Abstract
Background/objective
Post-traumatic headache is one of the most common and persistent symptoms following mild traumatic brain injury. The objective of this narrative review is to provide an update on the diagnostic criteria, clinical presentation, epidemiology, pathophysiology, and treatment of post-traumatic headache, and to identify future research priorities.
Methods
This is a narrative review of the literature regarding post-traumatic headache attributed to mild traumatic brain injury.
Results
Onset of post-traumatic headache within 7 days of injury is the only evidence for a causal relationship between the injury and the headache included in the diagnostic criteria. Post-traumatic headache often resolves within the first few days of onset, whereas it persists for at least 3 months in 30–50%. The majority of insights into post-traumatic headache pathophysiology come from pre-clinical animal studies and human imaging studies, which implicate structural, functional, metabolic, and neuroinflammatory mechanisms for post-traumatic headache. There is a paucity of quality evidence for how to best treat post-traumatic headache.
Conclusions
Although meaningful progress has been made in the post-traumatic headache field, priorities for future research are numerous, including the optimization of diagnostic criteria, a greater understanding of post-traumatic headache pathophysiology, identifying mechanisms and predictors for post-traumatic headache persistence, and identifying safe, well-tolerated, effective therapies.
Introduction
Post-traumatic headache (PTH) is a secondary headache disorder that is attributed to an underlying head and/or neck injury. PTH is one of the most common symptoms following a traumatic brain injury (TBI) and is one of the most likely symptoms to persist following such an injury (1). Acute and persistent PTH result in substantial pain and disability (2,3). The high prevalence and impact of PTH underscore the need for its recognition, an understanding of its pathophysiologic mechanisms, accurate diagnosis and prognostication, and the discovery of effective therapies.
In this narrative review, we discuss the diagnostic criteria, epidemiology, clinical presentation, pathophysiology, and treatment of PTH, with a focus on PTH attributed to mild TBI (mTBI). Areas of greatest need for future research are proposed.
Diagnostic criteria and clinical presentation
The International Classification of Headache Disorders 3rd edition (ICHD-3) recognizes three injury types that can lead to PTH: TBI, whiplash, and craniotomy (4). TBI is further subcategorized into mTBI versus moderate to severe TBI. According to the ICHD-3 criteria, PTH must begin within 7 days of: i) the injury, ii) regaining consciousness following the injury, or iii) discontinuation of medications that impair the ability to sense or report headache following the injury. PTH is considered “acute” if it has resolved within 3 months after its onset or if has not yet resolved but 3 months have not yet passed since its onset, and “persistent” if it has been present for greater than 3 months since its onset. Three months was chosen as the threshold for considering PTH to be “persistent” to be consistent with the definition for “persistent” used for other secondary headaches in the ICHD-3.
Post-traumatic headache research priorities.
In the ICHD-3 criteria, the only evidence for causality between the underlying injury and PTH is the time interval between the injury and headache onset. The ICHD-3 classification committee chose to keep this interval relatively short, to maximize the specificity of the criteria and since data suggest that the majority of PTH does indeed begin within 7 days of injury (5–7). However, it is recognized that some PTH might begin after the first 7 days (6,8,9). Thus, the ICHD-3 Appendix includes diagnostic criteria for “delayed onset” PTH, which allow for the headache to begin up to 3 months following the inciting injury. In the author’s opinion, it is unlikely that headaches that begin within 7 days of injury and those that begin in delayed fashion closer to 3 months after the injury have the same underlying pathophysiologic mechanisms. In at least some situations, delayed onset headaches after an injury could be due to other post-traumatic symptoms such as sleep problems, autonomic dysfunction, vestibulo-ocular dysfunction and changes in normal physical and mental activities. Criteria for delayed-onset PTH need to be field tested to determine the most appropriate interval between injury and headache onset.
There are no specific headache characteristics that contribute to the diagnosis of PTH, since it can manifest with different symptoms that vary from patient to patient, and there are not yet any symptoms or headache characteristics that have been demonstrated to be specific for PTH. It is common to describe PTH according to another primary or secondary headache that it most resembles. For example, if PTH has a lot of migraine features, it could be described as “PTH of a migraine-like phenotype”. The most common PTH phenotypes are that of migraine and tension-type headache (8,10–12). Whether migraine or tension-type headache is the most common phenotype varies amongst studies (3,12,13). Differences in phenotypic proportions might be due to the most common types and severities of injuries leading to PTH in different studies, whether the data are being collected from patients who have sought care for their PTH versus being collected during TBI screening, differences in proportions of individuals who had primary headaches prior to the injury that caused PTH, and different approaches to assigning headache diagnoses. Since migraine and PTH often share symptoms, a few studies have compared symptoms associated with PTH to those associated with migraine. These studies have found substantial overlap in symptoms, perhaps with some differences in the severity of sensory hypersensitivities such as allodynia, photophobia, and phonophobia, nausea, insomnia, and symptoms of autonomic nervous system dysfunction (14–18).
Whereas the diagnosis of PTH is fairly straightforward in the patient who had no recent history of headaches prior to the TBI and then had headache onset very soon after sustaining a head and/or neck injury, in the author’s opinion the diagnosis is more challenging when the patient was having headaches prior to the injury. For example, the patient who has regular migraine attacks prior to a TBI who then has a worsening pattern of headaches following the TBI might be experiencing a migraine exacerbation versus new onset of PTH with a migraine-like phenotype. These two diagnostic possibilities can be difficult to differentiate in the clinical setting. The ICHD provides guidance for this situation, instructing that both the primary headache diagnosis (e.g. migraine) and the secondary headache diagnosis of PTH be given: “When a pre-existing primary headache becomes chronic or is made significantly worse (usually meaning a twofold or greater increase in frequency and/or severity) in close temporal relation to such a causative disorder, both the primary and the secondary headache diagnoses should be given, provided that there is good evidence that the disorder can cause headache” (4). However, further investigations are needed to determine if the pathophysiology, clinical presentation, prognosis, and response to treatment are similar in individuals who have PTH with no history of prior headaches versus those who had pre-existing headaches that seem to have worsened or changed following injury.
Although PTH might be the only symptom following head injury, it is commonly accompanied by other TBI-related symptoms, including cognitive dysfunction, imbalance or vertigo, personality changes, depression, anxiety, post-traumatic stress, sleep problems, oculomotor dysfunction, autonomic nervous system dysfunction, and extracranial pain (16,17,19–21). Evaluation and management of the patient with PTH must consider the entire complex of symptoms and how they might interact with one another.
Epidemiology
Globally, new TBIs affect nearly 70 million people each year, with about 60–80% being classified as mTBI (22). PTH is amongst the most common symptoms following mTBI, although the prevalence of acute PTH following mTBI is likely underreported and not well established. PTH often resolves within the first several days of onset. However, in about 30–50% of cases, PTH persists for at least 3 months and thus is considered “persistent” (8,23). A systematic review of the literature found a 58% prevalence of “chronic” headache following TBI, perhaps more strongly associated with mTBI compared to moderate or severe TBI (24). Several risk factors for persistence of post-TBI symptoms, perhaps including persistent PTH, have been suggested: Gender, age, marital status, education, alcohol/drug/tobacco use, history of neurological disease, history of psychiatric disease, pre-injury headaches, prior TBIs, abnormal head CT findings, injury mechanism, extracranial injuries, duration of post-traumatic amnesia, and symptoms of anxiety and depression (25–27). However, a recent systematic literature review found only three published studies on risk factors for the development of acute and/or persistent PTH attributed to TBI that had used ICHD diagnostic criteria, none of which assessed risk factors for the development of persistent PTH (28). Thus, there is a need for well-designed studies that identify risk factors for PTH persistence. Development of a predictive model that accurately estimates the likelihood of early PTH resolution versus PTH persistence would be valuable for clinical practice and for identifying individuals who are the most appropriate for enrolment into PTH therapeutic clinical trials. A predictive model might include clinical characteristics, blood-based biomarkers, imaging findings, or other data. A recent study developed a predictive model for persistence of PTH at 6 months following a motor vehicle accident (25). Baseline data were collected at a median of 10 days after injury. A model including the following characteristics had 95.5% (95% CI: 91.1–97.8%) specificity and 27.2% (95% CI: 20.4–35.2%) sensitivity for PTH persistence in a validation cohort: Older age, not employed, severe headache pain intensity, symptoms in arms or hands, dizziness or unsteadiness, neck stiffness, pre-existing headache, and lower recovery expectations by the patient. Studies that use the ICHD criteria for PTH and its persistence are needed to determine if this predictive model is valid and generalizable.
Pathophysiology
Much of the understanding about PTH pathophysiology comes from pre-clinical animal studies of pain behaviors and sensitization following mTBI, and from human neuroimaging studies of those with PTH. mTBI results in structural and functional changes that lead to post-TBI symptoms including PTH (29). TBI causes alterations in pain-processing and modulating brain regions and networks that result in an imbalance between pain facilitation and pain inhibition. In support of this mechanism, animal studies have demonstrated a loss of diffuse noxious inhibitory controls (DNIC) after mTBI, while human studies have demonstrated a loss of conditioned pain modulation, findings reflective of a reduction in pain inhibition (30–32). As reviewed more comprehensively in an article by H Ashina and colleagues, cellular injury due to mTBI results in loss of normal ionic homeostasis, energy depletion, oxidative stress, and axonal damage, processes that could contribute to the development and persistence of PTH (29,33,34). TBI, including mTBI, might trigger cortical spreading depression (CSD) at the time of injury and be associated with the extent of brain tissue injury and functional outcomes (35). As has been theorized for migraine, CSD might directly activate the trigeminocervical system resulting in headache (36–38). In fact, some patients with PTH are reported to experience auras with their PTH (12,18,39). Microglia proliferation in the trigeminal nucleus caudalis could contribute to PTH via pro-inflammatory actions and altered neuron-microglial signaling (40,41). Finally, neurogenic inflammation after mTBI, including the release of calcitonin gene-related peptide (CGRP), might contribute to the initial development of PTH as well as the development of trigeminal and extratrigeminal sensitization, which facilitate PTH persistence (42,43). Pre-clinical studies that block CGRP have demonstrated benefits for reducing post-TBI pain behaviors, early sensitization, late latent sensitization, and photophobia (44–46). Preclinical and human studies suggest that CGRP might play more of a role early after TBI and PTH onset, with a diminished role after PTH has become persistent (45,47). However, experimental CGRP infusions in people with persistent PTH can result in headache exacerbations with migraine-like features, suggesting that persistent PTH is associated with hypersensitivity to CGRP (48).
In humans, PTH imaging studies have begun to provide insights into brain regions and connections that might be important in PTH pathophysiology, investigated similarities and differences to other headache types, and have used imaging findings to predict patient outcomes. Results suggest that PTH is associated with alterations in structure of several brain regions, many within the frontal lobes (49–52). Furthermore, there is abnormal structural and functional connectivity associated with PTH, measured via diffusion tensor imaging and resting state functional connectivity studies (52–54). Imaging findings suggest that those with PTH have changes in the ability to properly modulate painful stimuli, likely in the direction of having less capability for pain inhibition. Some imaging findings also suggest that PTH is associated with alterations in affective components of pain processing and in multisensory integration, perhaps explaining why PTH is commonly associated with other symptoms such as light and sound hypersensitivity. Even when accounting for the effects of the TBI (by comparing those with PTH to those who had TBI but do not have PTH), many differences in brain structure and function persist, suggesting that there are brain structural and functional findings that are attributable to PTH instead of the underlying TBI (49,55). Comparisons of PTH to migraine demonstrate overlap in regions involved in the two headache types, but also demonstrate differences in brain structure and functional connectivity, even when those with PTH have a migraine phenotype (50,53,56). Finally, brain imaging findings might be predictive of future clinical outcomes, as has been demonstrated in a study showing that periaqueductal gray resting state functional connectivity measured within 7 days of injury accurately predicts persistent PTH at three months (57).
Treatment
There are no established guidelines on when to start symptomatic and preventive therapy for PTH. Thus, at the current time, the decision is made subjectively. PTH severity, frequency, and impact on patient function should be considered, as should the longitudinal course of PTH. Those who are demonstrating meaningful improvement without therapy might not need therapeutic intervention. Symptomatic therapy is often needed, especially for those with at least moderate intensity headaches. The frequency of symptomatic medication use should be monitored, so as to initiate preventive therapy prior to a patient developing medication overuse/medication overuse headache.
There is a lack of published evidence indicating which symptomatic and preventive therapies are useful for PTH treatment. There are no large, prospective, randomized, controlled clinical trials of PTH pharmacotherapy. A recent systematic review of the literature concluded that published studies provide low-grade evidence on the pharmacological treatment of PTH (58). The low-grade evidence includes study results for symptomatic therapies including NSAIDs, acetaminophen, triptans, anti-emetics, diphenhydramine, and peripheral nerve blocks, and preventive therapy with gabapentin, tricyclic antidepressants, propranolol, valproate, topiramate, flunarizine, and melatonin (58). Since the systematic review was performed, there are additional data on the use of erenumab and transcranial magnetic stimulation for treatment of persistent PTH (59,60). Because there is a lack of evidence for how to best treat PTH, it is often suggested to treat PTH like the other headache type it most resembles. According to this recommendation, PTH that has a lot of migraine symptoms would be treated like migraine, and PTH that resembles tension-type headache would be treated like tension-type headache. Although this approach is often used in clinical practice, there are not sufficient data to suggest that this is an effective approach.
When treating PTH, it is essential to recognize that it is often accompanied by other post-injury symptoms such as problems with sleep, cognition, mood, and balance, symptoms of autonomic nervous system dysfunction, and cervical pain and muscle tension. Since these other symptoms can contribute to worsening of PTH, treatment of these other symptoms, as indicated, can be essential for improving the overall health of the patient and for obtaining improvements in PTH. Such treatments might include cognitive rehabilitation, vestibular rehabilitation, physical and occupational therapy, and treatment of autonomic dysfunction, mood, and sleep disorders. When starting treatments for PTH, it is important to ensure that the treatments and their side effects will not exacerbate other symptoms such as sleep and cognitive problems.
Future directions
Clinical diagnostic criteria for PTH need to be tested, refined, and validated. The optimal interval between injury and PTH onset should be further studied and modified if supported by emerging data. This includes testing of the ICHD-3 Appendix diagnostic criteria for “delayed onset” PTH. Arguments in favor of extending the 7-day interval need to be balanced with the possible negative impact of reducing the specificity of the criteria. In addition to investigating similarities and differences in injury characteristics, symptoms, and prognosis amongst those who have PTH onset within 7 days versus those who have onset after 7 days, similarities and differences in pathophysiology should also be investigated. Consideration could be given to sub-classifying PTH according to headache frequency, as has been done in migraine with chronic migraine diagnostic criteria. At the current time, it does not seem as if specific headache characteristics or associated symptoms will be useful for differentiating PTH from other headache types, but this also needs to be further explored. It is possible that a constellation of symptoms could be more indicative of PTH compared to other headaches (e.g. a combination of headache characteristics with psychological symptoms and symptoms of autonomic nervous system dysfunction). Longitudinal research is needed to determine whether 3 months as the threshold for defining ‘persistent’ PTH is optimal, or if there is another time threshold that better represents an inflection point in PTH longevity that is associated with a low likelihood of natural resolution and a high likelihood of continued headache. Research is needed to determine the extent to which non-headache symptoms observed in individuals with PTH are part of a PTH syndrome versus being independent from the PTH and attributed to the underlying injury. Objective biomarkers for PTH diagnosis and prognosis need to be identified and then used to refine and optimize the diagnostic criteria and prognosticate outcomes.
Further insights into the pathophysiology of acute and persistent PTH are needed. There are now several animal models of PTH that have contributed meaningfully to our mechanistic understanding of PTH. Validation of these models, translation of findings to humans, and a greater understanding of how each model might differentially contribute to understanding specific aspects of TBI and PTH are needed. Research that yields insights into differences that might be attributable to injury mechanism (e.g. blast vs. direct impact), pre-injury existence of primary headaches such as migraine, and the impact of comorbid symptoms and conditions (e.g. anxiety, sleep problems) is much needed. A research approach that combines different data types, such as clinical data, genetics, imaging, body-fluid-based biomarkers, and longitudinal outcomes might yield important insights into PTH pathophysiology. Work that identifies the similarities and differences in pathophysiology between PTH and primary headaches is needed to confirm that PTH is a headache type distinct from the primary headaches, even when it has similar symptoms.
A substantial amount of work is still needed to identify existing and new therapies that provide the most benefits for patients with PTH. Guidelines on when to initiate symptomatic and preventive therapy for PTH would assist the clinician. Treatment strategy guidelines would be particularly useful if they were based not only on expert opinion, but also on data that show the optimal timing for initiating symptomatic and preventive therapies with the goal of optimizing benefits while limiting side effects, adverse events, and the number of individuals who receive treatment that would have otherwise had natural resolution. Guidelines for trial design and outcomes assessment for PTH clinical trials should be developed. There are inherent challenges in PTH clinical trial designs due to difficulties estimating rates of natural PTH resolution, varying injury mechanisms, and differences in post-TBI symptoms amongst potential participants. Therapies that provide the most benefit in the acute (first few days after PTH onset), subacute (first few weeks after PTH onset), and persistent phases of PTH need to be identified. Optimally, a therapy would treat PTH in the acute or subacute phase and prevent it from becoming persistent. Models that predict who will have PTH persistence would assist the clinician when determining the need for PTH therapy and the researcher when determining which individuals are most appropriate for therapeutic clinical trials. Future studies that aim to identify predictors for persistent PTH should use ICHD criteria for defining PTH and its persistence. Although predictive models based on clinical features alone (e.g. demographics, injury features, post-injury symptoms, pre-injury medical conditions, patient expectations) would be optimal since data would be easy to collect, the added value of body fluid-based biomarkers, and imaging and neurophysiology results should be determined. Fortunately, there are numerous PTH clinical trials that have been registered in publicly available repositories, studying interventions such as CGRP mAbs, botulinum toxin, triptans, amitriptyline, anti-emetics, nerve blocks, cognitive behavioral therapy, acupuncture, neuromodulation, prazosin, and others (clinicaltrials.gov, accessed 27 August 2020). These new studies should test the often-stated clinical advice that PTH should be treated using therapies that are effective for the headache type that the PTH most closely resembles. Future studies should help determine if such advice is valid, if there are other predictors of treatment response, or if new therapies need to be developed that are specific for treating PTH, even if not currently known to be effective for the treatment of other headache types (Table 1).
Conclusions
PTH is a common and often persistent symptom following mTBI. Although meaningful advances have been made in defining the epidemiology, clinical presentation, diagnostic criteria, and pathophysiology of PTH due to mTBI, much more work is needed. Emphasis should be placed on determining the frequency of acute and persistent PTH, prediction of PTH persistence, optimizing PTH diagnostic criteria, translating pre-clinical animal study results to humans, developing clinical trial and treatment guidelines, and identifying safe and effective symptomatic and preventive therapies. Fortunately, recent years have seen a greater focus on PTH, providing hope to the millions of individuals with PTH.
Key findings
Post-traumatic headache is common and often persistent following mild traumatic brain injury. Several mechanisms for the development and persistence of post-traumatic headache have been proposed including neuro-inflammation, microglial proliferation, cellular injury with loss of normal homeostasis, axonal injury, loss of normal pain processing and modulation, and cortical spreading depression. There is a lack of high-quality evidence for when to initiate PTH treatment and for which therapies are safe and effective when treating acute and persistent PTH. Further work is needed to better define PTH epidemiology, diagnostic criteria, pathophysiology, and treatment.
Footnotes
Declaration of conflicting interests
The author declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: TJS has received research support from the American Migraine Foundation, Amgen, Henry Jackson Foundation, National Institutes of Health, Patient-Centered Outcomes Research Institute, and the US Department of Defense. Within the past 12 months, he has received personal compensation for serving as a consultant or advisory board member for Abbvie, Alder, Allergan, Biohaven, Click Therapeutics, Eli Lilly, Equinox, Lundbeck, and Weber & Weber. He holds stock options in Aural Analytics and Nocira. He serves as an Associate Editor for Headache and Cephalalgia. He receives royalties from UpToDate.
Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: TJS’ time writing this manuscript was partially supported by the Office of the Assistant Secretary of Defense for Health Affairs, through the Peer Reviewed Medical Research Program under Award No. W81XWH-19-1-0534. Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the Department of Defense. The U.S. Army Medical Research Acquisition Activity, 820 Chandler Street, Fort Detrick MD 21702–5014 is the awarding and administering acquisition office.
