Study Design and Participants
This single-institution, double-blind, randomized controlled study was approved by the Research Ethics Committee of National Taiwan University Hospital and was registered at ClinicalTrials.gov with the identifier NCT02878707. We enrolled patients older than 20 years who had undergone elective cranial surgery for brain tumor resection, aneurysm clipping, intracranial bypass procedure, and microvascular decompression surgery between April 2017 and April 2020. Patients who met any of the following criteria were excluded: age > 80 years; metastatic brain tumor; revision surgery; history of arrhythmia or New York Heart Association Functional Classification class III or higher heart failure; liver cirrhosis > Child B class, renal insufficiency with an estimated creatinine clearance of <60 mL min−1 1.73 m−2.
Before starting this trial, stratified randomization was performed by an independent statistic expert using a block size of 4. At an inclusive 1:3 ratio, patients who underwent cerebrovascular and brain tumor surgeries were equally randomized to the dexmedetomidine and control groups. All patients provided written informed consent on the day before surgery to an investigator who was unaware of the randomization result. The masked drug was provided by an independent pharmacy, thereby concealing the allocation from investigators and clinicians.
An infusion of dexmedetomidine (0.5 μg.kg−1.h−1) or equivalent dose of saline placebo was started immediately before the anesthesia induction and was maintained throughout the surgery. General anesthesia was performed and maintained through total intravenous anesthesia by using target-controlled infusion of propofol based on the Schnider model. Remifentanil and fentanyl were infused during the surgery per the attending anesthesiologist’s discretion. Intraoperative opioid consumption was calculated in fentanyl equivalents for comparison . Anesthesia was titrated to maintain the bispectral index (BIS; employed at the contralateral to the surgical side) between 40 and 60. The patients were ventilated with a tidal volume of 6–8 mL kg−1 and air:oxygen ratio of 1:1 and a positive end-expiratory pressure of 5 cmH2O. For analgesia, each patient received a scalp nerve block containing 10 mL of 0.5% levobupivacaine with 1:200,000 epinephrine mixture for each side of the scalp before skin incision . Neurophysiological monitoring techniques were used to enhance surgical safety.
Cardiac Output Monitoring and Goal-Directed Hemodynamic Therapy Protocol
After anesthesia induction, a 20-G radial arterial line was inserted and connected to the fourth-generation Vigileo/FloTrac system (Edwards Lifesciences, Irvine, CA, USA) to obtain the stroke volume index and cardiac index. A 16 or 14 G intravenous catheter was inserted in the forearm for the GDHT protocol. Details of the GDHT protocol and hemodynamic goals are provided in Appendix File 1. Briefly, the aim of the hemodynamic protocol was to optimize the stroke volume index while maintaining a cardiac index ≥ 2.2 L min−1 m−2 by repeatedly administrating 250 mL of colloid (Voluven; Fresenius Kabi, Uppsala, Sweden). The decision to administer the colloid was based on a stroke volume index increase of ≥5% after the mini-fluid challenge test of 100-mL crystalloid infusion within 1 min because this test remains valid during low tidal volume ventilation [18-20]. In addition, intraoperative blood pressure was maintained to mean arterial pressure ≥75 mm Hg and ≥85% of baseline state. The intraoperative parameters—including BIS, heart rate, mean arterial pressure, and cardiac index—were recorded every 15 min and were compared using the average values in the two study groups.
After surgery, all patients were immediately transferred to the same neurosurgical intensive care unit (ICU) and received identical postoperative care without using the study drug. If tracheal tube could not be extubated in the operating theatre, a weaning program was initiated to prevent prolonged mechanical ventilation (defined as the failure of extubation within 3 h of arrival in the ICU). Patients were considered to have delayed emergence if they did not either spontaneously open their eyes in response to speech or have motor response to command within 30 min after surgery. The hemodynamic protocol and criteria for discharge from the ICU and hospital are listed in Appendix File 2.
Postoperative Neurological Function and Outcome Assessment
This study’s primary outcome was the proportion of patients who developed new neurological deficit. Nurses specializing in neurosurgical care and neurosurgeons oversaw the patients’ assessment at least twice daily throughout the whole hospital stay. Diagnosis of neurological deficit was based on the criteria of the International Statistical Classification of Diseases and Related Health Problems, 10th Revision. The severity of disability of neurological deficit was evaluated by using in-hospital modified Barthel index change and perioperative modified Rankin Scale (mRS) [21-23]. In-hospital modified Barthel indices, assessed at admission and discharge, were obtained from institutional medical records, and the mRS scores were determined by an independent neurosurgeon who was unaware of the group allocation and other outcomes. The mRS, a 7-point scale in which 0 represents no disability, 1-2 represents slight disability, and 3-6 represents moderate to severe disability or death (mRS=6), was employed at preoperative baseline and 30 days after surgery. In addition, Postoperative change in mental state was assessed at least twice daily to document any delirium signs. One trained physician blinded to the group allocation reviewed these medical records and diagnosed the delirium sign using the criteria of the Intensive Care Delirium Screening Checklist (ICDSC) , a scale scored from 0 to 8 [6, 24]. The mRS and ICDSC scores were measured by another independent physician to determine interrater agreement.
Serum Biomarker Analysis
Serum lactate level was analyzed at baseline (after anesthesia induction) and at the end of surgery to assess the hemodynamic influence of dexmedetomidine. Plasma levels of biomarkers—including HMGB1, glial fibrillary acidic protein (GFAP), and S100β—were measured to assess the biochemical effects of dexmedetomidine on the attenuation of neuroinflammation. These serum biomarker levels at baseline (after anesthesia induction) and on the first postoperative day were compared. Serum concentrations of HMGB1 (Chondrex Inc., Redmond, WA, USA), GFAP (BioVendor LLC, Candler, NC, USA), and S100β (BioVendor LLC) were measured using enzyme-linked immunosorbent assay kits.
According to previous studies, postoperative new neurological deficits can occur in 40% of cases [13, 21]. Accordingly, group sample sizes of 80 in each group achieves 80% power to detect a difference between the group proportions of 20% by using the two-sided Z test with pooled variance. The significance level of the test was targeted at 0.05. Fisher’s exact test or Chi-square test was performed to analyze dichotomous data, Student’s t-test and the Mann–Whitney U test were used for normally distributed continuous data and nonparametric ordinal data respectively. The paired t-test was used to compare biomarker levels measured before and after surgery. Interrater agreement was analyzed by calculating the kappa coefficient. Logistic regression including all investigated preoperative and intraoperative variables was used to identify the risk factors of new postoperative neurological deficits and delirium. Preoperative and intraoperative variables were considered for the multivariate logistic regression model if p < 0.05 in the univariate analysis. Statistical analyses were performed using PASS Sample Size Software (NCSS, LLC, Kaysville, UT, USA) and MedCalc Statistical Software version 19.3.1 (MedCalc Software Ltd, Ostend, Belgium).
In our group sequential design, one interim analysis of severe adverse events was planned to investigate whether dexmedetomidine-induced cardiac output reduction resulted in lactate accumulation. Dexmedetomidine reportedly induces a heart rate reduction of 16.4 beats per min  which may comprise an average cardiac index reduction of 0.5 L·min−1 m2 given our previous report into patients receiving GDHT undergoing cranial surgery . Accordingly, 24 patients in each group would enable detection of differences in intraoperative averaged cardiac index. Therefore, the planned interim analysis was performed once 50 patients had been enrolled, and the study safety board suggested termination of the trial once this cardiac index difference resulted in significant lactate accumulation (p < 0.05). Because the interim analysis did not assess the primary outcome, and thus, the significance level (p < 0.05) was not adjusted.