Predictors of vagus nerve stimulation complications among pediatric patients with drug-resistant epilepsy

View More View Less
  • 1 Department of Neurological Surgery, University of Pittsburgh;
  • | 2 Institute for Clinical Research Education, University of Pittsburgh; and
  • | 3 Department of Bioengineering, University of Pittsburgh, Pittsburgh, Pennsylvania
Free access

OBJECTIVE

Complications from vagus nerve stimulator (VNS) procedures are common and can have important implications for morbidity and seizure control, yet predictors of complications are poorly understood. The objective of this study was to assess clinical factors associated with minor and major complications from VNS procedures among pediatric patients with drug-resistant epilepsy.

METHODS

The authors performed an 11-year retrospective review of patients who underwent VNS procedures for drug-resistant epilepsy at age < 21 years. The primary outcome was complications (minor or major) following VNS surgery. Preoperative and surgery characteristics were compared between patients who developed versus those who did not develop complications. Multivariable Poisson regression was performed to determine the association between preoperative characteristics and infection.

RESULTS

Of 686 surgeries, 48 complications (7.0%) developed; there were 7 minor complications (1.0%) and 41 major complications (6.0%). Surgeries with minor complications were an average of 68 minutes longer than those without minor complications (p < 0.001). The incidence rate of infection was 1 per 100 person-years, with 3% of procedures complicated by infection. Poisson regression revealed that after adjusting for age at surgery, duration of surgery, and primarily motor seizure semiology, the incident rate of infection for revision surgeries preceded by ≥ 2 procedures was 19 times that of first-time revisions.

CONCLUSIONS

The overall minor complication rate was 1% and the overall major complication rate was 6% for VNS procedures. Longer surgery duration was associated with the development of minor complications but not major complications. Repeat incisions to the VNS pocket may be associated with higher incident rate of infection, highlighting a need for longer-lasting VNS pulse generator models.

ABBREVIATIONS

AE = adverse event; DBS = deep brain stimulation; MSSA = methicillin-sensitive Staphylococcus aureus; VNS = vagus nerve stimulator.

OBJECTIVE

Complications from vagus nerve stimulator (VNS) procedures are common and can have important implications for morbidity and seizure control, yet predictors of complications are poorly understood. The objective of this study was to assess clinical factors associated with minor and major complications from VNS procedures among pediatric patients with drug-resistant epilepsy.

METHODS

The authors performed an 11-year retrospective review of patients who underwent VNS procedures for drug-resistant epilepsy at age < 21 years. The primary outcome was complications (minor or major) following VNS surgery. Preoperative and surgery characteristics were compared between patients who developed versus those who did not develop complications. Multivariable Poisson regression was performed to determine the association between preoperative characteristics and infection.

RESULTS

Of 686 surgeries, 48 complications (7.0%) developed; there were 7 minor complications (1.0%) and 41 major complications (6.0%). Surgeries with minor complications were an average of 68 minutes longer than those without minor complications (p < 0.001). The incidence rate of infection was 1 per 100 person-years, with 3% of procedures complicated by infection. Poisson regression revealed that after adjusting for age at surgery, duration of surgery, and primarily motor seizure semiology, the incident rate of infection for revision surgeries preceded by ≥ 2 procedures was 19 times that of first-time revisions.

CONCLUSIONS

The overall minor complication rate was 1% and the overall major complication rate was 6% for VNS procedures. Longer surgery duration was associated with the development of minor complications but not major complications. Repeat incisions to the VNS pocket may be associated with higher incident rate of infection, highlighting a need for longer-lasting VNS pulse generator models.

In Brief

Researchers evaluated vagus nerve stimulator (VNS) complications in pediatric patients with drug-resistant epilepsy. Overall, 1% and 6% of VNS surgeries resulted in minor and major complications, respectively; 3% of procedures resulted in infection. The incident rate of infection for VNS revisions preceded by ≥ 2 VNS procedures was 19 times that of first-time revisions. VNS is a safe treatment option for pediatric patients with drug-resistant epilepsy, although this study highlights a need for transcutaneously chargeable or longer-lasting pulse generators.

Drug-resistant epilepsy, defined as the "failure of adequate trials of two tolerated, appropriately chosen and used antiepileptic drug schedules,"1 affects approximately 150,000 children in the United States.2 When resection of seizure foci is not an option, vagus nerve stimulation can be considered for epilepsy treatment.38 Major complications of vagus nerve stimulator (VNS) implantation, such as infection or hardware complications, are common911 and can have important implications for morbidity and seizure control, yet predictors of major complications are poorly understood.

Mild adverse events (AEs) associated with VNS activation, including hoarseness and coughing, occur in 62% of patients.12 Such AEs appear to be dose-dependent, as decreasing vagus nerve stimulation parameters often mitigates these symptoms.13 Rarely, vagus nerve stimulation parameter adjustment does not alleviate symptoms. Less is known about how to modify the risk of these non–stimulation-dependent minor complications or of major complications, such as VNS infection or hardware failure. Major complications occur in up to 17% of patients.11 To date, many studies have assessed the incidence of complications following vagus nerve stimulation.5,7,9,11,1317 Still, it is unclear whether specific clinical factors are associated with complications from vagus nerve stimulation. The primary objective of this analysis was to assess clinical factors associated with minor and major complications from vagus nerve stimulation among pediatric patients with drug-resistant epilepsy.

Methods

We performed a retrospective review of electronic medical records from the University of Pittsburgh Medical Center’s Children’s Hospital of Pittsburgh between January 1, 2009, and January 1, 2020. Patients were included if they underwent VNS implantation for drug-resistant epilepsy when they were 21 years of age or younger. This study was approved by the University of Pittsburgh institutional review board.

The primary outcome of this analysis was the number and type of complications following VNS surgery. Both initial VNS placement procedures and pulse generator revisions were considered. A subset of patients had undergone multiple VNS or pulse generator revision surgeries. Each procedure was considered a unique observation for the purposes of complication analysis.

Surgical complications were defined in concordance with the definition proposed by Sokol and Wilson as an undesirable and unintended result of an operation.18 Three authors (N.M., N.S., and L.V.) reviewed medical records to determine whether complications were present, feasibly attributed to the VNS, and occurred after the patient’s encounter VNS surgery but before any subsequent surgeries. If the same patient experienced complications after multiple independent procedures (i.e., a complication after initial VNS insertion and another complication after pulse generator revision), then each complication was attributed to the surgical procedure that directly preceded it.

This analysis distinguished between expected AEs of VNS procedures and true minor complications. Expected AEs were defined as VNS-associated symptoms (e.g., coughing, vocal quality changes, and neck pain) which were either tolerated by the patient or were alleviated by vagus nerve stimulation parameter adjustment. AEs were not considered to be minor complications for statistical analysis. If the VNS was required to be turned off or the patient opted to remove the device entirely to mitigate such VNS-associated symptoms, then this was considered a minor complication. The most common minor complications included coughing with VNS activation, gagging/vomiting with VNS activation, vocal changes (including hoarseness and vibratory quality of voice), and neck pain or discomfort. Major complications were also separately analyzed. Major complications included infection (of any part of the VNS or of the incision itself), fractured lead wires, permanent anatomical damage (e.g., internal jugular vein injury), and any other complication associated with the VNS requiring reoperation. Reoperations for complications themselves (e.g., irrigation and debridement for infections, lead wire revisions, and device explantation) were not considered surgical encounters for this analysis.

Preoperative characteristics, including patient sex, age at epilepsy onset, age at surgery, presence of a gastrostomy or tracheostomy tube at the time of surgery, and operative characteristics such as duration of surgery, were recorded. Duration of surgery was defined as the number of minutes from "surgery start" to "surgery end" (rather than the total duration of time in the operating room) as recorded in intraoperative records. For pulse generator revisions, number of prior VNS surgeries included any surgery that required reopening of any initial VNS incision: prior pulse generator revisions, prior lead wire revisions, and prior irrigation and debridements. All pulse generator revisions were presumed to be preceded by VNS implantation, so prior VNS implantation was not included in the "number of prior surgeries" count. In our data set, few surgical encounters were preceded by more than 2 prior VNS surgeries. For this reason, "number of prior surgeries" was collapsed into a trichotomous variable with values of 0, 1, or ≥ 2. Each patient’s predominant seizure type was also determined and classified by the International League Against Epilepsy’s 2017 guidelines.19 From these guidelines, patients were categorized as having either predominantly motor seizures (focal motor, generalized motor, focal to tonic-clonic, or unknown motor seizures) or nonmotor seizures (focal nonmotor, generalized nonmotor, or unknown nonmotor seizures).

Statistical Analysis

All statistical analyses were performed using Stata software for Mac, version 17.0 (StataCorp 2021. Stata Statistical Software: Release 17. StataCorp LLC). Basic demographics and epilepsy characteristics were compared between surgeries with versus those without any (minor or major) complications using chi-square tests or independent-samples t-tests. The data were subsequently analyzed in subgroups, comparing 1) surgeries with minor complications with those without minor complications and 2) surgeries with major complications with those without major complications. Fisher’s exact test was used when expected cell counts were < 5. Statistical testing was done with a type I error rate of 5%, and p values were two-sided.

Given that infection is the most common and among the most serious complications following VNS procedures, we performed multivariable Poisson regressions to determine the association between preoperative characteristics (duration of surgery, age at surgery, and motor seizure semiology) and infection among VNS insertions. This analysis was repeated for pulse generator revisions and included the same independent variables as the model for VNS insertions, with the addition of the variable "number of prior VNS surgeries." Elective surgeries for VNS insertions that were preceded by complete VNS explantation were not considered an initial VNS implantation and were excluded from statistical analysis. These procedures were still included in demographic analysis.

Results

Demographics

This analysis included a total of 365 patients, representing 686 VNS surgeries. All patients had VNS implanted on the left side of their neck to stimulate the left vagus nerve. Of the 686 surgeries, 48 (7.0%) developed either minor or major complications. There were 7 minor complications (1.0%) and 41 major complications (6.0%). Table 1 demonstrates differences between surgeries with versus without complications. Surgeries with and without complications exhibited similar operating room times (82 minutes vs 71 minutes, p = 0.06). Among pulse generator revisions, surgeries with complications were a mean of 22 minutes longer than those without complications (69 minutes vs 47 minutes, p = 0.03). However, there were no significant differences in the durations of new VNS implantation operation between surgeries with versus those without complications. When reanalyzed in subgroups by complication type, surgeries with minor complications (compared to those without minor complications) were a mean of 68 minutes longer (139 minutes vs 71 minutes, p < 0.001). Otherwise, there were no significant differences between surgeries with versus those without minor complications or between surgeries with versus without major complications.

TABLE 1.

Demographic characteristics of the study sample

Total Sample (n = 686)Surgeries w/o Complications (n = 637)Surgeries w/ Complications (n = 48)*
Female sex324 (47.2)301 (47.3)23 (47.9)
Etiology of epilepsy
 Genetic160 (23.6)147 (23.4)13 (27.1)
 Structural254 (37.5)240 (38.2)14 (29.2)
 Infectious5 (0.7)4 (0.6)1 (2.1)
 Metabolic3 (0.4)3 (0.5)0 (0)
 Unknown255 (37.7)234 (37.3)20 (41.7)
Overall duration of op, mins71.7 ± 40.971.0 ± 40.582.0 ± 44.3
 New VNS implantation84.2 ± 26.983.8 ± 27.388.0 ± 20.9
 Pulse generator revision48.5 ± 41.147.0 ± 38.769.1 ± 63.7
Year of VNS op2013 ± 4.22013 ± 4.22013 ± 3.3
Age at op, yrs11.8 ± 5.111.8 ± 5.111.7 ± 4.8

Categorical variables are reported as n (%) and continuous variables are reported as mean ± SD. All tests were either chi-square tests or independent-samples t-tests assuming equal variances unless otherwise noted. Nonparametric tests were used when appropriate. Boldface type indicates statistical significance.

There were 7 encounters (1.0%) with minor complications and 41 (6.0%) with major complications in this sample.

Etiology is missing for 9 procedures.

In this sample, there were expected AEs associated with VNS in 25 encounter surgeries (3.6%). Of these, vagus nerve stimulation parameter adjustment was required in 7 cases (28%) to mitigate symptoms. Symptoms were tolerated in the remaining 18 cases (72%), and, in most cases, the treatment team was eventually able to up-titrate stimulation parameters to reach goal stimulation.

Table 2 compares the incidence of specific minor and major complications between new VNS insertions and pulse generator revisions. There were 7 surgeries (1.0%) with minor complications in this cohort. Approximately 0.6% of surgeries (n = 4) were complicated by coughing with VNS activation, 0.2% (n = 1) by gagging or vomiting with VNS activation, 0.3% (n = 2) by hoarseness or vocal changes, and 0.3% (n = 2) by neck pain or discomfort. Two of these surgeries were complicated by multiple minor complications simultaneously: cough and neck pain (n = 1) and cough and vocal quality changes (n = 1). There were no statistically significant differences between the incidence of minor complications and procedure performed (new VNS vs pulse generator revision). There were 41 surgeries (6.0%) with major complications. Approximately 3% of surgeries in this analysis (n = 18) were complicated by infections of either the VNS or incision, 2.7% (n = 17) by fractured lead wires requiring surgical revision, 0.2% (n = 1) by permanent anatomical damage secondary to VNS surgery, and 1.0% (n = 6) by some other VNS-associated complication requiring reoperation. There were no significant differences in major complications between new VNS insertions and pulse generator revisions.

TABLE 2.

Postoperative minor and major complications from encounter VNS surgeries in this sample

No. of Procedures (%)*
New VNS (n = 366)PG Revision (n = 255)
Total minor complications6 (1.6) 3 (1.2)
 Coughing w/ VNS activation2 (0.5)2 (0.8)
 Gagging/vomiting w/ VNS activation1 (0.3)0 (0)
 Hoarseness or vocal changes1 (0.3)1 (0.4)
 Neck pain2 (0.5)0 (0)
Total major complications24 (6.6)§18 (7.1)
 Infection10 (2.7)8 (3.1)
 Fractured lead wire11 (3.0)6 (2.4)
 Permanent anatomical damage1 (0.3)0 (0)
 Other complication requiring reop2 (0.5)4 (1.6)

PG = pulse generator.

Fisher’s exact test was used when 25% of expected cell counts were < 5. No statistical comparisons were significant. Surgeries for lead wire revisions and VNS reimplantation following prior complete explantation are not included in this analysis.

Data are missing for 65 procedures.

One surgery was complicated by both coughing and neck pain.

One surgery was complicated by both coughing and vocal changes.

One surgery was complicated by both lead wire fracture and anatomical damage.

As mentioned above, one surgery was associated with injury and sacrifice of the internal jugular vein. In this case, a 7-year-old girl underwent lead revision for dysfunctional VNS leads. At the time of surgery, the lead coils were attached to the jugular vein. In the process of dissecting the leads off the vein with the otolaryngology team, the vein was lacerated and began bleeding. After an attempt at primary repair, the vein was sacrificed. The patient made a full recovery without any apparent lasting complications.

Infection was the most common major complication in our analysis. Table 3 delineates select clinical details among the encounters in our sample complicated by infection. The average age at encounter surgery was 12 years (range 2–19 years). Wound infections were most common (n = 11), and the most common causative organism was methicillin-sensitive Staphylococcus aureus (MSSA; n = 11). Gastrostomy or tracheostomy tubes were present at the time of 6 encounters (33%) complicated by infection (Table 3). Two patients required two independent surgical treatments for infection from a single battery replacement (these were considered as a single infection for statistical analysis). The first patient (encounter 144) developed a device infection 3 weeks after pulse generator replacement, which was treated with wound washout and device removal. The neck electrode was maintained in situ with the hope of reconnecting a new apparatus after the infection cleared. One month later, the patient developed wound dehiscence and the electrode became exposed. The other patient (encounter 547) similarly developed an infection of the pulse generator site 5 weeks after a pulse generator revision. He was treated with surgical irrigation, debridement, and removal of the pulse generator. The electrode was, again, kept in place. The patient subsequently developed a delayed infection (10 months later) and was treated with irrigation, debridement, and full device explantation. Neither patient had their VNS subsequently reimplanted.

TABLE 3.

Description of the 18 surgeries resulting in infections

Encounter No.Age (yrs) at Encounter OpGastrostomy/Tracheostomy TubeInfection LocationCultureInfection Antibiotic Regimen
86YesWoundMSSAVancomycin IV × 1 wk
10418NoWoundPseudomonasCiprofloxacin PO × 2 wks
1268NoWoundMSSACefadroxil × 2 wks
144*6YesDeviceMSSACefepime IV × 2 wks
27614NoWoundMSSACefazolin × 2 wks, then cefalexin × 2 wks
28816NoDeviceMSSATMP-SMX (duration unclear)
34115YesWoundMSSAOxacillin IV × 6 wks
3518NoDeviceMSSACefazolin IV × 2 wks
37914NoDeviceMSSACefadroxil × 2 wks
38213NoWoundStaphylococcus epidermidisTMP-SMX indefinitely
43819NoWoundMSSACephalexin PO × 3 wks
4682YesDeviceSerratia sp., light Proteus sp.Cefepime × 1 wk
4835NoWoundMRSACiprofloxacin PO × 2 wks
54016YesDeviceMRSAAmoxicillin-clavulanate PO × 1 wk + doxycycline PO × 2 wks
547*18NoWoundMRSAClindamycin IV × 3 wks
6296YesWoundMSSAOxacillin/cefepime × 6 wks
65019NoWoundMSSACefazolin × 1 wk
67213NoDeviceProteus, Serratia, & Morganella spp.Cefepime IV × 2 wks

IV = intravenously; MRSA = methicillin-resistant Staphylococcus aureus; PO = by mouth; TMP-SMX = trimethoprim-sulfamethoxazole.

Patients who required two independent surgical treatments for the same infection. These 2 cases were considered as a single infection for statistical analysis.

Table 4 details the 6 patients in this series who required reoperation following VNS surgery for a reason other than infection or lead wire fracture. Most commonly, the VNS was replaced because of high lead impedance without evidence of lead wire fracture on radiography (n = 2) or lead wire discontinuity/poor positioning (n = 2). One encounter required reoperation for a nonfunctioning VNS autostimulation feature. The most common reoperation was pulse generator replacement (n = 5).

TABLE 4.

Description of the 6 surgeries which required reoperation for VNS for reasons other than infection or lead wire fracture

Encounter No.Age (yrs) at Encounter SurgeryReason for ReopReop Procedure
2313High lead impedance, possible microfracturePulse generator replacement
17811Autostimulation feature was not working, which worsened patient’s seizure controlPulse generator replacement
2938Poorly seated lead wirePulse generator replacement & hardware repositioning
4557High lead impedance, no fracture observedPulse generator replacement
55210Lead wire discontinuityPulse generator replacement
64320Exposed VNS hardware, no evidence of infection (cultures negative)Hardware removal

Overall, the incidence rate of infection in this sample was 1 per 100 person-years (1%), with an overall 3% of procedures being complicated by infection. Table 5 displays the results of multivariable Poisson regressions to assess the association between various pre- or intraoperative factors and development of infection after initial VNS implantations or VNS pulse generator revisions. In both subgroups, duration of operation, age at surgery, and motor seizures were not associated with a higher incident rate ratio for infection. For pulse generator revisions, having one prior VNS surgery (not counting initial VNS implantation) was non-significantly associated with 2.5 times the incident rate of infection compared with no prior VNS surgeries. Having at least two prior VNS surgeries (not counting initial VNS implantation) was significantly associated with 19 times the incident rate of infection compared with no prior VNS surgeries (p = 0.04).

TABLE 5.

Multivariable Poisson regressions to determine the association between surgery/patient characteristics and infections following new VNS implantations and pulse generator revisions

IRRCI
New VNS insertions (n = 329 observations included)
Duration of op1.00(0.98–1.03)
Age at op1.06(0.91–1.24)
Motor seizures0.49(0.11–2.18)
Pulse generator revisions (n = 219 observations included)
Duration of op1.00(0.98–1.02)
No. of prior VNS surgeries*
 0Ref
 12.51(0.20–31.37)
 ≥219.1(1.12–326.68)
Age at op0.75(0.55–1.03)
Motor seizures 0.50(0.08–3.09)

IRR = incident rate ratio.

Significance was assessed at alpha = 0.05 level. Boldface type indicates statistical significance. Surgeries for lead wire revisions and VNS reimplantation following prior complete explantation are not included in this analysis.

Number of prior VNS surgeries includes only those other than an initial VNS implantation.

Seventeen surgeries in this analysis were complicated by lead wire fracture. The average time to lead wire fracture was 45 months (range 1–124 months). In this sample, 8 encounter surgeries were for reimplantation of a VNS that had previously been completely explanted. Table 6 depicts the characteristics of the respective 8 patients. Patient age at the time of reimplantation surgery ranged from 2 to 16 years. Seven of 8 VNS systems were removed because of confirmed infection, while 1 was removed for suspected infection. For this reason, these 8 patients were typically given a longer course of intravenous antibiotics before proceeding with VNS reimplantation. One patient (encounter 277 in Table 6) developed subsequent lead wire fracture 4 months after VNS reimplantation due to repetitive rotation of the pulse generator. The family, despite counseling to remove all hardware, opted for pulse generator removal alone. Two weeks after pulse generator removal, the patient developed an infection of the VNS pocket, which required irrigation and debridement. During this procedure, neck dissection was not attempted to remove the lead wires because they were not visible intraoperatively. Since then, this patient’s drug-resistant epilepsy has been managed medically, and no subsequent complications have occurred.

TABLE 6.

Description of the 8 patients in this analysis who underwent complete VNS explantation and subsequent reinsertion of VNS

Encounter No.Age (yrs) at Reimplantation OpPrior Non-VNS OpsPrior VNS-Related OpsReason for VNS RemovalSubsequent InfectionFollow-Up Duration (mos)
58412No1 VNS placement, 2 abscess drainagesInfectionNo118
56516No1 VNS placementSuspected infectionNo36
53212No1 VNS placement, 1 PG revisionInfectionNo59
3542No1 VNS placementInfectionNo32
35210Lt temporal lobe resection of epileptic focus1 VNS placement, 1 PG revisionInfectionNo76
27714No1 VNS placement, 1 debridementInfectionYes80
24415No1 VNS placementInfectionNo5
1289Anterior 2/3 corpus callosotomy1 VNS placementInfectionNo39

Discussion

In this analysis, we assessed factors associated with complications from VNS procedures among pediatric patients with drug-resistant epilepsy. Our analysis demonstrated that 1% of surgeries were associated with minor complications and 6% were associated with major complications. Furthermore, surgeries associated with minor complications were, on average, 68 minutes longer than those without minor complications. Interestingly, we found that duration of surgery was not associated with major complications. The average duration of VNS operation in our cohort was 71.7 minutes, which is consistent with the duration of VNS procedures reported in the literature.20

The mechanism of VNS action can partially explain its associated expected AEs. Importantly, 10% of the vagus nerve is efferent.21,22 It provides motor input to the larynx, pharynx, and viscera of the thorax and abdomen. For example, stimulation of the vagus nerve activates the recurrent laryngeal nerve, potentially causing contraction of antagonistic laryngeal muscles, resulting in vocal quality alterations.15 It is not surprising, then, that vocal alterations due to VNS activation are common, with rates estimated to be between 1%23 and 62%7 considering both pediatric and adult patients. Activation of the recurrent laryngeal nerve with VNS is so consistent that researchers like Vespa et al. have sought to use laryngeal motor evoked potentials to serve as a biomarker for VNS activation.24 As such, vocal quality changes are thought to be dose-dependent, since lowering vagus nerve stimulation parameters has been associated with symptom mitigation.13 In our analysis, we found that only 2.3% of surgical procedures were associated with vocal quality changes, and only 0.3% of vocal quality changes in this sample were not tolerated or alleviated by stimulation parameter adjustment. This rate is at the low end of the range reported in the literature, although Smyth et al. reported that only 1.4% of their patients (in a purely pediatric sample) experienced vocal quality changes.23 The pediatric population of VNS patients is unique; some are infants and some are nonverbal. Therefore, our results may still be a relatively accurate representation of the incidence of vocal quality changes among the population of pediatric patients with antiseizure medication–resistant epilepsy. Another potential explanation for expected AEs is prolonged vagus nerve manipulation during surgery, which may lead to stretching or bruising of nerve fibers.

Postoperative incision or device infections are among the most common major complications of VNS surgery, with estimated incidence rates of approximately 1% to 8%.10,13,2529 In this analysis, the incidence rate of infection was 1 per 100 person-years, with 3% of operations complicated by infection. Hasegawa et al. found that among 808 VNS-related procedures in adult patients at their institution, 12 were associated with infection.29 Similar to our results in pediatric patients, Hasegawa et al. found no difference in infection rates between VNS insertions and pulse generator revisions. Horowitz et al., in a retrospective analysis of 100 consecutive patients who underwent VNS implantation, demonstrated an infection rate of 6% by a mean of 24 months of follow-up.30 In 2017, Selner et al. performed an 11-year retrospective review of the American College of Surgeons National Surgical Quality Improvement Program database. The authors demonstrated an infection rate of 1.3% (1 of 77 adult patients) at 30 days postoperatively.20 However, both Horowitz et al. and Selner et al. assessed complications among initial VNS implantations only. Furthermore, these authors, and the majority of other studies reporting VNS complication rates, have included either only adult or both adult and pediatric patients in their analysis. As reported in the American Academy of Neurology’s evidence-based guideline update in 2013, the odds of infection among pediatric patients at the VNS implantation site were 3.4 times that of adult patients.8 In a sample of 74 pediatric patients (representing 85 surgical procedures) who underwent VNS for antiseizure medication–resistant epilepsy, Smyth et al. reported an infection rate of 8.1%.23 As mentioned above, our analysis, which included pulse generator revisions as well as initial VNS implantations among pediatric patients exclusively, revealed an overall infection rate of 3%. Two infections in our sample required two surgical washouts for infection control. In both cases, VNS hardware was only partially removed during the first washout with the hope of future pulse generator–lead wire reconnection. This may suggest that complete hardware removal is important for patients who develop VNS infections to prevent the need for subsequent surgical washouts.

Poisson regression analyses demonstrated that duration of surgery, age at surgery, and motor seizures were not statistically associated with infectious complications from surgery. These overall results remained true for both new VNS implantations and pulse generator revisions. However, our results demonstrated clinically meaningful associations between having had a greater number of prior incisions to the VNS insertion site and the incident rate of infection. Specifically, we found that after adjusting for duration of surgery, age at surgery, and primarily motor seizures, one prior VNS surgery (other than initial implantation) was non-significantly associated with 2.5 times the incident rate of infection, while having at least two prior VNS surgeries was significantly associated with 19 times the incident rate of infection. With that stated, it is important to note that while the effect sizes of these analyses are large, so too are the confidence intervals. Thus, while the incident rate of infection is likely truly higher among multiple repeat VNS surgeries, the exact magnitude of this effect is largely uncertain.

Given that the average life spans of a VNS pulse generator and VNS leads are about 4.9 and 5.2 years, respectively,31 many VNS patients will require revision surgeries after VNS implantation. If patients continue with vagus nerve stimulation therapy, they may require multiple revisions. Couch et al. found that 46% of their 644-patient sample required at least one VNS revision within 14 years of implantation.31 Interestingly, Révész et al. demonstrated that among pediatric patients, 7.3% of infections (n = 4) were associated with initial VNS implantations, while 0 infections were associated with pulse generator replacements.25 Still, these sample sizes are small. It is possible that the more times on which a given incision is reoperated, the more likely an infection is to develop. Indeed, in the deep brain stimulation (DBS) literature, authors have shown that the incidence of infection after implantable pulse generator replacement was greater than three times that of de novo DBS surgery (10% vs 3%).32 The underlying reason for the authors’ results is unclear, although researchers have speculated that fibrous scar tissue associated with revision surgeries may not support as robust an inflammatory response as native tissue and that antibiotics may be less able to penetrate fibrous tissue.32 Our results demonstrated no difference in the proportion of surgeries that developed infections between de novo implantations and pulse generator revisions. Even with our study’s overall large sample size, the subgroup sample sizes for this analysis (175 pulse generator revisions with no prior VNS procedures, of which only 3 developed infection; 57 pulse generator revisions with 1 prior VNS procedure, of which 2 developed infections; and 17 pulse generator revisions with ≥ 2 prior VNS procedures, of which 1 developed infection) may have limited the ability to detect a statistically significant difference.

A potentially important interpretation of our results is that VNS pulse generator revisions are associated with a progressively higher risk of infection as a function of the number of prior pulse generator revision surgeries. Interestingly, while other device companies have invested in transcutaneously rechargeable pulse generators, this is not yet available for children with VNS. Investing in this technology would make VNS not only more cost-effective,33 but also safer, as our results demonstrate.

Limitations

This study has key limitations that should be carefully considered. Primarily, the study design is retrospective in nature. Therefore, it is subject to the information and selection biases associated with all retrospective cohort analyses. Specifically, the rate of true minor complications, like coughing and hoarseness, identified in this analysis is low. While it is possible that the minor complication rate is truly 1%, it is also possible that minor complications were underrepresented. For example, some patients in our database were nonverbal at baseline, so assessment of their vocal quality changes was not possible. That said, our analysis was able to distinguish between AEs (which were transient or ameliorated with stimulation parameter adjustment) and true minor complications. With major complications, like infections, underreporting is less of a concern given their overt nature. Finally, our analysis can only reveal correlations and cannot establish causation. Specifically, our results revealed that longer operative duration was associated with minor complications, but longer surgery may be a proxy for more complex anatomy or dense scar tissue. Such limitations can be addressed with future prospective studies (with standardized data collection forms) assessing real-time incidence of complications from vagus nerve stimulation.

Conclusions

In this retrospective analysis of 686 VNS surgeries performed in 365 pediatric patients with drug-resistant epilepsy, longer surgery duration was associated with the development of minor complications but not major complications. The overall minor complication rate was 1%, and the overall major complication rate was 6%. The incidence of infection was 3%. VNS surgery is a safe treatment option for pediatric patients with drug-resistant epilepsy.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Muthiah, Abel. Acquisition of data: Muthiah, Sharma, Vodovotz. Analysis and interpretation of data: Muthiah, White. Drafting the article: Muthiah, Sharma. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Muthiah. Statistical analysis: Muthiah, White. Study supervision: Abel.

References

  • 1

    Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia. 2010;51(6):10691077.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2

    Zack MM, Kobau R. National and state estimates of the numbers of adults and children with active epilepsy—United States, 2015. MMWR Morb Mortal Wkly Rep. 2017;66(31):821825.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Tzadok M, Harush A, Nissenkorn A, Zauberman Y, Feldman Z, Ben-Zeev B. Clinical outcomes of closed-loop vagal nerve stimulation in patients with refractory epilepsy. Seizure. 2019;71:140144.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    González HFJ, Yengo-Kahn A, Englot DJ. Vagus nerve stimulation for the treatment of epilepsy. Neurosurg Clin N Am. 2019;30(2):219230.

  • 5

    Klinkenberg S, Aalbers MW, Vles JS, et al. Vagus nerve stimulation in children with intractable epilepsy: a randomized controlled trial. Dev Med Child Neurol. 2012;54(9):855861.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Englot DJ, Chang EF, Auguste KI. Efficacy of vagus nerve stimulation for epilepsy by patient age, epilepsy duration, and seizure type. Neurosurg Clin N Am. 2011;22(4):443448.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Handforth A, DeGiorgio CM, Schachter SC, et al. Vagus nerve stimulation therapy for partial-onset seizures: a randomized active-control trial. Neurology. 1998;51(1):4855.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Morris GL III, Gloss D, Buchhalter J, Mack KJ, Nickels K, Harden C. Evidence-based guideline update: vagus nerve stimulation for the treatment of epilepsy: report of the Guideline Development Subcommittee of the American Academy of Neurology. Neurology. 2013;81(16):14531459.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Fahy BG. Intraoperative and perioperative complications with a vagus nerve stimulation device. J Clin Anesth. 2010;22(3):213222.

  • 10

    Giordano F, Zicca A, Barba C, Guerrini R, Genitori L. Vagus nerve stimulation: surgical technique of implantation and revision and related morbidity. Epilepsia. 2017;58(suppl 1):8590.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Kahlow H, Olivecrona M. Complications of vagal nerve stimulation for drug-resistant epilepsy: a single center longitudinal study of 143 patients. Seizure. 2013;22(10):827833.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Englot DJ, Chang EF, Auguste KI. Vagus nerve stimulation for epilepsy: a meta-analysis of efficacy and predictors of response. J Neurosurg. 2011;115(6):12481255.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    DeGiorgio CM, Schachter SC, Handforth A, et al. Prospective long-term study of vagus nerve stimulation for the treatment of refractory seizures. Epilepsia. 2000;41(9):11951200.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Dandurand C, Champagne PO, Elayoubi K, Weil AG, Lespérence P, Bouthillier A. Vagus nerve stimulator-related speech/exercise induced cough. J Clin Neurosci. 2017;37:4748.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Kersing W, Dejonckere PH, van der Aa HE, Buschman HP. Laryngeal and vocal changes during vagus nerve stimulation in epileptic patients. J Voice. 2002;16(2):251257.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16

    Lundgren J, Ekberg O, Olsson R. Aspiration: a potential complication to vagus nerve stimulation. Epilepsia. 1998;39(9):9981000.

  • 17

    Mallereau CH, Ollivier I, Valenti-Hirsch MP, Hirsch E, Proust F, Chaussemy D. Vagus nerve stimulation in epilepsy: efficiency and safety of outpatient practice. Neurochirurgie. 2020;66(4):270274.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18

    Sokol DK, Wilson J. What is a surgical complication?. World J Surg. 2008;32(6):942944.

  • 19

    Fisher RS, Cross JH, D’Souza C, et al. Instruction manual for the ILAE 2017 operational classification of seizure types. Epilepsia. 2017;58(4):531542.

  • 20

    Selner AN, Rosinski CL, Chiu RG, et al. Vagal nerve stimulation for epilepsy in adults: a database risk analysis and review of the literature. World Neurosurg. 2019;121:e947e953.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21

    Clarke BM, Upton AR, Kamath M, Griffin HM. Electrostimulation effects of the vagus nerve on balance in epilepsy. Pacing Clin Electrophysiol. 1992;15(10 Pt 2):16141630.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 22

    Crumrine PK. Vagal nerve stimulation in children. Semin Pediatr Neurol. 2000;7(3):216223.

  • 23

    Smyth MD, Tubbs RS, Bebin EM, Grabb PA, Blount JP. Complications of chronic vagus nerve stimulation for epilepsy in children. J Neurosurg. 2003;99(3):500503.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 24

    Vespa S, Stumpp L, Bouckaert C, et al. Vagus nerve stimulation-induced laryngeal motor evoked potentials: a possible biomarker of effective nerve activation. Front Neurosci. 2019;13:880.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25

    Révész D, Rydenhag B, Ben-Menachem E. Complications and safety of vagus nerve stimulation: 25 years of experience at a single center. J Neurosurg Pediatr. 2016;18(1):97104.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26

    Ramsay RE, Uthman BM, Augustinsson LE, et al. Vagus nerve stimulation for treatment of partial seizures: 2. Safety, side effects, and tolerability. Epilepsia. 1994;35(3):627636.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27

    Patel NC, Edwards MS. Vagal nerve stimulator pocket infections. Pediatr Infect Dis J. 2004;23(7):681683.

  • 28

    Ben-Menachem E. Vagus nerve stimulation, side effects, and long-term safety. J Clin Neurophysiol. 2001;18(5):415418.

  • 29

    Hasegawa H, Van Gompel JJ, Marsh WR, et al. Outcomes following surgical management of vagus nerve stimulator-related infection: a retrospective multi-institutional study. J Neurosurg. 2021;135(3):783791.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30

    Horowitz G, Amit M, Fried I, et al. Vagal nerve stimulation for refractory epilepsy: the surgical procedure and complications in 100 implantations by a single medical center. Eur Arch Otorhinolaryngol. 2013;270(1):355358.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 31

    Couch JD, Gilman AM, Doyle WK. Long-term expectations of vagus nerve stimulation: a look at battery replacement and revision surgery. Neurosurgery. 2016;78(1):4246.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32

    Pepper J, Zrinzo L, Mirza B, Foltynie T, Limousin P, Hariz M. The risk of hardware infection in deep brain stimulation surgery is greater at impulse generator replacement than at the primary procedure. Stereotact Funct Neurosurg. 2013;91(1):5665.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33

    Abel TJ, Remick M, Welch WC, Smith KJ. One-year cost-effectiveness of callosotomy vs vagus nerve stimulation for drug-resistant seizures in Lennox-Gastaut Syndrome: a decision analytic model. Epilepsia Open. 2022;7(1):124130.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation

Illustration from Cinalli et al. (pp 330–341). Printed with permission from © CC Medical Arts.

  • 1

    Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia. 2010;51(6):10691077.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2

    Zack MM, Kobau R. National and state estimates of the numbers of adults and children with active epilepsy—United States, 2015. MMWR Morb Mortal Wkly Rep. 2017;66(31):821825.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Tzadok M, Harush A, Nissenkorn A, Zauberman Y, Feldman Z, Ben-Zeev B. Clinical outcomes of closed-loop vagal nerve stimulation in patients with refractory epilepsy. Seizure. 2019;71:140144.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    González HFJ, Yengo-Kahn A, Englot DJ. Vagus nerve stimulation for the treatment of epilepsy. Neurosurg Clin N Am. 2019;30(2):219230.

  • 5

    Klinkenberg S, Aalbers MW, Vles JS, et al. Vagus nerve stimulation in children with intractable epilepsy: a randomized controlled trial. Dev Med Child Neurol. 2012;54(9):855861.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Englot DJ, Chang EF, Auguste KI. Efficacy of vagus nerve stimulation for epilepsy by patient age, epilepsy duration, and seizure type. Neurosurg Clin N Am. 2011;22(4):443448.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Handforth A, DeGiorgio CM, Schachter SC, et al. Vagus nerve stimulation therapy for partial-onset seizures: a randomized active-control trial. Neurology. 1998;51(1):4855.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Morris GL III, Gloss D, Buchhalter J, Mack KJ, Nickels K, Harden C. Evidence-based guideline update: vagus nerve stimulation for the treatment of epilepsy: report of the Guideline Development Subcommittee of the American Academy of Neurology. Neurology. 2013;81(16):14531459.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Fahy BG. Intraoperative and perioperative complications with a vagus nerve stimulation device. J Clin Anesth. 2010;22(3):213222.

  • 10

    Giordano F, Zicca A, Barba C, Guerrini R, Genitori L. Vagus nerve stimulation: surgical technique of implantation and revision and related morbidity. Epilepsia. 2017;58(suppl 1):8590.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Kahlow H, Olivecrona M. Complications of vagal nerve stimulation for drug-resistant epilepsy: a single center longitudinal study of 143 patients. Seizure. 2013;22(10):827833.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Englot DJ, Chang EF, Auguste KI. Vagus nerve stimulation for epilepsy: a meta-analysis of efficacy and predictors of response. J Neurosurg. 2011;115(6):12481255.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    DeGiorgio CM, Schachter SC, Handforth A, et al. Prospective long-term study of vagus nerve stimulation for the treatment of refractory seizures. Epilepsia. 2000;41(9):11951200.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Dandurand C, Champagne PO, Elayoubi K, Weil AG, Lespérence P, Bouthillier A. Vagus nerve stimulator-related speech/exercise induced cough. J Clin Neurosci. 2017;37:4748.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Kersing W, Dejonckere PH, van der Aa HE, Buschman HP. Laryngeal and vocal changes during vagus nerve stimulation in epileptic patients. J Voice. 2002;16(2):251257.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16

    Lundgren J, Ekberg O, Olsson R. Aspiration: a potential complication to vagus nerve stimulation. Epilepsia. 1998;39(9):9981000.

  • 17

    Mallereau CH, Ollivier I, Valenti-Hirsch MP, Hirsch E, Proust F, Chaussemy D. Vagus nerve stimulation in epilepsy: efficiency and safety of outpatient practice. Neurochirurgie. 2020;66(4):270274.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18

    Sokol DK, Wilson J. What is a surgical complication?. World J Surg. 2008;32(6):942944.

  • 19

    Fisher RS, Cross JH, D’Souza C, et al. Instruction manual for the ILAE 2017 operational classification of seizure types. Epilepsia. 2017;58(4):531542.

  • 20

    Selner AN, Rosinski CL, Chiu RG, et al. Vagal nerve stimulation for epilepsy in adults: a database risk analysis and review of the literature. World Neurosurg. 2019;121:e947e953.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21

    Clarke BM, Upton AR, Kamath M, Griffin HM. Electrostimulation effects of the vagus nerve on balance in epilepsy. Pacing Clin Electrophysiol. 1992;15(10 Pt 2):16141630.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 22

    Crumrine PK. Vagal nerve stimulation in children. Semin Pediatr Neurol. 2000;7(3):216223.

  • 23

    Smyth MD, Tubbs RS, Bebin EM, Grabb PA, Blount JP. Complications of chronic vagus nerve stimulation for epilepsy in children. J Neurosurg. 2003;99(3):500503.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 24

    Vespa S, Stumpp L, Bouckaert C, et al. Vagus nerve stimulation-induced laryngeal motor evoked potentials: a possible biomarker of effective nerve activation. Front Neurosci. 2019;13:880.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25

    Révész D, Rydenhag B, Ben-Menachem E. Complications and safety of vagus nerve stimulation: 25 years of experience at a single center. J Neurosurg Pediatr. 2016;18(1):97104.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26

    Ramsay RE, Uthman BM, Augustinsson LE, et al. Vagus nerve stimulation for treatment of partial seizures: 2. Safety, side effects, and tolerability. Epilepsia. 1994;35(3):627636.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27

    Patel NC, Edwards MS. Vagal nerve stimulator pocket infections. Pediatr Infect Dis J. 2004;23(7):681683.

  • 28

    Ben-Menachem E. Vagus nerve stimulation, side effects, and long-term safety. J Clin Neurophysiol. 2001;18(5):415418.

  • 29

    Hasegawa H, Van Gompel JJ, Marsh WR, et al. Outcomes following surgical management of vagus nerve stimulator-related infection: a retrospective multi-institutional study. J Neurosurg. 2021;135(3):783791.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30

    Horowitz G, Amit M, Fried I, et al. Vagal nerve stimulation for refractory epilepsy: the surgical procedure and complications in 100 implantations by a single medical center. Eur Arch Otorhinolaryngol. 2013;270(1):355358.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 31

    Couch JD, Gilman AM, Doyle WK. Long-term expectations of vagus nerve stimulation: a look at battery replacement and revision surgery. Neurosurgery. 2016;78(1):4246.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32

    Pepper J, Zrinzo L, Mirza B, Foltynie T, Limousin P, Hariz M. The risk of hardware infection in deep brain stimulation surgery is greater at impulse generator replacement than at the primary procedure. Stereotact Funct Neurosurg. 2013;91(1):5665.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33

    Abel TJ, Remick M, Welch WC, Smith KJ. One-year cost-effectiveness of callosotomy vs vagus nerve stimulation for drug-resistant seizures in Lennox-Gastaut Syndrome: a decision analytic model. Epilepsia Open. 2022;7(1):124130.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 970 970 44
Full Text Views 460 460 388
PDF Downloads 244 244 140
EPUB Downloads 0 0 0