Drug-induced sedation endoscopy (DISE) classification systems: a systematic review and meta-analysis

Croft et al. proposed the first form of a DISE classification system which consists of three categories: Group A, Group B and Group C. Group A is simple palatal snoring. No obstructive event was observed. There was notable palatal vibration, occasional prolapse of the uvula into the nasopharynx and circumferential narrowing at the velopharyngeal level associated with the palatal vibration. Group B is single-level palatal obstruction. There was palatal obstruction at the velopharyngeal level. The obstructions were usually circumferential closure. Group C was multisegment obstruction. There were snoring and obstructive episodes simultaneously at the velopharyngeal and oropharyngeal level. The proposed classification system is easy to understand. However, it is not sufficient for detailed analysis and reporting of DISE findings.

Pringle et al. proposed a new DISE classification system based on Croft et al. DISE classification system [7]. Grade 1 represents simple palatal snoring. There is no episode of upper airway obstruction; however, noisy snoring is observed. The noise is caused by the vibration of the soft palate and vibration of the walls of the velopharyngeal sphincter and upper oropharynx. The first observable pattern is the flapping of the soft palate during inspiration. The second observable pattern is a narrowing of the velopharyngeal area in an anteroposterior or circumferential configuration. Furthermore, the tonsil can be observed to bulge into the narrowing of the airway. Grade 2 represents single-level palatal obstruction. Upper airway obstruction occurs only at the velopharyngeal level. The severity of the obstruction can vary from mild to very severe obstruction. The configuration of the obstruction can either be anteroposterior or circumferential obstruction configuration. Grade 3 represents multisegmental obstruction with intermittent oro-hypopharyngeal collapse. Palatal level obstruction occurs with intermittent oro-hypopharyngeal involvement with each inspiratory effort. Grade 4 represents sustained multisegment obstruction. There is sustained multisegment obstruction throughout the obstructive episode. Grade 5 represents tongue base obstruction. The tongue base obstruction occurs either because the tongue base falls back onto the posterior pharyngeal wall or the posterior pharyngeal walls appearing to close down onto the tongue base. In addition, the tongue base could be forward but the epiglottis obscured the laryngeal inlet to obstruct the airway. Pringle and Croft grading system proposed a more detailed DISE classification system for assessing upper airway obstruction; however, it can be quite challenging to understand at an early onset especially for new DISE examiners.

Camilleri et al. proposed a simple grading system as a DISE classification system with three grading levels. Grade 1 represents palatal snorers. Grade 2 represents mixed snorers. Grade 3 represents non-palatal (tongue base). Uvulopalatopharyngoplasty (UPPP) was the recommended treatment for grade 1. Medical treatment combined with UPPP was recommended for grade 2. Medical treatment and other surgical treatment except UPPP were recommended for grade 3. Medical treatments include weight loss, mandibular advancement or continuous positive airway pressure. Other surgical treatments include nasal surgery, tonsillectomy laser to tongue base, maxillary or mandibular osteotomy, hyoid advancement and tracheotomy. The proposed classification system provided a framework for planning surgical treatment based on classified DISE. However, the proposed treatment model is not sufficiently robust.

Quinn et al. proposed a DISE classification system based on airflow dynamic, sound generation and obstruction at the palate, tonsils, epiglottis and tongue base. Palatal obstruction is characterised as massive reduction in cross-sectional area of the velopharynx during inspiration to either a funnel-shaped aperture or a transverse. Also, palatal obstruction was categorised as the narrowing of the velopharynx due to a circular aperture as the uvula swing back and forth. Tonsil obstruction was characterised as intermittent obstruction with vibration of the tonsils marked by medialization of the tonsils on inspiration with the formation of a narrow slit in the sagittal direction. Tongue base obstruction is characterised as the narrowing of the airway at the level of the tongue base with visible vibration of the tongue against the pharyngeal wall. Epiglottis obstruction was categorised an omega-shaped epiglottis obstruction or epiglottic obstruction due to the tip of the epiglottis falling back onto and vibrating against the posterior pharyngeal wall. This was the first study to relate upper airway obstruction in conjunction with upper airway obstruction classification based on DISE. Without a well-developed 3D airflow model, it is hard to appreciate the significance of DISE classification in conjunction with upper airway obstruction.

Sadaoka et al. proposed a DISE classification system based on three groups (Group A, Group B and Group C) of sites of upper airway obstruction. Grade A represents single-level palatal obstruction. Palatal vibration and velopharynx obstruction are observed. Grade B represents a single-level tongue base obstruction. Only tongue base obstruction is observed. Grade C is multisegment obstruction. Circumferential closure from the velopharynx to the tongue base is observed. This DISE classification system provides a simplistic and easy to understand upper airway obstruction classification. However, it is limiting when a robust analysis is required.

Higami et al. proposed a DISE classification system based on different pharyngeal stenosis patterns during sleep. Pharyngeal stenosis patterns were classified into three categories: falling type, all-round type and bilateral type. The falling type is an obstruction type caused by the soft palate and tongue sinking due to gravity. All-round type is an obstruction type caused by submucosa of the pharynx. Bilateral type is an obstruction type where the palatine tonsils were hypertrophied. A simple treatment plan was noted based on the DISE classification and attempt to evaluate surgical treatment outcomes. However, long-term surgical outcomes were not explored.

Iwanaga et al. proposed a DISE classification system to assess changes in airway morphology before and after surgery. The DISE classification system consists of five types of obstructions: soft palate obstruction, circumferential obstruction, tonsillar-type obstruction, root-of-tongue obstruction and mixed-type (palate and tongue) obstruction. Soft palatal type of obstruction was defined as an obstruction caused by a uvula and soft palate posteriorly in contact with the pharyngeal. Circumferential palatal obstruction was defined as full-circumference soft palatal obstruction involving the posterior and lateral pharyngeal walls during expiration. The tonsillar obstruction was defined as an obstruction where the palatine tonsils come in contact at the midline during expiration. Root-of-tongue obstruction was defined as a single-level obstruction at the back of the tongue only. Mixed obstruction was defined as an obstruction where the soft palate obstruction and the root of the tongue obstruct the airway simultaneously.

Kezirian et al. proposed VOTE classification for analysing and reporting DISE findings. VOTE classification is a DISE classification system for reporting DISE findings based on analysis of the severity and configuration of obstruction at four anatomical levels: velum, oropharynx, tongue base and epiglottis. The severity of obstruction can be classified as 0, no obstruction (no vibration, < 50%); 1, partial obstruction (vibration 50–75%) and 2, complete obstruction (collapse, > 75%).

The configuration of obstruction at the velum level can be in anteroposterior, lateral or concentric direction. Oropharynx obstruction occurs in the lateral and concentric direction. Tongue base obstruction occurs only in the anteroposterior configuration. The configuration of epiglottis obstruction occurs in the anteroposterior direction. VOTE classification provides a clear and simple framework for robustly and accurately analysing upper airway obstruction. However, VOTE can be perceived to have oversimplified upper airway obstruction.

Vicini et al. proposed the nose oropharynx hypopharynx and larynx (NOHL) as a DISE classification system for diagnostic standardised examination for OSA patients. NOHL classification system is focussed on assessing the severity and configuration of upper airway obstruction based on four anatomical zones: nose, oropharynx, hypopharynx and larynx. The severity of obstruction was categorised in conjunction with the Müller manoeuvre grading system. Grade 1 is an upper airway obstruction between 0 and 25% airway collapse and also defined as no collapse during Müller manoeuvre. Grade 2 is an upper airway obstruction between 25 and 75% airway collapse. Grade 3 is an upper airway obstruction between 50 and 75% airway collapse. Grade 4 is an upper airway obstruction greater than 75% airway collapse and also defined as a total pharyngeal wall collapse during Müller manoeuvre. A generalist approach of anteroposterior, lateral and concentric configuration of obstruction was applied to the nose, oropharynx and hypopharynx. The laryngeal obstruction was assessed as positive or negative at the supraglottic and glottic level. The key advantage of NOHL classification is the provision of a scoring framework for upper airway obstruction during awake and sleep state. Furthermore, NOHL classification can potentially aid therapeutic decision-making and surgical outcome analysis. The key limitation is that two of these anatomical zones (nose and larynx) play significantly reduced roles in upper airway obstruction. [31] Also, there is no sufficient evidence on the potential of NOHL classification aiding therapeutic decision-making and surgical outcome analysis.

Bachar et al. 2012 proposed endoscopic grading system and severity index for accurately understanding upper airway obstruction patterns. The endoscopic upper-airway grading system is a DISE classification system that comprises of five anatomical obstruction sites: nose and nasopharynx (N), palatine plane or tonsils (P), tongue base (T), larynx (L) and hypopharynx (H). All forms of obstruction are categorised as either partial (1) or complete (2). It is important to note that anatomic level without obstruction is not labelled. The endoscopic upper-airway grading system uses a severity index (SI). The endoscopic sleep index is calculated by adding the digits as representation of the obstruction pattern at each level. The key novelty of the endoscopic upper-airway grading system is the concept of a formulaic calculated cumulative severity of obstruction for precision classification. However, severity index does not account for the configuration of an upper airway obstruction.

Victores et al. proposed upper airway obstruction classification system assessing surgical outcomes on OSA patients called upper airway obstruction classification system. The upper airway obstruction classification system consists of four anatomic zones: palate (velum), lateral wall/tonsillars, tongue base and epiglottis [29]. The degrees of obstruction for all anatomic zones are partial (≤75% obstruction) and complete (≥75%). The sustainability of obstruction for tongue base and epiglottis is intermittent and sustained obstruction [24]. Intermittent obstruction is an obstruction only during inspiration in an apnoeic episode. Sustained obstruction is an obstruction throughout an apneic episode. Intermittent and sustained obstructions of the palate and lateral wall/tonsillar pillars were not visualised. This was the first study to combine degree of obstruction and sustainability of obstruction. Its key limitation was classification based on only nasal surgery effect.

Gillespie et al. proposed DISE index as DISE classification to provide a clear framework for clinical airway function analysis, assisting surgical planning and standardised DISE techniques, training and interpretation. DISE index classifies the degree of anatomic collapse of the upper airway on an ordinal scale from 0 (no collapse) to 12 (multilevel, complete collapse).The anatomical zones examined are as follows: palate, lateral walls of hypopharynx, tonsils, tongue base and epiglottis. Obstruction at the palate, lateral walls of hypopharynx, tonsils and epiglottis can be categorised into no collapse, partial collapse and complete collapse. Obstruction at the tongue base is categorised as no collapse, partial collapse with lingual tonsils, partial collapse without lingual tonsils, complete collapse without lingual tonsils and complete collapse with lingual tonsils. The key contribution of the DISE index is showing a relationship between classification and treatment planning. However, the relationship between classification and treatment plan is not well defined.

Koo et al. proposed a different DISE classification system based on modifying the VOTE classification. The modified VOTE classification system is based on two anatomic levels: retropalatal level (the region of posterior to the soft palate) and retrolingual level (the region of the pharynx posterior to the vertical portion of the tongue). The retropalatal level was subdivided into the palate (anteroposterior diameter), lateral pharyngeal wall (lateral diameter) and tonsil (specific structure contributing to obstruction). The retrolingual level consists of the tongue base (anteroposterior diameter), lateral pharyngeal wall (lateral diameter) and epiglottis (specific structure contributing to obstruction). The severity of obstruction was categorised as no obstruction (0), partial obstruction (1, 50–75%) and complete obstruction (2, >75%). The key contribution of the proposed classification system is the attempt to relate upper airway obstruction to BMI and AHI. Further, a population-driven classification system is proposed for the first time (Korean population). Its key limitation small population size and scope.

Lee et al. proposed a DISE classification system to evaluate upper airway obstruction based on sleep position. This study proposed a DISE classification based on four anatomic zones: soft palate, lateral wall including palatine tonsils, tongue base and larynx including epiglottis. The degree of obstruction is as follows: 0, no obstruction; 1, partial obstruction (vibration with desaturation) and 2, complete obstruction (total collapse of airway with desaturation). Tongue-based obstruction was categorised as grade 1 (more than 50% obstruction displacement compared to the supine state) and grade 2 (more than 75% obstruction displacement compared to the supine state). This was the first study to propose a classification system combining upper airway obstruction and sleep position. However, the characterisation of upper airway obstruction and sleep position was not well defined.

Herzog et al. proposed a DISE classification system based on the vibration and collapsibility of the upper airway. Four vibration and collapsibility levels of the upper airway were examined: pharyngeal collapse at velum level, uvula/palate contact to posterior pharyngeal wall, dorsal movement of the tongue base and pharyngeal collapse base level. Pharyngeal collapse at velum level was graded 1–5 based on lateral or circular pattern of collapsibility. The contact of the uvula/soft palate to the posterior pharyngeal wall is classified as grades 1–4 based on anteroposterior pattern of vibration. The dorsal movement of the tongue base is classified as grades 1–4 depending on the visibility of the valleculae and contact to the posterior pharyngeal wall based on an anteroposterior pattern of movement. The pharyngeal collapse of the tongue base level is graded 1–5 depending on a lateral and circular pattern of collapsibility. The key novelty of this DISE classification system is having well-defined classification boundaries and relating vibration and collapsibility of the upper airway obstruction during simulated sleep and DISE. The key limitation of this classification system that simulated snoring is significantly not representative on sleep.

Carrasco-Llatas et al. proposed another modified VOTE classification system very similar to the original VOTE classification. The proposed modification is lateral and concentric configuration of tongue base obstruction which is not considered in the original VOTE classification.

Veer proposed P-T-L-Tb-E classification system as DISE classification system that is simple to learn, objective and easy to remember. P-T-L-Tb-E classification system is characterised by severity and configuration on upper airway obstruction based on five anatomical structures: palate (P1, P2, P3), tonsils (T1, T2, T3), lateral pharyngeal wall (L1, L2, L3), tongue base (Tb1, Tb2, Tb3) and epiglottis (E1, E2). P1 grade represents no fluttering of the palate, soft palate not obstructing and open airway between the oropharynx and nasopharynx. P2 grade represents a palatal anteroposterior obstruction and/or uvula obstruction due to dropping into a collapsed airway inferiorly. P3 grade represents a palatal circumferential collapse. This is often seen with lateral wall collapse inferiorly. T1 grade represents no tonsil obstruction. T2 grade represents less than 50% tonsil obstruction of the airway using the anterior commissure as the horizontal midpoint. T3 grade represents greater than 50% tonsils obstruction of the airway. L1 grade represents no lateral wall obstruction. L2 grade represents less than 50% lateral pharyngeal wall obstruction of the airway using the anatomical midline and the anterior commissure as the horizontal midpoint. L3 grade represents greater than 50% lateral pharyngeal wall obstruction of the airway. Tb1 grade represents no tongue base obstruction, tongue base not altering the normal resting position of the epiglottis and vallecula seen. Tb2 grade represents less than 50% tongue base obstruction where the airway is defined as the space between the anterior commissure and the pharyngeal wall. Tb3 grade represents more than 50% tongue base obstruction. E1 grade represents no epiglottic trapdoor phenomenon. E2 grade represents epiglottis trapdoor phenomenon. P-T-L-Tb-E classification system provides a robust classification system, and scoring framework for analysis DISE results with well-defined boundary conditions. Furthermore, P-T-L-Tb-E classification system provides a well-defined classification framework for computational analysis. The visualisation of the anterior commissure during DISE might be quite challenging.