Medical refractory epilepsy — also called drug-resistant epilepsy (DRE) or intractable epilepsy — is defined by the International League Against Epilepsy (ILAE) as the failure of adequate trials of two or more appropriate antiepileptic drugs, used alone or in combination at tolerated doses, to achieve sustained seizure freedom. Approximately 30–40% of all people with epilepsy have drug-resistant epilepsy. For many, epilepsy surgery is the most effective available treatment — offering seizure freedom in 60–80% of carefully selected candidates.
Drug-resistant epilepsy (DRE) · intractable epilepsy · pharmacoresistant epilepsy — all synonyms for the same condition.
Failure of two appropriate antiepileptic drugs at adequate doses — whether alone or in combination.
30–40% of all epilepsy patients. Approximately 3.5 million people in India have drug-resistant epilepsy.
Hippocampal sclerosis · focal cortical dysplasia (FCD) · low-grade tumour (DNET, ganglioglioma) · cavernous malformation · genetic epilepsy syndrome.
65–80% seizure-free (Engel Class I) after temporal lobectomy for hippocampal sclerosis. Evidence: Wiebe NEJM 2001 RCT.
Immediately after the second drug fails — not after 5, 10, or 20 years. Earlier surgery produces better cognitive outcomes.
Video-EEG monitoring unit · sEEG with Stealth AutoGuide robotic guidance · Dr. Sita Jayalakshmi (neurology) · Senior Consultant, Neurosurgery.
Epilepsy Surgery · Neurology · Neurosurgery
040-44885000
⚠ EMERGENCY: A seizure lasting more than 5 minutes · a second seizure with no recovery in between · any seizure with no breathing or blue lips: call 040-46324365 immediately. This is status epilepticus — a medical emergency.
DefinitionMedical refractory epilepsy — also called drug-resistant epilepsy (DRE) or intractable epilepsy — is defined by the International League Against Epilepsy (ILAE) as the failure of adequate trials of two or more appropriate antiepileptic drugs, used alone or in combination at tolerated doses, to achieve sustained seizure freedom. Approximately 30–40% of all people with epilepsy have drug-resistant epilepsy. For many, epilepsy surgery is the most effective available treatment — offering seizure freedom in 60–80% of carefully selected candidates.
Expert epilepsy diagnosis and surgical evaluation at KIMS Neuroscience Hyderabad.
040-44885000International enquiries neuroscience@kimshospitals.comEpilepsy is a condition in which the brain’s electrical system produces sudden, abnormal bursts of activity that cause seizures. For most people with epilepsy, the right antiepileptic medication stops the seizures. For approximately one in three, it does not — no matter how many medications are tried, at whatever doses, in whatever combinations. This is medical refractory epilepsy.
The word “refractory” comes from the Latin for “resisting treatment.” It does not mean untreatable. It means that antiepileptic drugs alone are not sufficient — and that other interventions, including epilepsy surgery, vagus nerve stimulation, deep brain stimulation, or dietary therapy, need to be considered. The tragedy of refractory epilepsy in India is not that it cannot be treated — it is that most patients who could benefit from surgery are never referred for evaluation.
At KIMS Hyderabad, the comprehensive epilepsy programme provides the full evaluation-to-treatment pathway for patients across Telangana and Andhra Pradesh whose seizures have not responded to medication.
Some patients’ bodies metabolise antiepileptic drugs faster than normal, so drug levels in the blood never reach therapeutic concentrations in the brain. P-glycoprotein overexpression in the blood-brain barrier actively pumps certain drugs out of the brain before they can work. This is a biological mechanism, not a compliance problem.
Sodium channels — the target of drugs like carbamazepine and lamotrigine — can mutate in the epileptogenic zone so that the drugs no longer block them effectively. Adequate drug levels are present; the seizures continue because the target has changed.
Many refractory cases are caused by a structural abnormality: hippocampal sclerosis, focal cortical dysplasia, a low-grade tumour, or a cavernous malformation. Drugs suppress seizures globally but cannot correct the specific structural lesion generating them. Surgery to remove the lesion is the only curative intervention for these cases.
Medical refractory epilepsy is a functional category applied across many underlying epilepsy syndromes. The most clinically important distinction is:
| Type | Definition | Surgical Outcome | Most Common Cause |
|---|---|---|---|
| Lesional DRE | Structural lesion visible on MRI and corresponds to seizure onset zone | Best outcomes: 65–85% seizure freedom | Hippocampal sclerosis · FCD · low-grade tumour · cavernoma |
| MRI-negative DRE | MRI appears normal · requires advanced evaluation | 50–70% seizure freedom with full evaluation | Subtle FCD below MRI resolution · genetic causes |
| Temporal lobe epilepsy | Most common surgically treatable type | 65–80% seizure freedom after temporal lobectomy | Mesial temporal sclerosis (hippocampal sclerosis) |
| Extratemporal epilepsy | Frontal, parietal, or occipital lobe onset | 40–75% depending on lesion type | FCD · post-traumatic scar · cavernoma |
| Generalised DRE | Whole-brain onset · not typically resectable | Neuromodulation (VNS, DBS, CC) · no curative surgery | Lennox-Gastaut · Dravet syndrome · genetic epilepsies |
Medical refractory epilepsy is defined by two failed drug trials — not ten, not twenty. At that point, surgical evaluation is the clinically correct next step, and it offers seizure freedom in 65–80% of selected candidates.
Medical refractory epilepsy is not a sudden new symptom — it is a pattern that develops over time. The moment to recognise it and act is when two appropriate medications at adequate doses have failed to stop the seizures. Many patients and families accept continued seizures as “just how epilepsy is” for years, missing the window when surgical intervention is most effective.
The ILAE definition is specific: two adequately trialled drugs, not two total drugs. A drug is adequately trialled if it is appropriate for the seizure type, taken at a therapeutic dose, taken for a sufficient duration, and stopped for clear reasons of failure (continued seizures) rather than side effects alone.
The patient stops, stares, does not respond, and may make repetitive automatic movements (lip-smacking, hand-picking, chewing). Lasts 1–2 minutes, followed by confusion. No memory of the episode. Most common in temporal lobe epilepsy. Can occur multiple times per week.
A focal seizure that spreads to both hemispheres: body stiffens (tonic), then jerks rhythmically (clonic). Loss of consciousness, possible tongue bite, bladder incontinence. Highest SUDEP risk and highest injury risk. Carries the most immediate daily danger.
Patient remains conscious during the seizure. Content reflects the focus location: temporal → rising epigastric sensation, fear, déjà vu · motor focus → rhythmic limb jerking · occipital → flashing lights. Often the warning sign before a larger seizure.
Sudden loss of muscle tone causing an instant, unprotected fall to the ground. No warning. No loss of consciousness. Lasts 1–2 seconds. Characteristic of Lennox-Gastaut syndrome. Repeated head injuries are the primary physical consequence.
Medical refractory epilepsy causes harm not only through the seizures themselves but through cumulative neurological, cognitive, psychosocial, and safety consequences:
• SUDEP risk: Approximately 1 in 1,000 patient-years in refractory epilepsy — significantly higher than in controlled epilepsy.
• Cognitive decline: Repeated tonic-clonic seizures cause progressive hippocampal atrophy. The longer refractory epilepsy goes untreated, the less function there is to preserve.
• Driving prohibition: Active seizures legally prohibit driving in India. Refractory epilepsy permanently restricts independence and livelihood for working-age patients.
• Employment and education: Unpredictable seizures make sustained employment and academic completion extremely difficult for most patients with refractory epilepsy.
• Medication burden: 3–4 concurrent antiepileptic drugs cause fatigue, cognitive slowing, mood changes, liver toxicity, and teratogenicity in women of childbearing age.
• Psychosocial disability: Depression affects 30–40% of refractory epilepsy patients — two to four times the general population rate.
The average time from epilepsy onset to epilepsy surgery in India is 15–20 years. In well-functioning international centres, it is 5–8 years. The gap represents avoidable cumulative harm.
Refractory epilepsy causes harm not only through seizures but through progressive cognitive decline, SUDEP risk, and lost independence — every year without evaluation adds to this damage and cannot be recovered.
Before surgical evaluation begins, the KIMS epilepsy team confirms that the patient truly has drug-resistant epilepsy — and that the events are epileptic rather than psychogenic non-epileptic attacks (PNEA), which affect approximately 20–30% of patients referred with apparent drug-resistant epilepsy and require a completely different management approach.
Psychogenic non-epileptic attacks (PNEA) can appear identical to epileptic seizures. The only reliable way to distinguish them is long-term video-EEG monitoring, in which a captured event is correlated with simultaneously recorded EEG. PNEA produce no EEG change during the event.
Drug resistance is defined by failure of two adequate trials. The KIMS team reviews which drugs were used, at what doses, whether drug levels were measured, and whether the drugs chosen were appropriate for the specific seizure type.
Admitted to the KIMS Epilepsy Monitoring Unit for 5–10 days. Medications gradually reduced to provoke habitual seizures. Each seizure recorded simultaneously on video and EEG. Confirms that events are epileptic · characterises seizure semiology · EEG-localises the ictal onset zone. KIMS has three continuous long-term telemetry video-EEG units.
1 mm isotropic T1 MPRAGE · FLAIR and T2 coronal cuts perpendicular to hippocampus · SWI / susceptibility-weighted imaging · DWI. Read by a neuroradiologist with epilepsy imaging expertise. MRI-negative epilepsy (20–30% of cases) requires the full evaluation pathway.
Measures glucose metabolism. The seizure onset zone shows reduced metabolism in the interictal state. Particularly valuable in MRI-negative epilepsy where the focus may not be visible on structural MRI. SISCOM provides complementary perfusion data.
Verbal and non-verbal memory · language lateralisation · attention, processing speed, executive function. Serves two purposes: predicts post-operative cognitive risk (verbal memory after dominant temporal surgery) and establishes a pre-operative baseline for comparison.
All Phase 1 data reviewed by: Dr. Sita Jayalakshmi (epilepsy neurology) · Senior Consultant, Neurosurgery (epilepsy surgery) · neuroradiologist · neuropsychologist. Decides: proceed to surgery (concordant focus identified) or proceed to Phase 2 invasive EEG.
sEEG implants 8–15+ depth electrodes through small skull burr holes to record directly from specific cortical and subcortical targets. Patient monitored for 1–2 weeks capturing spontaneous seizures from depth electrodes. Provides millimetre-resolution localisation of the seizure onset zone. At KIMS, sEEG is performed using the Stealth AutoGuide™ robotic cranial guidance platform — the first in India — for precise electrode placement at each planned trajectory. Can sample deep structures (hippocampus, insula, orbitofrontal cortex, thalamus) that surface grids cannot reach.
| Phase 1 Finding | Implication | Next Step |
|---|---|---|
| MRI lesion + concordant EEG | Seizure focus confidently identified | Proceed directly to resective surgery |
| MRI lesion + discordant EEG | Lesion may not be the sole focus | Phase 2 sEEG to confirm · then surgery if confirmed |
| MRI negative + focal EEG + focal PET | Focus likely in hypometabolic region | Phase 2 sEEG to define · FCD likely on histology |
| MRI negative + non-focal EEG | No focal evidence from Phase 1 | Phase 2 sEEG required · or consider neuromodulation |
| Generalised EEG + no focal semiology | Generalised epilepsy — not resectable | Corpus callosotomy (drops) · VNS · DBS · dietary therapy |
| Events without EEG change | PNEA confirmed — not drug-resistant epilepsy | Psychiatric assessment · CBT · not a surgical candidate |
The KIMS evaluation answers one clinical question: does this patient have a seizure focus that can be safely removed? Phase 1 answers this non-invasively; Phase 2 sEEG answers it definitively when Phase 1 is inconclusive. Many patients told surgery is “not possible” have never had Phase 2 evaluation.
The treatment of medical refractory epilepsy at KIMS covers the full spectrum: resective surgery, disconnective surgery, Gamma Knife, neuromodulation, and dietary therapy. The recommendation for each patient is made by the multidisciplinary epilepsy conference based on seizure type, focus location, and individual circumstances.
Seizure-Free Rate (Temporal Lobe)
Drug-Resistant Epilepsy Patients in India
Removes the anterior temporal lobe including the hippocampus, amygdala, and parahippocampal gyrus on the side of the seizure focus. The landmark NEJM trial (Wiebe et al., 2001) demonstrated 64% seizure freedom at 1 year with surgery versus 8% with continued medical therapy. At KIMS, temporal lobectomy achieves Engel Class I (seizure freedom) in 65–80% of patients with hippocampal sclerosis. Surgery: 2–3 hours. Hospital stay: 5–7 days.
For strictly mesial temporal seizure onset: removes only the hippocampus and amygdala, preserving the lateral temporal neocortex. Equivalent seizure-free rates to temporal lobectomy with better preservation of lateral temporal cortex function. Choice between temporal lobectomy and SAH guided by intraoperative ECoG.
Surgical resection of the lesion (FCD, DNET, ganglioglioma, cavernoma) and surrounding epileptogenic cortex. Achieves seizure freedom in 50–85% depending on completeness of resection and histopathology. Performed across all brain regions including eloquent cortex, using neuronavigation, awake craniotomy, intraoperative ECoG, and continuous neurophysiological monitoring.
More challenging than temporal lobe surgery because seizure onset often involves eloquent cortex. Phase 2 sEEG is more commonly required. When a resectable focus is identified, surgery achieves seizure freedom in 40–75%. Intraoperative brain mapping used for all operations near eloquent cortex at KIMS.
Divides the corpus callosum, preventing seizures from spreading bilaterally. Does not eliminate seizures but reduces severity: primary indication is atonic seizures (drop attacks) causing repeated head injuries, most commonly in Lennox-Gastaut syndrome. Reduces drop attacks in 70–85% of patients. KIMS performs anterior two-thirds callosotomy as the standard approach.
Functional disconnection of one hemisphere. Used for Rasmussen encephalitis, hemimegalencephaly, or large perinatal porencephaly. Achieves 70–85% seizure freedom. Appropriate only when the affected hemisphere is already non-functional (severe established hemiplegia). KIMS performs the modified lateral approach that reduces operative blood loss.
Posterior disconnection (also called posterior quadrant disconnection or posterior quadrantotomy) is a surgical procedure in which the posterior cerebral cortex — the occipital lobe, posterior parietal lobe, and posterior temporal lobe — is functionally disconnected from the anterior brain by dividing the white matter tracts, the posterior corpus callosum, and the thalamo-cortical connections supplying the posterior quadrant. Unlike resection, the cortex is not removed but is rendered electrically isolated from the seizure-propagating circuitry of the rest of the brain.
Indication: Posterior quadrant epilepsy — seizures arising from the occipital lobe, parieto-occipital junction, or posterior temporal regions — where the epileptogenic zone is too extensive for localised resection, but where the posterior cortex is already functionally compromised. In most candidates, a fixed contralateral visual field defect (homonymous hemianopia) is already present from the underlying epilepsy or pathology, meaning the expected visual consequence of the surgery does not add functional loss.
Common underlying pathology: Posterior cortical dysplasia (focal cortical dysplasia Type II in the occipital or parieto-occipital region) · Sturge-Weber syndrome with posterior-predominant leptomeningeal angiomatosis · Hemimegalencephaly with predominant posterior involvement · Posterior perinatal infarct or porencephaly · Posterior-predominant Rasmussen encephalitis.
How it differs from hemispherotomy: Hemispherotomy disconnects the entire hemisphere. Posterior disconnection disconnects only the posterior quadrant, preserving motor and language cortex in the anterior hemisphere. It is the appropriate choice when seizure onset is confined to the posterior quadrant and the anterior (frontal and sensorimotor) cortex is functioning normally and generating no seizures.
Surgical outcomes: Published series report seizure freedom (Engel Class I) in approximately 60–75% of patients. Outcomes are best when pre-surgical evaluation with sEEG has confirmed posterior quadrant onset and when the underlying pathology is a well-circumscribed cortical dysplasia or vascular malformation. KIMS performs posterior disconnection using the modified lateral approach, with intraoperative neurophysiological monitoring to protect the motor strip boundary.
Expected deficits: Contralateral homonymous hemianopia — loss of the visual field on the side opposite the surgery. This is an expected and accepted consequence of the procedure, and in most candidates is already established before surgery. Motor function is preserved because the motor strip is not involved. Language function is preserved when the dominant hemisphere's posterior language cortex is not part of the disconnection (guided by pre-operative fMRI and sEEG language mapping).
For patients who refuse open surgery or are at elevated surgical risk. Seizure-free rates: approximately 50–65% at 2–3 years — lower than surgical temporal lobectomy, but without open surgery recovery burden. Radiation effect develops over 6–24 months. Available at KIMS.
For cavernous malformations, low-grade tumours, and cortical dysplasias in surgically inaccessible or high-risk locations. Considered case-by-case when resection carries unacceptable risk.
Implantable device: pulse generator under the left clavicle, connected to the left vagus nerve. Delivers intermittent stimulation that modulates cortical excitability. Reduces seizure frequency by ≥50% in approximately 50–60% of patients, improving over 5+ years. Magnet mode can abort or shorten a seizure on activation. Appropriate when no resectable focus exists or resection was unsuccessful.
Bilateral DBS of the anterior nucleus of the thalamus (ANT-DBS). The SANTE trial (2010) demonstrated 69% seizure reduction at 2 years. Responsive neurostimulation (RNS) — a closed-loop system that detects and responds to seizure onset — is an emerging option for patients with a well-localised but unresectable focus. Both available at KIMS.
High-fat, very low-carbohydrate diet producing ketone bodies with anticonvulsant effect. Reduces seizures by ≥50% in approximately 50% of drug-resistant patients. Most effective in: Dravet syndrome, Lennox-Gastaut syndrome, GLUT-1 deficiency, pyruvate dehydrogenase deficiency. Medically supervised at KIMS. Requires regular blood and urine monitoring.
Less restrictive alternatives for patients who cannot maintain the classical KD. Comparable efficacy for adolescents and adults. Modified Atkins Diet allows up to 15–20g carbohydrate per day versus the classical KD’s near-zero carbohydrate threshold.
Many patients labelled “treatment-resistant” are on inappropriate drugs for their seizure type · some have never had drug level measurement · rational polytherapy (combining drugs with complementary mechanisms) has not been attempted. A formal medication review by Dr. Sita Jayalakshmi is the first clinical action for every patient referred for surgical evaluation at KIMS.
Seven treatment options are available at KIMS for drug-resistant epilepsy — from surgery that cures to neuromodulation that controls. The right choice depends on seizure type, focus location, and individual evaluation, not on which option is most familiar to the referring physician.
| Surgery Type | Engel Class I (Seizure-Free) | Patient Group | Evidence Level |
|---|---|---|---|
| Temporal lobectomy (hippocampal sclerosis) | 65–80% at 1 year · 55–70% at 5 years | TLE with MRI-confirmed hippocampal sclerosis | Class I · NEJM RCT (Wiebe 2001) |
| Temporal lobectomy (tumour/FCD) | 70–85% at 2 years | Low-grade tumour or cortical dysplasia | Class II–III |
| Lesionectomy (extratemporal) | 50–75% at 2 years | FCD · ganglioglioma · cavernoma | Class II |
| Hemispherotomy | 70–85% at 2 years | Catastrophic unilateral hemispheric epilepsy | Class II–III |
| Posterior Disconnection | 60–75% at 2 years | Posterior quadrant epilepsy (occipital/parieto-occipital onset) with established/expected visual field defect | Class II–III |
| Corpus callosotomy | 70–85% drop attack reduction | Lennox-Gastaut · drop attacks | Class II |
| VNS | 50–60% achieve ≥50% reduction | Non-resectable DRE | Class II RCT |
| ANT-DBS | 69% seizure reduction at 2 years (SANTE trial) | Focal DRE · non-resectable | Class I RCT |
| Ketogenic diet | 50% achieve ≥50% reduction · 5–10% seizure-free | DRE generalised epilepsy | Class II RCT |
One year of confirmed seizure freedom after surgery, verified by the epileptologist, makes the patient eligible to apply for a driving licence in India. For a patient who had seizures for 10–20 years, regaining the ability to drive is often described as the single most life-changing outcome of surgery.
Seizure freedom removes the primary barrier to sustained professional and academic life. Many patients return to employment or complete education within 12–18 months of successful surgery.
By 2–3 years after surgery, approximately 30–40% of seizure-free patients are completely off antiepileptic medication. This eliminates years of side effects, costs, and teratogenic risk for women planning pregnancy.
Measurable improvements in attention, memory, mood, and quality of life occur in the majority of seizure-free patients within the first year of surgery — independent of any cognitive effect from the surgery itself.
Clinical review at 6 weeks, 3 months, 6 months, and 12 months · EEG at 3 and 12 months · MRI at 3 months and 12 months · Neuropsychological re-assessment at 12 months · Antiepileptic drug tapering begins at 12 months of seizure freedom.
One drug reduced at a time, at 25% of dose per 3 months, with close clinical monitoring. If seizures recur during tapering, the drug is increased back to the last effective dose. Complete drug freedom achieved in approximately 30–40% of seizure-free patients by year 3.
For patients from Vijayawada, Nellore, Karimnagar, and other AP/Telangana cities: 3-month MRI and EEG performed locally, images sent to KIMS electronically · video consultations with the neurology team for routine reviews · annual in-person assessments at KIMS. Call 040-44885000 to arrange.
Epilepsy surgery at KIMS achieves seizure freedom in 65–80% of temporal lobe cases. Seizure freedom means driving, working, and living without the constant anticipation of the next seizure — outcomes that no antiepileptic drug has achieved for these patients.
Related Conditions
Explore related epilepsy and seizure disorders at KIMS Neuroscience:
The following questions are structured for Google People Also Ask and AI Overview extraction. Each begins with a direct declarative answer before providing supporting detail.
The single most important answer on this page: two failed drugs is the threshold for surgical referral, not ten. The evaluation does not commit a patient to surgery — it determines whether surgery is an option. That is worth one phone call to 040-44885000.
KIMS Neuroscience Epilepsy Programme
Two medications have failed. The seizures are still happening. The evaluation has not been done. At KIMS Hyderabad, epilepsy surgery, the complete evaluation pathway, and senior neurology and neurosurgery consultants are available. The evaluation — not the surgery, just the evaluation — is the first step. Call 040-44885000 · Email: neuroscience@kimshospitals.com
neuroscience@kimshospitals.com