Real-World Study Tracks Discontinuation and Optimization in Parkinson’s Device Therapies

A prospective multicenter real-world investigation has tracked how often and why individuals with Parkinson’s disease discontinue advanced device-aided therapies, finding that nearly one-third of infusion treatments are stopped during routine clinical care. Analyzing prospective registry data, investigators observed an overall discontinuation proportion of 28.8 percent across all examined modalities, with 170 out of 591 initiated treatments discontinued. In contrast to infusion modalities, deep brain stimulation exhibited a substantially lower discontinuation frequency, highlighting distinct real-world utilization trajectories among different therapeutic approaches in advanced disease management.

The findings derive from the prospective multicenter DATs-PD GETM Spanish Registry, which systematically accrued clinical data between January 2024 and May 2026 to address the recognized scarcity of comparative real-world evidence regarding therapy persistence. The cohort comprised 565 patients who received a cumulative total of 591 device-aided interventions across five therapeutic categories. These modalities included surgical deep brain stimulation, continuous subcutaneous apomorphine infusion, foslevodopa/foscarbidopa, levodopa–carbidopa intestinal gel, and levodopa–entacapone–carbidopa intestinal gel. Researchers employed Kaplan–Meier survival estimates alongside Cox proportional hazards regression models to evaluate longitudinal persistence and identify demographic and clinical factors independently linked to treatment cessation.

Therapy cessation varied markedly between surgical intervention and continuous infusion systems across the registry. Deep brain stimulation demonstrated the lowest cessation rate in the cohort, with discontinuation occurring in 4.1 percent of instances, representing 4 cessations out of 97 interventions. Conversely, continuous subcutaneous apomorphine infusion recorded a cessation rate of 34.1 percent, accounting for 45 discontinuations among 132 treatments. Similarly, foslevodopa/foscarbidopa was discontinued in 34.7 percent of cases, representing 101 cessations out of 291 monitored therapies. Among intestinal infusion formulations, levodopa–carbidopa intestinal gel had a discontinuation rate of 28.6 percent, based on 4 cessations among 14 therapies, while levodopa–entacapone–carbidopa intestinal gel showed a discontinuation rate of 28.1 percent, reflecting 16 stops out of 57 treatments.

When investigators pooled all continuous infusion therapies together, the collective discontinuation rate stood at 33.6 percent, demonstrating comparable persistence across the different infusion approaches. Statistical evaluations confirmed that the observed differences in discontinuation rates across infusion types did not attain significance, yielding a p-value of 0.774. Furthermore, comparative Kaplan–Meier persistence analyses revealed no statistically significant divergence in retention timelines among the various infusion groups, returning a p-value of 0.701. Across the entire study population, the predominant driver of therapy termination was the emergence of adverse events, which accounted for 38.2 percent of all discontinuations, followed by autonomous patient decision, which prompted 25.3 percent of treatment cessations.

The registry also documented the duration required to achieve clinical dose optimization, recording an overall mean optimization timeframe of 61.6 ± 97.2 days across all examined therapies. Marked divergence characterized the optimization intervals required by different therapeutic mechanisms. Patients receiving levodopa–entacapone–carbidopa intestinal gel achieved regimen stabilization rapidly, registering a mean optimization duration of 19.6 ± 33.1 days. By comparison, individuals undergoing deep brain stimulation required a substantially more prolonged adjustment interval, with the surgical modality registering a mean optimization duration of 115.2 ± 118.3 days across the clinical cohort.

Multivariable Cox regression analyses identified several independent baseline characteristics associated with an increased hazard of therapy discontinuation. The living arrangement emerged as a prominent predictor, with patients residing in institutional caregiving settings exhibiting an elevated hazard ratio of 5.162, with a 95 percent confidence interval spanning from 1.441 to 18.490. Clinical disease severity also independently elevated discontinuation risk: each incremental increase in ON-state Hoehn and Yahr stage corresponded to an adjusted hazard ratio of 1.814, with a 95 percent confidence interval between 1.110 and 2.964. Additionally, daily motor complications contributed to treatment withdrawal, as each additional unit of daily OFF time yielded a hazard ratio of 1.093, with a 95 percent confidence interval ranging from 1.002 to 1.193.

The study authors concluded that while roughly one-third of patients receiving continuous infusion device-aided therapies discontinue treatment, persistence trajectories do not statistically differ between subcutaneous and intestinal infusion modalities. Greater disease severity, prolonged daily OFF periods, and institutional caregiving environments serve as significant risk factors associated with premature therapy discontinuation. Crucially, the researchers noted that inherent observational differences in clinical patient selection and cumulative therapy exposure preclude drawing definitive direct comparative-effectiveness conclusions across the various treatment modalities.

Subject of Research: Persistence, discontinuation factors, and optimization timing of device-aided therapies in Parkinson's disease

Article Title: Discontinuation and optimization of device-aided therapies in Parkinson’s disease: a prospective comparative real-world study

Article References: Santos García, D., Solleiro Vidal, Á., Blázquez Estrada, M., Mir, P., Alonso Modino, D., Legarda, I., López Ariztegui, N., García Ramos, R., Caballol, N., Ávila Rivera, A., Cabo, I., Hernández Vara, J., Sánchez Alonso, P., Ruiz Martínez, J., Sánchez Ferro, Á., DATs-PD GETM Spanish Registry Group, Adarmes Gómez, A. D., Aldaz, A., Alonso Modino, D., … Espinosa Rosso, R. (2026). Discontinuation and optimization of device-aided therapies in Parkinson’s disease: a prospective comparative real-world study. Journal of Neurology, 273(11), Article 703. https://doi.org/10.1007/s00415-026-14234-w

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14234-w

Keywords: Parkinson's disease, device-aided therapies, deep brain stimulation, continuous subcutaneous apomorphine infusion, foslevodopa/foscarbidopa, levodopa-carbidopa intestinal gel, treatment persistence, clinical neurology, Discontinuation, optimization, device-aided, therapies