Carolina Ferreira-Atuesta
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Carolina Ferreira-Atuesta has a total of 204 co-authors. Here are the top co-authors by number of shared publications:
- Marian Galovic
Department of Neurology, Clinical Neuroscience Center University Hospital and University of Zurich Zurich Switzerland
8 shared publications - John Sidney Duncan
Department of Experimental Psychology, University of Oxford, Life and Mind Building
4 shared publications - Josemir Wanderley Sander
Department of Clinical and Experimental Epilepsy UCL Queen Square Institute of Neurology London UK
4 shared publications - Kai Michael Schubert
Department of Neurology, Clinical Neuroscience Center University Hospital and University of Zurich Zurich Switzerland
4 shared publications Ana Rita Peralta
Department of Neurology, Hospital de Santa Maria, Lisbon, Portugal
3 shared publicationsCarla Bentes
Department of Neurosciences and Mental Health (Neurology), Hospital de Santa Maria-CHULN. Centro de Estudos Egas Moniz, Faculdade de Medicina, Universidade de Lisboa
3 shared publications
Latest Publications
- Open Access
- 30 September 2026
- Published by Elsevier in SSRN Electronic Journal
Association of Hearing Aid Use and Risk of Dementia in Adults with Hearing Loss
Summary Background: Hearing loss is a major modifiable risk factor for dementia, but whether hearing aids lower dementia risk, and in whom, is uncertain. We assessed this across comorbidities.Methods: In this target trial emulation using TriNetX health records, we included adults with hearing loss first diagnosed at age 60 years or older and no previous cognitive impairment. Hearing aid initiators (within 3 years of diagnosis) were propensity score matched 1:1 to non-users, overall and in seven non-overlapping cohorts (epilepsy, stroke, type 2 diabetes, heart failure, chronic kidney disease, migraine, osteoarthritis). The outcome was incident dementia. We also examined dementia, cognition, and MRI in 16 985 adults with hearing impairment in the National Alzheimer's Coordinating Center (NACC).Findings: In 6422 matched pairs, dementia was diagnosed in 396 (6·3%) of 6271 hearing aid users and 423 (6·8%) of 6244 non-users (hazard ratio [HR] 0·89, 95% CI 0·77–1·02); in an intention-to-treat analysis it was 0·92 (0·76–1·11). The HR decreased with older age at diagnosis (ratio 0·92 per 10 years, 0·88–0·97) and was 0·90 (0·85–0·96) at 75 years or older. Only the epilepsy cohort (701 pairs) showed a nominally significant association (HR 0·64, 0·45–0·90). In NACC, hearing aid users had greater temporal-lobe volume and slower ventricular expansion on brain MRI higher Mini-Mental State Examination scores (0·41 points, 0·18–0·65) but not lower dementia risk (HR 0·92, 0·82–1·03).Interpretation: Hearing aid use was not associated with lower dementia risk in the overall population. Any benefit of hearing rehabilitation appears concentrated in older adults and in those with comorbid epilepsy, in whom hearing screening and hearing aid provision warrant consideration.NACC participants gave written informed consent at their Alzheimer's Disease Research Centers under protocols approved by local institutional review boards.Funding None.
- 30 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessAssociation of Hearing Aid Use and Risk of Dementia in Adults with Hearing Loss
Summary Background: Hearing loss is a major modifiable risk factor for dementia, but whether hearing aids lower dementia risk, and in whom, is uncertain. We assessed this across comorbidities.Methods: In this target trial emulation using TriNetX health records, we included adults with hearing loss first diagnosed at age 60 years or older and no previous cognitive impairment. Hearing aid initiators (within 3 years of diagnosis) were propensity score matched 1:1 to non-users, overall and in seven non-overlapping cohorts (epilepsy, stroke, type 2 diabetes, heart failure, chronic kidney disease, migraine, osteoarthritis). The outcome was incident dementia. We also examined dementia, cognition, and MRI in 16 985 adults with hearing impairment in the National Alzheimer's Coordinating Center (NACC).Findings: In 6422 matched pairs, dementia was diagnosed in 396 (6·3%) of 6271 hearing aid users and 423 (6·8%) of 6244 non-users (hazard ratio [HR] 0·89, 95% CI 0·77–1·02); in an intention-to-treat analysis it was 0·92 (0·76–1·11). The HR decreased with older age at diagnosis (ratio 0·92 per 10 years, 0·88–0·97) and was 0·90 (0·85–0·96) at 75 years or older. Only the epilepsy cohort (701 pairs) showed a nominally significant association (HR 0·64, 0·45–0·90). In NACC, hearing aid users had greater temporal-lobe volume and slower ventricular expansion on brain MRI higher Mini-Mental State Examination scores (0·41 points, 0·18–0·65) but not lower dementia risk (HR 0·92, 0·82–1·03).Interpretation: Hearing aid use was not associated with lower dementia risk in the overall population. Any benefit of hearing rehabilitation appears concentrated in older adults and in those with comorbid epilepsy, in whom hearing screening and hearing aid provision warrant consideration.NACC participants gave written informed consent at their Alzheimer's Disease Research Centers under protocols approved by local institutional review boards.Funding None.
- Open Access
- 31 July 2026
- Published by Wiley in European Journal of Neurology
- Vol. 33 (8) , e70711
Predictors of Multiple Sclerosis After Clinically Isolated Syndrome: A Systematic Review and Meta‐Analysis
Background: Most people with multiple sclerosis (MS) present with a clinically isolated syndrome (CIS); however, not all individuals with CIS are subsequently diagnosed with MS. Methods: A systematic literature review was conducted until December 2025. Observational studies of adults with CIS that were later diagnosed with MS were included. Odds ratios (ORs) were pooled using random‐effects meta‐analysis. Heterogeneity was assessed with I 2 , sensitivity analyses with leave‐one‐out procedures, and publication bias with funnel plots and Egger's test. Meta‐regression was performed to explore further sources of heterogeneity when appropriate. Results: Seventy‐two studies with 9915 adults were included. In the meta‐analysis, younger age (OR = 1.6, p < 0.01) and multifocal presentation (OR = 1.55, p = 0.02) were associated with a diagnosis of MS after a CIS. Magnetic resonance imaging findings, including a higher number of T2 lesions (OR = 7.46, p = 0.02), periventricular lesions (OR = 4.08, p < 0.01), corpus callosum lesions (OR = 14.89, p = 0.02), infratentorial lesions (OR = 2.16, p = 0.03), spinal cord lesions (OR = 1.4, p = 0.03), and gadolinium‐enhancing lesions (OR = 1.91, p = 0.01), as well as cerebrospinal fluid inflammatory markers such as oligoclonal bands (OR = 3.57, p < 0.01) and pleocytosis (OR = 3.34, p = 0.02), were also associated with a subsequent diagnosis of MS. Conclusions: This meta‐analysis identified factors associated with an increased likelihood of MS after a CIS. These findings may help identify high‐risk individuals and guide personalized treatment strategies.
- 31 July 2026
- Published by Wiley in European Journal of Neurology
- Vol. 33 (8) , e70711
Open AccessPredictors of Multiple Sclerosis After Clinically Isolated Syndrome: A Systematic Review and Meta‐Analysis
Background: Most people with multiple sclerosis (MS) present with a clinically isolated syndrome (CIS); however, not all individuals with CIS are subsequently diagnosed with MS. Methods: A systematic literature review was conducted until December 2025. Observational studies of adults with CIS that were later diagnosed with MS were included. Odds ratios (ORs) were pooled using random‐effects meta‐analysis. Heterogeneity was assessed with I 2 , sensitivity analyses with leave‐one‐out procedures, and publication bias with funnel plots and Egger's test. Meta‐regression was performed to explore further sources of heterogeneity when appropriate. Results: Seventy‐two studies with 9915 adults were included. In the meta‐analysis, younger age (OR = 1.6, p < 0.01) and multifocal presentation (OR = 1.55, p = 0.02) were associated with a diagnosis of MS after a CIS. Magnetic resonance imaging findings, including a higher number of T2 lesions (OR = 7.46, p = 0.02), periventricular lesions (OR = 4.08, p < 0.01), corpus callosum lesions (OR = 14.89, p = 0.02), infratentorial lesions (OR = 2.16, p = 0.03), spinal cord lesions (OR = 1.4, p = 0.03), and gadolinium‐enhancing lesions (OR = 1.91, p = 0.01), as well as cerebrospinal fluid inflammatory markers such as oligoclonal bands (OR = 3.57, p < 0.01) and pleocytosis (OR = 3.34, p = 0.02), were also associated with a subsequent diagnosis of MS. Conclusions: This meta‐analysis identified factors associated with an increased likelihood of MS after a CIS. These findings may help identify high‐risk individuals and guide personalized treatment strategies.
- Open Access
Association of antiseizure medication with lower amyloid and tau burden
Network hyperexcitability is increasingly implicated in prodromal Alzheimer’s disease and may be suppressed by antiseizure medications (ASMs). ASMs are widely prescribed to older adults, yet whether their use relates to Alzheimer’s-disease biomarkers at the population level is unknown. In 52,537 participants in the National Alzheimer’s Coordinating Center (NACC) study, we compared cerebrospinal-fluid biomarkers, amyloid and tau positron emission tomography (PET) between ASM users and non-users using inverse-probability-of-treatment weighting with gradient-boosted propensity scores. ASM users showed directionally lower amyloid across multiple brain regions, amplifying markedly in APOE ε4 carriers (Centiloid β = −25.7, p = 0.007). All three temporal tau-PET composites were significantly lower in users (META-temporal β = −0.05, p = 0.01). The amyloid finding replicated independently in the the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset (Centiloid β = −8.6, p = 0.01), whereas four comparator drug classes showed no amyloid signal. These convergent observational findings provide a quantitative framework for evaluating ASMs as candidate disease-modifying agents in Alzheimer’s disease.
Open AccessAssociation of antiseizure medication with lower amyloid and tau burden
Network hyperexcitability is increasingly implicated in prodromal Alzheimer’s disease and may be suppressed by antiseizure medications (ASMs). ASMs are widely prescribed to older adults, yet whether their use relates to Alzheimer’s-disease biomarkers at the population level is unknown. In 52,537 participants in the National Alzheimer’s Coordinating Center (NACC) study, we compared cerebrospinal-fluid biomarkers, amyloid and tau positron emission tomography (PET) between ASM users and non-users using inverse-probability-of-treatment weighting with gradient-boosted propensity scores. ASM users showed directionally lower amyloid across multiple brain regions, amplifying markedly in APOE ε4 carriers (Centiloid β = −25.7, p = 0.007). All three temporal tau-PET composites were significantly lower in users (META-temporal β = −0.05, p = 0.01). The amyloid finding replicated independently in the the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset (Centiloid β = −8.6, p = 0.01), whereas four comparator drug classes showed no amyloid signal. These convergent observational findings provide a quantitative framework for evaluating ASMs as candidate disease-modifying agents in Alzheimer’s disease.
- Open Access
Sodium channel blockers are associated with reduced dementia risk in late-onset unexplained epilepsy
Late-onset unexplained epilepsy is a potential harbinger of dementia, likely driven by network hyperexcitability that facilitates amyloid-β release and tau propagation. Dampening this activity with antiseizure medications offers a potential disease modifying strategy, yet whether specific agents differentially alter this neurodegenerative trajectory remains unknown. Here, we emulated a target trial using global real-world federated data on patients with late-onset unexplained epilepsy to compare dementia risk across antiseizure monotherapies. Using data from over 75 million adults aged 55 years or older, we found that sodium channel blockers were associated with a 27% lower hazard of incident all-cause dementia (hazard ratio = 0.73, 95% confidence interval 0.61- 0.88) and 34% lower hazard of Alzheimer's disease (hazard ratio = 0.66, 0.49 -0.88), compared with levetiracetam/brivaracetam. While the class effect was protective, individual agents such as phenytoin, carbamazepine, and lamotrigine showed divergent safety and efficacy profiles. We replicated these findings in both a Down syndrome cohort and the external National Alzheimer's Coordinating Center dataset. Our results suggest that targeting neuronal excitability with sodium channel blockers is associated with lower risk of dementia, prioritizing the repurposing of these agents for dementia prevention trials.
Open AccessSodium channel blockers are associated with reduced dementia risk in late-onset unexplained epilepsy
Late-onset unexplained epilepsy is a potential harbinger of dementia, likely driven by network hyperexcitability that facilitates amyloid-β release and tau propagation. Dampening this activity with antiseizure medications offers a potential disease modifying strategy, yet whether specific agents differentially alter this neurodegenerative trajectory remains unknown. Here, we emulated a target trial using global real-world federated data on patients with late-onset unexplained epilepsy to compare dementia risk across antiseizure monotherapies. Using data from over 75 million adults aged 55 years or older, we found that sodium channel blockers were associated with a 27% lower hazard of incident all-cause dementia (hazard ratio = 0.73, 95% confidence interval 0.61- 0.88) and 34% lower hazard of Alzheimer's disease (hazard ratio = 0.66, 0.49 -0.88), compared with levetiracetam/brivaracetam. While the class effect was protective, individual agents such as phenytoin, carbamazepine, and lamotrigine showed divergent safety and efficacy profiles. We replicated these findings in both a Down syndrome cohort and the external National Alzheimer's Coordinating Center dataset. Our results suggest that targeting neuronal excitability with sodium channel blockers is associated with lower risk of dementia, prioritizing the repurposing of these agents for dementia prevention trials.
- 24 February 2026
- Published by Wolters Kluwer Health in Neurology
- Vol. 106 (4) , e214613
Timing for Starting Antiseizure Medication Withdrawal After Epilepsy Surgery in Adults
More than half of people undergoing epilepsy surgery become seizure-free and may consider withdrawing antiseizure medications (ASMs). Withdrawal practices vary, and the optimal timing remains unclear. We aim to compare seizure relapse risk among individuals initiating ASM withdrawal at different time points after epilepsy surgery. We conducted a multicenter observational cohort study of adults who underwent resective epilepsy surgery between 1990 and 2016 at 12 tertiary centers. Participants were seizure-free before medication withdrawal and had at least 1 year of follow-up. Seizure relapse risk was compared among those initiating withdrawal 1, 2, 3, 4, or 5 years postoperatively vs later. We used propensity score matching for each comparison to adjust for treatment selection bias. Of the 964 people included (51% female; median age at surgery 34 years [interquartile range 26-44]), 446 (46%) began ASM withdrawal in the first year after surgery, 255 (26%) in the second, 110 (11%) in the third, 58 (6%) in the fourth, 29 (3%) in the fifth, and 66 (7%) after the fifth year. After matching, those starting withdrawal in the first (hazard ratio [HR] 1.4; p = 0.003) or second (HR 1.18; p < 0.001) year had a higher risk of relapse than those who withdrew later. Starting withdrawal in the third (HR 1.7; p = 0.12), fourth (HR 1.3; p = 0.45), or fifth (HR 0.17; p = 0.82) year after surgery showed no increase in risk compared with later withdrawal. Long-term outcomes, such as seizure freedom and being entirely off ASMs at the final follow-up, were not substantially associated with withdrawal timing. Initiating ASM withdrawal within the first 2 postoperative years was linked to a higher initial risk of seizure relapse compared with later withdrawal, although long-term outcomes were similar regardless of withdrawal timing. Waiting more than 2 years did not confer additional benefit in reducing seizure risk. Deciding whether and when to withdraw ASMs is a shared process involving individuals, caregivers, and clinicians, balancing preferences, risk of injury, social factors (e.g., driving, work, and supervision), and clinical judgment. Transparent information on risks and benefits is essential. Our findings offer real-world evidence that may inform future evidence-based withdrawal protocols and follow-up strategies.
- 24 February 2026
- Published by Wolters Kluwer Health in Neurology
- Vol. 106 (4) , e214613
Timing for Starting Antiseizure Medication Withdrawal After Epilepsy Surgery in Adults
More than half of people undergoing epilepsy surgery become seizure-free and may consider withdrawing antiseizure medications (ASMs). Withdrawal practices vary, and the optimal timing remains unclear. We aim to compare seizure relapse risk among individuals initiating ASM withdrawal at different time points after epilepsy surgery. We conducted a multicenter observational cohort study of adults who underwent resective epilepsy surgery between 1990 and 2016 at 12 tertiary centers. Participants were seizure-free before medication withdrawal and had at least 1 year of follow-up. Seizure relapse risk was compared among those initiating withdrawal 1, 2, 3, 4, or 5 years postoperatively vs later. We used propensity score matching for each comparison to adjust for treatment selection bias. Of the 964 people included (51% female; median age at surgery 34 years [interquartile range 26-44]), 446 (46%) began ASM withdrawal in the first year after surgery, 255 (26%) in the second, 110 (11%) in the third, 58 (6%) in the fourth, 29 (3%) in the fifth, and 66 (7%) after the fifth year. After matching, those starting withdrawal in the first (hazard ratio [HR] 1.4; p = 0.003) or second (HR 1.18; p < 0.001) year had a higher risk of relapse than those who withdrew later. Starting withdrawal in the third (HR 1.7; p = 0.12), fourth (HR 1.3; p = 0.45), or fifth (HR 0.17; p = 0.82) year after surgery showed no increase in risk compared with later withdrawal. Long-term outcomes, such as seizure freedom and being entirely off ASMs at the final follow-up, were not substantially associated with withdrawal timing. Initiating ASM withdrawal within the first 2 postoperative years was linked to a higher initial risk of seizure relapse compared with later withdrawal, although long-term outcomes were similar regardless of withdrawal timing. Waiting more than 2 years did not confer additional benefit in reducing seizure risk. Deciding whether and when to withdraw ASMs is a shared process involving individuals, caregivers, and clinicians, balancing preferences, risk of injury, social factors (e.g., driving, work, and supervision), and clinical judgment. Transparent information on risks and benefits is essential. Our findings offer real-world evidence that may inform future evidence-based withdrawal protocols and follow-up strategies.
- Top Cited PapersOpen Access
- 1 July 2025
- Published by Wolters Kluwer Health in Stroke
- Vol. 56 (7) , 1748-1757
Association of the Timing and Type of Acute Symptomatic Seizures With Poststroke Epilepsy and Mortality
BACKGROUND: Acute symptomatic seizures (ASyS) increase the risk of epilepsy and mortality after a stroke. The impact of the timing and type of ASyS remains unclear. METHODS: This multicenter cohort study included data from 9 centers between 2002 and 2018, with a final analysis in February 2024. The study included 4552 adults (2005 female; median age, 73 years) with ischemic stroke and no seizure history. Seizures were classified using International League Against Epilepsy definitions. We examined ASyS occurring within 7 days after stroke. The main outcomes were all-cause mortality and epilepsy. Validation of the updated SeLECT score (SeLECT-ASyS) was performed in 3 independent cohorts (Switzerland, Argentina, and Japan) collected between 2012 and 2024, including 74 adults with ASyS. RESULTS: The 10-year risk of poststroke epilepsy ranged from 41% to 94%, and mortality from 36% to 100%, depending on ASyS type and timing. ASyS on stroke onset day had a higher epilepsy risk (adjusted hazard ratio [aHR], 2.3 [95% CI, 1.3–4.0]; P =0.003) compared with later ASyS. Status epilepticus had the highest epilepsy risk (aHR, 9.6 [95% CI, 3.5–26.7]; P P P =0.004) and status epilepticus (aHR, 14.2 [95% CI, 3.5–58.8]; P P =0.02) and in the validation cohort (0.70 versus 0.50; P =0.18). CONCLUSIONS: ASyS timing and type significantly affect epilepsy and mortality risk after stroke, improving epilepsy prediction and guiding patient counseling.
- 1 July 2025
- Published by Wolters Kluwer Health in Stroke
- Vol. 56 (7) , 1748-1757
Top Cited PapersOpen AccessAssociation of the Timing and Type of Acute Symptomatic Seizures With Poststroke Epilepsy and Mortality
BACKGROUND: Acute symptomatic seizures (ASyS) increase the risk of epilepsy and mortality after a stroke. The impact of the timing and type of ASyS remains unclear. METHODS: This multicenter cohort study included data from 9 centers between 2002 and 2018, with a final analysis in February 2024. The study included 4552 adults (2005 female; median age, 73 years) with ischemic stroke and no seizure history. Seizures were classified using International League Against Epilepsy definitions. We examined ASyS occurring within 7 days after stroke. The main outcomes were all-cause mortality and epilepsy. Validation of the updated SeLECT score (SeLECT-ASyS) was performed in 3 independent cohorts (Switzerland, Argentina, and Japan) collected between 2012 and 2024, including 74 adults with ASyS. RESULTS: The 10-year risk of poststroke epilepsy ranged from 41% to 94%, and mortality from 36% to 100%, depending on ASyS type and timing. ASyS on stroke onset day had a higher epilepsy risk (adjusted hazard ratio [aHR], 2.3 [95% CI, 1.3–4.0]; P =0.003) compared with later ASyS. Status epilepticus had the highest epilepsy risk (aHR, 9.6 [95% CI, 3.5–26.7]; P P P =0.004) and status epilepticus (aHR, 14.2 [95% CI, 3.5–58.8]; P P =0.02) and in the validation cohort (0.70 versus 0.50; P =0.18). CONCLUSIONS: ASyS timing and type significantly affect epilepsy and mortality risk after stroke, improving epilepsy prediction and guiding patient counseling.
- Open Access
Association of the timing and type of acute symptomatic seizures with post-stroke epilepsy and mortality
Background Acute symptomatic seizures (ASyS) increase the risk of epilepsy and mortality after a stroke. The impact of the timing and type of ASyS remains unclear. Methods This multicenter cohort study included data from nine centers between 2002 and 2018, with final analysis in February 2024. The study included 4,552 adults (2,005 female; median age 73 years) with ischemic stroke and no seizure history. We examined ASyS occurring within seven days after stroke. Main outcomes were all-cause mortality and epilepsy. Validation in three separate cohorts included 74 adults with ASyS. Results The ten-year risk of post-stroke epilepsy ranged from 41% to 94%, and mortality from 36% to 100%, depending on ASyS type and timing. ASyS on stroke onset day had a higher epilepsy risk (aHR 2.3, 95% CI 1.3-4.0, p=0.003) compared to later ASyS. Status epilepticus had the highest epilepsy risk (aHR 9.6, 95% CI 3.5-26.7, pConclusions ASyS timing and type significantly affect epilepsy and mortality risk after stroke, improving epilepsy prediction and guiding patient counseling.
Open AccessAssociation of the timing and type of acute symptomatic seizures with post-stroke epilepsy and mortality
Background Acute symptomatic seizures (ASyS) increase the risk of epilepsy and mortality after a stroke. The impact of the timing and type of ASyS remains unclear. Methods This multicenter cohort study included data from nine centers between 2002 and 2018, with final analysis in February 2024. The study included 4,552 adults (2,005 female; median age 73 years) with ischemic stroke and no seizure history. We examined ASyS occurring within seven days after stroke. Main outcomes were all-cause mortality and epilepsy. Validation in three separate cohorts included 74 adults with ASyS. Results The ten-year risk of post-stroke epilepsy ranged from 41% to 94%, and mortality from 36% to 100%, depending on ASyS type and timing. ASyS on stroke onset day had a higher epilepsy risk (aHR 2.3, 95% CI 1.3-4.0, p=0.003) compared to later ASyS. Status epilepticus had the highest epilepsy risk (aHR 9.6, 95% CI 3.5-26.7, pConclusions ASyS timing and type significantly affect epilepsy and mortality risk after stroke, improving epilepsy prediction and guiding patient counseling.
- Open Access
Antiseizure medication withdrawal risk estimation and recommendations: a survey of American Academy of Neurology and EpiCARE members
Background Choosing candidates for antiseizure medication (ASM) withdrawal in well- controlled epilepsy is challenging. We evaluated 1) the correlation between neurologists’ seizure risk estimation (“clinician predictions”) versus calculated predictions, 2) how viewing calculated predictions influenced recommendations, and 3) barriers to using risk calculation. Methods We asked neurologists to predict two-year seizure risk after ASM withdrawal for hypothetical vignettes. We compared withdrawal recommendations ASMs before versus after viewing calculated predictions using generalized linear models. Results Three-hundred forty-six responded. There was moderate correlation between clinician and calculated predictions (Spearman coefficient 0.42). Clinician predictions varied widely, e.g., predictions ranged 5%-100% for a two-year seizure-free adult without epileptiform abnormalities. Mean clinician predictions exceeded calculated predictions for vignettes with epileptiform abnormalities (e.g., childhood absence epilepsy: clinician 65%, 95% confidence interval [CI] 57%-74%; calculated 46%) and surgical vignettes (e.g., focal cortical dysplasia six-months seizure-free mean clinician 56%, 95% CI 52%-60%; calculated 28%). Clinicians overestimated the influence of epileptiform EEG findings on withdrawal risk (26%, 95% CI 24%-28%) compared with calculators (14%, 95% 13%-14%). Viewing calculated predictions slightly reduced willingness to withdraw (−0.8/10 change, 95% CI -1.0 to -0.7), particularly without epileptiform abnormalities. The greatest barrier to calculator use was doubting its accuracy (44%). Conclusions Clinicians overestimated the influence of abnormal EEGs particularly for low-risk patients and overestimated risk and the influence of seizure-free duration for surgical patients, compared with calculators. These data may question widespread ordering of EEGs or time-based seizure-free thresholds for surgical patients. Viewing calculated predictions reduced willingness to withdraw particularly without epileptiform abnormalities. What is already known on this topic For the two-thirds of patients with epilepsy who become seizure-free on antiseizure medications (ASMs), a key question is whether ASMs are necessary indefinitely after attaining seizure-freedom. An individualized post-withdrawal seizure risk calculator exists,[1,2] which has demonstrated moderate external validity, and one 1993 study suggested that viewing results from an older version of such a calculator slightly reduced patients’ willingness to withdraw. What this study adds Before our study, it was unknown how closely clinicians’ intuitive estimates of post-withdrawal seizure risk (“clinician predictions”) align with model (“calculated”) predictions, and no study had previously evaluated the influence of viewing calculated risk on clinician recommendations. We found 1) moderate correlation between clinician and calculated predictions (Spearman coefficient 0.42), 2) clinician predictions and recommendations varied widely, 3) clinicians tended to overestimate risk compared with calculators for vignettes with epileptiform abnormalities and surgical cases, 4) viewing calculated predictions slightly reduced willingness to withdraw particularly without epileptiform abnormalities, and 5) the greatest endorsed barrier to calculator use was doubting its accuracy. How this study might affect research, practice or policy Our study suggests patient scenarios in which viewing calculator results may be most useful or be most likely to change clinical practice. These data may also question widespread ordering of EEGs or time-based seizure-free thresholds for surgical patients given calculated risk differences were smaller than clinician predictions. We also provide preliminary data guiding future work to improve calculator implementation such as targeting educational efforts to improve knowledge of its development, improving its accuracy, and integrating it into the electronic medical record.
Open AccessAntiseizure medication withdrawal risk estimation and recommendations: a survey of American Academy of Neurology and EpiCARE members
Background Choosing candidates for antiseizure medication (ASM) withdrawal in well- controlled epilepsy is challenging. We evaluated 1) the correlation between neurologists’ seizure risk estimation (“clinician predictions”) versus calculated predictions, 2) how viewing calculated predictions influenced recommendations, and 3) barriers to using risk calculation. Methods We asked neurologists to predict two-year seizure risk after ASM withdrawal for hypothetical vignettes. We compared withdrawal recommendations ASMs before versus after viewing calculated predictions using generalized linear models. Results Three-hundred forty-six responded. There was moderate correlation between clinician and calculated predictions (Spearman coefficient 0.42). Clinician predictions varied widely, e.g., predictions ranged 5%-100% for a two-year seizure-free adult without epileptiform abnormalities. Mean clinician predictions exceeded calculated predictions for vignettes with epileptiform abnormalities (e.g., childhood absence epilepsy: clinician 65%, 95% confidence interval [CI] 57%-74%; calculated 46%) and surgical vignettes (e.g., focal cortical dysplasia six-months seizure-free mean clinician 56%, 95% CI 52%-60%; calculated 28%). Clinicians overestimated the influence of epileptiform EEG findings on withdrawal risk (26%, 95% CI 24%-28%) compared with calculators (14%, 95% 13%-14%). Viewing calculated predictions slightly reduced willingness to withdraw (−0.8/10 change, 95% CI -1.0 to -0.7), particularly without epileptiform abnormalities. The greatest barrier to calculator use was doubting its accuracy (44%). Conclusions Clinicians overestimated the influence of abnormal EEGs particularly for low-risk patients and overestimated risk and the influence of seizure-free duration for surgical patients, compared with calculators. These data may question widespread ordering of EEGs or time-based seizure-free thresholds for surgical patients. Viewing calculated predictions reduced willingness to withdraw particularly without epileptiform abnormalities. What is already known on this topic For the two-thirds of patients with epilepsy who become seizure-free on antiseizure medications (ASMs), a key question is whether ASMs are necessary indefinitely after attaining seizure-freedom. An individualized post-withdrawal seizure risk calculator exists,[1,2] which has demonstrated moderate external validity, and one 1993 study suggested that viewing results from an older version of such a calculator slightly reduced patients’ willingness to withdraw. What this study adds Before our study, it was unknown how closely clinicians’ intuitive estimates of post-withdrawal seizure risk (“clinician predictions”) align with model (“calculated”) predictions, and no study had previously evaluated the influence of viewing calculated risk on clinician recommendations. We found 1) moderate correlation between clinician and calculated predictions (Spearman coefficient 0.42), 2) clinician predictions and recommendations varied widely, 3) clinicians tended to overestimate risk compared with calculators for vignettes with epileptiform abnormalities and surgical cases, 4) viewing calculated predictions slightly reduced willingness to withdraw particularly without epileptiform abnormalities, and 5) the greatest endorsed barrier to calculator use was doubting its accuracy. How this study might affect research, practice or policy Our study suggests patient scenarios in which viewing calculator results may be most useful or be most likely to change clinical practice. These data may also question widespread ordering of EEGs or time-based seizure-free thresholds for surgical patients given calculated risk differences were smaller than clinician predictions. We also provide preliminary data guiding future work to improve calculator implementation such as targeting educational efforts to improve knowledge of its development, improving its accuracy, and integrating it into the electronic medical record.
- 23 November 2022
- Published by Oxford University Press (OUP) in Brain
- Vol. 146 (6) , 2389-2398
Predictive models for starting antiseizure medication withdrawal following epilepsy surgery in adults
More than half of adults with epilepsy undergoing resective epilepsy surgery achieve long-term seizure freedom and might consider withdrawing antiseizure medications. We aimed to identify predictors of seizure recurrence after starting postoperative antiseizure medication withdrawal and develop and validate predictive models. We performed an international multicentre observational cohort study in nine tertiary epilepsy referral centres. We included 850 adults who started antiseizure medication withdrawal following resective epilepsy surgery and were free of seizures other than focal non-motor aware seizures before starting antiseizure medication withdrawal. We developed a model predicting recurrent seizures, other than focal non-motor aware seizures, using Cox proportional hazards regression in a derivation cohort (n = 231). Independent predictors of seizure recurrence, other than focal non-motor aware seizures, following the start of antiseizure medication withdrawal were focal non-motor aware seizures after surgery and before withdrawal [adjusted hazard ratio (aHR) 5.5, 95% confidence interval (CI) 2.7–11.1], history of focal to bilateral tonic-clonic seizures before surgery (aHR 1.6, 95% CI 0.9–2.8), time from surgery to the start of antiseizure medication withdrawal (aHR 0.9, 95% CI 0.8–0.9) and number of antiseizure medications at time of surgery (aHR 1.2, 95% CI 0.9–1.6). Model discrimination showed a concordance statistic of 0.67 (95% CI 0.63–0.71) in the external validation cohorts (n = 500). A secondary model predicting recurrence of any seizures (including focal non-motor aware seizures) was developed and validated in a subgroup that did not have focal non-motor aware seizures before withdrawal (n = 639), showing a concordance statistic of 0.68 (95% CI 0.64–0.72). Calibration plots indicated high agreement of predicted and observed outcomes for both models. We show that simple algorithms, available as graphical nomograms and online tools (predictepilepsy.github.io), can provide probabilities of seizure outcomes after starting postoperative antiseizure medication withdrawal. These multicentre-validated models may assist clinicians when discussing antiseizure medication withdrawal after surgery with their patients.
- 23 November 2022
- Published by Oxford University Press (OUP) in Brain
- Vol. 146 (6) , 2389-2398
Predictive models for starting antiseizure medication withdrawal following epilepsy surgery in adults
More than half of adults with epilepsy undergoing resective epilepsy surgery achieve long-term seizure freedom and might consider withdrawing antiseizure medications. We aimed to identify predictors of seizure recurrence after starting postoperative antiseizure medication withdrawal and develop and validate predictive models. We performed an international multicentre observational cohort study in nine tertiary epilepsy referral centres. We included 850 adults who started antiseizure medication withdrawal following resective epilepsy surgery and were free of seizures other than focal non-motor aware seizures before starting antiseizure medication withdrawal. We developed a model predicting recurrent seizures, other than focal non-motor aware seizures, using Cox proportional hazards regression in a derivation cohort (n = 231). Independent predictors of seizure recurrence, other than focal non-motor aware seizures, following the start of antiseizure medication withdrawal were focal non-motor aware seizures after surgery and before withdrawal [adjusted hazard ratio (aHR) 5.5, 95% confidence interval (CI) 2.7–11.1], history of focal to bilateral tonic-clonic seizures before surgery (aHR 1.6, 95% CI 0.9–2.8), time from surgery to the start of antiseizure medication withdrawal (aHR 0.9, 95% CI 0.8–0.9) and number of antiseizure medications at time of surgery (aHR 1.2, 95% CI 0.9–1.6). Model discrimination showed a concordance statistic of 0.67 (95% CI 0.63–0.71) in the external validation cohorts (n = 500). A secondary model predicting recurrence of any seizures (including focal non-motor aware seizures) was developed and validated in a subgroup that did not have focal non-motor aware seizures before withdrawal (n = 639), showing a concordance statistic of 0.68 (95% CI 0.64–0.72). Calibration plots indicated high agreement of predicted and observed outcomes for both models. We show that simple algorithms, available as graphical nomograms and online tools (predictepilepsy.github.io), can provide probabilities of seizure outcomes after starting postoperative antiseizure medication withdrawal. These multicentre-validated models may assist clinicians when discussing antiseizure medication withdrawal after surgery with their patients.
- 1 December 2021
- Published by Wiley in Alzheimer's & Dementia
- Vol. 17 (S4) , e057478
Meta‐analysis of the associations of biomarkers of amyloid and tau with cognition in cognitively normal older adults
Background: Conflicting evidence exists on the in vivo relationship of amyloid and tau levels to cognition among cognitively normal older adults. We conducted a meta‐analysis of reported relationships of biomarkers of amyloid or tau levels with baseline and change in cognition in cognitively normal older adults.Method: We performed a literature search that returned studies between 2005 and January 2021 for studies reporting a relationship between amyloid or tau biomarkers measured via PET ligands, CSF, or plasma assays, with episodic memory or global cognition. Effect sizes were obtained for each study and weighted by sample size. For studies examining the same cohort, we selected effect sizes from the largest sample size available for each relationship of interest.Result: One hundred and seven studies (89% reporting cross sectional measures and 62% reporting longitudinal) representing 26968 adults (82% characterized with amyloid measures and 23% with tau) were included. Amyloid levels showed a significant correlation with baseline and longitudinal global cognition (r=0.05, p<0001; r=0.05, p=0.021, respectively), and with baseline episodic memory (r=0.05, p=0.001). Tau was significantly correlated with baseline and longitudinal global cognition (r=0.08, p=0.004; r=0.08, p=0.01, respectively), and with longitudinal episodic memory (r=0.02, p=0.04), whereas the interaction of amyloid and tau was significantly correlated only with baseline global cognition (r=0.09, p=0.024). The overall estimated correlation between amyloid and tau was r=0.17.Conclusion: In older adults with normal cognition, there is a very small but significant relationship between baseline global cognition and amyloid, tau, and their interactive effect. Episodic memory had a very small but significant relationship with amyloid at baseline and with tau longitudinally. These findings could help determine power estimations for future clinical trials evaluating cognition in asymptomatic older adults. Effect size and power estimations from this meta‐analysis may be limited by unmeasured variability among studies in the timing of biomarker measurement, differences in cognitive assessments, or other cohort‐specific differences.
- 1 December 2021
- Published by Wiley in Alzheimer's & Dementia
- Vol. 17 (S4) , e057478
Meta‐analysis of the associations of biomarkers of amyloid and tau with cognition in cognitively normal older adults
Background: Conflicting evidence exists on the in vivo relationship of amyloid and tau levels to cognition among cognitively normal older adults. We conducted a meta‐analysis of reported relationships of biomarkers of amyloid or tau levels with baseline and change in cognition in cognitively normal older adults.Method: We performed a literature search that returned studies between 2005 and January 2021 for studies reporting a relationship between amyloid or tau biomarkers measured via PET ligands, CSF, or plasma assays, with episodic memory or global cognition. Effect sizes were obtained for each study and weighted by sample size. For studies examining the same cohort, we selected effect sizes from the largest sample size available for each relationship of interest.Result: One hundred and seven studies (89% reporting cross sectional measures and 62% reporting longitudinal) representing 26968 adults (82% characterized with amyloid measures and 23% with tau) were included. Amyloid levels showed a significant correlation with baseline and longitudinal global cognition (r=0.05, p<0001; r=0.05, p=0.021, respectively), and with baseline episodic memory (r=0.05, p=0.001). Tau was significantly correlated with baseline and longitudinal global cognition (r=0.08, p=0.004; r=0.08, p=0.01, respectively), and with longitudinal episodic memory (r=0.02, p=0.04), whereas the interaction of amyloid and tau was significantly correlated only with baseline global cognition (r=0.09, p=0.024). The overall estimated correlation between amyloid and tau was r=0.17.Conclusion: In older adults with normal cognition, there is a very small but significant relationship between baseline global cognition and amyloid, tau, and their interactive effect. Episodic memory had a very small but significant relationship with amyloid at baseline and with tau longitudinally. These findings could help determine power estimations for future clinical trials evaluating cognition in asymptomatic older adults. Effect size and power estimations from this meta‐analysis may be limited by unmeasured variability among studies in the timing of biomarker measurement, differences in cognitive assessments, or other cohort‐specific differences.