The need to be alert to complications of peri‐lead cerebral edema caused by deep brain stimulation implantation: A systematic literature review and meta‐analysis study

Abstract Background The compatibility of deep brain stimulation (DBS) hardware and MRI scans has greatly improved the diagnostic rate of postoperative peri‐lead edema (PLE). However, the etiology, incidence, and prognostic outcomes of this complication have not been established. Objective The incidence of PLE and associated symptoms, the process of occurrence and progression of this complication, as well as treatment strategies were evaluated. Methods We conducted a Preferred Reporting Items for Systematic Reviews and Meta‐Analyses compliant systematic review of all studies that reported on incidences of PLE and associated symptoms after DBS implantation. Through systematic literature review, we evaluated its causes, neuropsychiatric symptoms, duration, treatment methods, and prognostic outcomes. Results Our search retrieved 10 articles, including 5 articles on PLE and 10 articles on symptomatic PLE. The incidence of PLE was 35.8% (95% CI: 17.0%–54.6%), while the incidence of symptomatic PLE was 3.1% (95% CI: 1.5%–4.7%) accounting for 8.7% of PLE. Conclusions This complication is not as rare as previously reported. Therefore, it requires significant attention after DBS implantation. The correlation between its causes, duration, symptoms, and the area involved in edema should be assessed in long‐term prospective clinical studies with large sample sizes.


| INTRODUC TI ON
Deep brain stimulation (DBS) is a surgical procedure in which electrodes or leads that generate electrical impulses targeted towards specific brain regions are implanted at specific locations in the brain. A brain pacemaker initiates the electrical stimulation for the treatment of movement disorders or neuropsychiatric disorders. The DBS surgery is not associated with major structural damages to the brain. However, several complications can occur during clinical treatment. 1 The main complications include hardware-related complications (such as lead malposition or migration, component frac- Current studies do not support PLE as an infectious process and speculated to be related to (1) mechanical damage to the punctures lead or accumulation of cerebrospinal fluid (CSF) to electrode puncture needle tract: the blood-brain barrier (BBB) can be damaged by a needle-stab injury or the transplantation of electrodes, activation of the inflammatory cells can lead to the production and secretion of cytokines, and the cytokines further aggravate the inflammatory process and angiogenesis within hours of the insult. 2 (2) Venous infarction: neurological deficits that occur shortly after the opening of the dura with interruption of the cortical vein can lead to acute neurological changes with clinical symptoms similar to those of symptomatic PLE. 3 (3) Electrode tissue compatibility and immune factors or neurotoxicity of the implant: this was attributed to a fact that implant autopsy revealed a foreign body multinucleated giant celltype reaction present in all patients with DBS, and this reaction was present irrespective of the duration of implantation and may be a response to the polyurethane component of the electrodes' surface coat, 4 but the study showed that the DBS leads did not contain any neurotoxins. 5 (4) Short-term inflammatory irritation: lead may stimulate an inflammatory tissue response, including microglial activation and astrogliosis, which may participate in inflammatory PLE. 6 In addition, no studies have reported any relationship between gender, age, disease duration, score, comorbidities, surgical target, types of electrodes, number of implantations, whether microelectrode recording, or fibrin glue was used or not, and whether stimulation was turned on postoperatively. Therefore, prospective studies with large sample sizes are needed to clarify any relationship.
Taken together, these findings suggest that the etiology, susceptibility factors, and prognosis of this complication are still unclear. We conduct a systematic literature review and meta-analysis study based on different viewpoints of this complication in different medical centers with the aim of clarifying the occurrence and development of this complication and identifying appropriate treatment strategies, so as to raise the academic attention to this complication. unilateral, or bilateral DBS (Intervention), none (Comparison), and brain edema, or symptomatic brain edema (Outcome). All study designs were eligible for inclusion. However, experimental studies, letters, comments, and editorials were excluded. Cross-checking was done to avoid data duplication. All studies were evaluated for eligibility in the first screening phase based on a quick review of the title and the abstract. In the second screening phase, the full-text articles were reviewed to evaluate eligibility criteria.

| Dataextraction
All articles were reviewed by three independent investigators. The investigators consulted among themselves in cases where there were differing opinions or consulted other experts if an agreement could not be reached. Each study contained the following features: first author, publication year, study type, sample size, imaging modality, the incidence of PLE cases, the incidence of symptomatic PLE cases, time duration to PLE onset postoperatively, time duration to symptomatic PLE onset postoperatively, results, and conclusions.
Cross-checking was further carried out after data extraction.

| Qualityassessmentofincludedstudies
The methodological quality of all studies was determined by one author, while two other authors did the discussion and verification to achieve consensus and establish an agreement with overall rating scores. The quality of the included studies was assessed by three independent investigators using the strengthening the Methodological Index for Non-randomized Studies (MINORS) method whereby items were scored 0 if not reported, 1 if the reporting was inadequate, and 2 if the reporting was adequate. The global ideal score was 16 for noncomparative studies and 24 for comparative studies.
The final assessment result was decided after the discussion. The maximum score was 16 since none of the studies included was a comparative study.

| Evaluationofpublicationbias
The Peters' test was performed to evaluate the publication bias.
The Peters' test for the 10 studies was 0.763, indicating no significant publication bias. To further evaluate potential publication bias, we performed a sensitivity analysis. No obvious bias was revealed among the studies.

| Statisticalanalysis
Meta-analyses were conducted using R software 4.1.1 with Meta package 4.18-2 and the R package "Metaprop" for determining the proportion of PLE and symptomatic PLE. The results were presented as proportions at 95% confidence intervals (CIs). Forest plots were used to show the difference between the proportion of PLE and the symptomatic PLE after DBS operation. Heterogeneity among the studies was evaluated by I² and Cochrane's Q statistics. For I² < 40%, a fixed-effects approach was used. The between-study variance was determined by the random-effects method.

| Studyselection
The literature search yielded a total of 138 articles after removing duplicates. Eighty-two articles were excluded after the first screening phase, leaving 56 articles that underwent full-text review.
Furthermore, an additional 46 articles were excluded after the second screening phase. Therefore, 10 articles reporting on peri-lead cerebral edema were included in this study. The study selection process is represented in a PRISMA flowchart ( Figure 1). Quality Assessment of included studies is represented in Figure 2.

| Studycharacteristics
Characteristics of the articles included in the final literature review are presented in Table 1

| Mainfindings
In five studies, PLE was detected in 61 of 301 patients (incidence of 1.5%-4.7%). The incidence rate was reported as 30% in the retrospective studies and 50% in the prospective studies. However, in the 10 studies, symptomatic PLE was detected in 51 of 1365 patients (incidence of 0%-6.8%), with an incidence rate of 2.7% in the retrospective studies and 3.4% in the prospective studies. The duration of PLE ranged from one to 70 days, with an average of 31.6 days.
However, the symptomatic PLE ranged from 4.4 to 48.9 days, with an average of 11.6 days. The included articles did not report any symptoms of long-term neurological damage.
The imaging findings revealed that PLE mainly occurs as localized mechanical vasogenic cerebral edema, with a good prognosis and no damage to the brain parenchyma. Furthermore, PLE occurs as a delayed reaction with no cases reported on the first day after surgery. In addition, PLE can be located near the tip, in the subcortical region, or around the whole electrode. However, it is not reported which form is more common. Despite most DBS procedures being implanted with bilateral electrodes, PLE tends to be unilateral. The literature review did not reveal the laterality of edema occurrence and motor symptom laterality in patients with Parkinson's disease, but relevant studies show that the laterality of PLE is not related to the number of electrodes implanted, the order of bilateral implantation, or whether microelectrode recording is used or not.

| Meta-analysis
Ultimately, we analyzed the difference in the incidence of PLE and symptomatic PLE, according to the study methods and objectives of the included studies. For the incidence of PLE, five studies (three retrospective and two prospective case series) were included in the random-effects model of single-arm meta-analysis with a total of 275 patients. Overall, the pooled PLE incidence was 35.8% (95%CI: 17.0%-54.6%; Figure 3A). Heterogeneity between studies was assessed using the I 2 index and Cochrane's Q test. The I 2 and Q statistics were 93.7% (95%CI: 88.2%-96.6%) and 66.32(p < 0.0001). The studies showed moderate to high heterogeneity.  Abbreviations: d, day; m, month; w, week; "-" represents not needed, "\" represents not described in the literature.

| Causesandoccurrenceofpostoperativeperileadedema
The patients did not present with fever after surgery, had the blood routine and inflammatory factors, within the normal range and had negative fungal and bacterial cultures. This shows that the patients did not have any infection. Although infections are common after DBS implantation, tracking infection along the DBS lead in brain parenchyma is uncommon. In most cases, PLE is self-limiting, and the resolution without administration of any antibiotics suggests that PLE does not occur due to an infectious process. 9,10 Therefore, there is no sufficient evidence to support that an infection causes PLE.
Moreover, PLE may occur due to microhemorrhage or mechanical trauma to the brain tissue during the DBS lead implantation.
Most imaging scans during target verification after DBS implantation do not show evidence of hemorrhage. However, implantation of the microelectrodes or macroelectrodes may cause local microhemorrhage and mechanical trauma to the brain, leading to impaired BBB and increased permeability resulting in "luxury perfusion.'' 9 However, the "luxury perfusion" does not explain the confinement of the PLE around the electrode tip. In addition, PLE may be caused by the flow of CSF from the subarachnoid space into the brain parenchyma along the electrode puncture needle tract, thus accumulation after electrode implantation.
In literature, symptomatic PLE has also been attributed to venous infarction due to obstruction of the venous return. 11  Other possible causes of PLE include acute inflammatory reactions after DBS implantation. 6 Biran et al. 16 reported that the method of anchoring silicon microelectrode arrays to the skull significantly led to shearing and/or compression of the adjacent tissue. This could arise from a stiffness mismatch between the implant and brain tissue caused by the relative motion between the skull and the brain parenchyma. Most studies demonstrate spontaneous healing of PLE without any treatment. However, studies do not explain the unilaterality of PLE or confinement around the electrode tips. A prospective study conducted by Borellini et al. 17 reported 100% incidence of PLE in the 19 patients included. However, this study was excluded from our meta-analysis due to the small sample size and the occurrence of peri-electrode hemorrhage in six patients.
It is not known whether PLE is related to changes in the vascular structure due to chronic diseases such as hypertension and diabetes mellitus, and to spasm of small blood vessels in the brain due to implant irritation, causing exudation and edema. However, according to the study, this complication was not associated with the presence of vascular disease, hypertension, diabetes, or the use of anticoagulants/antiplatelet agents, but due to the small sample size and the fact that the MRI scan was performed only 6 weeks after surgery, no accurate conclusions can be drawn. 18 So, it remains to be investigated whether comorbidities are related to PLE and whether the related changes can be alleviated by drugs to improve microcirculation and electrode surface irrigation.

| Symptoms,course,prognosis,and treatmentofpostoperativeperi-leadedema
The meta-analysis and literature review showed that the incidence of PLE was 35.8%, while the incidence of symptomatic PLE was 3.1%, accounting for an overall incidence of 8.7% of the PLE cases.

| Reasonsfordifferentviews
This review reveals that PLE is not an uncommon complication. Asymptomatic PLE is self-limiting. Patients with clinical symptoms might be treated with dehydration and hormonal therapy to achieve better prognostic outcomes without neurological sequelae.

CO N FLI C TO FI NTE R E S T S
The authors have no conflicts of interest to declare that are relevant to the content of this article.

DATAAVA I L A B I L I T YS TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.