How Do You Treat Traumatic Brain Injury? One Symptom at a Time

When General Creighton Adams was asked how to tackle a difficult problem, he reputedly answered, “When eating an elephant, take one bite at a time.” In our view, the problem of treating traumatic brain injury (TBI) should be addressed the same way, one symptom or subdomain at a time. In this editorial, we will make the case for the approach and lay the outline for a path forward.

TBI is no longer a silent pandemic. Military health systems, sports organizations, and the general public are clamoring for answers. We have very few solutions based on solid scientific evidence; most of what we do is based on “clinical experience,” which has been defined as “making the same mistakes with increasing confidence over an impressive number of years.”¹

So, what are our options?

We could keep doing large, acute phase clinical trials in moderate-to-severe TBI with broad inclusion criteria and global outcome measures, hoping to find a “magic bullet.” Unfortunately, this has not worked in the past, despite substantial effort and expense.^2–4

We could focus all of our efforts on prevention. Certainly, car safety, better helmets, and balance training are worthy endeavors, but an exclusive focus on prevention would leave the millions of patients with chronic symptoms and deficits in the lurch.

We could slump into nihilism. Nothing will ever fix the “squashed bug” or “unscramble the egg.” For us, this is not an option. Our patients demand more.

We could double down on basic science research, looking long and hard in the laboratory for the magic bullet that will cure TBI. Clearly, basic science is important and should continue. In the domain of cancer treatment, finer and finer pathophysiological subdivisions punctuated by occasional breakthrough discoveries have led to effective treatments for a few specific cancers.⁵ However, substantial progress in cancer has mainly come from systematically trying a lot of ideas in a lot of clinical trials to figure out empirically what works and what does not.^6–8

Fortunately, a few bright sparks have been spotted. There have been several examples of successful clinical trials focused on one issue, one symptom, or one subdomain at a time, conducted in specific TBI subpopulations most likely to benefit. These trials have shown the following:

1. Phenytoin and levetiracetam reduce early seizures in adults with acute, moderate-to-severe TBI.^9,10

2. Amantadine hastens recovery of consciousness for adults with subacute, very severe TBI.¹¹

3. Methylphenidate improves mental processing speed and attentiveness in adults with postacute attention deficit following TBI.¹²

4. Specific types of repetitive transcranial magnetic stimulation alleviate headache and depression in adults with chronic concussive/“mild”-to-moderate TBI.^13,14

5. Sustained-release melatonin 2 hours before bedtime modestly improves sleep quality, anxiety, and fatigue in chronic TBI patients with insomnia.¹⁵ This also exemplifies the important idea that treatment of one symptom can improve others.

Although there are more examples, the list of evidence-based treatments in the field of TBI is far too short.^16,17 Nonetheless, these modest successes give us hope that this is the right path, especially because we are at the beginning of what may be a “golden age” for TBI research, with more progress in the past few years than in the previous 5,000.¹⁸ Certainly, we would not argue that we know enough about the fundamental pathophysiology, natural history, and epidemiology of TBI; the topic is still relatively immature. However, there has been tremendous progress in recent years with the substantial international TBI research efforts underway, including TRACK-TBI in the United States and CENTER-TBI in Europe.¹⁹ These and other efforts have provided a substantial body of scientific information (eg, Seabury et al,²⁰ Stein et al²¹), clinical experience, and US Food and Drug Administration–approved assessment approaches (https://www.fda.gov/medical-devices/neurological-devices/medical-devices-assessing-head-injury) that will help guide a systematic program of clinical trials to test specific treatments. Clearly, we are in a much better position to define specific subgroups of patients for clinical trials and monitor them appropriately than we were just 3 to 5 years ago.

Here, we lay out examples of specific TBI subdomains and candidate treatments (Table); each of these could, in our view, be the focus of a rigorously designed, multicenter, randomized, blinded, controlled clinical trial. The trials we propose and the domains we highlight are not meant to be an inclusive, fully validated, or prioritized list. They are primarily intended to exemplify what we mean by “one bite.”

TABLE.

Examples of Domain-Specific, Candidate Treatments for TBI

	Mood Disorders following TBI	Sleep Disorders and Fatigue following TBI	PTH Disordersa	TBI-Related Cognitive Disorders
Therapy	CBT delivered by smart-phone app for depression	Brief behavioral therapy for subacute insomnia	CBT for TBI-related migraine delivered by smart-phone app	Computer-based brain fitness training for working memory
Neuromodulation	rTMS targeted using resting-state fMRI network mapping vs sham stimulation for depression	Theta frequency transcranial alternating current stimulation during sleep for daytime fatigue	Vagal nerve stimulation for chronic PTH	Transcranial direct current stimulation for sustained attention
Pharmacology (primarily testing drugs approved for other indications)	Lamotrigine for chronic mood instability and irritability	Sodium oxybate for refractory chronic insomnia	CGRP antagonists for acute PTH	Long-acting stimulants for cognitive endurance
Lifestyle	Intense daily cardiovascular exercise for chronic mood instability	Progressive gentle daily cardiovascular exercise for chronic fatigue	Elimination of specific food triggers for chronic PTH	Low refined sugar diet for working memory
Hybrid/combination	rTMS paired with CBT for depression	Exercise plus low-dose stimulants for chronic fatigue	Botulinum toxin plus CGRP antagonists for refractory PTH	Vagal nerve stimulation paired with cognitive rehabilitation for memory

These are meant only as illustrative examples of trials that could be performed. They are not meant to imply any recommendation, prioritization, or well-validated equipoise.

^aThe domain of PTH disorders could be further subdivided into, for example, migrainous, cervicogenic, neuropathic, and tension-type PTHs; each of these could be subject to separate treatment trials (eg, CGRP antagonists for migrainous headaches, nerve blocks for cervicogenic headaches, etc).

CBT = cognitive behavioral therapy; CGRP = calcitonin gene–related peptide; fMRI = functional magnetic resonance imaging; PTH = post-traumatic headache; rTMS = repetitive transcranial magnetic stimulation; TBI = traumatic brain injury.

Running these trials will be a lot of work and cost 10s of millions of dollars. We will need good partners and an efficient network of clinical trial sites ideally overseen by a single institutional review board. Once the network is set up, Bayesian adaptive platform designs can be implemented to bring multiple candidate therapeutics into each ongoing trial in much the same way as the I-SPY2 breast cancer platform trial.²² As with breast cancer, it is likely that some subsets of TBI patients will benefit from specific treatments and some will not. Explicitly sorting out these subsets based on demographics and/or biomarkers will be an important goal and may be accomplished relatively efficiently using Bayesian adaptive designs, especially for the sort of large effect sizes our patients need. Overall, this platform approach will reduce the times, costs, and logistical barriers associated with each individual trial. It will bring us more quickly to the end goal of figuring out what works and what does not for TBI patients. Our “big hairy audacious goal”²³ is to test 30 candidate treatments over the next 10 years at less than the cost of 10 standalone trials. Then, in 10 years, we will look back and see how much of the elephant we have eaten.