- Email: [email protected]
Intracranial and extracranial arteries in migraine
Correspondence
Intracranial and extracranial arteries in migraine It is a pity when valuable research such as that described by Faisal Mohammad Amin and colleagues1 is incorrectly interpreted and used to justify misleading conclusions. The investigators—who aimed to measure extracranial and intracranial arteries during attacks of migraine without aura—stated that migraine pain was not accompanied by extracranial dilatation. At first glance this assumption seems to be accurate, but on closer inspection it is misleading. The authors measured the proximal portion of the superficial temporal artery, in the retromandibular area, even though the retromandibular area is not the site of migraine pain. This part of the superficial temporal artery has not been implicated in migraine, so what purpose is served by showing that it does not dilate? On the contrary, the extracranial arteries that have repeatedly been shown to be affected during migraine are the superficial scalp arteries.2 One of the authors was previously an investigator in a study by Iversen and colleagues, 3 who showed changes in the frontal branch of the superficial temporal artery during migraine. In that study the investigators stated the following: “With the ultrasound techniques that we used diameters can be measured accurately and reproducibly. The patients were investigated during unilateral migraine attacks. The luminal diameter of the frontal branch of the superficial temporal artery was significantly larger on the symptomatic than on the nonsymptomatic side during but not between attacks”.3 Furthermore, in another article,4 the authors of which included two investigators from The Lancet Neurology study,1 the aim was to show that sodium nitroprusside, a powerful vasodilatory agent,
would induce both headache and vasodilatation. When it was important to show a correlation between headache and vasodilatation, they did not measure the proximal portion of the superficial temporal artery, but chose instead to measure the change in diameter of its frontal branch, which indeed dilated substantially (33%) with sodium nitroprusside.4 In view of the fact that at least one of the investigators in Amin and colleagues’ study1 has shown that the frontal branch of the superficial temporal artery does dilate during migraine, why did they not include this vessel in their study? How can they justify their conclusion that the extracranial vessels do not dilate during migraine? It would go a long way towards clarifying the situation if the authors were to publish their measurements of the frontal branch. One of the treatment modalities that I use is surgical cauterisation of the terminal branches of the external carotid artery (including the superficial temporal artery) in patients with migraine.
Elliot Shevel [email protected] The Headache Clinic, Johannesburg, South Africa 1
2
3
4
Amin FM, Asghar MS, Hougaard A, et al. Magnetic resonance angiography of intracranial and extracranial arteries in patients with spontaneous migraine without aura: a crosssectional study. Lancet Neurol 2013; 12: 454–61. Shevel E. The extracranial vascular theory of migraine—a great story confirmed by the facts. Headache 2011; 51: 409–17. Iversen HK, Nielsen TH, Olesen J, Tfelt-Hansen P. Arterial responses during migraine headache. Lancet 1990; 336: 837–39. Guo S, Ashina M, Olesen J, Birk S. The effect of sodium nitroprusside on cerebral hemodynamics and headache in healthy subjects. Cephalalgia 2013; 33: 301–7.
After reading the interesting Article by Faisal Mohammad Amin and colleagues,1 we were confused about the conclusion made by the authors, namely that ”future migraine research should focus on the peripheral and central pain pathways rather than simple arterial dilatation”. This conclusion was based on their observation
www.thelancet.com/neurology Vol 12 September 2013
that during a migraine attack, “no statistically significant dilatation was seen of the extracranial arteries on the pain side” and “only slight dilatation” of cerebral arteries was detected. Although the authors correctly comment that they could not exclude possible dilatation of dural branches of the middle meningeal artery (MMA) given that these small branches could not be analysed due to technical limitations, they still felt justified to draw the abovementioned conclusion. We perceive this limitation as an important issue and fear that their interpretation is premature. Amin and colleagues disregard the fact that an intracranial branch of an artery, such as the intracranial dural branches of the MMA, can differ from its extracranial branches.2,3 In particular, they did not measure intracranial MMA circumference in any of their human magnetic resonance angiography studies, but instead based their findings on assessments of only the extracranial MMA. By 1961, Hassler3 showed structural changes in the basilar artery wall between the open and closed skull after craniotomy. In the open skull preparation, the basilar artery had profoundly changed its vessel wall properties; hence we argue that the structure (and consequently also function) of the extracranial MMA might differ substantially from the intracranial MMA. We are further surprised by a previous report by the same group, in which they concluded that “migraine without aura is associated with dilatation of extra- and intracerebral arteries and that the headache location is associated with the location of the vasodilatation”.4 Admittedly, the migraine attacks seen in this study4 were induced by systemic infusion of calcitonin generelated peptide (CGRP). Their finding that spontaneous migraine attacks are not associated with dilation of 847
Correspondence
extracranial arteries suggests that provocation with CGRP infusion is a flawed migraine model. Alternatively, this discrepancy might be related to to some concerns regarding the study population and the procedures applied in the study by Amin and colleagues.1 The migraine attacks reported were of moderate severity and time between attack and attack-free scans was 5–274 days, as opposed to only a few hours in the CGRP infusion study.4 The arterial circumference changes measured between attack and attack-free scans were slight and had large SD. These limitations could have resulted in an underpowered study measuring minor changes from variable baseline values owing to the long period between scans. We acknowledge the sheer complexity of doing magnetic resonance angiography scans during spontaneous migraine attacks and value the study by Amin and colleagues.1 However, we contest their, in our view, premature conclusion. We sincerely approve of findings that help our understanding of migraine and migraine pain, but such findings should be based on solid arguments and data. We declare that we have no conflicts of interest
*Antoinette MaassenVanDenBrink, Khatera Ibrahimi, Lars Edvinsson a.vanharen-maassenvandenbrink@ erasmusmc.nl Erasmus Medical Center, Rotterdam, Netherlands 1
2
3
4
848
Amin FM, Asghar MS, Hougaard A, et al. Magnetic resonance angiography of intracranial and extracranial arteries in patients with spontaneous migraine without aura: a cross-sectional study. Lancet Neurol 2013; 12: 454–61. MaassenVanDenBrink A, Duncker DJ, Saxena PR. Migraine headache is not associated with cerebral or meningeal vasodilatation—a 3T magnetic resonance angiography study. Brain 2009; 132: e112. Hassler O. Morphological studies on the large cerebral arteries, with reference to the aetiology of subarachnoid haemorrhage. Acta Psychiatr Scand Suppl 1961; 154: 1–145. Asghar MS, Hansen AE, Amin FM, et al. Evidence for a vascular factor in migraine. Ann Neurol 2011; 69: 635–45.
Authors’ reply We thank Elliot Shevel and Antoinette MaassenVanDenBrink and colleagues for their interest in our study of spontaneous migraine using magnetic resonance angiography,1 and for their thoughtful comments. Findings from Wolff and colleagues’ pioneering studies suggested an association between migraine and the branches of external carotid artery.2 Findings from experimental headache studies by Ray and Wolff3 showed that throbbing pain associated with nausea and sickness could result from experimental distension of the main trunk of the superficial temporal artery (STA). During the 1960s and 1970s, the frontal branch of the STA was commonly measured, because clinicians could easily identify and measure it under the skin using old techniques. In the late 1980s, clinicians started using ultrasonography to measure the frontal branch of the STA. This more superficial branch is measured in these studies because it is easier to identify than the main trunk using ultrasonography. The frontal branch is not visible on the magnetic resonance images we obtained in our study1 but we do not consider this as a limitation because, to the best of our knowledge, there is no evidence for STA dilatation during spontaneous attacks. In his letter, Shevel refers to a study done by Iversen and colleagues,4 who used ultrasonography to directly measure the luminal diameter of the frontal branch of the STA. In that study, the investigators reported a constriction of the frontal STA on the non-headache side during attacks, compared with outside of attacks. However, more importantly, the frontal STA on the headache side was unchanged during attacks (during attack 1·10 mm; outside of attack 1·14 mm; p=0·75).4 The study by Guo and colleagues5 was of the physiological and headache-inducing effects of intravenous infusion of
nitroprusside in five people with no history of migraine. We consider that study irrelevant to the discussion of the mechanisms of spontaneous migraine pain. MaassenVanDenBrink and colleagues suggest that the extracranial and intracranial (dural) middle meningeal artery (MMA) might react differently and that a long period between the attack and the attack-free scans might result in a higher variance of arterial circumference compared with our previous study.4 At present, there is no evidence for differences in vasoreactivity between the human intracranial and extracranial MMA. The references provided by the authors do not report any studies on differences of the intracranial and extracranial portions of the MMA in man. The potential effects of a long period between attack and attackfree scans is worth considering. More importantly, the side-to-side comparison during and outside attack, which is a more powerful approach, showed no difference between the pain and non-pain sides, confirming the lack of extracranial dilatation. In light of the comments from MaassenVanDenBrink and colleagues we have excluded patients who had more than 90 days between the attack and attack-free scans (n=6); MMA circumference change in the remaining 12 patients was 1·9% (95% CI –3·3% to 7·2%; p=0·46) on the pain side and –1·1% (–4·8% to 2·5%; p=0·54) on the non-pain side, confirming the finding of no difference reported in our original paper.1 In addition, the SD values were almost the same in our previous4 and more recent1 studies. If the imaging resolution was too small to detect differences between the attack and attack-free scans, it should be more difficult to detect a decrease of vessel size than an increase, but we were able to detect decrease of the MMA circumference after administration of sumatriptan.1
www.thelancet.com/neurology Vol 12 September 2013
Intracranial and extracranial arteries in migraine It is a pity when valuable research such as that described by Faisal Mohammad Amin and colleagues1 is incorrectly interpreted and used to justify misleading conclusions. The investigators—who aimed to measure extracranial and intracranial arteries during attacks of migraine without aura—stated that migraine pain was not accompanied by extracranial dilatation. At first glance this assumption seems to be accurate, but on closer inspection it is misleading. The authors measured the proximal portion of the superficial temporal artery, in the retromandibular area, even though the retromandibular area is not the site of migraine pain. This part of the superficial temporal artery has not been implicated in migraine, so what purpose is served by showing that it does not dilate? On the contrary, the extracranial arteries that have repeatedly been shown to be affected during migraine are the superficial scalp arteries.2 One of the authors was previously an investigator in a study by Iversen and colleagues, 3 who showed changes in the frontal branch of the superficial temporal artery during migraine. In that study the investigators stated the following: “With the ultrasound techniques that we used diameters can be measured accurately and reproducibly. The patients were investigated during unilateral migraine attacks. The luminal diameter of the frontal branch of the superficial temporal artery was significantly larger on the symptomatic than on the nonsymptomatic side during but not between attacks”.3 Furthermore, in another article,4 the authors of which included two investigators from The Lancet Neurology study,1 the aim was to show that sodium nitroprusside, a powerful vasodilatory agent,
would induce both headache and vasodilatation. When it was important to show a correlation between headache and vasodilatation, they did not measure the proximal portion of the superficial temporal artery, but chose instead to measure the change in diameter of its frontal branch, which indeed dilated substantially (33%) with sodium nitroprusside.4 In view of the fact that at least one of the investigators in Amin and colleagues’ study1 has shown that the frontal branch of the superficial temporal artery does dilate during migraine, why did they not include this vessel in their study? How can they justify their conclusion that the extracranial vessels do not dilate during migraine? It would go a long way towards clarifying the situation if the authors were to publish their measurements of the frontal branch. One of the treatment modalities that I use is surgical cauterisation of the terminal branches of the external carotid artery (including the superficial temporal artery) in patients with migraine.
Elliot Shevel [email protected] The Headache Clinic, Johannesburg, South Africa 1
2
3
4
Amin FM, Asghar MS, Hougaard A, et al. Magnetic resonance angiography of intracranial and extracranial arteries in patients with spontaneous migraine without aura: a crosssectional study. Lancet Neurol 2013; 12: 454–61. Shevel E. The extracranial vascular theory of migraine—a great story confirmed by the facts. Headache 2011; 51: 409–17. Iversen HK, Nielsen TH, Olesen J, Tfelt-Hansen P. Arterial responses during migraine headache. Lancet 1990; 336: 837–39. Guo S, Ashina M, Olesen J, Birk S. The effect of sodium nitroprusside on cerebral hemodynamics and headache in healthy subjects. Cephalalgia 2013; 33: 301–7.
After reading the interesting Article by Faisal Mohammad Amin and colleagues,1 we were confused about the conclusion made by the authors, namely that ”future migraine research should focus on the peripheral and central pain pathways rather than simple arterial dilatation”. This conclusion was based on their observation
www.thelancet.com/neurology Vol 12 September 2013
that during a migraine attack, “no statistically significant dilatation was seen of the extracranial arteries on the pain side” and “only slight dilatation” of cerebral arteries was detected. Although the authors correctly comment that they could not exclude possible dilatation of dural branches of the middle meningeal artery (MMA) given that these small branches could not be analysed due to technical limitations, they still felt justified to draw the abovementioned conclusion. We perceive this limitation as an important issue and fear that their interpretation is premature. Amin and colleagues disregard the fact that an intracranial branch of an artery, such as the intracranial dural branches of the MMA, can differ from its extracranial branches.2,3 In particular, they did not measure intracranial MMA circumference in any of their human magnetic resonance angiography studies, but instead based their findings on assessments of only the extracranial MMA. By 1961, Hassler3 showed structural changes in the basilar artery wall between the open and closed skull after craniotomy. In the open skull preparation, the basilar artery had profoundly changed its vessel wall properties; hence we argue that the structure (and consequently also function) of the extracranial MMA might differ substantially from the intracranial MMA. We are further surprised by a previous report by the same group, in which they concluded that “migraine without aura is associated with dilatation of extra- and intracerebral arteries and that the headache location is associated with the location of the vasodilatation”.4 Admittedly, the migraine attacks seen in this study4 were induced by systemic infusion of calcitonin generelated peptide (CGRP). Their finding that spontaneous migraine attacks are not associated with dilation of 847
Correspondence
extracranial arteries suggests that provocation with CGRP infusion is a flawed migraine model. Alternatively, this discrepancy might be related to to some concerns regarding the study population and the procedures applied in the study by Amin and colleagues.1 The migraine attacks reported were of moderate severity and time between attack and attack-free scans was 5–274 days, as opposed to only a few hours in the CGRP infusion study.4 The arterial circumference changes measured between attack and attack-free scans were slight and had large SD. These limitations could have resulted in an underpowered study measuring minor changes from variable baseline values owing to the long period between scans. We acknowledge the sheer complexity of doing magnetic resonance angiography scans during spontaneous migraine attacks and value the study by Amin and colleagues.1 However, we contest their, in our view, premature conclusion. We sincerely approve of findings that help our understanding of migraine and migraine pain, but such findings should be based on solid arguments and data. We declare that we have no conflicts of interest
*Antoinette MaassenVanDenBrink, Khatera Ibrahimi, Lars Edvinsson a.vanharen-maassenvandenbrink@ erasmusmc.nl Erasmus Medical Center, Rotterdam, Netherlands 1
2
3
4
848
Amin FM, Asghar MS, Hougaard A, et al. Magnetic resonance angiography of intracranial and extracranial arteries in patients with spontaneous migraine without aura: a cross-sectional study. Lancet Neurol 2013; 12: 454–61. MaassenVanDenBrink A, Duncker DJ, Saxena PR. Migraine headache is not associated with cerebral or meningeal vasodilatation—a 3T magnetic resonance angiography study. Brain 2009; 132: e112. Hassler O. Morphological studies on the large cerebral arteries, with reference to the aetiology of subarachnoid haemorrhage. Acta Psychiatr Scand Suppl 1961; 154: 1–145. Asghar MS, Hansen AE, Amin FM, et al. Evidence for a vascular factor in migraine. Ann Neurol 2011; 69: 635–45.
Authors’ reply We thank Elliot Shevel and Antoinette MaassenVanDenBrink and colleagues for their interest in our study of spontaneous migraine using magnetic resonance angiography,1 and for their thoughtful comments. Findings from Wolff and colleagues’ pioneering studies suggested an association between migraine and the branches of external carotid artery.2 Findings from experimental headache studies by Ray and Wolff3 showed that throbbing pain associated with nausea and sickness could result from experimental distension of the main trunk of the superficial temporal artery (STA). During the 1960s and 1970s, the frontal branch of the STA was commonly measured, because clinicians could easily identify and measure it under the skin using old techniques. In the late 1980s, clinicians started using ultrasonography to measure the frontal branch of the STA. This more superficial branch is measured in these studies because it is easier to identify than the main trunk using ultrasonography. The frontal branch is not visible on the magnetic resonance images we obtained in our study1 but we do not consider this as a limitation because, to the best of our knowledge, there is no evidence for STA dilatation during spontaneous attacks. In his letter, Shevel refers to a study done by Iversen and colleagues,4 who used ultrasonography to directly measure the luminal diameter of the frontal branch of the STA. In that study, the investigators reported a constriction of the frontal STA on the non-headache side during attacks, compared with outside of attacks. However, more importantly, the frontal STA on the headache side was unchanged during attacks (during attack 1·10 mm; outside of attack 1·14 mm; p=0·75).4 The study by Guo and colleagues5 was of the physiological and headache-inducing effects of intravenous infusion of
nitroprusside in five people with no history of migraine. We consider that study irrelevant to the discussion of the mechanisms of spontaneous migraine pain. MaassenVanDenBrink and colleagues suggest that the extracranial and intracranial (dural) middle meningeal artery (MMA) might react differently and that a long period between the attack and the attack-free scans might result in a higher variance of arterial circumference compared with our previous study.4 At present, there is no evidence for differences in vasoreactivity between the human intracranial and extracranial MMA. The references provided by the authors do not report any studies on differences of the intracranial and extracranial portions of the MMA in man. The potential effects of a long period between attack and attackfree scans is worth considering. More importantly, the side-to-side comparison during and outside attack, which is a more powerful approach, showed no difference between the pain and non-pain sides, confirming the lack of extracranial dilatation. In light of the comments from MaassenVanDenBrink and colleagues we have excluded patients who had more than 90 days between the attack and attack-free scans (n=6); MMA circumference change in the remaining 12 patients was 1·9% (95% CI –3·3% to 7·2%; p=0·46) on the pain side and –1·1% (–4·8% to 2·5%; p=0·54) on the non-pain side, confirming the finding of no difference reported in our original paper.1 In addition, the SD values were almost the same in our previous4 and more recent1 studies. If the imaging resolution was too small to detect differences between the attack and attack-free scans, it should be more difficult to detect a decrease of vessel size than an increase, but we were able to detect decrease of the MMA circumference after administration of sumatriptan.1
www.thelancet.com/neurology Vol 12 September 2013