Even When Right Is All That’s Left: There Are Still More Options for Recovery from Aphasia

Two prominent hypotheses inform our current understanding of how the brain recovers speech-motor function after an insult: (1) recruitment of perilesional cortex to form a new network of regions in the affected hemisphere (the perilesional hypothesis) (Fridriksson et al., 2012; McKinnon et al., 2017), (2) engagement of homotopic regions and associated fiber tracts in the undamaged hemisphere to facilitate recovery of speech-motor function (Geranmayeh et al., 2014; Heiss and Thiel 2006; Zipse et al., 2012; Pani et al., 2016; Turkeltaub et al., 2011; Xing et al., 2016), and (3) a combination of both. This “right-shift hypothesis” is based on the neural redundancy theory that both hemispheres have the capacity to support speech-motor function (e.g., Hickok and Poeppel 2004, 2007; Saur et al., 2006, 2008; Marchina et al., 2018), although the degree to which the right hemisphere becomes involved after damage to the left may differ from person to person and/or depend upon factors that are not yet fully understood. Among the potential determinants of which route is chosen and which leads to success, size and site of left-hemisphere lesions and lesion load of the arcuate fasciculus (AF) or other speech/language-relevant structures (Marchina et al., 2011; Hillis et al., 2018), may be the most important.

Whether or not recruitment of the right hemisphere (a) is a sign of “incomplete recovery” or “less effective compensation” (Heiss and Thiel, 2006), (b) occurs only as a transient phenomenon during the recovery process (Saur et al., 2006), or (c) indicates use of built-in redundancy (Turkeltaub et al., 2011; Marchina et al., 2018; Ozdemir et al., 2006) to facilitate recovery in case of an insult to the left hemisphere is still a matter of debate.

In this volume of the Annals of Neurology, Northam and colleagues (page XXX) present the results of a study evaluating stroke lesions acquired in infancy and their long-term impact on anatomically-defined dorsal and ventral language streams. They bring to the fore an important discussion of the roles of these parallel streams during development, and highlight the unique contributions of the dorsal language stream and the arcuate fasciculus in speech-motor function. They report findings from a prospective cohort of infants who were recruited at London’s Hammersmith Hospital and followed-up many years later using structural and functional neuroimaging. Their results provide valuable insights into the adaptations being made by highly plastic, developing brains over an extended period of time. Interestingly, they showed that damage to the dorsal stream is associated with persistent speech deficits that are closely analogous to those that result from adult-onset conduction aphasia.

Yet some individuals performed within normal limits on tests of language despite severe damage to left-hemisphere speech-motor/language networks. Since all infants’ injuries occurred during the neonatal period prior to the acquisition of language, the differential effect seen in some individuals (who, despite the transition of speech-motor function to their right hemisphere, performed within normal range on tests), must be related to variability in the anatomical and/or functional hemispheric capacity for supporting speech and language function.

What could account for that variability? On one hand, variability could be genetically determined or based on very early experiences that result in asymmetric pruning of the right- more than the left AF. Supporting this hypothesis is published work (e.g. Song et al., 2016) suggesting that some components of the AF in typically-developing neonates and infants appear more symmetrical, only showing lateralizing effects later in development (years later). On the other hand, AF asymmetry might arise because localizing fast auditory-motor feedforward and feedback processing (required for repeating words/phrases and propositional speech) and mediated through the AF to one hemisphere is faster and more efficient than processing the same information bihemispherically through the corpus callosum due to the time constraints of transcallosal information transfer (Ringo et al., 1994). Finally, given the finding of Northam et al. (2018) of normal performance on behavioral tests despite severe damage to the left AF, the site and size of the left-hemisphere lesion may interfere with the normal transcallosal inhibitory system, allowing the right AF to develop via experience-dependent or compensatory plasticity and leading to greater engagement and recruitment of the right hemisphere and its AF during development but also possibly in the mature brain after insults.

In the adult brain, it has been shown that the size of the right AF (Forkel et al., 2015) and regional fractional anisotropy (FA) values in homotopic, right-hemisphere speech-motor regions are possible indicators of differences in local connectivity (Pani et al., 2016). Developing a deeper understanding of a specific lesion’s impact on left-hemisphere speech-motor structures and network connections would enable adaptation of interventions to meet individual patients’ needs, lead to more personalized treatments, and ultimately create the possibility of enhancing a patient’s outcome. For example, a patient with a smaller left-hemisphere lesion and/or low AF lesion load, might derive greater benefit either from approaches that strengthen left-hemisphere function/connectivity and create compensatory left-hemisphere pathways (e.g., Meinzer et al., 2016), or from approaches that interfere with potential inhibitory transcallosal influences from the right hemisphere (Martin et al., 2009; Hamilton et al., 2011). By contrast, patients with large left-hemisphere lesions or lesions that critically impact relevant structures (i.e., high AF lesion load) may benefit more from approaches that engage and recruit right-hemisphere structures (e.g., intonation-based speech-therapy; Norton et al., 2009; Zipse et al., 2012; Wan et al., 2014) or enhance activity/connections of right-hemisphere regions via approaches that could positively interfere with deleterious inhibitory influence of the affected left-hemisphere hemisphere on right hemisphere structures (Hamilton et al., 2011). In either case, it appears that due to its built-in bihemispheric organization and inherent redundancy, the speech-motor/ language system has more paths to recovery than the upper extremity motor system, for which recovery seems to be mostly dependent on the integrity and residual capacity of the lesional or affected motor output system.

In sum, the development of interventions for the speech-motor/language system can benefit greatly from studies examining natural outcomes and recovery following stroke even (or perhaps, especially) when those studies stretch across many years. Findings from such investigations underscore the brain’s resilience as it recruits healthy cortex surrounding smaller left-hemisphere lesions for new speech networks, highlight its remarkable ability to re-route speech-motor function to homotopic right-hemisphere regions/networks when left-hemisphere pathways are no longer viable, and reveal potentially untapped resources that may hold many more options for recovery through the unaffected right hemisphere which may play an important role in that process.