Posttraumatic Headache: Basic Mechanisms and Therapeutic Targets

Joshua Kamins, MD; Andrew Charles, MD


Headache. 2018;58(6):811-826. 

In This Article

Basic Mechanisms of PTH as Targets for Therapy

The main elements of the physiology of acute concussion include cellular ionic fluxes, cell swelling, axonal injury, excessive release of multiple neurotransmitters, a mismatch between cerebral metabolism and cerebral blood flow, and disruption of the blood brain barrier. These are followed by upregulation of inflammatory signals, more chronic changes in cerebral blood flow and metabolism, and neuroendocrine responses.[30,31] In many respects (except for the tissue injury), the acute cellular and neurochemical effects of TBI parallel those of cortical spreading depression (CSD), the slowly propagated wave of brain activity that is believed to be the pathophysiological substrate of the migraine aura.[32] In humans with focal brain injury, repetitive spreading depolarizations similar to CSD have been observed to emanate from the site of injury.[33] Whether or not spreading depolarization occurs in concussion remains unclear. While a variety of visual symptoms are reported following mild TBI, some of which may be related to deficits in cortical visual processing,[34] it is not particularly common for patients to develop typical slowly propagated migraine visual aura symptoms in the setting of acute concussion, or other slowly progressing neurological symptoms suggestive of CSD. It may be that TBI evokes similar mechanisms as those that cause CSD, but because in the case of concussion the injury is more diffuse there may be no opportunity for focal initiation of a propagated event to occur.

Significant evidence regarding the cascade of events in TBI has comes from animal models.[35] In humans, however, direct cellular and metabolic monitoring has occurred primarily in the ICU settings for more severe brain injuries. Techniques including EEG, cerebral blood flow measurements, structural and functional MRI, MR spectroscopy, and serum studies have been used to evaluate the nature and duration of physiological dysfunction after mild TBI.[36] Imaging and serum biomarkers confirm that metabolic and cerebral blood flow derangements typically last up to 30 days post injury. MRI studies, however, have shown persistent white matter tract changes, hyperconnectivity, and altered metabolism at 3 months after injury, indicating that brain changes may outlast clinical symptoms.[37] At this stage, none of these acute or sustained changes have been specifically correlated with headache in a manner that could guide therapeutic intervention. Nonetheless, the initial period following concussion might represent an important therapeutic window during which mechanisms underlying the development of chronic headache could be prevented. It is, therefore, important to consider which basic mechanisms of the response to TBI represent potential targets for therapy.

Neurovascular and Metabolic Function

Acute symptomatology of concussion may be caused in part by impairment of neurovascular coupling and mismatch between metabolic demand and blood flow. Both reduced and increased cerebral blood flow, as well as altered vascular reactivity have been reported following mild traumatic brain injury.[38] The variability in cerebral blood flow studies may be related to different techniques, regional differences in the vascular response, the timing relative to trauma, and the presence or absence of symptoms. While alterations in neurovascular function typically resolve by 30 days following trauma,[36] in some cases changes may persist. One study reported persistently increased perfusion in the left dorsal anterior cingulate cortex 6 weeks posttrauma in athletes with mild TBI compared with controls, with greater hyperperfusion seen in those with symptoms as compared with those without.[38] Similarly, a prospective cohort study of pediatric patients using arterial spin labeling MRI to evaluate cerebral perfusion after concussion found that at approximately 40 days post injury, global CBF was higher in the symptomatic group and lower in the asymptomatic group compared to controls.[39] Whether these blood flow changes are playing any causative role in symptomatology vs simply representing a physiological response to ongoing pathology remains unclear. Regardless, this study does indicate that prolonged alteration in cerebrovascular function could occur with prolonged PTH, and that further study of blood flow in chronic PTH may be productive.

Disruption of normal cellular energy metabolism represents another mechanism with potential overlap between migraine and posttraumatic headache. In migraine, there has been longstanding speculation (with little supporting evidence) that changes in mitochondrial function and metabolism could play a causative role[40,41] This hypothesis is part of the rationale for some nonprescription migraine preventive therapies such as riboflavin and CoQ10. Magnetic resonance spectroscopy studies have demonstrated alterations in brain chemistry suggesting impaired energy metabolism in migraine patients vs controls, but thus far no coherent picture has emerged regarding consistently altered metabolic function.[42,43] In the setting of concussion, derangements in metabolic function are also believed to play an important role in clinical symptoms, particularly in the acute phase following injury. Depression of normal glucose metabolism has been found both in animal and human mild TBI,[44–46] and animal studies indicate that impaired glucose delivery across the blood brain barrier may play a role.[46] As with migraine magnetic resonance spectroscopy studies indicate impaired energy metabolism in mild traumatic brain injury with substantial variability across different studies.[47–49] Whether or not alterations in energy metabolism play any role in posttraumatic headache remains unclear, but if they do, they represent an appealing therapeutic target.

Apart from glucose, ketones are considered to be the primary endogenous fuel that can contribute significantly to cerebral metabolism. Preclinical models of mild to severe traumatic brain injury have demonstrated improved outcomes with pre and post injury administration of the ketogenic diet.[50] These studies raise the possibility that dietary approaches could have therapeutic benefit for posttraumatic headache (although there is currently no evidence to support this possibility). Exogenous creatine supplementation is another possible therapy that may be worthy of further study. Creatine supplementation is widely used for athletic performance.[51] Creatine's role in energy metabolism involves the transfer of N–phosphoryl groups from phosphorylcreatine to adenosine diphosphate (ADP) to regenerate adenosine triphosphate (ATP). Given that creatine has its own Na+ Cl– dependent transporter across the blood brain barrier, it is possible that creatine supplementation could have therapeutic benefit in the setting of cerebral metabolic depression due to concussion,[52] and by extension, benefit for posttraumatic headache.


Glutamate is the primary excitatory neurotransmitter in the central nervous system. There is strong animal evidence that TBI causes substantial release of glutamate and alteration of glutamate receptors, and glutamate has also been implicated in migraine.[53] In migraine, elevated serum and CSF levels of glutamate have been reported both ictally and interictally, suggesting that acute and sustained increases in glutamate levels could play a role in headache.[54] Cortical glutamate levels have also been reported to be increased in migraine patients.[55] Topiramate, an inhibitor of the kainate receptor subtype of glutamate receptors, is widely used as a migraine preventive therapy.[56] Memantine, an activity dependent inhibitor of the NMDA receptor subtype of glutamate receptors, has also been used for the prevention of migraine (although it is not approved for this indication),[57] and has also shown benefit in animal models of traumatic brain injury.[58] Other medications that inhibit NMDA receptors, including dextromethorphan and ketamine, are available and in clinical use for other indications. There is some evidence that ketamine may inhibit prolonged migraine aura,[59] but little evidence that it has any durable benefit for migraine.[60,61] Magnesium is also an activity–dependent inhibitor of NMDA receptors, and while oral and parenteral magnesium have been used in the treatment of migraine,[62,63] it is not clear whether administration of magnesium actually influences NMDA receptor function. One challenge in considering glutamate receptors as a target for PTH is the timing of treatment. Since an increase in glutamate occurs in the early response to trauma, early treatment may therefore be important. Given its tolerability, early therapy with magnesium supplementation following concussion would be reasonable to consider. Because of its potential effects on cognitive function and mood, topiramate may not be indicated for posttraumatic headache. Memantine, by contrast, is generally well tolerated, and although not an approved treatment for any headache disorder, may be worthy of consideration as an early preventive therapy for posttraumatic headache

ATP and Adenosine

Although not extensively characterized, it is likely that ATP is released in substantial quantities in response to traumatic brain injury, as it is with CSD.[64] ATP is released in response to mechanical stimulation of astrocytes,[65] and can activate microglia in animal models of trauma.[66] ATP can act on purinergic receptors including P2X receptors, which are believed to play a significant role in both pain and inflammation.[67] ATP is metabolized extracellularly to adenosine, which may have a variety of effects depending on the receptor subtypes that are activated.[67] Activation of the A1 subtype of receptors has been reported to have neuroprotective effects in animal models.[68] Conversely, activation of the A2A subtype of receptors may have deleterious effects.[69] Caffeine is a nonselective adenosine receptor antagonist, and is commonly used by migraine patients as an acute therapy, either alone or in combination with an analgesic. There is uncertainty regarding the potential role of caffeine in migraine, in part because of concerns regarding overuse or exacerbation of migraine associated with withdrawal from regular use.[70] The effects of caffeine on post–traumatic headache have not been well characterized but given the circumstantial evidence that adenosine may play a role, further investigation of caffeine as either a therapeutic or exacerbating factor may be warranted. Receptor–subtype adenosine receptor modulators have been developed and are being investigated for other indications. There is a good rationale to consider investigation of these compounds as a therapy in TBI.

Swelling and Intracranial Pressure

Brain swelling and elevated intracranial pressure is a common complication of severe traumatic brain injury. Rare cases of significant brain edema, elevated intracranial pressure, and alteration in brain metabolism have been reported to occur in in the setting of repetitive mild head trauma,[45,71] but there is no evidence that significant elevations in intracranial pressure occur in mild TBI in general. Nonetheless, more subtle localized or diffuse cellular swelling could occur, and could represent a therapeutic target. Based in part upon this hypothesis, one study evaluated the effects of hypertonic saline on headache and other symptoms following concussion in children. This study reported that hypertonic saline infusion after concussion reduced headache both acutely and 3 days later as compared with normal saline infusion, although there was no significant change in other concussion symptoms.[72] Recent studies in animal models have identified a novel strategy for the reduction of intracranial hypertension based on targeting the GLP–1 receptor.[73] These findings raise the possibility that brain swelling and its potential adverse consequences including PTH could be treated acutely and/or chronically with well tolerated oral or parental approaches.


A variety of inflammation–related markers are altered following concussion,[74,75] and some of these may also altered in headache disorders. However, apart from the beneficial effect of nonsteroidal antiinflammatory therapies in the acute treatment of migraine, and the substantial evidence for a role for neuropeptides that may be considered broadly as part of an inflammatory response, the evidence supporting a role for inflammation in migraine is surprisingly weak. Nonetheless, antiinflammatory therapies should be considered as potential approaches for the treatment of PTH. Nonsteroidal antiinflammatory medications are widely prescribed in the setting of PTH, but there is no high–quality evidence that these medications have efficacy as either acute or preventive therapies for PTH. Despite the lack of evidence, given their generally good tolerability a trial of NSAID's as an acute or preventive therapy for PTH is clearly reasonable.


CGRP. There is now substantial evidence that calcitonin gene related peptide plays a primary role in migraine, and existing basic and clinical evidence for this role has been confirmed by the efficacy of small molecule CGRP receptor antagonists or antibodies targeting CGRP or its receptor.[76] CGRP is released in response to head trauma in animal models and has been shown to mediate migraine–related behaviors in this setting including hyperalgesia and light avoidance[77–80] These data from animal models of traumatic brain injury, in combination with the substantial data regarding the role of CGRP in migraine, indicate that CGRP may also play an important role in PTH. Quantification of CGRP levels in the acute posttraumatic period, or in patients with chronic posttraumatic headache could represent a valuable biomarker. Such assays are challenging to perform, however, and require rigorous standardization and control. Regardless of these levels, it will be important to investigate the potential efficacy for PTH of CGRP peptide or receptor antibodies, or small molecule CGRP receptor antagonists, when these medications become available.

PACAP. PACAP release has been demonstrated in severe traumatic brain injury in humans In 38 human patients with severe TBI, PACAP was found to be elevated in both CSF and plasma during the first week after injury.[81] Increased expression of PACAP was also found in glia surrounding traumatic lesions in humans.[82] In the setting of TBI, particularly severe TBI, PACAP has been proposed to have a protective function, and in animal models exogenous administration of PACAP has been reported to reduce damage[83–85] associated with TBI. Elevated blood levels of PACAP have been reported in women with posttraumatic stress disorder,[86] suggesting a possible role in this condition. Elevated levels of PACAP have also been reported during migraine attacks,[87] and it has been hypothesized that PACAP and its receptors (specifically the PAC1 receptor) may represent a therapeutic target in migraine. Based on this hypothesis, monoclonal antibodies to the PAC1 receptor are in early phase clinical trials as a potential therapy for migraine. As with the CGRP antibodies, there is a strong rationale for considering antibodies to the PAC1 receptor or to PACAP itself as potential therapies for post traumatic headache.

Substance P. Substance P is released in response to traumatic brain injury and has been proposed to play a role in edema, increased permeability of the blood brain barrier, and functional consequences of TBI.[88] It is a primary mediator of the phenomenon of neurogenic inflammation via its activation of neurokinin (NK) receptors.[89] Based on the role of substance P in animal models it was investigated as a potential therapeutic target for migraine, with negative results. Multiple NK receptor antagonists have failed in clinical trials investigating their efficacy for migraine.[90] Nonetheless, given the evidence for a role for Substance P in the physiological and functional consequences of TBI, and given that at least 2 NK receptor antagonists (aprepitant and rolapitant) are currently FDA approved for the treatment of nausea,[91] it would be reasonable to consider a trial of this class of medications as a therapy for PTH.

Brain Derived Neurotropic Factor (BDNF)

BDNF is the most abundant neurotrophin in the brain and has been hypothesized to play a significant role in the recovery from traumatic brain injury.[92] In animal models, exercise induced enhancement of cognitive recovery after TBI has been reported to be dependent on BDNF.[93] BDNF may also play a significant role in central sensitization and neuropathic pain.[94] In migraine patients, blood BDNF levels have reported to be decreased interictally,[95,96] but increased during attacks.[97] Two polymorphisms in the BDNF gene have been identified as potentially increasing susceptibility to migraine, including the rs6265[98,99] and rs2049046 polymorphisms.[99] Taken together, these results suggest that BDNF may be a therapeutic target in posttraumatic headache, and including representing a potential mechanism involved in the therapeutic benefit of exercise.

Hypothalamus/Neuroendocrine Mechanisms

Multiple derangements in neuroendocrine function may occur as a consequence of TBI, particularly severe TBI, and this dysfunction may persist for extended periods of time.[100,101] Abnormalities of hypothalamic, pituitary, adrenal, and gonadal function may occur. In migraine, there is growing evidence based on functional imaging studies that alteration in hypothalamic function occurs in the premonitory phase of a migraine attack, supporting a primary role of the hypothalamus in migraine.[102,103] Sex hormones clearly play a significant role in migraine[104] and other aspects of neuroendocrine function may represent important migraine mechanisms.[105] This raises the possibility that neuroendocrine function should be more closely examined as a potential therapeutic target in posttraumatic headache.

Cervical Nerve Roots

Most mechanisms of head injury also exert significant forces on the cervical spine, thus potentially causing alteration in the function of or injury to the cervical nerve roots, cervical spinal cord, and caudal brainstem. While headache is often considered to be transmitted primarily through the trigeminal system, it is clear that the upper cervical nerve roots and rostral cervical spinal cord may play a critically important role in the generation of headache. The anatomy of the upper cervical nerve roots is highly variable from individual to individual. Branches of the C1, C2, and C3 nerve roots form a plexus with variable anastomoses amongst themselves and with other nerves.[106] Also, the anatomy of the C1 nerve root, particularly its sensory component, is highly variable. While the C1 root has traditionally been considered to have no sensory function, stimulation of this root can produce frontal and periorbital pain, particularly in patients with migraine.[107] This raises the intriguing possibility that cervical nerve root anatomy and alteration of its function could play an important role in PTH. If this is the case, then the upper cervical nerve roots could represent a therapeutic target. Injections of anesthetics and steroids over the greater and lesser occipital nerves is used as a therapeutic approach in primary headache disorders, although there is conflicting evidence about the benefit.[108–110] Tenderness over the occipital nerves, and accompanying radiation of pain to the frontal/periorbital region, may be predictors of response to occipital injections.[111] Therefore, examination for occipital nerve tenderness and frontal radiation of pain may be useful in the assessment of PTH, and occipital/suboccipital injections of anesthetics and/or steroids may represent an effective therapeutic approach for some patients. High–resolution imaging studies of the upper cervical nerve roots may provide important new insight into their role in headache disorders, particularly PTH.