Direct laryngoscopy in this case was more difficult than predicted and was Cormack-Lehane grade IV. Among the preoperative findings, Mallampati class III was predictive of difficult intubation. However, it has been noted that a single bedside screening test is not sufficient to predict difficult intubation [2]. The sensitivity of Mallampati class III or higher for predicting difficult intubation derived from 47 studies involving 23,396 patients was reported to be 0.55 [3]. Therefore, several physical findings are combined in common clinical practice. In the “JSA airway management guideline 2014” [4], the Japanese Society of Anesthesiologists proposed 12 preoperative assessment factors that predict combined difficult facemask ventilation and direct laryngoscopy. In the present case, four of the 12 items were consistent: Mallampati class III, age over 46 years, presence of teeth, and limitation of forward mandibular movement (mandibular protrusion test class B). The likelihood of simultaneous occurrence of difficult mask ventilation and difficult direct laryngoscopy in this model was 0.47%, indicating that it was difficult to predict based on physical findings alone. A recent systematic review listed inability of upper lip bite, a short hyomental distance, and retrognathia as the best predictors of difficult endotracheal intubation [3]. Thus, mandibular size and forward mobility are strongly associated with difficulty in direct laryngoscopy. Although the upper lip bite was untested in this case, the forward mobility limitation of the mandible was not severe.

Ul Haq et al. reported that patients with Mallampati class III and IV have a 6.38-fold increased risk of difficult intubation, and patients with mandibular protrusion test class B and C (lower incisors cannot be protruded edge to edge with upper incisors) have an 8.32-fold increased risk of difficult intubation. Of the 760 patients enrolled in the study, 9.3% had Mallampati class III or IV, and 8.2% had mandibular protrusion test class C; however, none had Cormack-Lehane grade IV [5]. Given that the present patient had a moderate risk of difficult intubation, our findings of Cormack-Lehane grade IV made this an unusual case.

Another significance of this case report is that the present case had difficulty with direct laryngoscopy, probably due to a bone deformity of the upper cervical vertebra, despite normal neck extension. Visualization of the glottis during direct laryngoscopy requires alignment of the operator’s gaze with the glottic axis, which has been shown to be approached by extension of the occipito-atlantal joint in awake healthy volunteers [6]. A study using lateral fluoroscopy to record the movement of the occiput to the cervical spine in patients under general anesthesia also reported that most of the cervical motion during direct laryngoscopy occurs at the occipito-atlantal and atlanto-axial joints [7]. RA is associated with cervical spine involvement in up to 80% of patients, with the upper cervical spine being particularly susceptible [8]. Many anesthesiologists are aware of the possibility of airway compromise due to limited mobility of the cervical spine in patients with RA. However, joint damage is commonly associated with instability, such as subluxation and neurological symptoms. Therefore, the asymptomatic course of this patient may have led to an underestimation of cervical spine deformity. Takenaka et al. reported the limitations of the physical cervical extension test, such as the Bellhouse test, for limited upper cervical spine mobility, and proposed the usefulness of the ratio of hyomental distance in head extension and neutral position (hyomental distance ratio; HMDR) as an airway assessment in RA patients [9]. Although HMDR may be a good predictor of decreased cervical extension capacity, accurate measurement may be difficult to attain. In this case, neck rotation was limited despite preserved cervical retroflexion. The range of motion of neck rotation in adults without cervical spine involvement is approximately 70° to one side [10, 11], so this case is clearly small. Considering that at least 60% of neck rotation is produced by the upper cervical spine [11], a range of motion of 18° in this case may raise suspicion of limited upper cervical motion. Although there is no evidence to support neck rotation as a screening for limited upper cervical motion, it may be useful as a reference to prompt further evaluation by radiography or CT.

Although the patient did not appear to have impaired extension of the entire neck, radiographic examination in the retroflexed position was decisive in diagnosing limited upper cervical motion. The presence of a shortening of the occipito-atlantal or -axial distance in the retroflexed position is directly diagnostic of the extension capacity of the occipito-atlanto-axial complex. CT scans are useful for detailed evaluation of bone and joint deformities.

In conclusion, we present a case of Cormack-Lehane grade IV that was not predicted preoperatively. Anesthesiologists should be aware that RA patients may have limited upper cervical spine motion, which is essential for direct laryngoscopy, despite normal cervical retroflexion. Neck rotation may be useful as a reference to facilitate further assessment of upper cervical spine mobility.