To the Editor,

We wish to present a novel cause of hypoxemia during induction of anesthesia in a patient with a nasocutaneous fistula. Written consent was obtained from the patient to report this case.

Anesthetic management of the patient undergoing complex midface and ophthalmologic surgery presents unique challenges to anesthesiologists. In this setting, nasocutaneous fistulae are known complications of a variety of primary pathologies such as granulomatosis with polyangiitis (GPA)1 and post-surgical course.2 Despite this, evidence is lacking to guide airway management in this setting. These fistulae establish a pathway for flow from the airway to the atmosphere with an external orifice that may be located outside the footprint of a standard anesthetic face mask. Consequently, preoxygenation may be compromised because of entrained room air through the fistula and bag-mask ventilation may be difficult because of circuit leak allowed by the fistula.

We report the case of a 61-yr-old female, ASA physical status III, with GPA who presented for orbital biopsy. She presented with a one-year history of right periocular pain associated with right maxillary and ethmoid sinusitis requiring sinus debridement and six months of antimicrobial therapy. Six weeks prior to surgery, she developed diplopia followed by development of a 2-mm hole between the medial canthus and the nose through which she felt air exiting and shower water entering the nasal cavity. Her computed tomography scan showed destructive sinusitis of the right ethmoid and maxillary sinuses with missing bone in the roof of the right ethmoid, medial wall, and floor of the right orbit (Figure).

FIGURE
figure 1

Computed tomography scan of right nasocutaneous fistula. The large arrow depicts the cutaneous orifice of the nasocutaneous fistula.

Full size image

The patient was positioned supine with standard non-invasive monitors. It was noted that the fistula fit within the footprint of a medium sized face mask. Consequently, no difficulty with bag-mask ventilation was anticipated. After placement of a peripheral intravenous line, preoxygenation was performed with four deep breaths over 30 sec with oxygen flow of 8 L·min-1. General anesthesia was induced with propofol and rocuronium. After loss of consciousness, bag-mask ventilation was attempted but resulted in an air leak through the nasocutaneous fistula and a consequent loss of the capnogram. The patient’s oxygen saturation decreased to 87% within one minute. An occlusive dressing was applied over the aperture of the fistula with successful restoration of effective bag-mask ventilation. Once the saturation had returned to 99%, the patient was intubated easily using a video laryngoscope and surgery was completed without further incident. The patient continued to have an unremarkable perioperative course.

Our case shows that location of the nasocutaneous fistula within the footprint of a face mask is not guaranteed to facilitate bag-mask ventilation. Consequently, emphasis should be on increasing safe apneic time with 100% oxygen flows at 5 L·min-1 with tidal volume breathing for three minutes or eight deep breaths within one minute at 10 L·min−1.3 Airway adjuncts and temporary closure of the fistula should also be considered. Case reports of temporary closure of fistulous tracts have been previously described with both dry pad packing2 and occlusive transparent dressing.4

Our experience suggests that similar cases require a detailed clinical assessment, evaluation of the lesion’s extension on imaging, followed by a management plan that should consider methods of managing loss of bag-mask ventilation. Whereas occlusive dressing was successful in our case, it may be necessary to consider elective nasal packing plus occlusive dressing prior to induction of anesthesia. Airway adjuncts including supraglottic airways and video laryngoscopy should be readily available and appropriate preoxygenation should be ensured prior to induction of general anesthesia.