Sleep Apnea Surgical Treatment

Uvulopalatopharyngoplasty (UPPP), introduced by Fujita in 1981, was the first widely adopted surgical treatment for obstructive sleep apnea (OSA). Early enthusiasm was tempered by recognition of modest success rates (approximately 40-50% achieving AHI reduction >50%) and significant complications including velopharyngeal insufficiency and dysphagia. The Sher meta-analysis (1996) confirmed that UPPP as a standalone procedure had limited and unpredictable efficacy, largely because it addressed only one level of obstruction in a disease that typically involves multiple anatomic sites (palate, tongue base, lateral pharyngeal walls, epiglottis). This recognition of multi-level obstruction drove the development of multi-level surgical approaches and the search for better methods to identify the site and pattern of upper airway collapse.

The STAR trial (Stimulation Therapy for Apnea Reduction) was the pivotal multicenter prospective study that established hypoglossal nerve stimulation (HNS) as a transformative treatment for moderate-to-severe OSA in CPAP-intolerant patients. Published in the New England Journal of Medicine in 2014, the STAR trial implanted the Inspire Upper Airway Stimulation system in 126 patients and demonstrated a 68% reduction in median AHI (from 29.3 to 9.0 events/hour) and significant improvements in oxygen desaturation index and quality of life at 12 months. The response rate (AHI reduction ≥50% and AHI <20) was 66%. Crucially, the trial used drug-induced sleep endoscopy (DISE) to exclude patients with complete concentric collapse at the palatal level, establishing DISE as a patient selection tool. This trial led to FDA approval of HNS and represented the most significant advance in OSA surgery in decades.

The ADHERE (Adherence and Outcomes of Upper Airway Stimulation for OSA) international registry confirmed the STAR trial findings in real-world practice, with 5-year data from over 640 patients demonstrating sustained AHI reduction (median AHI 6.4 at 5 years) and high adherence rates (mean nightly use 6.5 hours — substantially higher than typical CPAP adherence). Drug-induced sleep endoscopy (DISE) evolved from a research tool to a standard preoperative assessment, allowing personalized surgical planning based on the specific pattern and level of pharyngeal collapse. Multi-level surgery approaches combining palatal, tongue base, and skeletal procedures were evaluated in the SAMS trial (2020), which demonstrated that multi-level surgery in selected patients could achieve outcomes comparable to CPAP in terms of quality of life improvement. Maxillomandibular advancement (MMA) remained the most effective traditional surgical option for severe OSA but was reserved for select patients due to its complexity.

The AASM clinical practice guideline (2021) positions CPAP as first-line therapy for moderate-to-severe OSA, with surgical options recommended for CPAP-intolerant patients. HNS is recommended for select patients with moderate-to-severe OSA (AHI 15-65), BMI ≤32, absence of complete concentric palatal collapse on DISE, and documented CPAP failure. The European Position Paper on OSA surgery recommends DISE as a standard preoperative assessment tool and supports multi-level surgery for appropriately selected patients. The AAO-HNS position statement supports HNS as an effective treatment for CPAP-refractory moderate-to-severe OSA.

OSA surgery has evolved from the one-size-fits-all UPPP era to a phenotype-guided approach. DISE-based patient selection, hypoglossal nerve stimulation for appropriate candidates, multi-level surgery for complex obstruction patterns, and MMA for severe cases represent the current surgical armamentarium. HNS adoption is growing rapidly, with over 50,000 implants worldwide. BMI criteria are being expanded based on emerging data in patients with BMI 32-35. Research frontiers include bilateral HNS, titratable/adaptive stimulation algorithms, expansion to patients with concentric palatal collapse (previously excluded), and combination therapy approaches. The OSA phenotyping concept (PALM model — Pcrit, arousal threshold, loop gain, muscle responsiveness) is moving toward personalized treatment selection that matches therapy to the patient's dominant pathophysiological mechanism.