Surgery Frees 58% of Drug-Resistant Epilepsy Patients from Seizures; Stimulation Offers Safer Alternative

Temporal lobe epilepsy is the most common form of focal epilepsy, and for a substantial fraction of patients, it resists all available medications. These individuals - often suffering dozens of seizures per month for years - face two main surgical options. One involves physically removing the brain tissue driving the seizures. The other involves implanting a device that detects seizure activity and delivers electrical stimulation to interrupt it. Both approaches use stereoelectroencephalography (SEEG) to map seizure origins with implanted electrodes before any treatment decision is made.

A systematic review following PRISMA 2020 guidelines, published in early 2026, analyzed 15 studies that directly compared or reported outcomes for SEEG-guided temporal lobe resection (TLR) and SEEG-guided responsive neurostimulation (RNS) in adults with drug-resistant temporal lobe epilepsy. Study sizes ranged from 10 to 440 participants. The findings make the tradeoffs between the two approaches starkly clear.

The Numbers on Seizure Control

SEEG-guided temporal lobe resection achieved an average seizure freedom rate of 58.5%, with results spanning 32% to 85% depending on the study population and surgical technique. Mean seizure reduction across all TLR patients was 75%, ranging from 60% to 90%. Quality of life improvements were reported in 72% to 82% of TLR patients.

Responsive neurostimulation produced average seizure freedom in 12.85% of patients - substantially lower than resection - with a mean seizure reduction of 63.2%. Quality of life improvements were seen in 44% of RNS patients.

The seizure freedom gap is large: roughly 58.5% versus 12.85%. But interpreting that gap requires understanding what kind of patients receive each treatment. Resection is generally offered when seizures originate from a well-localized zone in tissue that can be removed without unacceptable functional consequences. RNS is more commonly chosen when the seizure focus is in eloquent cortex - brain regions responsible for language, memory, or movement - where resection carries unacceptable cognitive or functional risk, or when seizures arise from multiple locations. The patient populations are therefore not directly comparable.

Safety and Cognitive Outcomes

Both interventions showed strong overall safety profiles across the reviewed studies. TLR was associated with transient memory deficits in 12% of patients and mild infections in 8%. These complications are consistent with the known risks of open brain surgery and generally resolve or improve over time.

RNS had higher device-related complication rates: lead revision was required in 10% of patients, and minor infections occurred in 4%. Crucially, cognitive outcomes - including memory and language - were better preserved with RNS than with resection. This reflects the fundamental difference between the two approaches: RNS adds hardware to the brain without permanently altering its structure, while resection irreversibly removes tissue.

For patients with dominant hemisphere temporal lobe epilepsy - where the affected tissue is on the same side as language processing - resection carries the highest cognitive risk. Memory deficits can be substantial and lasting. For these patients, RNS offers meaningful seizure reduction while avoiding permanent cognitive change.

What SEEG Adds to Both Approaches

Both treatment options in this review were specifically SEEG-guided, distinguishing them from older approaches that used surface or grid electrodes. SEEG involves placing slender depth electrodes directly into brain tissue through small burr holes, allowing three-dimensional mapping of where seizures begin and how they spread. This precision improves patient selection for resection - it allows surgeons to confirm exactly which tissue is driving seizures before committing to removal - and it improves targeting for RNS electrode placement.

The review highlights SEEG-guided TLR as offering superior seizure control for patients with well-localized epileptogenic zones. For patients where RNS is selected because resection is too risky, the 63.2% average seizure reduction - even with only 12.85% full freedom from seizures - can still represent a substantial improvement in quality of life compared with remaining on ineffective medications.

Limitations of the Evidence Base

The systematic review draws on 15 studies, a modest number given the complexity of the question. Study populations varied in age, epilepsy duration, prior surgical history, and the exact SEEG protocols used. The absence of randomized controlled trials comparing TLR and RNS head-to-head is a persistent gap in the epilepsy surgery literature - randomizing patients to brain resection versus implantable stimulation raises obvious ethical and practical challenges.

Quality assessment using the Cochrane Risk of Bias Tool and the Newcastle-Ottawa Scale identified variability in study quality across the included research. Short follow-up durations in some studies may underestimate late adverse events or delayed cognitive effects. Long-term RNS efficacy data extending beyond five years remain limited.

For clinicians and patients making treatment decisions, the review confirms that resection offers the best chance of seizure freedom when the epileptogenic zone is clearly localized and resection is safe. RNS provides a meaningful alternative when it is not, with the added benefit of leaving the door open to future adjustments in stimulation parameters as the patient's condition evolves over time.