Abstract
Objective
To investigate clinical characteristics and treatment patterns in persistent post-traumatic headache attributed to mild traumatic brain injury.
Methods
A total of 100 individuals with persistent post-traumatic headache attributed to mild traumatic brain injury were enrolled between July 2018 and June 2019. Deep phenotyping was performed using a semi-structured interview while allodynia was assessed using the 12-item Allodynia Symptom Checklist.
Results
In 100 subjects with persistent post-traumatic headache, the mean headache frequency was 25.4 ± 7.1 days per month. The most common headache phenotype was chronic migraine-like headache (n = 61) followed by combined episodic migraine-like and tension-type-like headache (n = 29) while nine subjects reported “pure” chronic tension-type-like headache. The most frequent trigger factors were stress, lack of sleep, and bright lights. A history of preventive medication use was reported by 63 subjects, of which 79% reported failure of at least one preventive drug, while 19% reported failure of at least four preventive drugs. Cutaneous allodynia was absent in 54% of the subjects, mild in 23%, moderate in 17%, and severe in 6%.
Conclusions
The headache profile of individuals with persistent post-traumatic headache most often resembled a chronic migraine-like phenotype or a combined episodic migraine-like and tension-type-like headache phenotype. Migraine-specific preventive medications were largely reported to be ineffective. Therefore, there is a pressing need for pathophysiological insights and disease-specific therapies.
Introduction
Post-traumatic headache (PTH) is a secondary headache disorder (1) and a common sequela of mild traumatic brain injury (TBI) (2). It accounts for ∼4% of all symptomatic headache disorders (3) and has a lifetime prevalence of 4.7% in men and 2.4% in women (4). Thus, PTH is a leading cause of injury-related disability and imposes a huge burden on patients, their families, and society.
In the International Classification of Headache Disorders 3rd edition (ICHD-3) (1), PTH attributed to TBI is defined by onset of headache within 7 days following trauma to the head. If the headache persists beyond 3 months of onset, it is characterized as persistent PTH. According to one prospective cohort study, ∼30% of mild TBI patients continued to complain of headache at 3 months post-trauma (5). Despite the widespread prevalence and disability associated with persistent PTH, it remains an under-recognized and not well characterized secondary headache disorder (6). In fact, the ICHD-3 criteria for persistent PTH do not mention any criteria for headache frequency, headache phenotype, or other clinical characteristics (1). Therefore, deep phenotyping of clinical characteristics and treatment patterns are much needed to improve disease characterization and wider recognition of persistent PTH by both physicians and patients.
We aimed to conduct a deep clinical phenotyping study to assess trauma and headache characteristics as well as treatment patterns in individuals with persistent PTH attributed to mild TBI.
Methods
Study population
We enrolled 100 individuals with persistent PTH attributed to mild TBI. Study participants were recruited from the outpatient clinic of the Danish Headache Center, the Danish post-concussion syndrome support group website (www.hjernerystelsesforeningen.dk), and from neurological departments and rehabilitation centers in the Capital Region of Denmark.
This article is part of a larger parental study that was approved by the Regional Health Research Ethics Committee of the Capital Region of Denmark (H-18011477). All participants gave written consent after receiving detailed oral and written information. The study was conducted at the Danish Headache Center in accordance with the Declaration of Helsinki (7), with later revisions. The study was registered on ClinicalTrials.gov (identifier: NCT03791515) and approved by the Danish Data Protection Agency.
Inclusion and exclusion criteria
The diagnosis of persistent PTH attributed to mild TBI was made according to the 3rd edition of the International Classification of Headache Disorders (ICHD-3) (1). Other inclusion criteria for individuals with persistent PTH were: a) Mild TBI to have occurred at least 12 months prior to study participation and b) age between 18–65 years. Exclusion criteria were a) any history of primary headache disorder (except infrequent tension-type headache); b) any history of more than one TBI or whiplash injury; c) pregnant or nursing women; d) cardiovascular or cerebrovascular disease; e) pre-trauma psychiatric disorder unless well-regulated, defined as no need of treatment within the past 12 months; and f) medication-overuse headache, as defined by ICHD-3 (1). Furthermore, we required no intake of analgesics within 24 hours prior to study participation. To further avoid inclusion of subjects with a pre-existing primary headache disorder (except infrequent tension-type headache), medical records were used to cross-check for any formal headache diagnosis code.
Study procedures
An in-person semi-structured interview was performed by a trained locum doctor (AI) and under supervision of two headache specialists (MA and HWS). The semi-structured interview was used to record data on demographics, medical history, and full clinical course. Data on demographics included questions about age, gender, height, weight, education, and self-rated health. To assess self-rated health, subjects were asked how they would rate their overall health according to the following response categories: Excellent, very good, good, rather poor, and poor. Medical history included psychiatric and somatic history of any kind (including pre-trauma chronic pain conditions) and was assessed as a series of “yes/no” questions, with a “yes” to any item scored as a positive history. To assess the full clinical course since the mild TBI, we collected data on the following: a) Trauma characteristics, b) headache characteristics, c) post-trauma neck pain and/or low back pain, d) medico-legal issues/litigation, and e) treatment patterns. Furthermore, the severity of cutaneous allodynia was assessed using the 12-item Allodynia Symptom Checklist (ASC-12).
Headache phenotype classification
Classifications of headache phenotypes were performed based on ICHD-3 criteria for primary headache disorders (1). A migraine-like phenotype was defined as any headache lasting at least 4 hours that fulfilled the C and D criteria for 1.1 Migraine without aura in accordance with ICHD-3 (1). In some study participants, photophobia and phonophobia may occur independent from their headache, as part of their post-concussive symptoms. Therefore, we decided that study participants who reported continuous photophobia and/or phonophobia (defined as continuously ongoing – daily symptoms) would only fulfill criteria D if they reported an increased severity of photophobia and phonophobia concurrent with an increase in headache intensity. Moreover, we defined a chronic migraine-like phenotype as headache occurring on 15 or more days per month for more than 3 months, with at least 8 days fulfilling the C and D criteria for 1.1 Migraine without aura in accordance with ICHD-3 (1). A tension-type-like headache phenotype (TTH-like) was defined as any headache that fulfilled the A to D criteria for one of the following: 2.1 Infrequent episodic tension-type headache, 2.2 Frequent episodic tension-type headache, or 2.3 Chronic tension-type headache. A combined episodic migraine-like/TTH-like phenotype was present if a study participant fulfilled the criteria for both a migraine-like phenotype and a TTH-like phenotype while not fulfilling the criteria for a chronic migraine-like phenotype; that is, the study participant experiences migraine-like headache at times and TTH-like at other times.
Definition of trigger factors for migraine-like headache exacerbations
Trigger factors were defined as events associated with an increased probability of migraine-like headache exacerbations. Identification of trigger factors were based on the self-perception of the study participants following an instruction in what defined a trigger factor.
Statistical analysis
Continuous data are presented as means and standard deviations (SD) while categorical data are presented as frequency counts and/or percentages. All statistical analyses were performed using IBM SPSS Statistics for Windows, Version 25.0, Armonk, NY, USA.
Results
Study population
Characteristics of the study population.
Localization of mild TBI was located to more than one region in some subjects.
Chronic pain was defined as persistent or recurrent pain that lasted longer than 3 months.
Pre-trauma psychiatric history and chronic pain conditions
Of 100 subjects with persistent PTH, 23% reported a pre-trauma psychiatric history (Table 1), with 14 subjects reporting a history of major depressive disorder and 12 subjects reporting a history of anxiety disorders. Four subjects also reported a history of eating disorders.
Pre-trauma chronic pain conditions were present in 12 of 100 subjects, with three subjects reporting a history of chronic low back pain while one subject reported a history of neuropathic pain (carpal tunnel syndrome). Nine subjects reported a history of “other chronic pain condition” (i.e. unspecific musculoskeletal pain in the shoulder, hip, or leg region), of whom one subject had also reported a history of chronic low back pain.
Trauma characteristics
Causes of injury included falls (42%), “other unintentional injury” (30%), motor vehicle collision (18%), sports-related injury (8%), and violence/assault (2%) (Table 1). Frontal impact was reported by 41% of the subjects, while temporal, parietal, and occipital impact was reported by 27%, 15%, and 27%, respectively (Table 1). The most common immediate symptoms following head injury were dizziness/vertigo (87%), impaired memory/concentration (84%), nausea (82%), and transient confusion, disorientation, or impaired consciousness (82%) (Table 1). Moreover, 83% reported onset of headache within 1 day after mild TBI while 13% did so between day 2 and 3; lastly, 4% reported onset of headache between day 4 and 7.
Headache characteristics
Headache and other clinical characteristics of the study population.
Mild headache intensity = does not impair ability to work and/or other activities; moderate headache intensity = impairs but does not prevent ability to work and/or other activities; severe headache intensity = prevents ability to work and/or other activities.
Positive family history of primary headache disorders = first degree relative with any primary headache disorder other than infrequent tension-type headache.
Continuous photophobia = continuously ongoing – daily – symptoms of photophobia.
Continuous phonophobia = continuously ongoing – daily – symptoms of phonophobia.
Cutaneous allodynia scores were defined as follows: none (scores 0–2), mild (3–5), moderate (6–8), and severe (≥9).
Classification of headache phenotypes were based on ICHD-3 criteria for primary headache disorders (5). The most common phenotype was a chronic migraine-like headache (n = 61) followed by a combined episodic migraine-like/TTH-like headache (n = 25), and a “pure” chronic TTH-like headache (n = 9), (Table 1).
Headache characteristics of the study population with migraine-like headache phenotype.
Trigger factors were defined as events associated with an increased probability of migraine-like headache exacerbations.
In nine subjects with a “pure” chronic TTH-like phenotype, the mean headache frequency was 29.7 ± 0.7 days per month, with most subjects reporting their usual headache to be of moderate intensity (Table 3). Headaches were most often bilateral (66%) and exclusively of pressing quality (100%). Continuous photophobia was reported by 44%, while none reported continuous phonophobia.
Neck pain, low back pain, and allodynia
Neck pain occurred in 78% of the subjects; approximately half of these experienced neck pain more than 180 days within the past 12 months (Table 2). Low back pain was reported by 37% of the subjects, with nearly one-fourth of these experiencing low back pain more than 180 days within the past 12 months (Table 2). Cutaneous allodynia was absent in 54% of the subjects, mild in 23%, moderate in 17%, and severe in 6% (Table 2).
Treatment patterns
Medication use of the study population.
A history of preventive medication use was reported by 63 subjects, of whom 55 subjects still used it. Among the former, 79% reported failure of at least one preventive drug, while 19% reported failure of at least four preventive drugs. For all preventive medications currently or previously used, lack of efficacy ranged from 67–100%. In terms of non-pharmacological treatments, the following were reported: acupuncture (n = 4), mindfulness therapy (n = 1), physical therapy (n = 1). All six subjects reported that they had benefited from these non-pharmacological treatments, although this was not assessed in further detail.
Discussion
Headache characteristics
The present study provides much-needed research into the clinical presentation of persistent PTH and its relation to primary headache disorders, such as migraine and TTH. Efforts to systematically investigate clinical characteristics will facilitate utilization of symptom assessment in the clinical management of individuals with persistent PTH. To our knowledge, no previous study has conducted deep phenotyping of clinical characteristics and treatment patterns in individuals with persistent PTH attributed to mild TBI.
In 100 subjects with persistent PTH, we found a mean headache frequency of 25.4 ± 7.1 days, with 92% defining most of their headache days as moderate or severe. In addition, the headaches were more often reported to be bilateral (64%), in the frontal region (70%), and of combined pressing and throbbing quality (45%). In comparison, one clinic-based study examined clinical characteristics at 7 to 10 days after mild TBI in 66 individuals with acute PTH (8). Typical characteristics of acute PTH were bilateral localization (56%), moderate or severe intensity (59%), and pressing quality (69%). However, ∼29% of these subjects had pre-existing headaches. Furthermore, none of the subjects reported headache at the follow-up assessment (at least 90 days post-trauma). Thus, it would be problematic to draw any direct comparisons to the present study population that had at least a 1-year history of persistent headache attributed to mild TBI and no pre-existing primary headache disorders (except infrequent TTH).
In terms of headache frequency, it is interesting that our study participants suffered from headaches on almost a daily basis. In comparison, Lucas et al. reported that only 27% experienced headaches several times per week to daily in a prospective cohort of mild TBI patients with persistent PTH (9). Although no solid data is available on individuals with persistent PTH, the discrepancy could in part be explained by differences in gender distribution. In the present study, 83% were females while only 14% were females in the study by Lucas et al. (9). In individuals with migraine, one population-based study found that females suffered from more frequent migraine attacks compared to males (10). As such, it could be speculated that in individuals with persistent PTH, females are more prone to have a higher headache frequency than their male counterparts. However, this hypothesis would have to be confirmed in a large population-based sample before causality can be inferred. In the present study, we found that 91% of the subjects had a migraine-like phenotype. This is consistent with findings from most studies of persistent PTH patients (9,11), albeit evidence to the contrary does exist (12). We would argue that reported differences in headache phenotype distribution is most likely due to methodological differences in classification of specific headache phenotypes. Also, headache phenotype classification may be further complicated by the myriad of symptoms that constitute post-concussion syndrome (5). Thus, individuals with persistent PTH may suffer often from post-concussion symptoms – such as photo- and phonophobia – independent from their headache. Among 91 subjects with a migraine-like phenotype, we found that 46% reported continuous photophobia and 60% reported continuous phonophobia independent from their headache. Therefore, we believe that any future migraine-like phenotype classification should consider including a requirement of increased severity of photo- and phonophobia concurrent with an increase in headache intensity. In support, such an association has previously been reported in population-based migraine studies (13,14). Moreover, we found that cutaneous allodynia was absent in 54% of individuals with persistent PTH, mild in 23%, moderate in 17%, and severe in only 6%. In a population-based study (15), it was found that cutaneous allodynia was absent in 37% of migraine sufferers while 42% reported moderate or severe cutaneous allodynia. Thus, sensory profiles appear to differ between individuals with persistent PTH and individuals with migraine, which may be suggestive of distinct pathophysiological mechanisms. To further explore disease-specific sensory profiles, future studies should use a validated quantitative sensory testing (QST) protocol. To date, only two studies have performed QST in individuals with persistent PTH and both studies included subjects with moderate to severe TBI (16,17). As such, future studies should emphasize QST assessment in a large sample of individuals with persistent PTH attributed to mild TBI. In order to better understand the clinical similarities between PTH and migraine, we assessed trigger factors of migraine-like headache exacerbations, with the most common ones being stress (∼73%), lack of sleep (∼69%), and bright lights (∼60%). Interestingly, trigger factors have been reported in individuals with migraine (18,19). In 91 subjects with a migraine-like phenotype, we found that the three most frequently reported associated symptoms were photophobia (∼96%), phonophobia (∼96%), and nausea (∼71%). In conjunction with the reported trigger factors, our findings provide new data to the ongoing debate of shared disease mechanisms between migraine and PTH (6). To further explore this topic, more studies are needed, including biochemistry and neuroimaging studies.
The optimal management approach to individuals with persistent PTH is a unique challenge for clinicians since there are no acute or preventive medications approved for PTH. A recent systematic review highlighted the complete lack of randomized clinical trials (RCTs) in the PTH field (20). Current management strategies are often based on treating PTH according to the primary headache phenotype it resembles the most (21). In the present study, we found that 98 of 100 subjects had tried acute medication(s) while 63 subjects had a history of preventive medication use. Although 87 of 100 the subjects reported dissatisfaction with their current treatment status, 89 subjects still reported current use of acute medication while 55 subjects reported current use of preventive medication. Thus, future studies should assess more optimal treatment strategies and whether current clinical strategies involve unnecessary exposure to ineffective treatments. Moreover, another interesting finding was that 39 of 100 subjects currently or previously used triptans, with approximately half of these reporting lack of triptan efficacy. However, a placebo-controlled study would be needed to reliably assess the efficacy of triptans as an acute medication for persistent PTH. Moreover, of 63 subjects who had a history of previous preventive medication use, 79% had failed at least one preventive drug while almost one-fifth had failed four or more migraine-specific preventive drugs. Based on our data, it seems that conventional migraine preventive drugs lack efficacy, which raises the question of whether disease mechanisms underlying PTH are different from those underlying migraine, despite the phenotypic similarities. Thus, novel therapeutic options are needed to address the huge burden and disability associated with persistent PTH.
An interesting finding from the present study was that only 11 of 100 subjects reported a chronic pain condition prior to the mild TBI. In addition, no subjects reported chronic pre-trauma neck pain and only three subjects reported pre-trauma chronic low back pain. These findings may likely be an underestimation of the actual number and subject to recall bias. In support, one Swedish population-based study reported substantially higher prevalences of chronic neck pain and low back pain (22). Nevertheless, it remains highly interesting that a considerable proportion of subjects in the present study reported frequent occurrence of post-trauma neck pain and low back pain. Chronic pain conditions have previously been independently associated to low self-rated health (23). In this work, we found that 52% of the subjects reported being in “rather poor” or “poor” health. In comparison, 72% of the Danish people reported being in good health (24). Thus, our data suggest that persistent PTH results in an enduring burden of detrimental effects on self-rated health.
Besides chronic pre-trauma pain conditions, we also assessed any history of pre-trauma psychiatric disorders, with the most frequently reported disorders being major depression (14%), anxiety disorders (12%), and eating disorders (4%). In comparison, summary data from the European Union found that the 12-month prevalence of major depression was 6.9% while the 12-month prevalence of anxiety disorders was 14% and less than one percent for eating disorders (25). Future large-scale studies should assess the prevalence of post-trauma psychiatric comorbidities in individuals with persistent PTH and discuss a possible cause-effect relationship.
Limitations
A limitation of the present study was that the study population was recruited from three different settings: a) The specialized outpatient clinic of the Danish Headache Center (n = 55), b) various neurology and rehabilitation centers (n = 24), and c) the website of a patient-support group (n = 21). This may result in some heterogeneity between study participants that were recruited from different settings, although of the 55 subjects recruited from the Danish Headache Center, 89% had a migraine-like phenotype compared to 91% in the overall study population of 100 individuals. Nevertheless, the study population might be skewed towards individuals more burdened by persistent PTH and consequently the results may not be representative for individuals with persistent PTH who are less burdened by their headache. Another limitation was that family history of primary headache disorders was assessed by study participant report. In migraine, the number of affected relatives with migraine was considerably underestimated when assessed by study participant report compared to clinical interviews of first-degree relatives (26). Moreover, data collection was performed using a semi-structured interview instead of a prospectively collected headache diary. Consequently, our data is subject to recall bias. In addition, assessment of trigger factors of migraine-like headache exacerbations might also be influenced by recall bias, personal beliefs, and reverse causality (27) as well as the likelihood of false attribution given the high frequency of both headache and migraine-like exacerbations. As such, there remains a pressing need for long-term prospective deep phenotyping studies in the PTH field. This will facilitate multidimensional outcome assessments, better treatment strategies, and potentially a robust prognostic model that can be used in mainstream clinical practice. Moreover, it is possible that at least some study participants suffered from a whiplash injury in relation to their mild TBI. However, we assessed trauma history thoroughly and used the ICHD-3 definition of whiplash injury for exclusion. Also, none of the study participants reported being involved in a rear-end collision, which is the most common cause of whiplash injury.
Conclusions
Persistent PTH is a common sequela of mild TBI. In most individuals, we found that PTH resembles a migraine-like headache phenotype, with similar trigger factors and associated symptoms. However, migraine-specific preventive medications were reported to lack efficacy. Consequently, current management strategies appear to cause unnecessary patient exposure to ineffective treatments. Therefore, there is a pressing need for pathophysiological insights and PTH-specific treatments that seek to maximize effectiveness by taking into account clinical characteristics of individual patients.
Footnotes
Clinical implications
Persistent PTH often resembled a migraine-like headache phenotype with similar trigger factors and associated symptoms.
Migraine-specific preventive medications lacked efficacy and consequently current therapeutic strategies may involve unnecessary exposure to ineffective treatments.
Future studies should investigate whether phenotype-guided therapeutic approaches improve treatment outcomes.
Author contributions
HA: Study concept and design, acquisition of data, analysis (including statistical analyses) and interpretation, drafting the manuscript. AI: Acquisition of data and critical revision of the manuscript for important intellectual content. HMA-K: Acquisition of data and critical revision of the manuscript for important intellectual content. SA: Analysis (including statistical analyses) and interpretation, critical revision of the manuscript for important intellectual content. JO: Critical revision of the manuscript for important intellectual content. RHJ: Critical revision of the manuscript for important intellectual content. FMA: Critical revision of the manuscript for important intellectual content. MA: Study concept and design, critical revision of the manuscript for important intellectual content. HWS: Study concept and design, critical revision of the manuscript for important intellectual content, supervision.
Declaration of conflicting interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: SA has received honoraria from Allergan and Teva and consultant fees from Amgen and Allergan; and honoraria for consulting for Allergan, Amgen, Eli Lilly, Novartis, Promius, Satsuma, Supernus, and Theranica. FMA is a lecturer or scientific advisor for Teva and Novartis. MA has received personal fees from Alder BioPharmaceuticals, Allergan, Amgen, Eli Lilly, Novartis, and Teva, and also participated in clinical trials as the principal investigator for Alder, Amgen, electroCore, Novartis, and Teva. MA also serves as an associate editor of Cephalalgia, associate editor of Headache and co-editor of the Journal of Headache and Pain. MA reports research grants from Lundbeck Foundation, Research Foundation of the Capital Region of Copenhagen, and Novo Nordisk Foundation. RHJ has given lectures for Pfizer, Allergan, Merck, ATI, TEVA, Novartis and Conducted clinical trials for Electrocore, ATI, and Eli-Lilly. HWS received speaking fees from Novartis and Teva. The other authors declare no conflicts of interest.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was supported by a grant from the Rigshospitalet Research Foundation (F-23340-02).