Results

Table 1 shows the characteristics of the studied attacks.

During Queckenstedt’s manoeuvre eight out of 20 patients reported unchanged headache intensity compared with baseline rating, five out of 20 reported a worsening, and seven out of 20 an improvement of their headache. Ten out of 20 patients had unchanged intensity during placebo, two reported a worsening, and eight patients reported an improvement of headache intensity compared with baseline. Seven patients reported a lower headache intensity during the placebo procedure compared to that during Queckenstedt’s manoeuvre, three patients a lower intensity during Queckenstedt’s manoeuvre compared with placebo, whereas 10 patients rated the intensity the same during both procedures.

No significant difference in headache intensity was found between the manoeuvre of Queckenstedt and the placebo manoeuvre (p=0.22, marginal homogeneity test) (table 2).

Discussion

The aim of the present study was to evaluate the effect of venous dilatation on migraine pain. Queckenstedt’s manoeuvre increases the pressure in the extracranial and intracranial veins and sinuses and increases the intracranial pressure, the latter being the reason for the long known clinical value of the test. We chose Queckenstedt’s manoeuvre rather than Valsalva’s manoeuvre, to achieve a more constant and equal venous dilatation in all patients, without simultaneous cardiovascular changes. Valsalva’s manoeuvre causes marked changes in heart rate and blood pressure and also causes straining of respiratory and possibly pericranial muscles. Pure breath holding would not cause a sufficient dilatation. The pressure applied on to the upper portion of the sternomastoid muscles (placebo) did not affect the veins due to the underlying transverse processes of the cervical vertebrae.

An effort was made to objectively describe whether a sufficient applied pressure was performed during Queckenstedt’s manoeuvre. In colleagues without headache, reddening of the facial skin and protruding facial vessels were evident and a pounding feeling in the head related to heartbeat was reported. However, in patients with migraine, these changes were inconsistent, while some already felt a throbbing pain in the head and were flushing. Bentsen et alreported a small within observer and a larger between observer variation in palpation of muscles of the head and neck.9To reduce variability we therefore used the same investigator throughout the study. The intracranial pressure or venous pressure was not measured and it has been reported that the intracranial pressure may vary between patients during Queckenstedt’s manoeuvre.10

Although the characteristics of the studied migraine attacks were similar in characteristics to attacks found in a population based study, there were small differences.2 Thus the frequency of photophobia, nausea, and worsening by physical activity were almost identical, whereas vomiting was less frequent in the present study. The intensity of pain was lower and less patients had throbbing headache compared with the population based, epidemiological study. The explanation for the differences is probably that clinical characterisation of the migraine attacks in the present study described only a short moment of a single attack, and that the patients were lying down in a quiet, dark laboratory.

The results show no significant difference between Queckenstedt’s manoeuvre and placebo. Thus the obtained venous dilatation did not aggravate migraine pain.

Neurogenic inflammation can be elicited in animals by electrical stimulation of sensory nerves and ganglia. This causes plasma extravasation and vasodilatation, and this has been shown in rat dura mater.6 This phenomenon can be blocked by sumatriptan and ergotamine.11 Neurogenic inflammation is currently the most widely used and accepted experimental animal model of migraine. In this model, the most conspicuous changes occur around the venules, which show morphological changes; The surface of the endothelium shows irregular elevations and there are numerous endothelial vesicles and cytoplasmic microvillous projections, the second containing pinocytic vesicles—all of which reflect an increased transendothelial transport. It has been suggested that liberation of potent vasodilators from surrounding perivascular nerves, such as substance P, may initiate the described permeability changes seen in neurogenic inflammation. The perivascular sensory nerves surrounding the intracranial veins are mainly of trigeminal origin and seem to react to vascular distention with increased firing. Thus Kaube et alfound that mechanical distention of the superior sagittal sinus (in cats) caused activation of the trigeminovascular system.12

We found no effect of the increase in venous pressure on headache intensity after the Queckenstedt’s manoeuvre. Our study, therefore, does not indicate an important role for the cranial venous system in migraine pain. As the venules are predominantly involved in neurogenic inflammation our results also suggest that neurogenic inflammation is not present or is not of major importance during migraine attacks. Recently, other clinical studies contradicting the neurogenic inflammation model have become available. A direct study of gadolinium enhancement around dural and brain blood vessels using MRI during attacks of migraine without aura, showed no evidence of extravasation.13 In addition, the endothelin receptor antagonist bosentan and the substance P antagonist RPR100893–201 both block neurogenic inflammation, but have shown no effect in aborting migraine attacks in humans.14 15

In conclusion, our study argues against a major involvement of the cranial venous system in the pain mechanisms of migraine and does not support the importance and presence of neurogenic inflammation in migraine.