Individual Studies
While all studies selected for the systematic review and metanalysis offered informative pre- and post-operative PSG data, the goals and designs of the studies themselves were not ubiquitously tailored to demonstrating DISE-directed tongue surgery efficacy. Some data sets required careful extraction of relevant data for the purposes of this review and meta-analysis.
Esteller et al conducted a prospective cohort study with 20 pediatric patients comparing pre- and post- operative AHI and snoring intensity visual analog scale (VAS) data. Only 7 of the 20 patients had DISE-guided tongue surgery, all of which were multi-level (i.e. radiofrequency lingual tonsillectomy plus revision adenoidectomy, pharyngoplasty, and/or revision tonsillectomy). Determined by DISE, tongue base obstruction was present in 40% of the patients. AHI improved with tongue surgery, LSAT was not reported (Table 1). The study reported significant decrease in the VAS score for snoring after (1.55 ± 1.701) compared to before (6.75 ± 2.337) DISE-directed surgery (p < 0.05), with 80% of the patients’ parents reporting VAS > 6 prior to surgery, but this data is confounded with upper airway surgery not involving the tongue.10
He et al’s retrospective study also assessed a variety of multi-level airway surgeries for persistent/recurrent OSA; including lingual tonsillectomy, midline posterior glossectomy, or hyoid myotomy/suspension. Similar to Esteller et al, the article does not directly link PSG data to the specific type of surgery performed so it is unclear how tongue surgery independently affected AHI. However, the study described overall outcomes by surgery: the responder data showed four (57%) patients who underwent lingual tonsillectomy were responders, and three (43%) were non-responders, as previously defined. For partial midline glossectomy, there was one (25%) responder and three (75%) non-responders. Hyoid suspension/myotomy included one responder (50%) and one non-responder (50%).
Ulualp et al conducted a small retrospective series with ten patients who underwent combined tongue base reduction and lingual tonsillectomy, half of which were syndromic, and the majority were overweight or obese. There was overall statistically significant improvement in AHI post-operatively, but 60% of the patients had residual OSA, and there was no statistically significant improvement in LSAT. It is unsurprising there was residual OSA in most patients given the high proportion of overweight-obese patients, and it is well-understood in the literature that this is a risk factor for persistent OSA.16
Williamson et al evaluated post-operative PSG outcomes of 168 children. Of these patients, 101 underwent lingual tonsillectomy alone; 25, midline posterior glossectomy alone; and 42, both. Concurrent procedures included turbinate reduction, revision adenoidectomy, and/or supraglottoplasty. Their mean AHI improvement was 3.52 ± 8.39, 2.55 ± 5.59, and 3.70 ± 6.07 events/hr, respectively. Each surgical group experienced significant improvement in sleep apnea when pre- and postoperative AHI was compared (P < 0.01). Overall, 126 (75%) patients experienced surgical success.14