A retrospective analysis of perioperative complications during intracranial neuroendoscopic procedures: Our institutional experience
Correspondence Address: DOI: 10.4103/0028-3886.91368
Background: Neuroendoscopic procedures are now being performed more frequently, and with advancement in technology, complications related to the procedure and equipments have also minimized or changed. We report our experience with 223 patients who underwent intracranial neuroendoscopic procedures. Materials and Methods: The rates of various perioperative complications, both surgical and anesthesia related, during intracranial neuroendoscopic surgeries were studied. Data collected included demographics, patient's medical history and any associated comorbid conditions, diagnosis, procedure performed, anesthetic management, intraoperative and postoperative complications and outcomes. Results: Of the 223 patients studied, 119 were pediatric (age <14 years) and 104 were adults. Hypothermia (25.1%) and cardiovascular complications (such as tachycardia 18.8%, bradycardia 11.3%, hypertension 16.1%, and hypotension 16.6%) were the commonly observed complications during intraoperative period both in pediatric and adult patients. At the end of the procedure, delayed arousal was observed in 17 patients and 19 patients required postoperative ventilatory support. Postoperative frequent complications included: fever (34.1%), tachycardia (32.7%), nausea and vomiting (18.8%). Potentially fatal complications such as intraoperative hemorrhage, air embolism, etc. were rare. Most of the complications were transient and self-limiting. Conclusion: Although endoscopic procedures are considered minimally invasive, at times may lead to life-threatening complications and one should be aware of them.
Keywords: Anesthesia, cardiovascular complications, intracranial surgery, minimally invasive, neuroendoscopy, surgical complications
In recent years, endoscopic techniques have been applied successfully in neurosurgery for various intracranial and spinal lesions, and this has been possible mainly due to improvements in the optics and advances in technology.  Due to less invasive nature of endoscopic surgeries compared to the traditional neurosurgical techniques, the incidence and rate of perioperative complications are expected to be low.  Nevertheless, this procedure is also associated with host of complications: Infection, hemorrhage, subdural hygroma, subdural hematoma, epidural hematoma, pneumocephalus and increased intracranial pressure (ICP). ,, Complications associated with the use of endoscope and irrigation fluid include: Rise in ICP due to rapid run or inadequate egression of irrigation fluid, ,, dilution and electrolyte abnormalities, , injury to basilar artery lying beneath the third ventricular floor  and injury to structures such as the fornix, hypothalamus, or cranial nerves. Few studies in the anesthesia literature have discussed perioperative problems in patients undergoing neuroendoscopic surgeries. ,,,, We report our experience of various perioperative complications, both surgical and anesthesia related, in 223 patients who underwent neuroendoscopic procedures at our institute.
This is a retrospective review of case records of all patients who underwent endoscopic neurosurgical procedures between January 2006 and August 2009. Data collected included demographics, patient's medical history and any associated comorbid condition, diagnosis and procedure performed, anesthetic management, intraoperative and postoperative complications and the condition of patient at discharge. This study was approved by institutional ethical committee.
A standard management protocol was followed for all patients. Most patients received phenytoin sodium and acetazolamide in the preoperative period which was continued till the day of surgery. In addition, patients receiving drugs for any comorbid conditions also continued to receive them. All patients received glycopyrrolate or atropine an hour prior to surgery. General anesthesia was induced with fentanyl 2 mcg kg and thiopentone 4-5 mg/kg. Tracheal intubation was facilitated using rocuronium 1 mg/kg. Anesthesia was maintained with either isoflurane or sevoflurane in a mixture of nitrous oxide (60%) and oxygen (40%) along with intermittent boluses of fentanyl and vecuronium or rocuronium. Mechanical ventilation was instituted to maintain the end-tidal carbon dioxide concentration at 35 ± 1 mm of Hg. Invasive blood pressure monitoring was performed to record the hemodynamic changes in most patients by cannulating either the dorsalis pedis or the radial artery. Nasopharyngeal temperature was monitored in all the patients. Normal saline was used as the replacement and maintenance fluid during the intraoperative period. During endoscopy, Ringer's solution warmed to 37°C was used for rinsing. At the end of the surgery, anesthesia was discontinued and neuromuscular blockade reversed with neostigmine and glycopyrrolate. All patients, whether tracheally extubated or requiring postoperative ventilation, were shifted to neurosurgical intensive care for supportive care and further management.
Intraoperative and postoperative records of the patients were recovered to collect the information on perioperative events during the hospital stay. Various intraoperative and postoperative events recorded included: tachycardia (sudden increase in heart rate by 20% of baseline value), bradycardia (fall in heart rate by 20% or more), hypertension and hypotension (rise or fall in blood pressure by more than 20%, respectively), hypothermia (fall in core temperature below 35.5°C), electrolyte disturbances (hyponatremia or hypernatremia), nausea and vomiting, fever (temperature above 99°F), increased ICP (suggested by the signs like nausea and vomiting, bradycardia, drowsiness or altered sensorium and CT findings), failed endoscopic procedures (either due to severe hemorrhage or inability to reach the site of lesion requiring open craniotomies or other alternative procedure), arrhythmias (like premature ventricular contractions, bigemini and asystole), venous air embolism, hemorrhage, delayed awakening, meningitis, hydrocephalus, pneumocephalus, subdural hygroma, diabetes insipidus, cerebrospinal fluid fistula, shivering, convulsion, anisocoria, respiratory complications (like upper or lower respiratory tract infection, pneumonia, pleural effusion, stridor, wheezing, bronchospasm, laryngospasm, tachypnea, desaturation, reintubation, postoperative mechanical ventilation), psychiatric complications (like hallucinations, delirium, restlessness, agitation, somnolence), paresis of various intracranial nerves and new neurological deficits. The duration of postoperative ventilation, ICU stay, hospital stay and condition of patient at discharge were also noted. Data are presented as mean (SD), median (range), or number (percentage).
A total of 240 patients underwent neuroendoscopic surgeries during the study period. Of these, data of 17 patients were found to be incomplete and were excluded from the final analysis. Of the 223 patients included in the study, 143 were males and 80 were females; median age was 13 years (14 days to 69 years), 119 patients were in the pediatric age group (<14 years), while 104 were adults. The mean weight was 34.6 (25.2) kg. The associated comorbidity in children was delayed milestones (21%), while in adults hypertension was the most common comorbidity (8.7%). Other associated comorbidities were: Diabetes mellitus, tuberculosis, asthma, meningomyelocele, encephalocele, convulsions, cranial nerve palsies, atrio-septal defect, tracheoesophageal fistula, stridor, history of preterm birth, pneumonia, and neonatal sepsis. The average duration of anesthesia and surgery was 147.3 (48.9) min and 86.9 (42.6) min, respectively. All patients had received general anesthesia with the inhalation agents, either isoflurane ( n = 102) or sevoflurane ( n = 121), in oxygen (40%) and nitrous oxide (60%) mixture. Various indications and surgeries performed in pediatric age group and adults are listed in [Table 1]. Different intraoperative and postoperative complications observed are tabulated in [Table 2] and [Table 3], respectively. The average ICU stay was 1 (1-20) day and the average hospital stay was 4 (2-24) days. Of the 223 patients, 19 patients (16 children and 3 adults) could not be extubated and required postoperative ventilation. Eight of these patients had poor respiratory efforts, while 11 patients were drowsy and did not follow simple verbal commands. Delayed arousal was observed in 17 patients of whom 4 were hypothermic.
One patient developed decreased vision in the immediate postoperative period, but the vision improved to normal on the second postoperative day. We did not have any mortality, and all but one patient were discharged with no fresh neurological deficits. One patient who had multiple complications in the postoperative period: hydrocephalus, intracranial bleed, seizures, fever, tachycardia, hypertension, and vomiting, and required ventriculo-peritoneal shunt for drainage of CSF and blood. He was discharged on postoperative day 24 with a Glasgow Coma Score of 12.
The real incidence of complications for neuroendoscopic procedures is difficult to estimate because not all complications are reported , and there is a lack of clear definition of what should be considered a complication.  The reported complication rates for neuroendoscopic surgeries varied from 6 to 20%. ,
Intraoperatively, hypothermia was the common finding in our study. Causes of hypothermia include low environment temperature of the operation theater, inadequate warming measures used for the patient after induction of anesthesia or wetting of drapes with egression of irrigation fluid. Hypothermia can also result if too large a quantity of irrigation fluid at room temperature is used.  In our patients, hypothermia was observed in spite of using warming measures and pre-warmed fluid (at 37°C) for irrigation. The frequency of various cardiovascular changes found in our series was similar to the observations made in other series. , In our study, tachycardia was the commonest finding, with nearly half of the patients having associated hypertension, a finding similar to what has been observed by Ganjoo et al.  On the contrary, El-Dawlatly et al.  observed bradycardia to be the commonest arrhythmia in their study. These hemodynamic changes probably reflect the change in the ICPs which occur with the use of irrigation fluid to improve visibility. ,,,,,[ 19] Inadvertent stimulation or injury of the posterior hypothalamus (which modulates the cardiac regulatory function of the brain stem via descending autonomic pathways) or of the third cranial nerve, both of which lie in close proximity to the floor of third ventricle, can also produce significant hemodynamic responses.  Bradycardia along with hypertension noted in some of our patients (described as Cushing reflex) is the clear sign of raised intracranial pressure. Prabhakar et al.  also demonstrated hemodynamic changes and increase in ICP with concurrent fall in cerebral perfusion pressures during Valsalva maneuver in patients undergoing neuroendoscopic procedures. Anandh et al.  also found a significant slowing of the heart rate during the fenestration of the floor of the third ventricle. Similarly, El-Dawlatly et al.  reported an incidence of 41% of bradycardia in pediatric patients undergoing endoscopic third ventriculostomy (ETV). In another study by Kalmar et al.,  of the 17 patients who underwent neuroendoscopic procedures, almost every one had hypertension and tachycardia and Cushing reflex (in those patients in whom cerebral perfusion pressure dropped below 15 mmHg), further emphasizing that simultaneous onset of hypertension and tachycardia is a better and early indicator of impaired cerebral perfusion.
Massive intraoperative bleeding in the ventricular system due to damage to the ependymal vessels or the basilar artery lying beneath the floor of the third ventricle has been reported by different authors.  It is a major complication that may necessitate urgent craniotomy to prevent any catastrophe. In this series, intraoperative bleeding was observed in 12 patients, and in 7 of them, it was associated with various hemodynamic changes. In three patients, the procedure had to be abandoned and open craniotomy had to be performed. In another patient, endoscopy was converted to open craniotomy due to inability to assess the site of lesion.
Handler et al.  have implicated the rapid and forceful irrigation to be the cause of cardiac arrest either due to the direct distortion of hypothalamic nuclei or secondary to the increase in intraventricular pressure by irrigation fluid and recommended regulating the speed of irrigation to less than 10 ml/min. Though we did not witness asystole in any of our patients, arrhythmias like premature ventricular contractions and bigemini were observed; probably all these reflect the increase in ventricular pressure and/or stimulation of hypothalamic nuclei.
Another potentially fatal complication is the air embolism, either during craniotomy or while withdrawing the endoscope, and it accounted for 4% in the series by Fabregas et al.  We had air embolism in one patient while doing a burr hole which was detected immediately and informed to the surgeon who flooded the field with saline. There was tachycardia and hypertension lasting for 2-3 min and sustained fall in EtCO 2 which gradually returned to normal over 90 sec and the further course of the procedure was uneventful. During the event, nitrous oxide was discontinued and 100% oxygen delivered.
Delayed arousal was noted in 17 (7.6%) patients. Large amount of irrigation fluid used during the procedure, which alters the composition of CSF, or the injury to the brain structure during surgical manipulation or the sustained high pressure levels may be the reasons for this complication.  There was associated hypothermia in four of these patients. Incidence of delayed arousal as high as 15% has been reported by Fabregas et al.  and they attributed it some degree of neurological impairment. Most of our patients were extubated at the end of procedure and those who could not be extubated were mostly children. These patients either did not have adequate respiratory efforts or they were not alert enough (delayed awakening) to follow simple commands. This may be related to prolonged action of muscle relaxants or sedatives used (especially in children) as many a times the procedure ends abruptly, causing difficult titration of anesthetics.
Fever was the most common postoperative event in our study and is probably the result of aseptic irritation of the ependyma or manipulation of the hypothalamus.  Because of the differences in chemistry between the irrigation fluid and CSF, toxic reactions have been reported which manifest as fever, headache, chemical meningitis and increased CSF cell count.  Irrigation with normal saline has been reported to cause inflammatory neurological complications including high-grade fever and headache.  In our study, the irrigation fluid used was Ringer lactate, and in most of our patients, fever was transient and resolved with cold sponging or by antipyretics.
Transient change in the ionic composition of CSF by the irrigation fluid may cause stimulation of respiratory centers located in the brain stem and this has been suggested as the possible cause for respiratory complications such as hyperventilation.  Enya et al.  described two infants who developed respiratory arrest after endoscopic surgery for hydrocephalus and suggested postanesthetic apnea monitoring of young infants who undergo ventriculoscopic surgery. One of our patients had tachypnea during the postoperative period with no obvious explainable cause. One patient had bilateral wheeze and two patients developed stridor postoperatively.
Injury to brain structures or cranial nerves in the adjacent region may result in various postoperative complications. Diabetes insipidus and electrolyte imbalance may result from the damage to hypothalamus,  and various nerve palsies may result from injury to the cranial nerves or their nuclei. Postoperative electrolyte imbalance is most often with no clinical significance.  However, permanent diabetes insipidus following third ventriculostomy has been documented.  Other rare complications associated with this procedure in our series included: Pneumocephalus, subdural hygroma, postoperative hydrocephalus, restlessness and confusion. Seizures can be a complication of this procedure; , five of our patients had seizures during the postoperative period. Postoperative seizures may be the result of either Pneumocephalus, intraventricular bleed or electrolyte imbalance. ,, We had no mortality, and the morbidity at discharge was also very low (0.44%).
The retrospective nature of our study has its limitation. We did not monitor ICP in our patients, so the reasons provided for hemodynamic changes secondary to change in intracranial pressures can be considered hypothetical. Long-term outcome of our patients is also not known as it was beyond the aim of our study.
[Table 1], [Table 2], [Table 3]