1 INTRODUCTION

2 PATIENT SELECTION

Patients diagnosed with obstructive sleep apnea in whom the preoperative evaluation confirms or suggests tongue base collapse as the cause of the hypopharyngeal obstruction are candidates for RF. Patients need to be in stable medical condition, understand that the procedure may be a multiple stage process, and appreciate the potential complications. Patients with preoperative complaints of dysphagia or dysarthria should undergo a modified barium swallow and/or a speech therapy evaluation prior to considering these therapeutic options, so as not to exacerbate these complaints post treatment. Those patients with underlying diabetes mellitus should be counseled regarding the increased risk of tongue abscess formation.

Two treatment approaches exist: (1) performing RF alone; (2) performing RF with other base of tongue procedures, such as genioglossus advancement, hyoid suspension or repose procedure.

3 OUTLINE OF PROCEDURE: TONGUE BASE RADIOFREQUENCY

The technique can be performed as an outpatient or as an inpatient. Depending on the severity of the patient’s sleep-disordered breathing, post-treatment airway monitoring and/or protection with nasal continuous positive airway pressure (CPAP) or a tracheotomy may be considered.

First, the oral cavity is prepped by the patient gargling with approximately 30 ml of 0.12% oral chlorhexidine (Peridex). Oral or parenteral cephalexin is administered prior to placing the electrode. If under local anesthesia, the tongue may be sprayed with 20% benzocaine as a topical anesthetic. In addition, a local anesthetic as a gel (2% lidocaine) may be placed on a sterile cotton tipped applicator at the tongue injection sites. A 25-gauge needle injects 5.0 ml of local anesthesia such as 0.25% bupivacaine (Marcaine) with 1:200,000 epinephrine into each site. A different sterile needle is used at each site of injection. Lingual nerve blocks may be used, but are not necessary. The author does not inject saline into the treatment area. There is an increased risk of infection with each additional injection at the lesion site by bringing potential superficial tongue debris and bacteria into the area to be ablated.

Currently, a dual 22-gauge RF needle electrode with a10 mm active length and a 10 mm protective sheath in a custom-fabricated device allows placement of the electrode under the superficial tongue musculature in the area selected for treatment. Potential sites of treatment include the midline and paramedian dorsal base of tongue areas and the anterior ventral tongue areas (Figs 40.1 and 40.2). Treatment sites should be spaced a minimum of 1.5 cm apart if a single needle probe is used or at least 2.5 cm using the dual needle probe device (Fig. 40.3). Continual pressure on the electrode and visualization that the insulation sheath is not retracting out of the tongue tissue reduce the risk of superficial tissue injury. Using the most current Gyrus radio frequency generator machines and a dual probe at 85°C, 600 joules (J) of rapid energy radiofrequency is delivered to two base of tongue sites with a total of 1200 J delivered. Treatment sessions may be repeated every 4–6 weeks.

Fig. 40.1 Tongue cross-sectional view. Note that the probe should be seated firmly against the dorsal tongue tissue to prevent withdrawal of the probe and injury to the superficial tissue. Also, the hypoglossal neurovascular bundle is located approximately 2.0–2.5 cm deep from the dorsum and should be safe from injury if the probe is placed no more lateral than the paramedian position.

Fig. 40.2 Tongue dorsum view. Note the position of the probe is at the midline or paramedian position of the dorsal tongue. The probe can be placed at the base of tongue, middle tongue area or ventral surface of the tongue to achieve tongue reduction.

Fig. 40.3 Tongue radiofrequency handpiece – double or dual probe. Each probe has a 1 cm distal active tip and a 1 cm proximal insulated component.

Internal Gyrus research comparing the lesions created by the previous generator equipment and the new generators revealed comparable lesion sizes with 1500 J delivered at the two single probe sites from the previous machine compared to 600 J delivered using the dual probe with the new generator. If tongue base RF is being performed simultaneously with other base of tongue procedures, such as the genioglossus advancement, the surgeon must understand that additional tongue swelling will occur. Appropriate airway monitoring and airway protection should be considered perioperatively.

4 POSTOPERATIVE MANAGEMENT AND COMPLICATIONS

Patients undergoing the procedure under local anesthesia, who have a prominent gag reflex, may benefit from a 10–20 mg oral dose of oral diazepam. Corticosteroids may be used postoperatively, but the author does not use them, because of the concern of tongue abscess formation. If postoperative swelling and airway compromise are a concern, the surgeon can consider using nasal CPAP or auto-titrating positive airway pressure devices and/or monitoring the patient as an inpatient with pulse oximetry or intensive care monitoring. Ice by mouth is recommended for the first 2 days to decrease tongue edema. Pain as a result of this technique is usually mild with a 2–3 day duration treated with acetaminophen, non-steroidal anti-inflammatory medications or low-dose opioid narcotic analgesics.

The main potential postoperative complications are superficial tongue ulceration, tongue abscess formation, upper airway obstruction from immediate post-treatment swelling or hypoglossal nerve injury. This later complication can result in temporary or permanent dysphagia and dysarthria.

To diminish the risk of infection in patients undergoing tongue base radiofrequency, a number of precautions and directions can be followed (Table 40.1). The incidence of tongue abscess formation in nearly every study ranges from 0 to 1.5%. Following the recommendations inTable 40.1 may limit the incidence of tongue superficial ulcerations and infection.

Table 40.1 Technical factors to decrease the risk of infection

To avoid hypoglossal nerve injury, it is best to limit the ablation sites to no more than 2 cm lateral to midline. The neurovascular bundle is approximately 2–3 cm from the surface epithelium and approximately 2–3 cm lateral from midline in a normal sized tongue (Figs 40.1 and 40.2).

Tongue discomfort, odynophagia and dysphagia may occur for 3–5 post treatment. To lower the incidence of significant post treatment discomfort one should follow these recommendation: (1) limit the rapid energy delivery per each dual probe to 600 J; (2) limit the treatment to two dual probe sites per treatment session; (3) use crushed ice by mouth as much as possible for the first 24–48 hours; and (4) follow the treatment techniques to limit the risk of infection.

5 TONGUE BASE RADIOFREQUENCY RESULTS

In the first published study evaluating RF for sleep-disordered breathing, a porcine study1 evaluated the relationship of lesion size to total RF energy delivery and subsequent tissue volume reduction. The study showed that tongue musculature volume could be safely reduced using RF energy in a precise and controlled manner with minimal risk of mucosal injury or infection. As the total energy delivered increased from 500 to 2400 J the lesion cross-sectional area increased from 0.7 cm² to 1.4 cm². After 3 weeks of scar formation and subsequent contraction the overall tissue volume reduced by 26.3%.

The first human tongue pilot study2 investigated the feasibility, safety, and possible efficacy in the treatment of sleep-disordered breathing. Eighteen patients were enrolled in the study. These patients were incompletely treated with upper airway reconstruction procedures, all of which underwent at least an uvulopalatopharyngoplasty. The mean preoperative Apnea/Hypopnea Index (AHI) of 39.6 improved to a mean of 17.8 (55%) post treatment; the mean Apnea Index (AI) of 22.1 improved to a mean of 4.1 (80%); and the mean preoperative SAO₂ nadir of 81.9% improved to 88.3% (12%) post treatment. The mean energy delivery per treatment session was 1543 J for a mean total energy delivery of 8490 J. Magnetic resonance imaging revealed a tongue volume reduction of 17% with an improvement of the posterior airway space by 15%. Speech and swallowing evaluated by VAS scores were unchanged. The only complication was one tongue abscess in 181 treatment sessions, which resolved after incision and drainage.

A long-term follow-up study3 on the patients in the initial human study revealed 16 patients completed the follow-up at a mean of 28 months after the initial treatment. Sleep studies were performed using a new nasal cannula by Sleep Solutions (Redwood City, CA), which can quantitate airflow. This technology is thought to be more accurate in identifying hypopneas. The follow-up data revealed that the AI remained the same at 5.4, but the overall AHI relapsed to 28.7. The SAO₂ nadir relapsed to 85.8%. There were no changes in the quality of life scale (SF36), Epworth Sleepiness Scale, speech or swallowing VAS scores. The conclusion of the authors was that the success of tongue base RF may reduce with time with the primary relapse in the AHI. The limitation of the study was the use of different monitoring devices to document hypopneas.

A prospective, non-randomized study4 evaluated 10 patients who underwent uvulopalatopharyngoplasty (UPPP), nasal surgery and tongue base RF as an inpatient. Two additional RF sessions were performed postoperatively in an outpatient setting. Nine of the ten patients received a cumulative dose of approximately 12,000 J.Five of these patients were successfully treated, defined as an AHI <20 with at least a 50% reduction in AHI. The mean preoperative AHI of 29.5 ±14.8 improved to a mean AHI 18.8 ± 14.6 and the mean AI of 8.7 ± 6.4 improved to a mean AI of 3.7 ± 4.9.

Another study5 compared the results of treatment with UPPP alone compared to tongue RF along with UPPP showing that the combined approach significantly improved the success rate in patients with level II and III sites of upper airway obstruction. The success rate increased in level II patients by 17.2% (37.9% vs. 55.1%) and in level III patients by 24.9% (8.1% vs. 33.0%) with the addition of tongue base RF to the UPPP procedure.

A multi-institutional study6 evaluated 73 patients with underlying obstructive sleep apnea; 56 (76.7%) of these patients completed tongue base RF and follow-up polysomnography. Sixty-five (92.9%) had prior pharyngeal surgery. A mean of 5.4± 1.8 treatment sessions were performed with a mean of 3.1 ± 0.9 lesions per treatment session and an overall energy delivery of 13,394± 5459 J.The mean AHI of 40.5 ± 21.5 decreased to a mean of 32.8±22.6 post treatment. The authors concluded that RF diminishes the severity of obstructive sleep apnea with subjective outcomes comparable to nasal CPAP.

In an investigation to utilize higher energy levels, 20 patients under local anesthesia were treated with two treatment sessions delivering 2800 J per session (700 J to each of four tongue base sites) for a total of 5600 J.7 The results revealed a reduction of the mean AHI of 27.7 to 22.6 and showed a 45% success rate, defined as an AHI less than 20 with at least a 50% improvement. A subsequent study by the same authors using 2800 J per treatment session and a mean number of 3.4 treatment sessions revealed a mean reduction of the AHI from 32.1 to 24.9 and a 33% success rate.8

Steward9 evaluated 22 patients who underwent both palatal and tongue RF treatments (mean 4.8 tongue and 1.8 palate sessions) and revealed statistical improvement in daytime sleepiness, and a significant improvement in the AHI from 31.0 to 18.7. The surgical success was 59% defined as an AHI less than 20 with at least a 50% improvement. There were no significant postoperative complications.

Twenty-six patients with mild to moderate sleep apnea were randomized to RF therapy for three tongue treatments and then two additional tongue and palate treatments.10 The initial three tongue treatments significantly improved the quality of life and the additional two tongue and palate RF treatments further significantly improved the quality of life as determined by the Functional Outcomes of Sleep Questionnaire (FOSQ), daytime sleepiness and reaction times.

A study evaluating the efficacy of ventral tongue probe placement combined with dorsal RF delivery in 20 patients revealed no significant complications and a mean AHI decreased from 35.1 ± 18.1 to 15.1 ± 17.4.11 The authors noted the limitations of the study with lack of a control group, small sample size and short-term follow-up, but had significant results of the combined treatment of the ventral and dorsal surfaces of the tongue.

A randomized trial compared tongue and palatal RF and nasal CPAP to sham-placebo. Both nasal CPAP and multi-level RF revealed improvement in quality of life parameters as compared to placebo and there were no differences between the RF and nasal CPAP groups.12

Tongue base radiofrequency has been shown to improve the amount of obstructions in patients with sleep-disordered breathing. The technique can be performed as an outpatient or inpatient, under local or general anesthesia with minimal risks of complications and minimal quality of life changes. Tongue radiofrequency volumetric tissue reduction yields improvement in the AHI of obstructive sleep apnea patients in all studies to date and is usually a low-risk procedure; however, following the preceding recommendations will aid in minimizing possible complications.13 If there is persistent obstructive sleep apnea confirmed by a postoperative sleep study, other stand-alone surgical procedures or in combination with tongue RF may be considered.