Disorganization Of Language And Working Memory Systems in Frontal Versus Temporal Lobe Epilepsy Part 2

Correlation of fMRI measures with clinical variables

For language tasks in FLE, longer epilepsy duration related to (i) higher verbal fluency fMRI activity in a rostral left middle frontal area bordering the classical activation map (PFWE<0.05), possibly reflecting compensatory recruitment; (ii) lesser deactivation of the whole DMN (sperm =0.28, Punc=0.049; Fig. 7B); and (iii) lesser deactivation at the gradient apex (bins 18–20, sperm range=0.31–0.37, Punc=0.011–0.029; Fig. 7C). For verb generation, longer duration related to lesser posterior temporal and angular deactivation (PFWE< 0.05; Fig. 7A); longer time since last seizure correlated with lower anterior temporal activation (PFWE<0.05), while higher seizure frequency was associated with lower deactivation at the gradient apex (bins 19–20, rpermrange=0.28–0.29, Punc=0.046–0.040). For working memory in FLE, we found a significant association between lower bilateral parietal activation during the verbal task and longer disease duration (PFWE<0.05; Fig. 7D). This was paralleled by positive correlations between age at onset and (i) frontoparietal control activity (sperm =0.33, Punc=0.025; Fig. 7E); as well as (ii) gradient-based profiles (bins 12–18, sperm range=0.30–0.35, Punc=0.026–0.048; Fig. 7F) during verbal working memory. History of FBTCS related to lower right dorsolateral frontal activation (PFWE<0.05).

There is indeed a strong connection between frontoparietal control and memory, a connection that has been confirmed by many scientific studies. The frontoparietal area is part of the cerebral cortex. It is responsible for processing and controlling our cognition, thinking, emotional regulation, and other functions. It also has a crucial impact on memory.

Scientists have found through experiments that patients with damage to the frontoparietal area usually experience a significant memory decline. After training, the frontoparietal area can be effectively exercised and improved, which not only helps enhance the brain's cognitive ability, and improves emotional regulation ability, but also helps improve memory.

So, how to effectively exercise the frontoparietal area? Here are some ways:

1. Change the way of thinking: Gradually develop positive thinking habits and cultivate an optimistic attitude. Positive thinking and positive emotions stimulate the frontoparietal area, thereby improving the brain's cognitive and memory abilities.

2. Memory training: Creative memory training can help improve the brain's cognitive ability and memory. Common memory training includes associative memory, location memory, logical memory, image memory, etc.

3. Exercise your brain: Intellectual games, puzzles, mathematical games, etc. can exercise your brain's cognitive ability and memory. These games require a lot of thinking and judgment and give the frontoparietal area a good workout.

In short, the impact of the frontoparietal area on memory is very significant. In daily life, we should pay attention to maintaining positive thinking and optimism, and perform regular memory training and mental exercises, to help improve our memory. Let’s be positive together and have better brain health and memory! It can be seen that we need to improve memory, and Cistanche deserticola can significantly improve memory, because Cistanche deserticola can also regulate the balance of neurotransmitters, such as increasing the levels of acetylcholine and growth factors. These substances are very important for memory and learning. In addition, Meat can also improve blood flow and promote oxygen delivery, which can ensure that the brain receives sufficient nutrients and energy, thereby improving brain vitality and endurance.

Click know supplements to improve memory

For language tasks in TLE, earlier age at onset was associated with lower inferior frontal activation during verbal fluency. Complementing our prior work,73 history of FBTCS was associated with lower activation of temporal, angular and DMN areas during verbal fluency (PFWE<0.05), mapping onto (i) dorsal attention, frontoparietal control, DMN, limbic, visual and somatomotor systems (sperm = −0.33/−0.29/−0.36/−0.55/−0.34/−0.45, PFDR=0.010/0.026/ 0.010/<0.001/0.010/0.010); and (ii) most gradient bins (bins 2–20, rpermrange= −0.28 to −0.39, PFDR=0.016–0.033). For verb generation, the history of FBTCS is positively related to right posterior temporal activation (PFWE<0.05), possibly reflecting compensatory recruitment. For working memory in TLE, FBTCS were associated with effects across premotor/precentral areas (PFWE<0.05) during the verbal task. For visual working memory (1–0 Back), longer duration related to lesser deactivation of posterior DMN areas (PFWE<0.05); for the 2–1 Back contrast, longer time since last seizure related to higher activity of (i) dorsal attention, salience and somatomotor systems (ρperm =0.33/0.36/0.38, PFDR=0.034/0.020/0.020); and (ii) unimodal to intermediate gradient sections (bins 3–9, ρperm range= 0.35– 0.39, PFDR=0.019–0.023).

Sensitivity analyses

Extensive sensitivity analyses (detailed in the Supplementary material) corroborated the robustness of the above results to (i) different systems-level parcellation; (ii) pathology-informed subgroup allocation (dysplasia-related FLE); (iii) lesional status; (iv) frontal language laterality; and clinical characteristics such as (v) laterality of seizure focus; (vi) seizure frequency; (vii) history of FBTCS; and (iii) time since last seizure.

Discussion

Knowledge of the neural substrates of cognitive impairment in FLE is scarce. Here, we profiled the neural correlates of language and working memory impairment in a large FLE sample. Additional analysis of a TLE group allowed decoding of shared and syndrome-specific effects. Our findings indicate impaired fronto-temporo-parietal activation and impaired DMN deactivation in FLE, with global disorganization of task-related recruitment during working memory, and implicate areas across a broad spectrum of functional specialization. While patterns of impairment largely overlapped across syndromes, we found more prominent alterations of DMN deactivation in FLE, and more marked alterations of temporal activity in TLE during verb generation, which entails semantic processing. Task-related functional signatures were detrimentally modulated by clinical characteristics both in FLE and TLE. This study conveys a comprehensive characterization of the neural correlates of cognitive impairment in FLE, paving the way for future analyses of disease subgroups. We identify neural targets that may aid future investigations into cognitive prognostics and inform the development of novel rehabilitation and therapeutic approaches.

Neuropsychological profiling indicated generalized cognitive impairment in FLE, with poorer performance for functions typically ascribed to the frontal lobes, including working memory, verbal fluency, and mental flexibility, as well as weaker verbal memory and naming, processes that rely more markedly on temporal lobe function. A comparison of FLE and TLE showed more prominent executive dysfunction in the former and more marked impairment of verbal learning and semantic knowledge in the latter. Our findings corroborate evidence of dysexecutive traits and memory difficulties in FLE8,9,11 and indicate that cognitive profiles in FLE and TLE differ. Echoing prior work,4,7,14 we suggest that functions that more strongly recruit areas overlapping with the epileptic network are affected more pervasively.

Traditional voxel-based fMRI maps provide fine-grained accounts of regional group differences but cannot capture large-scale effects. Using a multiscale approach, we attained systems-level inference35,71,76 and profiled the landscape of brain activation and deactivation patterns on the backbone of the principal gradient of intrinsic connectivity.36,77 The gradient describes a continuous transition from unimodal sensory to transmodal areas, offering a compact framework to probe task-related effects in the context of cognitive system hierarchies and detect global differences in cognition-related brain activity.39,40,66 Our analyses thus enabled the decoding of the neural signatures of cognitive processes and related functional reorganization in epilepsy, capitalizing on a ‘local-to-global’ perspective.

In controls, language tasks elicited left-lateralized activation of the middle frontal, inferior frontal, and temporoparietal cortices, which positively correlated with fluency and naming scores. Systems-level effects captured frontoparietal control and opercular involvement, and activity shifts along intermediate-to-transmodal gradient segments implicated a combination of attentional and high-level executive processing. Attenuation of DMN activity was extensive for verbal fluency, which requires executive control,15,78 was tracked by negative effects at the gradient apex, and was neurobehaviourally relevant, as demonstrated by correlations analyses with out-of-scanner performance. Temporo-parietal activation was marked during verb generation, owing to its semantic demands,79,80, and correlated with naming scores.

Across language tasks, we found reduced left inferior frontal and middle frontal activation in FLE. Frontal language lateralization was also weaker in FLE than controls, corroborating findings of prior case series.81,82 Repeat comparisons controlling for language laterality, however, excluded a substantial influence of interhemispheric frontal language organization on the observed group differences. Thus, dysfunction of the left frontal language core may be a key feature underlying impaired expressive language in FLE. From a neurobiological perspective, such dysfunction may stem from epileptic activity of frontal origin, which may have adverse effects on synaptic connections and local computation. Exploratory analyses identified similar disruptions of left-hemispheric activation in the left and right FLE. Involvement of left-sided language areas and verbal deficits were previously described in people with right TLE, which led to reconsidering earlier views on the sparing of verbal functions in right hemisphere pathology.15,83–85 Similarly, it is possible that left frontal language alterations in right FLE may be a consequence of transcallosal propagation of epileptic activity,86 given the rapid bilateral spread of frontal lobe seizures.1,87

FLE and TLE groups had similar frontal abnormalities during verbal fluency. Frontal language dysfunction thus represents a shared trait, possibly a downstream consequence of both proximal (frontal) and more distal (temporal) pathology. Notably, we and others previously showed abnormal frontal activation and frontotemporal connectivity during expressive language in TLE,21,22,88,89 which may emerge as a propagated abnormality, mediated by microstructural alterations in perisylvian white matter tracts.58,84,85 Differences between FLE and TLE in the semantic processing stream during verb generation, involving posterior temporoparietal and occipital areas,80,90,91 indicate that impairment during tasks with semantic demands is TLE-specific, providing a correlate to neuropsychological findings. While frontal language areas represent a common substrate of expressive language difficulties in TLE and FLE, temporal pathology may primarily affect areas in its vicinity, such as posterior language centers, leading to more pervasive dysfunction during tasks that specifically rely on these.

In both working memory tasks, we detected bilateral frontoparietal activation30,92 that mapped on dorsal attention and frontoparietal control systems, resulting in activity shifts in the intermediate-to-transmodal gradient sections, and strongly correlated with cognitive scores. For both conditions entailing a 2 Back working memory span, individuals with FLE exhibited bilateral attenuation of frontoparietal activation and altered gradient profiles. Activation reductions were particularly marked for the 2–1 Back contrast and affected cognitive systems beyond those directly implicated in working memory. While indicating global dysfunction, such widespread effects may also reflect loss of motivation and participant disengagement. The verbal working memory task, however, was less challenging and, though performed slightly less accurately by people with FLE than controls, resulted in satisfactory performance (>80%) in all groups. Patients with FLE still presented with frontoparietal hypoactivation, more selective for dorsal attention and cognitive control systems. Thus, we suggest that attenuated frontoparietal activation may more parsimoniously reflect inefficient recruitment of areas required for successful task performance. The low-demand visual working memory contrast highlighted enhanced frontoparietal activation and reduced DMN deactivation in FLE. This sequence of higher and lower activation for easy and difficult task conditions thus points to cognitive system saturation already occurring for low-level task demands, followed by defective additional recruitment for higher task difficulty. Notably, by encompassing altered deactivation of DMN regions, neural processes underlying working memory in FLE already proved inefficient for easier task conditions.

A comparison of FLE and TLE for working memory activation did not indicate marked group differences. Working memory processing engages distributed bilateral fronto-temporo-parietal networks.34,93,94 Propagation of ictal and interictal epileptic activity may lead to long-lasting neural derangements within multiple sites relevant for performance, which all result in less efficient working memory networks, independent of the location of the epileptic focus. Future investigation of recent-onset focal epilepsy may clarify whether the involvement of working memory hubs may be sequential, with earlier effects close to the seizure onset zone.

Across tasks, we showed impaired deactivation of DMN areas in FLE, and to a lesser extent in TLE. The DMN subserves processes including self-awareness, mind-wandering, and cognition supported by internal representations.95–97 DMN deactivation and anticorrelation between DMN and frontoparietal activity were described for executive tasks.32,98,99 Alterations in such processes were previously identified in psychiatric disorders, including autism100,101 and schizophrenia,102,103, and may reflect suboptimal distribution of neural resources during goal-directed cognition.104 We and others previously showed altered DMN deactivation during working memory in juvenile myoclonic epilepsy and pediatric TLE.55,105 Our current findings in FLE underscore the vulnerability of the DMN across the epilepsy spectrum and indicate altered DMN deactivation profiles as a possible trans-syndromic marker. Notably, while intergroup differences encompassed DMN areas, global alterations, affecting the whole DMN, only emerged for visual working memory. The spatially widespread, distributed nature of the DMN in the seven-system partition, along with recent evidence that DMN subsections have distinct functional roles,106 are possible reasons for the lack of group differences cohesively involving the DMN in language tasks.35 In addition, we and others previously documented the influence of ASMs on task-related deactivation during language and working memory.46,47,107–111 Such effects, however, inconstantly involved those midline anterior and posterior DMN areas that were sites of prominent differences between FLE and controls and between FLE and TLE in this study.
Interestingly, alterations of task-related deactivation, mostly encompassing DMN nodes, were more marked in FLE than in TLE. A possible explanation may lie in the differential connectivity profiles of specific DMN hubs, such as the medial prefrontal cortex, that is strongly embedded within midline DMN, and also likely to be involved in the propagation network of frontal seizures.87 Covarying for clinical characteristics did not alter working memory findings but modulated angular and posterior DMN deactivation differences between FLE and TLE during verb generation. Thus, our findings suggest that epilepsy severity may represent an additional contributing factor to differences in language system architecture in FLE and TLE.

Correlation analyses showed that disease load and factors associated with severity, as tracked by age at onset, epilepsy duration, history of FBTCS, and time since last seizure, may modulate the degree of both activation and deactivation profiles in FLE, extending prior work in TLE.73,74 In FLE, effects across areas of activation were particularly evident for working memory, while clinical variables more strongly influenced deactivation patterns during language. Prior work identified detrimental associations between cognitive function and early age at epilepsy onset,112–115 linking them to disrupted white matter maturation and connectivity.116,117 In our study, associations in FLE were marked for functions with a prolonged maturational trajectory, such as working memory, which depends on the myelination of long-range fiber tracts.118,119 Future longitudinal work may shed further light on associations between clinical, cognitive, and neural profiles and their joint modulation by ASMs. Finally, we note that the co-occurrence of impaired frontoparietal activation and DMN deactivation in FLE mirrors prior comparisons of healthy older adults (∼70 years) to younger individuals.120,121 As the mean age in our FLE sample was ∼30 years, it is tempting to speculate that the identified traits may reflect a functional marker of accelerated brain aging. If proven by future work, such a phenomenon would dovetail with recent evidence of accelerated grey matter loss in people with focal epilepsy, which revived the ‘disease progression’ hypothesis.122–125

Our study has several strengths, including the use of large samples, robust methodology, comprehensive mapping of function from local to global perspectives, direct comparison of patient groups, and several sensitivity analyses. Our study also has limitations. The FLE group was heterogeneous in terms of etiology and MRI findings but representative of the spectrum of FLE patients assessed in tertiary centers.3,8,126 Sensitivity analyses in the focal cortical dysplasia-related FLE subgroup largely corroborated our main findings. Further subgroup analyses showed similar differences in task-related effects in lesional and non-lesional FLE compared to controls. However, a direct comparison of patient subgroups provided preliminary evidence for a less efficient functional architecture in lesional FLE, with the relative enhancement of lateral temporal activity during language and higher dysfunction of frontal circuitry subserving working memory, possibly in light of its more prolonged maturational trajectories.118,119 Future work is encouraged to disentangle cognitive alterations specific to distinct aetiologies, grouping larger patient samples based on sub-lobar lesion location or seizure semiology.87 The language tasks were covert, which prevented online performance monitoring. However, these tasks were previously validated,58,88 are extensively used clinically56,127 and captured interindividual differences in out-of-scanner fluency and naming performance in our study. Future studies may benefit from using overt language paradigms that allow investigators to monitor compliance and task performance,20,21,128 which would provide assessments of functional reconfigurations directly subserving task execution. Notably, systems-level analyses did not capture differences between FLE and controls during language tasks. While localized frontal language dysfunction in FLE may simply not translate to large-scale, bilateral abnormalities, thus lacking a systems-level correlate, we also note that the landmark functional atlas we utilized lacks a dedicated language partition, which may have decreased the sensitivity of our analyses. Future work may benefit from employing recently developed functional parcellations that also incorporate a language system.129 Individuals with FLE and TLE were drug-resistant, taking ASMs in various combinations. We and others previously described compound- and syndrome-specific effects of ASMs on cognitive networks.107,109–111,130 Patient groups were however balanced for medications with known detrimental (topiramate/zonisamide) or more favorable (levetiracetam) cognitive network profiles.46,47 As for comparisons with controls, it is possible that impaired task-related activation in FLE (and TLE) may be partially influenced by ASMs. As discussed earlier, however, ASM-related effects on task deactivation patterns inconsistently involve the midline areas that showed marked intergroup differences in this study. Finally, we identified associations of age at seizure onset, disease duration, FBTCS, and time since the last seizure with task-related imaging phenotypes, which suggests disease-related mediating factors other than ASMs. Future longitudinal work, including the assessment of drug-naïve patients, may better characterize the contribution of ASMs to cognitive system reorganization in epilepsy.

In conclusion, our study decodes neural processes underlying language and working memory impairment in FLE, showing local, systems-level, and global abnormalities that indicate an altered interplay between task-related cognitive system activation and deactivation. While patterns of dysfunction in FLE and TLE largely overlap, the activity of posterior language centers is more affected in TLE relative to FLE, and profiles of default-mode deactivation are more impaired in FLE. This work bridges a substantial knowledge gap in the epilepsy literature and delivers neural markers that can be validated in the context of cognitive prognostics, serving as targets for the future development of targeted treatments.

Acknowledgments

We thank our patients and controls for their participation in this study, and the radiographers at the Epilepsy Society MRI Unit, Philippa Bartlett, Jane Burdett, and Elaine Williams for their help during data acquisition. We also thank Jason Stretton for data acquisition. Dr. Jonathan O’Muircheartaigh and Professor Mark P. Richardson are acknowledged for previous collaborative research initiatives. Alexander Lowe is acknowledged for insightful discussions during prior work. Drs Sara Larivière and Reinder Vos De Wael are acknowledged for helpful conversations on functional gradients. Drs Linden Parkes and Julia K. Brynildsen are acknowledged for insightful discussions.

Funding

Participant recruitment was funded through the Wellcome Trust (Project Grants No 079474 and 083148 to M.J.K. and J.S.D.). We are grateful to the Epilepsy Society for supporting the Epilepsy Society MRI scanner. This research was also supported by the NIHR University College London Hospitals Biomedical Research Centre. L.C. acknowledges support from a Brain Research UK PhD scholarship (Award 14181) and from a Berkeley Fellowship jointly awarded by University College London and Gonville and Caius College, Cambridge. X.H. acknowledges grant support from the American Epilepsy Society. K.T. was supported by fellowships from the European Academy of Neurology and the Austrian Society of Neurology (OEGN). G.P.W. was supported by the MRC (G0802012, MR/M00841X/1). L.C., X.H., and D.S.B. acknowledge support from the NINDS (R01-NS099348); D.S.B. acknowledges support from the John D. and Catherine T. MacArthur Foundation, the Alfred P. Sloan Foundation, the Paul Allen Family Foundation and the Institute for Scientific Interchange (ISI Foundation). B.C.B. acknowledges research funding from the SickKids Foundation (NI17-039), the National Sciences and Engineering Research Council of Canada (NSERC; Discovery-1304413), CIHR (FDN-154298), Azrieli Center for Autism Research (ACAR-TACC), BrainCanada the Montreal Neurological Institute and Hospital, as well as Fonds de la recherche du Quebec– Santé (FRQ-S) and Canada Research Chairs. The funders had no role in study design, data collection, and analysis, the decision to publish, or the preparation of the manuscript.

Competing interests

The authors report no competing interests.

Supplementary material

Supplementary material is available at Brain online.

References

1. Jobst BC, Siegel AM, Thadani VM, Roberts DW, Rhodes HC, Williamson PD. Intractable seizures of frontal lobe origin: clinical characteristics, localizing signs, and results of surgery. Epilepsia. 2000;41:1139–1152.

2. O’Muircheartaigh J, Richardson MP. Epilepsy and the frontal lobes. Cortex. 2012;48:144–155.

3. Lee JJ, Lee SK, Lee SY, et al. Frontal lobe epilepsy: clinical characteristics, surgical outcomes, and diagnostic modalities. Seizure. 2008;17:514–523.

4. Lin JJ, Mula M, Hermann BP. Uncovering the neurobehavioural comorbidities of epilepsy over the lifespan. Lancet. 2012;380: 1180–1192.

5. Stretton J, Thompson PJ. Frontal lobe function in temporal lobe epilepsy. Epilepsy Res. 2012;98:1–13.

6. Hermann BP, Struck AF, Busch RM, Reyes A, Kaestner E, McDonald CR. Neurobehavioural comorbidities of epilepsy: Towards a network-based precision taxonomy. Nat Rev Neurol. 2021;17:731–746.

7. Upton D, Thompson PJ. General neuropsychological characteristics of frontal lobe epilepsy. Epilepsy Res. 1996;23:169–177.

8. Helmstaedter C, Kemper B, Elger CE. Neuropsychological aspects of frontal lobe epilepsy. Neuropsychologia. 1996;34:399–406.

9. Patrikelis P, Angelakis E, Gatzonis S. Neurocognitive and behavioral functioning in frontal lobe epilepsy: a review. Epilepsy Behav. 2009;14:19–26.

10. Risse GL. Cognitive outcomes in patients with frontal lobe epilepsy. Epilepsia. 2006;47:87–89.

11. Centeno M, Thompson PJ, Koepp MJ, Helmstaedter C, Duncan JS. Memory in frontal lobe epilepsy. Epilepsy Res. 2010;91:123–132.

12. Exner C, Boucsein K, Lange C, et al. Neuropsychological performance in frontal lobe epilepsy. Seizure. 2002;11:20–32.

13. Patrikelis P, Gatzonis S, Siatouni A, et al. Preoperative neuropsychological presentation of patients with refractory frontal lobe epilepsy. Acta Neurochir (Wien). 2016;158:1139–1150.

14. McDonald CR, Delis DC, Norman MA, Wetter SR, Tecoma ES, Iragui VJ. Response inhibition and set-shifting in patients with frontal lobe epilepsy or temporal lobe epilepsy. Epilepsy Behav. 2005;7:438–446.

15. Metternich B, Buschmann F, Wagner K, Schulze-Bonhage A, Kriston L. Verbal fluency in focal epilepsy: a systematic review and meta-analysis. Neuropsychol Rev. 2014;24:200–218.

For more information:1950477648nn@gmail.com