Click here for the PDF version

Contents

1. To report that lingual tonsils may cause recurrent episodes of choking and pseudo-hemoptysis, as well as obstructive sleep apnea (OSA).
2. To discuss new CO₂ laser technology for lingual tonsillectomy.

A 14-year-old male with Trisomy 21 and medically controlled hypothyroidism presented with an intense fear of swallowing solid food after a severe choking incident. He had awoken at night and tried to eat a hotdog in a dark room. His mother "heard him choking and found him turning blue and unable to breath". She performed the Heimlich maneuver, which expelled a large piece of hotdog. He never lost consciousness.

Over the next few weeks, he had several other, less severe episodes of choking on solid food. Eventually, he began to refuse to eat solid food altogether. He also had chronic purulent rhinorrhea and obstructive sleep apnea episodes lasting 4–5 s. His tonsils and adenoids had been removed 10 years previously at another hospital.

Flexible endoscopy revealed large adenoids and abundant lymphoid tissue on the base of tongue, completely filling the vallecula. The posterior hypopharynx was asymmetrical such that the mucosa was touching the right arytenoid. The larynx appeared normal (Fig. 1).

Videofluoroscopy revealed that when he swallowed semi-solid foods, a portion became trapped within the lingual tonsils (Fig. 2). He was unaware of this and had to be given liquid to clear the semi-solid food from the hypopharynx. He had no aspiration or esophageal obstruction.

Computerized tomography revealed no evidence of lingual thyroid or neoplasia (Fig. 3). The patient was treated with topical nasal saline and steroid sprays and ranitidine for several months, without significant improvement. Adenoidectomy helped his rhinorrhea but not his other symptoms. Subsequent overnight pulse oximetry suggested mild to moderate obstructive sleep apnea (OSA), with 115 oxygen desaturations below 90% over 1% of total sleep time and an oxygen saturation nadir of 76%.

Partial lingual tonsillectomy was performed with the CO₂ laser set at 4 W in superpulse mode, with the beam slightly defocused. Char unfortunately limited the depth of the procedure. He had no oxygen desaturations postoperatively. His pain was well controlled with acetaminophen.

Three months later, his mother requested complete removal of his lingual tonsils, as she had noted partial improvement in his choking and OSA after the partial lingual tonsillectomy. A new Sharplan CO₂ laser was used, in flashscan mode at 20 W with a 3-mm spot size. Relatively char-free ablation of all of the lingual tonsils was achieved. Again, the patient had an uneventful postoperative course.

Remarkably, he had complete resolution of his choking and OSA within a week. This was confirmed with a normal videofluoroscopic swallow study and overnight pulse oximetry. He has remained symptom-free for 1 year.

A 13-year-old girl presented with recurrent mild hemoptysis, which eluded diagnosis for 2 years. She would typically cough up about 5–10 ml of fresh blood mixed in saliva a few times a day for about a week. The hemoptysis would abate spontaneously but then return a few months later. She had no other history to suggest coagulopathy and clotting profiles and chest radiography were normal.

Over the preceding 2 years, she had undergone laryngobronchoscopy three times – twice with brush biopsy – and all findings were normal. At the time of her fourth laryngobronchoscopy, 1–2 ml of blood was noted in the middle of her lingual tonsil, which was hypertrophic and mildly inflamed. A biopsy was taken and minor bleeding was controlled with a suction cautery.

When interviewed postoperatively, the patient admitted that she tended to clear her throat rather than cough deeply before each episode of hemoptysis (or more accurately, pseudo-hemoptysis). She also complained of fluctuating odyntophagia, globus sensation, and snoring. She was prescribed amoxicillin for 1 week. The biopsy was reported as a mildly inflamed lymphoid papilla. Her pseudo-hemoptysis resolved for 3 months but then recurred.

At this point, flexible laryngoscopy revealed moderately inflamed lingual tonsils, with a small amount of white exudate and a few prominent blood vessels just right of the midline (Fig. 4). Bacterial and fungal cultures were negative. Over the next few weeks, she had several more episodes of pseudo-hemoptysis. She was therefore booked for lingual tonsillectomy.

The CO₂ laser was used in flashscan mode with a 3-mm spot size. Power was gradually increased from 20 to 50 W. This achieved good hemostasis of several small blood vessels as well as char-free ablation of the entire lingual tonsil.

Her recovery was uneventful. Her snoring, odyntophagia, and globus sensation resolved and she has been free of pseudo-hemoptysis for 9 months.

For all three procedures, the patients were intubated with a metal orotracheal tube and standard laser safety precautions were taken. Exposure was obtained with a Benjamin laryngoscope. About one-third of the lingual tonsil could be exposed at a time. The laryngoscope, suspension apparatus, and orotracheal tube were therefore repositioned periodically to expose other thirds of the lingual tonsil.

The lingual tonsils were ablated with the CO₂ laser microscope and not dissected as a single mass. Care was taken not to injure the tongue musculature or the epiglottic mucosa. Mild bleeding was controlled with neurosurgical patties soaked in oxymetazoline. Blood loss for each case was about 10 cm³.

The patients were treated intraoperatively with dexamethasone (0.2 mg/kg IV).

After each surgery, a post-tonsillectomy diet was ordered. The head of the bed was elevated to at least 30°. Oxygen saturation was continuously monitored overnight. No supplemental oxygen was needed.

The patients were treated with amoxicillin (40 mg/kg/day, divided BID) for 1 week and ranitidine (5 mg/kg/day, divided BID) for 6 weeks. Pain was well controlled with acetaminophen. No narcotics were needed outside the recovery room.

The patients were advised to avoid non-steroidal anti-inflammatory medications for 2 weeks pre- and post-operatively. No postoperative bleeding occurred.

In their series of 25 adults with symptoms of chronic lingual tonsillar hypertrophy (LTH), Golding-Wood and Whittet [1] reported that one patient experienced "choking on food". A review of the recent English language literature reveals no previous case in which the lingual tonsils were found to cause choking in the pediatric age group or to cause pseudo-hemoptysis at any age [1, 2, 3, 4, 5, 6, 7, 8]. Patients with hemoptysis should have a careful evaluation of the entire upper aerodigestive tract, including the lingual tonsils.

OSA from LTH is rare in children and "extremely rare" in adults [2]. Most children with OSA from LTH are close to or in their teenage years; however, OSA from LTH has been reported in a child as young as 5 years [3]. Several previous studies have reported that LTH may cause OSA in older children with Trisomy 21 [3, 4].

Lingual tonsils may also cause pain, globus sensation, and chronic cough. They are the last structure of Waldeyer's ring to involute. Unfortunately, they have often been overlooked as a cause of upper aerodigestive symptoms [1].

Many papers have chronicled the technical evolution of lingual tonsillectomy from sharp dissection to cryotherapy to the use of cutting diathermy, suction cautery, and finally, the CO₂ laser [1, 5, 6]. By minimizing thermal injury to the tongue and epiglottis, the CO₂ laser causes the least airway edema, pain, and bleeding [5]. Disadvantages of the CO₂ laser include cost and need for a metal endotracheal tube to prevent airway fire. Whether the CO₂ laser will eventually be replaced by radiofrequency technology for lingual tonsillectomy remains to be seen.

In 1988, Wouters et al. [5] reported that exposure for CO₂ laser lingual tonsillectomy was facilitated by nasotracheal intubation and the use of a specially designed wide laryngoscope (Medin-Groningen). This laryngoscope was not available for the patients in this study. Although the orotracheal tube and the Benjamin laryngoscope required periodic repositioning, this was a minor inconvenience. At such times, communication between the surgeon and anaesthesiologist is essential for patient safety.

In terms of surgical time and avoidance of char during CO₂ laser lingual tonsil ablation, the Sharplan flashscan mode was found to be vastly superior to the superpulse mode. With each firing of the CO₂ laser in flashscan mode, a 3-mm area of lingual tonsil was vaporized in rapid, spiral fashion without char or significant bleeding. With this new technology, it seems unnecessary to excise portions of the lingual tonsil with the CO₂ laser (as described by Wouters et al. [5]), unless a specimen is desired for histology. Likewise, with flashscan mode CO₂ laser ablation and the occasional use of neurosurgical patties soaked in oxymetazoline, a suction cautery is unlikely to be needed for hemostasis in lingual tonsillectomy.

Both patients were treated empirically with intraoperative dexamethasone and postoperative amoxicillin and ranitidine. Use of these medications seems to be prudent, especially in children, in order to minimize airway edema, pain, and the risk of secondary hemorrhage. It is remarkable that neither child in this study required oxygen or narcotics outside the recovery room.

These are the first cases reported in which the lingual tonsil was found to be the cause of choking in the pediatric population or pseudo-hemoptysis at any age.

CO₂ laser lingual tonsillectomy is a safe and effective procedure, which has been facilitated by recent advances in laser technology.