Dexmedetomidine target controlled infusion for awake craniotomy^†

Summary

A 73‐year‐old woman underwent an awake craniotomy for the resection of a supratentorial brain tumour. We provided sedation for the surgery using a dexmedetomidine target controlled infusion using the Dyck pharmacokinetic model. Using a target controlled infusion allowed more rapid titration to the desired plasma level compared with a manual infusion, without any unexpected cardiovascular, respiratory or other complications. Rapid titration of sedation during awake craniotomy is desirable, allowing deeper sedation during stimulating parts of the surgery, followed by lighter sedation – or absence of sedation – during cortical mapping. While this can be performed manually, we found utilising the Dyck model in this case simple and quick to use, avoiding the need to manually calculate infusion rates. We believe this is the first report of using a target controlled infusion model to administer dexmedetomidine for awake craniotomy, and suggest it could be considered as an alternative to administering a manual infusion.

Introduction

Dexmedetomidine is a highly selective α₂ adrenoreceptor agonist. It is widely used in anaesthesia and intensive care medicine for its sedative, anxiolytic and analgesic properties and it has a favourable side effect profile compared to other sedative agents including maintenance of ventilation and airway reflexes. Dexmedetomidine is licensed in the United Kingdom as a sedative agent in the intensive care unit (ICU) and for sedation of non‐intubated adult patients undergoing diagnostic or surgical procedures [1]. It has numerous indications in anaesthesia including as a sedating pre‐medication, sedation during invasive procedures such as awake fibreoptic intubation, sympatholysis and as an opioid‐sparing analgesic [2]. At our centre, dexmedetomidine is commonly used in neuroanaesthesia as a sedative for neurosurgical procedures requiring intra‐operative patient co‐operation, for example, awake craniotomy and deep brain stimulation.

In anaesthetic practice, dexmedetomidine is typically administered as an initial loading infusion of 1 μg.kg⁻¹ over 10–15 min followed by a maintenance infusion, titrated to effect, at a rate between 0.2 and 1 μg.kg⁻¹.h⁻¹. While numerous pharmacokinetic models for dexmedetomidine exist, including the Dyck, Dutta and Talke models, their use is not widespread. We could only identify a small number of case reports where a dexmedetomidine target controlled infusion (TCI) was used in anaesthetic practice, for example, for awake tracheal intubation [3], and no case reports where it had been used in neuroanaesthesia.

In this report, we describe administering a dexmedetomidine TCI using the Dyck pharmacokinetic model to provide sedation for a patient undergoing an awake craniotomy for tumour resection.

Case report

A 73‐year‐old woman with a past medical history of well‐controlled hypertension presented for elective resection of a supratentorial brain tumour. An awake craniotomy was planned so that intra‐operative cortical mapping could be performed. As per the protocol at our centre, the patient had undergone a full surgical, anaesthetic and psychological assessment prior to the day of surgery to ensure she would be an appropriate candidate for awake surgery; explanation and consent for anaesthesia had been carried out at this stage.

On the day of surgery, a routine pre‐operative anaesthetic visit was carried out and no new issues were identified. In the anaesthetic room, routine monitoring was applied as per the Association of Anaesthetists guidance and intravenous access was established. A dexmedetomidine TCI was administered using the Dyck pharmacokinetic model via a Volumed® μVP7000 infusion pump (Arcomed, Zurich, Switzerland) and was titrated to effect by altering the desired plasma concentration (ng.ml⁻¹) on the infusion pump. The set plasma concentration was titrated to clinical effect and then checked by noting the infusion rate in μg.kg⁻¹.h⁻¹ (also displayed on the pump), which we had more familiarity with. This allowed us to ensure that infusion rates remained within safe limits for sedation (up to 1 μg.kg⁻¹.h⁻¹). Within 10 min of starting the dexmedetomidine TCI, moderate to deep sedation was achieved with the patient still rousable to voice. An arterial line was inserted and a scalp block performed using 20 ml 0.5% bupivacaine before the patient was transferred into the operating theatre. Sedation was continued during the surgical set‐up (neuro‐navigation, Mayfield Clamp application, exposure of the cortex) and then titrated to a plasma level of 0 ng.ml⁻¹ once the surgical team were ready to carry out cortical mapping. Within 10 min, the patient was co‐operative enough to perform cortical mapping. Once cortical mapping had been carried out, the sedation was titrated to achieve deep to moderate sedation. Surgery proceeded uneventfully and sedation was stopped once the Mayfield Clamp had been removed. Airway and cardiorespiratory stability were maintained throughout the procedure without the need for any support other than routine oxygen administration and end‐tidal CO₂ monitoring via nasal cannulae. During the case, the set plasma concentration for the dexmedetomidine TCI ranged between 0 and 2 ng.ml⁻¹ with the infusion rate never exceeding 1 μg.kg⁻¹.h⁻¹. The patient was reviewed on the first and second postoperative days. She reported being very satisfied with the sedation she had received and had minimal recall of the surgery. Her pain was well controlled with simple analgesia and her postoperative course was uneventful.

Discussion

This is, to our knowledge, the first case report where the Dyck pharmacokinetic model has been used to administer a dexmedetomidine TCI to provide sedation for an awake craniotomy. Dexmedetomidine has many of the desirable characteristics of the ideal sedative agent including minimal effect on airway reflexes, maintenance of respiratory drive and production of an easily rousable sedative state [2]. However, even when an initial loading dose is administered, the onset of sedation can be slower than other sedative agents such as propofol or remifentanil. While this is acceptable for sedation in the ICU, faster titration is desirable in the theatre environment. For example, in this case being able to rapidly increase the level of sedation when the Mayfield Clamps were applied and then rapidly reduce the level of sedation when patient co‐operation was required for cortical mapping would be a desirable quality of the sedative agent used. Faster titration can be achieved using a TCI with the added benefit of reducing the risk of under‐ and overdosing that can be seen with a manual infusion [4], avoiding potential side effects; we believe that using the Dyck model in this case allowed us to achieve this.

The pharmacokinetics of dexmedetomidine can be adequately described by a three‐compartment model [5]. The Dyck pharmacokinetic model was developed using data from 16 healthy volunteers and uses height as the only covariate [5, 6], allowing a target plasma concentration to be set. The effective sedative plasma concentration of dexmedetomidine is believed to be between 0.2 and 3.2 ng.ml⁻¹ [7]. In this case, we maintained a plasma concentration between 0 and 2 ng.ml⁻¹ and, owing to a limited familiarity with the model, we ensured safety by maintaining the infusion rate below 1 μg.kg⁻¹.h⁻¹, which is typically the maximal rate we would use when running an infusion manually. At these plasma levels, we were able to rapidly achieve a deep, but easily rousable, level of sedation quickly without any adverse airway, breathing or circulation effects. For even more rapid and tighter titration, an effect‐site TCI is required. To our knowledge, there is no commercially available effect‐site TCI model for dexmedetomidine, and so plasma‐site TCI is the next best available thing. While sedation should always be titrated to effect rather than a target plasma concentration, continued use of the Dyck model will help us develop a better understanding of the typical plasma levels required for different levels of sedation, supporting us to achieve more rapid titration and particularly maintenance of sedation. Another benefit of TCI dexmedetomidine sedation is that it eliminates the need to manually calculate a loading dose as the model accounts for this when targeting the set plasma level, reducing the risk of error.

Limitations of the Dyck model include the fact that it was derived using data from a small sample of healthy male volunteers, it may underestimate plasma levels (especially at higher target levels) [6] and the lack of availability of a pre‐programmed model on many of the infusion pumps in commercial use. Our institution has recently changed to Volumed® μVP7000 infusion pumps (Arcomed, Zurich, Switzerland) with the Dyck model pre‐programmed, allowing us to use dexmedetomidine TCI. However, several of the newer infusion pumps also have a pre‐programmed dexmedetomidine model suggesting it may be much more widely available in the future. Our pumps require the patient's age, height, weight and gender to be input when setting up the Dexmedetomidine TCI although the Dyck model only uses height as a covariate. Contraindications to using a dexmedetomidine TCI are the same as for a manual infusion and include uncontrolled hypotension and second‐ or third‐degree heart block. Dexmedetomidine should be avoided in pregnancy and breastfeeding and used with caution in elderly patients due to the increased risk of hypotension, especially with the loading dose. Therefore, when using the Dyck model in elderly patients, it may be appropriate to initially target a lower plasma level and then titrate up slowly to achieve the desired level of sedation. Dexmedetomidine was initially only licensed as an ICU sedative agent in the United Kingdom and was used off licence for sedation during surgery at our centre because of its favourable side profile compared with other sedative agents. More recently, it has been approved for sedation of adult patients during surgical procedures [1], and we continue to use it routinely for neurosurgical procedures requiring intra‐operative patient co‐operation.

In conclusion, we present a case where a dexmedetomidine TCI was used to provide sedation for an awake craniotomy. The key benefits were more rapid titration of sedation and avoiding the need to manually perform a loading dose followed by having to calculate infusions rates. We suggest that as dexmedetomidine TCI models become more widely available, dexmedetomidine use for sedation in anaesthetic practice may increase.