gms | German Medical Science

Objective: The efficacy of deep brain stimulation (DBS) has remained similar since its introduction. Thus, its impact on postoperative quality of life is largely determined by adverse events (AE).

Method: Medical records (1289 documents) of 123 consecutive patients (female 56; median age 63 yrs) who had undergone DBS surgery between January 2007 and June 2011 were analyzed retrospectively. Indications were Parkinson's disease (81), dystonia (22), different types of tremor (18), Huntington's disease (1), and Tourette's syndrome (1). Electrodes were implanted into the subthalamic nucleus (65), ventrolateral thalamus (23), and internal globus pallidus (17). Mean and median follow-up was 4.7 yrs (578 patient-years). AE were graded according to the guidelines of CTCAE. Serious adverse events (SAE) according to FDA criteria were identified.

Results: A total of 442 AE were recorded in 105 patients. SAE (103) occurred in 59 patients. All AE and SAE that were related to surgery, infection, or the implanted devices were reversible and resolved completely. In two patients, routine CT scanning revealed small hemorrhages (1.6%) in the striatum along implanted electrodes. In four patients (3.3%; 0.7% per patient-years) IPG removal was required because of infection (3 patients replanted). Postoperative wound complications or secondary healing (3.3%) was treated by local revisions without sequelae. One defective electrode was replaced one week after implantation. There was no lead migration or surgical revision because of lead malplacement. Overall, neurological (46.1%) and psychiatric (17.5%) AE were most common. Most were associated with the underlying disease, i.e. gait disturbance (53.7%), dysarthria (42.3%), depression (14.6%), cognitive decline (11.4%), and confusion (8.1%). There was one suicide (GPI stimulation). Most AE (51.5%) were reversible. Non reversible AE (23.1%) included dysarthria, gait disturbance, tumor disease and other. Reversibility was unknown in 25.4%. There has been no reduction of AE over time (i.e. no 'learning curve'). Age was not correlated with the occurrence of AE.

Conclusions: We investigated a real world patient cohort treated by DBS in common surgical targets. Neurological and psychiatric AE were common and usually related to the underlying disease. Surgery and device-related AE are relatively rare and appear to represent residual risks.