Elsevier

Brain and Language

Volume 131, April 2014, Pages 11-19
Brain and Language

Longitudinal changes in brains of patients with fluent primary progressive aphasia

https://doi.org/10.1016/j.bandl.2013.05.012Get rights and content

Highlights

  • Longitudinal analysis of performance, brain structure, and function in PPA over 1 year.

  • Consistent high-level performance in a visual lexical word-pseudoword decision task.

  • Consistent signal increase for BOLD lexicality effect in left IFG/IFS and MTG.

  • Consistent and ongoing atrophy in the very same regions.

  • These findings hint at a mechanism for temporally compensating brain atrophy in PPA.

Abstract

Primary progressive aphasia (PPA) is a rare clinical dementia syndrome with predominant, progressive language impairment. Clinical symptoms, linguistic impairment and the course of the disease may vary considerably between patients. In order to capture these aspects, longitudinal assessments of neurofunctional changes in PPA including their relationship to behaviour and clinical symptoms are mandatory, ideally at intervals shorter than 1 year. Here, we report a longitudinal fMRI study investigating the development of lexical processing and their neural basis in PPA patients over 1 year. Four logopenic PPA patients and four matched controls were scanned 3 times (T1, T2, T3, at 6 months intervals) while performing a visual lexical decision task on German words and pseudowords. Group differences for the lexicality effect (pseudowords > words) were assessed at time point T1 and its longitudinal changes in the BOLD signal associated with the lexicality effect were analysed. Brain atrophy was assessed with a high-resolution MPRAGE sequence and analysed using deformation based morphometry (DBM). From the very beginning of the study, PPA patients showed reduced left-hemispheric and increased right-hemispheric activations compared to controls. During the progression of the disease, activation increased predominantly in left posterior middle temporal gyrus (pMTG) and inferior frontal junction area, whereas the same regions decreased in activity in control brains. Interestingly, DBM data showed that this increase in activation in PPA patients was accompanied by progressing atrophy in the same regions. At a behavioural level, the accuracy in the lexical decision task was comparably high for both groups during the whole period of examination, despite some large variability between patients. To conclude, the dissociation between (i) maintained high performance, (ii) increased activity in regions involved in lexical access such as pMTG, and (iii) progressive atrophy of the very same regions supports the notion of a compensatory mechanism in brains of PPA patients for maintaining language while brain atrophy is progressing. The activity increase within a left-lateralised fronto-temporal network seems vital for high-level performance, whereas initial right-hemispheric recruitment of homologue language regions, which is reminiscent of that in vascular aphasics, has no continuous impact on lexical performance.

Introduction

Primary progressive aphasia (PPA) is a rare type of dementia, belonging to neurodegenerative disorders with an onset between 60 and 70 years of age. It was first described by Mesulam (1982). The predominant impairment is in the language domain. During the course of PPA, other cognitive deficits can emerge after about 2 years, but aphasia remains the most severe impairment (Mesulam, 2001) and finally results in mutism (Gorno-Tempini et al., 2004). Because of its circumscribed deficit with relative slow progression of linguistic symptoms, PPA may reveal especially in its early phase, compensatory mechanisms of structure–function interrelation. Detecting the relationship between cognitive, neurofunctional, and neuroanatomical parameters in PPA may thus improve our understanding of the plasticity of the brain under pathological conditions. Consequently, this knowledge may help formulate a more general model which extends to other degenerative diseases and which may be the ground to understand the role of neurofunctional changes e.g. in other types of dementia.

In the literature, several attempts have been made to classify subtypes of PPA according to their predominant aphasic symptoms, e.g. syntactic/agrammatic deficits vs. semantic or phonological retrieval problems (e.g. Sonty et al., 2003). The most widely agreed-upon distinction pertains to (at least) one non-fluent/agrammatic type and one fluent type with word finding difficulties. Some authors (e.g. Snowden, Neary, & Mann, 1996; Neary et al., 1998; Hodges, Patterson, Oxbury, & Funnell, 1992) further distinguish between this fluent PPA variant and semantic dementia (SD): Whereas in SD, semantic memory loss is generally evident in verbal as well as non-verbal tasks, patients with fluent PPA only have circumscribed deficits in lexical semantics but are unimpaired in non-verbal semantic tasks. The distinct symptoms in subtypes of PPA are associated with different patterns of atrophy in frontal (non-fluent PPA) vs. temporal (fluent PPA) cortices (e.g. Mesulam et al., 2009). Often atrophy emerges in parallel with hypoperfusion (Perneczky, Diehl-Schmid, Pohl, Drzezga, & Kurz, 2007). Gorno-Tempini et al. (2011) have recently published a new guideline for the classification of PPA in non-fluent/agrammatic, semantic, and logopenic PPA.

The underlying neurofunctional mechanism of progression of aphasic symptoms in PPA remains largely unknown. Cross-sectional studies of cerebral blood flow showed differences between PPA patients in different stages of progression; however, these differences intermingled with regional perfusion changes in healthy aging subjects (Meinzer et al., 2009). Evidence for a reduced structural and functional connectivity in brains of PPA patients likewise rely on cross-sectional rather than longitudinal data (Oliveira et al., 2011, Sonty et al., 2007). In order to distinguish pathological changes from normal age-related changes, longitudinal examinations are mandatory (e.g. Rogalski et al., 2011), which should cover short time intervals. In addition, such studies are capable of disclosing the course of pathological changes in the individual patient, a necessary prerequisite towards personalised therapy (e.g. Dressel et al., 2010).

The present study was designed to address longitudinal changes in language performance and underlying neurofunctional changes in PPA. Over the course of 1 year, fluent (logopenic) PPA patients and matched controls were repeatedly examined at intervals of 6 months (time point T1: 0 months; time point T2: 6 months; time point T3: 12 months). In this study, we employed a lexical decision task including a battery of words and pseudowords. This lexical decision task, which was performed during fMRI scanning, tested for skills of receptive language processing, and avoided problems with expressive language in later stages of the disease. The “lexicality effect” (pseudowords vs. words), a well-established indicator of visual-lexical processing (Binder et al., 2005, Carreiras et al., 2007) was used to investigate lexical processing in fluent PPA. Based on findings in the literature, the following hypotheses were tested: (1) Considering that atrophy is often associated with hypoperfusion (Acosta-Cabronero et al., 2011, Perneczky et al., 2007) indicating reduced neuronal activity (Hughes, Nestor, Hodges, & Rowe, 2011), we tested the hypothesis that brains of PPA patients show hypoperfusion (and hence, reduced BOLD signal) relative to controls, which becomes more pronounced over time. (2) During aging (Meinzer et al., 2009) as well as in aphasic patients after stroke (e.g. Saur et al., 2006), an additional recruitment of right-hemispheric regions was observed for language processing. In this context we asked the question about comparable effects during disease progression. (3) Finally, we examined whether the differences in functional activations found in the aforementioned experiments were associated with local structural changes. To this end, relative changes of the activation clusters found in the fMRI experiments were computed by means of longitudinal deformation field morphometry (Pieperhoff et al., 2008).

Section snippets

Participants

Four fluent logopenic PPA patients1 (mean age ± SD: 65.8 ± 7.0 years; 2 women; all right-handed according to the Edinburgh Inventory by Oldfield, 1971) and four matched control subjects (66.5 ± 6.5 years; 2 women) were examined repeatedly on three occasions over the course of 1 year. The details about the individual patients and control subjects are provided in Table 1. The

Performance during lexical decisions

Response accuracy in the lexical decision task was at ceiling for both groups in all examinations. Consequently, there were no significant differences between PPA patients and controls at any examination (T1: Mann–Whitney U = 1.0; p = 0.057; T2: Mann–Whitney U = 4.5; p = 0.343; T3: Mann–Whitney U = 4.5; p = 0.343; all p-values not significant at thresholds uncorrected for multiple comparisons). In particular, the initial trend for a group difference disappeared at later examinations. However, there was

Activation differences between PPA patients and controls at T1

The comparison of PPA patients and control subjects at the initial examination (T1) revealed that PPA patients had higher amplitudes for the lexicality effect than controls in the right posterior middle temporal gyrus (pMTG), extending into posterior superior temporal gyrus, and in right inferior parietal lobule. A second maximum was found in the right occipital lobe (for details see Table 2 and Fig. 2a). Reversely, controls had higher amplitudes for the lexicality effect than PPA patients in

Discussion

The present fMRI study sought to elucidate neurofunctional changes in the course of fluent PPA by means of a longitudinal approach involving short between-test intervals. Investigating four PPA patients and the same number of matched control subjects for three times over the course of 1 year, we provide fine-grained quantitative data of neuronal responses and their temporal dynamics in brains of patients with PPA.

Conclusion

The present study demonstrates unique neurofunctional changes in PPA when examined longitudinally and at short between-examination intervals. The data show that the loci of neurofunctional dynamics also show progressive atrophy. Interestingly, longitudinal activation increase for lexical processing, rather than decrease, was observed in brain regions increasingly affected by neurodegeneration, hinting at a compensatory mechanism. Moreover, our study shows that right-hemispheric involvement in

Acknowledgments

We wish to thank Yosef Grodzinsky for inspiring discussions about the different facets of PPA. Moreover, we are grateful for support by the Department of Neurology at RWTH Aachen University. This project was supported by a JARA – Translational Brain Medicine grant to MG and SH, and the “Stiftung für Altersforschung”, Heinrich-Heine-University, Düsseldorf, Germany (Grant to MS). The project underlying this report “Helmholtz Alliance for Mental Health in an Ageing Society” (HelMA, HA-215) was

References (49)

  • J. Acosta-Cabronero et al.

    Atrophy, hypometabolism and white matter abnormalities in semantic dementia tell a coherent story

    Brain

    (2011)
  • K. Amunts et al.

    Cytoarchitectonic mapping of the human amygdala, hippocampal region and entorhinal cortex: Intersubject variability and probability maps

    Anatomy and Embryology

    (2005)
  • K. Amunts et al.

    Broca’s region: Novel organizational principles and multiple receptor mapping

    PLoS Biology

    (2010)
  • K. Amunts et al.

    Broca’s region revisited: Cytoarchitecture and intersubject variability

    Journal of Computational Neurology

    (1999)
  • M. Carreiras et al.

    Brain activation for lexical decision and reading aloud: Two sides of the same coin?

    Journal of Cognitive Neuroscience

    (2007)
  • S. Caspers et al.

    The human inferior parietal lobule in stereotaxic space

    Brain Structure and Function

    (2008)
  • J. Derrfuss et al.

    Neural activations at the junction of the inferior frontal sulcus and the inferior precentral sulcus: Interindividual variability, reliability, and association with sulcal morphology

    Human Brain Mapping

    (2009)
  • K. Dressel et al.

    Model-oriented naming therapy in semantic dementia: A single-case fMRI study

    Aphasiology

    (2010)
  • S. Geyer et al.

    Areas 3a, 3b, and 1 of human primary somatosensory cortex: 2. Spatial normalization to standard anatomical space

    Neuroimage

    (1999)
  • M.L. Gorno-Tempini et al.

    Cognition and anatomy in three variants of primary progressive aphasia

    Annals of Neurology

    (2004)
  • M.L. Gorno-Tempini et al.

    Classification of primary progressive aphasia and its variants

    Neurology

    (2011)
  • S. Heim et al.

    Modality-independent involvement of the left BA 44 during lexical decision making

    Brain Structure and Function

    (2007)
  • Heim, S., Seidel, B., Etcheverry, L., Kuijsten, W., Grande, M., Schulte, S., Fimm, B., Südmeyer, M., Pieperhoff, P.,...
  • G. Hickok et al.

    The cortical organization of speech processing

    Nature Reviews Neuroscience

    (2007)
  • Cited by (11)

    • A longitudinal study of speech production in primary progressive aphasia and behavioral variant frontotemporal dementia

      2019, Brain and Language
      Citation Excerpt :

      Additional work is needed to determine the basis for bvFTD patients’ decline on features of connected speech production. A number of recent studies have investigated the longitudinal progression of brain atrophy in patients with PPA and bvFTD (Brambati et al., 2015; Etcheverry et al., 2012; Faria et al., 2014; Heim et al., 2014; Knopman et al., 2009; Olm et al., 2016; Rogalski et al., 2014; Rohrer et al., 2013; Tetzloff et al., 2018; Tu et al., 2016), but few of these reports have related longitudinal language change to imaging. We identified areas of significant cortical thinning in the combined cohort of PPA and bvFTD patients at Time 1 and Time 2 compared to controls, and we evaluated the association of cortical thinning in those areas to decline on language measures.

    • Test-retest reliability in an fMRI study of naming in dementia

      2019, Brain and Language
      Citation Excerpt :

      As individuals with dementia experience the neurodegenerative process, which includes atrophy and other neuropathology over time (Brambati et al., 2015; Rogalski et al., 2011; Staffaroni et al., 2017; Whitwell et al., 2015), it is possible that they would show dissimilar brain activation patterns at two different points, even without intervention. For instance, a longitudinal study investigated functional brain changes in four logopenic variant primary progressive aphasia (lvPPA) participants over the course of 1 year, repeating fMRI measures every 6 months in the absence of any intervention (Heim et al., 2014). The participants performed a visual lexical decision task in the scanner.

    • Neuropsychological Profiles Differ among the Three Variants of Primary Progressive Aphasia

      2015, Journal of the International Neuropsychological Society
    View all citing articles on Scopus
    View full text