Cortical stimulation mapping and Wada results demonstrate a normal variant of right hemisphere language organization.

Cortical stimulation mapping (CSM) involves inducing a localized electrical blockage of cortical function, and has been used in neurosurgery to identify areas essential to language functioning since the pioneering work of Penfield and Roberts with epilepsy surgery patients during the mid-1950s (Penfield & Roberts, 1959). Identification of language cortex enables the surgeon to avoid language areas when formulating a surgical strategy and to expand the resection of epileptic regions with increased safety (Kim et al., 2009). CSM results have brought insight to the distribution of language, confirming the presence of the classical Broca’s and Wernicke’s areas and identifying nontraditional sites found elsewhere in the cortex (Ojemann et al 1989). These data have also validated the accuracy of the intracarotid amobarbital (Wada) procedure, confirming that essential language sites are present in patients that are left-hemisphere (LH) dominant for language. The Wada procedure has served as the gold-standard for determining language dominance in presurgical patients since the early days of epilepsy surgery. Nevertheless, little data have been published examining CSM in patients who are determined to be exclusively right-hemisphere (RH) dominant for speech. Such individuals are believed to be rare, and atypical language lateralization is thought to often develop in patients following an early life injury involving their left cerebral hemisphere (Moddel et al., 2009; Rasmussen & Milner, 1977). Surgical patients experiencing such pathological language reorganization may not undergo CSM, as neurosurgical intervention is more likely to involve the LH of these individuals. As the Regional Epilepsy Center at the University of Washington has over 30 years of data involving CSM and the Wada procedure, we sought to examine the frequency of RH language dominance, validate the Wada procedure in these cases, and determine how the CSM results of such patients compare to those of individuals with typical speech lateralization. Validation of the Wada Procedure The Wada procedure has traditionally been the primary method for determining language dominance in neurosurgical candidates, as it provides a means to assess each hemisphere’s contribution to language function independently (Loring et al., 1992). This procedure remains in use by most epilepsy centers for the presurgical evaluation, although increasing efforts have been made to employ a noninvasive alternative [e.g., fMRI] (Arora et al., 2009). The principle behind the Wada was demonstrated by W. James Gardner in 1941 (Gardner, 1941), who injected procaine directly into the cortex thought to potentially be involved in language. Juhn Wada more formally established this procedure in the area of epilepsy surgery, making separate injections of amobarbital into each internal carotid artery to produce hemispheric anesthesia (Wada, 1949). During this period of anesthesia, patients are asked to carry out tasks to evaluate language and memory. Anesthetizing the area renders it nonfunctional, causing the hemisphere with language representation to be revealed when language deficits appear during an injection. The validity of LH language dominance as predicted by the Wada procedure has been demonstrated in several group studies using CSM (Loring et al., 1992; Ojemann et al., 1989), yet validity data for RH language dominance in epilepsy have been restricted primarily to a handful of case studies with one or two subjects (Rosenbaum et al., 1989; Wyllie et al., 1990). Duffau and colleagues (2008) found RH language sites in a small series of tumor patients (n=9) who were all left-handed, but did not have Wada data on any of their patients (some percentage were likely bilateral language cases). One recent study examined CSM results in 15 neurosurgical patients (mostly tumor patients) that had exhibited either RH (n=7) or bilateral (n=8) language on the Wada (Chang et al., 2011). These researchers found language sites in the RH of all 15 patients, thoroughly delineated their neuroanatomical location, and noted that the pattern of language sites was consistent with findings from LH language dominant patients. There are also a few additional studies that demonstrate RH speech in a small number of patients identified as bilateral by the Wada (Jabbour et al., 2005; Loring et al., 1988). Prevalence of Atypical Language Lateralization Years of accumulated knowledge from the use of the Wada confirms that most neurosurgical candidates are LH dominant for language, with atypical language lateralization in a minority (Loring et al., 1990; Moddel et al., 2009). The prevalence of exclusive RH speech is believed to be infrequent, and right-sided surgery on such patients is believed to be rare. Most studies have suggested the prevalence of RH language representation to range from approximately 2 to 10% in epilepsy patients undergoing the Wada (Loring et al., 1990; Moddel et al., 2009), with the exception of a study by Rasmussen and Milner (1977) that estimated it to be as high as nearly 20%. These studies have estimated bilateral language representation to range from 5 to 25% in this population. Variability in reported prevalence rates is believed to primarily reflect differences in the implementation of the Wada across epilepsy centers, as many institutions have not routinely administered the Wada procedure to all individuals. Most centers select patients for the Wada based on the presumed clinical likelihood of atypical language occurrence. Variability in reported rates of atypical language may also be influenced by small sample sizes used in these studies, and general variability across Wada methodologies. Finally, while the Wada procedure is invasive and administered only to neurosurgical candidates, functional technologies (fMRI, functional transcranial Doppler sonography, MEG) have been used to estimate rates of atypical language prevalence in healthy subjects (Knecht et al., 2000; Szaflarski et al., 2002). These studies suggest that atypical language lateralization (RH or bilateral) will occur in 4 to 7.5% of right-handers and up to 22% of left-handers in the general population. Studies of “crossed-aphasia,” a term for the development of language impairment in normal right-handed individuals who sustained damage to the right hemisphere later in life, also provides evidence that atypical language lateralization may be a normal variant of language development (Hecaen & Albert 1978). Such studies suggest that atypical language may occur in .4% to 3% of healthy right handed adults (Marien et al 2004; Zangwill 1979), although most reports have not included the necessary functional imaging studies to determine whether this reflects bilateral or right only language (See Vandervliet et al 2008; Vitali et al 2011 for recent exceptions). More than two hundred cases appear in the research literature (Marien et al 2004); with more recent studies documenting similar aphasia types, lesion-behavior relationships, and recovery patterns as in uncrossed aphasia (Castro-Caldas & Confraria 1984; Coppens et al 2002; Yarnell 1981). The similarity of the resulting syndromes suggests that atypical language representation could parallel the more typical left-hemisphere variety. Right-hemisphere Speech Patients as Surgical Candidates Most neurosurgical patients found to have RH speech are undergoing surgical procedures involving their LH. This is because atypical lateralization of the neural language network appears to occur most frequently as a result of early childhood neuronal injuries leading to a shift of language functions from the left to the RH. A previous study from our Center (Miller et al., 2003) had also suggested that atypical speech lateralization may not occur in epilepsy patients with normal neurologic histories through the age of 15 years. In the current study, we retrospectively examined the CSM and Wada data of all patients found to have RH speech and who also underwent RH surgical procedures. This represents the largest published sample of epilepsy patients undergoing the Wada procedure, and is unique in that it is a near consecutive series requiring the obligatory completion of this study for all surgical candidates during the time span of the study regardless of hemisphere of seizure onset and the routine injection of both hemispheres. Additionally, the Wada procedure remained unchanged throughout the duration of the study. We predicted that CSM results for these RH patients would positively identify one or more essential language sites in each patient, providing validation for the sensitivity of the Wada procedure for detecting RH speech in a large, unselected sample of RH language patients. We also hypothesized that the general distribution of sites would be consistent with the patterns observed in neurosurgical patients with LH language dominance, as we believe that prior research suggests this group will predominantly reflect a sample of patients with an atypical yet normal variant of language organization.