Machine learning for predicting levetiracetam treatment response in temporal lobe epilepsy

Croce, Pierpaolo; Ricci, Lorenzo; Pulitano, Patrizia; Boscarino, Marilisa; Zappasodi, Filippo; Lanzone, Jacopo; Narducci, Flavia; Mecarelli, Oriano; Lazzaro, Vincenzo Di; Tombini, Mario; Assenza, Giovanni

doi:10.1016/j.clinph.2021.08.024

Cited by 23 publications

(25 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instance, while Pellegrino et al ( 37 ) showed no difference between groups after Bonferroni correction, Ricci et al ( 38 ) found differences only in theta power, which was lower in people with epilepsy. Another study by Croce et al ( 39 ) deployed machine learning tools to predict the clinical response to anti-seizure medications using the resting-EEG recordings of people with temporal lobe epilepsy. The authors found they could predict the clinical response to an anti-seizure medication (levetiracetam), from the patients' EEG phenotype.…”

Section: Discussionmentioning

confidence: 99%

Dynamic coupling between the central and autonomic cardiac nervous systems in patients with refractory epilepsy: A pilot study

Melo

Fiel

Milhomens³

et al. 2022

Front. Neurol.

View full text Add to dashboard Cite

The heart and brain are reciprocally interconnected and engage in two-way communication for homeostatic regulation. Epilepsy is considered a network disease that also affects the autonomic nervous system (ANS). The neurovisceral integration model (NVM) proposes that cardiac vagal tone, indexed by heart rate variability (HRV), can indicate the functional integrity of cognitive neural networks. ANS activity and the pattern of oscillatory EEG activity covary during the transition of arousal states and associations between cortical and autonomic activity are reflected by HRV. Cognitive dysfunction is one of the common comorbidities that occur in epilepsy, including memory, attention, and processing difficulties. Recent studies have shown evidence for the active involvement of alpha activity in cognitive processes through its active role in the control of neural excitability in the cortex through top-down modulation of cortical networks. In the present pilot study, we evaluated the association between resting EEG oscillatory behavior and ANS function in patients with refractory epilepsy. Our results show: (1) In patients with refractory epilepsy, there is a strong positive correlation between HRV and the power of cortical oscillatory cortical activity in all studied EEG bands (delta, theta, alpha, and beta) in all regions of interest in both hemispheres, the opposite pattern found in controls which had low or negative correlation between these variables; (2) higher heartbeat evoked potential amplitudes in patients with refractory epilepsy than in controls. Taken together, these results point to a significant alteration in heart-brain interaction in patients with refractory epilepsy.

show abstract

Section: Discussionmentioning

confidence: 99%

Dynamic coupling between the central and autonomic cardiac nervous systems in patients with refractory epilepsy: A pilot study

Melo

Fiel

Milhomens³

et al. 2022

Front. Neurol.

View full text Add to dashboard Cite

show abstract

“…The research team retrospectively reviewed data from newly diagnosed TLE patients enrolled at the epilepsy clinic of Department of Human Neurosciences of Policlinico Umberto I University Hospital of Rome and of Campus Bio-Medico University of Rome between January 2016 and January 2021. The data have been previously used for other studies from our group and selection criteria for patients in our cohort can be found elsewhere (Croce et al 2021 ). The study protocol received approval by the ethics committee of Policlinico Umberto I Ethic Board-Rome- and Campus Biomedico University Ethic Board-Rome.…”

Section: Methodsmentioning

confidence: 99%

“…Pharmaco-EEG studies in epilepsy have usually focused on assessing frequency modifications induced by old-generation ASMs, either visually or through quantitative analysis (Sannita et al 1989;Xiao 1996, 1997;Höller et al 2018). Only recently, previous works from our group showed that new-generation ASM therapy can induce a "normalization" of the EEG power spectrum and connectivity features in people with different types of epilepsy (Pellegrino et al 2018;Lanzone et al 2021;Ricci et al 2021) and that such modifications are also predictive of good clinical outcome in TLE (Croce et al 2021). Yet, to our best knowledge, microstate EEG analysis to evaluate the effects of ASMs in people with epilepsy has never been attempted before.…”

Section: Microstate Metricsmentioning

confidence: 99%

“…This work was partially supported by the 'Search of Excellence', initiative of the Department of Neuroscience, Imaging and Clinical Sciences (DNISC) of the University of Chieti-Pescara. The authors acknowledge that the cohort of patients used to estimate resting-state EEG microstates was also employed in a previous work from our group (Croce et al 2021).…”

Section: Acknowledgementsmentioning

confidence: 99%

“…Indeed, pharmaco-EEG has numerous advantages as an analytic tool in that it can provide a multidimensional approach for the evaluation of brain activity by assessing the dynamics of several features at the same time (i.e., frequency power, connectivity and complexity analysis) (Tong and Thankor 2009 ) and its use in epilepsy research is still far to be fully exploited (Höller et al 2018 ). Pharmaco-EEG analysis applied to epilepsy has the potential to effectively predict therapeutic efficacy (Croce et al 2021 ) and to objectively measure the neurotoxicity of ASMs (Saletu et al 1986 ).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Levetiracetam Modulates EEG Microstates in Temporal Lobe Epilepsy

et al. 2022

Self Cite

View full text Add to dashboard Cite

To determine the effects of Levetiracetam (LEV) therapy using EEG microstates analysis in a population of newly diagnosed Temporal Lobe Epilepsy (TLE) patients. We hypothesized that the impact of LEV therapy on the electrical activity of the brain can be globally explored using EEG microstates. Twenty-seven patients with TLE were examined. We performed resting-state microstate EEG analysis and compared microstate metrics between the EEG performed at baseline (EEGpre) and after 3 months of LEV therapy (EEGpost). The microstates A, B, C and D emerged as the most stable. LEV induced a reduction of microstate B and D mean duration and occurrence per second (p < 0.01). Additionally, LEV treatment increased the directional predominance of microstate A to C and microstate B to D (p = 0.01). LEV treatment induces a modulation of resting-state EEG microstates in newly diagnosed TLE patients. Microstates analysis has the potential to identify a neurophysiological indicator of LEV therapeutic activity. This study of EEG microstates in people with epilepsy opens an interesting path to identify potential LEV activity biomarkers that may involve increased neuronal inhibition of the epileptic network.

show abstract

Development and Validation of a Deep Learning Model for Predicting Treatment Response in Patients With Newly Diagnosed Epilepsy

et al. 2022

View full text Add to dashboard Cite

ImportanceSelection of antiseizure medications (ASMs) for epilepsy remains largely a trial-and-error approach. Under this approach, many patients have to endure sequential trials of ineffective treatments until the “right drugs” are prescribed.ObjectiveTo develop and validate a deep learning model using readily available clinical information to predict treatment success with the first ASM for individual patients.Design, Setting, and ParticipantsThis cohort study developed and validated a prognostic model. Patients were treated between 1982 and 2020. All patients were followed up for a minimum of 1 year or until failure of the first ASM. A total of 2404 adults with epilepsy newly treated at specialist clinics in Scotland, Malaysia, Australia, and China between 1982 and 2020 were considered for inclusion, of whom 606 (25.2%) were excluded from the final cohort because of missing information in 1 or more variables.ExposuresOne of 7 antiseizure medications.Main Outcomes and MeasuresWith the use of the transformer model architecture on 16 clinical factors and ASM information, this cohort study first pooled all cohorts for model training and testing. The model was trained again using the largest cohort and externally validated on the other 4 cohorts. The area under the receiver operating characteristic curve (AUROC), weighted balanced accuracy, sensitivity, and specificity of the model were all assessed for predicting treatment success based on the optimal probability cutoff. Treatment success was defined as complete seizure freedom for the first year of treatment while taking the first ASM. Performance of the transformer model was compared with other machine learning models.ResultsThe final pooled cohort included 1798 adults (54.5% female; median age, 34 years [IQR, 24-50 years]). The transformer model that was trained using the pooled cohort had an AUROC of 0.65 (95% CI, 0.63-0.67) and a weighted balanced accuracy of 0.62 (95% CI, 0.60-0.64) on the test set. The model that was trained using the largest cohort only had AUROCs ranging from 0.52 to 0.60 and a weighted balanced accuracy ranging from 0.51 to 0.62 in the external validation cohorts. Number of pretreatment seizures, presence of psychiatric disorders, electroencephalography, and brain imaging findings were the most important clinical variables for predicted outcomes in both models. The transformer model that was developed using the pooled cohort outperformed 2 of the 5 other models tested in terms of AUROC.Conclusions and RelevanceIn this cohort study, a deep learning model showed the feasibility of personalized prediction of response to ASMs based on clinical information. With improvement of performance, such as by incorporating genetic and imaging data, this model may potentially assist clinicians in selecting the right drug at the first trial.

show abstract

Machine learning for predicting levetiracetam treatment response in temporal lobe epilepsy

Cited by 23 publications

References 32 publications

Dynamic coupling between the central and autonomic cardiac nervous systems in patients with refractory epilepsy: A pilot study

Dynamic coupling between the central and autonomic cardiac nervous systems in patients with refractory epilepsy: A pilot study

Levetiracetam Modulates EEG Microstates in Temporal Lobe Epilepsy

Development and Validation of a Deep Learning Model for Predicting Treatment Response in Patients With Newly Diagnosed Epilepsy

Contact Info

Product

Resources

About