Authors :
Presenting Author: Michael Baumgartner, PhD – Perelman School of Medicine at the University of Pennsylvania
Samuel Tomlinson, MD – University of Pennsylvania
Kathleen Galligan, PA-C – Children's Hospital of Philadelphia
Benjamin Kennedy, MD – Children's Hospital of Philadelphia
Rationale:
Fever following hemispheric disconnection surgery is a well-known though poorly understood phenomenon, one which results in frequent unnecessary infectious workups (Kossoff et al. 2002). Prior studies have identified risk factors for post-hemispherotomy fever based on etiology, with Rasmussen encephalitis (Kamanth et al. 2015, Phung et al. 2013) and size of the disconnected hemisphere (de Almeida et al. 2006) being linked with higher rates of fever, as well as protective factors, such as placement of an external ventricular drain2. These studies have largely been based on batteries of univariate tests. No multivariate analyses have been conducted examining links between clinical characteristics, CSF parameters, and probability of post-hemispherotomy fever.Methods:
A retrospective chart review was conducted on all patients having undergone hemispherotomy at the Children’s Hospital of Philadelphia between May 2017 and July 2024. All surgeries were performed by the senior author (BCK). Clinical characteristics including seizure etiology, laterality, operative duration, blood loss, post-operative length-of-stay, and age and weight at surgery were collected. Daily maximum temperature was collected through POD21 (if applicable) as well as all CBC and CSF samples taken. Fever was defined as any Tmax >38.5C between POD0-14. Two multivariate logistical regressions were performed with whether or not the patient fevered postoperatively as the dependent variable. The first analysis was conducted on all patients with clinical characteristics as predictors, whereas the second was performed on those patients for whom CSF studies were performed.
Results:
70 patients met inclusion criteria for this study (35 males, average age: 7.2 years). Post-operative fever occurred in 30/70 patients (42.9%), with group-level Tmax occurring on POD2. Fevers were more common at the start of the surgeon’s practice, although there was no clear change in operative duration. Logistic regression analysis on clinical characteristics found that patients with Rasmussen’s encephalitis were likelier to fever (8/11, 72.7%, p=0.036, OR=18.258) but failed to find an association between age, blood loss, and blood transfusions and fever. Among patients for whom CSF was collected (52/70, 74.3%), a significant association was observed between CSF protein level and postoperative fever (0.029, OR=1.002 per unit mg/dL) but no association with glucose or cell counts. Conclusions:
Fevers following hemispherotomy are more common in patients with Rasmussen’s encephalitis and are positively associated with CSF protein levels. Rates of postoperative fever may decline with surgeon experience.Funding:
This work was not supported by any independent sources of funding.