Synergistic clinical outcome assessments for cervical dystonia - The dystonias overall are a rare neurologic disorder. Cervical dystonia (CD), sometimes called
“spasmodic torticollis”, is one of the most common forms of dystonia. CD is characterized by the partial
loss of voluntary control of the neck musculature producing abnormal postures and/or movements of
the head in the form of head tremor. In addition to these overt motor abnormalities, the disorder is also
associated with non-motor symptoms including pain, fatigue, anxiety, and depression. Treatment
options for CD are suboptimal. Many oral medications have been tried but their efficacy is minimal and
limited by dose-dependent adverse side effects. Botulinum neurotoxin (BoNT) injections repeated every
3-4 months are the primary treatment of choice. Although BoNT is highly efficacious for many patients,
for a variety of reasons about 1/3 of patients discontinue BoNT treatment, and of those who continue
treatment about 1/3 are unsatisfied with the response.
Because treatment options are suboptimal, there is an active effort to find better strategies for
treating CD, as evidenced by dozens of active trials listed on ClinicalTrials.gov. However, the most
common clinical outcome assessment used to measure motor abnormalities – the Toronto Western
Spasmodic Torticollis Rating Scale (TWSTRS) – does not capture head tremor. Also, as with most
clinical rating scales, the TWSTRS is an intrinsically subjective assessment and therefore suffers from
inter-rater variability. This reduces our power to detect treatment effects in clinical trials.
Technology-based objective measures have the potential to circumvent this variability.
Advances in computer vision technology have enabled the measurement of head orientation/rotation
from 2-D images of the face in conventional video recordings. One of the long-term objectives of our
group is to leverage these advances to develop software that can capture and quantify motor
abnormalities across multiple types of focal dystonia. We are calling this system the Computational
Motor Objective Rater (CMOR). In this project specifically targeting CD, our aims are 1) to evaluate
CMOR’s convergent validity with patient reports of severity of abnormal head posture and head tremor
and 2) to determine CMOR’s sensitivity to changes in severity associated with interventions. To
accomplish these aims, we will conduct CMOR analyses of motor symptoms from video recordings of
100 CD patients enrolled in a separate Dystonia Coalition project to evaluate the variability of efficacy
of BoNT. That project will also acquire patient reports in the form of a patient centered outcome with
specific questions about the two motor features of CD and the patient’s global impression of change
(PGIC) in response to each BoNT treatment. Collectively the results will provide important information
about CMOR’s validity and a quantitative basis for sample size estimates for future clinical trials in CD.