Responsive neurostimulation (RNS) is a promising treatment for patients for whom resective surgery is not an option, but less is known about the indications and safety of RNS in pediatric patients.1–9 In RNS surgery, a pulse generator is implanted into the cranium via a small craniectomy, and strip or depth electrodes are implanted at the stimulation target in the brain, often the epileptogenic zone.2,3 Implantation of these devices is thought to be associated with an acceptable rate of complications in adults, and emerging evidence also suggests safety in children.1–9 However, given that existing series in children have low numbers and that younger patients have a thinner skull diameter, which may increase the risk for device-related complications, more data are necessary to understand the relative risk in children.6–9 Additionally, children have different etiological and developmental considerations that may lead to RNS being used for different indications from what is observed in adults. In this article, we review our case series of pediatric RNS at University of Pittsburgh Medical Center Children’s Hospital of Pittsburgh to contribute to evidence of the safety and indications of RNS in the pediatric population.
Methods
We present a case series of 20 pediatric patients ranging in age from 8 to 21 years (mean age 15 [SD 4] years), who underwent RNS implantation of depth electrodes (n = 15), strip electrodes (n = 2), or both (n = 3), performed at Children’s Hospital of Pittsburgh between 2020 and 2022. Institutional review board approval was obtained as part of the global protocol for retrospective data collection in epilepsy patients (STUDY20070281). Indications for RNS placement were extracted from the electronic medical record. Data assessing RNS safety and effectiveness in reducing seizure frequency were collected. Considered RNS complications included infection, stroke, hemorrhage, and device malfunction. Engel and International League Against Epilepsy (ILAE) classifications at the latest follow-up visit were used to define postoperative seizure responsiveness to RNS. As an additional measure of effectiveness, antiepileptic drug (AED) prescriptions were evaluated pre- and post-RNS surgery (at latest follow-up) as a means of assessing RNS-gained seizure freedom.
Results
Patient Demographics and Clinical Characteristics
Of our 20-patient cohort, 8 patients were female (40%) (Table 1). In general, the indications for RNS were 1) multifocal epilepsy (e.g., bitemporal epilepsy) for which resection would not be possible, 2) epilepsy in which the epileptogenic zone involved eloquent cortex (e.g., motor cortex epilepsy), or 3) generalized epilepsy when resection was not possible and a decision was made for centromedian thalamic nucleus RNS. See Table 1 for descriptions of primary seizure semiology and epileptogenic zone locations. The primary seizure semiology varied across patients, including 3 patients (15%) with focal to bilateral tonic-clonic (FBTC) only, 9 patients (45%) with focal motor onset seizures with or without FBTC, 3 patients with (15%) focal nonmotor onset, and 5 patients with (25%) generalized onset motor or nonmotor seizures.
Patient characteristics
| Patient No. | Sex | Age at RNS (yrs) | Seizure Semiology | Prior Op | Primary Target | Implantation Strategy | Hospitalization (days) | Decrease in AEDs?* | FU (mos) | FU Engel Class | FU ILAE Class |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | M | 16 | Generalized onset motor & nonmotor | SEEG; rt MTL biopsy; MTL LITT | CMTN | Depth electrodes | 1 | Yes | 20 | IIB | 3 |
| 2 | M | 18 | Focal motor onset | Resection of rt frontal FCD; SEEG | Rt SMA & primary motor cortex | Depth electrodes | 2 | Mixed | 22 | IIIB | 5 |
| 3 | F | 13 | FIAS motor onset & FBTC | Oligodendroglioma resection; resection of SOZ; VNS | Lt primary motor cortex | Depth electrodes | 1 | No | 22 | IIIA | 4 |
| 4 | F | 15 | FIAS nonmotor onset | SEEG; LITT ablation of rt amygdala | Bihippocampal | Depth electrodes | 1 | No | 17 | IVA | 5 |
| 5 | M | 10 | FBTC | SEEG | Bihippocampal | Depth electrodes | 1 | No | 17 | IIIA | 4 |
| 6 | M | 9 | Generalized onset nonmotor | None | CMTN | Depth electrodes | 1 | No | 13 | IIIA | 5 |
| 7 | F | 11 | Focal motor onset | SEEG; LITT of rt frontal cortex | Rt SMA | Depth electrodes | 2 | Mixed | 15 | IIIA | 4 |
| 8 | M | 16 | FIAS motor onset | VNS; SEEG | Bihippocampal | Depth electrodes | 1 | No | 15 | IVA | 5 |
| 9 | M | 21 | FBTC | SEEG | Rt midfrontal & temporal gyri | Strip electrodes | 2 | No | 8 | IIIA | 5 |
| 10 | M | 12 | FIAS motor onset | SEEG | Bilat superior & midtemporal gyri | Depth electrodes | 1 | No | 6 | IVB | 5 |
| 11 | M | 19 | Generalized onset nonmotor & motor | SEEG | Rt hippocampus & rt frontal neocortex | Combination | 1 | Yes | 10 | IVC | 5 |
| 12 | M | 15 | Focal motor onset | SEEG × 2; lt temporal lobectomy; amygdalohippocampectomy | Lt motor cortex & lt superior parietal cortex | Depth electrodes | 1 | No | 7 | IIB | 4 |
| 13 | F | 15 | FBTC | SEEG; lt frontal resection; lt SMA resection | Posterior to lt SMA | Strip electrodes | 4 | No | 6 | IVB | 5 |
| 14 | F | 18 | FBTC | SEEG; rt FCD resection | Rt motor cortex | Depth electrodes | 1 | IFU | 0 (moved out of state) | IFU | IFU |
| 15 | F | 8 | FIAS motor onset | VNS + partial CC; completion CC | CMTN | Depth electrodes | 10 | No | 6 | IVB | 5 |
| 16 | M | 18 | FIAS motor onset & FBTC | SEEG × 2; VNS | Rt insula, rt temporoparietal | Combination | 1 | No | 4 | IIIA | 5 |
| 17 | M | 16 | FIAS nonmotor onset & FBTC | SEEG | Lt motor cortex & lt parietal cortex | Depth electrodes | 1 | No | 2 | IFU | IFU |
| 18 | F | 19 | Generalized onset motor & nonmotor | None | CMTN | Depth electrodes | 4 | No | 1 | IFU | IFU |
| 19 | M | 17 | Focal nonmotor onset & FBTC | LITT MTL; SEEG | Rt occipital cortex | Combination | 3 | No | 4 | IFU | IFU |
| 20 | F | 20 | Generalized onset motor & nonmotor | VNS; anterior CC; completion CC | CMTN | Depth electrodes | 1 | No | 9 | IVA | 5 |
CC = corpus callosotomy; CMTN = centromedian thalamic nucleus; FCD = focal cortical dysplasia; FIAS = focal impaired awareness seizures; FU = follow-up; IFU = insufficient FU; LITT = laser interstitial thermal therapy; MTL = mesial temporal lobe; SOZ = seizure onset zone; VNS = vagus nerve stimulation.
"Mixed" refers to a decrease in one medication and an increase in another medication.
As for the primary target of RNS (Table 1), 5 patients (25%) underwent electrode implantation to bilateral centromedian thalamic nuclei, 7 patients (35%) had implantation to the motor cortex, 4 patients (20%) had implantation to the mesial temporal lobe (3 bilateral and 1 right hippocampal head only), and 4 patients (20%) had implantation to cortical seizure foci. The implantation strategy was predominantly placement of depth electrodes, with 15 patients (75%) undergoing depth electrode implantation, 2 (10%) undergoing strip electrode implantation, and 3 (15%) with a combination of strip and depth electrodes. In the case of patient 13, stereo-electroencephalography (SEEG) exploration revealed the epileptogenic zone to be within the left supplementary motor area (SMA), immediately anterior to the primary motor cortex. The patient continued to have typical seizure semiology after initial resection, and additional resection and implantation of RNS electrodes on the primary motor cortex was offered to the family for postoperative seizure monitoring and therapeutic motor control. In general, depth electrode placement was preferred because it enabled stimulation of either specific cortical targets identified by SEEG or subcortical targets in cases of centromedian thalamic nucleus RNS. Strip electrodes were used when a gyral surface target or regional stimulation across a depth and strip electrode was preferred.
Complications
In our patient cohort, no instances of hemorrhage, stroke, or RNS device malfunction were observed. One patient experienced a postoperative infection (5%). Standard infection precautions and prophylaxis were employed, including preoperative Hibiclens shampoo and a 5-day course of Keflex (Pragma) postoperatively. Patient 4, a 15-year-old young woman with type 1 diabetes mellitus, returned to the clinic 1 month postoperatively with concerns of superficial infection at the inferior portion of her incision. She was initially sent home with oral antibiotics but was admitted to the hospital 1 week later for observation and intravenous antibiotic treatment after no improvement had been made. A small amount of exposed hardware was apparent, and the patient was brought to the operating room for wound debridement and reclosure. Cultures were positive for methicillin-susceptible Staphylococcus aureus and Enterobacter, and she was treated successfully with a course of Augmentin (USAntibiotics) and levofloxacin. Finally, pseudomeningocele (5%) developed in 1 case (patient 6), which resolved by postoperative day 6 without intervention.
Postoperative weakness was experienced by 4 patients (20%) in our cohort. One day postoperatively, patient 1 demonstrated a slight strength deficit in his left hamstring muscle and dynamic strength deficit in his left quadriceps femoris muscle in the stance phase of gait following bilateral centromedian nucleus implantation, which resolved spontaneously prior to follow-up. Patient 3 reported postoperative right leg weakness following left primary motor RNS depth electrode placement, but this resolved overnight and without intervention. Patient 13 experienced right facial and right upper- and lower-extremity weakness following extension of a prior resection of her left SMA and RNS strip electrode placement posterior to resection, considered to be an SMA syndrome, and was discharged to inpatient rehabilitation for 15 days. At the 5-month follow-up, weakness had improved and she was experiencing only mild right-handed weakness with fine motor deficit. Patient 18 experienced left-handed weakness following bilateral centromedian nucleus RNS, which has continued to improve with follow-up. Patient 2 reported 3 days of tremor and numbness of the left second, third, and fourth digits, which resolved spontaneously. While we have not observed transient weakness with our motor cortex SEEG implantations, it is clearly a risk with RNS implantation to motor cortex based on this series. One potential explanation is that patients have a lower threshold for developing weakness given that the implanted cortex is epileptogenic, but this is speculative. Given that transient hemiparesis is a risk of electrode implantations in motor cortex, risk of transient or permanent hemiparesis is important to discuss with families preoperatively.
Postoperative Response
We evaluated postoperative response using two scales, the Engel Epilepsy Surgery Outcome Scale and the ILAE classification scale, in patients with at least 3 months of follow-up (n = 16; Table 1). The mean follow-up time was 11 months (SD 7 months; range 4–22 months). At the most recent follow-up, 2 patients (13%), 6 patients (38%), 1 patient (6%), 3 patients (19%), 3 patients (19%), and 1 patient (6%) had Engel class IIB, IIIA, IIIB, IVA, IVB, and IVC outcomes, respectively.
Using the ILAE classification scale, 1 patient (6%) was classified as ILAE class 3, meaning that he only experienced 1 to 3 seizure days per year. Four patients (25%) were classified as ILAE class 4, indicating that these patients saw 4 seizure days per year to 50% reduction of baseline seizure days. Eleven patients (69%) were classified as ILAE class 5, implying that these patients reported less than 50% reduction in baseline seizure days. Furthermore, AED treatment was not changed in 15 patients (75%). However, 2 patients (10%) decreased the dose or discontinued at least one AED. Additionally, 2 patients (10%) decreased the dose of one AED but increased the dose of another medication or were started on a new AED.
Discussion
In this study, we evaluated the safety and effectiveness of RNS in 20 pediatric patients with drug-resistant epilepsy treated at the Children’s Hospital of Pittsburgh since 2020. Here, we observed rates of serious adverse events, including infection, hemorrhage, stroke, and device malfunction, similar to published adult and pediatric RNS studies (Table 2).6–9 No long-term complications were reported. In terms of effectiveness, nearly three-fourths of our patients reported worthwhile seizure reduction at their latest postimplantation follow-up. Our data support the currently limited literature on RNS as a surgical intervention for pediatric patients with drug-resistant epilepsy. The RNS system also provides chronic electrocortical monitoring in an ambulatory setting. This may be leveraged to further evaluate the epileptogenic zone, which is often more diffuse in the pediatric population and is therefore less precisely estimated.10,11 This information may be used to guide adjunct or potentially more successful surgical epilepsy interventions, even when initial RNS fails to produce adequate seizure freedom.12
Review of the literature
| Authors & Year | No. of Patients | Average FU (mos) | No. of Complications | Seizure Outcome | Other |
|---|---|---|---|---|---|
| Singhal et al., 20186 | 1 | 6 | 0 (0) | 1 (100), 100% reduction | |
| Kokoszka et al., 20188 | 2 | 20 | 0 (0) | 2 (100), 85%–100% reduction | |
| Nagahama et al., 20219 | 17 | 3 | 0 (0) | 4 (23), 90%–100% reduction | 3 (18) transient postop neurological symptoms |
| 6 (35), 50%–89% reduction | |||||
| 3 (18), <50% reduction | |||||
| 4 (24), no improvement | |||||
| Panov et al., 20207 | 27 | 22 | 3 (11) postop infections | 12 (55), 75%–100% reduction | 2 (7) infections required surgical intervention; 1 (4) infection required removal of device |
| 4 (18), 50%–74% reduction | |||||
| 4 (18), 25%–49% reduction | |||||
| 2 (9), 0%–24% reduction |
Values represent the number of patients (%) unless stated otherwise.
While the safety profile of RNS is favorable, it is not free from potential serious adverse events. Implantation site infection is the most common complication, recorded at rates ranging from 3.7% to 9.4% in the adult population.1,2,4 Others, including Panov et al., reported infection rates < 11% in pediatric patients, consistent with our infection rate of 5%.7 We do not report any instances of hemorrhage or stoke. This was expected, as these complications are relatively uncommon in both children and adults undergoing RNS implantation, occurring in < 5% of patients.1,2,4,7 Furthermore, we did not record any instances of device malfunction/migration, which previous studies have reported at < 5%.1,2,4 Although limited, our study supports the few other pediatric RNS case series showing similar rates of complications to that of adults.
RNS therapy yielded various degrees of seizure freedom in our pediatric cohort similar to that noted in other pediatric RNS studies (Table 2). We evaluated postoperative response using two scales: the Engel Epilepsy Surgery Outcome Scale and the ILAE classification scale in the 16 patients with at least 3 months of follow-up. At the latest visit, 56% of our cohort (n = 9) had an Engel class IIIB outcome or better, indicating that these patients saw worthwhile seizure reduction after RNS implantation, while 43% of our cohort (n = 7) experienced no appreciable change in their seizure frequency since RNS implantation (Engel class IVA or worse). It is important to note that 3 of the 7 patients with Engel class IV outcomes have the shortest follow-up at only 6 months. As the continued refinements in detection, increased charge density, and chronic neuromodulatory effect of RNS are thought to increase its effectiveness over time, it is not surprising that those patients with the shortest follow-up time reported the least improvement in seizures.1,4 Using the ILAE classification scale, 31% of our cohort (n = 5) saw 4 or fewer seizure days per year or at least a 50% reduction of baseline seizure days, while 69% of our cohort (n = 11) reported less than 50% reduction of baseline seizure days (Table 1). Again, as observed in the Engel classification, better outcomes tended to correlate with longer follow-up. As an additional measure of responsiveness, we evaluated changes in AED treatment. AEDs remained unchanged in 75% (n = 15) of patients. This is comparable to what has been reported in other RNS studies.5
As a single-institution retrospective case series (i.e., no control group), our study has limitations. Our sample is highly heterogeneous on multiple levels, including etiology, RNS treatment strategies, and follow-up length. However, few RNS studies have focused on young patients, who may have distinct complication profiles due to different comorbidities as adults and smaller skull diameter. Crucially, future studies should address adverse events and complication rates associated with RNS treatment in pediatric patients. In that way, this study can provide preliminary safety data to motivate future research funding proposal and also be included in meta-analyses of observational data.
Conclusions
RNS is a promising neuromodulatory intervention for pediatric patients with drug-resistant epilepsy, particularly in patients for whom resective or ablative surgery is not possible, or vagus nerve stimulation has failed. This case series suggests that RNS is a safe and potentially effective surgical intervention for pediatric patients with drug-resistant epilepsy. We observed comparable rates of serious adverse events to current RNS literature in pediatric and adult populations.1,2,4,7 We anticipate further improvement in effectiveness as follow-up continues due to programming and stimulation refinement as well as a chronic neuromodulatory effect hypothesized to occur in RNS patients.1,4 Further large-scale, multicenter case-series of RNS in pediatric patients with drug-resistant epilepsy are required to determine its long-term pediatric safety and effectiveness.
Acknowledgments
Dr. Abel reports NIH NIDCD funding from awards R21 DC019217-01A1 and R01 DC013315-07.
Disclosures
Dr. Abel is a consultant for Monteris Medical and receives research funding through Monteris Medical for the LAANTERN Trial.
Author Contributions
Conception and design: Abel. Acquisition of data: Curtis, Hect, Harford, Welch. Analysis and interpretation of data: Curtis, Hect. Drafting the article: Curtis, Hect, Harford. Critically revising the article: Abel, Hect, Welch. Reviewed submitted version of manuscript: all authors. Administrative/technical/material support: Harford, Welch. Study supervision: Abel.
References
- 1↑
Bergey GK, Morrell MJ, Mizrahi EM, et al. Long-term treatment with responsive brain stimulation in adults with refractory partial seizures. Neurology. 2015;84(8):810–817.
- 2↑
Heck CN, King-Stephens D, Massey AD, et al. Two-year seizure reduction in adults with medically intractable partial onset epilepsy treated with responsive neurostimulation: final results of the RNS System Pivotal trial. Epilepsia. 2014;55(3):432–441.
- 3↑
Morrell MJ. Responsive cortical stimulation for the treatment of medically intractable partial epilepsy. Neurology. 2011;77(13):1295–1304.
- 4↑
Nair DR, Laxer KD, Weber PB, et al. Nine-year prospective efficacy and safety of brain-responsive neurostimulation for focal epilepsy. Neurology. 2020;95(9):e1244–e1256.
- 5↑
Razavi B, Rao VR, Lin C, et al. Real-world experience with direct brain-responsive neurostimulation for focal onset seizures. Epilepsia. 2020;61(8):1749–1757.
- 6↑
Singhal NS, Numis AL, Lee MB, et al. Responsive neurostimulation for treatment of pediatric drug-resistant epilepsy. Epilepsy Behav Case Rep. 2018;10:21–24.
- 7↑
Panov F, Ganaha S, Haskell J, et al. Safety of responsive neurostimulation in pediatric patients with medically refractory epilepsy. J Neurosurg Pediatr. 2020;26(5):525–532.
- 8↑
Kokoszka MA, Panov F, La Vega-Talbott M, McGoldrick PE, Wolf SM, Ghatan S. Treatment of medically refractory seizures with responsive neurostimulation: 2 pediatric cases. J Neurosurg Pediatr. 2018;21(4):421–427.
- 9↑
Nagahama Y, Zervos TM, Murata KK, et al. Real-world preliminary experience with responsive neurostimulation in pediatric epilepsy: a multicenter retrospective observational study. Neurosurgery. 2021;89(6):997–1004.
- 10↑
McGovern RA, Knight EP, Gupta A, et al. Robot-assisted stereoelectroencephalography in children. J Neurosurg Pediatr. 2018;23(3):288–296.
- 11↑
Ansari SF, Maher CO, Tubbs RS, Terry CL, Cohen-Gadol AA. Surgery for extratemporal nonlesional epilepsy in children: a meta-analysis. Childs Nerv Syst. 2010;26(7):945–951.
- 12↑
DiLorenzo DJ, Mangubat EZ, Rossi MA, Byrne RW. Chronic unlimited recording electrocorticography-guided resective epilepsy surgery: technology-enabled enhanced fidelity in seizure focus localization with improved surgical efficacy. J Neurosurg. 2014;120(6):1402–1414.
