Teva study shows HCPs favour AUSTEDO attributes for older TD patients
Teva Pharmaceuticals has released findings from a discrete choice experiment (DCE) examining how healthcare providers (HCPs) weigh treatment attributes when selecting a vesicular monoamine transporter 2 (VMAT2) inhibitor for tardive dyskinesia (TD) in patients aged 55 and over. The data, presented at Psych Congress in New Orleans on 18 September 2026, found that AUSTEDO (deutetrabenazine) was the predicted choice across all four patient profiles modelled in the study.
The DCE enrolled 489 HCPs from a range of specialties. Participants ranked somnolence risk as their most heavily weighted consideration, accounting for between 26.8% and 36.2% of decision-making weight depending on patient profile. Short-term improvement on the Abnormal Involuntary Movement Scale came second at 24.4% to 29.5%, followed by dose formulation and drug-drug interaction risk. The study found that HCPs adjusted their priorities based on care setting: somnolence was weighted more heavily for community outpatients than for patients in long-term care facilities.
AUSTEDO's predicted preference share
Applying the elicited preferences to the available VMAT2 inhibitor options, AUSTEDO returned the highest predicted choice probability across all four patient scenarios modelled. Teva attributed this outcome primarily to the drug's somnolence profile, the duration of long-term response data available for the agent, and its drug-drug interaction characteristics. The company positioned these attributes as directly aligned with the tolerability and evidence-durability criteria HCPs placed at the top of their decision hierarchy.
Eric Hughes, Executive Vice President, Global R&D and Chief Medical Officer at Teva, said the findings reflect the company's commitment to "delivering clinical solutions for TD communities that help eliminate challenges to treat." The quote is largely promotional, but the underlying framing is consistent with the study design: a preference study commissioned by Teva will naturally surface attributes on which its own product performs well, and the release does not include head-to-head efficacy or safety data against named comparators.
Companion registry data on untreated TD
Alongside the DCE, Teva also presented an interim analysis from the IMPACT-TD Registry, which tracked individuals with probable TD who had not received any VMAT2 inhibitor therapy over 24 months. Between 89% and 96% of participants experienced at least a mild global impact of TD across the observation window, with 60% to 74% reporting moderate-to-severe impact. Most participants (55% to 69%) showed stable or worsening severity relative to baseline, with no evidence of spontaneous remission. Depression and anxiety scores fluctuated without sustained improvement over the two years.
The registry data adds a real-world burden angle to what is otherwise a market-research study. Taken together, the two datasets are designed to reinforce the case for early and sustained treatment rather than a watch-and-wait approach.
Competitive and regulatory context
The VMAT2 inhibitor space for TD is currently served by deutetrabenazine (AUSTEDO, Teva) and valbenazine (Ingrezza, Neurocrine Biosciences), both approved by the US Food and Drug Administration. A third agent, tetrabenazine, is approved for Huntington's chorea but carries a more demanding dosing schedule and broader drug interaction profile, which likely explains why it did not surface favourably in HCP preference modelling. The two primary competitors have differentiated mainly on dosing convenience and tolerability data, making preference studies of this kind an increasingly common commercial tool in the space.
For Teva, which has flagged AUSTEDO as a key growth asset as it pivots away from its generics heritage, sustaining HCP preference data ahead of any patent cliff pressures will be commercially important. The company did not disclose funding arrangements for the DCE beyond its own authorship, and independent replication of the preference weighting would strengthen the findings. Clinicians and payers evaluating the data should note its sponsored provenance.