What peer-reviewed clinical trials since 2023 show lasting remission of tinnitus symptoms?

1. The strongest peer‑reviewed signal: Lenire’s pivotal and follow‑on studies show symptom reduction, not cure

A multi‑site pivotal trial of bimodal neuromodulation using the Lenire device—published in Nature Communications—served as the controlled, single‑arm pivotal dataset supporting FDA De Novo clearance and documented clinically meaningful reductions in tinnitus severity when participants acted as their own controls (Stage 1 sound‑only then Stage 2 bimodal stimulation), with trial writeups and industry summaries reporting responder rates around 70.5% for clinically meaningful improvement in the trial cohorts ^{[1] [4]}. Subsequent real‑world and company‑linked reports claim high responder rates (91.5% in a retrospective chart review), but those real‑world data are either recent preprints or industry‑linked communications and are not evidence that tinnitus was eliminated or that remission is universal or permanent ^{[5] [6] [7] [8]}.

2. Targeted auditory‑somatosensory stimulation: randomized trial evidence of benefit for somatic tinnitus

A double‑blind randomized clinical trial led by Susan Shore and colleagues published in JAMA Network Open in 2023 tested targeted auditory‑somatosensory stimulation in 99 patients with somatic tinnitus and reported reductions in tinnitus measures using validated questionnaires (TFI, THI) over the study interval, indicating clinically relevant benefit for a defined subtype of tinnitus rather than proof of lasting, universal remission ^[2].

3. Small‑scale or open‑label implant approaches show promise but lack generalizable remission data

Experimental implant or repurposed cochlear‑stimulation approaches have produced striking individual improvements in tiny cohorts—e.g., a small study repurposing a cochlear implant system for long‑term stimulation reported improvements in all three treated patients over years of stimulation—but these are uncontrolled, tiny samples and do not establish a reproducible, causal therapy that produces lasting remission broadly ^[9].

4. Expert syntheses and consensus temper expectations: “no reliable elimination” since 2023

Multidisciplinary expert reviews and state‑of‑the‑field papers emphasize heterogeneity of tinnitus and conclude that, as of their publication, there is no treatment that reliably eliminates tinnitus or reduces loudness consistently across patients; case reports of remission exist but lack causal confirmation tied to specific interventions, underscoring that trial‑documented reductions in questionnaire scores are distinct from proven, lasting remission ^{[3] [10]}.

5. Conflicts of interest, regulatory context, and what “lasting remission” would require

The most prominent positive trials for device‑based approaches are linked to commercial developers (Neuromod/Lenire) and public summaries often highlight responder percentages and FDA De Novo clearance (which recognizes a device’s safety and probable benefit) rather than proof of permanent absence of tinnitus; regulatory approval and peer‑reviewed efficacy data demonstrate symptom reduction for many patients, but do not equate to documented, durable remission for the condition as a whole ^{[8] [1] [4]}. The literature also shows that some tinnitus improves spontaneously over years in a subset of patients, complicating attribution of remission to a therapy unless trials include long‑term randomized controls ^[10].

Bottom line

Peer‑reviewed randomized and controlled studies since 2023 report meaningful and sometimes durable symptom reductions for certain interventions—most notably Lenire’s bimodal neuromodulation trials (Nature Communications) and targeted auditory‑somatosensory stimulation (JAMA Network Open)—but none of the peer‑reviewed trials since 2023 provide robust, generalizable evidence that a therapy causes lasting, universal remission (total elimination) of tinnitus; expert reviews continue to state explicitly that no intervention reliably eliminates tinnitus across patients ^{[1] [2] [3]}.