Key words: Complication: difficult intubation. Technique: three-dimensional computed tomography scan.
SURGERY for malignant tumors of the neck or mandible frequently results in severe postoperative anatomic distortion of the pharynx and the larynx. When these patients undergo reoperation for recurrent tumors or reconstructive surgery, laryngoscopy and intubation may be difficult. Although numerous reports describe how to predict difficult intubation, [1-4]there is little mentioned about the preoperative evaluation of surgically distorted airways. Recently, helical scan computed tomography (CT) has become available to provide a three-dimensional imaging (3D-CT) of the skull, vessels, and tumors.* We have used this new technique to evaluate the airway of patients in whom distortion of airway was evident.
A 59-yr-old woman was scheduled for reconstruction of the mandible, half of which had been resected 6 yr earlier because of carcinoma of the oral floor and lower gum. The hard and soft palates, but not the uvula, were visible with mouth-opening (Mallampati classification III ). The patient received 2 mg midazolam 1 h before induction of anesthesia. Awake nasotracheal intubation was attempted after sedation with 200 micro gram fentanyl and 5 mg midazolam intravenously and topically applied 4% lidocaine. The epiglottis could not be visualized with direct laryngoscopy because of restricted tongue movement and a narrow pharyngeal space. Fiberoptic laryngoscopy via the tracheal tube was unsuccessful in identifying the epiglottis and vocal cords again, because of the narrow pharyngeal space, and surgery was postponed. The next week, a 3D-CT of the neck was obtained (SOMATOM PLUS-S, Siemens, Germany). During scanning, the patient was asked to open her mouth and extend her tongue as much as possible. 3D-CT of a healthy volunteer was taken to compare with this patient (Figure 1). Major findings of this case were (1) left pyriform fossae and left vallecula epiglottica almost nonexistent, (2) posterior displacement of the tongue, (3) narrow space between the epiglottis and posterior pharyngeal wall, and (4) anterior shift of the larynx (Figure 2). The following week, fiberoptic-assisted nasotracheal intubation was attempted again. Although the epiglottis was not seen, we noted two orifice-like structures. The first orifice was located to the upper right of the second one. The 3D-CT image suggested that the upper right orifice was a part of the vallecula epiglottica. Thus, we proceeded with fiberscopy via the left orifice to find the larynx. Thereafter, the trachea was intubated without difficulty.
A 54-yr-old man was scheduled for resection of the tongue and neck because of recurrent carcinoma of the oral floor. He had undergone hemimandibulectomy and neck resection 8 months earlier, followed by radiation therapy for several weeks. The patient could open his mouth about 2 cm. The tongue was distorted and could be extended only just beyond the lower teeth. The hard palate, but not the soft palate nor the uvula, was seen (Mallampati classification IV ). Preoperative 3D-CT revealed (1) narrow air space between the base of the tongue and posterior pharyngeal wall, (2) relatively wide space behind the epiglottis, (3) normally shaped epiglottis, and (4) no distortion of the larynx (Figure 3). After premedication with 5 mg midazolam and 0.5 mg atropine, nasotracheal intubation was attempted under sedation with 200 micro gram fentanyl and topically applied 4% lidocaine. Although the epiglottis could not be recognized using direct laryngoscopy, based on the CT scan showing that the pharynx was normally aligned with the larynx, which in turn was located in the middle of neck and was not distorted, we successfully intubated the trachea without further instrumentation.
Factors used to predict difficult tracheal intubation include increased body weight; decreased head, neck, and jaw movement; receding mandible; and protruding incisors. Radiologic measurements have been attempted to predict difficult intubation. [4,8]Few reports have evaluated intubation of patients with the pharynx or larynx distorted by surgery. Methods for preoperative evaluation of these patients are limited. Although preoperative fiberscopic examination provides some information, it is difficult to obtain clear views to reveal the distorted structures in the narrowed pharynx. Frontal and lateral x-ray views of the neck show the larynx as an air space. Once distorted, however, overlapping shadows make it difficult to obtain accurate detailed information. Conventional CT scan of the neck provides a clearer anatomic image than that provided by plain x-rays. However, such images are two-dimensional, and no sagittal view can be obtained 3D-CT overcomes these limitations.
For predicting difficult laryngoscopy according to Mallampati classification, Tham et al. reported that phonation while sitting improves the view of pharyngeal structures. However, it is difficult to continuously phonate for 30 s during scanning, especially for aged patients. Alternatively, we asked patients to open their mouths fully and to extend the tongue as much as possible. This maneuver creates an air space in the pharynx. The outline of the 3D-CT image is presented as a cast of air space of the pharynx and the larynx (Figure 1and Figure 3). In case 1, the main region of deformity was shown as a defect of the air space in the outline (Figure 2(A)). Information on the area from the mouth to the trachea is obtained via the sagittal plane (Figure 1, Figure 2, Figure 3). A narrowed space behind the epiglottis suggests the posterior shift of the base of tongue. Images cut incrementally from the top of the pharynx toward the larynx provide views similar to those seen during laryngoscopy or fiberscopy from the mouth to the larynx. A view through the posterior pharyngeal wall shows a posterior view of the uvula and the epiglottis. In the first case, the epiglottis was small and asymmetric. A small space above the epiglottis (vallecula epiglottica) was observed. This would have confused anatomic orientation from the limited view available using fiberscopy. A 3D-CT image in case 2 revealed that the epiglottis was neither distorted nor shifted toward either side. A straight line could be drawn from the nasopharynx toward the larynx. For that reason, we selected the blind nasal approach without fiberscopy. In contrast, the anterior larynx in case 1 made it difficult to intubate via either an oral or a nasal approach.
Helical scanning, a newly developed method, can scan a wide area of body quickly (30 s) compared with a conventional CT scan. Helical scan data are sequential, and a large number of tomographs can be reconstructed by interpolated algorithms.* High-resolution 3D-CT images are drawn from those volume data. Selecting the endoscopic mode, an air space is reconstructed as a cast with a certain wall width. Internal structures are also clearly shown. Once a three-dimensional image is available, the objects can be tilted, rotated, and cut freely and repeatedly. In the latest model, images of the pharynx and the larynx can be superimposed over translucent images of bones, such as the mandible, vertebra, and hyoid, which will enable us to evaluate difficult airways more precisely.
In summary, we had experiences dealing with two cases of difficult tracheal intubation with direct laryngoscopy because of previous surgery for carcinoma of the oral floor and the tongue. After evaluation with 3D-CT images of the pharynx and larynx, the trachea was intubated successfully in both cases. These results suggest that 3D-CT of surgically distorted pharynx and larynx may be helpful for tracheal intubation with or without fiberoptics.
* Moriyama N: Application of helical CT scanning to the examination of the abdomen and thorax. Medical Review 38:24-30, 1991.