Visuospatial deficits in patients with early left-hemispheric lesions and functional reorganization of language: Consequence of lesion or reorganization?

Abstract

Introduction

Patients with early left-hemispheric lesions have repeatedly been reported to show spared language functions (which are then mediated by the right hemisphere), but exhibit deficits in visuospatial functions. The “crowding hypothesis” explains these deficits by a neuronal scarcity in the right hemisphere for the original right hemispheric functions. An alternative hypothesis suggests direct lesion effects as the sole reason for the visuospatial impairments. The scope of this study was to examine the relations between visuospatial skills and hemispheric preference for language, and between visuospatial skills and cerebral lesion size.

Methods

In a sample of young adults with pre- or perinatally acquired focal lesions of the left hemisphere and unilateral spastic cerebral palsy on the right side, we assessed intelligence and neuropsychological functions in the verbal and visuospatial domains. The behavioural data was correlated with structural MRI information and language preference as assessed using fMRI.

Results

Patients with right hemispheric language production were impaired in visuospatial functions, compared not only to normal controls, but also to patients without right hemispheric language preference. Additionally, the degree of right hemispheric language correlated negatively with performance in visuospatial tasks. Lesion size correlated negatively only with performance IQ, which again correlated negatively with motor impairment of the patients and thus does not seem to reflect true cognitive deficits.

Conclusion

Visuospatial deficits in patients with early left-hemispheric lesions are a consequence of lesion-induced right hemispheric language organization, thus lending further support to the “crowding hypothesis”.

Introduction

Brain lesions acquired before birth or within the first years of life induce different patterns of neuropsychological deficits than similar brain lesions acquired in adult life. The developing brain shows a remarkable plasticity and flexibility in dealing with focal damage. In an adult patient, the site of a lesion predicts deficits in certain neuropsychological functions: lesions in the territory of the middle cerebral artery of the left hemisphere (LH) often lead to severe aphasias (Damasio & Geschwind, 1984). Lesions of the temporo-parietal region of the right hemisphere (RH) induce problems in visual attention, known as the spatial neglect phenomenon (Karnath, Fruhmann-Berger, Küker, & Rorden, 2004). In contrast, the correlation between neuropsychological deficits and lesion site is much less clear in patients with early lesions. Deficits in visuospatial skills after early RH lesions are in line with the adult model (Muter, Taylor, & Vargha-Khadem, 1997; Vargha-Khadem, Isaacs, & Muter, 1994), although a contrasting finding is that children with RH lesions experience delays in language development (Aram, Ekelman, Rose, & Whitaker, 1985; Thal et al., 1991, Vicari et al., 2000). On the other hand, young children with LH lesions behave concordantly with the adult model, in showing delays in language acquisition (Chilosi, Cipriani, Bertuccelli, Pfanner, & Cioni, 2001; Thal et al., 1991, Vicari et al., 2000); however, by school-age or adolescence, these formerly language-delayed children show a remarkable catch-up, with only subtle, if any, deficits persisting (Bates & Roe, 2001). Even more in contrast to the adult model are reports on children with early LH lesions suffering from deficits in visuospatial tasks or showing a lower performance IQ (Muter et al., 1997, Vargha-Khadem et al., 1994), both being traditionally regarded as indicators of disturbed RH-mediated functions. These paradoxical data indicate that developing brains react differently to insults than adult brains. Communicative skills are often spared in early LH brain lesions by shifting language functions to the RH (Rasmussen & Milner, 1977). The paradoxical neuropsychological findings of impaired visuospatial skills in patients with early LH brain damage have been explained by the “crowding-hypothesis”: the RH is thought to have difficulties with the mediation of both nonverbal and verbal functions, which, due to a dominance of language, leads to deficits in RH nonverbal functions (Teuber, 1974). By now, a growing body of evidence has confirmed the notion that patients with early LH lesions show deficits in visuospatial functions and spared language skills. An alternative explanation, however, could be that brain lesions have a nonspecific effect on visuospatial skills: neuropsychological tests which are sensitive to RH brain damage are also sensitive to diffuse brain injury (Lee & Hamsher, 1988). Therefore, simple lesion size could be the determining factor in producing visuospatial impairments in patients with early LH brain lesions.

Experiments examining the relationship between language reorganization and neuropsychological profiles have yielded controversial results. The relations between language dominance and neuropsychological profiles have been examined most comprehensively in patients with intractable epilepsy, who have undergone the Wada-test. In these studies, patients with atypical (RH or bilateral) language representation have shown larger deficits in visuospatial functions than patients with normal LH language organization (Billingsley & Smith, 2000; Loring et al., 1999; Strauss, Satz, & Wada, 1990). In hemiplegic children without epilepsy, language dominance has also been assessed by the dichotic listening task. Here, children with atypical left ear advantage did not show any deficits in nonverbal tasks when compared to children with normal right ear advantage (Korkman & von Wendt, 1995). Both types of studies, however, have caveats which limit the interpretability of the results: In epileptic patients, cognitive representations within the brain (Janszky et al., 2003) as well as cognitive functions (Henkin et al., 2005, Lindgren et al., 2004) are influenced by seizure activity; thus it is unclear whether inferences drawn from data on epileptic patients are valid also for non-epileptic subjects. Furthermore, the dichotic listening task assesses language perception, which is, to some extent, bilateral (Hickok & Poeppel, 2000; Staudt et al., 2001) and thus not as prone to be reorganized due to LH brain damage. It is also influenced by attentional factors and thus less reliable in assessing language dominance than the Wada procedure. Using positron emission tomography (PET), language can be localized reasonably reliably not only in terms of hemispheric dominance, but also with regard to brain structures involved. As the PET procedure involves application of radioactively labelled substances, its use in non-clinical studies has been rather restricted. As with the Wada-test, it has been applied mainly in patients with epilepsy to test language reorganization after early lesions (Müller et al., 1998a, Müller et al., 1998b, Müller et al., 1999a, Müller et al., 1999b). RH language activation in these epileptic patients more or less mirrors the normal LH pattern in healthy controls. Functional MRI has the advantage of being non-invasive and it can localize language representation reliably and with a high spatial resolution. However, literature on language representation assessed by fMRI in patients with early unilateral brain lesions is still limited. Reorganized language representation has been evaluated, again, mainly in patients with epilepsy (Hertz-Pannier et al., 1997, Liegeois et al., 2004, Pataraia et al., 2004, Springer et al., 1999). Studies on language representation in non-epileptic patients with early LH lesions using non-invasive fMRI are scarce, but consistent: a study on patients with unilateral white-matter lesions revealed a significant correlation between left facial motor-tract involvement and RH activation for language production (Staudt et al., 2001). Other studies showed that RH language production recruits areas homologous to LH language areas in normal right-handers (Booth et al., 1999, Staudt et al., 2002). Both studies have been restricted to the description of functional activation patterns in the patients, with little information on neuropsychological profiles. Patients showed normal nonverbal intelligence with slight language impairments (Booth et al., 1999), or normal overall and verbal intelligence with slightly lower performance IQ than the control group (Staudt et al., 2002). However, due to small groups and limited neuropsychological tests, language (re-)organization patterns were not related directly to neuropsychological outcome.

To summarize, two competing hypotheses try to explain visuospatial deficits in patients with lesion-induced RH language:

H1: Language reorganization induces scarcity of neural substrate in the RH, thus impairing visuospatial functions (Teuber, 1974).

H2: Brain lesions have a nonspecific effect on cognition, with visuospatial functions being most vulnerable to impairment (Lee & Hamsher, 1988).

In the present work, we examined the relationship between anatomy (structural and functional) and neuropsychological profiles in patients with early LH focal brain lesions.