Awake craniotomy for operative treatment of brain gliomas – experience from University Medical Centre Ljubljana

We included all consecutive patients who underwent surgery by awake craniotomy due to gliomas from years 2015 to 2019 at the Department of Neurosurgery of Ljubljana University Medical Centre. The study was approved by the National Medical Ethics Committee of the Republic of Slovenia. All the procedures were performed in accordance with the Declaration of Helsinki. Awake craniotomy was considered in selected patients with the following preoperative morphological images: I) a tumour near or within the language brain areas (Broca's and Wernicke's areas, angular gyrus), II) all cases of insular tumours and III) selected patients with lesions near or within primary motor cortex or corticospinal tract.

Each patient was assessed preoperatively by the anaesthesiologist, neuropsychologist, and neurosurgeon. A collective decision was made whether the patient was suitable for awake craniotomy. The inclusion criteria encompassed a good clinical, physical, and affective condition. The contraindications for awake craniotomy were a non-compliant patient (e.g., due to old age or unfavourable psychosocial factors), the potential for breathing problems during surgery (e.g., known sleep apnoea, significant obesity), and important preoperative dysphasia (i.e., the patient names less than 80% of objects presented at four-second intervals). Informed consent was obtained from all study participants.

The preoperative planning consisted of the acquisition and analysis of preoperative morphological and functional images, the selection of surgical approach and trajectory to the lesion, and the planning of the extent of the resection. MRI was performed using a 3-Tesla clinical scanner (Siemens Trio, Siemens). The diagnostic imaging included T1- and T2-weighted sequences. For 3D-imaging, we used a post-contrast T1-weighted 3D-fast spoiled gradient recalled (FSPGR) sequence (a series of 124 images, 1.4 mm thick with a matrix up to 512x512 of 240mm field of view). A fMRI was used to define the primary motor and speech areas in all patients. We used single shot echo planar imaging in a transverse plane (TR 3000/TE 40, a series of 43 images, 3 mm thick with a 64x64 matrix with 200 mm field of view) during which the patient performed a motor task (a self-paced sequential tapping of the thumb against each finger) or a speech task (verbal fluency and verb generation). A general model-based statistical analysis (SPM 12) was used to measure the extent of cortical activation.

To plan the surgical procedure, we used computer assisted 3D-visualization by neuronavigational software (Stealth Station S7 Surgical Navigation System, Medtronic) and 3D-Slicer software (http://www.slicer.org).¹⁴ The 3D-visualization of the medical images allowed us to perform preoperative 3D-planning, i.e., to interactively present relevant anatomic structures as 3D-objects in virtual space on the computer screen, to define surgical targets, to perform quantitative measurements, and finally to select the most suitable surgical approach or trajectory to the lesion.¹⁵ Interactive 3D-preoperative planning started with defining the tumour model characteristics (size, volume, extension) and their relationship to the models of the cortical surface and functional data. Based on the available data, we planned the position and size of the trepanation opening, the sites of intraoperative electrophysiological testing, and the preliminary EOR.

We used awake-awake-awake technique of awake craniotomy - i.e. procedural sedation and analgesia with dexmedetomidine in combination with scalp block. Initially deksmedetomidine infusion (200 μg in 50 ml of 0.9% NaCl; rate 0.2–1.5 4 μg/kg/h) was started to sedate the patient followed by scalp block. Scalp block was performed by local infiltration of the scalp nerves (i.e. n. supratrochlearis, n. supraorbitalis, n. auriculotemporalis, n. zygomaticotemporalis, n. occipitalis minor and major, n. auricularis major) to enabled effective analgesia for skin incision and craniotomy.¹⁶ For local infiltration of the nerves we used 1–3 ml of the mixture of levobupivacaine (Chirocaine 5 mg/ml solution, Abbvie Pharmacy) and xylocaine with adrenaline (Xylanaest 2% with epinephrin 1: 200000 solution, Kemofarmacija d.d.). Adrenaline causes vasoconstriction and thereby delays absorption of local anaesthetics, increases duration of anaesthesia and prevents local bleeding after skin incision.

In addition, we introduced therapeutic communication with medical hypnosis leaving patients less sedated, more competent during the entire surgical procedure without stress.¹⁷ Hypnosis session was usually carried out 1 to 3 days before surgery, to gain patients approval and confidence and to teach the patient how to construct an imaginary place where they can feel safe and protected.

Anaesthesia was conducted by two senior anaesthesiologists. The standard intraoperative monitoring was used. All patients had dedicated intravenous and arterial lines, a Foley catheter, and a nasal oxygen catheter. Hypothermia was prevented with use of warming blankets. Cefazolin 2g was used intravenously for perioperative antibiotic prophylaxis.

The patients lay supine with their head rotated approximately 45 degrees to the side. They were encouraged to find the most comfortable position and were generously padded as necessary. The head was fixed in a Mayfield clamp under a scalp block combined with dexmedetomidine. Additional local anaesthesia was applied along the skin incision line and between the dural sheets before the skin and dura incision. Sedation during the procedure was maintained with dexmedetomidine, and painful phases were treated with boluses of remifentanil. Bispectral Index Scale (BIS) was used for sedation monitoring. After the surgical procedure, all patients were transferred to the intensive care unit for overnight observation.

Tumour removal was performed in accordance with classical microsurgical techniques. In all cases, microsurgical microscope was employed. During the surgical procedure, frameless neuronavigation, intraoperative neurophysiological testing and augmented malignant tumour visualization with 5-aminolevulinic acid (5-ALA) were used.

For precise intraoperative localization of the tumour, we used neuronavigation with a Stealth Station S7 (Medtronic) navigation station, thus defining the size and position of the trepanation opening and the morphological tumour boundaries on the cortical surface. During tumour removal, neuronavigation was used for approximate depth orientation, and as a help to assess the EOR intraoperatively.

Injury to the eloquent regions was avoided after they had been identified by intraoperative neurophysiological testing. For cortical stimulation, a bipolar electrode was used with biphasic electrical current (50 Hz, 0.5 ms, 4 to 10 mA). For subcortical simulation, a monopolar electrode on the tip of aspirator was used, with a train of five pulses in row (a train of five pulses at 200 Hz, 0.5 ms, from 2 to maximum 20 mA). The patient's speech was monitored by an examiner, who talked to the patient and detected potential speech difficulties. During the stimulation of the cortical surface, the patient named objects on pictures or repeated words after the examiner. When stimulating the speech area, the speech changed or ceased. The stimulation of the primary motor area on the cortical surface resulted in motor evoked potentials (MEPs) detected in the corresponding muscles of the patient's body, and discomfort and/or movements, as reported by the patient. Subcortical stimulation was performed to define the distance from the corticospinal tract in patients where the tumour was in its vicinity. During tumour removal, the subcortical white matter was continuously stimulated with a monopolar electrode, and MEPs were monitored. At the beginning of the resection, the stimulation current was set at 20 mA. When MEPs were detected during the progression of the resection, the stimulation current was gradually reduced. When the MEP response was already induced by a 3–5 mA current, the resection in that direction was stopped, due to the proximity of the corticospinal tract.¹⁸

In cases where the preoperative diagnostic images showed the characteristics of malignant tumours, 5-ALA was used. The fluorescence of the tumour tissue was observed under the operating microscope (OPMI Pentero 900 Microscope, Carl Zeiss).

The demographic data of the patients were collected before surgery. According to the American Society of Anaesthesiologists, their physical status classification system ASA score was noted and KPS was assessed in all patients. In 12 patients, health-related quality of life (HRQoL) was assessed by 36-Item Short Form (SF-36) health survey.^19,20

During the surgery, we measured duration of anaesthesia, duration of the operative procedure and duration of the neurophysiological testing. Patient cooperation during surgery was assessed by a neurosurgeon and neurophysiologist and graded on scale from 0 to 10 (modified 11-point numerical rating scale, where 0 represented poor cooperation and 10 excellent cooperation). Patient comfort and pain were assessed by the anaesthesiologist. Pain was assessed by visual analogue scale (VAS) from 0–10 (0 meaning no pain, 10 meaning intolerable pain). Comfort was assessed by modified VAS scale from 0–10 (0 being the least and 10 being the most comfortable). Intraoperative complications were recorded.

After surgery, we monitored postoperative neurological status and noted non-neurological complications. The level of effort needed for the whole procedure, as reported by the patients, was assessed by anaesthesiologist 2–5 days after surgery, by modified VAS scale from 0–10 (0 being the least and 10 being the most difficult). KPS was assessed 1 week after surgery and the patients were asked to describe what was the worst experience during the awake craniotomy. A contrast enhanced brain MRI scan was performed the day following the surgery to assess EOR in all patients. In cases of malignant gliomas (AA, GBM) EOR was assessed on T1 weighted contrast enhanced MRI images. In those cases, gross total resection (GTR) was defined as no residual enhancement on postoperative images, near total resection (NTR) was defined as rim enhancement of the resection cavity, and subtotal resection (STR) was defined as residual nodular enhancement. In cases of LGG, EOR was assessed on T2 and FLAIR weighted postoperative MRI images. In those cases, GTR was defined as no hyperintense signal changes on postoperative images, STR was defined as persistence of hyperintense signal around resection cavity, and biopsy resection (BR) when approximately more than 50% of hyperintense signal remained on postoperative images. Length of hospital stay (LOS) was noted.

Three months after the surgery, the KPS score was assessed in all patients and the HRQoL was assessed by SF-36 health survey in the same 12 patients as before surgery. All patients were asked, if they would be willing to undergo awake craniotomy again if necessary and their answers were noted.

The means or ranges of continuous variables are presented, and categorical data are summarized as counts. The differences between the KPS scores before, after and 3 months after surgery were evaluated using One Way ANOVA. The differences between the SF-36 health survey subscales before and 3 months after surgery were evaluated using Wilcoxon test. A p-value of less than 0.05 was considered statistically significant. Data were analysed by SPSS 13.0 software package (IBM Corp., Armonk, NY, USA).

The demographic and preoperative data are showed in Table 1. During the 5-year study period, we performed awake craniotomies in 24 patients with glial brain tumours. They were 22 to 60 years old. In all cases, the main presenting symptom was epileptic seizure. In addition, in two patients, a mild dysphasia, and in one a disorientation was present. The main comorbidities in patients were well controlled diabetes and arterial hypertension. Four more patients were considered for awake craniotomy but were later found to be unsuitable and underwent surgery under general anaesthesia, one due to anxiety, one due to morbid obesity, one due to asthma and one due to psychosis.

The preoperative fMRI was successful in all patients and revealed speech dominance on the left side in all but one case in which it was bilateral. In 4 cases with tumour in frontal-central region (i.e., near or within primary motor region) the fMRI was used also to define primary motor regions (Figure 1).

FIGURE 1.

Intraoperative and early postoperative outcomes are presented in Table 2. The longest surgical procedure lasted 5 hours and 20 minutes. Patient cooperation during surgery was in majority of the cases excellent. Discomfort during surgery reported by the patients was generally described as just a little uneasy and/or just starting to bother. Pain perceived during surgery was regarded as moderate. In two patients, however, we observed significant fatigue during surgery that started about 2.5 hours after the start of the procedure, and it was so intense that further neurophysiologic testing was not feasible. During cortical stimulation, we observed epileptic seizure activity in 8 out of 24 patients (33% of the patients). In one of these patients, further testing was not possible, due to limited compliance after the seizure. Neurophysiological testing by cortical stimulation was therefore successful in identifying the cortical language areas or primary motor cortex in all but three cases. Subcortical stimulation was performed in cases with insular and frontocentral located tumours. The corticospinal tract upon stimulation with a 20 mA current or less was detected in 8 out of 11 tested patients (72% of the patients).

After the surgery, the neurological worsening was observed in 3 out of 24 patients (13% of the patients) and the other postoperative complications were present in 6 out of 24 patients (25% of the patients) (Table 2). Postoperative pain was mild in almost all cases. Level of the effort during surgery experienced by the patients was assessed and reported to be mild to moderate. When asked “what was the worst experience during the awake craniotomy”, 15 patients said “nothing”, four patients said, “bone cutting”, two patients said, “Mayfield clamp placement”, and the remaining three patients said, “scalp block”, “craniofix bone flap fixation” or “waiting for the end of surgery”. GTR was achieved in 29% (4 out of 14 cases) in cases of LGG, and in 60% (6 out of 10 cases) in cases of GBM and AA (Figure 2).

FIGURE 2.

All patents were discharged home. The time of postoperative hospital stay was from 3 to 9 days for most of the patients, and 23 days for one patient.

Three months after surgery all patients were under postoperative oncological treatment with radiotherapy and/or chemotherapy. All patients were still on a sick leave, and none yet returned to work. The average KPS scores after surgery and 3 months after surgery were slightly lower and are shown in Figure 3A. However, the differences between the KPS scores before, after and 3 months after surgery were not significantly different (p = 0.14). Patient HRQoL was assessed by SF-36 health survey before surgery and 3 months after surgery, the results for each SF-36 subscale are shown in Figure 3B. We found some increase in patient perceived problems with work or other daily activities because of physical health (i.e., “role limitations due to physical problems” SF-36 subscale) three months after surgery in comparison to before surgery, however, the difference was not statistically significant (p = 0.07). The scores in other SF-36 subscales (i.e., physical functioning, bodily pain, general health, vitality, social functioning, role limitations due to emotional problems, mental health) were very similar before and after surgery, and not significantly different (p > 0.1). Therefore, the surgery did not have important negative impact on functional status or quality of life as assessed by KPS and SF-36 health survey, respectively. Three months after surgery, the patients were asked, “if they would be willing to undergo awake craniotomy again if needed”. One patient answered “no”, one answered, “only if absolutely necessary” and the other 22 patients answered “yes”.

FIGURE 3.