Delayed Intracerebral Hemorrhage after Deep Brain Stimulation for Parkinson’s Disease Open Access

Parkinson’s disease (PD) is the second most prevalent neurodegenerative disorder globally and represents a significant public health challenge [1, 2]. The disease follows a progressive course, with hallmark symptoms such as bradykinesia, tremor, and rigidity initially managed effectively with pharmacotherapy. However, alternative treatment strategies must be considered the therapeutic window narrows [3]. In the mid-1980s, interventions for PD focused on ablative procedures [4, 5]. A breakthrough came with the development of deep brain stimulation (DBS) of the subthalamic nucleus (STN), offering an adjustable alternative [6]. DBS has since become a well-established treatment for advanced PD, providing significant symptomatic relief for patients suffering from severe motor fluctuations and dyskinesia [7, 8].

Large-scale studies provide comprehensive insights into DBS complications, highlighting an overall adverse event rate that aligns with previous meta-analyses. Their findings indicate that infection and intracranial hemorrhage remain among the most prevalent perioperative risks, emphasizing the need for stringent patient selection and postoperative management [9]. Additionally, criteria for DBS candidacy, as outlined in recent trials, underscore the benefits of early intervention in appropriately selected patients. Evidence suggests that earlier DBS implantation can lead to improved long-term outcomes, particularly in motor function and quality of life, reinforcing prior findings from studies such as the EARLYSTIM trial [10].

While DBS is relatively safe, it does carry risks of perioperative complications such as infections and hemorrhage, as well as postoperative issues like confusion, delirium, and cognitive decline [11, 12]. Recent advancements, including closed-loop DBS systems that dynamically adjust stimulation in response to real-time neural activity [13], and directional leads that offer more targeted modulation of pathological circuits [14], represent significant steps toward enhancing both the efficacy and safety of neuromodulation. These innovations promise to reduce side effects and optimize therapeutic outcomes by offering more individualized and adaptable stimulation strategies.

Postoperative symptomatic intracerebral hemorrhage (ICH) is a neurosurgeon’s worst nightmare, especially in DBS procedures. As a cornerstone of functional neurosurgery aimed at enhancing quality of life rather than addressing life-threatening conditions, such complications can be particularly devastating.

The incidence of postoperative symptomatic intracerebral hemorrhage reported in the literature ranges from 0.2% to 5.6%, with risk varying by target and surgical approach [15, 16]. This complication may be associated with factors such as hypertension, the number of lead passes, the use of microelectrode recording (MER), and whether the procedure was performed under general anesthesia or with the patient awake [17, 18]. STN-DBS has an ICH rate of 2.5%, whereas GPi-DBS presents a higher rate at 6.7% [19]. Regarding surgical trajectory, transventricular approaches have been associated with a slightly higher hemorrhage rate (3.7%) compared to transcerebral approaches (2.7%), though modifications such as using a guiding cannula may mitigate this risk [20]. Furthermore, 49.6% of ICH cases are symptomatic, often leading to neurological deficits, while asymptomatic cases are frequently detected on imaging with minimal clinical impact [16]. These distinctions help refine the risk assessment and clinical management of ICH in DBS patients.

This article presents a rare case of delayed ICH after STN DBS, emphasizing the possibility of late-onset complications. We also propose strategies to mitigate these risks, underscoring the importance of vigilant postoperative monitoring and individualized management plans to prevent such adverse outcomes.

We present a case of a 67-year-old male diagnosed with PD in 2016, initially presenting with a left-hand tremor. His symptoms rapidly progressed, extending to his left leg and right arm. While dopaminergic therapy with levodopa initially provided relief, he eventually required additional treatments. Unfortunately, he experienced adverse effects, including leg edema with pramipexole and hallucinations with amantadine. Despite increasing his levodopa dosage, he developed motor fluctuations and dyskinesia. With a total MDS-UPDRS (ON) score of 39 (I 0, II 12, III 24, IV 3) and deteriorating quality of life, he was considered a suitable candidate for DBS as he had minimal non-motor symptoms, an NMSQ score of 1/30, and normal neuropsychological testing. His previous medical history also included diabetes mellitus (DM) type 2 and surgery for colon carcinoma, which was without recidivism in the last 5 years.

The patient underwent bilateral STN DBS implantation using standard stereotactic coordinates for the STN target. MER was not utilized in this case. Intraoperative impedance measures confirmed appropriate lead placement, with values within the expected range. The procedure was performed under general anesthesia, with the patient fully asleep throughout the surgery. The anesthetic protocol included a balanced regimen, ensuring hemodynamic stability, and no dexmedetomidine or remifentanil was used. Intraoperative blood pressure monitoring showed stable systolic pressure, with values not exceeding 130/80 mm Hg. After the first stage of surgery, approximately 30 min after the lead implantation, an immediate perioperative frame-based head CT scan was performed to confirm precise lead positioning and assess for acute complications, such as hemorrhage or pneumocephalus (shown in Fig. 1a). CT was chosen for initial imaging due to its rapid acquisition time and superior sensitivity in detecting acute hemorrhages compared to MRI. The second stage involved implanting an extension and a pectoral pulse generator. On the first postoperative day, the patient exhibited no neurological deficits and showed a micro-lesion effect of STN DBS. At the same time, a routine MRI was conducted using T2-weighted and susceptibility-weighted imaging sequences to evaluate for microhemorrhages and revealed the hemorrhage around the left lead (shown in Fig. 1b). The potential role of intraoperative imaging modalities, such as MRI or O-arm 2D/3D imaging system, in reducing ICH risk has been explored in recent literature. While intraoperative MRI offers excellent soft tissue contrast, its routine use in DBS remains limited due to logistical constraints and prolonged operative time [21]. O-arm, an alternative intraoperative imaging method, provides real-time verification of lead placement and potential early detection of hemorrhage, though it was not employed in this case [22].

On the second postoperative day, the patient experienced sudden neurological deterioration, presenting with moderate right-sided hemiparesis. A CT scan confirmed ICH alongside the left lead, with minor perifocal edema (shown in Fig. 1c). During hospitalization, the patient underwent daily physical therapy, showing significant improvement. Initially, the left lead was not stimulated due to the hemorrhage, while the right lead was, resulting in significant contralateral improvement of PD symptoms.

After 2 weeks, the right-sided hemiparesis improved to a mild condition, and control CT scans showed complete regression of the ICH, with no perifocal edema or other structural damage (shown in Fig. 1d). To objectively assess hemorrhage evolution, we performed a quantitative volumetric analysis using pixel-based segmentation, providing arbitrary area units as a relative measure of the size of hemorrhage. The hemorrhage showed an ∼80% reduction by 2 weeks, correlating with the patient’s clinical recovery. While arbitrary area unit does not provide absolute volumetric measurements, it reliably demonstrates progressive resorption of the hemorrhage over time. The patient was transferred to our referring neurological department for further monitoring and stimulation optimization. At the 1-month follow-up, the patient had fully recovered, showing no signs of lateralizing neurological deficits. At that time, stimulation of the left lead was initiated, further improving the patient’s Parkinson’s symptoms. Six months after the implantation, his total UPDRS score was reduced to 19 (I 0, II 5, III 13, IV 1), with no visible consequence of the hemorrhage.

ICH is a well-known but relatively infrequent complication of DBS, with an incidence of approximately 2.5% per patient and 1.4–1.6% per implanted lead, a significant portion of which is asymptomatic [16, 23‒25]. Although recent literature suggests that most ICH cases are detected within 24 h postoperatively [25], it is important to report and discuss delayed hemorrhages, particularly in patients with risk factors, to determine the appropriate timing for neuroimaging control [23, 25‒27]. Despite its low incidence, ICH can have serious clinical and neurological consequences, with about half of the cases being symptomatic and potentially causing deficits while accounting for a small percentage of the overall low mortality rate associated with DBS [20, 23, 24, 28]. Several risk factors for ICH after DBS include the use of MER, hypertension, older age, cardiovascular comorbidities, and transventricular surgical trajectories [20, 23‒25, 28]. The use of MER during DBS has been linked to a higher rate of ICH [25], though some studies report no significant difference in bleeding rates associated with its use [16, 28]. Hypertension and older age, along with cardiovascular comorbidities, are substantial risk factors for ICH during DBS surgery, with affected patients typically being an average of 5 years older, likely due to increased vascular fragility and higher comorbidity rates [16, 20, 23, 28]. Diabetes, particularly type 1, may be a risk factor for ICH as it has been associated with an increased likelihood of hemorrhage, though the exact incidence remains uncertain [29]. In addition, chronic hyperglycemia contributes to endothelial dysfunction and cerebral microangiopathy, significantly increasing the risk of cerebrovascular complications, including ICH [30]. Persistent hyperglycemia induces oxidative stress, inflammatory responses, and blood-brain barrier (BBB) disruption, leading to increased vascular permeability and a higher likelihood of hemorrhagic events [31]. Additionally, genetic predisposition may play a role as studies suggest that the APOE ε4 allele is associated with increased risk of ICH, likely due to its effects on cerebral vessel integrity and impaired response to vascular insults [32]. Furthermore, transventricular approaches may slightly increase the risk of hemorrhage compared to transcerebral approaches, although both can be performed safely with appropriate precautions [20]. Surgeon experience also plays a critical role, with a weak inverse correlation observed between the rate of ICH per lead and the number of leads implanted per year by a single surgeon, indicating that increased expertise may slightly reduce the risk of hemorrhage [24]. To further enhance clarity, we have included a summary table of key studies reporting ICH incidence, risk factors, and outcomes in DBS surgery. These additions aim to provide clinicians with a concise reference for understanding and managing ICH in DBS patients (Table 1).

A comparative analysis of ICH risk between STN-DBS and GPi-DBS remains an important consideration. While some studies suggest that STN-DBS is associated with a slightly higher hemorrhage rate due to its deeper and more vascular location, recent meta-analyses have found no significant difference in ICH risk between the two targets [35]. Additionally, lead design and implantation techniques have gained attention as potential modifiers of hemorrhage risk. Segmented electrodes offer improved current steering and may reduce unintended tissue activation, potentially lowering procedural risks [36]. Similarly, the choice between stylet-assisted and stylet-free implantation techniques has been examined, with some studies suggesting that stylet-free methods may reduce brain tissue microtrauma [37]. Further research is warranted to establish the impact of these technical variables on ICH risk in DBS patients. Due to similarities in surgical approach, it is beneficial to compare the bleeding risks associated with DBS and external ventricular drainage. Both procedures involve parenchymal penetration via burr hole access, and while their indications differ, the mechanical risks share certain features. Reported rates of symptomatic hemorrhage following DBS range from 0.5% to 2.5%, whereas in external ventricular drainage, the rate can reach up to 5%, particularly in emergent or traumatic settings. Key risk factors common to both include hypertension, coagulopathy, and multiple insertion attempts. Although DBS targets deeper structures and is more often elective, understanding these parallels may aid in refining surgical strategies and improving perioperative risk assessment [38, 39].

Regarding the mechanisms of delayed hemorrhage, extravasation of blood along the electrode tract due to incomplete fibrin seal formation could potentially contribute to delayed ICH. More commonly, delayed hemorrhage arises from the rupture of small vessels caused by microtrauma or postoperative increases in intracranial pressure. Additionally, cerebral autoregulatory dysfunction in PD patients may play a role, particularly in the context of dopamine agonist withdrawal and orthostatic hypotension. Sudden perioperative blood pressure fluctuations can impair cerebral perfusion, leading to hypoperfusion-reperfusion injury and increasing the likelihood of vessel rupture in the microtraumatized area [40]. This risk is further elevated by sudden spikes in systemic blood pressure during early mobilization, particularly in patients who experience autonomic dysregulation due to the withdrawal of antiparkinsonian medications preoperatively, leading to significant blood pressure fluctuations [41].

Moreover, microtrauma from lead insertion may contribute to delayed hemorrhage through BBB disruption. DBS lead implantation, even when performed with precision, induces localized tissue injury that can compromise the integrity of the BBB. This disruption may facilitate delayed extravasation of blood components, triggering an inflammatory response and subsequent edema, which can exacerbate hemorrhage formation (Fig. 2). Studies suggest that BBB breakdown post-DBS may persist for several days, further supporting the need for vigilant monitoring and blood pressure control in the early postoperative period [25, 42].

In our case, hemorrhage resolved over approximately 2 weeks, likely facilitated by CSF pulsations aiding in blood product clearance [43, 44] or macrophagic activity in the periventricular white matter contributing to hemorrhage/hematoma breakdown and absorption (Fig. 2) [45, 46]. These observations underline the unpredictable nature of such complications and emphasize the need for vigilant postoperative monitoring, follow-up imaging, and the implementation of preventive strategies to ensure patient safety. Preoperative evaluation should address patient-specific risk factors, including blood pressure control and coagulation status, and use CT and MRI to plan safe lead trajectories. Intraoperative measures, such as immediate response to bleeding, brain shift prevention, and fibrin glue application, are essential [38]. Immediate intraoperative CT ensures proper lead placement and absence of hemorrhaging, while postoperative MRI within 1 to 2 days confirms lead accuracy and detects delayed ICH [23]. In addition, strict postoperative blood pressure control and careful monitoring are crucial to minimize complications.

ICH is a serious DBS complication that requires understanding its risk factors, mechanisms, and management to reduce its occurrence. Symptomatic cases are often managed conservatively with a tailored, multidisciplinary approach, while continuous follow-up is crucial for recovery, complication management, and optimizing treatment plans. This case highlights the importance of diligent patient follow-up as certain complications can arise later or in a delayed matter and may lead to permanent neurological deficits. To improve outcomes and mitigate risks in high-risk DBS patients, a stepwise clinical algorithm can be employed. Preoperative optimization should include strict blood pressure control, cessation of antiplatelet agents when appropriate, and careful selection of lead trajectories to minimize vascular injury. Intraoperative strategies should focus on real-time imaging to confirm lead placement and rule out immediate hemorrhagic complications, as well as maintaining hemodynamic stability throughout the procedure. Postoperatively, serial imaging and neurological assessments are critical to detect and manage potential delayed complications. The primary aim of functional neurosurgery is to enhance patients’ quality of life, and preventing surgical complications is essential to preserving the therapeutic benefits of DBS.

Furthermore, to improve the understanding of ICH risk factors and outcomes in DBS, prospective registries collecting granular data on patient characteristics, surgical variables, and postoperative complications are needed. Such datasets would enable more accurate risk stratification and facilitate the development of predictive models for hemorrhagic complications. Additionally, biomarkers such as S100B, GFAP, and neurofilament light chain show potential for early detection of hemorrhagic events, with S100B and GFAP aiding in intracranial injury assessment and neurofilament light chain indicating neuroaxonal damage after DBS implantation. Future research should explore the feasibility of incorporating these biomarkers into clinical workflows to enhance the early detection and management of ICH in DBS patients.

The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000546056).

The authors respectfully thank the patient.

Written informed consent was obtained from the patient for the publication of this case report and accompanying images. Ethical approval was not required for this study, following local guidelines.

The authors declare no conflicts of interest.

This research received no external funding.

Conceptualization: H.C., M.R., and P.M.; writing – original draft, H.C., M.R., and D.C.; and writing – review and editing, V.R., E.P., M.H., and V.V. All authors have read and agreed to the published version of the manuscript.

Hana Chudy and Marina Raguž contributed equally to this work.

The patient’s data are protected according to the GDRP regulations of Croatia and the EU. All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.