Breathing Before Death: Are We Closer to Identifying Premortem Biomarkers of Sudden Unexpected Death in Epilepsy (SUDEP)?

Abstract

Risk Markers for Sudden Unexpected Death in Epilepsy: An Observational, Prospective, Multicentre Cohort Study

Ochoa-Urrea M, Luo X, Vilella L, Lacuey N, Omidi SJ, Hupp NJ, Talavera B, Hampson JP, Rani MRS, Tao S, Li X, Miyake CY, Cui L, Hampson JS, Chaitanya G, Vakilna YS, Sainju RK, Friedman D, Nei M, Allen L, Scott CA, Oliveira J, Gehlbach B, Schuele SU, Ogren JA, Harper RM, Diehl B, Bateman LM, Richerson GB, Yamal JM, Zhang GQ, Devinsky O, Lhatoo SD. Lancet. 2025 Oct 4;406(10511):1497-1507. doi: 10.1016/S0140-6736(25)01636-8. Epub 2025 Sep 17. Erratum in: Lancet. 2025 Oct 4;406(10511):1472. doi: 10.1016/S0140-6736(25)01970-1. PMID: 40975113.

Background: Sudden unexpected death in epilepsy (SUDEP) is the leading cause of epilepsy-related mortality. Generalised-particularly nocturnal-convulsive seizures, longstanding epilepsy, and solitary living have been identified retrospectively as risk factors. No definitive electroclinical biomarkers have been prospectively ascertained. This study aimed to identify SUDEP risk markers using multimodality data with long-term follow-up. Methods: This prospective, multicentre, observational cohort study, conducted at nine centres (eight in the USA and one in the UK), recruited children and adults with epilepsy who were undergoing prolonged video-electroencephalographic (EEG) monitoring. Inclusion criteria were diagnosis of epilepsy by an epilepsy specialist, with or without drug resistance; age older than 2 months; admission to the epilepsy monitoring unit of a participating centre, with video-EEG monitoring; and completion of at least one 6-month follow-up. Demographic, electroclinical, and cardiorespiratory data were collected at baseline. Participants were followed up long-term through routine clinic visits, review of electronic health records, and telephone interviews to collect information about seizure frequency, medication status, and mortality. The primary endpoint was time to SUDEP. Cox proportional hazards models were used to assess significant risk factors. Findings: Between September 17, 2011, and December, 30, 2021, 2632 children and adults with epilepsy were enrolled in this study; 164 were lost to follow-up. 38 (1.54%) of 2468 participants died from SUDEP (12 definite, 18 probable, and eight possible SUDEP cases), and two had near-SUDEP events. Incident SUDEP mortality rate was 4.76 (95% CI 3.37-6.53) cases per 1000 person-years, from a cohort of 7982 person-years. Living alone (hazard ratio 7.62, 95% CI 3.94-14.71), three or more generalised convulsive seizures in the previous year (3.1, 1.64-5.87]), longer ictal central apnoea (1.11, 1.05-1.18), and longer postictal central apnoea (1.32, 1.14-1.54]) were significant predictors of increased SUDEP risk. In a subanalysis excluding possible and near-SUDEP cases, longer ictal central apnoea was not significant. Interpretation: This study shows an association between premortem peri-ictal apnoea and increased SUDEP risk. Cardiorespiratory monitoring during seizures might benefit assessments of epilepsy mortality risk. Together with solitary living and convulsive seizure frequency, peri-ictal apnoea (>14 s for postictal central apnoea and >17 s for ictal central apnoea) could inform the development of a validatable SUDEP risk index.

Commentary

Sudden unexpected death in epilepsy (SUDEP) remains the leading direct cause of epilepsy-related mortality. Among neurological causes, it is second only to stroke in years of potential life lost, with more than 100 thousand years lost in the US in 2010. ¹ Despite this impact, individual risk prediction remains imprecise. Current counseling and prediction tools rely on clinical factors derived from retrospective data.^2,3 These factors include generalized convulsive seizures (GCS), nocturnal seizures, living alone, substance and alcohol abuse, male sex, epilepsy duration, and antiseizure medication (ASM) nonadherence.^4–6 The field has long sought prospectively validated physiological markers that could refine risk stratification and guide prevention.

In this context, Ochoa-Urrea et al ⁷ report the first large multicenter prospective cohort study evaluating electroclinical and cardiorespiratory markers of SUDEP. Over 10 years, 2632 patients undergoing video-electroencephalogram evaluation were enrolled across 9 epilepsy monitoring units (EMUs), contributing nearly 8000 person-years of follow-up. Thirty-eight (1.54%) patients died from SUDEP, and 2 experienced near-SUDEP events in the EMU. The observed SUDEP incidence rate was 4.76 per 1000 person-years. The study prospectively confirms 2 major risk factors. Living alone was associated with a markedly elevated risk of SUDEP (hazard ratio [HR] = 7.62), and ≥3 GCS in the prior year was associated with a 3-fold increase in risk. These findings reinforce what retrospective case–control studies have long suggested: convulsive seizure burden and lack of supervision remain central, modifiable drivers of SUDEP risk.

The novel contribution lies in understanding the role of respiratory physiology in SUDEP. Prolonged peri-ictal central apnea emerged as a potential premortem biomarker. In unadjusted analyses, longer postictal central apnea (HR 1.32 per 10-s increment) and longer ictal central apnea (HR 1.11 per 10-s increment) were associated with increased SUDEP risk. Using median cutoffs, postictal central apnea ≥15 s and ictal central apnea >17 s stratified patients into higher 5-year cumulative risk categories. These findings move the field beyond descriptions of the terminal cascade. Prior work, including MORTEMUS, demonstrated that postictal apnea often precedes terminal asystole. ⁸ What remained unclear was whether respiratory dysfunction during nonfatal seizures signals long-term vulnerability. This study suggests that peri-ictal apnea may reflect an underlying vulnerability rather than merely a terminal event. If validated, it could represent an intermediate phenotype linking seizure activity to fatal outcome. This signal is biologically plausible and clinically measurable.

However, a signal is not proof. The HRs for peri-ictal apnea were modest. In multivariable models adjusted for clinical covariates, the associations were attenuated and did not consistently remain statistically significant. Excluding possible and near-SUDEP cases further weakened the respiratory signals. These findings do not negate the signal. They support the biological plausibility, but temper confidence in their independent predictive validity. Moreover, only 38 SUDEP events were available for analysis. Wide confidence intervals and model sensitivity are expected in this context. Low event rates constrain prospective SUDEP research and require cautious interpretation. The generalizability of these findings to at-home seizures also warrants consideration. Seizures recorded in EMU often occur during ASM reduction. Seizure severity and physiological stress may differ from home settings. Whether apnea duration measured under such conditions reflects outpatient risk remains uncertain.

Despite these limitations, the study has some translational implications. Living alone and frequent GCS remain dominant, potentially modifiable risks. The addition of measurable respiratory parameters opens new avenues. Cardiorespiratory monitoring during EMU evaluation is feasible and could be standardized. ⁹ Systematic quantification of peri-ictal apnea duration could be incorporated into structured EMU reporting. However, current evidence does not justify the inclusion of apnea thresholds in clinical risk calculators. The key question is whether apnea duration adds meaningful discrimination beyond seizure frequency and living status. Demonstration of incremental predictive value and calibration will be essential before clinical adoption.

Even modest HRs may be clinically significant in the context of a rare but catastrophic outcome. The goal is not to identify a single decisive marker but to incrementally improve risk stratification. Prospective physiological measures represent 1 such step. At the same time, imprecise prediction continues to complicate SUDEP counseling. Clinicians often hesitate to quantify risk in the absence of individualized markers. Population-level data guide practice but limit personalization. Measurable physiological signals may narrow this gap, even if validation remains pending.

The search for SUDEP biomarkers is accelerating. A recent prospective nested case–control study in drug-resistant focal epilepsy identified extratemporal epileptogenic zones, particularly perisylvian and frontal, as independent risk factors. ¹⁰ Notably, the REPO₂MSE study and Ochoa-Urrea et al ⁷ found no association between peri-ictal oxygen saturation below 80% and SUDEP risk. Oxygen desaturation and central apnea are related but distinct phenomena. Together, these findings suggest that SUDEP risk is multidimensional. Network-level vulnerability, as reflected in seizure-onset localization, may interact with physiological instability, as reflected by prolonged central apnea. No single biomarker is likely sufficient. The field will likely move toward composite models that integrate convulsive burden, living status, respiratory control, and brain localization.

Beyond its specific findings, the study by Ochoa-Urrea et al ⁷ establishes a methodological milestone. Large-scale, long-term, collaborative physiological monitoring is feasible. Standardized adjudication of SUDEP and systematic capture of cardiorespiratory data can be achieved across centers. The REPO₂MSE study similarly demonstrates the value of prospective, multicenter phenotyping. Together, these efforts represent a blueprint for future SUDEP research.

SUDEP prevention remains grounded in established principles. Optimizing control of GCS is paramount. Addressing nocturnal supervision and living circumstances remains essential. Emerging evidence regarding body mass index and epileptogenic zone localization suggests additional modifiable or targetable domains. ¹⁰ Respiratory physiology may eventually inform intervention strategies, but such applications remain investigational.

The central advance of Ochoa-Urrea et al ⁷ is conceptual. SUDEP risk may reflect not only seizure counts and circumstances but also quantifiable physiological vulnerability. Whether peri-ictal central apnea becomes a validated component of risk stratification will depend on replication, external validation, and demonstration of incremental predictive value. For now, the field stands at an inflection point. Prospective biomarkers are emerging. Signals are accumulating. Yet caution remains essential. Breath, convulsions, and solitude together shape risk. The challenge ahead is to integrate these dimensions into reliable, clinically actionable models that move SUDEP prevention from aspiration toward precision.