pacchionian granulations) allowing the fluid to return into the venous circulation (Fig. 2). Other parts leave the subarachnoid space alongside larger cranial nerves and elsewhere, reaching the peripheral lymphatic system and then the venous circulation [16,17,18].

Fig. 2
figure 2

Pathophysiological mechanisms after intraventricular hemorrhages including particular changes after intraventricular rt-PA treatment. a An overview of the brain and the ventricular system including the CSF flow. b A schematic illustration of the most important functional parts of the parenchymal CSF/ISF pathways and the so-called glymphatic system. a Intraventricular blood and rt-PA predominately affect different brain barriers, the glymphatic system, and other brain structures clearing the CSF/ISF. The CSF is produced by the choroid plexus and initially circulates in the ventricular system before inflowing the interstitial space of the brain parenchyma via certain perivascular routes and through more or less tight brain barriers. At this stage, the CSF is called ISF building the distinguished neuronal and glial extracellular microenvironment, which may also contain waste products. ISF leaves the parenchyma via multiple routes such as the perivascular glymphatic system, thereby reappearing as CSF (external CSF). CSF of the subarachnoid space near the sagittal sinus is taken up by the pacchionian granulations that direct the fluid into the venous circulation, guaranteeing a steady state of CSF production and resorption. b After intraparenchymal and/or intraventricular hemorrhage, blood including its solutes and degradation products enters the interstitial space thus disturbing the microenvironment and the neuronal and glial functions. This “blood cocktail” destabilizes brain barriers to a certain extent, therefore entering the interstitial space by various short cuts. This cerebral barrier dysfunction is further enhanced by the use of intraventricular rt-PA application. More importantly, therapeutic rt-PA resolves the (intraventricular) blood clot in a shorter time period compared to the physiologic pathways, thereby increasing the concentration of the “blood cocktail” in the CSF/ISF including all devastating consequences for the microenvironment such as cellular dysfunction and death. Please see further detailed explanation in the main text body. Arrows signify direction of CSF/ISF flow

Full size image

Cerebrospinal and Interstitial Fluid Flow after Intraventricular Hemorrhage

During and after IVH, blood-derived solutes, e.g., thrombin, potassium, and other vaso-/endothelium-active substances, affect the distinct environment of the brain parenchyma, particularly the neurovascular unit (NVU) [11, 19, 20] and the CSF-generating and CSF-clearing mechanisms [11, 21]. The NVU is the conceptual unity of brain cells (neurons, astrocytes, etc.), immune cells, pericytes, endothelial cells, and other blood vessel cells [21]. Overall, it is now recognized as a core signaling unit within the cerebrovascular network, regulating blood supply to the brain depending on functional demand, thereby guaranteeing brain homeostasis. Similarly, the CSF serves a variety of functions including supplying nutrients and clearing waste products. In this, it works in conjunction with the glymphatic system (Fig. 2b), which appears to be a crucial paravascular clearing mechanism for brain waste products [17, 18, 22]. Indeed, the glymphatic system may help to clear blood derivatives from the CSF/ISF after IVH in general, but particularly after therapeutically boosted plasmin-mediated clot lysis [17, 18]. However, the function of the glymphatic system, as well as of the pacchionian granulations and choroid plexuses, could be impaired in severe IVH by blood derivatives including albumin and fibrin resulting in disturbed CSF flow and various forms of hydrocephalus [2, 4, 5, 18, 20, 22].

Pathophysiologic Sequelae After Intraventricular Hemorrhage and rt-PA Treatment

Several clinical trials on IVH with rt-PA treatment have been conducted in the last decade. Table 1 provides an overview of important clinical trials investigating treatment options, occurrence of hydrocephalus, and outcome after IVH. After these trials, particularly after CLEAR III and the recently published related sub-studies [11,12,13,14,15], however, it still remains unclear whether fast clot lysis may decrease the pathophysiologic problem of blood plus its solutes flooding the ventricular system and subsequently the brain parenchyma/ISF, or whether fast clot lysis may augment this problem. One could speculate that intraventricular rt-PA may add to the initial damage after IVH by quickly overwhelming physiological clearing routes and specific structures in the short time frame of iatrogenic intensified clot lysis. Although our hypothesis appears to be fairly straightforward from a physiologic point of view, more basic science and clinical studies are needed to draw final conclusions. On the other hand, the rationale behind the CLEAR III trial was strong and the positive results regarding mortality were promising [7, 11]. Moreover, other authors discussed further benefits and other indications of rt-PA perhaps supporting the glymphatic system [22]. At the same time, the neutral results regarding the primary outcome measure after rt-PA in IVH appear disappointing [11].

Table 1 Overview of clinical trials investigating rt-PA treatment and outcome of patients suffering from intraventricular hemorrhage
Full size table

The question still remains whether blood of any aggregate status should be therapeutically cleared by using the physiologic parenchymal CSF/ISF routes, such as the glymphatic system [18], thereby potentially inducing secondary brain cell damage and barrier impairment (Fig. 2b). Alternatively, it may be better to remove the blood burden before its components affect the ISF, the glymphatic system, and thereby directly the brain parenchyma cells including the NVU and the blood-brain barrier (BBB).

Fisher considered an organ-specific (including a brain-specific) coagulation and plasmin activity [23]. Particularly in the brain, the endogenous plasmin-mediated clot lysis system is known to function efficiently in extravascular cerebral compartments [23, 24]. In addition, it is a well-recognized clinical observation, and most neurosurgeons know exactly that CSF interferes with the coagulation, which barely reaches its final stage in the brain parenchyma or cerebral fluids: a persistently stable clot [23, 24]. This fact may be due to a lack of cerebral transglutaminase activity in CSF/ISF, such as seen in acquired coagulation factor XIII deficiency in chronic subdural hematoma/hemorrhage patients [24]. Application of intraventricular rt-PA further amplifies the endogenous plasmin system with anti-coagulation characteristics of the CSF, leading not only to timely intraventricular clot lysis but also to an increased cerebral inflammatory response [11, 13]. The question arises of whether fast clot lysis in the CSF beyond the endogenous lytic rate helps or harms IVH patients. Plasmin-/plasminogen-boosted clot lysis using rt-PA treatment subsequently opens and impairs and may thus harm important brain barriers such as the BBB [25] or the BCSFB [26], thereby affecting brain cells as well [27]. It is unknown whether the endogenous, i.e., intrinsic plasmin activity, represents a physiological strategy to moderately open certain brain barriers, thereby increasing fluid turnover and clearance of the CSF/ISF from harmful blood solutes. On the other hand, rt-PA itself is a powerful drug, but it is also known to induce serious side-effects in brain parenchyma and in particular at the NVU; for a review, see [27]. CSF inflammation after IVH is not only dependent on the initial volume of the IVH; intraventricular rt-PA further augments this inflammatory response, even after correcting for initial IVH volume [13]. However, although this augmented inflammatory response may harm the NVU, it was not associated with a noticeably worse modified Rankin scale score 180 days after IVH [13].

Moreover, various additional pathophysiologic mechanisms are relevant for ICH/IVH such as microglia activation, alterations of the interleukin levels, and/or other neuroimmunological responses to blood and its derivatives in the brain and CSF compartments. These specific aspects are out of the scope of this focused review and should be discussed elsewhere.

Future Directions

In general, and for future research, another fact seems to be important: Dissolution of blood (clots) in a fluid (CSF/ISF) does not imply its subsequent removal. In clinical practice, a unique blood removal approach would require a second CSF catheter. This could facilitate volume-neutral addition of artificial CSF, resulting in higher fluid turnover with increased external drainage, thereby leading to clearance of blood from the ventricles before entering the brain parenchyma/ISF. In a recent clinical trial of IVH patients, intraventricular rt-PA treatment plus additional lumbar drainage was shown to reduce permanent shunt dependency after ICH/IVH [28] and thus may be protecting the endogenous “clearance” systems of the brain. Theoretically, even CSF dialysis or apheresis might be feasible with or without intraventricular rt-PA treatment. This technique could be considered as advantageous, because the use of larger CSF catheters alone did not improve the outcome of IVH patients [29]. Other interesting therapeutic options may be (a) the pharmacologic support of the glymphatic system by dexmedetomidine to further increase clearance of blood-derived waste products particularly in the parenchyma [30] or (b) the pharmacologic support of key mechanisms of the NVU  e.g. with low-dose lithium to stabilize endothelial  cytosolic calcium homeostasis, decrease myosin light chain phosphorylation,  thereby improving  endothelial barrier, and vascular smooth muscle function [31,32,33]. Using these approaches means potentially mitigating IVH- and rt-PA-induced impairment of the glymphatic system and the NVU dysfunctions. Therefore, further basic research is needed in this field. One should keep in mind, however, that structure and function of the glymphatic system are still not entirely understood and discussed. Most authors believe, but a few doubt the widespread continuum of the glymphatic system or the plausibility of bulk flow therein [34, 35]. Clinical trials may not fully answer important questions regarding the human glymphatic system and the best treatment options after IVH (Table 1). Ultimately, only further translational cerebrovascular research together with subsequent well-designed clinical trials, such as the recent ones [11,12,13,14,15, 28,29,30, 32, 36], may verify that individual IVH patients benefit from therapeutic CSF clearing and removal strategies [37, 38].