Slide Slam P6 Sandbox Series
Right hemisphere structural connectivity predicts naming outcomes in chronic aphasia
Lynsey Keator1, Natalie Busby Hetherington1, Samaneh Nemati1, Leigh Ann Spell1, Argye Hillis2, Gregory Hickok3, Christopher Rorden1, Leonardo Bonilha4, Julius Fridriksson1; 1University of South Carolina, 2Johns Hopkins School of Medicine, 3University of California Irvine, 4Medical University of South Carolina
Background Aphasia is a language disorder resulting from damage to eloquent language regions in the left hemisphere (LH). Diffusion tensor imaging (DTI) assesses the microstructural integrity of white matter pathways in the brain. Most DTI studies in aphasia focus on network connectivity in the LH and suggest that damage to LH structural connections in remote regions may substantially impact language in the chronic stages (Yourganov et al., 2016; Bonilha et al., 2014). Fewer investigate connections between right hemisphere (RH) homologues (Wan et al., 2014; Schlaug et al., 2009; Pani et al., 2016; Keser et al., 2019). The role of the RH in recovery, and the interaction with LH integrity, remains inconclusive and it is still unclear which neural mechanisms support aphasia recovery. The aim of the current study is to determine if RH structural connectivity is associated with naming outcomes in poststroke aphasia. To do this, we model severity at two different time points and then model longitudinal changes across these time points. Methods First, we used baseline DTI data and naming abilities (as measured by the Philadelphia Naming Test; Roach et al., 1996) at two time points: baseline (n = 101) and six months (n = 80). NiiStat toolbox (https://www.nitrc.org/projects/niistat/) was used to conduct DTI analyses (p < 0.05, correction for multiple comparisons with permutation thresholding [5000 permutations]). To control for the LH lesion, we regressed out damage to LH dual stream regions (sum of proportional damage to dorsal and ventral regions; Fridriksson et al., 2016) using the AICHA atlas (Joliot et al., 2015). Next, longitudinal differential tractography was performed using DSI Studio (http://dsi-studio.labsolver.org/) between baseline and 6-month DTI scans (Yeh et al, 2013). Changes in naming scores were calculated as proportion of maximal gain (PMG): (score at 6 months-baseline)/(maximum score-baseline). Tracking only occurs along trajectories where changes in anisotropy are apparent between repeat scans in the same individual. PMG scores were associated with these connectomes using correlational tractography. Results Naming performance at baseline was predicted by the integrity of structural connections between anterior and posterior RH homologues (IFG : Middle Occipital Gyrus (MOG), z = 4.33, 4.62; MOG : Anterior Insula, z = 4.24; MOG : Superior Temporal Gyrus (STG), z = 3.93, 3.91; Anterior Insula : STG, z = 4.31). Similar RH anterior-posterior connections predicted outcomes at six months. Longitudinal connectometry analyses revealed that connectivity in the RH inferior longitudinal fasciculus (ILF) positively correlated with change in naming performance between baseline and 6-months (p < 0.01). Conclusion Results suggest that the integrity of RH anterior-posterior homologue regions and the structural connections between them are correlated with naming at baseline and six months post-treatment. Consistent with this, longitudinal analyses reveal the right ILF positively correlates with change scores. Results are largely consistent with a recent study examining similar associations in subacute aphasia (Blom-Smink et al., 2020). Determining the role of the RH, or more broadly, considering predictive neural mechanisms that may predict language outcomes in chronic aphasia, will be important to advance treatment modalities for aphasia.