In a study of experimental head pain in which capsaicin was injected into the forehead of volunteers without headache (May et al., 1998b), there was a bilateral activation pattern in midline structures over several planes, slightly lateralized to the left, anterior to the brain stem and posterior to the chiasmatic region. Superimposed on a magnetic resonance imaging template, the location of the activation covers intracranial arteries as well as the region of the cavernous sinus bilaterally but more marked on the ipsilateral side.
Similarly, strong activation was observed in the same region in the cluster headache study (the cavernous sinus) (May et al., 1998a). This change might be interpreted as increased venous inflow from the superior ophthalmic vein, which drains the ophthalmic artery (Waldenlind et al., 1993), or a longer transit time for the tracer in this region, which could be due to impeded venous drainage.
Another, more exciting possibility is that the observed increase in activation is due to bilateral dilation of the internal carotid artery, since such a change would suggest vasodilation mediated by the ophthalmic division of the trigeminovascular system, a neurally driven vasodilation. It is difficult to assess the contribution of these two sources to the activity measured using PET.
To further address this question, magnetic resonance angiography was performed using the same design as the PET study (May et al., 1998a). Using an image calculation tool, the angiographies were subtracted from each other, leaving only structural changes between conditions. It was demonstrated that in the condition nitroglycerine (NTG) inhalation without headache there was dilation of the basilar artery and the internal carotid arteries bilaterally compared to the rest state (May et al., 1999c).
These vessels stayed dilated during the third condition (cluster headache attack), some 20 to 30 minutes later. Given that we have observed vasodilation in large vessels in cluster headache and increased signal in the region of the cavernous sinus after capsaicin injection to the forehead in a PET study (May et al., 1998b), it seems clear that the vascular changes are an epiphenomenon of activation of the trigeminovascular system (Goadsby and Duckworth, 1987) and that cluster headache is not a disorder of the carotid vasculature or cavernous sinus. We have recently seen the very same change in this region in acute migraine studied with PET (Bahra et al., unpublished data).
The data suggest that activation of the trigeminal system, as such, triggers the impeded arterial or venous drainage or increases flow in the region of these vessels. At the physiological level, the common link is the involvement of the ophthalmic division of the trigeminal nerve by a neurally driven dilation of the carotid vessels; on the basis of the data, these headaches should be regarded as neurovascular since vessel change occurs generically with first division of trigeminal pain.
While most people probably think that they understand tension-type headache (TTH), it seems to the author that this is one of the greatest obstacles to its study. One is generally less questioning about what is true than about what is not thought to be true; TTH implies a process underlying the syndrome, but is this accurate?
The Clinical Problem
Perhaps one of the main stumbling blocks to the study of TTH is the clinical definition. The definition is as much about what TTH is not as what it is. It should not be pulsating, should not be severe, should not be unilateral, should not be aggravated by physical activity, and should have no nausea. It may have photophobia or phonophobia, but not both. Yet are any of these so-called features distinguishing, when they may be seen in migraine. Moreover, chronic TTH offers the possibility of nausea or photophobia or phonophobia but is otherwise phenotypically the same. What is the rationale to allow nausea in chronic TTH but not episodic? What is the possible reason to allow either photophobia or phonophobia in TTH but not both?
There are no answers, and there can be no answers until the study of TTH parcels out all of the features prospectively. The phenotype TTH can probably be produced by a number of biological changes. The fundamental clinical difference between TTH and migraine is that TTH lacks features of sensory sensitivity of any description, so absolutely no nausea and no sensitivity to light, sound, or movement should be permitted; also, it lacks the usual triggering associations of migraine, such as aggravation by menses, skipping meals, or changing sleep patterns. One must ask, how many patients who have contributed to the TTH literature have migrainous biology? This issue must be resolved if we are to understand TTH.
Given our current knowledge of TTH, it is useful to talk about possible mechanisms rather than identifying one primary factor as causative. It is worth setting out what has been noted as a basis for thinking about the disorder. What follows uses the IHS system, and readers should evaluate the data in light of the clinical remarks above.
Epidemiology and Genetics
The 1-year prevalence of TTH is somewhere between 30% and 80%, most likely on the higher end. In a very thorough general population-based study in Denmark, Rasmussen and colleagues (1991) found a 1-year prevalence of TTH of 69% for men and 88% for women, giving a slight predominance in women. In a telephone-based population sample in the United States, TTH was less common, running at 38% of the population (Schwartz et al., 1998). Chronic TTH is seen in about 2% of the population (Castillo et al., 1999; Scher et al., 1998).
In a general population, most patients with TTH do not consult their physicians (Rasmussen et al., 1992). One study suggests that there are some genetic factors at play in chronic TTH (Ostergaard et al., 1997). Stress is just as often identified as a trigger for migraine as TTH (Scharff et al., 1995) and as such has no diagnostic utility when patients report the symptom.
The dull, bilateral, generalized nature of typical TTH provides the impetus to study peripheral sources for the pain. Indeed, TTH has been called “muscle contraction headache.” Injection of saline into cranial muscles certainly induces pain (Kellgren, 1938), and this usually takes 10 to 15 seconds to build up (Jensen and Norup, 1992). Pain from muscle injection is not necessarily restricted to the muscle injected, so injections into the temporalis muscle can cause pain in neck muscles (Simmons et al., 1943). Alogenic substances, such as bradykinin and serotonin, have been injected into human cranial muscles, with minimal pain from the bradykinin and no pain from the serotonin, although together they are more painful than isotonic saline (Jensen et al., 1990). Direct electrical stimulation of muscle causes cramp-like local pain (Laursen et al., 1997), while a combination of chewing and temporal artery ischemia will produce a dull bifrontal headache (Gobel and Cordes, 1990).
Consistent with these observations, pericranial muscles in patients with TTH are more tender than in healthy controls when studied blinded, although this was found mainly for patients with frequent headache of greater than 25 days a month (Langemark and Olesen, 1987). Later studies have shown tenderness in both episodic and chronic TTH (Jensen et al., 1993, 1998).
Pressure pain detection thresholds have been used to study whether this cranial pain is due to some local factor or associated with a more generalized disorder. Pressure pain detection thresholds are normal in episodic TTH (Jensen et al., 1993), whereas they are abnormal in chronic TTH (Bendtsen et al., 1996; Schoenen et al., 1991a). Moreover, patients with chronic TTH are more sensitive to other stimuli, such as thermal stimuli (Langemark et al., 1989). Muscle hardness, which has recently begun to be measured with an ingenious noninvasive technique (Sakai et al., 1995), requires further study. Whether changes represent local pathology or reflexly induced change needs consideration.
Given the very interesting role of nitric oxide generation in spinal cord nociception (Meller and Gebhart, 1993), the recent result of a double-blind study that showed reduction in pain intensity in chronic TTH when the nitric oxide synthase inhibitor L-NG-methylarginine hydrochloride was given (Ashina et al., 1999) suggests a central mechanism for some part of this tenderness.
Because muscle contraction has been thought to play a key role in TTH and is indeed enshrined in the current IHS classification, electromyography (EMG) has been of interest in the condition. A review of the data to the mid-1980s (Pikoff, 1984) concluded that in about half of the studies the muscles were normal and in the other half muscular activity was increased. In chronic TTH, EMG activity tends to be higher in patients than controls, but this is unrelated to headache severity or pressure pain thresholds (Schoenen et al., 1991b). Similarly, there is no increase in EMG activity during a headache compared with EMG activity in the absence of a headache (Clark et al., 1995; Jensen, 1995).
Neurophysiological methods have allowed the study of brain stem pathways that have processing functions in headache. The most extensively studied reflex has been that involving exteroceptive suppression produced as a suppression of voluntary masseter and temporalis contraction with electrical stimulation of trigeminal nerve fibers. Exteroceptive suppression is divided into two periods, the early period ES1, mediated by an oligosynaptic pathway, and the late period ES2, mediated by a polysynaptic pathway (Desmedt and Godaux, 1976).
The interneurons responsible for ES2 are likely to be part of the bulbar reticular formation (Hopf, 1994) and receive modulatory projections from limbic structures and the periaqueductal gray matter (Schoenen, 1993b). The ES2 duration is reduced in chronic TTH (Schoenen et al., 1987) but normal in episodic TTH and migraine (Schoenen, 1993a). The conventional blink reflex is normal in TTH (Sand and Zwart, 1994), although the modified nociceptive-specific reflex has not been studied (Kaube et al., 2000). Contingent negative variation, an event-related potential that is recorded over the frontal cortex, is normal in TTH (Schoenen and Timsit-Berthier, 1993).
Given the changes described and the biochemical changes that revolve around central opiate activity (Langemark et al., 1995), its seems that the bulk of the pathophysiological processes in TTH fall within the central nervous system. A vicious cycle of peripheral activation with sensitization and further peripheral activation may underlie some of the muscle observations. Until the clinical material is clarified more closely, our understanding of TTH will progress very slowly.
Peter J Goadsby
Editors: Silberstein, Stephen D.; Lipton, Richard B.; Dalessio, Donald J.