SEARCH

SEARCH BY CITATION

Keywords:

  • Cortical dysplasia;
  • Epilepsy surgery;
  • MRI;
  • EEG;
  • Neuropsychological assessment

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

Purpose:  Prenatal and perinatal adverse events are reported to have a pathogenetic role in focal cortical dysplasia (FCD). However, no data are available regarding the prevalence and significance of this association. A cohort of children with significant prenatal and perinatal brain injury and histologically proven mild malformations of cortical development (mMCD) or FCD was analyzed.

Methods:  We retrospectively evaluated a surgical series of 200 patients with histologically confirmed mMCD/FCD. Combined historical and radiologic inclusion criteria were used to identify patients with prenatal and perinatal risk factors. Electroclinical, imaging, neuropsychological, surgical, histopathologic, and seizure outcome data were reviewed.

Results:  Prenatal and perinatal insults including severe prematurity, asphyxia, bleeding, hydrocephalus, and stroke occurred in 12.5% of children with mMCD/FCD (n = 25). Their epilepsy was characterized by early seizure onset, high seizure frequency, and absence of seizure control. Patients with significant prenatal and perinatal risk factors had more abnormal neurologic findings, lower intelligence quotient (IQ) scores, and slower background EEG activity than mMCD/FCD subjects without prenatal or perinatal brain injury. MRI evidence of cortical malformations was identified in 74% of patients. Most patients underwent large multilobar resections or hemispherectomies; 54% were seizure-free 2 years after surgery. Histologically “milder” forms of cortical malformations (mMCD and FCD type I) were observed most commonly in our series.

Conclusions:  Surgically remediable low-grade cortical malformations may occur in children with significant prenatally and perinatally acquired encephalopathies and play an important role in the pathogenesis of their epilepsy. Presurgical detection of dysplastic cortex has important practical consequences for surgical planning.

Focal cortical dysplasia (FCD) and mild malformation of cortical development (mMCD) are common causes of focal intractable epilepsy, cognitive disability, and other neurologic disorders. mMCD and FCD are frequent etiologies in pediatric and adult epilepsy surgery series (Harvey et al., 2008; Lerner et al., 2009; Blümcke et al., 2009).

The etiopathogenesis of mMCD/FCD has not been satisfactorily clarified. Prenatal and perinatal events are suspected to be important in the pathogenesis of different types of malformations of cortical development (MCD) (Palmini et al., 1994; Rakic, 2000; Schwartzkroin & Walsh, 2000; Crino et al., 2002). However, only anecdotal reports and small patient series have identified mMCD/FCD in epileptic patients with significant prenatal and perinatal brain injury (Wyllie et al., 1996; Kremer et al., 2002; Govaert et al., 2006). There are no data regarding the prevalence and significance of this association.

Previous studies analyzed prenatal and perinatal factors in various MCDs including lissencephaly, polymicrogyria, hemimegalencephaly, and nodular heterotopia (Palmini et al., 1994; Raymond et al., 1995; Montenegro et al., 2002, 2005). Other studies focused only on the relationship of histopathology to etiopathogenetic factors (Marin-Padilla, 1996, 1997, 1999, 2000). To our knowledge, there has been no systematic study of the clinical features of patients with mMCD/FCD and associated prenatally and perinatally acquired encephalopathies.

The present study aims to analyze a cohort of children with significant prenatal and perinatal insults and histologically proven mMCD/FCD. We sought to identify distinctive characteristics of epileptic and neurologic syndromes in these patients in order to achieve earlier diagnosis and enhance surgical management.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

Patient selection

Data of patients with intractable epilepsy who underwent resective epilepsy surgery at the Miami Children’s Hospital between 1986 and 2006 were retrospectively reviewed. We selected patients who had (1) a histologically proven diagnosis of mMCD/FCD and (2) unequivocal history of a prenatal and perinatal brain insult. Rigorous combined inclusion criteria based on both medical histories and the presence of typical radiologic findings suggestive of prenatal and perinatal insults were used for identification of prenatal and perinatal adverse events. Patients’ data were compared with a control group of 30 children operated on for intractable epilepsy caused by mMCD/FCD, but without any known prenatal and perinatal insults. The controls had the same proportions of individual histologic mMCD/FCD subtypes as the patient group.

Neuropathologic analysis and classification

Brain tissue analysis was performed at the Department of Pathology, Miami Children’s Hospital, Miami, Florida (1986–2003) and at the Department of Pathology and Laboratory Medicine (Neuropathology), David Geffen School of Medicine at University of California, Los Angeles, California (2003–2006). The same neuropathologic methodology was used at both institutions and has been reported in detail previously (Mischel et al., 1995; Lawson et al., 2005). All neuropathologic findings were reclassified by a neuropathologist (HV) and two experienced epileptologists (PK, MD) according to the current classification scheme (Palmini et al., 2004; Hildebrandt et al., 2005). Only definite features of cortical malformation distinct from consequences of prenatal and perinatal insults were accepted in our series. Subjects with malformations of cortical development other than FCD/mMCD (such as tuberous sclerosis complex, polymicrogyria, nodular heterotopia, or hemimegaencephaly) were excluded.

Prenatal/perinatal risk factor assessment

Pregnancy, delivery, and postnatal histories of all subjects were obtained from obstetrical and birth records and analyzed in detail. Only unequivocal prenatal and perinatal risk factors were accepted. Perinatal asphyxia was diagnosed according to the standard criteria of Sarnat and Sarnat (Finer et al., 1981). We excluded subjects with questionable historical adverse events including mild prematurity, maternal infection, or drug ingestion during pregnancy if their significance was not supported by radiologic findings in the child.

MRI evaluation

MRI scans of 23 subjects imaged at 1.5-Tesla including fluid attenuated inversion recovery (FLAIR) sequences were rigorously reevaluated. We searched for both radiologic consequences of prenatal and perinatal insults and changes typical of dysplastic lesions. MRI data were reevaluated independently by two experienced investigators (PK, BM) who were aware of the diagnosis but who were not informed about patient history or histologic subtype. Discrepancies between reviewers led to a case being rereviewed together until a consensus was reached. If disagreement remained, the MRI feature in question was omitted from the analysis.

Two children did not have good-quality MRI data (Nos 18 and 22). One was imaged on a 0.3-Tesla scanner and the other could not undergo MRI because of a metallic pellet in the skull. In both cases the available neuroimaging revealed unequivocal signs of prenatal and perinatal injury, but we did not analyze features of mMCD/FCD.

The following features were considered to be reliable markers of prenatal and perinatal adverse events: encephalomalatic lesions, periventricular leukomalacia, hydrocephalus with VP shunt, ventriculomegaly, cortical atrophy, and atrophy of the corpus callosum. Nonspecific gliosis was not considered a typical consequence of prenatal or perinatal risk factors.

We also retrospectively identified possible changes caused by the presence of cortical malformations. We used a protocol of MRI evaluation described previously (Krsek et al., 2008, 2009b). The following typical MRI characteristics of mMCD/FCD were searched for: the presence of increased cortical thickness, transmantle sign, gray–white matter junction blurring, abnormal gyral or sulcal patterns, lobar hypoplasia, and gray and white matter signal change in individual MR sequences (Colombo et al., 2009). Reviewers tried to distinguish changes of cortical malformations from the above-mentioned consequences of prenatal and perinatal adverse events.

Other presurgical evaluation

A complete history, a neurologic examination, and preoperative scalp video-EEG (electroencephalography) were evaluated in all patients. Interictal epileptiform discharges and ictal EEG patterns were classified as regional (appearing exclusively over a single lobe or in two contiguous regions such as centroparietal discharges), multilobar, hemispheric, or generalized. Selected children also underwent additional diagnostic tests including positron emission tomography (PET) and single photon emission computed tomography (SPECT); evaluation of these results was not included in the present study.

Neuropsychological evaluation

Presurgical neuropsychological data for reevaluation were available in 18 patients. Intellectual test results and/or overall adaptive functioning questionnaires of the patients were reviewed by two independent experts in the psychological assessment of children with neurocognitive disorders (BK and GR). Global functional ranking was determined because heterogeneous neuropsychological batteries had been utilized for assessment and because subjects with pervasive intellectual or developmental impairments could not be examined with common psychometric instruments. The following categories of the ranking based on Standard Scores (SS) were distinguished: SS 1: Moderate to Severe Impairment, IQ <59; SS 2: Mild Impairment, IQ 60–69; SS 3: Borderline Intelligence, IQ 70–79; SS 4: Low Average Intelligence and Above, IQ >80. The remaining seven cases had no quantitative data available for this ranking process.

Surgical procedures and outcome

All patients underwent resective surgery for intractable epilepsy at the Department of Neurological Surgery, Miami Children’s Hospital, Miami, Florida. Type and extent of surgery were evaluated.

Postoperative seizure outcome 2 years after the final surgery was analyzed. Surgical outcome was classified according to Engel’s classification scheme: (I) completely seizure-free, auras only or only atypical early postoperative seizures; (II) ≥90% seizure reduction or nocturnal seizures only; (III) ≥50% seizure reduction; and (IV) <50% seizure reduction.

Statistical analysis

Twelve clinical parameters (e.g., age at seizure onset, seizure frequency, duration of epilepsy, incidence of infantile spasms, status epilepticus and abnormal neurologic finding, neuropsychological ranking, EEG and surgery characteristics, and seizure outcome at 2 years) were compared between two groups of mMCD/FCD patients with (n = 25) and without (n = 30) prenatal and perinatal insults. Investigated parameters consisted of three interval variables and nine categorical variables. Categorical parameters were analyzed with contingency (cross-tabulation) table analysis. Statistical comparisons between the groups were evaluated with Pearson chi-square statistic, maximum-likelihood chi-square test, Spearman R (rank order correlation) statistic, and (in the cases of two-way tables) also with Fisher’s exact test. Interval variables were compared with t-test for independent samples.

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

From a population of 567 children undergoing resection at the Miami Children’s Hospital from March 1986 to June 2006, 200 subjects had a histologic diagnosis of mMCD/FCD. Twenty-five patients (12.5%; 9 females, 16 males) had a definite history of prenatal or perinatal risk factors according to the above-mentioned combined historical and radiologic criteria and were included in the study.

Prenatal and perinatal insults

Prenatal and perinatal adverse events in patients are summarized at Table 1. Unequivocal risk factors (frequently in combinations) were identified in the medical histories of 22 patients. The most prevalent adverse events in our series were prematurity (n = 8, 32%), perinatal asphyxia (n = 9, 36%), intraventricular hemorrhage (n = 6, 24%), hydrocephalus with ventriculoperitoneal (VP) shunt (n = 7, 28%), and prenatal stroke (n = 4, 16%). Other risk factors included congenital heart defects (n = 2), neonatal seizures (n = 3), VP shunt infections (n = 2), and sepsis (n = 1). In all cases except three (Nos 4, 11, and 23), the insults was confirmed by computed tomography (CT)/MRI findings. Two cases without an apparent MRI prenatal and perinatal lesion were born premature and had mild perinatal asphyxia,; one was born at term, but had severe asphyxia due to a congenital heart defect.

Table 1.   Clinical data and neurologic and neuropsychological findings in patients
Case/sexKnown prenatal and perinatal insultsNeurologic findingsAge at seizure onsetSeizure frequencyNeuropsychological ranking
  1. M, male; F, female; R, right; L, left; Prem., prematurity; w, week; mo, month; y, year; IVH, intraventricular hemorrhage; IUGR, intrauterine growth restriction; DM, diabetes mellitus; NA, not available.

1/F0Normal6 moDailyIQ 60–69
2/FPrem. 27 w, IVH, hydrocephalus, VP shunt+infectionsMicrocephaly, spastic diparesis3 yWeeklyIQ ≤ 59
3/MBorn at term, L prenatal infarctR hemiparesis1 wDailyNA
4/MPrem. 32 w, mild perinatal asphyxiaNormal1.5 yDailyIQ 60–69
5/M0R hemiparesis1 moDailyIQ ≤ 59
6/FBorn at term, perinatal asphyxiaR hemiparesis10 yDailyIQ 60–69
7/FPrem. 25 w, IVH, hydrocephalus, VP shunt+infectionsNormal3 moDailyIQ ≤ 59
8/F0Normal3 moDailyIQ ≤ 59
9/MPrem. 32 w, IVH, hydrocephalus, VP shunt, neonatal seizuresR hemiparesis, R visual field cut4 yDailyIQ ≤ 59
10/FIUGR, born at term, perinatal asphyxiaL hemiparesis, microcephaly7 moDailyIQ ≤ 59
11/MBorn at term, perinatal asphyxia due to congenital heart defectNormal1 yMonthlyIQ 60–69
12/FBorn at term, bilateral prenatal infarct, preeclampsia, meconium aspirationMicrocephaly, quadriparesis more on the R1 wDailyIQ ≤ 59
13/MBorn at term, R prenatal infarctL hemiparesis, L visual field cut1 moDailyNA
14/MBorn at term, perinatal asphyxia, neonatal seizuresNormal8 yWeeklyIQ 70–79
15/FBorn at term, perinatal asphyxia, hypoglycemia, failure to thriveNormal2 yDailyIQ ≤ 59
16/MBorn at term, perinatal asphyxia, meconium aspirationMicrocephaly, spastic diparesis6 moDailyIQ ≤ 59
17/MBorn at term, congenital heart defect, R prenatal infarct, IVH, hydrocephalus, VP shuntL hemiparesis2 wDailyIQ 60–69
18/FBorn at term, L prenatal infarctR hemiparesis11 moDailyNA
19/MTriplet, prem. 26 w, IVH, necrotizing enterocolitis, sepsisMicrocephaly, spastic diparesis6 moDailyIQ ≤ 59
20/MBorn at term, congenital hydrocephalusSpastic diparesis2 yDailyIQ ≥ 80
21/MPrem. 26 w, IVH, hydrocephalus, VP shuntNormal6 moDailyIQ 60–69
22/MBorn at term, congenital hydrocephalusR hemiparesis8 moDailyNA
23/MPrem. 33 w, mild perinatal asphyxiaNormal1 yDailyNA
24/M0Microcephaly, R hemiparesis,1 wDailyNA
25/MGestational DM, prem. 34 w, perinatal asphyxia, neonatal seizuresNormal3 yDailyNA

Neurologic and neuropsychological findings

Data concerning neurologic and neuropsychological findings are shown in Table 1. Only nine subjects (36%) had normal neurologic findings. Most frequent neurologic deficits included hemiparesis (10 of 25, 40%), spastic diparesis (4 of 25, 16%), and microcephaly (6 of 25, 24%). One child with bilateral encephalomalacic lesions was quadriparetic.

Neuropsychologic data were available for 18 patients. More than half (10 of 18, 56%) evidenced severe mental retardation. Mild intellectual impairment (IQ 60–69) was found in 6 of 18 patients (33%) and borderline intelligence (IQ 70–79) in one patient. Only one subject had normal intelligence (IQ >80). Several specific developmental disorders such as behavioral disorders, pervasive developmental disorders, or developmental language disorders were reported in some cases (data not shown).

Epilepsy and EEG characteristics

Mean age at seizure onset was 1.6 years (range 1 week to 10 years). Twenty-two patients (88%) had daily seizures. West syndrome occurred in three cases (12%) and status epilepticus in seven (28%). No subject evidenced prolonged periods of seizure control.

All children had abnormal scalp EEG. Slow background activity was present in 15 cases (60%), either as intermittent or continuous focal slowing in 20 subjects (80%). Interictal epileptiform activity was present in all patients and was classified as regional in 4 (16%), multiregional in 17 (68%), and hemispheric in 4 (16%) subjects. Continuous epileptiform activity (EEG status epilepticus) was found in two patients. Only seven cases (28%) had regional ictal EEG patterns. Multiregional or hemispheric ictal EEG patterns occurred in 11 cases (44%). The remaining seven patients (28%) had nonlateralized or multiple ictal patterns.

MRI findings

Neuroimaging data are provided in Table 2. Neuroimaging data to evaluate the consequences of prenatal and perinatal insults were available in all patients. They were identified in 22 cases (88%). The most frequent findings included encephalomalacia in 12 (48%), periventricular leukomalacia in 9 (36%), hydrocephalus in 6 (24%), ventricular dilation in 2 (8%), cortical atrophy in 4 (16%), and corpus callosum atrophy in 3 (12%) subjects. Four children (Patients 1, 5, 8, and 24) had no prenatal or perinatal risk factor in their medical records but evidenced clear radiologic proof of prenatal or perinatal insults (encephalomalacia in three, periventricular leukomalacia in two, and cortical atrophy in one subject).

Table 2.   MRI findings in patients
Case/sexConsequences of prenatal and perinatal insultsFindings suggestive of mMCD/FCD and HS
  1. PVL, periventricular leukomalacia; CC, corpus callosum; G/W mat. junct. blur., gray–white matter junction blurring; sig. ch., signal change; HS, hippocampal sclerosis .

1/FEncephalomalacia, cortical atrophy0
2/FPVL, shunted hydrocephalus, CC, and cortical atrophy0
3/MEncephalomalacia, porencephalyG/W mat. junct. blur.; lobar hypoplasia; WM-FLAIR, T2W sig. ch.
4/M0Lobar hypoplasia; WM-FLAIR, T2W sig. ch.
5/MEncephalomalacia, porencephalyLobar hypoplasia; HS
6/FEncephalomalacia, CC atrophyLobar hypoplasia
7/FPVL, shunted hydrocephalusG/W mat. junct. blur.; GM-FLAIR, T2W sig. ch.; WM-FLAIR, T2W, T1W sig. ch.
8/FPVL0
9/MEncephalomalacia, shunted hydrocephalusG/W mat. junct. blur.; lobar hypoplasia, HS
10/FPVL, right ventricle dilatationG/W mat. junct. blur.; lobar hypoplasia; WM-FLAIR, T2W, T1W sig. ch.; HS
11/M0G/W mat. junct. blur.; WM-FLAIR, T2W, T1W sig. ch.; HS
12/FEncephalomalaciaG/W mat. junct. blur.; lobar hypoplasia; WM-FLAIR, T2W, T1W sig. ch.
13/MEncephalomalacia, porencephaly, PVL, cortical atrophyG/W mat. junct. blur.; lobar hypoplasia; WM-FLAIR, T2W, T1W sig. ch.
14/MBilateral encephalomalacia and porencephaly0
15/FPVLG/W mat. junct. blur.; lobar hypoplasia; GM-FLAIR sig. ch.; WM-FLAIR, T2W, T1W sig. ch.; HS
16/MEncephalomalaciaLobar hypoplasia
17/MEncephalomalacia, porencephaly, CC atrophyG/W mat. junct. blur.; lobar hypoplasia; GM-FLAIR, T2W sig. ch; WM-FLAIR, T2W, T1W sig. ch.; HS
18/FEncephalomalacia (CT scan only)NA
19/MPVL, ventriculomegaly0
20/MShunted hydrocephalusG/W mat. junct. blur.; lobar hypoplasia; GM-FLAIR, T2W sig. ch.; WM-FLAIR, T2W, T1W sig. ch.
21/MPVL, shunted hydrocephalus0
22/MShunted hydrocephalus (0.3 Tesla MRI)NA
23/M0G/W mat. junct. blur.; lobar hypoplasia; +GM-FLAIR sig. ch.; WM-FLAIR, T2W, T1W sig. ch.
24/MEncephalomalacia, PVLG/W mat. junct. blur.; WM-T2W, T1W sig. ch.
25/MCortical atrophyGM-FLAIR sig. ch.;

Changes suspicious of cortical malformations were revealed in 17 of 23 cases (74%) with available high-resolution MRI scans: gray–white matter junction blurring in 12 (52%), lobar hypoplasia or atrophy in 13 (56%), gray matter signal change on FLAIR sequences in seven (30%) or on T2-weighted (T2W) in three (13%), and white matter signal changes in FLAIR, T2W, and T1 in 14 (61%), 13 (56%), and 10 (43%) cases, respectively. Typical MRI features of FCD type II such as increased cortical thickness, transmantle sign, and abnormal gyral or sulcal patterns were absent from our series. Three patients without MRI changes typical for prenatal or perinatal insults had MRI abnormalities typical for cortical malformations that were similar to those of other cases in the group. Hippocampal atrophy and signal changes characteristic of hippocampal sclerosis were observed in six patients (Patients 5, 9, 10, 11, 15, and 17). Regarding the extent of changes suspicious of mMCD/FCD, three cases had unilobar, five patients multilobar, and the remaining nine had hemispheric abnormality.

Surgery, histology, and outcome

Details about the surgical procedures, histology, and outcome are given in Table 3. The mean age at surgery was 9.85 years (range 1–25.6 years). The average duration of epilepsy was 8.1 years (range 1–17.6 years). Fifteen patients underwent one-stage excisional procedures. All except two hemispherectomy cases had intraoperative electrocorticography. Ten subjects underwent chronic invasive monitoring with implanted subdural electrodes. Eight cases had unilobar resections, seven had multilobar resections, and 10 had hemispherectomy. Three children required reoperation.

Table 3.   Surgical data, histology, and postsurgical seizure outcome in patients
Case/sexAge at surgeryType of surgeryHistologySeizure outcome at 2 years
  1. Surgical outcome was classified according to Engel’s classification scheme: Engel I, seizure-free, auras only; Engel II, >90% seizure reduction; Engel III, >50% seizure reduction; Engel IV, <50% seizure reduction.

  2. L, left; R, right; F, frontal; T, temporal; P, parietal; O, occipital; SZ, seizure.

1/F14.7L F corticectomymMCDIV
2/F5.6R TPO corticectomymMCDII
3/M2L functional hemispherectomymMCDI
4/M12R F corticectomymMCDIII
5/M20.3L functional hemispherectomymMCDI
6/F15.3, 16.6L FP corticectomy, L PO corticectomyFCD IaI
7/F7.9R T lobectomyFCD IaNA
8/F9.8L F lobectomy and callosotomyFCD IaII
9/M10.9L functional hemispherectomyFCD IaI
10/F5R functional hemispherectomyFCD IaIV
11/M16.2L functional hemispherectomyFCD IaIII
12/F7.2R functional hemispherectomyFCD IaI
13/M1.6R T and orbitofrontal resectionsFCD IaI
14/M25.6L F corticectomyFCD IaIII
15/F10.6, 11.1L T lobectomy, L TO lobectomyFCD IaNA
16/M13.5R F corticectomyFCD IaIV
17/M4.7R functional hemispherectomyFCD IaNA
18/F16.4L functional hemispherectomyFCD IaI
19/M7.2L TPO corticectomyFCD IbI
20/M3.3, 9.8R TPO corticectomy, R functional hemispherectomyFCD IbI
21/M17.8R F corticectomyFCD IbI
22/M7.7L T lobectomyFCD IbI
23/M3.9L functional hemispherectomyFCD IbIV
24/M1L functional hemispherectomyFCD IIbI
25/M3.8L FP corticectomyFCD IIbIV

mMCD type II was histologically proven in 5 patients (20%), FCD type Ia in 13 patients (52%), FCD type Ib in 5 patients (20%), and FCD type IIb was present in 2 subjects (8%).

Seizure outcome data 2 years from the last surgery was available for 22 subjects. Seizure freedom was achieved in 12 subjects (54%). Engel class IV outcome was encountered in five cases (23%).

Comparison with mMCD/FCD patients without prenatal and perinatal brain injury

Comparison of selected parameters of mMCD/FCD patients with and without prenatal and perinatal insults is depicted in Table 4. Null hypothesis declaring no difference between the two groups was refused at a highly significant level (greater than p = 0.0255 within all statistical tests used) for parameters “Incidence of abnormal neurological finding,”“Neuropsychological ranking,”“Incidence of slow background EEG activity,” and “Extent of resection”—that is, subjects with prenatal and perinatal risk factors had more abnormal neurologic findings, lower IQ scores, higher incidence of slow background EEG activity, and greater likelihood of a larger resection. There was no statistical significance for other parameters.

Table 4.   Comparison of findings in patients with mMCD/FCD with and without prenatal and perinatal risk factors
 mMCD/FCD with prenatal and perinatal insults (%)mMCD/FCD without prenatal and perinatal insults (%)
  1. *p < 0.05.

  2. Note that not all patients had available neuropsychological data and seizure outcome.

Mean age at seizure onset1.632.99
Infantile spasms 3/25 (12) 3/30 (10)
Status epilepticus 7/25 (28) 8/30 (27)
Daily seizures22/25 (88)21/30 (70)
Abnormal neurologic finding17/25 (77) 6/30 (20)*
Mental retardation (IQ <70)16/18 (89) 7/22 (32)*
Slow background EEG activity15/25 (68) 9/30 (30)*
Focal EEG slowing20/25 (80)22/30 (73)
Hippocampal sclerosis 6/25 (24) 7/30 (23)
Mean age at surgery9.7610
Mean duration of epilepsy8.137.15
Resection (one lobe-multilobar-hemispherectomy)8 (32) – 7 (28) – 10 (40)21 (70) – 7 (24) – 2 (8)*
Seizure-free seizure outcome12/22 (55)18/30 (60)

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

Our study, conducted on the largest population of pediatric patients with histologically proven mMCD/FCD, demonstrated a frequent association with significant prenatal and perinatal risk factors. We further showed that children with acquired prenatal or perinatal brain lesions associated with cortical malformations benefit from epilepsy surgery. The size of the dataset, all collected at one institution, facilitated the identification of distinct clinical features that may assist the surgical management of these patients.

There is consensus that the pathogenesis of mMCD/FCD involves events taking place during human corticoneurogenesis. The development of the human cerebral cortex can be divided into three overlapping stages: (1) proliferation of stem cells into neuroblasts or glial cells; (2) migration from the periventricular germinal matrix to the developing cortex; (3) cortical organization of the six-layer neocortex associated with synaptogenesis and apoptosis (Raymond et al., 1995; Barkovich et al., 1996, 2005). Regarding the timing of these events, the proliferative stage is believed to range from the 5th or 6th until the 16th or 20th gestational week, migration from the 6th or 7th until the 20th or 24th gestational week, and organization from the 16th until approximately the 24th gestational week (Rakic, 2000). However, there is growing evidence that some migration and organization may occur during the third trimester of pregnancy (Cepeda et al., 2006).

It is a matter of ongoing debate as to when developmental alterations accounting for individual types of MCD might occur. Although previous studies assumed these events take place before the 24th week of intrauterine life (Barkovich et al., 1996; Rakic, 2000), recent observations suggest that much later insults may be important. It was proposed that the presence of abnormal cell types in dysplastic tissue could be explained by the failure of prenatal cell degeneration before birth. According to this hypothesis, histologically severe types of FCD such as type II are caused by alterations in the late second or early third trimester, whereas events occurring closer to birth account for milder forms of cortical malformations such as FCD type I and mMCD (Cepeda et al., 2006). A recent report described neuropathologic changes typical for FCD in two infants who survived a severe brain injury (shaken baby syndrome) shortly after birth (Marin-Padilla et al., 2002). These observations lend support to the hypothesis that besides prenatal insults, perinatal insults may also play a causal role in the pathogenesis of mMCD/FCD.

The migration of neuroblasts to their final destination and their organization within the cortical mantle may be disturbed by different genetic and environmental factors. The latter are more frequently suspected in mMCD/FCD, since there is no description of familial cases of these malformations, except for those associated with specific syndromes, such as tuberous sclerosis complex. Observations from experimental models and anecdotal reports suggest the possible importance of several candidate extraneous insults such as toxins (Baraban & Schwartzkroin, 1995), intrauterine infections (Iannetti et al., 1998), and ionizing radiation (Marin-Padilla et al., 2003).

In the clinical setting, the relevance in the pregnancy history of events potentially harmful to the ultimate cortical malformation remains speculative. Prenatal and perinatal risk factors were identified in a series of 40 patients with different types of MCD (22 had FCD). Potentially harmful prenatal environmental events were reported in 58% of cases; however, they included questionable factors such as maternal bicornuate uterus, ingestion of laxative during the first trimester, or maternal hyperglycemia during pregnancy. The significance of the insults was not verified by MRI or histologic findings (Palmini et al., 1994). Other reports of patients with cortical malformations found the following incidence of prenatal and perinatal risk factors: 32% in MCD (Raymond et al., 1995), 29% in FCD (Widdess-Walsh et al., 2005), and 10% in FCD (Krsek et al., 2009a). Adverse events were reported less frequently in the histories of patients with FCD than in those with other types of MCD (Montenegro et al., 2002). However, the types of insults were usually not specified.

We verified the significance of prenatal and perinatal insults using strictly combined clinical and radiologic criteria and verified the diagnosis of mMCD/FCD by direct histopathologic confirmation. Hypoxic–ischemic and hemorrhagic brain lesions were frequently associated with shunted hydrocephalus in our patients. The impact of these insults on cortical development was neuropathologically studied in children who survived perinatally acquired encephalopathies (Marin-Padilla, 1996, 1997, 1999). It was demonstrated that intrauterine insults could produce cytoarchitectural alterations of the developing neocortex that eventually give rise to neuropathologic findings compatible with “acquired” cortical dysplasia. These postinjury alterations represent dynamic processes that begin after the brain injury and continue to evolve over weeks, months, or even years after the original insult. It was also suggested that if a child eventually develops epilepsy, further secondary alterations might be caused by the seizures themselves (Marin-Padilla, 2000).

A second possible explanation for the association between cortical malformations and prenatal and perinatal risk factors is that MCD itself predisposes to perinatal distress. Montenegro et al. (2005) reported that patients with cortical malformations frequently experience intrapartum complications. It was hypothesized that MCD could predispose to decreased fetal movements due to preexisting neurologic impairment during pregnancy that in turn predispose to intrapartum complications. However, their series included mostly severe types of MCD such as lissencephaly, schizencephaly, and polymicrogyria. The majority of mMCD/FCD cases have an uncomplicated birth, normal neurologic findings, and normal early psychomotor development (Widdess-Walsh et al., 2005; Krsek et al., 2008; Lerner et al., 2009). In our series, prenatal and perinatal adverse events were significantly associated with histopathologically “milder” forms of cortical malformation (mMCD and FCD type I) than FCD type II, in accord with our previous observations (Krsek et al., 2009a). Findings of FCD type II in the setting of severe prenatal and perinatal brain damage are anecdotal (Wyllie et al., 1996; Kremer et al., 2002). We suggest that our results support the above-mentioned hypothesis that mMCD/FCD might be caused by the third trimester and perinatal adverse events. Notwithstanding, the possibility that the presence of some mMCD/FCD predisposes to perinatal distress cannot be excluded.

In addition to the high frequency of abnormal neurologic findings and mental retardation, clinical syndromes in our series were characterized by severe pharmacoresistant epilepsy with very early seizure onset, high seizure frequency, and absence of long periods of seizure control. We proved that children with significant prenatal and perinatal risk factors exhibited a higher frequency of abnormal neurologic findings, lower IQ scores, and higher incidence of slow EEG background activity compared to patients with mMCD/FCD without known prenatal and perinatal brain injury. These observations could be explained by the presence of widespread cortical and subcortical brain injury. However, our patients had the same epilepsy syndrome characteristics as control subjects. These results might suggest that cortical malformations represent the pathologic substrate of epilepsy that is independent of prenatally and perinatally acquired encephalopathy.

The incidence of epilepsy in patients with cerebral palsy has been reported to range from 15–41.8% (Steffenburg et al., 1995; Hadjipanayis et al., 1997). Seizures could be controlled by antiepileptic drugs in approximately 60% of cases (Singhi et al., 2003), but catastrophic drug-resistant cases referred for epilepsy surgery are well recognized (Battaglia et al., 2005; Guzzetta et al., 2006; Oguni et al., 2008). There are only a few anecdotal reports attributing the pathogenesis of drug-resistant epilepsy in these children to cortical malformations, and these are usually different from mMCD/FCD: polymicrogyria (Bordarier & Robain, 1992), focal pachygyria (Watanabe et al., 1990), and neuronal heterotopias (Reutens et al., 1993). Our results suggest that mMCD/FCD could play an important role in the genesis of epilepsy in children with acquired neonatal encephalopathies. Prospective studies including comparisons with a control group of subjects with perinatally acquired encephalopathy without associated cortical malformation should provide further information about the role of mMCD/FCD in the pathogenesis of epilepsy and neurologic deficits in these children.

The preoperative identification of mMCD/FCD in children with significant prenatally or perinatally acquired brain damage is highly challenging. Retrospective MRI review in our population consistently revealed lobar atrophy, blurring of the gray–white matter junction, increased signal intensities on T2W and FLAIR sequences, predominating in the white matter always ipsilateral to the resection site. Although nonspecific, these radiologic findings are suggestive of cortical malformations (Palmini et al., 2004; Widdess-Walsh et al., 2006). They were recently recognized as distinctive radiologic features of mMCD and FCD type I (Krsek et al., 2008, 2009a; Colombo et al., 2009), and are consistent with our pathologic results. However, similar radiologic findings in the absence of histopathologic examination could be interpreted as a consequence of ischemic insults. We speculate that since the MRI features of cortical malformations could easily be overlooked or misinterpreted in these patients, the true incidence of mMCD/FCD in children with acquired neonatal encephalopathies might be considerably underestimated.

Our observation that cortical malformations associate with severe prenatal and perinatal brain injury has practical consequences for surgical planning. The dysplastic cortex is likely to be at least a part of the epileptogenic zone and its complete resection critical for obtaining a seizure-free outcome (Paolicchi et al., 2000; Krsek et al., 2009b; Lerner et al., 2009). Presurgical delineation of associated cortical malformations in patients with prenatally and perinatally acquired brain lesions may, therefore, considerably influence the type and extent of the resection. Because MRI detection of the dysplastic cortex is not always possible, we suggest that intracranial EEG (either intraoperative electrocorticography or chronically implanted intracranial electrodes) may be required to assess the extent of dysplastic cortical areas. In our series, all cases except for two hemispherectomy cases had at least intraoperative electrocorticography, and 10 subjects underwent chronic invasive monitoring using implanted subdural electrodes. The standard criteria for evaluating intracranial EEG data were used (Jayakar et al., 1994; Krsek et al., 2009b). Although mMCD/FCD was detected retrospectively in the majority of our subjects, use of a definition of completeness of surgery based on careful evaluation of both MRI and intracranial EEG data enabled us to achieve seizure freedom in 54% of the patients.

Our patients underwent larger resections than subjects with isolated mMCD/FCD, but the ultimate postoperative seizure outcome of both groups was comparable. This result lends support to indications for epilepsy surgery in children with cortical malformations in the setting of severe prenatal and perinatal brain injury. Further studies are necessary to optimize diagnostic and surgical approaches in these subjects. We hope our observation might promote an interest in the association between prenatally and perinatally acquired brain lesions and cortical malformations as well as initiate multicenter cooperation that could ultimately improve prognosis of individual patients.

Acknowledgments

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

Supported by grants MZOFNM2005 and Kontakt Program ME09042. HV Vinters supported by the Daljit S. & Elaine Sarkaria Chair in Diagnostic Medicine. Statistical analysis was conducted by Zbynek Hrncir, MSc.

Disclosure

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

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

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References