1. Endoscopic third ventriculostomy
On February 6, 1923, William Jason Mixter performed the first endoscopic third ventriculostomy in medical history in a child with congenital hydrocephalus using a urethroscope and a flexible probe [1]. Endoscopic third ventriculostomy was initially used only for obstructive hydrocephalus [2]. In recent years, with the continuous advancement of endoscopic equipment and endoscopic operating techniques, as well as the continuous research on the effects of endoscopic third ventriculostomy, the indications for endoscopic third ventriculostomy have been gradually expanded to include: aqueductal stenosis, idiopathic lateral and median foramen stenosis of the fourth ventricle, isolated four ventricles, DandyCWalker malformation, hydrocephalus due to intraventricular hemorrhage, tumor hydrocephalus (including: intracerebroventricular tumors, tumors of the pineal region and tetraspanic region, tumors of the posterior cranial fossa), hydrocephalus due to arachnoid cysts (including: arachnoid cysts of the suprasellar region, arachnoid cysts of the pineal region, arachnoid cysts of the tetraspanic region), hydrocephalus after infection or hemorrhage, ventricular syndrome, normal pressure hydrocephalus, non-traffic spinal cord cavity combined with Chiari I malformation and hydrocephalus, brain bulge combined with hydrocephalus, spinal closure insufficiency combined with hydrocephalus, and obstructive hydrocephalus due to tuberculous meningitis [3,4].
Most of the current studies have used the improvement of signs and symptoms of hydrocephalus and the absence of postoperative dependence on shunt as the criteria for successful endoscopic third ventriculostomy. According to the literature, the success rate of endoscopic third ventriculostomy ranges from 22% to 93.8% [5,6,7,8].Schroeder et al. performed an 8-year follow-up of 198 patients who underwent endoscopic third ventriculostomy with an overall success rate of 66% and a complication rate of 15.6% [8].Gangemi et al. showed that 86.4% of patients who underwent endoscopic did not need to rely on shunts after subtotal ventriculostomy, and 93.5% of these patients were patients with aqueductal stenosis [5]; Beems T et al. found that the success rate of endoscopic third ventriculostomy was 87% in patients with aqueductal stenosis [9]. Other related reports have also confirmed the high success rate of endoscopic third ventriculostomy for the treatment of hydrocephalus due to aqueductal stenosis [10,11,12].
The current study found that the complication rate of endoscopic third ventriculostomy is 0-22% [6,8,13]. Complications include intraventricular hemorrhage [14], hypothalamic dysfunction [15], bradycardia [16], apnea [17], intraoperative cardiac arrest and transient postoperative Horner’s syndrome, psychiatric symptoms [18], enuresis [19], epilepsy, pneumothorax [20], subdural effusion and hematoma [21], and progressive hypertrophic poly neuritis [22]. However, endoscopic third ventriculostomy is still a relatively safe procedure if the patient’s condition is carefully evaluated preoperatively, the diagnosis is clear, the indications for the procedure are strictly grasped, and attention is paid to intraoperative operations and postoperative management.
In a retrospective study by Tuli et al, the failure rate of endoscopic third ventriculostomy was 44%, while the failure rate of lateral ventriculoperitoneal shunt for hydrocephalus in children was 45%, which was not statistically different [23].In a study by Dieter H et al, 75% of failed endoscopic third ventriculostomy cases occurred within 6 months after surgery. Reasons for failure included: too small a fistula opening; failure to open the Liliequist membrane; increased protein and fibrinogen levels in the cerebrospinal fluid due to bleeding and infection, resulting in decreased cerebrospinal fluid uptake; and scar formation at the fistula opening.
Risk factors that may lead to failure of endoscopic third ventriculostomy include: 1. intracranial infection: this includes preoperative infection as well as postoperative infection, as infection can lead to subarachnoid adhesions, which in turn can affect cerebrospinal fluid circulation and absorption; 2. younger patient age: the postoperative failure rate is higher in younger patients with the same etiology, with some reports considering 2 years of age as the cut-off point, and others reporting 1 year or 6 months as the cut-off point. Because cerebrospinal fluid absorption is imperfect in younger patients, it has also been suggested that cerebrospinal fluid circulation is different in younger children than in adults [13], and these patients usually have both abnormal absorption and obstructive factors [24].3, Postoperative meningitis: the development of meningitis during follow-up can lead to the proliferation of fibrous tissue around the fistula and blockage of the fistula [1].
2. Lateral ventriculoperitoneal shunt
Nulsen and Spitz performed the first shunt in 1949, and since then, improvements have been made to the shunt device. In the 1960s and 1970s, shunt devices were widely used in clinical practice. Initially, cerebrospinal fluid was shunted into the venous system, but the risk of thromboembolism from this method was so high that it was replaced by the lateral ventriculoperitoneal shunt (VPS).
The advent of the lateral ventriculoperitoneal shunt has changed the treatment of hydrocephalus. However, lateral ventriculoperitoneal shunts still have a high failure rate. In a large clinical survey, the incidence of death due to shunts was found to be 1.4% [25]. The failure rate is about 40% in the first year after surgery, 53% in the second year and 60-62% in the fourth year [26,29,30]. The failure rate within 10 years after bypass is about 63-70% [23,29,30]. Related studies have found that the longer the bypass tube is in normal operation, the lower the incidence of bypass failure. 14% of bypass failures occur within 1 month after surgery, but only 5% of bypass failures occur 2 years after surgery [25].
Reasons for shunt failure include: overshunt, undershunt, shunt-associated infection, blockage of the ventricular or ventral end of the shunt, and valve failure.
McGirt MJ et al. showed that patients younger than 1 year of age, and infants born prematurely, were more likely to have shunt failure early after shunt surgery. They concluded that the age of the patient, whether the birth was premature or not, and the number of shunt replacements were all related to the length of time the shunt worked properly, whereas the cause of hydrocephalus was not related to shunt survival time. The duration of the shunt procedure is also a risk factor for infection. To reduce the failure rate, some scholars believe that the shunt procedure should be completed within 40 minutes [25]. In addition, shunts are generally not used in children with cerebrospinal fluid protein > 1,000 mg/dl and weight < 2 kg because of poor cerebrospinal fluid absorption in younger children, which can easily lead to shunt failure.
The incidence of shunt-associated infections is 5.1-15.2% [27,31,32,33,34]. Among patients who experienced shunt failure within the first month after shunt surgery, 45% were due to shunt-associated infections. However, only 6% of patients who experienced shunt failure after 2 years postoperatively were due to shunt-associated infection [25]. In 70C 90% of cases of shunt-associated infections, the infections were caused by Staphylococcus aureus and coagulase-negative staphylococci [35]. Thus, shunt failure occurs mainly in the early post-shunt period, and shunt-associated infections are the main cause of early shunt failure.
Both low age and history of prematurity are risk factors for shunt-associated infections due to: underdeveloped immune system in younger children and preterm infants, poor general condition of the skin, high density of bacteria on the skin surface, and high chance of sepsis.
To address the problem of shunt-associated infections, it was found that preoperative prophylactic antibiotics were not effective in preventing their occurrence, and that soaking the shunt with antibiotics was effective in killing the flora on the shunt surface and reducing the incidence of shunt-associated infections. This approach was first used in South Africa to prevent shunt-associated infections, and related randomized controlled trials showed that the incidence of shunt-associated infections decreased to 1.2-5% after the use of antibiotic-soaked shunts [31,32,33,34].
A recent study at Johns Hopkins University found that 74 patients under 1 year of age underwent a total of 108 shunts, all with shunts soaked with 0.054% rifampin and 0.15% clindamycin. The incidence of shunt-associated infections was 4.6% at a postoperative follow-up of more than 9 months. No antibiotic-related toxicities were found in this study. Therefore, it was concluded that antibiotic-soaked shunts used in patients with hydrocephalus younger than 1 year of age and premature birth were effective in reducing the incidence of shunt-associated infections [31]. In an Australian study, the rate of shunt infections was reduced from 6.5% to 1.2% with the use of antibiotic-soaked shunts. They concluded that this method only soaks the shunt and the reservoir capsule and valve are not treated, which leaves the possibility of shunt-associated infections in the shunt. However, soaking the shunt with antibiotics only reduces the incidence of early postoperative infections, not late infections, because late postoperative infections are primarily hematogenous and are not related to the number of bacteria on the surface of the shunt or the surgical procedure [35].
The incidence of excessive shunts has been reported in the literature to be approximately 1.5%-6.7%, with the highest reported by some to be 37% [36]. The incidence of excessive shunts leading to subdural effusion or hematoma is 3.4% [28].
Robinson et al. reported a 17% incidence of shunt overflow in patients with no or low pressure valves at 1 year. the incidence of shunt overflow reached 43% at 5 years and 57% at 7 years. In patients with medium- and high-pressure valve shunts, the incidence of excessive shunts was 2% after 1 year and 17% after 5 years. In contrast to patients with low-pressure shunts, patients with medium- and high-pressure shunts rarely experienced shunt overload. One possible reason for this is that low-pressure shunts can shunt more cerebrospinal fluid than high-pressure shunts. Excessive shunting of cerebrospinal fluid can result in smaller ventricles, which can lead to blockage of the ventricular end of the shunt as well as further ventricular shrinkage [27].
Intracranial pressure in children is changing with age, and currently, the use of low-pressure shunts is preferred for infants, thus avoiding persistent ventricular oversizing before the infant’s fontanelle is closed. As they age, they require the use of high pressure shunts to avoid excessive shunting, hypocranial pressure, and medically induced craniosynostosis [38].
Conventional shunts control the flow of cerebrospinal fluid through the valve flap and its associated components, and the valve pressure can be pre-set to high, medium, or low pressure, and the valve can open when the cerebrospinal fluid pressure is higher than the valve pressure, at which time the valve flap has very little resistance to the flow of cerebrospinal fluid, and if the patient is in the standing position, then gravity will cause a continuous flow of cerebrospinal fluid resulting in low intracranial pressure, which is known as siphoning. Siphoning is one of the causes of excessive shunting, and several different types of anti-siphoning valves are used in the clinic to address this problem.
The Delta Valve (Medtronic PS Medical, Goleta, CA) adds an anti-siphon device to the original valve, which consists of a pair of flexible diaphragms that narrow the opening between the diaphragms to reduce cerebrospinal fluid flow when the cerebrospinal fluid pressure decreases. Drake JM et al. showed no statistical difference in shunt failure rates between the conventional shunt device and the two anti-siphon shunts in a clinical comparison study. However, the anti-siphon shunt device was effective in reducing excessive shunting of cerebrospinal fluid due to siphoning while the patient was standing. No excessive shunting occurred in patients using the Orbis-Sigma Valve. This study showed that the design of the valve only made a difference in the cause of shunt failure, but did not reduce the incidence of shunt failure [28,37].
The advent of adjustable shunts in the 1980s allowed physicians to noninvasively adjust the shunt valve pressure according to the patient’s age and ventricular size, thus avoiding excessive shunts and the risk of reoperation [38]. Thus, in theory, adjustable pressure shunts can reduce the incidence of shunt overflow, shunt underflow, and blockage of the ventricular end of the shunt, thereby reducing the failure rate of lateral ventriculoperitoneal shunts.
An analysis of 279 shunt patients at Johns Hopkins Hospital showed that shunt failure occurred in 48% (135/279) of patients within 17±13 months postoperatively. Blockage of the ventricular end of the shunt accounted for 24%, shunt-associated infection for 6.4%, blockage of the ventral end for 5%, shunt valve failure for 6.1%, and other causes such as shunt breakage or dislodgement for 5.7%. In contrast, the use of adjustable pressure shunts reduced the incidence of blockage at the ventricular end of the shunt and the rate of shunt replacement [38].
However, Kestle et al. found that the success rate of shunts within 1 year of initial placement of an adjustable-pressure shunt was 67% compared with 61% for a nonadjustable-pressure shunt, with no statistical difference between the two. Reasons for shunt failure included obstruction (17%), excessive shunt (1.5%), ventricular separation (2%), and infection (10.6%). The results showed that adjustable pressure shunts did not reduce the rate of shunt failure in all patients with hydrocephalus [39], and similar findings were obtained in a study by Florian et al [40], and a study by Drake et al similarly confirmed that neither valve reduced the incidence of shunt failure [28].
Aschoff et al. found that even when the Codman Hakim shunt pressure was adjusted to 200 mmH2O, cerebrospinal fluid shunts due to siphoning when the patient was standing could reach 436 ml/h, which was attributed to the lack of an anti-siphoning device in the Codman Hakim shunt [41]. In addition, it has been reported that magnetic fields can alter the pressure of the Codman Hakim adjustable shunt valve while the patient is undergoing MRI, leading to shunt failure [42].
In recent years, to solve this series of problems, the ProGAV adjustable pressure shunt (Aesculap-Miethke, Tuttlingen/Potsdam, Germany) has been applied clinically, and its valve device consists of a gravity valve and an adjustable valve gate, and the opening pressure of the gravity valve can be set to 15, 20, 25, or 30 cmH2O, but its pressure cannot be readjusted after surgery. The pressure of the adjustable valve can be adjusted in the range of 0-20 cmH2O and can be re-adjusted outside the body after surgery. This design allows the valve pressure to remain constant when the patient is lying down and standing up, thus ensuring the stability of cerebrospinal fluid flow. In addition, the ProGAV adjustable shunt has a locking device that prevents valve pressure changes due to exposure of the valve to magnetic fields [43,44]. Current studies have confirmed its safety and effectiveness, but further research is needed to determine whether the incidence of shunt failure is lower in ProGAV adjustable pressure manifolds than in other manifolds [44].
The ideal shunt should have the following points: 1. the valve opening pressure is close to the pressure of cerebrospinal fluid outflow under normal physiological conditions. 2. the flow of cerebrospinal fluid should remain stable under different pressure conditions. 3. the flow of cerebrospinal fluid and the opening and closing pressure of the valve should not change due to changes in body position, body temperature, external pressure and cerebrospinal fluid pulsatile pressure. 4. the valve can stop the reverse flow of cerebrospinal fluid [42]. Unfortunately, however, there is still no shunt valve with clear advantages [26].
Enger et al. showed that the shunt device replacement rate was 69% 10 years after the first shunt [45]. a retrospective study by Berry et al. found that 37% (1307 patients in total) of patients required at least 1 shunt device replacement and 20% required more than 2 shunt device replacements within 5 years after shunt [27].
Berry et al. found that the low age of the patient at the time of first shunt placement was a risk factor for the need for shunt device replacement. In patients younger than 30 days, 49.2% required shunt replacement. 35.8% of patients between 30 days and 1 year of age required shunt replacement. However, in patients older than 1 year of age, the shunt replacement rate was 28.6% within 5 years of the initial shunt placement. In patients with spina bifida, 22% of patients required multiple shunt replacements after the initial shunt. The authors concluded that age less than 80 days was a risk factor for the need for multiple shunt device replacements. Patients with obstructive hydrocephalus had a higher rate of shunt device replacement. Patients with spina bifida are at potential risk for multiple shunt device replacements [46]. Therefore, shunt device replacement is not a good treatment for shunt failure.
Cinalli et al. evaluated the efficacy of treating shunt failure with endoscopic third ventriculostomy in patients with obstructive hydrocephalus, and in 76.7% of 30 patients, no further shunts were required [47].
Buxton et al. analyzed 88 patients who underwent endoscopic third ventriculostomy after shunt failure. In the 3-year follow-up after endoscopic third ventriculostomy was performed, the overall success rate was 52%, and in patients with obstructive hydrocephalus, the success rate could reach 73%. Comparison showed that patients undergoing endoscopic third ventriculostomy after shunt were as safe and reliable as those undergoing endoscopic third ventriculostomy for the first time [48].
O’Brien DF et al. reported that the success rate of endoscopic third ventriculostomy for shunt failure was 70%. They suggested that this is due to the fact that lateral ventriculoperitoneal shunts can alter the kinetics of cerebrospinal fluid circulation, which can lead to narrowing of the conduit after shunt, which makes obstructive factors the main cause of hydrocephalus after shunt failure, so endoscopic third ventriculostomy can be an effective alternative treatment [49].
Similar results were obtained in a study by Liana Beni-Adani et al. They concluded that hydrocephalus in children usually has both obstructive and abnormal absorption causes, and abnormal cerebrospinal fluid absorption is often due to factors such as infection and hemorrhage, but these factors that cause abnormal cerebrospinal fluid absorption gradually disappear with age, so endoscopic third ventriculostomy can achieve good results [24].
3. 3D-CISS and Cine PC in the diagnosis of hydrocephalus
In recent years, the use of 3D-CISS (3D-constructive interference in the steady state) sequence has provided important information for the diagnosis of hydrocephalus, the selection of treatment plan and the determination of the patency of the fistula after endoscopic third ventriculostomy.T2-weighted images cannot show the fine structure of the ventricular system well due to the poor spatial resolution and the influence of cerebrospinal fluid flow. The 3D-CISS sequence uses the effect of heavy T2WI to highlight the signal of cerebrospinal fluid and enhance the signal contrast between cerebrospinal fluid and brain tissue, and the higher spatial resolution of the 3D-CISS sequence can achieve the effect of noninvasive ventriculography and can clearly show the membrane structures in the ventricles and brain pools, such as Liliquist’s membrane, aqueduct, septum, and interventricular pore septum. septum, interventricular foramen septum, etc., and most of these membranous structures cannot be detected on T1 and T2 images. This provides an accurate basis for determining the cause of obstruction, the site of obstruction, and the choice of treatment options [11,50].Laitt RD et al. found that in patients with post-infectious and posthemorrhagic hydrocephalus, the presence of some membranous structures between the slope and the basilar artery (the same structures can be seen endoscopically), which interfere with the normal flow of cerebrospinal fluid, can be seen by 3D-CISS sequence images [ 10]. Based on this information, disruption of these membranous structures while performing endoscopic third ventriculostomy was performed, and postoperative follow-up revealed complete resolution of hydrocephalus symptoms.
A study by A. Dinçer et al. similarly found that only 57 membranous structures could be identified on T1 and T2 images of 134 patients, whereas 157 membranous structures could be identified on 3D-CISS sequence images of these patients. Based on the information provided by T1 and T2 images, 46 patients were diagnosed with traffic hydrocephalus, and based on the information from 3D-CISS sequence images, membranous structures were present in the ventricular or ventricular pool system in 26 of these 46 cases diagnosed with traffic hydrocephalus, and the authors performed ventriculoscopy in all of these 26 patients and found the presence of septa in the ventricular or ventricular pool system at sites similar to those shown on 3D- CISS sequence images, and all of these patients underwent endoscopic third ventriculostomy and had effective control of hydrocephalus [50].
In addition, the patency of the fistula opening is an important component of postoperative follow-up. Imaging evaluation of the effect of endoscopic third ventriculostomy can be done by the size of the ventricles, the fluid flow-void effect at the fistula and the quantitative measurement of cerebrospinal fluid flow at the fistula according to the Cine PC sequence.
Bargalló N et al. found no correlation between ventricular size and postoperative symptom improvement after endoscopic third ventriculostomy. They found that only 38% of patients with postoperative symptomatic improvement had a smaller ventricle than before surgery. However, in patients with aqueductal stenosis due to tumor compression, the ventricles shrank to varying degrees in all patients after surgery [12]. This suggests that the ventricles can be significantly smaller after endoscopic third ventriculostomy in patients with acute hydrocephalus, but in patients with chronic hydrocephalus, the postoperative changes in the ventricles are not significant. Therefore, the patency of the fistula opening cannot be accurately determined by the ventricular size alone.
On the 3D-CISS sequence and T2-weighted images, we can determine the patency of the stoma based on the signal of fluid flow at the stoma. In up to 50% of the patients with a patent stoma based on fluid flow signals, the clinical symptoms did not improve. In contrast, quantitative measurement of cerebrospinal fluid flow at the stoma using the Cine PC sequence revealed that in these patients, the stoma was patent but their cerebrospinal fluid flow was low [12].
The Cine PC sequence allows for a noninvasive study of cerebrospinal fluid flow and a more accurate determination of cerebrospinal fluid flow rate, flow and direction.The Cine PC sequence can be used to select the stoma as the area of interest and to determine the direction and flow rate of cerebrospinal fluid flow at the stoma during a cardiac cycle using relevant software. The total amount of cerebrospinal fluid entering and exiting the fistula during a cardiac cycle is calculated from the relevant data. The patency of the fistula is then determined.
Using the Cine PC sequence, Bargalló N et al. found that patients with cerebrospinal fluid flow >75 mm3 at the fistula opening during one cardiac cycle showed improvement in their hydrocephalus symptoms; if the flow was <75 mm3, their postoperative hydrocephalus symptoms did not improve significantly [12]. Therefore, for measuring the patency of the fistula opening after endoscopic third ventriculostomy, the use of the Cine PC sequence provides more and more accurate information than other methods.