Abstracts

Rationale: Classification of hemispheric language representation prior to epilepsy surgery is critical to avoid postoperative dysphasia. There is risk of language impairment even in patients with bilateral language representation undergoing surgery involving the minor language hemisphere. Atypical hemispheric language representation is more commonly seen in patients who are left-handed and those with a history of early left hemisphere seizure onset or structural abnormality. Right-handers are given little consideration regarding the potential for language in the right hemisphere despite a reported incidence of 13% in one large epilepsy sample. This study sought to investigate the possible value of family history of left-handedness (FHLH) in predicting right hemisphere language in right-handed patients independent of other predictive factors in a pediatric sample. Methods: The records of 248 consecutive pediatric epilepsy patients ages 7-18 years referred for the IAP were reviewed. Patients who underwent only a single amobarbital injection, or for whom hemispheric language dominance could not be classified, were excluded. Language was classified as bilateral if tboth hemispheres evidenced meaningful speech production or disruption of language despite preserved alertness.  The final sample included 117 left, 41 bilateral, and 16 right hemisphere dominant patients. Bilateral and right dominant patients comprised the atypical language (AL) group. AL was examined in relation to presence (FHLH+) or absence (FHLH-) of a first degree left-handed relative, patient handedness, side of seizure onset , structural abnormality,  and age of seizure onset. Data were analyzed by chi-square, Fisher exact test, one-way ANOVA and t-test. Results: Left-handed patients (62%) were significantly more likely to evidence AL than right-handed patients (26%). Among left-handed patients, AL was more common with left (76%) than right (25%) hemisphere seizure onset (p=.03), but not disproportionately more likely with left (69%) than right (43%) structural abnormality or in FHLH+ (45%) versus FHLH- (65%).  Among right-handed patients, AL was more common with left (38%) than right (12.5%) hemisphere seizure onset (p=.002) and with left (45%) than right (25%) structural abnormality, the latter falling slightly short of statistical significance (p=.13).  In right handed patients, AL was more common in FHLH+ (48%) than in FHLH- (20%,p=.001), whether FHLH+ patients had a right (31%) or left (58%) seizure onset, or had a right (50%) or left (63%) structural abnormality. AL was highly unlikely in right-handed/right seizure onset patients in the absence of FHLH (3%) or with seizure onset earlier than two years (12%). Seizure onset was earlier in those with right dominance (1.7 years) than in those with left (5.2 years) or bilateral language (5.5 years; p=004). Conclusions: As in previous studies, left-handedness and left hemisphere EEG/MRI abnormalities are associated with increased incidence of atypical language on IAP. In right-handed patients, there is an independent contribution of family history of left-handedness to atypical hemispheric organization of language. Atypical langauge occurs in a substantial proportion of right-handed patients with a family history of left-handedness regardless of side of seizure onset or side of structural abnormality. These findings have important implications for predicting likelihood of right hemisphere involvement in language processing in those who cannot undergo IAP or noninvasive language lateralization procedures prior to epilepsy surgery, or deciding whether a patient should undergo IAP rather than noninvasive hemispheric language assessment. Funding: No funding was received.