Early Celecoxib use in Patients with Traumatic Brain Injury

Debarati Bhanja, David R. Hallan, Jacob Staub, Elias Rizk, Joseph Christopher Zacko

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Background: Traumatic brain injury (TBI) can cause rapid brain inflammation. There is debate over the safety and efficacy of anti-inflammatory agents in its treatment. With a particular focus on cyclooxygenase 2 (COX2) selective inhibition, we sought to determine the impact of celecoxib versus no celecoxib treatment on outcomes in patients with TBI and compare these with outcomes associated with nonselective COX inhibition (ibuprofen) and corticosteroid (dexamethasone) treatment. Methods: This retrospective cohort study used TriNetX, a large publicly available global health research network, to gather clinical data extracted from the electronic medical records. Using International Classification of Diseases, Tenth Revision and pharmacy codes, we identified patients with TBI who were and were not treated with celecoxib, ibuprofen, and dexamethasone. Analysis was performed on propensity-matched and unmatched cohorts, which were matched on demographics, comorbidities, and neurological injuries. Our primary end point was 1-year survival. Secondary end points were ventilator and tracheostomy dependence, gastrostomy tube placement, seizures, and craniotomy. Results: After propensity score matching, a total of 1443 patients were identified in both the celecoxib and no celecoxib cohorts. Ninety-two (6.4%) patients in the celecoxib cohort died within 1 year following TBI versus 145 (10.0%) in the no celecoxib cohort (odds ratio 0.61; 95% confidence interval 0.46–0.80; p = 0.0003). The 1-year survival rate was 96.1% in the celecoxib cohort versus 93.1% in the no celecoxib cohort (p < 0.0001). At the end of the 1-year period, celecoxib was associated with significantly lower gastrostomy tube dependence (p = 0.017), seizure activity (p = 0.027), and myocardial infarction (p = 0.021) compared with the control cohort. Ibuprofen was also associated with higher 1-year survival probability and lower rates of post-TBI complications. Dexamethasone was broadly associated with higher morbidity but was associated with higher 1-year survival probability compared with the no dexamethasone cohort. Conclusions: Early celecoxib and ibuprofen use within 5 days post TBI was associated with higher 1-year survival probabilities and fewer complications. With emerging yet controversial preclinical evidence to suggest that COX inhibition improves TBI outcomes, this population-level study offers suggestive support for these drugs’ clinical benefit, which should be pursued in prospective clinical studies.

Original languageEnglish (US)
JournalNeurocritical Care
StateAccepted/In press - 2023

All Science Journal Classification (ASJC) codes

  • Critical Care and Intensive Care Medicine
  • Clinical Neurology

Cite this