CO₂ laser lingual tonsil reduction as a treatment for tongue discomfort during hypoglossal nerve stimulation: a case report

Abstract

The purpose of this investigation is to demonstrate a multimodality approach to the surgical management of obstructive sleep apnea. Hypoglossal nerve stimulator (HGNS) implantation has been a life-changing procedure for many patients with obstructive sleep apnea. When activated it produces tongue protrusion via electrical stimulation of the hypoglossal nerve. This advances the lingual tonsil, making the pharynx diameter greater. Unfortunately, for some patients the electrical stimulation required is too high and awakens the patient. In such cases the patient’s fragmented sleep is not improved with the hypoglossal nerve stimulator. Here we present a case where hypoglossal nerve stimulator and CO₂ laser lingual tonsil reduction are used in conjunction to reduce the hypoglossal nerve stimulator setting required for airway patency, thereby allowing the patient to sleep through the night. For those patients who are unable to tolerate hypoglossal nerve stimulator settings, a combined approach with lingual tonsil reduction may be an alternative.

Citation:

Fontenot A, Liu SYC, Dewan K. CO₂ laser lingual tonsil reduction as a treatment for tongue discomfort during hypoglossal nerve stimulation: a case report. J Clin Sleep Med. 2024;20(11):1857–1861.

BRIEF SUMMARY

Current Knowledge/Study Rationale: Some patients are awakened by the impulse delivered by the hypoglossal nerve stimulator. For them, restorative sleep remains elusive.

Study Impact: This case report presents one potential method of surgically managing the patient who cannot tolerate high settings of the hypoglossal nerve stimulator, lingual tonsil reduction.

INTRODUCTION

Obstructive sleep apnea (OSA) is a chronic sleep disorder estimated to affect up to 38% percent of the population worldwide.^1,2 As a result of partial or complete upper airway collapse, ventilation is reduced or halted entirely, leading to impaired gas exchange and arousal from sleep. Adverse cardiovascular and neurocognitive outcomes have been strongly associated with this disorder.^3,4 Continuous positive airway pressure (CPAP) is the gold standard of treatment, but it has a low adherence rate due to poor tolerance.⁵ Fortunately, several alternative options exist, one of which is surgical implantation of a hypoglossal nerve stimulator (HGNS). The HGNS achieves upper airway stimulation through 3 main components: a pressure-sensing lead, a stimulating electrode, and a pulse generator. The pressure-sensing lead is placed within the intercostal space to monitor respiratory patterns. When inhalation is detected, a pulse is sent through an electrode encircling the hypoglossal nerve branches responsible for tongue protrusion. Stimulation helps reduce obstruction and dilate the airway through increased neuromuscular tone and tongue protrusion. The pulse generator is placed in the subcutaneous tissue of the chest and stores programmed settings. Commonly titrated settings include electrode configuration (bipolar or unipolar), amplitude (0.0–5.0 V), stimulation (20–40 Hz), and pulse width (60–210 µs). Clinicians can individualize and optimize settings in the generator, and patients can conveniently turn the device on or off with an external remote control.⁶

In 2014 this device was approved by the US Food and Drug Administration and is an option for individuals who meet the following eligibility criteria: (1) intolerant to positive airway pressure (defined as use < 4 hours per night, < 5 nights per week, or unwillingness to use) or have failed positive airway pressure (defined as apnea-hypopnea index [AHI] >20 events/h on positive airway pressure), (2) age ≥ 22 years, (3) moderate to severe OSA (defined as AHI of 15–65 events/h), (4) absence of complete concentric collapse of the velopharyngeal, (5) absence of anatomical abnormalities that prevent device performance, and (6) predominantly obstructive events (defined as central and mixed apneas ≤ 25% of AHI).³ Furthermore, HGNS implantation is recommended for individuals with a body mass index (BMI) ≤ 32 kg/m².^3,7,8 Diagnostic sleep endoscopy (DISE) is a sedative-induced sleep nasendoscopy performed to determine the site of the airway collapse during sleep. Patients with OSA can have airway collapse at 1 or more sites including the soft palate, base of the tongue (BOT), or larynx. Patients who have the best results with HGNS implantation have isolated BOT collapse. The hypoglossal nerve is responsible for BOT contraction and tongue protrusion. Therefore, stimulation of this nerve will cause airway dilation at the level of the BOT. Lingual tonsil or BOT hypertrophy is assessed during the DISE procedure as well as during the preoperative clinic visit.

Treatment response is generally determined by the Sher criteria, which require a ≥ 50% decrease in AHI and an overall AHI ≤ 20 events/h at the 12-month follow-up appointment.⁹ Appropriate selection of candidates is considered critical to achieve an adequate response. Still, even with careful selection, nearly one third of individuals are ultimately considered nonresponders.¹⁰ Although the cause of inadequate response in appropriately selected candidates is largely unknown, there are reports of treatment-resistant patients attaining a sufficient response by means of further intervention with advanced titrations postoperatively and through use of external devices such as a chin strap.^11,12

Here we report the case of a patient who was intolerant to CPAP therapy and HGNS despite numerous titrations. The patient became responsive to HGNS therapy with subsequent CO₂ laser reduction of the lingual tonsil.

REPORT OF CASE

A 74-year-old, nonobese male with a history of tonsillectomy, gastroesophageal reflux disease, seasonal allergies, and OSA initially presented to the sleep surgery clinic for fragmented sleep due to breathing interruptions. He was having difficulty with unrefreshing sleep, excessive daytime sleepiness, fatigue, and decreased attention during activities. Although CPAP therapy provided symptomatic improvement, he remained intolerant to multiple devices due to positional discomfort and insomnia. After 3 years of various CPAP trials, he was referred to the sleep surgery clinic for further evaluation. At that time, he had an AHI of 37.7 events/h and an O₂ nadir of 82% with a mandibular advancement device in place and a BMI of 24.96 kg/m². Physical examination was notable for bilateral septal spurs, a modified Mallampati tongue protrusion of 4, the presence of a long uvula, and a grade 4 lingual tonsil. To find an alternative treatment option, evaluation with DISE was completed and revealed anteroposterior collapse of the velum, stable lateral pharyngeal walls, and posterior displacement of the epiglottis by the BOT that would otherwise not collapse on its own. Considering his CPAP intolerance, appropriate BMI, and the absence of complete concentric collapse of the velopharynx, the patient was a candidate for HGNS therapy and proceeded with the surgery.

There were no complications with the procedure or during the postoperative course. After adequate healing, the HGNS was activated and titrated. Initial postoperative device settings were an electrode configuration of [+/−/+] with an amplitude range of 2.2–3.2 V. These settings produced bilateral tongue protrusion and adequate waveform excursion. Unfortunately, the patient was unable to tolerate the device for more than 1 hour per night due to tongue discomfort. He reported overstimulation causing interrupted sleep with any amplitude above 2 V. Multiple titrations were later performed to reach a tolerable and effective setting, yet the patient was bothered by overstimulation at most settings. An electrode configuration of [−/−/−] at 0.5 V was found to be most endurable. Whereas this titration stiffened the tongue with moderate protrusion, the lingual tonsils became an evident risk factor for obstruction with adequate stiffening but moderate advancement. The patient reported sleeping better with the stimulator on rather than off with this programming, but he remained symptomatic and still required daily naps.

Due to continued inadequate response, the patient was referred to a laryngologist and received further evaluation through fiberoptic endoscopic evaluation of swallowing and laryngostroboscopy, an indirect laryngoscopy in conjunction with intermittent light used to augment observation of true vocal cord motion. Fiberoptic endoscopic evaluation of swallowing was unremarkable and demonstrated a grossly normal swallow function. However, laryngostroboscopy was notable for moderate BOT hypertrophy with significant tufts on tonsillar tissue near the midline, and the patient was determined to have grade 4 lingual hypertrophy (Figure 1). Given the obstructive risk posed by the hypertrophied tonsil and the higher stimulation required for appropriate tongue protrusion, it was decided that the patient might benefit from lingual tonsil reduction. The patient agreed to and proceeded with CO₂ laser reduction of lingual tonsil using direct laryngoscopy, DISE, and electrical analysis of the implant with titrations under anesthesia. Following BOT reduction, optimal HGNS settings were established at an electrode configuration of [+/−/+] with an amplitude range of 0.8–2.0 V. The operation went well and was without complications. The patient was monitored overnight and remained hemodynamically stable without any acute events. At the 1-month postoperative appointment, he reported experiencing no pain, bleeding, or dysphagia. The patient reported a slight alteration in his sense of taste but stated that it was improving. Postoperative laryngostroboscopy demonstrated a 50% reduction of the lingual tonsil along with a well-healed tongue base. Following reduction, the epiglottic frenulum and lingual surface of the epiglottis were visible on laryngostroboscopy (Figure 2). There was no change in swallow function noted with fiberoptic endoscopic evaluation of swallowing. Thus, the patient was cleared to resume HGNS use. A home sleep study was completed 2.5 months after the surgery and indicated an improved AHI of 15.1 events/h and an O₂ nadir of 87% with the electrode configuration of [+/−/+] at 1.1 V or 1.2 V. Additionally, at the 4-month postoperative appointment the patient reported overall symptomatic improvement with the same programming.

Figure 1. Preoperative views of the lingual tonsil.

(A) Preoperative laryngoscopy view of the lingual tonsil. The lingual tonsil is noted to be large, pushing the epiglottis forward. (B) Preoperative view of the lingual tonsil with tongue protrusion.

Figure 2. Postoperative laryngoscopy view of the lingual tonsil.

The lingual tonsil is reduced in size by 50%. There is more room in the midline as well as laterally.

DISCUSSION

The degree of stimulation required to alleviate airway obstruction also prevented the patient from getting restful sleep, which drastically limited device use and ultimately the benefit of therapy. Despite many disappointing titrations over the course of 2 years, the patient eventually gained symptomatic relief and improvement in AHI by way of CO₂ laser reduction of the lingual tonsil paired with HGNS titration. The patient’s AHI decreased from 37.7 to 15.1 events/h, fulfilling therapy response criteria. To our knowledge, this is the first report of lingual tonsil reduction enabling sufficient HGNS treatment response.

Identifying prognostic indicators for successful HGNS therapy is gaining attention as these devices continue to evolve and prove to be effective. The parameters used to predict outcome response are largely based on the feasibility study conducted by Van de Heyning et al, in which they found positive predictors of a successful response to be a BMI ≤ 32 kg/m², an AHI between 20 and 50 events/h, and the absence of complete concentric collapse of the velopharyngeal upon DISE evaluation.⁸ There are many causes of intolerance to the HGNS. In the case presented here, repeat DISE was able to elucidate the cause. For this patient the HGNS settings required to displace this tissue during sleep were so high that they awakened the patient. CO₂ laser lingual tonsil reduction may be helpful in those patients who have lingual tonsil hypertrophy on preoperative examination. Additionally, these patients must demonstrate normal bolus propulsion on swallow evaluation. It would be detrimental to swallow function to reduce the volume of the lingual tonsil or tongue base in a patient who is preoperatively having difficulty propelling the food bolus. For those patients without lingual tonsil hypertrophy, in addition to HGNS intolerance, reduction of the tongue base is unlikely to improve their OSA or their HGNS tolerance.

Response being limited due to complete concentric collapse of the velopharyngeal was echoed by Vanderveken et al, who found it be a strong negative predictor of treatment success. In their study, all participants with the anatomic feature failed to achieve treatment success, whereas 86% of individuals without complete concentric collapse at this level responded to treatment.¹³ The poor response has been attributed to obstruction involving collapse of the lateral pharyngeal, which is less likely to respond to tongue protrusion.¹⁴ Obstruction can occur at multiple levels along the upper airway, causing airflow impediment, and elevated AHI values correlate with a higher percentage of multilevel airway collapse.¹⁵ Because the HGNS addresses BOT and palatal collapse, it is thought that upper airway stimulation would provide limited benefit in more severe cases of OSA due to residual obstruction. However, there are data to support a superior benefit in individuals with a higher AHI, even with a baseline AHI above 65 events/h.^16–18 Conversely, there are also findings that lack a significant association between an increased AHI and greater treatment benefit.^19,20 There are mixed reports within the literature regarding the impact elevated BMI has on HGNS therapy outcome as well. Multiple studies support that elevated BMI negatively affects HGNS treatment response.^19,20 As BMI increases, redundant soft tissue creates mechanical forces that augment pharyngeal collapsibility.²¹ Yet, Sarber et al found that BMI ≥ 32 kg/m² held no significant value for predicting treatment response.¹⁷ Other studies have reported similar findings.²²

When selecting candidates and accounting for the above-mentioned characteristics, the rate of response ranges from 66–100%.^3,23,24 Although advancements have been made, the varied response to therapy along with the conflicting data implies that when using the current method of preselecting patients there remains some unpredictability in achieving surgical success. Moreover, it highlights the need for deeper evaluation of current prognostic markers and for further studies to identify novel clinical predictors that can advance patient selection. Recently, Boroosan et al investigated the prognostic value of the change in airway cross-sectional area with awake tongue protrusion and identified a decreased cross-sectional area with awake tongue protrusion as a negative predictor for treatment success.¹⁸ Rather than widening the airway, tongue protrusion exacerbates airway obstruction in some of the population. This phenomenon was true for 38% of individuals regardless of OSA diagnosis.²⁵

Decreased cross-sectional area on tongue protrusion played a role in this case. At the time of upper airway stimulation implantation, the patient had satisfied the published selection criteria, yet he remained nonresponsive. The hypertrophic lingual tonsil, found above the BOT musculature bilaterally, added to the BOT bulk and decreased the oropharyngeal cross-sectional area when the tongue was moderately protruded. Consequently, greater hypoglossal nerve stimulation was required to adequately protrude the tongue and overcome the obstruction, but at the cost of arousal from sleep. In the past, reduction of the lingual tonsil for OSA treatment has been challenging and avoided due to risks of severe postoperative pain, postoperative hemorrhage, postoperative dysphagia, and dysgeusia. However, in this notable case, use of a CO₂ laser safely reduced the lingual tonsil and improved airflow, negating the need for higher stimulation and preventing arousal from sleep. The unique therapy combination of lingual tonsil reduction with CO₂ laser and upper airway stimulation is significant because it offers surgeons an additional, reliable option to create a more patent airway when managing OSA. The lingual tonsil is assessed during the preoperative examination and during the DISE procedure. The patient was noted to have some lingual tonsil hypertrophy (grade 4). Moreover, each patient undergoes preoperative swallow assessment. The lingual tonsil and BOT are necessary for adequate bolus propulsion in swallowing. Therefore, it is important to avoid lingual tonsil reduction when possible. Therefore, it is a first-line treatment, in a patient without concentric airway collapse, to treat OSA with HGNS implantation. Given the potential negative impact on swallow function, lingual tonsil reduction is reserved for those who fail or do not qualify for HGNS. Although additional investigation for better clinical predictors of HGNS response are necessary, the authors encourage clinicians to consider CO₂ laser BOT reduction as a treatment option for patients with lingual hypertrophy decreasing the cross-sectional area of the oropharynx on tongue protrusion when preselecting candidates as well as when managing nonresponsive patients with HGNS. Furthermore, we encourage clinicians to keep in mind the option of combined surgical intervention and device use when managing these patients.

CONCLUSIONS

As demonstrated by the above-mentioned case, surgical intervention through CO₂ lingual tonsil reduction paired with the use of a hypoglossal nerve–stimulating device has proven beneficial by reducing AHI and O₂ nadir in managing a patient with OSA. As a result, this patient is now responsive to OSA therapy and has improved quality of sleep.

DISCLOSURE STATEMENT

All authors have seen and approved the manuscript. Work for this study was performed at Stanford University. The authors report no conflicts of interest.

ABBREVIATIONS

AHI

apnea-hypopnea index

BMI

body mass index

BOT

base of the tongue

CPAP

continuous positive airway pressure

DISE

drug-induced sleep endoscopy

HGNS

hypoglossal nerve stimulator

OSA

obstructive sleep apnea