Paroxysmal kinesigenic dyskinesia and infantile convulsions

Article abstract—Objective: To clinically characterize affected individuals in families with paroxysmal kinesigenic dyskinesia (PKD), examine the association with infantile convulsions, and confirm linkage to a pericentromeric chromosome 16 locus. Background: PKD is characterized by frequent, recurrent attacks of involuntary movement or posturing in response to sudden movement, stress, or excitement. Recently, an autosomal dominant PKD locus on chromosome 16 was identified. Methods: The authors studied 11 previously unreported families of diverse ethnic background with PKD with or without infantile convulsions and performed linkage analysis with markers spanning the chromosome 16 locus. Detailed clinical questionnaires and interviews were conducted with affected and unaffected family members. Results: Clinical characterization and sampling of 95 individuals in 11 families revealed 44 individuals with paroxysmal dyskinesia, infantile convulsions, or both. Infantile convulsions were surprisingly common, occurring in 9 of 11 families. In only two individuals did generalized seizures occur in later childhood or adulthood. The authors defined a 26-cM region using linkage data in 11 families (maximum lod score 6.63 at U 5 0). Affected individuals in one family showed no evidence for a shared haplotype in this region, implying locus heterogeneity. Conclusions: Identification and characterization of the PKD/infantile convulsions gene will provide new insight into the pathophysiology of this disorder, which spans the phenotypic spectrum between epilepsy and movement disorder.

cific abnormalities, with rare exceptions. 14 Neuroimaging studies are unremarkable, and two autopsies have shown no neuropathologic abnormalities. 7,15 Accumulating evidence supports altered function or excitability of basal ganglia and related structures. 13,16 This entity is to be distinguished from paroxysmal nonkinesigenic dyskinesia (OMIM 118800, locus 2q), [17][18] paroxysmal choreoathetosis with episodic ataxia and spasticity (OMIM 601042, locus 1p), 19 paroxysmal exertional dyskinesia, 11,[20][21] and paroxysmal hypnogenic dyskinesia. 11,[22][23][24] Szepetowski et al. first reported an association of paroxysmal choreoathetosis with infantile convulsions (IC) in four French families, designating it the ICCA syndrome (IC with choreoathetosis; OMIM 602066). 25 Affected individuals had IC and spontaneous dystonia or dystonia on exertion; however, the movement disorder was not clearly kinesigenic. They demonstrated linkage to a pericentromeric chromosome 16 locus. A Chinese family with IC and a kinesigenic movement disorder was subsequently reported. 26 Recently, Tomita et al. confirmed linkage of paroxysmal kinesigenic choreoathetosis to this locus. 27 IC were present in six of eight Japanese families. Linkage was also demonstrated in an African American family without IC. 28 Thus, clinical and linkage data suggest that PKD is a disorder with age-dependent, variable phenotypic expression of abnormal movements as well as generalized or partial seizures.
We report clinical and linkage data in 11 families with PKD. Ethnic background is diverse and includes two African American families, a Taiwanese family, an Ashkenazi Jewish family, a Dutch family, a Mexican American family, and five families of mixed European ancestry including English, German, Yugoslavian, Irish, and Polish heritage. IC were present in nine families, and a history of generalized seizures persisting to adulthood was present in two individuals inheriting the at-risk haplotype. These families had 51 affected individuals (kinesigenic dyskinesia, IC, or both). Linkage data confirm the chromosome 16 locus as a major locus in families with PKD from a variety of ethnic groups. However, our data also suggest that locus heterogeneity is likely, as affected members of one family with classic PKD and IC fail to share a common haplotype.
Methods. Patient identification and DNA isolation. We evaluated 11 families with PKD, including 44 individuals with kinesigenic dyskinesia, IC, or both. This study was approved by the University of Utah Medical Center Institutional Review Board and informed consent was obtained from all participating family members. We collected DNA samples and clinical data on all 44 of these individuals, on an additional 33 at-risk individuals, and on 16 individuals related by marriage. Affected and unaffected family members underwent an interview and completed a questionnaire to confirm details regarding a possible movement or seizure disorder, age at onset, precipitating and alleviating factors, and family and medical history. A neurologic ex-amination was completed on the proband in each family. Medical records and EEG reports were reviewed. Venous blood samples were collected and genomic DNA was isolated from whole blood lysates using the Puregene DNA isolation kit (Gentra Systems; Minneapolis, MN).
Genotyping and linkage analysis. A locus-specific linkage analysis was performed using 14 markers spanning 30 cM across the pericentromeric region of chromosome 16. Markers were labeled fluorescently and used to amplify 50 ng of template DNA with standard PCR procedures of the Genomics Core Facility, Huntsman Cancer Institute, University of Utah School of Medicine. Genotypes were visualized using a PE Biosystems model 373S and analyzed by the Genotyper peak-calling software. Computer-generated results were checked by one of the authors (K.J.S.). Kindreds 3323, 3534, 4962, and 3446 had had previous linkage studies to exclude the 2q locus associated with the nonkinesigenic form of the disorder (data not shown).
Diagnostic criteria. For linkage analysis, individuals were definitely affected if they had movement-related dystonia, choreoathetosis, or other involuntary movement or posturing. They were unknown if they had only IC, were younger than 20 years, or reported nonspecific symptoms such as leg cramping, muscle spasm, chest tightness, or syncope. They were definitely unaffected if they were 20 years or older, had no evidence of a movement or seizure disorder by personal interview and questionnaire, a history during infancy was available, and they denied having children with a movement disorder or nonfebrile IC.
Linkage analysis. Two-point linkage analysis was performed using MLINK of the LINKAGE program, assuming an autosomal dominant mode of inheritance. Multipoint linkage analysis was performed using LINKMAP. Normal and disease allele frequencies were 0.999 and 0.001, and allele frequencies were assumed equal. Lod scores were calculated over a range of penetrance values from 70 to 95%.

Results. Clinical findings.
Of 44 individuals with dyskinesia, IC, or both, we had complete information regarding a movement disorder in 42 and IC in 37. The latter denominator reflects a lack of adequate history during infancy in some cases. Among affected individuals, 36 of 42 (86%) had kinesigenic dyskinesia, 23 of 37 (62%) had one or more IC, and 16 of 32 (50%) had a combined phenotype. Penetrance for individuals at risk by haplotype for either PKD or IC, including only those in whom we had complete clinical information, was 84% in women (26/31) and 86% in men (19/22).
Episodes of IC were brief and spontaneous and remitted by age 3 years in the absence of treatment. Onset was between 3 and 18 months of age. They did not occur in association with fever. In cases where information was available, EEG within the 24 hours following an event was normal or showed nonspecific slowing. Episodes were associated with altered consciousness, staring, upward eye deviation, or apnea with or without unilateral or bilateral tonic stiffening. Duration varied from a few seconds to 1 to 2 minutes. In one case, episodes occurred during attempts to have a bowel movement, especially if the infant was constipated. Prolonged crying may have been a trigger for episodes in a second infant. A clear history of a kinesigenic trigger generally could not be elicited; however, history was remote from these events for most individuals for whom this information was obtained. Gelastic-type epi-sodes (staring, followed by smiling and giggling, or sometimes frank laughter) were documented in one 6-month-old girl with video-EEG monitoring (figure 1, K5471:34712). EEG recording during events was without change. The number of IC varied from a single episode to several dozen. In contrast to the movement disorder, recurrent episodes in a single day were unusual. Anticonvulsant therapy when instituted was deemed beneficial, but episodes were usually infrequent and rather benign even in the absence of treatment. None had prolonged seizure activity or required intubation. Only four adults with a history of IC denied subsequent development of dyskinesia; two of these individuals reported nonspecific symptoms such as foot cramping or back spasm. Febrile convulsions (FC) were reported in two individuals (designated FC in figure 1; K5471 and K4874). One man had a single FC as an infant with high fever and systemic illness; he clearly did not inherit the haplotype associated with a positive affection status. One woman had recurrent FC; she also appeared to inherit an alternative haplotype, although several markers in the critical region were uninformative in this individual.
Neither manifested a subsequent seizure or movement disorder.
Two individuals had a history of generalized seizures (GS) persisting into adulthood (figure 1, designated GS; K4962 and K5471). In both, onset of generalized convulsions occurred in early adolescence. Seizures were infrequent on anticonvulsant treatment. Neither individual was alive at the time of the study. Medical records were unavailable, details were limited to secondary reports, and information was lacking regarding IC in both cases.
Age at onset for episodes of paroxysmal dyskinesia was 6 to 14 (average 10.4) years in men and 6 to 23 (average 11.5) years in women. Episodes were triggered by sudden movement in all cases. Three individuals also reported episodes during physical exertion such as running or swimming; in one, onset of symptoms first occurred in the midst of a track event, during maximal physical exertion. Onset occurred within the first minute or two of exertion but was short-lived as long as exertion was curtailed. Episodes were uniformly less than 5 minutes, and the majority lasted only seconds. Stress and anxiety were precipitants in virtually all individuals; symptoms frequently occurred when affected persons were required to do a performance-related task in front of an audience.
Jumping up to answer a ringing telephone was also a commonly reported trigger. A few individuals felt that caffeine was a trigger, but the majority did not. Alcohol was neither a precipitating nor alleviating factor. Several individuals reported falling during particularly severe episodes, but consciousness was never altered. Most individuals remained symptomatic at the time of participation in the study; virtually all reported improvement with age over 25 years. All who were placed on anticonvulsant medications reported clinical benefit, with reduction in frequency and severity of episodes. Carbamazepine or phenytoin were significantly helpful in individuals in nine families; members of one family received only phenobarbital, which also ameliorated symptoms.
In three families, PKD symptoms in men seemed subjectively more severe. However, men were not more likely than women to report symptoms. Although a greater total number of women in these families had IC, prevalence of PKD was evenly distributed among men and women. There was no apparent worsening of symptoms during pregnancy, although some women reported exacerbation of symptoms around the time of their menses. A few women noted that wearing high heels made symptoms more manifest, particularly when walking across an uneven surface. Many individuals reported an aura of increased muscle tension that failed to result in a visibly apparent abnormal movement in addition to more typical episodes. Moving slowly, waiting to move, or forcibly holding themselves in a stiff posture could sometimes circumvent an episode that would otherwise have been visible. One individual with a history of IC who denied subsequent development of a movement disorder reported occasional toe-cramping during initiation of movement, while wearing heels, when anxious, and intermittently during sleep (figure 1, K5471: 35260). Another reported intermittent "back spasm" spontaneously and with sudden movement (figure 1, K4962: 35367).
EEG data were available on at least one individual with PKD in each kindred. The majority of these studies were normal. In one individual, a teenager (figure 2, K3446: 24129), 6-Hz spike and wave activity was noted; a subsequent EEG a year later on anticonvulsant therapy was normal. No other studies demonstrated evidence of epilep-

Figure 2. Paroxysmal kinesigenic dyskinesia (PKD) pedigree with and without infantile convulsions (IC). Circles represent women, squares represent men. Affection status for PKD, IC (under age 2 years), GS (generalized seizures occurring after age 2 years), and febrile convulsions (FC) is as noted. Genotypes for the microsatellite markers are listed top to bottom as shown. A rectangle surrounds the shared haplotype segregating with PKD and IC. Arrowheads indicate recombination events.
tiform activity, although occasional nonspecific abnormalities including mild slowing or occipital sharp transients were noted. Two children thought to be at risk for symptoms on the basis of haplotype analysis had a single presumed syncopal episode, but no other symptoms (figure 1, K5471:34717 and 34721). The father of two affected boys in K3391 (figure 2, 23899) had experienced episodes of "chest tightness" during exertion or sudden movement during adolescence, but denied other symptoms. These individuals were considered of unknown status for purposes of linkage and clinical analysis.
Linkage analysis. Two-point lod scores are shown in the table. A maximum lod score of 6.63 was achieved at D16S3131 at ⍜ ϭ 0, 95% penetrance. Lod scores were significant for a range of penetrance values from 70 to 95% (5.71 to 6.63; for additional information, please visit our Web site at www.neurology.org and click on the title link for this article). Multipoint lod score analysis revealed a maximum lod score of 6.60 between the closely spaced markers D16S3100 and D16S753. If linkage analysis was performed designating all those with nonfebrile IC as definitely affected, maximum two-point lod score increased to 8.11, and maximum multipoint lod score increased to 8.82. Figure 3 shows a map of the critical region on chromosome 16 delineated in our data as well as previously published data. 25,27,28 Critical recombination events were present in individual 35088 in K3323 (figure 2) and in individuals 34767 and 34720 in K5471 (figure 1), placing the disease gene in a 26-cM interval from D16S3131 to D16S3396. These individuals all had PKD symptoms, with or without IC. Recombination data in an individual with a history of a single infantile convulsion (35121, K4874, figure 1), whose mother was asymptomatic, would place the disease gene proximal to marker D16S3080. However, we did not use data from this individual to determine our critical region.
Linkage and haplotype data for K3349 are depicted in figure 1 (for additional information, please visit our Web site at www.neurology.org and click on the title link for this article). Haplotype analysis reveals no common haplotype among the three affected individuals. Because a lod score of ϽϪ2.0 is only achieved at ⍜ ϭ 0, and because these markers span a relatively large region, we cannot definitively exclude the chromosome 16 locus. Nonetheless, the data suggest that locus heterogeneity exists for this phenotype. Because several small families are included in this study, one or more of them may prove to have pathogenetic mutations at another locus.

Discussion.
Our observations confirm the relationship between nonfebrile IC and kinesigenic dyskinesia in families with autosomal dominant PKD and suggest that this association is usual rather than extraordinary. The majority of those with seizures in infancy ultimately developed the movement disorder. This supports the concept that a lowered threshold for seizure activity may occur as a mani-  25 ; and ***critical region for PKD in a single large African American family. 28 Arrowheads indicate recombination events in individuals with a single IC. festation of variable expression of a gene or related gene products during CNS development and maturation, ultimately manifesting in the characteristic movement disorder. The phenotype encompassed by the ICCA syndrome and PKD/IC syndrome are almost certainly identical.
We have confirmed linkage of the PKD/IC phenotype to the pericentromeric chromosome 16 locus, and provide evidence in support of locus heterogeneity. Our critical region, defined conservatively by using recombination data only from individuals with PKD, lies between markers D16S3131 and D16S3396, a distance of 26 cM. These data, in conjunction with those reported for the ICCA syndrome, would narrow the critical region to a 3.2-cM region spanning the centromere (see figure 3). However, these data should be viewed with caution, as the boundary on 16q is defined by a recombinant haplotype in an individual reported by Szepetowski et al. to have had only a single infantile convulsion. 25 Because seizures in infancy are not uncommon, and at least one individual among our families with infantile FC did not inherit the haplotype of interest, the certainty of these data remains unclear at present.
Our critical region shows significant overlap with that recently delineated by other authors ( figure  3). 27,28 In the larger region defined by this interval, there are several intriguing candidates, including adenylate cyclase 7, GNAO1, and CNCG3L. GNAO1 produces the alpha subunit of a guanine nucleotide binding protein, part of a large family of signal transducing molecules. Disruption of GNAO1 by homologous recombination in mice reduced opioid receptor-mediated calcium channel currents and reduced voltage threshold for calcium channel activation. 29 CNCG3L encodes the beta subunit of a cyclic nucleotide gated cation channel. Griggs and Nutt have noted that PKD is an excellent candidate for an ion channel disorder, as paroxysmal kinesigenic choreoathetosis has been observed in patients with episodic ataxia type 1 due to mutations in the potassium channel gene KCNA1 (12p). 30 Ion channels have been implicated in several other paroxysmal disorders manifesting in seizures or abnormal movements, including episodic ataxia type 2 (calcium channel mutations), 31 febrile seizures (sodium channel mutations), 32 and benign familial neonatal convulsions (potassium channel mutations). 33 Whether abnormalities in ion channel function will prove pathogenetic remains uncertain. However, identification and characterization of the PKD/IC gene will provide new insight into the pathophysiology of this interesting disorder, which spans the phenotypic spectrum between epilepsy and movement disorder. Article abstract-Objective: To assess intellectual functioning (IQ) in 54 children and adolescents with intractable epilepsy who later underwent cortical resection due to unilateral malformations of cortical development acquired in utero. Methods: Lesion type was classified into circumscribed mass lesions and diffuse cortical dysplasia based on histopathologic analysis of surgical tissue. Cortical dysplastic lesions were further graded as mild, moderate, or severe according to specific microscopic features. Laterality of lesion was determined through neurologic examination and electrophysiologic and neuroradiologic procedures. Classification of lesion type was corroborated by its significant relationship with other disease-related variables known to be related to clinical severity (age at seizure onset, age at resection, and extent of lesion). Results: Analyses of covariance revealed that circumscribed lesions had a less deleterious effect on nonverbal IQ than did diffuse cortical dysplasia, after controlling for age at seizure onset and extent of lesion. This effect was also found on verbal IQ measures, but only in subjects with right-sided lesions. Subjects with left-sided lesions performed significantly more poorly on verbal IQ measures than those with right-sided lesions. Additionally, younger age at onset and greater extent of lesion were associated with poorer cognitive outcome. Conclusions: Cortical dysplasia and early left hemisphere lesions have a significantly worse impact on cognitive functioning than circumscribed lesions or right hemisphere developmental lesions in children with epilepsy. Although there is abundant research documenting evidence of the consequences of acquired brain injury, the consequences of malformations of cortical development (MCD) are relatively unknown. The types of MCD are diverse, and neural tissue is rarely available for study. Children with intractable epilepsy who undergo surgery provide an important opportunity to investigate the relationship between lesions acquired in utero and cognitive outcome. Although studies of prenatally acquired malformations in children with epilepsy have continued to expand, [1][2][3][4][5][6][7][8][9][10][11] few studies have integrated information on clinical status with abnormal neurodevelopmental processes in a systematic manner. 12 Studies that link underlying pathology with clinical outcome may provide direct evidence for or against theories regarding the development of cerebral lateralization (e.g., equipotentiality 13 versus functional specialization, 14 the crowding hypothesis 15 ).
To define further the relationship between MCD and cognitive outcome, we studied intellectual functioning in children with medically resistant partial epilepsy due to disorders of cortical development as a function of etiology and laterality of the lesion. We characterized subjects according to the histopathologic features of specimens removed at epilepsy surgery. We were specifically interested in investigating a possible relationship between histopathology