Agenesis of the Arcuate Fasciculi in Congenital Bilateral Perisylvian Syndrome: A Diffusion Tensor Imaging and Tractography Study

Objective  To describe the absence of the arcuate fasciculi in 2 cases of congenital bilateral perisylvian syndrome (CBPS).

Design  Case series.

Setting  Pediatric referral hospital–based study.

Patients  Two patients with CBPS, referred to our institution as candidates for surgical treatment of epilepsy.

Intervention  Diffusion tensor imaging (1.5-T scanner; 15 encoding directions; b = 800 s/mm²) and deterministic tractography of the main projection and association tracts.

Main Outcome Measures  Neuropsychology evaluation; fractional anisotropy, apparent diffusion coefficients, and anatomical aspect of the tracts.

Results  Absence of the arcuate fasciculus was observed in both subjects. Ancillary findings were complete absence of the superior longitudinal fasciculi in 1 case and underdevelopment in the other. Low fractional anisotropy of the left inferior occipitofrontal fasciculus was found in both cases. The same tract was maloriented in 1 of the cases.

Conclusion  Agenesis of the arcuate fasciculus may accompany CBPS.

The congenital bilateral perisylvian syndrome (CBPS) is a type of cortical developmental abnormality characterized by poor operculation of the parietal and frontal lobes, wide lateral sulcus, polymicrogyria, orofacial diplegia, epilepsy, and developmental delay.^1,2 Seizures are present in 65% of cases.^3,4 To our knowledge, CBPS has not been studied to date with diffusion tensor imaging and fiber tractography.

We describe 2 cases with absence of the arcuate fasciculus (as part of agenesis or hypoplasia of the superior longitudinal fasciculus) in CBPS using diffusion tensor imaging and fiber tractography in relation to clinical and neuropsychological findings. To our knowledge, there are no existing reports describing bilateral agenesis of the arcuate fasciculus in this condition. The correlation of this finding with the clinical analysis of the language/speech deficit may contribute to the understanding of the arcuate fasciculus function.

The clinical and neuroradiological findings of 2 cases with CBPS are summarized in the Table. Case 2 has no arcuate fasciculus but has the remnant fibers of the superior longitudinal fasciculus. Normally, the superior longitudinal fasciculus has, in addition to the arcuate fasciculus, a bundle of short fibers connecting the parietal areas (supramarginal gyrus) with frontal areas. In addition, the fractional anisotropy of the left inferior occipitofrontal fasciculus was found to be low in both cases and bilaterally in the cingulum of case 2.

A single-shot, spin-echo, echo-planar imaging sequence with diffusion weighting consisting of 15 encoding directions was performed in a 1.5-T scanner (Figure 1). A diffusion weighting (b) of 800 s/mm² was used. Fractional anisotropy and tractography was performed using Volume-One software (http://www.volume-one.org/). The superior longitudinal fasciculus containing the arcuate fasciculus fibers was sought in a coronal plane at the level of the rostral aspect of the splenium. The tract appears normally as a green triangle lateral to the blue descending fibers of the corticospinal tracts (Figure 2), from where the arcuate fasciculus can be tracked (Figure 3). This area was contoured bilaterally defining the seeding region of interest (ROI). Tract propagation was terminated when the tract trajectory reached a voxel with fractional anisotropy less than 0.13 or when the angle between 2 consecutive steps was greater than 45°. Fractional anisotropy and apparent diffusion coefficients values were obtained from the inferior occipitofrontal fasciculi, cingulum (single coronal ROIs at the level of the anterior commissure), the inferior longitudinal fasciculi in conjunction with the inferior occipitofrontal fasciculi (single coronal ROI at the level of the retrosplenial surface), and the internal capsules (single axial ROI at the level of the thalamus) (Figures 4, 5, and 6).

The superior longitudinal fasciculus consists mainly of the long curved fibers with posterior end points in the temporal cortex and a bundle of rather horizontal fibers whose posterior end points are located in the parietal lobe (Figure 3). The curved fibers correspond to the arcuate fasciculus, a tract considered crucial for the communication between receptive and expressive language brain areas.^5-7 Lesions of the arcuate fasciculus result in a deficiency in the capacity to repeat, a syndrome that has been coined “conduction aphasia.” Other authors have proposed that the arcuate fasciculus also plays a role in intelligence⁸ and nonlanguage cognitive functions.⁹

The absence of the arcuate fasciculus in our 2 cases provides an opportunity to look into its role. We were more concerned in what has been preserved as opposed to the deficit, since our cases have many other cortical abnormalities that could be the cause of any cognitive or motor deficiency. Looking at what has been preserved gives us an idea of what the arcuate fasciculus is not involved in.

Automatized language (eg, reciting automatic series) and delayed recall of verbal and nonverbal material was preserved in both patients. Strikingly, no report of conduction aphasia was mentioned. Therefore, at least for these 2 patients, the arcuate fasciculus was not needed for these functions. The common clinical findings in these cases with arcuate fasciculus agenesis were delayed speech development with poor articulation and poor prosody and other aspects of speech. In addition, both patients showed poor phonemic and semantic word generation, and difficulties in visuospatial, organization/assembly skills, that may prompt a diagnosis of constructional apraxia.

The idea to attribute the phonological difficulties of our patients merely to the arcuate fasciculus absence seems supported by a recent report of intraoperative electrophysiological studies that have shown the arcuate fasciculus transmits phonological cues.^10,11 However, concomitant cortical and connectivity findings confound this observation.

The role of the arcuate fasciculus (and the entire superior longitudinal fasciculus) in speech and language is not completely understood. Normal volunteers show a longer left arcuate fasciculus, with more fibers, and higher fractional anisotropy values. Moreover, in many cases, the right arcuate fasciculus is nonexistent.^12-14 Strikingly, left arcuate fasciculus dominance has been reported in subjects with right hemisphere language dominance.¹⁵ More recently, lateralization of the arcuate fasciculus has been found correlated with the lateralization index of language determined by functional magnetic resonance imaging on patients with right but not left temporal lobe epilepsy.¹⁶

We present for the first time, to our knowledge, 2 cases of CBPS with bilateral absence of the arcuate fasciculi. This finding analyzed in the context of the associated clinical findings may help to understand the clinical presentation of the condition and further expose language organization.

Correspondence: Byron Bernal, MD, 3100 SW 62nd Ave, Miami, FL 33176 (byron.bernal@mch.com).

Accepted for Publication: June 30, 2009.

Author Contributions:Study concept and design: Bernal. Acquisition of data: Bernal, Rey, Dunoyer, and Shanbhag. Analysis and interpretation of data: Bernal, Rey, and Altman. Drafting of the manuscript: Bernal, Rey, and Dunoyer. Critical revision of the manuscript for important intellectual content: Bernal, Rey, Shanbhag, and Altman. Administrative, technical, and material support: Dunoyer. Study supervision: Bernal, Rey, and Altman.

Financial Disclosure: None reported.