We read with great interest the article by Prabhakar et al. regarding the use of bispectral index (BIS) monitoring for an elective frontal craniotomy and fenestration in a 2-year-old boy with bilateral frontal lobe porencephalic cysts. During the procedure, the patient presented low preanaesthetic BIS values (39) that showed no significant change (between 30 and 40), despite a reduction in the inspired concentration of sevoflurane from 1 to 0.2%. No burst suppression pattern was observed, and BIS values remained low after anaesthetic recovery. In the discussion, the authors suggest that intracranial pressure increments could be responsible for the low BIS values.
We believe that we present the first case of BIS monitoring in hydranencephaly, an extreme form of porencephaly. With it we hope to throw light on the origin of low BIS values recorded in these patients.
A 5-month-old boy, weighing 4.5 kg, with the diagnosis of hydrocephalus, was scheduled for endoscopic ventriculostomy and coagulation of choroid plexus.
Hydranencephaly is a rare condition in which the brain's cerebral hemispheres are absent and replaced by sacs filled with cerebrospinal fluid. It is characterized by near total or total absence of cerebral cortex and basal ganglia. The thalami, pons, cerebral peduncles, and cerebellum are usually present, as may be a small amount of tissue from the occipital, frontal, and temporal lobes. At birth, children with hydranencephaly may appear normal, spontaneous reflexes such as sucking, swallowing, crying, and moving the arms and legs may all seem normal, but developmental delay becomes apparent within a few weeks. Common manifestations of this condition include irritability, abnormal muscle tone, seizures, and increased head size (as a result of hydrocephalus). Swallowing difficulties lead to malnutrition, aspiration, and pneumonia.
Our patient presented global hypotonia but had developed social smile, was able to hold up his head, managed to focus and track an object with both eyes and reacted to sounds.
MRI revealed bilateral multiple porencephalic cysts replacing the cerebral hemispheres (Fig. 1). Electroencephalogram (EEG) showed a slow activity pattern with low amplitude waves compatible with severe damage to cerebral electrogenesis.
Anaesthesia was induced with an increasing sevoflurane/oxygen mixture (1–3–6%) delivered by facemask, with atropine, fentanyl (10 μg) and atracurium. Tracheal intubation was performed using a 4-mm uncuffed Portex single lumen tube.
General anaesthesia was maintained using an oxygen mixture with sevoflurane (2%). In order to monitor anaesthetic depth, brain electrical activity was continuously measured with an A 2000-XP BIS Monitor (Aspect Medical Systems, Newton, Massachusetts, USA). Paediatric low-impedance electrodes were attached to the patient's frontal area as specified by the manufacturer.
Initial average BIS values (AVGBIS) were above 80, probably due to a low signal-quality index (SQI); however, during anaesthetic maintenance recorded values were persistently below 30, with a SQI higher than 90%. Interestingly, average electromyography values (AVGEMG) were practically the same as AVGBIS (Fig. 1).
BIS records cortical electroencephalographic signals. However, in the absence of EEG signals or in extreme EEG suppression situations (as occurs in brain death or patients with deep hypothermia) the BIS algorithm becomes very vulnerable to a wide range of artefacts . As a consequence, the ECG or high electromyographic signals  can be misinterpreted as EEG activity .
The EMG frequencies overlap the BIS algorithm's beta ratio in the 30–47 Hz range, associated with awake or light levels of anaesthesia. Thus, EMG values within this range can be misinterpreted by the BIS algorithm as EEG activity and a spuriously increased BIS value can be recorded.
The BIS variable was created following collection of clinical data from healthy volunteers with normal EEG. It is therefore not surprising that neurological disorders that manifest abnormal EEG patterns can affect BIS monitoring. It has been described that patients with Alzheimer's type dementia and children with West syndrome and lissencephaly show low preanaesthetic BIS values. In the same way, children with cerebral palsy present lower values during and after anaesthesia than healthy ones. In porencephaly, the EEG recorded over porencephalic cysts is characterized by increased theta and delta bands in the areas surrounding the injury identified by computed tomography . Taking into account that the BIS algorithm is based on electroencephalogram processing, it is not unexpected that low BIS values were recorded in this disease.
In this case, due to the cysts, the cerebral cortex and consequently the cerebral electrogenesis are diminished in such a manner that EMG activity might be the main component of the BIS value. If we analyse the variance of AVGBIS and AVGEMG values recorded during surgery, the AVGEMG value might explain a significant amount of AVGBIS variability (F Snedecor = 529 637; df = 1; P < 0001). The corrected determination coefficient (r2 corrected) indicates that the AVGEMG value explains 91.2% of AVGBIS variation.
Intracranial pressure (ICP) increase caused by the cysts can provoke a reduction in cerebral perfusion pressure and global cerebral ischaemia. This fact would also explain the marked decrease in BIS values. However, in our opinion, this was not the main physiopathological mechanism of BIS reduction, neither in our case report nor in the one reported by Prabhakar et al.. This reduction cannot be explained by increased ICP because it was not measured, and there was no clinical evidence of a sufficient rise in ICP that could explain a significant decrease in cerebral perfusion pressure. Furthermore, there was no record of an increase of burst suppression paralleling a BIS decrease in any of the cases, as we could expect in this situation . It seems more reasonable that low BIS values were explained by cerebral cortex loss.
There is evidence that placement of BIS monitor leads in positions different from those advised by the manufacturer may still provide interpretable readings. A statistically significant correlation of BIS values between frontal and occipital placement of electrode strips has been demonstrated . In dolphins, it has been proven that a BIS monitor can detect interhemispheric asymmetry in electroencephalographic activity during sleep. For this reason, it is possible that the position of the electrodes might make a difference in cases of neurological injury, such as the case reported by Prabhakar et al.. In this case, most of the cerebral mass was located in the occipital area; therefore, placement of BIS monitor leads in this zone could have increased the amount of data obtained.
Cerebral function monitoring is a new application of BIS, which is different from its initial purpose. Although BIS is a well accepted monitor for anaesthetic depth, several factors, unrelated to anaesthesia, can modify it. Several clinical observations suggest that BIS monitoring can be used as an ‘alarm signal’ to detect repercussions of hypoxic, ischaemic or other similar injuries on cerebral function as long as the depth of anaesthesia is controlled. However, further studies are needed to validate the use of BIS for this purpose and address its potential applications and limitations.
In contrast, patients with neurological disorders who manifest abnormal EEG patterns will probably have significant changes in BIS monitoring readings. Accordingly, in these patients one has to be careful in interpreting BIS values because, as in the hydranencephaly case, recordings correspond more to EMG activity than to the EEG of cerebral cortex.
1 Prabhakar H, Ali Z, Bithal PK, et al
. Low bispectral index values in a 2-yr-old with a large bifrontal proencephalic cyst. Eur J Anaesthesiol 2008; 25:513–514.
2 Dahaba AA. Different conditions that could result in the bispectral index indicating an incorrect hypnotic state. Anesth Analg 2005; 101:765–773.
3 Myles PS, Cairo S. Artifact in the bispectral index in a patient with severe ischemic brain injury. Anesth Analg 2004; 98:706–707.
4 Vivien B, Paqueron X, Le Cosquer P, et al
. Detection of brain death onset using the bispectral index in severely comatose patients. Intensive Care Med 2002; 28:419–425.
5 Cerquiglini A, Seri S, Sturniolo MG, et al
. Computerized electroencephalographic assessment of congenital brain infarction. Childs Nerv Syst 1994; 10:252–258.
6 Morimoto Y, Monden Y, Ohtake K, et al
. The detection of cerebral hypoperfusion with bispectral index monitoring during general anesthesia. Anesth Analg 2005; 100:158–161.
7 Shiraishi T, Uchino H, Sagara T, et al
. A comparison of frontal and occipital bispectral index values obtained during neurosurgical procedures. Anesth Analg 2004; 98:1773–1775.