Imaging Predictors of Cryolysis Efficacy for Treatment of Obstructive Sleep Apnea

Previous research has shown there is increased fat at the base of the tongue and other locations of the oropharynx in patients with Obstructive Sleep Apnea (OSA) compared to match control non-OSA patients. Subsequently, Cryosa developed the Cryosa Procedure, which is further detailed in the ARCTIC-3 protocol (IRB #854182). This procedure involves the Cryosa System, a device intended to induce adipose cryolysis, a non-surgical removal of by inducing cell death (apoptosis) with a controlled freezing of the soft tissue, in the upper airway.

Whereas the objective of the ARCTIC-3 study is to determine the efficacy and safety of the Cryosa Procedure, the co-primary aims of the ICE-OSA study are to (1) evaluate predictors of successful treatment with the Cryosa Procedure and (2) evaluate a potential mechanism of action of this novel therapy. The study team hypothesizes that (1) higher baseline quantity of oropharyngeal fat and higher baseline upper airway neurotonic activity are correlated with a successful reduction in OSA severity, and (2) responders will have a decrease in quantity of oropharyngeal fat (as measured by post-operation magnetic resonance imaging, MRI). Investigating the mechanisms and predictors of this novel therapy is necessary to inform future clinical trials and patient selection for the Cryosa Procedure. To assess these metrics, the ICE-OSA study utilizes MRI, point-of-care ultrasound (POCUS), and the addition of pharyngeal manometry and ultrasound (US) during drug-induced sleep endoscopy (DISE) for ARCTIC-3 participants.

Study Statistics:

These prognostic data are going to be collected as part of a small pilot study to inform the upcoming pivotal trial. For this reason, investigators seek independent variables with large effect sizes to update and enhance patient selection criteria for the future trial.

The study team propose use of Student's t-test for responders/non-responders (responder defined as reduction of AHI by \>50%) for the surgical intervention. The expected ratio of responders to non-responders is 1:1 based on data provided by the study sponsor (unpublished).

Previous data examining tongue fat in MRI demonstrated values of quantity of fat tissue in fatty tongues to be roughly 14,000 mm3 and fat tissue in non-fatty tongues to be roughly 7,000 mm3. Using these parameters, it was calculated that 10 subjects total would be required (assuming a standard deviation of 4,000 mm3) to detect significant differences between responders/non-responders with an alpha of 0.05 and a power of 80%.