Ibuprofen Versus Acetaminophen for Treatment of Mild Traumatic Brain Injury

OBJECTIVES The long-term goal of the proposed project is to develop an effective pharmacological therapy for patients with mild TBI to improve functional outcomes. The main objective is to compare the effects of ibuprofen versus acetaminophen for reducing the short-term development post-concussion symptoms in patients with mild TBI. We will conduct a double-blinded randomized clinical trial (RCT) with the specific aim of measuring the comparative effects of equipotent doses of ibuprofen versus acetaminophen on the incidence of concussion symptoms to be measured 7 to 10 days after mild TBI in emergency department patients with isolated closed head injury. The main study hypothesis of our research project is that patients with mild TBI who are treated with ibuprofen will have a lower incidence of post-concussion symptoms compared to patients treated with acetaminophen. The proposed investigation is novel because it will be the very first RCT to compare analgesic medications for the specific treatment of mild TBI. BACKGROUND Pathophysiology of Mild TBI. TBI results from external mechanical force applied to the cranium leading to parenchymal brain damage. This resultant brain injury can range from mild temporary impairment to severe disability. TBI transpires in two phases: (1) primary brain injury occurs at the moment of the application of mechanical force on the cranium and results in lacerations, contusions, hematomas and shearing injuries of the brain; and (2) secondary brain injury begins immediately after the primary brain injury and results from the cellular mediation of neuroinflammation.15,16 The principal injury of mild TBI is diffuse axonal injury from shearing forces that is not usually identified by CT brain scan.17-20 Secondary brain injury activates multiple cellular pathways that are initially adaptive, but become pathological with overexpression and persistence.21-23 The biochemical consequences of these responses develop over a period of hours with the accumulation of arachidonic acid from cellular membrane stores and the induction of cyclooxygenase-2 (COX-2) gene expression and enzyme activity.15,16 Arachidonic acid is then converted to detrimental vasoactive prostanglandins and free radicals by COX-2 enzyme leading to neuronal cell death (Figure 1 see appendix). Neural COX-2 enzyme activity remains elevated for 1 to 3 days.24 Severity of neuronal injury is correlated to COX-2 overexpression which results in a "vicious cycle" of neuroinflammation when secondary injury propagates further COX-2 activity.15,16 Potential Effect of COX-2 Inhibition. If COX-2 overexpression causes neuronal cell damage after brain injury, then COX-2 inhibition may provide neuroprotective effects through two important mechanisms: (1) by reducing detrimental vasoactive prostanglandins and free radical synthesis; and (2) shunting of arachidonic acid down alternate pathways that produce beneficial eicosanoids.15 These cellular mechanisms are based on findings from several animal studies over the past 30 years that strongly suggest potential beneficial effects of NSAID inhibition of the COX-2 enzyme as a treatment for TBI-activated neuroinflammation.25-35 Although these studies examined different NSAID COX-2 kinhibitors and utilized different animal models of TBI, they found an overall consensus of beneficial effects (i.e., reduced prostaglandin synthesis, reduced brain edema, improvement in cognitive and motor function, improvements in memory, and reduced mortality) suggesting COX-2 inhibition may have effects beyond analgesia in patients with mild TBI. In fact, COX-2 inhibition has been shown to be beneficial for animal models for other types of brain insults including ischemic brain injury.36-40 If we identify a beneficial effect of ibuprofen for patients with mild TBI that is confirmed with validation studies, then new standards and guidelines will be developed to improve the ED management of this common form of brain injury. Future studies will be conducted to elaborate the effects of ibuprofen and other NSAIDS on both the short-term and long-term complications of mild TBI and possible other forms of TBI. Discovery of a beneficial effect of NSAIDs for mild TBI could be especially useful for combat-related and sports-related brain injuries. STUDY DESIGN Study subjects will be identified and enrolled in the emergency department. They will be followed for a 7 to 14 day period only. The proposed project will be completed over a one-year period with tasks and milestones as follows: Month 1 - develop protocol, create data collection instrument and codebook, test and modify instrument, create electronic database, and train assistants; Months 2 and 10 - identify and enroll eligible study subjects in the ED setting, create database, enter and edit data; and Months 11 and 12 - analyze data and interpret findings, prepare and submit papers to journals. Study Design and Participants. We will conduct a double-blinded RCT to compare the effects of ibuprofen versus acetaminophen on the incidence of post-concussion symptoms measured 7 to 14 days after mild TBI. Treatment Variable. The treatment variable will be equipotent oral doses of either ibuprofen (800 mg TID) or acetaminophen (1000 mg TID). Ibuprofen is a nonselective NSAID that inhibits both COX-1 and COX-2 isoenzymes. COX-2 inhibition prevents arachidonic acid from converting to vasoactive prostaglandins and reactive oxygen species in brain cell. The analgesic, antipyretic, and antiinflammatory activity of ibuprofen operates mainly through inhibition of COX-2.55,56 Acetaminophen is a poor inhibitor of both COX isoenzymes in the CNS and has significantly weaker antiinflammatory effects than NSAIDs. Acetaminophen does not inhibit COX in peripheral tissues and is less effective in the presence of peroxides.57,58 Both ibuprofen and acetaminophen are frequently prescribed for headache pain related to mild TBI in the ED and other primary care settings and have excellent safety profiles for short-term use. In fact, both ibuprofen and acetaminophen have been used for the antipyretic treatment of children with severe traumatic brain injury.59 Subjects will receive the first medication dose in the ED and will be given the remaining 5 doses to take over 48 hours as outpatients. This time period was selected based on animal models identifying maximum COX-2 activity. Both medications will be identically prepared to prevent subjects from identifying the type of treatment regimen. Treatment assignment will be randomized by the NMH research pharmacy and blinded to both the patients and the investigators. Standard methods for blinding will be implemented by using tablets that are identical in size, shape, color and taste for both ibuprofen and acetaminophen. Data Collection and Variable Measurement. Information concerning the outcome and secondary study variables will be obtained by research assistants using standardized data collection instruments. Follow up outcome assessment will be conducted through telephone interviews at 7 to 14 days after head injury. We will implement best practice methods for accurate and complete data collection, including the following: (1) training assistants in data collection methods; (2) blinding assistants to the subjects' treatment assignment; and (3) conducting meetings with assistants to review data collection processes and difficulties. Subjects with persistent mild TBI symptoms at time of follow up will be referred to the Rehabilitation Institute of Chicago Concussion Clinic for further treatment. Research Location. Northwestern Memorial Hospital is a state-of-the art, two million square foot, 873-bed tertiary teaching hospital in downtown Chicago that is the primary teaching hospital affiliated with the Northwestern University Feinberg School of Medicine. The Northwestern Memorial Hospital Emergency Department is an urban, academic, 56 bed, Level I trauma center with an annual patient volume of over 85,000 patients per year or approximately 230 patients per day. The Emergency Department is staffed exclusively by board-certified emergency medicine physicians that also supervise the care of 48 residents and four fellows. The reported demographic distribution of the overall Emergency Department patient population is as follows: 7% Hispanic; 2% Asian; 35% Black; and 63% White. SATATISTICAL ANALYSES Data Analyses Plan. We will conduct all statistical analyses based on the American Statistical Association Ethical Guidelines for Statistical Practice67,68 and include the following: 1. Univariate Analyses. We will first conduct univariate analyses to assess potential demographic and clinical differences between the two treatment groups. We will also use ANOVA to compare the effects of ibuprofen versus acetaminophen treatment on the incidence of mild TBI symptoms using continuous measures of pain scores and Neural-QOL measures of pain and cognitive function. 2. Multivariate Analyses. We will use log binomial regression modeling to adjust for potential residual confounding to estimate the effects of ibuprofen versus acetaminophen on the incidence of mild TBI symptoms. We will also conduct survival analyses using Kaplan-Meier graphs and Cox proportional hazards regression modeling to estimate adjusted hazards ratios for the effects of ibuprofen versus acetaminophen on the main outcomes of mild TBI symptoms. We will also conduct likelihood ratio and goodness-of-fit tests to identify and include specific covariates for the regression models and diagnostic methods to evaluate overall model fit and plausibility of model assumptions. 3. Advanced Methods for Causal Inference. We will use advanced statistical methods to allow assessment of causation: (1) instrumental variables to estimate the effects of ibuprofen versus acetaminophen with potential treatment regimen noncompliance by using treatment randomization as a perfect instrument;69,70 (2) marginal structural modeling to account for time-varying confounders and to assess for direct versus indirect effects in the presence of intermediate factors.71-74 Sample Size Estimation. Sample size estimation is especially challenging because there are no prior human studies measuring the effect of COX-2 inhibition on the incidence of mild TBI symptoms. Notwithstanding this limitation, we will use the presence of headache pain at 7 to 14 days as a binary outcome. Sample size estimation is based on the following: (1) 40% expected frequency of headache pain in the ibuprofen group versus 60% expected frequency in the acetaminophen group; (2) the use of two-sided tests performed at a significance level of 0.05; (3) a power of 0.80 to detect a true difference in headache pain between subjects receiving ibuprofen versus acetaminophen; and (4) one to one distribution of study subjects receiving ibuprofen versus acetaminophen. With these assumptions, the estimated number of study subjects needed to statistically assess ibuprofen versus acetaminophen for the treatment of mild TBI is a total of 214 subjects.60 Our plan to enroll over 400 subjects will provide an adequate sample size to conduct our statistical analyses. The Northwestern Memorial Hospital Emergency Department provides care to about 2,000 head injured patients with mild TBI each year.