Neuromodulation: Technology at the Neural Interface Received: July 29, 2019 Revised: September 29, 2019 Accepted: October 30, 2019 (onlinelibrary.wiley.com) DOI: 10.1111/ner.13081 Reliability of Intraoperative Testing During Deep Brain Stimulation Surgery Francesco Sammartino, MD ; Rahul Rege, BA; Vibhor Krishna, MD, SM Introduction: Deep brain stimulation (DBS) is an effective treatment for medically refractory Parkinson’s disease (PD). During DBS surgery, intraoperative testing is performed to confirm optimal lead placement by determining the stimulation thresholds for symptom improvement and side effects. However, the reliability of intraoperative testing in predicting distant postopera- tive thresholds is unknown. In this study, we hypothesized that intraoperative testing reliably estimates postoperative thresh- olds for both symptom improvement and side effects. Methods: We retrospectively analyzed a prospective database with intraoperative and postoperative thresholds for symptom improvement and side effects from a cohort of 66 PD patients who underwent STN DBS. We recorded the stimulation loca- tions relative to the mid-commissural point. Within-patient stimulation pairs were generated by clustering the intraoperative stimulation locations closest to the DBS contacts. We computed the distance between stimulation locations and atlas-based pyramidal tract (PT) and medial lemniscus (ML) masks. A leave-one-out cross-validation analysis was performed to determine the reliability of intraoperative testing in predicting postoperative thresholds while controlling for the distance from the rele- vant tracks. Results: Intraoperative testing reliably predicted (area under ROC >0.8) postoperative thresholds for tremor and rigidity improvements, as well as stimulation-induced motor contractions and paresthesias. The reliability was poor for improvement in bradykinesia. Conclusion: Intraoperative testing reliably predicts postoperative thresholds. These results are relevant during the informed consent process and patient counseling for DBS surgery. These will also guide the development of future methods for intraoperative feedback, especially during asleep DBS. Keywords: Deep brain, reliability, stimulation, stimulation, testing, thresholds Conflict of Interest: The authors do not have conflicts of interest. INTRODUCTION Deep brain stimulation (DBS) improves the motor symptoms of Parkinson’s disease (PD). During implantation surgery, microelec- trode recording (MER), and stimulation are performed to guide and refine the final location of the DBS lead (awake DBS) (1). In particular, macroelectrode stimulation (macrostimulation) is per- formed to determine the thresholds (minimum stimulation ampli- tude required to elicit a clinical effect) and calculate a therapeutic window (the difference between the threshold for side effect and symptom improvement). In general, optimally positioned macro- electrode induces motor benefits at lower thresholds and side effects at higher amplitudes (i.e., a wide therapeutic window). If the therapeutic window is suboptimal, the macroelectrode is then repositioned further away from the neural substrates presumed to be causing the side effects and eventually determine the final DBS location; however, there is no standard way to estimate this distance between actual and optimal macroelectrode location. While this process can be executed seamlessly by experienced teams, the reliability can be adversely affected by factors such as patient-related (e.g., level of patient cooperation, duration since last levodopa dose, high anxiety associated with awake brain sur- gery), procedure-related (e.g., intraoperative sedation, supine posi- tioning during tremor assessment with written testing), and clinician-related (e.g., experience and training). Besides, a lower aspect ratio and higher impedance of the macroelectrode can result in significantly larger current spread than the DBS electrode (2). Thus, the thresholds measured by macrostimulation can be different from those derived by DBS. Given these limitations, the overall utility of intraoperative testing needs to be evaluated in a “real world” scenario. While a comprehensive determination of the utility of intraoperative testing can be complicated, the ability to accurately predict postoperative thresholds can establish its reliability. If per- formed separately for each of the commonly observed stimulation-induced clinical improvements (i.e., reduction in tremor, bradykinesia, and rigidity) and side effects (i.e., persistent paresthesias and involuntary motor contractions), such analysis 1 Address correspondence to: Vibhor Krishna, MD, SM, Center for Neuromodulation, 480 Medical Center Dr., S1019, Columbus OH, 43210, USA. Email: vibhor.krishna@osumc.edu Department of Neurosurgery, The Ohio State University, Columbus, OH For more information on author guidelines, an explanation of our peer review process, and conflict of interest informed consent policies, please go to http:// www.wiley.com/WileyCDA/Section/id-301854.html Source(s) of financial support: This work was supported by the Neuroscience Research Institute at The Ohio State University and the OSU College of Medi- cine Roessler Research Scholarship. Neuromodulation 2019; ••: ••–•• © 2019 International Neuromodulation Society www.neuromodulationjournal.com