Study objectives: To evaluate pain, swallowing, speech, edematous response, tissue shrinkage, sleep, snoring, and safety (energy limits and adverse effects) following radiofrequency (RF) treatment to the palate in 22 subjects with sleep-disordered breathing.
Design: This investigation is a prospective nonrandomized study. Polysomnography, radiographic imaging, and infrared thermography, along with questionnaires and visual analog scales, were used to evaluate the effects of RF treatment to the palate.
Setting: Treatments were delivered on an outpatient basis at Stanford University Medical Center.
Patients: Twenty-two healthy patients (18 men), with a mean age of 45.3±9.1 years, were enrolled. All were snorers seeking treatment and met predetermined criteria: a respiratory disturbance index ≤15, oxygen saturation ≥85%, and a complaint of daytime sleepiness.
Intervention: RF was delivered to the submucosa of the palate with a custom-fabricated electrode for a mean duration of 141±30 s with a mean of 3.6±1.2 treatments per patient. Reduction of their snoring scores determined the end point of the study.
Results: Neither speech nor swallowing was adversely affected. Pain was of short duration (0 to 48 h) and was controlled with acetaminophen. There were no infections. Although there was documented edema at 24 to 48 h, there were no clinical airway compromises. Polysomnographic data showed improvement in esophageal pressure measurements of the mean nadir and the 95th percentile nadir (p=0.031, p=0.001) respectively, as well as the mean sleep efficiency index (p=0.002). Radiographic imaging showed a mean shrinkage of 5.5±3.7 mm (p≤0.0001). Subjective snoring scores fell by a mean of 77% (8.3±1.8 to 1.9±1.7, p=0.0001) accompanied by improved mean Epworth sleepiness scores (8.5±4.4 to 5.2±3.3, p=0.0001).
Conclusion: The results of this investigation allowed the formulation of safety parameters for RF in this defined population with mild sleep-disordered breathing. There was a documented tissue reduction and improvement in symptoms in all subjects. However, given the small sample size and short-term follow-up, these results should be confirmed by further investigation.