gms | German Medical Science

Ventricular anatomy might provide difficult circumstances for accurate placement of cerebro-spinal fluid (CSF) catheters for ventricular peritoneal shunting in few cases. Slit-like ventricles are often associated with pseudotumor cerebri, also known as benign intracranial hypertension [1]. Preferable treatment in those cases is medical. However, CSF shunts might be necessary, if medical management fails, severe headaches persists or if neurological dysfunction, like deterioration of vision or visual field loss, develops. Placement of CSF shunts was shown to significantly decrease neurological deficits [2], [3]. Due to narrow internal CSF spaces, preferably, lumbo-peritoneal shunts were implanted. However, high incidence of shunt failure were reported in these patients [4], [5]. Moreover, in cases of intracranial mass lesions anatomical displacement of the lateral ventricles may occur. If CSF drainage is necessary, these conditions may restrict correct implementation of CSF drainage. To achieve optimal placement of ventricular catheters in these patients the technique of stereotactic placement was introduced [6]. This approach is restricted because of a complex intra-operative procedure with a longer duration of shunt placement with a higher risk for shunt infections. Since computer assisted procedures like frameless stereotaxy are nowadays available in most neurosurgical departments, this technique might be superior to conventional stereotaxy in these patients. Recently, the computer assisted approach was described using a pointer based placement of catheter with peel-away sheaths [7], [8]. We report of a simple solution for computer assisted catheter placement using a neuronavigation system to evaluate its feasibility and accuracy in an experimental setup and peri-operative experience.

Our department is equipped with a BrainLab^TM Vector Vision Square neuronavigation unit (Heimstetten, Germany). The system enables integration of different tools within the navigated procedure. We developed a stainless steel tube with an inner diameter slightly larger than the outer diameter of frequently used ventricular catheters (Christoph Mietke, Potsdam, Germany). The outside diameter of the tube was chosen to achieve optimal stability of the probe. A dynamic reference frame with passive, reflective markers is mounted to the probe. Using the calibration matrix the tube is registered for navigation of its trajectory (Figure 1 [Fig. 1]).

We investigated the feasibility of the tube in an experimental condition using a skull model. A computer-tomography (CT) image data set of the skull model was acquired. The planning workstation was used to define fiducial marker position and to plan an optimal trajectory to approach the area of the right frontal portion of the lateral ventricle. The surgical procedure was simulated in the operation theatre. Following registration of the skull, the tubes navigated position was validated in comparison to the pointers orientation and tip position. The tube was mounted indirectly to the operation table using a flexible retractor system. Within the navigation screen the probes view was used to bring the tube in the correct position in accordance to the pre-planned trajectory (Figure 2 [Fig. 2]).

Following achievement of a good work flow with the entire experimental procedure and after reaching sufficient accuracy, we performed navigated ventricular catheter placement in four patients. In all patients a 2 mm non contrast enhancing cranial CT scan (Lightspeed, General Electrics, USA) was performed. The data transfer to the workstation was achieved by an in-house network. The planning procedure enabled the definition of the optimal trajectory for a catheter placement. Therefore, the foramen monroi was used as target point, while the entry point was chosen individually on the surface of the frontal skull to receive the longest possible intra-ventricular position of the trajectory. Data transfer after the planning procedure was performed using a zip drive. Intra-operatively, the patients head was placed in a rigid position using a mayfield clamp. The position of the head was slightly turned and extended to achieve optimal conditions to place the distal parts of the ventriculo-peritoneal shunt. Spatial registration was performed using six fiducial markers placed on the convexity of the heads surface. Localisation of the skin incision and the burr whole was achieved using the pointer navigation. According to our experimental setup the navigated tube for catheter placement was placed in the correct position after exposure of the dura mater. Following dura incision and a small corticotomy the catheter was placed via the fixed tube. The length of catheter placement was chosen individually according to the length of the pre-planned trajectory. In all patients a postoperative cranial CT scan was performed to prove accurate catheter positioning.

The feasibility to integrate the tube as a navigated tool within the procedure was proven in the experimental setting. Accurate registration of the tube was achieved compared to the pointers orientation and tip positioning. The work flow for the entire setup within the operation theatre was easily established and took a total of 6 minutes. As far as navigation is familiar to the surgeon no significant additional learning curve was necessary for the technical setup. Hence, the setting for navigated placement of the ventricular catheter could be transferred intra-operatively in patients without any difficulties. So far this technique was used in four patients. Three patients suffered from pseudotumor cerebri, one patient had an intraventricular mass displacing the lateral ventricles. Two representative cases are demonstrated in the following:

Case 1: A 40 years old female patient suffering a pseudotumor cerebri. She presented with headache, bilateral papilloedema and an incomplete loss of the visual field. Angiography did not show any signs of sinus thrombosis. Preoperative cranial CT scan and T2 weighted MR imaging visualized bilateral slit-like ventricles. Conservative management including repeated lumbar taps and drainage did ameliorate the symptoms only temporarily. Postoperative CT scan verified correct intraventricular catheter placement in the right frontal horn (Figure 3 [Fig. 3]).

Case 2: The second patient was a 63 years old female patient. A compartmental hydrocephalus in the isolated left occipito-temporal horn due to an intraventricular tumour mass was visualized in a preoperative magnet resonance imaging (MRI) as well as in the thin sliced, reconstructed CT scans. The patient's progressing drowsiness was explained by a brainstem compression due to fluid collection within the isolated occipito-temporal portion of the left ventricle. We decided to implant a bilateral interventricular shunt from the occipito-temporal part of the left ventricle to the right frontal horn displaced by a midline shift of about 1.5 cm. Postoperative CT scan showed accurate localisation of the right frontal CSF catheter (Figure 4 [Fig. 4]).

In all patients the optimal placement of the catheter within the ventricle was achieved instantly. Clear CSF could be collected after the first puncture. Mean time of surgery was 55.5 minutes in the four cases. In the post-operative cranial CT imaging we could verify the correct catheter position compared to the pre-planned trajectory (Figure 3 [Fig. 3] and Figure 4 [Fig. 4]). Clinical improvement in all patients proved the reliability performance of the shunts in the immediate and later follow-up 6 months after surgery.

We report of a simple technical solution to achieve CSF catheter placement by using a computer assisted frameless stereotactic system. Our experience showed a feasible solution without exaggerated additional technical needs. In the experimental setup accuracy was reported and intra-operatively catheter placement was achieved immediately in all patients. We proved correct catheter placement by postoperative CT scans.

The navigated CSF catheter placement within shunt procedures is only necessary in selected patients with slit like or displaced ventricles. In the time period in which the surgery of the presented cases was performed (5/05 - 11/05), 42 ventriculo-peritoneal shunts were implanted in our department. Thus, we suggest navigated CSF catheter placement is necessary in only 9.5% of the patients according to our series.

Compared to other groups our solution with a navigated tube, which is able to be sterilized, no additional disposable items are necessary for the procedure. McGirt and co-workers reported of navigated catheter placement using peel-away sheath. The sheath is placed in the brain parenchyma with the help of the navigated pointer. After withdrawing the pointer the catheter was placed in the ventricle and the peel away technique enables the retraction of the sheath [7]. This technique was also feasible and easy to use according to the authors report. However, with the peel away sheath additional disposable items are necessary for this procedure.

In contrast to stereotactic procedures, neuronavigation does not need the time consuming intra-operative imaging with the anaesthetized patients. Hence, operation time can be reduced to a reasonable duration, compared to regular ventricular peritoneal shunt procedures and no significant prolongation is necessary. Since the amount of time for operation is a predictive factor for infection in ventricular peritoneal shunting [9], navigated placement of CSF catheters might decrease shunt infection compared to stereotactic placement of ventricular drainage in shunt implantation. As alternative for stereotactic placement of ventricular catheters the endoscopic shunt insertion trail (ESIT) investigated the potential benefit of endoscopic guided catheter placement [10]. Herein the notion was represented that endoscopic placement did not show any benefit compared to conventional catheter insertion, which was also not affected by the surgeons learning curve [10]. Compared to the endoscopic procedure, a navigated catheter placement has the clear advantage of a less invasive surgical approach.

Accurate placement of CSF catheters might overcome the need of lumbar peritoneal shunting, which is associated with a higher rate of shunt failures and the danger of overdrainage [4], [5]. Moreover, our philosophy proposes optimal catheter placement of the tip of the CSF catheter within frontal horn of the lateral ventricle to prevent contact with the plexus and thus dysfunction due to catheter occlusion. It has been reported that stereotactic CSF catheter placement in patients with pseudotumor cerebri could significantly reduce the rate of occlusive shunt failures in long term follow up [7]. For navigated catheter placement this data could not be delivered so far, since only a small amount of patients received this procedure. Long term follow up in a prospective investigation is needed to clarify this uncertainty.

We developed a simple and easy-to-use solution for exact catheter placement assisted by neuronavigation. Only in few patients the navigated procedure of ventricular catheter placement is necessary, if slit-like ventricles or distorted anatomy complicates the procedure of optimal placement. Further investigations are necessary to clarify if this technique is able to decrease the rate of catheter occlusions in hydrocephalus patients with slit-like or displaced ventricles.