Abstracts

Rationale: Some epilepsy surgery candidates show right hemisphere or bilateral language representation. However, language consists of many abilities, including confrontation naming, reading, comprehension, and repetition. Pathognomonic signs such as paraphasic errors are also important in interpreting language results. The current study examines patterns of non-left language dominance from testing of multiple language domains during Wada testing in a large, multi-center epilepsy sample. Methods: All of the Wada reports from 2003 to 2008 (n=175) with uncomplicated bilateral ICA injections performed at the Northeast Regional Epilepsy Group, Long Island Jewish Medical Center, and NY Presbyterian/Weill-Cornell were reviewed. All of the Wada tests utilized the same protocol administered by the same examiner (KP). Immediately after amobarbital injection and confirmation of adequate anesthesia, the patient was shown objects to name aloud and memorize. Language was further tested by serially conducting tasks of naming, reading sentences, following written and oral commands, auditory verbal comprehension of reversed syntax, and repetition. Patients were identified as left language dominant if the left intracarotid (ICA) injection produced partial or complete mutism, anomia for the test stimuli, alexia for oral reading and following commands, impaired comprehension, and impaired repetition with paraphasic errors. All patients who did not meet these criteria were determined to be non-left language dominant. Results: Sixteen of 175 patients had non-left language. Only two patients showed completely reversed right hemisphere language dominance. Five patients were aphasic after both injections, and one patient was not aphasic after either injection (but with adequate confirmation of anesthesia). The remaining patients showed complex patterns of bilateral language. Two patients were briefly mute after left injection and 6 were mute after right injection. Right ICA injections produced anomia in 6 patients, alexia in 8 patients, comprehension deficits in 10 patients, and paraphasic errors in 12 patients. Left ICA injections produced anomia in 9 patients, alexia in 8 patients, comprehension deficits in 6 patients, and paraphasic errors in 9 patients. Generally, if one language function was unimpaired after one injection, it was impaired with the contralateral injection. Conclusions: Language is not a unitary entity. Unless multiple language functions such as naming, reading, comprehension, repetition, and presence of paraphasic errors are examined, false negative errors regarding language could occur and result in post-operative impairment. Completely reversed right hemisphere language is rare, while other non-left language dominant patients show complex patterns of separate language functions present in each hemisphere that should be considered when planning surgery near eloquent cortex.