Is Endoscopic Third Ventriculostomy a Feasible Option or Ventriculoperitoneal Shunt a Safer Bet for the Treatment of Posttraumatic Hydrocephalus? A Gap Time Model-based Algorithm
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.294832
Source of Support: None, Conflict of Interest: None
Keywords: Endoscopic third ventriculostomy, hydrocephalus, post traumatic hydrocephalus, traumatic brain injury, ventriculoperitoneal shunt
Traumatic brain injury (TBI) can lead to a grave complication like posttraumatic hydrocephalus (PTH). Its incidence has been reported to be 0.7%–51.4%., PTH can lead to progressive accumulation of cerebrospinal fluid (CSF) in the ventricles, leading to their enlargement at the expense of normal brain parenchyma. PTH adds to the morbidity and mortality of TBI, and even more so, if left untreated. Management options include observation, ventriculoperitoneal shunt (VPS), thecoperitoneal shunt (TPS), or endoscopic third ventriculostomy (ETV). These surgical procedures have a high failure rate in the treatment of PTH. The reason ranges from communicating hydrocephalus (HCP) (low-pressure gradient), infection, and presence of blood products in the CSF or CSF fistulae formation following trauma.,,,,
After going through literature, we felt that the clinical data are still insufficient to provide practicing guidelines to clinicians for the management of this difficult condition. There is not enough evidence to suggest usefulness of VP shunt over ETV in PTH or vice versa. We did an ambispective analysis of PTH patients to evaluate the usefulness of VPS and ETV in the treatment of PTH and to establish the indications for their usage. We present our findings in the backdrop of searched literature and attempt to formulate an algorithm, which may help clinicians in treating these cases.
During the study period (September 2012 to December 2016), around 2500 cases of TBI were treated at our institute. 141 patients of consecutively diagnosed PTH were included in this study. It is an ambispective study of nonrandomized, consecutively treated patients of PTH in which patients were analyzed according to the choice of procedure which was based on surgeon's discretion. These patients were segregated into two groups: ETV group comprised patients in whom ETV was done as a primary procedure and VPS group patients had undergone VPS as a primary procedure for PTH. The two groups were compared to analyze the differences in radiological and clinical outcome.
Informed consent for respective procedures was taken for all patients. Ethical clearance for the study was taken from the institute's ethical committee. Patients without follow-up imaging, electronic treatment records, and those unwilling to participate were excluded from the study. The patients in whom HCP was not a consequence of TBI were also excluded.
All patients meeting the inclusion criteria underwent a detailed clinical and radiological evaluation. The findings were recorded in a structured pro forma. All patients underwent mandatory preoperative noncontrast computerized tomography (NCCT) head as per our institute protocol. No patient was prescribed extra NCCT head for this study. NCCT images were used for comparing the changes in ventricular size. We considered the change in the maximum temporal horn diameter >2 mm, maximum third ventricular (TV) diameter >2 mm, and maximum frontal horn (FH) diameter >5 mm as significant.
Since many patients with decompressive craniectomy (DC) developed PTH, calculating Evan's index was going to be a challenge. Therefore, we used a modification of it and measured the index as maximum distance of FHs divided by maximum anteroposterior diameter at the same level. It was considered as significant when the change was >0.03 [Figure 1].
Patients were followed up as per the advice of the treating physician. All patients who came for follow up were evaluated by a blinded observer, who was not part of the operating team at the time of surgery. Those who were unable to come for follow-up due to illness and social or economic reason were telephonically interviewed and answers by their immediate relatives or caretakers were recorded in the form of a self-designed and administered questionnaire.
The data were analyzed by Stata 14, StataCorp LLC, Texas, USA and presented in mean/standard deviation/median (min–max) and frequency (percentage). Continuous variables were compared using independent t-test (following normal distribution)/Wilcoxon rank-sum (Mann–Whitney) test (skewed data). Chi-square test/Fisher's exact test was used for categorical variables. For the sake of analysis, every procedure was considered as an individual event. Thus, 175 procedures which were performed in 141 patients were included as separate events [Figure 2]. A gap time model was used to calculate the failure-free survival of each procedure taking into consideration the repeated nature of the outcome (procedure failure).
A total of 141 patients were included in the study. As mentioned in methodology, we presumed each procedure as a separate one; thus, a total of 175 procedures (ETV/VPS) were performed during this procedure. ETV group had 30 procedures and VPS had 145 procedures [Figure 2].
In the present study, the mean age in the ETV group population was 37 years, whereas, in the VPS group, the mean age was 32 years (P = 0.085). The ETV group had 28 males (93%) versus 125 male patients (87%) in the VPS group. There was no statistical difference in the two groups (P = 0.376). The most common cause of injury was road traffic accidents (RTAs) (83%). The pattern of injury in both the groups was similar and no statistically significant differences were observed (P = 0.760). The most common finding in CTs for all patients was of subarachnoid hemorrhage (SAH) (52%), followed by contusion (45%). No difference in radiological injury pattern was observed between the two groups (P = 0.25). SAH and intraventricular hemorrhage (IVH) are known to impair CSF absorption which can compound hydrocephalus. On comparing the two groups, there was no difference in the occurrence of SAH and IVH (P = 0.26). Other important radiological prognostic factors such as midline shift and effacement of basal cistern were also compared between the two groups; their incidence was also similar (P = 0.86 and 0.291, respectively). Bulging flap was the most common presenting symptom (70%). Presentations in both the groups were similar and had no statistical difference. Out of 141 patients in whom 175 procedures were performed, the number of patients undergoing DC was 118 (82%). Their distribution among the two groups was similar (P = 0.542). The comparison of the various group characteristics is summarized in [Table 1].
Previous surgical procedure for posttraumatic hydrocephalus
In the ETV group and VPS group, external ventricular drain (EVD)/ventricular access device (VAD)/TPS were done as primary CSF diverting procedure in 17% and 16% of cases, respectively. There was no statistically significant difference in the two groups (P = 0.76).
Postoperative fever was the most common complication. Postoperative sepsis was observed in a statistically higher number of patients in the ETV group [Table 2].
Clinical improvement after surgery
In the ETV group, 37% of cases showed improvement versus 73% of cases in the VPS group. A statistically significant number of patients had improvement in presenting symptoms in the VPS group as compared with the ETV group (P = 0.001).
Need for resurgery/other surgeries
In the ETV group, 18 (60%) patients required secondary CSF diversionary procedure. In the VPS group, 26 (17.9%) patients required secondary CSF diversionary procedure. This shows that the requirement of secondary procedure following ETV for PTH is significantly higher (P = 0.001).
Glasgow coma scale at discharge and Glasgow outcome score at 6 months
There was no significant difference in Glasgow coma scale (GCS) between the two groups at the time of discharge from the hospital. Similarly, on comparison of the Glasgow outcome score at discharge and at 6 months of follow-up, there was no significant difference between the two groups.
Cox regression and gap time model analysis
Cox regression analysis was used to understand the effects of multiple (clinical) variables on the survival probability and risk of failure. The analysis in [Table 3] shows that the hazard ratios for poor GCS, previous CSF infection, and postoperative meningitis were 2, 2.1, and 7, respectively, and these were found to have a significant effect on the failure-free period of the procedure. Sex, primary CT scan findings, presence of SAH/IVH/mass effect, primary management, or primary CSF diverging procedure failed to show a similar effect. On comparing the probabilities of failure-free period of ETV versus VPS, the chances of VPS failure are 61% less than ETV. The factors affecting the probability of failure free period of the procedures have been summarized in [Table 3].
Kaplan–Meier survival estimate curves
Kaplan–Meier survival analysis was done and illustrated as under for various variables such as preoperative GCS [Figure 3]a, previous CSF infection [Figure 3]b, and postoperative meningitis [Figure 3]c, which show a significant effect on the failure-free period of the procedure.
PTH occurs due to imbalance in production of CSF and its drainage, leading to progressive distension of the ventricular system. Gudeman et al. reported the incidence of PTH between 1.5% and 29%. De Bonis et al. after their literature search had reported an incidence of 0.7%–51.4%. In our experience of treating >500 cases of TBI every year, the incidence of PTH in our center is around 20%–30%. However, there might be a reporting bias in it, because our center is a tertiary care center of the country and we tend to admit severe TBI cases only.
Like previous studies, the most common mode of injury in our study was RTA (83%), followed by fall from height (15%). Most of the patients in our study were male (87%) and in young age group (mean = 32.6 years). Similar findings have been reported by the study by Kamal et al. These findings only corroborate the fact that most victims of RTA are men. The incidence of PTH as such does not depend on sex. Since most patients in our study had already undergone DC (82%), the most common presenting feature was bulging flap (72%). Waziri et al. also described 88.2% incidence of PTH in DC. Magnetic resonance imaging (MRI) is the investigative modality of choice, when evaluating patients for PTH as it can help in determining CSF dynamics, especially cine MRI. One of the limitations of the current study is that MRI CSF flow studies could not be done due to resource constraints and poor clinical conditions of the patient. If available, MRI should be considered as the modality of choice for imaging, but its use should be tempered with clinical judgment because many sick patients require general anesthesia for MRI, which adds to cost, morbidity, duration of stay, and logistical challenges.
Treatment options for posttraumatic hydrocephalus
Surgical management of PTH depends on timely CSF diversion, which can be achieved by ventricular shunts (VP/ventriculoatrial/ventriculopleural, etc.), ETV, TPS, EVD, lumbar drain, and placement of Ommaya reservoir/VAD.
Role of ventriculoperitoneal shunt in treatment of posttraumatic hydrocephalus
There is overwhelming evidence to suggest that VPS is the treatment of choice for patients with communicating HCP.,,,, In cases of PTH, VPS has been shown to be as an effective treatment. Phuenpathom et al. reported DC as an important causative factor for PTH and advocated VPS as a treatment modality of choice that improved clinical and radiological picture. We also believe that VPS is an effective treatment modality. Our study has also shown greater improvement in patients of PTH undergoing VPS. However, there are general risks related to shunt surgery, which leads to shunt failure.
Role of endoscopic third ventriculostomy in the treatment of posttraumatic hydrocephalus
ETV has been a successful modality for treating obstructive hydrocephalus. Usefulness of ETV in communicating HCP is questionable due to very low success rate. Although a few studies have described usefulness of ETV in communicating HCP (e.g. NPH) as well. In patients with PTH, ETV may not be beneficial because of blood and its products in ventricles after TBI which can lead to blockage of arachnoid granulations.
In addition, the pressure gradient with every heartbeat is not created due to absence of bone flap, leading to impaired absorption. However, in PTH patients with shunt malfunction due to infection, ETV may be considered as an option. The basic advantage of ETV in such settings, where the rate of infection is high, is that it can provide shunt independence and chances of infection in such cases are low because there is no foreign body is in situ.
Factors affecting the outcome
Our study has shown that most important factors, affecting the outcome of patients with PTH, are preoperative clinical status (GCS), previous CSF diverging procedure, previous infection, and postprocedural meningitis, which is similar to the findings of previous studies.,
CSF infection is an important determinant of failure of procedure for PTH. This has been documented in our study where we have shown that in the presence of infection, chances of failure are statistically higher. This is probably due to higher scarring of absorptive pathways and blockage of alternative pathways (VPS or ETV stoma) due to coagulum or high protein contents. Performing ETV in such cases is also challenging, because occasionally TV floor walls are thickened and fibrotic. We do not perform shunt, until CSF values are nonmeningitic and two cultures are sterile in suspected cases. We have shown that preoperative and postoperative infection plays a major role in outcome of the patients.
Management of posttraumatic hydrocephalus An algorithmic approach
Based on our existing knowledge and the experience gained by this study, we have created an algorithm that can help clinicians in deciding the best line of management for cases of PTH [Figure 4]. We recommend that if patient of TBI has any deterioration in sensorium, bulging flap (in case of DC), memory, visual, motor disturbances, etc., we should keep the suspicion of PTH and investigate accordingly. Proper clinical evaluation including fundus examination, lumbar pressure monitoring, NCCT head, and MRI brain should be done. In case of suspected meningitis, CSF cytobiochemical analysis should be done to know the presence of infection.
In case of CSF values being meningitic, EVD either medicated or nonmedicated should be considered and antibiotics therapy should be continued. VPS should be considered only after two consecutive sterile CSF cultures and nonmeningitic CSF values. In case of persistent infection of CSF or blood in ventricles, consider VAD and once values become nonmeningitic and cultures are sterile or blood has resolved, then if necessary, convert to shunt.
VPS is the treatment of choice for diagnosed cases of PTH. Preferable option is programmable VPS with antibiotic impregnated tubing. If a person is unable to afford it, we recommend the use of a fixed pressure valve system with silicone elastomer tubing.
In case of failure, try to identify cause and treatment should be tailored accordingly. Suitable options can be revision of VPS shunt, TPS, or ETV. We recommend the use of ETV as a last resort, when multiple shunt failures and TPS fail to improve the clinical condition.
VPS is an effective modality for the management of PTH and has a much lower failure rate as compared to ETV. We recommend ETV's use as a salvage procedure, not as a procedure of choice. ETV can be considered in difficult situations of recurrent shunt malfunction where infection due to a presence of foreign body is a cause of concern.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3]