Efficacy of neurosurgical interventions for epilepsy in polymicrogyria: A systematic review

Abstract

Polymicrogyria (PMG) is a rare malformation of cortical development (MCD) characterized by abnormal neuronal architecture, often associated with epilepsy. Neurosurgical interventions have been explored, but their effectiveness remains a subject of ongoing research and debate. This systematic literature review aims to provide a comprehensive analysis of surgical interventions for epilepsy in PMG individuals, their outcomes, complications, and factors influencing surgical decision‐making. This systematic review, conducted according to PRISMA guidelines and registered on PROSPERO (CRD420251010221), analyzed studies on individuals with PMG and epilepsy undergoing neurosurgical treatment. A structured search across five databases was performed. Two reviewers independently selected and extracted data on clinical characteristics, surgical techniques, and outcomes, with seizure control assessed through Engel and ILAE classifications. Statistical analyses included Chi‐square and Mann–Whitney U tests. Study quality was evaluated. A total of 47 studies involving 225 patients were included, mostly consisting of case reports or series. The majority of patients underwent resective surgery or hemispherotomy. Postsurgical outcomes were evaluated using Engel or ILAE classifications in 178 cases: 64.61% achieved seizure freedom (Engel/ILAE I). A shorter epilepsy duration was significantly associated with better outcomes. Only a minority received neuropsychological assessments, with cognitive improvements observed mainly in patients operated at a younger age. Postoperative complications were discussed. This review highlights the potential effectiveness of neurosurgical interventions in selected patients with PMG and drug‐resistant epilepsy. Despite the overall encouraging seizure outcomes, the limited and heterogeneous reporting of cognitive and quality‐of‐life measures underscores the need for more comprehensive and standardized pre‐ and postsurgical assessments to better inform clinical decision‐making and long‐term management.

Plain Language Summary

This study looked at the effects of brain surgery in people with epilepsy caused by polymicrogyria, a condition where the brain does not form properly before birth and develops an unusual surface. Many patients became seizure‐free or had fewer seizures after surgery, especially when performed at an earlier stage. More research is needed to understand how surgery affects mental functions (e.g., memory, attention) and quality of life of these patients.

Key points.

• Surgery led to seizure freedom in 64.61% of patients with PMG‐related drug‐resistant epilepsy.

• Better outcomes were associated with a shorter duration of epilepsy.

• Early surgery was associated with improved seizure control and cognitive outcomes.

• Only 33.78% of patients received neuropsychological assessments after surgery.

• Postsurgical complications included motor, cognitive, and visual deficits in some patients.

1. INTRODUCTION

Polymicrogyria (PMG) is a rare malformation of cortical development (MCD) characterized by an excessive number of abnormally small cerebral gyri with cortical overfolding.^{1, 2} This condition is often associated with epilepsy, exhibiting variation in presentation, underlying pathology, and treatment response.^{3, 4} Recent studies^{3, 5} revealed a population‐based prevalence of PMG of 2.3 per 10 000 children and an estimated annual incidence of 1.9 per 10 000 person‐years; moreover, PMG shows a high prevalence of epilepsy (78%), with early onset seizures in 44%, and a substantial proportion (56%) of patients experiencing drug‐resistant epilepsy (DRE). Neurodevelopmental symptoms were observed in 94% of individuals. Several causes may lead to the occurrence of PMG: genetic mutations, congenital infections, metabolic disorders, vascular injuries, and perinatal events. ³ However, the correct prevalence of different etiologies is still uncertain. In a cohort of 109 patients, Kolbjer et al. ⁵ found genetic mutations as the most common etiology (32% of them reported pathogenic variants, while 12% revealed variants of uncertain significance), followed by vascular etiology (5.5%), infection (4.5%), and the etiology in the remnant of cases remained unknown.

Neurosurgical interventions have been explored as potential treatments for epilepsy in PMG patients, but their effectiveness remains under investigation.

For patients with PMG‐related DRE, the decision to opt for surgical intervention hinges on several factors: the accurate localization of the epileptogenic zone (EZ), the specific brain region responsible for generating seizures, and the mapping of the patient's eloquent brain areas and cognitive functions are pivotal for surgery decision‐making.^{6, 7} These presurgical evaluations involve advanced neuroimaging techniques, such as brain magnetic resonance imaging (MRI), diffusion tensor imaging (DTI), functional magnetic resonance imaging (fMRI), positron emission tomography (PET), single‐photon emission computed tomography (SPECT), electroencephalogram (EEG), stereoelectroencephalography (SEEG), and an exhaustive neuropsychological test protocol ⁸ to inform the risks and benefits of resection, disconnection, or selected minimally invasive procedures aimed to remove (or disconnect) the EZ.

However, important knowledge gaps remain specific to epilepsy related to PMG. The effectiveness of epilepsy surgery in patients with PMG remains uncertain, in part because concordance between the EZ and the PMG is often not demonstrated or incompletely reported. Neuropsychological and quality‐of‐life outcomes are infrequently and nonuniformly assessed, limiting conclusions on cognitive trajectories after surgery.

Therefore, this systematic review aims to: (i) review seizure outcomes after epilepsy surgery in PMG using standardized classifications, when available, (Engel/ILAE) across surgical techniques; (ii) explore potential predictors of outcome, with emphasis on EZ‐PMG correspondence and epilepsy duration; (iii) summarize postoperative complications and changes in anti‐seizure medications (ASMs); and (iv) investigate the available evidence on neuropsychological and quality‐of‐life outcomes to highlight priorities for future research.

2. METHODS

This systematic review was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) guidelines, ⁹ and has been registered in the PROSPERO database (registration number CRD420251010221).

2.1. Databases and search strategy

A systematic literature search was conducted on October 30, 2024, across the following databases and registers: PubMed/MEDLINE, SCOPUS, Web of Science, clinicaltrial.gov, and clinicaltrialsregister.eu. A structured search strategy was conducted incorporating pertinent keywords, as follows: (Polymicrogyria OR Polymicrogyrias OR PMG) AND (Epilepsy OR Epilepsies) AND (Surgery OR “Surgical Treatment” OR “Surgical Intervention” OR “Surgical Management”).

As shown in Figure 1, a total of 391 scientific papers were retrieved from all databases. After eliminating duplicates, 200 unique records were assessed in accordance with predefined inclusion and exclusion criteria. The references of the articles found were considered.

FIGURE 1.

PRISMA flowchart illustrating the systematic review process adopted in this study.

2.2. Eligibility criteria and study selection

Inclusion criteria: studies published in English; studies involving both pediatric and adult patients with confirmed diagnoses of PMG and concurrent epilepsy who underwent neurosurgical treatment; posttreatment outcome reported by the authors; randomized control trials (RCT) or nonrandomized control trials (n‐RCT) studies, case reports/series.

Exclusion criteria: studies not published in English; studies not involving patients with confirmed diagnoses of PMG and epilepsy; studies lacking a reported postsurgical outcome; studies with a probable or confirmed overlap of participant cohorts; reviews, book/chapter, letter to the editor, and other papers with not available data.

2.3. Data extractions and synthesis

Two reviewers (SR, MLB) independently screened the selected articles. Conflicts were resolved through discussion with a third reviewer (FOC) to ensure unanimous agreement in the studies selection. All researchers confirmed the final evaluation. The relevant information was extracted using a predefined table designed in accordance with all the authors to ensure an objective and standardized methodological approach. The features collected from the studies were: study design; patient demographics; age at seizure onset; age at surgical intervention; epilepsy duration; correspondence between EZ and PMG; surgical techniques; outcome measurements; postsurgery outcomes; postsurgery antiepileptic drugs (AEDs); follow‐up duration.

Procedures were grouped as: resection/lobectomy; hemispheric disconnection (hemispherectomy and hemispherotomy); callosotomy; targeted disconnection (disconnection of a limited epileptogenic network without complete hemispheric disconnection); SEEG‐guided ablation (e.g., RF‐TC, RFA).

The features were pooled according to the efficacy/outcome reported by the authors, the type of surgical treatment and the PMG‐to‐EZ correspondence. Studies that reported an official classification, such as the Engel or ILAE classification,^{10, 11} were treated separately. Since our study focused on the postsurgical outcomes, we referred to the two official classifications, Engel and ILAE, whenever applicable. Both the Engel and ILAE classification categorize outcomes based on seizure reduction, ranging from seizure‐free to no improvement or worsening of seizures.^{10, 11} Since only a minority of studies reported Engel Class I subclasses (i.e. 1A and 1B), we analyzed these together, as Engel I, meaning “freedom from disabling seizures” (FDS). To harmonize with ILAE classification, FDS was defined as ILAE I–II (ILAE I = no seizures; ILAE II = auras only). Accordingly, all comparative analyses use FDS (Engel I/ILAE I–II) vs. non‐FDS (Engel II–IV and ILAE III–VI). This approach minimizes selective data loss due to inconsistent subclass reporting.

Given the nonnormally distributed nature of the data in our sample, the median and the interquartile range (IQR) at the 25th and 75th percentiles were calculated and reported. When possible, to identify possible associations between Engel/ILAE classes, age (at epilepsy onset and at surgery), duration of epilepsy, AEDs intake after surgery, and correspondence between EZ and PMG, the Chi‐square test and the nonparametric Mann–Whitney U test were performed. The significance level was set at <0.05.

Finally, the level of evidence was assigned for each retrieved work, in accordance with the Oxford Centre for Evidence‐based Medicine (OCEBM), which categorizes studies into five levels, ranging from Level 1 (highest quality) to Level 5 (lowest quality). ¹² Since most of the studies retrieved from the literature consisted of case reports or case series, the “Quality Assessment Tool for Case Series Studies” by the National Heart, Lung and Blood Institute (NHLBI) ¹³ was used to score the quality of evidence, based on a 0‐to‐9 points scale, where 9 represents the highest level of evidence quality. Based on the assigned scores, the studies were consequently categorized by the authors as follows: Poor (0‐to‐3 points); Fair (4‐to‐6 points); Good (7‐to‐9 points). References were handled with Zotero (ver.6.0.27).

3. RESULTS

As illustrated in Figure 1, as of October 30, 2024, a total of 391 articles were retrieved from databases and online registries. After duplicate removal, the screening process based on eligibility criteria resulted in the selection of 47 articles, including two studies retrieved from the references of included works. Among these, no RCTs from online registries met the inclusion criteria. The retrieved articles spanned from 1991 to 2024.

As presented in Table 1, out of the 47 studies,^{14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60} 45 were case reports/series (95.75%) and two observational studies (4.25%). A total of 45 out of 47 articles (95.75%) were assigned an OCEBM‐based evidence level of 4, while the remaining (2/47) included studies were assigned a level of 3 (4.25%). Furthermore, based on the NHLBI scale, the quality of the studies was categorized as follows (Table 1): 7/47, Poor (14.89%); 22/47, Fair (46.81%); 18/47, Good (38.3%).

TABLE 1.

Level of evidence and quality assessment of the included studies.

Reference	Type of study	Level of evidence (OCEBM)	NHLBI quality assessment score (0–9 pts.)	NHLBI quality rating (poor/fair/good)
Aung et al., 2024	Observational study	3	8/9	Good
Baba et al., 2018	Case series	4	4/9	Fair
Barba et al., 2013	Case series	4	6/9	Fair
Caraballo et al., 2013	Case series	4	5/9	Fair
Caraballo et al., 2016	Case series	4	6/9	Fair
Chassoux et al., 2008	Case series	4	6/9	Fair
Chugani et al., 2011	Case series	4	7/9	Good
Chun‐Wei Chang et al., 2019	Case series	4	7/9	Good
Cossu et al., 2016	Case series	4	8/9	Good
Dan‐Dan Wang et al., 2014	Case series	4	5/9	Fair
Doris Wang et al., 2014	Case series	4	7/9	Good
Elliott et al., 2022	Case report	4	3/9	Poor
Ferri et al., 2023	Case report	4	4/9	Fair
Fohlen et al., 2019	Case series	4	7/9	Good
Fohlen et al., 2022	Case series	4	5/9	Fair
Isik et al., 2007	Case report	4	3/9	Poor
Iwasaki et al., 2012	Case series	4	8/9	Good
Jalloh et al., 2018	Case series	4	7/9	Good
Jansen et al., 2005	Observational study	3	7/9	Good
Jeong et al., 2017	Case series	4	7/9	Good
Kawai et al., 2004	Case series	4	6/9	Fair
Kuo et al., 2023	Case series	4	3/9	Poor
Kuzniecky et al., 1997	Case series	4	5/9	Fair
Leblanc et al., 1991	Case series	4	4/9	Fair
Liu et al., 2019	Case report	4	4/9	Fair
Maillard & Ramantani, 2018	Case report	4	2/9	Poor
Maillard et al., 2017	Case series	4	8/9	Good
Maton et al., 2010	Case series	4	6/9	Fair
Moletto et al., 2018	Case report	4	7/9	Good
Mullatti et al., 2019	Case series	4	8/9	Good
Nagai et al., 2016	Case report	4	2/9	Poor
Nishibayashi et al., 2007	Case report	4	5/9	Fair
Otsuki et al., 2013	Case series	4	8/9	Good
Podkorytova et al., 2016	Case series	4	5/9	Fair
RamachandranNair et al., 2006	Case report	4	3/9	Poor
Ramachandran Nair et al., 2008	Case series	4	7/9	Good
Ramantani et al., 2013	Case series	4	7/9	Good
Ravindra et al., 2023	Case report	4	7/9	Good
Säisänen et al., 2010	Case series	4	5/9	Fair
Saito et al., 2009	Case series	4	5/9	Fair
Sculier et al., 2021	Case series	4	6/9	Fair
Shamim et al., 2021	Case series	4	6/9	Fair
Urbach et al., 2004	Case report	4	1/9	Poor
Wang et al., 2016	Case series	4	8/9	Good
Wichert‐Ana et al., 2008	Case series	4	6/9	Fair
Zhang et al., 2020	Case report	4	3/9	Fair
Zsoter et al., 2012	Case series	4	4/9	Fair

The retrieved studies involved a total of 225 patients, of whom 117 (52%) were males, 96 (42.67%) were females, while in 12 cases gender was not specified (5.33%). The median age at seizure onset was 4.21 years (range 0.0–32.0; IQR: 0.55–12.0), while the median age at intervention was 15.0 years (range 0.25–51.0; IQR: 7.9–27.0), with a median time elapsed between seizure onset and surgical intervention of 9.95 years (range 0.25–42.0; IQR: 6.5–16.9).

All the relevant details for each included study are reported in Table 2.

TABLE 2.

Population and clinical features of the included studies.

Reference	No. patients (male/female)	Seizure age at onset (years; SD; range)	Age at surgical intervention (years; SD; range)	Epilepsy duration (years; SD)	PMG‐to‐EZ correspondence (No. patients)	Surgical treatment (No. patients)	Postoperative complications (No. patients)	Postsurgical outcome measures	Outcome [Engel/ILAE/clinical classification (No. patients)]	Follow‐up (SD; range; No. patients)
Aung et al., 2024	23 (11 M; 12F)	5 (IQR 1.67–17.0)	/	11 (IQR 3.5–22.0)	EZ = PMG (6) EZ ≈ PMG (7) EZ ≠ PMG (3) n.s. (7)	Resection (19); hemispherectomy (4)	/	Engel; Neurological examination	Engel I (12); Engel II (5); Engel III (3); DS (3)	31 months (IQR 17–60.15)
Baba et al., 2018	3 (F)	0.25; at birth; 3	7; 3.5; 8	6.75; 3.5; 5	EZ ≈ PMG (3)	Callosotomy (3)	/	EEG; Neurological examination	DS (3) + developmental progress: improvement in lethargy and hyperactivity subscales of Aberrant Behavior Checklist (1)	84 months (1); 36 months (1); 10 months (1)
Barba et al., 2013	1 (M)	2	17	15	EZ ≈ PMG (1)	Resection (1)	/	Engel; Neuropsychological assessment	Engel IV + improvement in cognitive functions (1)	60 months (1)
Caraballo et al., 2013	2 (n.s.)	/	/	/	/	Resection (1); hemispherectomy (1)	/	EEG; Neuropsychological assessment	SF (2)	144 months (36–264) (2).
Caraballo et al., 2016	1 (n.s.)	0.5	/	/	EZ = PMG (1)	Resection (1)	/	Neurological examination; Neuropsychological assessment	SF (1)	66 months (1)
Chassoux et al., 2008	3 (2 M; 1F)	22; 28; 19	31; 46; 33	9; 18; 14	EZ ≈ PMG (3)	Resection (3)	Contralateral superior quadrantopsia (1); contralateral inferior quadrantopsia worsening (1)	Neurological examination; MRI	SF (3)	24 months (1)
Chugani et al., 2011	3 (n.s.)	/	7.9	/	/	Resection (3)	/	Engel	Engel I (2); Engel IV (1)	104.4 months (±21.6) (3)
Chun‐Wei Chang et al., 2019	2 (F)	1; 12	23; 19	22; 7	EZ ≈ PMG (1); EZ ≠ PMG (1)	Cortical stimulation (2)	/	ECoG; Neurological examination	DS (2)	60 months (1); 36 months (1)
Cossu et al., 2016	24 (13F; 11 M)	5.6 (±5.7)	20.1 (±12.8)	14.4 (±11.7)	EZ = PMG (5); EZ ≠ PMG (11); n.s. (8)	Resection (24)	Left‐sided hemiparesis (1); tension pneumocephalus (1); transient neurologic deficits (11); permanent neurologic deficits (5)	Engel; Neurological examination; EEG; MRI; Neuropsychological assessment	Engel I (17); Engel II (5); Engel III (1); Engel IV (1) + improvement of cognitive performance (5)	66.5 months (±54.0) (24)
Dan‐Dan Wang et al., 2014	4 (3 M; 1F)	1; 0.6; 2; 0.6	35; 24; 11; 11	34; 23.4; 9; 10.4	/	Resection (3); hemyspherectomy (1)	/	Engel	Engel I (2); Engel III (1); n.s. (1)	14.04 months (2); 5.04 months (1)
Doris Wang et al., 2014	3 (1 M; 2F)	/	2; 2; 0.25	/	/	Hemispherectomy (3)	Left hemiparesis + aphasia + oropharyngeal dysfunction (1); right hemiparesis (1); right mild hemiparesis (1)	Engel; Neurological examinations	Engel I (3)	3.6 months (2); 6 months (1)
Elliott et al., 2022	1 (M)	5	16	11	EZ ≈ PMG (1)	NeuroPace RNS (1)	/	MRI; XR; NeuroPace device	DS (1)	22 months (1)
Ferri et al., 2023	1 (M)	18	31	13	EZ = PMG (1)	RF‐TC (1)	Increase in the frequency/intensity of seizures, change in semiology and psychotic episodes (1)	QOLIE‐31 questionnaire; EEG	WS; After CBD introduction: DS + improvement psychotic episodes, cognitive and work performance; quality of life (51.9/100 Vs. 38.4/100) (1).	12 months (1)
Fohlen et al., 2019	18 (6F; 12 M)	2.06 (±1.22)	7.29 (±1.79)	5.23 (±2.14)	EZ = PMG (18)	Hemispherotomy (17); callosotomy (1)	/	Neuropsychological assessment; Engel; EEG	Engel I (16); Engel II (2) + improvement in neuropsychological testing of at least 10 points (10).	153.6 months (6–247.2) (18)
Fohlen et al., 2022	3 (M)	0.5; 1.5; 1	8.5; 11; 5	8; 9.5; 4	EZ ≠ PMG (3)	Hemispherotomy (1); targeted disconnection ‐ endoscopic HH disconnection (2)	Development of obesity (1)	Engel	Engel I (2); Engel III (1) + cognitive improvement (1)	36 months (1); 60 months (1); 120 months (1)
Isik et al., 2007	1 (M)	0.33	5	4.67	EZ ≈ PMG (1)	Resection (1)	/	Neuropsychological assessment; Neurological examination	DS + improvement in language skills, attention span and eye contact (1)	12 months (1)
Iwasaki et al., 2012	1 (M)	0.33	14	13.67	/	Callosotomy (1)	/	MRI; EEG; Neurological examination	DS (1)	9.96 months (1)
Jalloh et al., 2018	12 (8 M; 4F)	7.2 (0.2–9.1)	8.8 (0.5–17.8)	/	EZ = PMG (3); EZ ≈ PMG (2); n.s. (7)	Resection (7); hemispherectomy (5)	Hemiparesis (7); upper limb weakness + visual field defect (2); hemiparesis+ facial weakness (1); visual field defect (2)	ILAE; MRI; Neurological examination	ILAE I (9); ILAE IV (1); ILAE V (2)	37.68 months (±32.76) (12)
Jansen et al., 2005	6 (4F; 2 M)	9; 19; 12; 0.42; 9; 0.25	38; 42; 40; 29; 51; 40	29; 23; 28; 28.58; 42; 39.75	/	Resection (2); callosotomy (4)	/	Neuropsychological assessment; neurological examination	SF (3); DS (3)	/
Jeong et al., 2017	4 (2 M; 2F)	3 (±0.71)	8.15 (±2.64)	5.15 (±2.69)	EZ ≈ PMG (3); n.s. (1)	Hemispherotomy (4)	/	Engel; EEG; neuropsychological assessment	Engel I (4) + reported significant speech/language improvement (1); IQ improvement at WPPSI‐III and VABS (1); reported academic improvements (1); aphasia (1)	31.2 months (1); 73.2 months (1); 55.2 months (1); 21.6 months (1)
Kawai et al., 2004	1 (F)	8	15	7	EZ ≈ PMG (1)	Callosotomy (1)	/	Parental survey; neurological examination	SF + improved daily function and social contact (1)	51.6 months (1)
Kuo et al., 2023	3 (1F; 2 M)	1; 0.08; 0.08	19	/	EZ = PMG (2); EZ ≈ PMG (1)	Callosotomy (2); VNS + callosotomy (1)	/	EEG; MRI	Callosotomy: short‐term DS (2); callosotomy + VNS: longer DS (1)	/
Kuzniecky et al., 1997	2 (F)	6; 6	/	/	EZ ≈ PMG (1); EZ = PMG (1)	Resection (2)	Incomplete homonimous quadrantic defect (1)	Engel; neurological examinations	Engel II (1); Engel III (1)	/
Leblanc et al., 1991	2 (M)	19; 13	24; 37	5; 24	EZ ≈ PMG (2)	Resection (2)	/	ECoG; neurological examination; neuropsychological assessment	DS (2) + no change in neuropsychological and neurological assessment (2)	12 months (1); 18 months (1)
Liu et al., 2019	1 (F)	19	27	8	EZ ≈ PMG (1)	RF‐TC (1)	/	MRI; Engel	Engel II (1)	3 + 6 + 12 months (1)
Maillard & Ramantani, 2018	1 (F)	18	27	9	EZ ≈ PMG (1)	Resection (1)	/	n.s.	SF (1)	96 months (1)
Maillard et al., 2017	39 (22 M; 17F)	9.1 (±7.3)	25.4 (±13)	16.5 (±11.4)	/	Resection (34); hemispherotomy (5)	/	sEEG; MRI; neurological examination; Engel	Engel I (28); Engel II (6); Engel III (4); Engel IV (1)	55.2 months (SD n.s.) (39)
Maton et al., 2010	3 (F)	0.25; 0.9; 0.08	11; 20; 0.33	10.75; 19.1; 0.25	EZ ≈ PMG (1); EZ ≠ PMG (1); n.s. (1)	Hemispherectomy (2); resection (1)	/	Engel; MRI	Engel II (1); Engel III (1); Engel IV (1)	36 months (1); 120 months (1); 36 months (1)
Moletto et al., 2018	1 (F)	/	5	/	/	Hemispherotomy (1)	/	Engel; neuropsychological assessment	Engel I (1)	144 months (1)
Mullatti et al., 2019	1 (M)	32	44	12	EZ = PMG (1)	RF‐TC (1)	Left arm hypoesthesia (1)	Neurological examination; neuropsychological assessment; Engel; MRI	Engel II (1)	60 months (1)
Nagai et al., 2016	1 (M)	1	2.75	1.75	EZ ≈ PMG (1)	Hemispherotomy	/	n.s.	SF (1)	>12 months (1)
Nishibayashi et al., 2007	1 (M)	13	24	11	EZ ≈ PMG (1)	Resection (1)	Transient hemiparesis (1)	EEG; neurological examination	DS (1)	24 months (1)
Otsuki et al., 2013	5 (n.s.)	/	/	/	/	Resection (2); disconnection (2); hemispherotomy (1)	/	ILAE	ILAE I (1); n.s. (4)	/
Podkorytova et al., 2016	2 (F)	0.2; 0.08	6.75; 1.8	6.55; 1.72	EZ ≈ PMG (2)	Hemispherectomy (1); resection (1)	/	EEG; neurological examination	SF (1); DS (1)	42 months (1); 6 months (1)
RamachandranNair et al., 2006	1 (M)	2	9	7	EZ ≈ PMG (1)	Resection (1)	Left hemiplegia (1)	Neurological examination	SF (1)	6 months (1)
RamachandranNair et al., 2008	1 (n.s.)	2.5	9	6.5	EZ ≈ PMG (1)	Resection (1)	/	Neurological examination	SF (1)	29.04 months
Ramantani et al., 2013	4 (2 M; 2F)	18; 12; 6; 13	27; 36; 44; 21	9; 24; 38; 8	EZ ≈ PMG (2); EZ ≠ PMG (2)	Resection (4)	Contralateral homonymous hemianopia (2)	Engel; neurological examination	Engel I (3); Engel II (1)	72 months (3); 36 + 60 (1)
Ravindra et al., 2023	1 (M)	/	2.8	/	/	Hemispherectomy (1)	/	Engel; EEG; DTI	Engel I (1)	/
Säisänen et al., 2010	1 (M)	/	5	/	/	n.s.	/	Engel	Engel III (1)	12 months (1)
Saito et al., 2009	1 (M)	0.08	3	2.92	EZ = PMG (1)	Callosotomy + hemispherectomy (1)	/	Neurological examination	SF (1)	7 months (1)
Sculier et al., 2021	5 (3 M; 2F)	4.18 (±3.99)	11.48 (±3.38)	7.30 (±3.40)	EZ = PMG (1); EZ ≈ PMG (4)	Targeted Disconnection – right fronto‐insular/frontal disconnection (2); Resection (2); Callosotomy (1)	Vascular injury during surgery, motor deficit (1)	Engel; EEG; neurological examination	Engel I (2); Engel III (1); Engel IV (2) + improvement motor autonomy and QoL (1).	32.6 months (7–98) (5)
Shamim et al., 2021	2 (1F; 1 M)	13; 21	27; 37	14; 16	EZ = PMG (2)	Radiofrequency Ablation (2)	Worsening naming performance (1)	Engel; neuropsychological assessment	Engel I (2) + improved memory (2)	35 months (1); 30 months (1)
Urbach et al., 2004	1 (F)	/	22	/	EZ = PMG (1)	Resection (1)	/	Engel	Engel III (1)	24 months (1)
Wang et al., 2016	12 (7 M; 5F)	5.5 (0–26)	13.1 (0.25–44)	/	/	Hemispherectomy (5); resection (7)	Mild right hemiparesis (1); Right arm>leg weakness+mild motor and cognitive delay (1); Mild cognitive, social, emotional delay (1); Left homonymous hemianopsia (1); Mild left hand apraxia (1); Left arm>leg weakness + language delay (1); right hemiparesis+ homonymous hemianopsia (1); Left arm>leg weakness, oromotor dysfunction, gross developmental delay, aphasia (1); very mild right hemiparesis (1)	Engel; neurological examination; neurpsychological assessment	Engel I (6); Engle II (1); Engel III (3); Engel IV (2)	84 months (SD n.s.) (12)
Wichert‐Ana et al., 2008	12 (6 M; 6F)	4.21 (±3.47)	24.08 (±8.85)	19.88 (±5.37)	EZ ≈ PMG (8); EZ = PMG (4)	Resection (7); hemispherotomy (1); callosotomy (3); resection + callosotomy (1)	/	Engel	Engel I (3); Engel III (5); Engel IV (4)	58.44 months (±27.72) (12)
Zhang et al., 2020	1 (M)	0.16	5	4.84	EZ ≠ PMG (1)	Resection (1)	/	Neurological examination; MRI	SF (1)	6 months (1)
Zsoter et al., 2012	4 (4 M)	3.7 (0.17–4)	6.9 (3–9)	/	EZ ≈ PMG (4)	Hemispherotomy (4)	/	Neurological examination	SF (4)	24–72 months (4)

Abbreviations: DS, decreased seizures; DTI, diffusion tensor imaging; ECoG, electrocorticography; EEG, electroencephalogram; EZ = PMG, correspondence; EZ ≠ PMG, no correspondence; EZ ≈ PMG, partial correspondence; EZ, epileptogenic zone; HH, hypothalamic hamartoma; MRI, magnetic resonance imaging; PMG, polymicrogyria; QOLIE‐31, quality of life in epilepsy‐31; RFA, radiofrequency ablation; RF‐TC, radiofrequency thermocoagulation; RNS, responsive neurostimulation system; sEEG, stereoelectroencephalography; SF, seizure‐free; VABS, vineland adaptive behavior scales; VNS, vagus nerve stimulation; WPPSI‐III, Wechsler Preschool and Primary Scale of Intelligence – Third Edition; WS, worsened seizures; XR, X‐rays.

Concerning potential etiological factors, in seven patients (3.11%), specific genetic mutations known to be related to PMG were reported. Perinatal events were presented in 15 (6.67%) patients (prolonged labor, precipitous delivery, fetal distress, threatened abortion, dystocic delivery), while only two studies mentioned congenital CMV infections associated with PMG. For the remaining patients, no data were reported.

In a total of 123 out of 225 patients (54.67%), the relationship between the EZ and the area affected by PMG was retrieved. A total of 46 out of 47 (97.87%) studies reported preoperative evaluation using EEG (33/46; 71.74%), MRI (45/46; 97.83%), and SEEG (13/46; 28.26%), to localize the EZ. The other study did not report any evaluation. Specifically, in 47 cases out of 123 (38.21%), the EZ corresponded to the PMG; in 54 cases (43.9%) there was a partial correspondence, while in 22 cases (17.89%) the EZ did not correspond to PMG.

Among the proposed treatments, a total of 132 out of 225 (58.67%) cases underwent resection or lobectomy, 56 (24.89%) hemispheric disconnection, 18 (8%) callosotomy, nine (4%) SEEG‐guided ablation, six (2.67%) targeted disconnection (nonhemispheric), two (0.89%) cortical stimulation, and one (0.44%) was treated with the Responsive Neurostimulation System (NeuroPace; RNS). In one case, the type of treatment was not specified.

The median follow‐up time in the included studies was 33.1 months (range 3.6–153.6; IQR: 12.5–60.0), providing insight into the long‐term trends of the examined outcomes.

3.1. Outcome/effectiveness of treatment

A total of 165 out of 225 patients (73.33%) received an Engel classification as outcome evaluation. Specifically, 63.64% (105/165) of these patients achieved an Engel Class I (FDS), 15.15% (25/165) Engel Class II, 13.94% (23/165) Engel Class III, and 7.27% (12/165) Engel Class IV. Additionally, another 10 patients were reported to have an ILAE Class I classification, one had Class IV, and two had Class V. There were no patients who were reported to have an ILAE II classification. Overall, considering the combined Engel and ILAE outcomes, 64.61% (115/178) of patients achieved seizure freedom.

When stratifying outcomes by type of surgical procedure (Table 3), among patients for whom both a postsurgical Engel/ILAE classification and the type of surgical procedure were reported (161/225; 71.55%), resection or lobectomy (119/161; 73.91%) showed Engel Class I/ILAE I outcomes in 73/119 patients (61.34%). Hemispheric disconnection procedures (hemispherotomy or hemispherectomy, n = 44/161; 27.33%) achieved Engel Class I in 37/44 patients (84.09%). Among callosotomy cases (6/161; 3.73%), seizure freedom (Engel I) was not reported, with outcomes distributed as Engel II–IV. SEEG‐guided ablation (4/161; 2.48%) resulted in Engel I in two patients (50%) and Engel II in another two (50%). Targeted disconnection (4/161; 2.48%) achieved Engel I in three cases (75%), and Engel IV in one case (25%).

TABLE 3.

Engel/ILAE class and surgical treatment.

TREATMENTS	ENGEL class				ILAE class
	I	II	III	IV	Class (No. patients)
Resection/lobectomy a	63	20	17	6	I (10); IV (1); V (2)
Hemispheric Disconnection (Hemispherotomy + Hemispherectomy b )	37 (27 + 10)	2 (1 + 1)	3 (1 + 2)	2 (0 + 2)	‐
Callosotomy	‐	1	2	3	‐
SEEG‐guided ablation (RF‐TC + RFA)	2 (0 + 2)	2 (2 + 0)	‐	‐	‐
Targeted Disconnection	3	‐	‐	1	‐
Not. specif.	‐	‐	1	‐	‐

Abbreviations: RF‐TC: radiofrequency thermocoagulation; RFA: radiofrequency ablation.

^a In this subgroup are included two patients, underwent to two different surgical intervention. Only the last intervention was considered for the statistical analysis.

^b In this subgroup are included four patients, underwent to two different surgical intervention. Only the last intervention was considered for the statistical analysis.

Among the remaining patients (47/225; 20.89%) who did not receive an Engel/ILAE classification but were clinically evaluated by the authors, 21/47 (44.68%) were declared seizure‐free, 21/47 (44.68%) experienced a reduction in the frequency/severity of seizures, and one (2.13%) exhibited a worsening of seizures associated with psychotic episodes.

Furthermore, when considering only the patients with a reported correspondence between EZ and PMG and whose surgical outcome was assessed using the Engel/ILAE classification (35/225; 15.56%), it was found that 65.71% (23/35) of them achieved Engel Class I.

Information on postsurgery drug management was provided for 49.78% (112/225) of the patients. Among them, 30.36% (34/112) discontinued treatment with AEDs, 41.96% (47/112) tapered treatment, 25% (28/112) maintained treatment, while only 2.68% (3/112) increased the dosage/number of drugs. As given in Table 4, only 68/112 (60.71%) of these patients received an Engel/ILAE classification. Among them, 35.29% (24/68) had an Engel/ILAE class I and discontinued the pharmacological treatment.

TABLE 4.

Postsurgery outcomes and AEDs.

Engel/ILAE class	Discontinued (%)	Tapered (%)	Unchanged (%)	Increased (%)
Engel I/ILAE I	24 (35.29)	17 (25.0)	8 (11.76)	3 (4.41)
Engel II	1 (1.47)	2 (2.94)	5 (7.35)	0 (0.0)
Engel III	0 (0.0)	1 (1.47)	1 (1.47)	0 (0.0)
Engel IV/ILAE IV‐V	1 (1.47)	2 (2.94)	2 (2.94)	1 (1.47)
Total (N. 68)	26 (38.24)	22 (32.35)	16 (23.53)	4 (5.88)

Abbreviation: AEDs: Antiepileptic drugs.

Furthermore, as expected, the Chi‐squared test showed a statistically significant association between discontinuation of AEDs and favorable surgical outcome, with seizure freedom occurring more frequently among patients who were able to withdraw medication. Moreover, the data suggest that age at surgery (adult vs. pediatric) may be related to surgical outcomes, although it did not reach conventional levels of statistical significance (p = 0.09). Finally, the correspondence between the EZ and PMG did not show any significant relationship (Table 5).

TABLE 5.

Associations between Engel classes and patient age at surgery (<18 and >18 years), correspondence between (EZ/PMG), and postsurgery AEDs administration.

	Engel I/ILAE Class I–II (N)	Engel II–IV/ILAE Class III–V (N)	χ2	df	p‐value
Age
<18	16	10	2.88	1	0.09
>18	9	15
EZ/PMG
EZ = PMG	23	12	0.94	2	0.62
EZ≈PMG	8	7
EZ≠PMG	13	6
Postsurgery AEDs
Discontinued	24	2	5.87	1	0.01
Ongoing	28	14

Abbreviations: AEDs, antiepileptic drugs; Engel/ILAE class was dichotomised as Class 1 vs. Classes II–IV; EZ, epileptogenic zone; PMG, polymicrogyria. Bold values indicate statistically significant results (p < 0.05).

In addition, statistical analyses showed that patients in Engel Class I had a significantly shorter epilepsy duration prior to surgery compared with those in classes II–IV [Mann–Whitney U = 253, p = 0.027 (rank‐biserial correlation = 0.306)] (Figure 2).

FIGURE 2.

Boxplot of epilepsy duration according to Engel outcome classes (I vs. II–IV).

Among all surgically treated patients, only 76/225 (33.78%) reportedly received a neuropsychological assessment, encompassing cognitive abilities (i.e., IQ, memory, attention, language, etc.), behavioral and social functioning, and quality of life. Of these, only 23/76 (30.26%) reported improvement in cognitive performance, one in behavioral skills, two in social abilities, and two in overall quality of life. For the remaining patients, no data were reported. Among patients who reported cognitive improvement after surgery, the median age at onset was 2.53 years (0.0–21.0; IQR: 0.63–7.4), at surgery was 9.75 years (3.50–37.0; IQR: 7.07–19.33), and the duration of epilepsy was 7.15 years (3.50–16.0; IQR: 5.04–13.75).

Notably, only two studies disclosed the specific neuropsychological assessment test used for postsurgical evaluation. In general, a comprehensive pre‐ and post‐neuropsychological evaluation was reported in only four studies.

As detailed in Table 2, several postoperative complications were observed among patients, including an increase in the frequency/severity of seizures and changes in seizure semiology. Additionally, some patients experienced worsening in cognitive, language, and behavioral skills, motor deficits, and the occurrence of psychotic episodes. Hemiparesis was a relatively common complication, ranging from mild‐to‐severe, and often accompanied by additional neurological deficits such as visual field defects (e.g., contralateral homonymous hemianopia and contralateral quadrantopsia) or facial/body weakness. Other complications included hemiplegia, tension pneumocephalus, apraxia, oropharyngeal dysfunction, and hypoesthesia.

4. DISCUSSION

4.1. Presurgery evaluation

The importance of preoperative investigations in epilepsy surgery is widely recognized.^{61, 62, 63}

In our systematic review, almost all included studies (46/47, 97.87%) performed instrumental preoperative investigations, while only a minority (12/47; 25.53%) performed a preoperative neuropsychological assessment.

As the ILAE Pediatric Epilepsy Surgery Task Force emphasizes, ⁶⁴ a multidisciplinary team should be involved in this evaluation. The limited administration of preoperative neuropsychological assessments raises concerns about the comprehensive understanding of cognitive and behavioral functions and potential preexisting deficits in these patients, as well as the potential impact of surgery on their outcomes.

Another crucial aspect in surgical decisions concerns the correspondence between the EZ and the PMG region, especially if they involve eloquent brain areas.

While in a significant number of cases a clear alignment between the EZ and PMG was demonstrated, these data were not systematically reported. It is worth noting that identifying the EZ can be a complex task, and it does not always correspond with the PMG. Interestingly, our findings closely align with the study conducted by Leventer et al., ⁶⁵ who reported 78% (vs. 82.11% in our data) of correspondence between EZ and PMG in their cases, supporting the frequent association between PMG and epileptogenicity. On the other hand, this indicates that in ~ 20% of cases the PMG does not coincide with the EZ, underlining the ambiguity of their relationship. For instance, Maillard et al. ⁵² stressed that the EZ may only partially overlap with the PMG, emphasizing that the extent of PMG should not necessarily deter consideration of epilepsy surgery.

Concerning the tools that should be selected for the preoperative decision‐making process, apart from routine presurgical EEG and MRI assessments, some studies highlighted the necessity of intracranial electroencephalography (iEEG), such as SEEG, subdural explorations (SDE), and electrostimulation mapping (EM) to inform surgical decisions and better delineate the EZ.^{20, 23, 31, 40, 54, 56, 58, 60}

Expanding on these insights, Wang et al. ⁵³ explored the potential of less invasive intracranial electrocorticography (ECoG) and fMRI in guiding surgical resections, particularly for patients with extensive or bilateral PMG malformations. In addition to fMRI, Zsoter et al. and Säisänen et al.^{41, 43} discuss the utility of transcranial magnetic stimulation (TMS) in evaluating sensorimotor function for guiding surgical planning. Nagai et al. ²² highlighted the utility of DTI in distinguishing epileptogenic networks, while Wichert‐Ana et al. ⁵⁷ suggested that SPECT may serve as a valuable tool in the evaluation and prognosis of epilepsy surgery for PMG. Lastly, few studies have suggested the use of magnetoencephalography (MEG).^{15, 25, 31}

4.2. Timing of surgical treatments

The gap between the onset of epileptic seizures (median age: 4.21 years) and surgical intervention (median age: 15.0 years) is wide, resulting in an average delay of ~ 13 years. This delay in surgical treatment is concerning as our analysis shows that the longer duration of epilepsy surgery is associated with worse postoperative outcome. ⁶⁶ This delay may be attributed to various factors, including diagnostic challenges, attempts at pharmacological treatment before considering surgical options, and potentially delays in accessing specialized services for epilepsy surgery.

Therefore, the timing of surgical intervention is a critical factor in optimizing the outcomes and seizure control. ⁶⁶

4.3. Surgical treatments and seizures outcomes

Our systematic review reveals that resection and hemispheric disconnection were the most widely applied surgical techniques. These choices are clearly influenced by the extent and the localization of the underlying epileptogenic lesion. ⁶⁷ Recent research papers included in this review have highlighted promising and innovative options for minimally invasive interventions and multimodal treatment approaches. Ravindra et al. ¹⁵ demonstrated the feasibility of magnetic resonance‐guided laser interstitial thermal ablation (MRgLITT) for hemispherotomy. Furthermore, Kuo et al. ²⁷ emphasized the benefits of combining callosotomy with vagus nerve stimulation (VNS), showcasing the potential of a multimodal treatment approach in specific epilepsy cases. Liu et al. ¹⁹ demonstrated the benefits of SEEG‐guided radiofrequency thermocoagulation (RF‐TC) as a therapeutic option for DRE. Additionally, Elliott et al. ¹⁷ underscored the advantages of the NeuroPace Responsive Neurostimulation system in tailoring treatment options, especially when dealing with eloquent cortical areas. Other methods are the open‐loop electrical cortical stimulation, described by Chang et al., ⁵⁹ and the network radiofrequency ablation (RFA) reported by Shamim et al. ³⁰ as a viable option for patients with focal epilepsy and PMG.

Considering the classical approaches to epilepsy surgery, Fohlen et al. ²⁹ emphasize the potential advantages of parasagittal hemispherotomy (PSH) in patients with extensive unilateral PMG. They stress the importance of considering PSH as an option when patients present with a preexisting motor deficit and cognitive or behavioral decline. Also, Jeong et al. ³³ provide further insight into the benefits of hemispherotomy, particularly in the context of refractory electrical status epilepticus during slow sleep (ESES) and structural lesions like PMG. Caraballo et al. ³⁵ focused on early resective surgery as a management option to prevent cognitive deterioration in patients with ESES. Finally, Kawai et al. ⁴⁸ present their findings on total corpus callosotomy, emphasizing its ability to halt drop attacks and preserve neuropsychological function in patients with bilateral PMG.

Overall, the fact that 64.61% of patients who received an Engel/ILAE classification achieved a Class I outcome suggests that surgical interventions can be highly effective in a substantial proportion of cases.

However, given that 95.75% of the included studies were case reports or small case series, the observed proportion of favorable outcomes (e.g., seizure freedom; Engel I/ILAE I–II) is likely inflated by publication and selective reporting bias; accordingly, these data should be interpreted as descriptive and hypothesis‐generating rather than as estimates of treatment effect.

Altogether, the presence of patients who have not achieved seizure freedom highlights the variability in treatment outcomes. Given the heterogeneity of the collected data, it proved challenging to elucidate the underlying reasons for this variability. It could have been influenced by several factors such as the extent of PMG, the precise location of the EZ, the timing of intervention, and individual patient characteristics.

Comparatively, epilepsy surgery outcomes reported for other structural etiologies vary. In focal cortical dysplasia (FCD) type II, ~ 67% achieve an Engel Class I outcome and in tuberous sclerosis complex (TSC) cohorts the figure is ~ 56%. These differences are likely driven by etiology and disease‐specific factors; therefore, cross‐etiology comparisons should be interpreted cautiously.^{68, 69, 70, 71}

Notably, based on statistical analysis, the prior duration of epilepsy appears to influence the outcome of epilepsy surgery. A possible explanation may stem from the cumulative impact of prolonged epileptic activity: a lengthier epilepsy duration suggests a chronic and potentially refractory course, leading to pronounced structural and functional changes in the brain.^{72, 73} Overall, this finding suggests that prompt and thorough presurgical evaluation should be scheduled for individuals who are considered suitable candidates for epilepsy surgery. ⁷⁴

Finally, it seems that a significant percentage of patients who are seizure‐free after surgery may not require AEDs. Considering other series reported, Kwon et al. ⁷⁵ described 40 patients with isolated mild MCD, of whom 36.8% received an ILAE Class I and discontinued all AEDs after surgical resection, which is similar to the proportion reported in our findings (35.29%).

4.4. Postsurgery complications

Although epilepsy surgery results in substantial improvement in patients' lives, postsurgery complications represent a crucial aspect of epilepsy management.

Most postsurgical complications were consistent with the type of surgery and the location of the procedure in the brain, except for two cases that exhibited unexpected complications. In their study, Ferri et al. ¹⁶ reported a remarkable case involving a patient who experienced an increase in the frequency and intensity of seizures after surgery. Surprisingly, this patient even presented with postsurgical psychotic episodes successfully resolved through the administration of cannabidiol. Moreover, Wang et al. ⁵³ reported unexpected cognitive, emotional, and social deficits in a patient following the resection of a right occipital epilepsy foci. Apart from these two cases of unexpected adverse events, in general surgery, resulted in expected complications. Studies included in this review^{30, 33, 53} highlighted the development of cognitive and language deficits (e.g., naming, memory, and attentional skills), behavioral, social, and emotional delays, pointing to the potential impact of surgery on individual functions. Furthermore, two studies^{31, 38} found a spectrum of motor deficits, with hemiparesis affecting 10 cases, upper limb weakness combined with visual field defects in two cases, and facial weakness in one case. Cossu et al. ⁵⁴ documented the persistence of left‐sided hemiparesis and reported transient neurologic deficits in 11 patients, with permanent neurologic deficits affecting five individuals. In their study, Mullatti et al. ⁵¹ reported left arm hypoesthesia, drawing attention to sensory disturbances complications. Several authors highlight visual complications following surgery, such as visual field defects, ³¹ hemianopia, ³⁹ quadrantopsia, ⁴⁶ and incomplete quadrantic defect. ⁴⁹

4.5. Neuropsychological functioning and quality of life after surgery

Surgical interventions have the potential to ameliorate cognitive and behavioral functioning,^{76, 77} thereby improving the overall quality of life. However, neuropsychological outcomes after surgery in PMG‐related epilepsy are poorly described. Regrettably, few included studies have incorporated neuropsychological and quality‐of‐life assessments into their postsurgical evaluations. Standardized instruments to gauge these critical aspects have rarely been employed. Moreover, due to the heterogeneity of the cohorts and the limited availability of data, it was not possible in our analyses to distinguish between pediatric and adult patients.

Some of the retrieved studies^{29, 40, 54} have reported a substantial improvement in cognitive performance in a significant portion of their patients, with no instances of cognitive decline or language deficits observed postsurgery.

As has previously been shown, the degree of cognitive improvement is closely linked to the effective control of seizures. ⁷⁶ For instance, Isik et al. ²³ documented remarkable improvements in language skills, vocabulary, attention span, and eye contact in a patient who experienced a more than 90% reduction in seizures after surgery. Moletto et al. ²¹ reported a unique case where verbal skills shifted from the right hemisphere to the contralateral side, demonstrating an intriguing adaptation of brain connections in response to surgery. Furthermore, surgical interventions can also yield positive outcomes in terms of behavioral and social skills. In the 2004 study by Kawai et al. ⁴⁸ one of their patients became seizure‐free following surgery and exhibited substantial improvements in their daily functioning and social interactions. Overall, our data suggest that patients who have cognitive improvement after surgery had a lower median age at epilepsy onset (2.53 vs. 4.21 years), shorter duration of epilepsy (7.15 vs. 9.95 years), and lower age at time of surgery (9.75 vs. 15.0 years). The latter finding is consistent with the research by Liu et., al. ⁷⁷ which showed that significant improvements in postoperative IQ scores were associated with an age at surgery of ≤10 years (p < 0.01), in patients with TSC and associated epilepsy. This may suggest that early surgical intervention in younger patients may play a critical role in improving cognitive outcomes, possibly due to more effective neuroplasticity supporting neurodevelopment in this cohort of patients. However, these findings seem to be in contrast with the work of Arfaie et al., ⁷⁸ which did not report any significant association with cognitive outcomes and age at epilepsy onset, surgery, and interval between them. As highlighted by the study of Skirrow et al., ⁷⁹ such findings underline the importance of further investigation considering age as a factor when assessing the potential for cognitive improvement following epilepsy surgery.

In conclusion, the absence of postsurgical neuropsychological assessments, together with the fact that the specific tests used were rarely reported when such evaluations were conducted, substantially limits the transparency and consistency of outcome assessment across the retrieved studies.

4.6. Limitations of the study

A notable limitation of this review is the overall low quality of the studies included. There was a preponderance of case reports and case series. Consequently, the review is particularly susceptible to publication bias and small‐study effects that preferentially highlight favorable or atypical outcomes.

This problem also limits the generalizability of the findings. The potential for selection bias and confounding variables to exert undue influence on the reported outcomes is high. This limitation prevented us from performing a meta‐analysis across the studies.

Outcome classification heterogeneity is another limitation: several studies reported only aggregate Engel Class I (without subclasses distinction e.g., IA/IB), while others used ILAE categories. This inconsistency limits comparability across studies. Therefore, we harmonized outcomes as freedom from disabling seizures (FDS), defined as Engel I or ILAE I–II. This approach minimizes selective data loss but may slightly overestimate the proportion with complete seizure freedom and limits outcome analyses.

Neuropsychological outcomes were rarely and inconsistently assessed (33.78%), with heterogeneous or unspecified tests, variable timing, and scarce baseline–postoperative pairs. Consequently, cognitive and health‐related quality‐of‐life effects cannot be reliably inferred, and any observed changes should be treated as indicative information rather than effect estimates. Moreover, it was not possible to analyze adult and pediatric populations separately with respect to neuropsychological and quality‐of‐life outcomes. This represents a limitation as these domains can differ substantially due to developmental and age‐related factors.

There is a possibility that not all relevant cases in the literature were retrieved using the selected keywords, which could introduce a bias in the sample of studies considered.

Finally, only a small number of studies report on whether there is concordance of the EZ and PMG.

These limitations underscore the importance of interpreting the findings from this review with caution and highlight the need to conduct meticulously designed studies and RCTs to definitively clarify the efficacy of surgery for PMG‐related epilepsy.

5. CONCLUSION AND FUTURE DIRECTIONS

This systematic review underscores the complexity of surgical decision‐making in the context of epilepsy associated with PMG.

The reviewed studies reveal that resection and hemispheric surgery are commonly employed surgical techniques, with the choice influenced by the extent and location of the epileptogenic lesion. The data indicate that many patients are reported to achieve favorable outcomes, with a substantial proportion experiencing Engel Class I seizure control, and most of them discontinuing medication. However, the variability in outcomes and evaluation tools used is notable. Despite the potential benefits of epilepsy surgery, the quality of life and neuropsychological outcomes of patients are not consistently evaluated across the included studies, making comprehensive assessment of outcomes challenging.

Future research should prioritize prospective, multicenter studies with standardized protocols for pre‐ and postsurgical evaluation, including cognitive, behavioral, and quality of life measures. Shared clinical guidelines and advanced neuroimaging and neurophysiological tools may improve surgical planning and patient selection. Long‐term follow‐up is also crucial to evaluate sustained benefits and late‐onset complications.

FUNDING INFORMATION

No financial assistance was received in support of this article.

CONFLICT OF INTEREST STATEMENT

None of the authors have any conflict of interest to disclose.

ETHICS APPROVAL STATEMENT

We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

DATA AVAILABILITY STATEMENT

The authors confirm that the data supporting the findings of this study are available from the corresponding author upon reasonable request.