Objective
To study retrospectively the effects of IT baclofen in treating 25 patients with severe segmental or generalized dystonia.

Setting
Neurological Institute, Columbia-Presbyterian Medical Center, New York, NY.

Patients
Twenty-five patients with severe segmental or generalized dystonia that was refractory to oral medications underwent IT baclofen test dosing. In addition to dystonia, 17 patients had spasticity or painful spasms. Thirteen of 25 patients responded to the test doses of IT baclofen, according to unblinded neurological assessments that included the patient's subjective report; all 13 underwent implantation of a pump for continuous IT baclofen infusion.

Results
In contrast to reports of patients with spasticity of spinal origin, those with dystonia in the present series had a lower response rate to bolus IT baclofen doses and a smaller degree of clinical improvement. For 10 of the 13 responders to the test doses of IT baclofen, dystonia rating scale scores of videotaped examinations by blinded observers detected no significant change (P<.07) in severity of dystonia. Retrospective data from 11 of 13 patients with implantable pumps, followed up for a mean interval of 21 months after pump insertion, showed continuing efficacy in 6 individuals (55%), based on a determination of patient satisfaction; however, only 3 patients (27%) reported a sustained improvement in functional capacity. Five (38%) of the 13 patients with implantable pumps experienced severe complications that required hospitalization.

Conclusions
Despite recent reports that have described the benefit in small numbers of patients with dystonia, we concluded that the role of IT baclofen in treating severe dystonia remains uncertain. Intrathecal baclofen may be more effective when dystonia is associated with spasticity or pain. In the present series, we detected no significant difference in the response to IT baclofen in patients with or without spasticity or pain, perhaps owing to the small sample size.