Baseline structural imaging correlates of treatment outcomes in semantic variant primary progressive aphasia

Heather Dial, Eduardo Europa, Stephanie M. Grasso, Maria Luisa Mandelli, Kristin M. Schaffer, H. Isabel Hubbard, Lisa Wauters, Lindsey Wineholt, Stephen M. Wilson, Maria Luisa Gorno‐Tempini, Maya L. HenryCortex 2022 · DOI

Introduction

Semantic variant primary progressive aphasia (svPPA) is a neurodegenerative condition marked by the progressive erosion of semantic memory, leading to significant difficulties in word retrieval and object recognition. Despite the degenerative nature of the disease, behavioral interventions like Lexical Retrieval Training (LRT) have shown promise in helping patients relearn lost vocabulary. However, treatment response varies significantly across individuals. This study investigates whether baseline brain structure, measured via magnetic resonance imaging (MRI), can predict which patients are most likely to benefit from language therapy and generalize those gains to untrained words.

Methodology

The study involved 19 individuals diagnosed with svPPA, representing one of the largest cohorts for this type of intervention research. Participants underwent Lexical Retrieval Training (LRT), a structured behavioral therapy that utilizes a hierarchy of semantic, phonemic, and orthographic cues to facilitate word retrieval.

To identify the neural substrates of treatment success, researchers employed two primary neuroimaging analyses:

  1. Voxel-Based Morphometry (VBM): A whole-brain approach used to identify correlations between regional gray matter density and post-treatment naming accuracy.
  2. Region of Interest (ROI) Analysis: A targeted approach focusing on four anatomical "families" hypothesized to support language and memory: the temporal, parietal, and frontal lobes, and the hippocampal complex.

Baseline MRI scans were compared against a control group of 60 age-matched healthy individuals to characterize the extent of atrophy before correlating structural integrity with behavioral outcomes.

Key Findings

The behavioral results confirmed the efficacy of LRT. Participants demonstrated a dramatic increase in naming accuracy for trained items, rising from a baseline of approximately 5.6% to 78.9% post-treatment. Notably, the therapy also led to generalization, with accuracy for untrained items increasing from 4.6% to 25%.

The structural analysis revealed specific predictors of these gains:

  • Left Parietal Lobe: Higher tissue density in the left supramarginal gyrus (SMG) and angular gyrus (AG) was strongly associated with better naming outcomes.
  • Left Hippocampus: Integrity of the left hippocampus emerged as a critical predictor not only for trained items but specifically for the generalization of gains to untrained items.

Discussion: The Neural Mechanisms of Recovery

The findings support the "Alternative Route" hypothesis in svPPA. As the anterior temporal lobes—the brain's primary semantic hub—atrophy, successful patients appear to recruit spared regions in the left parietal lobe. These regions are traditionally associated with phonological processing and working memory. By becoming "hyper-phonological," patients may bypass damaged semantic pathways to retrieve words.

Furthermore, the role of the left hippocampus suggests that episodic memory and strategic learning are vital for therapeutic success. The hippocampus likely facilitates the encoding and retrieval of the specific strategies learned during therapy, allowing patients to apply these techniques to novel, untrained stimuli. The "privileged connection" between the left hippocampus and the lateralized language network explains why its structural integrity is a more potent predictor than its right-hemisphere counterpart.

Conclusion and Future Directions

This research demonstrates that baseline structural MRI can serve as a biomarker for treatment response in svPPA. Identifying the left parietal lobe and hippocampus as key drivers of recovery provides a roadmap for understanding how the brain compensates for neurodegeneration.

While the sample size is substantial for this rare population, future studies utilizing functional MRI (fMRI) and Diffusion Tensor Imaging (DTI) are needed to explore real-time network activity and white-matter connectivity. Ultimately, these insights could lead to personalized rehabilitation programs tailored to a patient's unique neuroanatomical profile, optimizing the impact of behavioral therapy in the face of progressive decline.

Summary updated

BibTeX
@inproceedings{dial2022baseline,
  author = {Heather Dial and Eduardo Europa and Stephanie M. Grasso and Maria Luisa Mandelli and Kristin M. Schaffer and H. Isabel Hubbard and Lisa Wauters and Lindsey Wineholt and Stephen M. Wilson and Maria Luisa Gorno‐Tempini and Maya L. Henry},
  title = {{Baseline structural imaging correlates of treatment outcomes in semantic variant primary progressive aphasia}},
  booktitle = {Cortex},
  year = {2022},
  doi = {10.1016/j.cortex.2022.10.004},
}
Figure