Management Patterns in Deep Brain Stimulator Hardware Infection: Is Hardware Sparing Feasible?

INTRODUCTION: Deep brain stimulation (DBS) hardware infection is a serious complication with heterogeneous presentations and no standardized management guidelines. Management ranges from hardware-sparing (HS) washout to partial (PE) or complete explantation (CE). This study aimed to evaluate outcomes of initial management strategies and identify factors associated with treatment failure and escalation. METHODS: We conducted a single-center retrospective review of adult DBS patients (2002-2020) who required surgery for first-time and recurrent hardware infection or wound erosion. Treatment success was defined as infection clearance without further surgical intervention. Treatment failure was defined as escalation or conversion to a more extensive procedure among those who did not clear the infection.Outcomes were characterized using two complementary definitions - infection clearance success rate defined as absence of recurrent DBS-related infection or wound erosion after index management and non-escalation success rate, defined as avoidance to progression of a more advanced surgical strategy after index procedure. Recurrence freedom was evaluated using Kaplan-Meier analysis, and predictors of recurrence were assessed using multivariable Cox proportional hazards regression. RESULTS: Ninety-eight patients were included. Initial management consisted of HS in 39 patients, PE in 37, and CE in 22. Using the infection-clearance definition, success was observed in 38% of HS cases, 62% of PE cases, and 59% of CE cases. When success was defined by avoidance of escalation, HS achieved non-escalation success in 49% of cases and PE in 78%. Recurrence freedom differed significantly by management strategy, with PE and CE demonstrating greater durability than HS (log-rank χ² = 23.98, p < 0.001). In multivariable Cox regression, PE and CE were associated with significantly lower recurrence hazard compared with HS, while erosion at presentation independently predicted recurrence. Disease category was associated with conversion/escalation, with Parkinson's disease patients less likely to undergo escalation, but PD status did not predict recurrence freedom. Despite lower infection-control durability, HS preserved stimulation for a clinically meaningful interval in selected patients. CONCLUSION: DBS infection outcomes reflect the interaction between infection phenotype, management aggressiveness, and patient-centered goals of therapy. PE and CE provide greater recurrence freedom than HS, particularly when infection control is the primary objective. However, HS remains clinically relevant in selected patients because it may preserve therapeutic stimulation and delay or avoid escalation, even when definitive infection clearance is not achieved.