The role of neurologic pupil index in pediatric patients with severe traumatic brain injury

Introduction: Effective treatment of intracranial hypertension (ICH) in severe traumatic brain injury (sTBI) is essential to improve morbidity and mortality. Early identification of patients at risk for refractory ICH could improve outcomes. A Pupillometer non-invasively combines pupillary variables into a Neurological Pupil index (NPi) score. NPi < 3 is considered pathological. Studies in adults demonstrate an inverse correlation between NPi and intracranial pressure (ICP), with report that an abnormal NPi can precede maximum ICP by 15.9 hours. The aim of this study is to evaluate changes in NPi in the preceding 24 hours in pediatric patients with sTBI who have increasing ICP, with a goal to determine if pupillometry may assist in earlier identification of ICH. Methods: In this retrospective analysis of children ≤18 years with sTBI and ICP monitoring at a level 1 trauma center from 2018–2021, hourly ICP and NPi readings were collected up to 72 hours in controls (ICP < 20) and 120 hours in cases (≥ 1 ICP reading ≥20). NPi was reviewed for 24 hours preceding ICP spike. Descriptive statistics and scatter plots were generated. Secondary analysis assessed relationships between ICP and other pupillometry measurements, pupillometry response to pentobarbital, hyperosmolar therapy and decompressive craniectomy (DC). Results: Out of 27 patients, 81.5% had an ICP reading ≥20. The cohort was predominantly male (66.7%) and white (59.3%), with a mean age of 11.5 years (0.7-18). Median GCS was 4, with most common mechanism being blunt trauma (74.1%) and 55.6% had midline shift on CT. An EVD was placed in 63% and intraparenchymal bolt in 52% of patients. 16 patients had non-refractory ICH, while 6 had refractory ICH. Out of 287 ICP readings ≥20, only 17.8% had a NPi < 3 within the preceding 24 hours. In the DC cohort, 128/141(90.8%) ICP readings ≥20 were associated with NPi >3. Pentobarbital infusion initiation was associated with decrease in NPi. Conclusions: There was no association between ICP ≥20mmHg and a preceding NPi < 3, indicating that NPi cannot be used in isolation to predict ICH in sTBI pediatric patients. Patients who underwent DC had normal NPis with ICP ≥20mmHg, likely due to improved brain compliance after DC. Pentobarbital infusion effect on pupil reactivity may diminish utility of NPi in predicting ICH.