Acute Awake Fiberoptic Intubation in the ICU in a Patient with Limited Mouth Opening and Hypoxemic Acute Respiratory Failure

The incidence and survival of patients with head-and-neck cancer have been on the increase for decades. Following surgery or radiation therapy, complications such as difficult airways may evolve. These difficult airways may be unique and not manageable with conventional intubation methods as well as video laryngoscopes. Acute awake fiberoptic intubation may be a feasible option also for urgent emergency airway management of known difficult airways. The “cannot intubate–cannot oxygenate” (CI–CO) situation has to be avoided at all costs, since emergency cricothyrotomy has a fail ratio of more than 50% when performed by an anesthesiologist.


Introduction
In Denmark, the incidence of head-and-neck cancer (HNC) has more than doubled since 1980 and the relative 5-year survival has risen from 49% to 72% [1]. Previous HNC surgery or radiation therapy is a strong predictor of a future difficult airway [2]. Of all patients undergoing an emergency cricothyrotomy due to a "cannot intubate-cannot oxygenate" (CI-CO) scenario, 75% went through HNC surgery at some point prior to the event [2]. Hence, it is likely that the proportion of patients with potentially difficult airways will increase in hospitals without ear-nose-throat (ENT) expertise, with acute symptoms requiring immediate treatment. Our hypothesis is that all hospitals managing patients in acute respiratory failure should not only have backup plans for unexpected difficult airways but also for the expected ones [3]. is optimally includes skills in awake fiberoptic intubation (FOI), which preserves spontaneous breathing and thus can prevent the risk of critical desaturation or a CI-CO situation [4].

Case Presentation
Written informed patient consent for publication was obtained. Approval from e Central Denmark Region Committee on Health Research Ethics was unwarranted.
A 71-year-old man (height 166 cm, weight 60 kg) with previous HNC, presented in the emergency room due to fever and increasing dyspnea during the last 4 days.
On arrival, the patient was fully alert and oriented. Respiratory rate was 40 breaths/min, peripheral saturation was 80% with 10 L/min oxygen delivered on a Hudson mask, blood pressure was 91/64 mmHg, heart rate was 110 beats/min, and temperature was 38.5°C. e Hudson mask was replaced with a reservoir mask with an oxygen flow of 15 L/min. A computed tomography scan including intravenous contrast was performed, which showed bilateral pneumonia and excluded pulmonary embolism (Figure 1). e patient was transferred to the intensive care unit (ICU), where high-flow nasal oxygen (HFNO) was administered with a flow of 60 L/min and a fraction of inspired oxygen (FiO 2 ) of 100%. However, this was quickly amended to noninvasive ventilation (NIV) with a positive end expiratory pressure of 12 cm H 2 O and a FiO 2 of 100%. Oxygenation remained unacceptably low (Table 1). Accordingly, the decision to intubate was made.
In 1999, the patient's tonsil cancer had been treated with radiation therapy. Unfortunately, this caused mandibular osteoradionecrosis, which had necessitated hemi-mandibulectomy and comprehensive reconstruction surgery (free fibula flap combined with a pedicled pectoralis major myocutaneous flap) on the right side in 2003 and on the le side in 2015. e surgeries were performed following awake FOI and surgical tracheostomy. In 2014, at another hospital, a laparoscopic cholecystectomy necessitated intubation, which was attempted a er the induction of anesthesia. A nasal approach utilizing a video laryngoscope (VL) with a hyper angulated blade (McGrath™ MAC X-Blade™, Medtronic, MN) failed, and the esophagus was intubated twice. is was attributed to the extreme limited mouth opening, which made visualization of the relevant structures impossible. Further nasal attempts were stopped due to bleeding, and intubation finally succeeded orally with a fiberscope. Since this episode, only awake FOI attempts had been performed.
Other comorbidities included arterial hypertension (treated with an Angiotensin-II-receptor-blocker), chronic jaw  pain (treated with oral morphine 40 mg/day) and a history of smoking for 28 years (½ pack of pipe tobacco/week). Despite his comorbidities, he was physically active and played golf on a weekly basis. His airway examination revealed a mouth opening of 10 mm and neck movement <80° (due to the bilateral fibrous myocutaneous flaps), among other factors ( Figure 2 and Table  2). Most important, however, was the history of previous neck radiation and difficult intubation.
In the ICU, the patient was placed in the upright sitting position to provide airway patency and to ease his respiratory efforts. Glycopyrrolate 0.2 mg IV (Meda, Solna, Sweden) was administered once. Sedation was titrated slowly with midazolam 0.5 mg IV (Hameln, Hameln, Germany) and S-ketamine 2.5 mg IV (Pfizer, Ixelles, Belgium), ensuring slight sedation but cooperation and responsiveness corresponding to a level 2 on the Ramsey Sedation Scale. Airway topicalization was initiated with lidocaine 2% (Mylan, Canonsburg, PA) 4 mL nebulized using the NIV machine. is was done twice at 5-minute intervals (total of 160 mg). Lidocaine 10% (AstraZeneca, Södetälje, Sweden) was sprayed onto the base of the tongue and posterior pharynx, subsequently deeper on each of 3 times at 3-4-minute intervals (total of 90 mg). Transtracheal injection of lidocaine 4% 1 mL was also administered (total of 40 mg). As potential backup plans, the right nasal cavity was prepared with lidocaine 2% with epinephrine 5 µg/mL, using both a mucosal atomizing device (MAD Nasal™, Teleflex, Wayne, PA) and lidocaine soaked ribbon gauze (total of 20 + 60 mg), and infiltration in front of the cricothyroid membrane with 3 mL (total of 60 mg) was performed. A er 12 minutes of preparation, the patient was changed back from NIV to HFNO and an uneventful oral FOI was performed with a reinforced LMA Fastrach™ (Teleflex, Waine, PA) endotracheal tube with an internal diameter of 7.5 mm using the aScope™ Regular (Ambu, Ballerup, Denmark). e FOI in itself took less than 2 min. Correct tube placement was confirmed visually and with capnography, Propofol-infusion was initiated (70 µg/kg/min) and the patient was connected to the ventilator. Five weeks of ventilator therapy followed. Eight days a er admission, a surgical tracheostomy was performed at the ENT facility of another T 2: Some data on predicted difficulties with intubation and backup plans for oxygenation. Highlighted (with a "+"), which of these features our patient presented [3,6,7]. In a recent review on elective awake FOI protocols in the operating room, no specific strategy regarding premedication, local anesthesia, or sedation could be declared superior to the others [5]. In the critically ill patient, with minimal physiological reserves, airway obstruction (due to over-sedation) and respiratory or circulatory collapse would be dangerous. Sedation should thus be kept to a minimum (or no sedation at all) and nonopioid medication including ketamine may be the safest choice. In the end, familiarity of the anesthesiologist performing the FOI with their drug of choice is probably the most important point [5]. e disadvantage of FOI is that blood and secretions in the airway or at the tip of the fiberscope may obscure the vision. A thin suctioning catheter can be placed nasally in the pharynx for continuous suctioning and cautious low-flow oxygen through the working channel of the fiberscope can be administered to continuously flush the tip. Both of these techniques were prepared for, but not needed since visibility was obtained immediately. Some argue that FOI is extremely time consuming. However, with the right preparations undertaken simultaneously, this may not be true [8,11].

Predictors of difficult direct laryngoscopy Predictors of difficult facemask ventilation
Other theoretical options would have been classical blind techniques used for decades, such as blind nasal [18], retrograde [19] or light wand intubation [20]. Nevertheless, these are increasingly becoming a lost art and even if local expertise exists, success rates are not necessarily acceptable [21]. It is possible that an oral retromolar approach with a rigid intubating stylet (e.g., Bonfils™ [22]) or a nasal approach with the newer rigid video-intubating stylet (Trachway™ [23]) may have worked, but equipment and expertise are not available in all departments. Placing the patient on extracorporeal membrane oxygenation (ECMO) before performing a surgical tracheostomy has been reported in extreme circumstances [24]. However, besides potential risk of major hemorrhaging, highly specialized equipment and expertise in thoracic and ENT surgery are required. e airway plan was an awake oral FOI. It was preferable to avoid the nasal route, since it was probably necessary to decrease the tube size and since nasal bleeding would have been disastrous. Accordingly, an awake nasal FOI was the second choice. As backup plan in hospitals with ENT surgical expertise an awake surgical tracheostomy would have been a viable option or could have represented the primary choice instead of the awake FOI. In hospitals without ENT surgical expertise (as in this case), an awake percutaneous dilatational tracheostomy (if experienced intensivists are present) or an awake cricothyrotomy could have been the potential backup plan, keeping in mind that these procedures may be technically very challenging, when performed in local anesthesia in the critically ill patient on spontaneous breathing. e last option in either case would have been an emergency cricothyrotomy performed under adequate anesthesia and full relaxation.
hospital. e patient was discharged to his home 8 weeks a er admission.

Discussion
National guidelines on the management of difficult airways have been published in numerous countries including the USA, the United Kingdom and Canada [3]. e ultimate goal of all airway management is to avoid a CI-CO scenario, since emergency cricothyrotomy performed by anesthesiologists fails in more than 50% of these situations [2,5].
For the unexpected difficult airway focus should be on prediction and planning, as to decrease the incidence of its occurrence in the anesthetized and paralyzed patient [3]. Although several bedside tests can be used to anticipate the likelihood of a difficult airway, unfortunately no clinical finding can reliably exclude a difficult airway [6,7]. If difficulties arise a er the induction of anesthesia, rescue plans and algorithms should focus on oxygenation and not just center around intubation [3]. Emphasis should be on both technical and non-technical skills [3].
For the expected difficult airway an awake airway securing technique may be the safest option, since it facilitates airway patency, oxygenation and protection against aspiration [3,4]. Awake intubation is considered a safe option when difficulties with intubation are anticipated and backup plans for maintaining oxygenation are expected to be difficult (facemask ventilation, supraglottic device (SGD) placement and/or emergency cricothyrotomy) [3,4]. e presence of obstructing airway pathology, risk of rapid desaturation (e.g. respiratory failure) and increased risk of aspiration all favor an awake approach [3,8].
e first awake FOI was described by Murphy in 1967 [9], and since the 1970's awake FOI has been the "gold standard" for managing anticipated difficult airways [3,4,10]. In recent years, awake VL intubation has been described with similar success rates as awake FOI [10]. In cases of uniquely altered anatomy (e.g., limited mouth opening, HNC in the oral cavity or upper airway and previous radiation), insertion of the VL blade with free sight may, however, be impossible [11]. In our case, use of a VL had previously been impossible due to the extremely limited mouth opening. Usually, the following minimal mouth opening must be present for reliable airway maneuvers: Macintosh intubation > 25 mm, SGD usage > 25 mm and VL intubation > 20 mm [8,12,13]. Nevertheless, successful use of these devices in patients with less mouth opening has been reported [14].
Experts disagree as to whether awake FOI skills should be mastered by every anesthesiologist or only restricted to practitioners who deal with these difficult airways on a daily basis [15]. A survey among US anesthesiologists in 2003 found that only 59% reported to have skills in FOI [16]. Opportunities to acquire FOI skills in the operating room have decreased since the introduction of the VL in clinical practice in 2002. An acceptable level of technical skills may be acquired a er performing 10 FOI's in general anesthesia and 15-20 on awake patients [8,17], but true expertise probably requires more training. is necessitates heavy weighting of trainees performing supervised FOI's, optimally as part of a routine