|Home | About | Journals | Submit | Contact Us | Français|
The aim of this investigation was to evaluate the suitability and usefulness of the Stealth Station™ intraoperative guiding system (Medtronic Sofamor Danek, Memphis, TN) in a variety of indications. Eleven intraoperative image–guided procedures were performed for anterior or lateral skull base lesions. The most common neurosurgical approaches included frontal, coronal, and parietotemporal access. Neuronavigation reliably allowed the extent of tumor configuration and risk zones (e.g., blood vessels) to be visualized. Thus, gross tumor resection was achieved in 6 of 7 patients and facilitated reconstruction by the maxillofacial surgeon, resulting in radiologically symmetrical and clinically satisfying results. Postoperatively, one patient was blind from a continuity defect of the optic nerve caused by a bone fragment. Despite destruction of anatomical landmarks related to tumor invasion or intraoperative bone removal, neuronavigation proved helpful in the reconstruction of bony structures. Overall, the use of neuronavigation in interdisciplinary surgery for complicated tumors or trauma of the anterior or lateral skull base allows more radical resection associated with less morbidity.
Intracranial navigation and computer–assisted surgery are of major interest for several specialties. Their use is established in neurosurgery and orthopedics1 and recently they have been introduced in otolaryngology and craniomaxillofacial surgery. The fields of application have predominantly been sinus surgery, craniofacial malformation, and resection of tumors near important anatomical structures such as the anterior cranial base or optic nerve.
Different systems have been used and their accuracy has been documented. These systems are known to increase accuracy and to facilitate exposure and precise localization of tumors or traumatic lesions. Using computed tomographic (CT) data the navigation systems offer additional benefit for reconstructive surgery. The available systems use either electromagnetic (radiofrequency) or optical (infrared) signals to localize instruments within the surgical field.
Marmulla et al2, 3 showed that the accuracy of the Polhemus Inc. electromagnetic system4 is unsuitable for computer–assisted surgery ranged between 0.6 mm and 24.3 mm. It delivers serious spherical deviations in the presence of metal, rotating instruments, and circuits. Metson et al.5 compared two (InstaTrak® and Stealth Station™) image–guidance systems for sinus surgery. In all cases, intraoperative anatomical localization was accurate to within 2 mm at the start of surgery. During the operative procedure, accuracy degraded by 0.89 mm ± 0.20 mm. However, the authors concluded that both systems proved valuable for anatomic localization during sinus surgery.
This study attempted to evaluate the suitability and usefulness of the infrared–based Stealth Station™ (Medtronic Sofamor Danek, Memphis, TN) in skull base surgery.
CT or magnetic resonance imaging (MRI) data were acquired using a Somatom Plus 4 Volume Zoom and a Magnetom Open (both Siemens AG, Erlangen, Germany). CT/MR volume data (CT volume data using a spiral mode with reconstruction of 1 mm thickness and 1 mm increment, FOV 200 mm, 120 kV, 300 mAs; MR volume data using a FLASH 3–D gradient echo sequence: fast low–angle shot, TE 7.0 ms, TR 16.1 ms, flip angle 30°, slab 168 mm, 112 slices, FOV 250 mm, matrix 256 × 256) were transferred to the navigation workstation (SGI 02, Silicon Graphics, Mountain View, CA), where the data were reformatted and displayed in three orthogonal planes.
Stealth Station™ is an optical–based system, which uses an infrared camera array to monitor instrument and head position. This camera tracks the location of light–emitting diodes (LEDs) mounted to surgical instruments and to a reference arch attached to the Mayfield head clamp. Throughout the entire surgery, rigid fixation is necessary to maintain the correlation between space coordinates of the navigation system and the patient's coordinates as defined during registration.
The registration is initiated using surface landmarks, which are correlated with the same points on the three–dimensional CT/MRI surface reconstruction of the skin. The location of the navigation probe in the surgical field is displayed in the reformatted CT and MR data set. The quality of registration is very important for obtaining acccuracy from the system. Neuronavigational accuracy was documented by mean registration error and repeated landmark checks.
In 1999 11 patients (7 women, 4 men; mean age, 39.5 years; age range 18 to 61) underwent an interdisciplinary neurosurgical–craniomaxillofacial approach using neuronavigation guidance (Table 1). Seven patients had extensive anterior or lateral skull base tumors. Two of these patients had a recurrent meningioma. One had an eosinophilic granuloma and one had a fronto–orbital rhabdomyosarcoma. Two had an adenoid cystic carcinoma and one patient had an extensive fronto–orbitotemporal fibrous dysplasia. Two patients had comminuted trauma of the anterior or lateral cranial base, and intraoperative imaging supported decompression of the optic nerve. Two patients had secondary malformations related to displacement of bony fragments (malar bone) during primary resection. All patients underwent postoperative contrast–enhanced radiological studies to assess the degree of surgical resection, the outcome after reconstruction of the bony structures, or both.
The neurosurgical approaches included frontal, coronal, and parietotemporal access. Clinical features after surgical treatment varied according to the site of the lesion. The most common clinical findings were headache, facial malformation, visual loss, and seizures.
The neuronavigation allowed gross tumor resection in six of the seven patients, helping to preserve important anatomical structures surrounding the tumor such as the optic nerve. Use of the neuronavigation system facilitated bony reconstruction of the anterior and lateral skull base, resulting in symmetrical and clinically satisfying outcomes. Especially when bone substitutes were used, the contour could be corrected more easily by comparing the radiological data with the clinical impression. There was no permanent postoperative deficit in our patients, but one patient remained blind after the primary trauma displaced a bone fragment and caused a continuity defect in the optic nerve.
Setting up the neuronavigation system required an additional 15 minutes of anesthesia time. However, the neuronavigation guidance helped to save operation time by speeding the surgeon's orientation to the anatomy in the surgical field.
At the beginning of the operation, the neuronavigation system provided an accuracy of registration of 0.61 to 2.70 mm (mean, 1.21 mm ± 0.72 mm). Landmark checks during surgery documented that the Stealth System™ remained stable.
The quality of the initial registration is monitored by the root mean square (RMS)–error given by the software. The registration plausibility is then checked with known anatomical landmarks. Before craniotomy bone divets are drilled for continuous surveillance of the positional shift and their x–, y–, z– coordinates are checked for intraoperative changes.
A 20–year–old male presented with progressive visual loss. In early childhood, he had undergone irradiation of a retinoblastoma, and 4 years before presentation, an additional leiomyosarcoma of the temporal region was irradiated. Radiological studies showed an extensive recurrent leiomyosarcoma in the fronto–orbitotemporal region (Figs. (Figs.1A1A,,B).B). The neuronavigation system was used to guide tumor removal, to identify surrounding critical structures, and to reconstruct the malar bone, which had to be resected due to tumor infiltration (Figs. (Figs.2A2A,,BB).
A 51–year–old male who had experienced progressive facial asymmetry for many years presented to the neurosurgical department with seizures. Examination revealed a right supraorbital prominence and enophthalmus. A biopsy confirmed the diagnosis of fibrous dysplasia involving the frontal orbital bone (Fig. 3). Ophthalmologically, early visual deterioration was evident. The patient underwent surgical removal of the monostotic fibrous dysplasia (Figs. (Figs.4A4A,,B).B). Neuronavigation supported and facilitated identifying the optic nerve and the symmetrical reconstruction of the orbital roof and frontal bone using autologous split calvarial bone grafts. His postoperative course was uneventful and the cosmetic result was satisfactory (Figs. (Figs.5A5A,,BB and and6A6A,,BB,,CC).
Navigation systems are based on either mechanical, electromagnetic, or optical tracking. When electromagnetic tracking systems are used, ferromagnetic surgical tools must be avoided in the immediate vicinity of the sensor assembly to prevent field distortions. When these guidelines are followed, Watzinger et al6 believe that reliable navigation is possible and that deviations are below the overall repeatability of the electromagnetic tracking system (0.8 mm/0.15 degrees according to the specifications, Polhemus Inc.4). They used computer–aided navigation for secondary reconstruction of post–traumatic deformities of the zygoma. Successful applications of electromagnetic tracking devices in ENT surgery confirm this view.7
Optical systems require a clear line of sight to be maintained between the infrared camera, reference arch and pointer instruments.
Since the first description by Watanabe et al,8, 9 the accuracy and clinical usefulness of frameless systems based on an optical digitizer for neuronavigation have been assessed in various studies.10, 11, 12, 13 Germano et al14 found a mean accuracy of 2 mm in 170 neurosurgical procedures using the Stealth Station™.
Using a high–precision mechanical micromanipulator, Kaus et al15 detected a total technical error of 0.55 ± 0.64 mm when evaluating the different system–inherent sources of erroneous target localization of the LED–based neuronavigation system (Stealth Station™). In our study the precise intraoperative feedback proved helpful for orientation in the surgical site and increased the accuracy of recognition of tumor landmarks. It was especially helpful in the reconstruction of bony structures that had to be removed during tumor resection or that had to be repositioned due to malpositioning during earlier primary reduction of the fracture. Image–guidance systems can provide surgeons with accurate information for anatomical localization when surgical landmarks are poor due to extensive disease or prior surgery.
There have been several studies on the accuracy of navigation systems. Our findings correspond to those of Metson et al5 In our hands intraoperative anatomical localization at the start of surgery in all cases was accurate to within 1.21 mm and remained stable during the operative procedure. So–called brain shift,16 which may cause increasing inaccuracy during the procedure, is unimportant when treating patients with lesions of the anterior and lateral cranial base.
We conclude that image–guided surgery is particularly valuable for the treatment of anterior and lateral skull base tumors or trauma cases. Further indications must await future investigations.
The authors are to be commended for their clear demonstration of the value of navigational assistance during the resection of skull base lesions, particularly those associated with complex anatomy and indistinct margins. Their figures ((1,1, ,2,2, ,3,3, ,4,4, ,5,5, ,6)6) show excellent resections of fibrous dysplasia and at least one large skull base tumor. Their narrative convinces one that the Stealth™ unit was very helpful in indicating the location, particularly the margins, of the tumor to define the extent of resection. When the margins of tumors are visually distinct, the use of intraoperative navigation may not be so useful. However, in situations like those the authors present-when the edges of the abnormal anatomy are unclear or when critical structures traverse the areas of abnormality-navigational tools are likely to be beneficial, reducing operative time and increasing the safety of surgery.