TENT-A3 pivotal trial: sound-only vs bimodal stimulation (within-subject)

Record details

• Authors: Boedts M, et al.
• Journal / source: Nature Communications (2024)
• Design: Multi-site single-arm within-subject: 6 weeks sound-only, then 6 weeks bimodal
• Sample: 112 enrolled (105 analyzed)
• Primary endpoint: Responder rate (≥7-point THI improvement) during the bimodal stage vs the point estimate from the sound-only stage
• Randomized: No · Sham/placebo-controlled: No · Independent of the manufacturer: No

What this research found

The primary endpoint was NOT achieved in the full cohort: responder rate during the bimodal stage was 43.3% vs 63.3% during the preceding sound-only stage. Important: this is a sequential design — everyone got sound first, bimodal second, and the bimodal stage was measured from an already-improved baseline — so it is NOT evidence that sound beats bimodal, but neither does it show bimodal superiority. In the moderate-or-worse subgroup (THI≥38 at stage start — a threshold specified at a pre-planned interim analysis, not in the original protocol): bimodal 58.6% vs sound-only 43.2%, p=0.022.

96.8% of device-related adverse events were mild; no serious device-related AEs. This is the trial behind the FDA De Novo authorization — which rested on the severe-subgroup result. ITT analysis n=112 with imputation for dropouts.

Limitations and conflicts of interest

Single-arm sequential design (no parallel control); full-cohort primary endpoint missed; benefit confined to the more severe subgroup.

Funding / conflicts: Funded by Neuromod; co-author H.H. Lim has a financial interest in the manufacturer.

Primary sources

• PubMed record (39160146)
• Publisher (DOI 10.1038/s41467-024-50473-z)
• Source link

Used as evidence for

• Lenire (bimodal tongue stimulation)