Drug-Induced Sleep Endoscopy (DISE) with Target Controlled Infusion (TCI) and Bispectral Analysis in Obstructive Sleep Apnea

Summary

The aim of this study was to establish a standardized protocol for sleep endoscopy to differentiate obstruction patterns in obstructive sleep apnea (OSA). Target-controlled infusion (TCI) of the sedative was combined with real-time monitoring of the depth of sedation using bispectral analysis.

Abstract

The aim of this study was to establish a standardized protocol for drug-induced sleep endoscopy (DISE) to differentiate obstruction patterns in obstructive sleep apnea (OSA). Target-controlled infusion (TCI) of the sedative propofol was combined with real-time monitoring of the depth of sedation using bispectral analysis.

In an observational study 57 patients (mean age 44.8 years, ± SD 10.5; mean apnea hypopnea Index (AHI) 30.8/hr, ± SD 21.6, mean BMI 28.2 kg/m², ± SD 5.3) underwent cardiorespiratory polysomnography followed by DISE with TCI and bispectral analysis. Sleep was induced solely by the intravenous infusion of propofol with a TCI-pump, with an initial target plasma level of 2.0 µg/ml. Under continuous monitoring of the patient's respiration, state of consciousness and value of the bispectral analysis, the target plasma propofol level was raised in steps of 0.2 µg/ml/2 min until the desired depth of sedation was reached. The mean value of the bispectral analysis at the target depth of sedation was determined and the obstruction patterns during DISE-TCI-bispectral analysis then classified according to the VOTE-system. Subsequently the results were analyzed according to polysomnographic and anthropometric data. The occurrence of multilevel obstruction sites across all degrees of severity of OSA clarifies the need for sleep endoscopy prior to upper airway surgery.

The advantage of this technique is the reproducibility of the protocol even for heterogeneous groups of patients. In addition, the gradual controlled and standardized increase of the plasma level of propofol with real-time control of the bispectral index leads to a precisely controllable depth of sedation. The DISE-TCI-bispectral analysis procedure is a step towards a required reproducible protocol of sleep endoscopy — capable of standardization. However it is not yet known whether these observed obstruction patterns also correspond to findings in natural sleep.

Introduction

Obstructive sleep apnea (OSA) is characterized by repetitive phases of complete (apnea) or partial (hypopnea) collapse of the upper airways during sleep. These phases are often associated with arterial oxygen desaturation and a fragmentation of sleep caused by arousals. The reported prevalence of OSA with accompanying daytime symptoms in the general population is 3 - 7% in men and 2 - 5% in women¹. The gold standard in the treatment of moderate to severe OSA is nasal continuous positive airway pressure (nCPAP), for which compliance is internationally reported as about 40 - 60%². This treatment is used by 29 - 83% of OSA patients on a regular basis for less than 4 hr³. Both positive and negative predictors regarding long-term compliance of CPAP-use are well known nowadays⁴. In addition, emotional and clinical side effects often considerably reduce compliance. Alternative treatment options such as upper airway surgery therefore play an important role in OSA therapy. However, the fact that the success rate of surgery (responder rate) is relatively low in comparison with ventilation therapy is problematic ^5,6.

It was hoped that the introduction of sleep endoscopy by Croft and Pringle in 1991 would not only provide further insights into the pathophysiology of OSA but also might improve the responder rate through individualized surgical treatment⁷. As early as 2011, studies by De Vito et al. demonstrated the advantage of using investigative techniques based on target-controlled infusion (TCI) and bispectral analysis with respect to safety, stability and accuracy⁸. In the meantime, the validity and reliability of sleep endoscopy have been established and, ever since the 2014 European Position Paper, it is on the road to standardization^9-11. The aim of the present study is to establish a standardized protocol for sleep endoscopy by target-controlled infusion of the sedative propofol, combined with real-time monitoring of the depth of sedation using bispectral analysis, in order to differentiate obstruction patterns according to OSA-severity.

CASE PRESENTATION:

Study Design:

The retrospective study was conducted in the Department of Otorhinolaryngology, Head & Neck Surgery of the Friedrich-Alexander University Erlangen-Nürnberg between September 2012 and November 2014, following approval by the local Ethics Committee. All 57 participating patients, aged 20 to 73 years, were recruited by the Department of Otorhinolaryngology, Head and Neck Surgery. 52 patients were men and 5 women. In addition to a standardized interview, they were examined by an otorhinolaryngologist and underwent an awake endoscopy to assess the upper airways. Cardiorespiratory polysomnography was then undertaken in the Department's sleep laboratory to enable an exact classification of their sleep-related breathing disorder. The OSA severity was classified as mild (AHI 5 - 15/hr), moderate (AHI > 15 < 30/hr) or severe (AHI > 30/hr), according to the criteria of the American Academy of Sleep Medicine Task Force¹². The indication for sleep endoscopy was established in the context of a planned surgical intervention of the upper airways (primary indication) or in the case of nCPAP non-compliance (secondary or adjuvant indication).

Inclusion criteria for this study were men and women aged 18 - 75 years with mild, moderate or severe OSA diagnosed by polysomnography. Exclusion criteria were an American Society of Anesthesiologists Classification (ASA) IV/V, central sleep apnea, positive history of misuse of sedatives, alcohol or addictive drugs, allergy to propofol, pregnant women.

Diagnosis, Assessment, and Plan:

Cardiorespiratory polysomnography (PSG):

Polysomnography was carried out with a 33-channel cardiorespiratory diagnostic system. The technical procedure for the polysomnographic diagnostics followed the recommendations of the American Academy of Sleep Medicine (AASM) in the standardized technique using an electroencephalogram (EEG; F4-M1, C4-M1, O2-M1), right and left electro-oculogram, electromyogram of the mentalis and tibialis muscles, nasal pressure cannula, thoracic and abdominal respiratory effort sensors (inductive plethysmographs), body position sensors, pulse oximetry, snoring microphone, a one-channel ECG and an infrared video recording¹³. The evaluation was performed according to the AASM Criteria (Version 2.0, 2012) and was undertaken by an accredited medical sleep specialist of the German Sleep Society (DGSM)^{13, 14}. After polysomnography had confirmed OSA, all 57 patients underwent a standardized propofol-based Drug-Induced Sleep Endoscopy (DISE) with TCI and bispectral analysis (DISE-TCI-bispectral analysis).

Preoperative preparations:

Due to the muscle relaxant effect, no premedication with benzodiazepines was given in the case of sleep endoscopies undertaken solely for diagnostic purposes. If the sleep endoscopy was performed during a planned surgical procedure, clonidine was used for premedication as an alternative for benzodiazepines, taking the corresponding contraindications into account.

Classification of the obstruction:

The VOTE system was used for classification purposes¹⁵. The following sites of obstruction were considered: velum, oropharynx (lateral oropharyngeal walls, tonsils), tongue base and epiglottis. Obstruction severity was divided into three grades (no obstruction; partial and complete obstruction). The configuration of the obstruction was divided into anterior-posterior, lateral and concentric.

Target-controlled infusion (TCI):

Target-Controlled Infusion (TCI) describes the infusion of drugs using microchip-controlled infusion pumps to achieve a target concentration in the blood. The aim of this method is the rapid achievement and maintenance of a certain effect (e.g., sedation) based on a specified (defined) plasma level or an effect level in the case of effect-site TCI. Calculation of the plasma level or effect-site level is based on pharmacokinetic 3-compartment models (according to Marsh or Schnider) that use the pharmacological half-life values and distribution coefficients determined in a volunteer population^16-18. The infusion rates needed to rapidly achieve and maintain the specified target level of propofol in plasma are then automatically calculated and controlled by the infusion pump. The present study used a system consisting of an infusion pump and a data manager together with the pre-programmed pharmacokinetic model of Marsh. The objective depth of sedation was simultaneously monitored by bispectral analysis.

Bispectral analysis:

The bispectral analysis/index is correlated with electrical activity in the brain. The monitor of bispectral analysis records frontal EEG signals and, with the help of various proprietary algorithms, analyzes the distribution of the EEG power spectrum. The bispectral index is a dimensionless number between 0 and 100 ¹⁹. In general, a value around 90 reflects a preponderance of high-frequency beta-waves and indicates that the patient is awake. Bispectral analysis values below 10 are indicative of EEG suppression ^{20, 21}. This scale therefore provides an indirect measurement of the effect of sedatives on the brain. In order to maintain an adequate depth of anesthesia and avoid an undesirable intraoperative awakeness of the patient, a bispectral index of < 60 is recommended. On the other hand, values of the bispectral index of < 40 should be avoided, in order to prevent unnecessarily deep anesthesia.

All sleep endoscopies were performed by two experienced otorhinolaryngology consultants with additional qualifications in sleep medicine (T. M. 59.6% (34/57), A. F. 40.4% (23/57).

Protocol

The protocol follows the guidelines of the local ethics committee of the Friedrich-Alexander University Erlangen-Nürnberg (FAU).

1. DISE-TCI-Bispectral Analysis Procedure

1. Perform venipuncture and place an IV catheter in the patients left back of the hand or median cubital vein to obtain intravenous access to administer the sedative.
2. Carry out the investigation at room temperature (22 °C).
3. Bring the awake patient into the operating theater. Place the patient in the supine position on the operating table.
4. Following this, have the anesthesiologist connect the monitoring system of vital signs. Monitor the patient during the whole procedure by peripheral pulse oximetry, 3-channel ECG and non-invasive blood pressure measurement at 3-min intervals.
5. Ensure that both the anesthesiologist and the anesthetic equipment are on the left side of the patient.
6. Have the anesthesiologist place and fix the single-use sensor for bispectral analysis.
   1. To do this, wipe skin with disinfectant and let it dry.
   2. Position sensor (4 interconnected and adhesive electrodes) diagonally on the patient's forehead: one electrode (number 1) at center of forehead, electrodes 2 and 3 directly above eyebrow and electrode 4 on temple between corner of eye and hairline.
   3. Connect sensor via interface cable to the bispectral analysis-monitor. Press electrodes firmly and ensure that the automatic sensor test is passed.
7. Have the anesthesiologist enter patients' individual sex, age, height, weight and plasma target concentration of propofol into the data manager with the pre-programmed pharmacokinetic model of Marsh¹⁷.
   NOTE: According to this model, the TCI-data manager calculates the infusion rates of propofol and operates the connected TCI-infusion pump to establish and automatically maintain the set target plasma concentration.
8. Dim the light in the operating room.
9. Induce sleep solely by the intravenous infusion of propofol with the TCI-infusion pump with an initial target plasma level of 2.0 µg/ml.
10. Under continuous monitoring of the patient's respiration, state of consciousness and value of the bispectral analysis, raise the target plasma propofol level in steps of 0.2 µg/ml/2 min until the desired depth of sedation is reached.
    NOTE: The target depth of sedation is a sleeping patient (eyes closed), with maintained spontaneous breathing and / or starting to snore and / or obstructive apnea observed by the examiner with the endoscope or the anesthesiologist.
11. Record manually the mean value of the bispectral analysis at the target depth of sedation.
12. Have the examiner with the endoscope stand at head level on the right side next to the patient.
13. At the target depth of sedation insert a flexible fiber endoscope transnasally via the inferior nasal meatus.
14. Film the endoscopic examination and store digitally. Use the image acquisition system according to manufacturer's instructions.
    NOTE: In case of nasal septal deviation or hypertrophy of the inferior turbinates, the inferior meatus with the larger diameter should be chosen.
15. Examine the upper airways by videoendoscopy in a proximal to distal direction from the nasopharynx as far as the hypopharyngeal entrance.
16. Classify the obstruction site(s) (velum, oropharynx, tongue base, epiglottis) and configuration/obstruction pattern (anterior-posterior, lateral, concentric) and the degree/severity (no obstruction/no vibration, partial obstruction/vibration, complete obstruction/collapse) according to the VOTE system at the target depth of sedation¹⁵.
17. In the case of isolated sleep endoscopy, terminate sedation of the patient by stopping the propofol infusion after completing the endoscopic examination.
18. In the case of sleep endoscopy as part of a planned surgical procedure, let the anesthesiologist perform total intravenous anesthesia by starting infusion of opioids (e.g., using remifentanil in TCI mode) and increasing infusion of propofol (e.g., set TCI level to 4 µg/ml) until loss of consciousness. Inject muscle relaxant (e.g., rocuronium) and intubate the patient.

Results

Characteristics of the patients are shown in Table 1. The mean value of the bispectral analysis at the desired depth of sedation was 60 (± SD 10.4, range 35 - 80; 95% confidence interval 42.0 - 47.6). Single or multiple combined sites of obstruction were found across all levels of OSA severity. The identified sites of obstruction, in decreasing order of frequency, were: velar 59.6% (34/57), tongue base 43.9% (25/57), oropharynx 42.1% (24/57) or at the level of the epiglot...

Discussion

The first experiments using flexible fiber optic rhinopharyngolaryngoscopy during spontaneous sleep to detect the site of obstruction was described as early as the 1970s²². However, the roots of this investigative technique were initially of a purely descriptive pathophysiological nature. During the establishment of surgical techniques to treat OSA, such as uvulopalatopharyngoplasty (UPPP) by Fujita, it was, however, shown that the success of the operation was apparently dependent on a careful preoperative sel...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Cardiorespiratory polysomnography
SOMNOscreen	SOMNOmedics, Randersacker, Germany	SBT202
Sedation
Propofol-Lipuro 20 mg/ml; 2,6-diisopropylphenol	B. Braun Melsungen AG, Melsungen, Germany
Target-controlled infusion (TCI)
Infusion pump Orchestra Module DPS Visio	Fresenius Kabi, Germany GmbH	Z082420
Data manager Orchestra Base Primea	Fresenius Kabi Germany GmbH	Z081320
Bispectral analysis (BIS)
BIS single-use electrode BIS Quatro Sensor	Covidien, Neustadt/Donau, Germany GmbH	186-0106
BIS monitor BIS VISTA	Covidien, Neustadt/Donau, Germany GmbH	186-0210
Endoscope
Laryngo fiberscope, length 30 cm, diameter 3.5 mm	KARL STORZ GmbH & Co. KG, Tuttlingen, Germany	11101RP
Picture Archiving  System
AIDA	Karl STORZ GmbH & Co. KG, Tuttlingen, Germany	WD 200-XX
Premedication
Catapresan; Clonidin-HCl 0.075 mg/0.15 mg/0.3 mg	Boehringer Ingelheim Pharma, GmbH & Co. KG, Ingelheim am Rhein, Germany