Article Text

Download PDFPDF
I40 Number needed to harm (NNH) analysis of deutetrabenazine (DTB) and tetrabenazine (TBZ) from two pivotal trials: First-HD and Tetra-HD
  1. Daniel O Claassen1,
  2. Lisa De Boer2,
  3. Ravi Iyer3,
  4. Rajeev Ayyagari4,
  5. Fan Mu4,
  6. Sanjay Gandhi3,
  7. Benjamin Carroll3,
  8. David Stamler2
  1. 1Vanderbilt University Medical Centre, Nashville, Tennessee, USA
  2. 2Teva Pharmaceutical Industries, La Jolla, California, USA
  3. 3Teva Pharmaceutical Industries, Frazer, Pennsylvania, USA
  4. 4Analysis Group, Inc., Boston, Massachusetts, USA


Background DTB was efficacious and well tolerated in patients with Huntington’s disease (HD) chorea (First-HD). In the absence of a head-to-head trial with TBZ, we compared safety and tolerability outcomes of TBZ versus DTB using an indirect treatment comparison method and an NNH metric.

Methods Safety data from First-HD and TETRA-HD were used to calculate the NNH, where serious adverse events (SAEs), discontinuation (all cause- and adverse event [AE]-related), and specific AEs (in > 10% of patients) were defined as undesirable outcomes. The difference between the proportions of undesirable outcomes was estimated by subtracting the applicable placebo-adjusted risk from both studies (TETRA-HD minus First-HD). NNH estimates, defined as the number of patients to be treated with TBZ instead of DTB for an average of one additional patient to experience an AE, were calculated as the inverses of the risk differences. P-values were obtained from z-tests, which approximate normal distribution.

Results First-HD (N = 90) and TETRA-HD (N = 84) cohorts were of similar age (53.7 years vs 49.2 years) and gender (proportion of females 44% vs 62%), and baseline chorea scores were 12.7 vs 14.9. TBZ demonstrated a significant NNH vs DTB of 3 (95% CI: 1, 8; P < 0.01) for moderate to severe AEs, meaning that, if three patients were treated with TBZ instead of DTB, then one more patient, on average, would experience a moderate to severe AE. TBZ had a significant (all P ≤ 0.02) NNH vs DTB of 5 (95% CI: 3, 15) for depression, 5 (95% CI: 3, 15) for akathisia, 7 (95% CI: 4, 19) for Parkinsonism, 5 (95% CI: 3, 27) for somnolence, 4 (95% CI: 3, 12) for insomnia, and 8 (95% CI: 4, 46) for agitation. No other AEs analysed were significant.

Conclusions In this NNH analysis, DTB demonstrated significantly better outcomes than TBZ on several safety measures. Further analysis adjusting for demographic differences between trials should be conducted to confirm these findings.

  • Deutetrabenazine
  • Tetrabenazine
  • Chorea

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.