Tardive Dystonia Treatment & Management
The treatment of patients with tardive dystonia is difficult. Refractoriness to treatment is a substantial clinical concern. Several pharmacologic and other somatic interventions have been tried with variable results.
The first step after the diagnosis of tardive dystonia induced by neuroleptics or other drugs is to taper and then discontinue the causative drugs. Many times, a severe psychiatric illness makes this impossible, but carefully reconsidering the indications for dopamine antagonists in a given patient and considering alternate therapy are imperative. Unfortunately, it is not uncommon for the symptoms to worsen for a time after the offending medication is discontinued or reduced.
If the dystonia is focal and amenable to botulinum toxin therapy, this should be considered. [23, 24]
The primary pharmacological treatment for tardive dystonia is dopamine-depleting agents. Another option would be dopamine receptor blockers (ie, neuroleptics). [2, 3, 4] A common observation for all tardive syndromes is that the symptoms improve with an increase of dopamine blockade and worsen with a decrease. Thus, the goal is to add a medication that will provide dopamine blockade while minimizing the risk of worsening the tardive syndrome or creating new tardive syndromes. The treatments of choice are dopamine depleters such as tetrabenazine or reserpine, since these do not appear to cause tardive symptoms. However, their side effects can make these difficult to tolerate and they are not as effective at treating psychiatric illness as dopamine receptor blockers.
Another strategy, particularly for those with severe psychiatric illnesses or who are intolerant to the side effects of dopamine depleters, is to start a neuroleptic such as clozapine, which has minimal risk for creating tardive syndromes. Clozapine is frequently chosen because it has the least number of credible reports of inducing tardive symptoms and the most data for alleviating those symptoms. Other neuroleptics could also be considered; however, given that there are case reports of all neuroleptics causing tardive dystonia, other choices should be made with caution.
If starting a dopamine-depleting or a dopamine receptor blocking agent is only partially successful, treatment can be supplemented with alpha-methyl-para-tyrosine (AMPT) for additional dopamine blockade.
If dopamine blockade is not successful or only partially successful, anticholinergic medications such as artane can be tried. [2, 3, 4]
There are case reports of clonazepam being added to clozapine for additional benefit,
and this could be considered if a single medication is only partially successful. [25, 26]
There is also limited evidence of the effectiveness of electroconvulsive therapy in this condition, and this may be considered if pharmacological treatments have failed and there is reason not to consider deep brain stimulation (DBS). 
A comprehensive approach to patients with tardive dystonia includes patient education and supportive care. Physical therapy and well-fitted braces are designed primarily to improve posture and to prevent contractures. Although braces are tolerated poorly, particularly by children, they may be used in some cases as a substitute for sensory input. For example, in some patients with cervical dystonia, neck and head braces seem to provide sensory input by touching certain portions of the head or neck in a fashion similar to the patient's own sensory trick, thus enabling the patient to maintain a desirable head position. Some patients find various muscle relaxation techniques and sensory feedback therapy useful adjuncts to medical or surgical management.
Deep brain stimulation is probably the surgical treatment of choice at this time for those with severely disabling dystonia who have not responded to medical therapy.
Small studies by Trottenberg et al and Zhang et al initially showed some success in deep brain stimulation of the globus pallidus interna and bilateral subthalamic nuclei. [27, 28]
Thobois et al were able to show improvement by 63% in 5 patients with bilateral GPi stimulation compared to 8 controls . Gruber et al were able to show improvement in 9 patients by nearly 75% on the Burke-Fahn-Marsden Dystonia Rating Scale after 3-6 months. 
Gruber et al assessed the long-term effects, including motor function, quality of life, and mood, of bilateral globus pallidus internus DBS on patients with tardive dystonia and concluded it is a safe and effective long-term treatment. Patients were assessed 3 times using established movement disorder and neuropsychological scales. Results showed significant improvement in quality of life regarding physical components and affective state. They also noted that cognitive functions remained unchanged, and no permanent adverse effects occurred. 
Other treatments, such as thalamotomy and pallidotomy, have also been investigated, but as there is no evidence that these provide benefits beyond DBS (and certainly carry greater risk due to their irreversibility), these would be considered second-line.