Standard complete blood count to predict long-term outcomes in febrile infection–related epilepsy syndrome (FIRES) - A multicenter study

Sep 30, 2026

Uploaded by dabgoudw

Objective: Investigate whether complete blood count (CBC) analyses during an intensive care unit stay can predict 12-month outcomes in cryptogenic febrile infection–related epilepsy syndrome (FIRES). Methods: Outcomes at 12 months were classified as unfavorable (GOS 1–3) or favorable (GOS 4–5). Demographic, clinical, and serial CBC data were collected across three treatment phases and modeled with decision tree regression, with leave-one-patient-out validation. Results: Across 12 centers, 63 patients and 1530 CBCs were analyzed; key CBC ratios differed by outcome and treatment. Significance: Standard CBCs may enable rapid early prognostication, especially in children.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages15

Reading Time38 min

AccessFree to download

Abstract

Objective

Methods

Results

Significance

Introduction

Rationale and hypothesis

Key points

DOI: 10.1111/epi.18605 RESEAR CH ARTICLE Standard complete blood count to predict long-term outcomes in febrile infection–related epilepsy syndrome (FIRES): A multicenter study Martin Guillemaud1 | Aurélie Hanin1,2,3,4 | James J. Riviello5 | Mario Chavez1 | Ayush Batra6 | Megan Berry7 | Francesca Bisulli8,9 | Carlos Castillo-Pinto10 | Carla Cobos-Hernandez11 | Sophie Demeret12 | Krista Eschbach11 | Raquel Farias-Moeller13 | Madeline Fields14 | Nicolas Gaspard4,15 | Elizabeth E. Gerard6 | Teneille E. Gofton16 | Margaret T. Gopaul4 | Matthew D. Gruen4 | Anthony D. Jimenez4 | Karnig Kazazian16 | Minjee Kim6 | Marwa Mansour17 | Lara Marcuse14 | Clémence Marois12 | Mikaela Morales10 | Lorenzo Muccioli8,9 | Elena Pasini8 | Michelle M. Pham18 | Santiago Philibert Rosas7 | Aaron F. Struck7 | Nathan Torcida15 | Mark S. Wainwright10 | Ji Yeoun Yoo14 | Eyal Muscal5,19 | Vincent Navarro1,20 | Lawrence J. Hirsch4 | Yichen Lai17 Correspondence Aurélie Hanin, Institut du Cerveau, Paris Brain Institute, ICM, INRIA, Inserm, CNRS, AP-HP, Hôpital de la Pitié-Salpêtrière, Sorbonne Université, Paris, France. [Email:](Email: aurelie.hanin@icm-institute.org)[ aurelie.hanin@icm-institute.org](Email: aurelie.hanin@icm-institute.org) Funding information Daniel Raymond Wong Neurology Fund; NORSE/FIRES Research Fund at Yale; FFRE-Paratonnerre Abstract Objective: We investigated whether complete blood count (CBC) analyses during intensive care unit stay could predict 12-month outcomes in patients with cryptogenic febrile infection–related epilepsy syndrome (FIRES), a subset of newonset refractory status epilepticus (NORSE) . Methods: Outcomes at 12 months were classified as “unfavorable” (Glasgow Outcome Score [GOS] 1–3) or “favorable”(GOS 4–5). Demographic, clinical, and serial CBC data were collected across treatment phases: (1) no immunotherapy (before initiation or no treatment),(2) first-line immunotherapy, and (3) second-line immunotherapy. For each treatment phase, predictive models stratified outcomes based on CBC features using decision tree regression, with separate models for adults and children. Model performance was tested using a leave-one-patient-out approach. Results: We studied 63 patients (34 adults, 29 children) from 12 centers. Unfavorable outcomes occurred in 18 adults and 12 children. Children were more likely to receive second-line immunotherapy. We analyzed 1530 CBCs (adults: 997 CBCs, including 539 for unfavorable outcomes; children: 533 CBCs, including Martin Guillemaud and Aurélie Hanin are co-first authors. For affiliations refer to page 4791. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2025 The Author(s). Epilepsia published by Wiley Periodicals LLC on behalf of International League Against Epilepsy. 415 for unfavorable outcomes) . Subgroup analyses revealed differences in CBC levels according to the outcomes and the treatment received. Adults with unfavorable outcomes notably had higher neutrophil-to-lymphocyte ratios (NLRs) and monocyte-to-lymphocyte ratios (MLRs), whereas children with unfavorable outcomes had higher red cell distribution width. NLRs and MLRs increased when CBCs were collected after the initiation of immunotherapy for both adults and children. The variables of interest differed in the different predictive models but always included the proportion of at least one subtype of leukocyte. Prediction accuracy with our models was higher in children (87% overall, with the best performance in no-treatment and first-line phases) than in adults (83% overall, with the best performance during/after the initiation of second-line) . Significance: Findings suggest the potential for standard CBCs to serve as a rapid,

FAQs

What is the main objective of this study in FIRES patients?

The study evaluates whether complete blood count (CBC) analyses during an ICU stay can predict 12-month outcomes in cryptogenic FIRES.

How were outcomes classified at 12 months?

Outcomes were grouped as unfavorable (Glasgow Outcome Score 1–3) or favorable (GOS 4–5).

Which CBC features were linked to unfavorable outcomes?

Adults with unfavorable outcomes had higher neutrophil-to-lymphocyte ratios and monocyte-to-lymphocyte ratios, while children with unfavorable outcomes showed higher red cell distribution width. Ratios also increased after immunotherapy initiation.

Sep 30, 2026

Uploaded by dabgoudw