Blackwell Science, LtdOxford, UKCHACephalalgia1468-2982Blackwell Science, 2004249753757Original ArticleHypothalamic activation in trigeminal autonomic cephalgiaT Sprenger et al.

doi:10.1111/j.1468-2982.2004.00753.x

Hypothalamic activation in trigeminal autonomic cephalgia:

functional imaging of an atypical case
T Sprenger1, M Valet1, M Hammes1, P Erhard1, A Berthele1, B Conrad1 & TR Tolle1
1
Department of Neurology, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany

Sprenger T, Valet M, Hammes M, Erhard P, Berthele A, Conrad B & Tolle TR.

Hypothalamic activation in trigeminal autonomic cephalgia: functional imaging
of an atypical case. Cephalalgia 2004; 24:753–757. London. ISSN 0333-1024
We report headache induced BOLD changes in an atypical case of trigeminal
autonomic cephalgia (TAC). A 68-year-old patient was imaged using fMRI during
three attacks of a periorbital head-pain with a average duration of 3 min. During
the attacks, left sided conjunctival injection, rhinorrhea, lacrimation, facial sweat-
ing and hypersalivation were apparent. These attacks were usually partly respon-
sive to oxygen administration but otherwise refractory to any drug. The patient
described either attacks with a duration of one minute or less or longer attacks
persisting for maximum of 20 min with headaches occurring up to 100 times a
day. When considering the symptoms, frequency, duration and therapeutic
response of the patient’s headache, no clear-cut classification to one of the sub-
types of trigeminal autonomic cephalgias (cluster headache, paroxysmal hemic-
rania, SUNCT) or trigeminal neuralgia was possible. The cerebral activation
pattern was similar but not identical to those previously observed in cluster
headache and SUNCT with a prominent activation in the hypothalamic grey
matter. This case study underlines the conceptual value of the term TAC for the
group of headaches focusing around the trigeminal-autonomic reflex. Our results
emphasize the importance of the hypothalamus as key region in the pathophysi-
ology of this entity.
Trigeminal autonomic cephalgia, SUNCT, cluster headache,
CPH, trigeminal neuralgia, fMRI
Dr Till Sprenger, Department of Neurology, Technical University Muenchen, Moehlstr.
28, D-81675 Muenchen, Germany. Tel. +49 894140 4628, fax +49 894140 4659, e-mail
sprenger@lrz.tu-muenchen.de Received 16 September 2003, accepted 18 December 2003

terized by the most intense pain, continuing for 15–

Introduction
180 min, with one to eight attacks per day and pre-
Trigeminal autonomic cephalgia is a relatively new dominance in males (4 : 1) (2). In episodic and
term, first proposed by Goadsby and Lipton (1) for chronic paroxysmal hemicrania, headache periods
a group of primary headaches with pain and auto- are shorter (2–45 min), more women are affected,
nomic involvement in the facial area of the trigemi- and this disorder typically shows a good response to
nal nerve. Although the headache syndromes of this indomethacin administration (usually between 100
group, namely cluster headache, paroxysmal hemic- and 150 mg per day). Finally, short lasting unilateral
rania, and SUNCT, clearly share typical clinical fea- neuralgiform headache with conjunctival injection
tures, in most cases a subclassification is mandatory and tearing (SUNCT) is a syndrome with very short
as therapeutical regimen and response differ. There- headache attacks (15–60 s), but a high daily fre-
fore, the intensity, localization, frequency and dura- quency, and is often refractory to any medical and
tion of the pain, autonomic involvement, age and surgical treatment (3).
gender of the patients as well as triggering factors A typical differential diagnosis for SUNCT is
have to be considered. Cluster headache is charac- trigeminal neuralgia, where pain also occurs in the

© Blackwell Publishing Ltd Cephalalgia, 2004, 24, 753–757 753

754 T Sprenger et al.
distribution of the trigeminal nerve, however,
mostly in the area of the second or third branch and
usually lacking autonomic symptoms. The incidence
of daily attacks in trigeminal neuralgia is sometimes
up to 100 per day and therefore even higher than in
SUNCT (1).
Despite unquestionable differences between the
mentioned syndromes, some patients show a mix-
ture of symptoms, more than one clinical presenta-
tion on different occasions (e.g. cluster headache and
trigeminal neuralgia in the same patient) or a shift
of symptoms, respectively, attributed diagnosis with
time.
During the last decade, substantial progress has
been achieved in the understanding of the pathogen-
esis of TAC and other primary headache disorders.
Participation of the trigemino-parasympathetic
reflex, which has been first described in cats (4), in
the pathological process is now well accepted. More-
over, there is growing evidence of the hypothalamic
grey being also involved in the pathogenesis of TAC
as PET activation studies revealed ipsilateral hypo-
thalamic activation in cluster headache (5). The cere-
bral activation detected in the study by May et al. (6)
was not found in experimental pain studies and
therefore seems to be specific to cluster headache.
These results were taken further using voxel based
morphometry, suggesting an increased grey matter
density in the homolateral hypothalamus. Finally,
hypothalamic activation near to the one in cluster
headache was also found in an imaging case study
of a patient suffering from another TAC, namely
SUNCT attacks (7).
We present the clinical characteristics and func-
tional imaging data of an interesting patient with
TAC. In this patients case we felt, that no further
classification was feasible, with the patient showing
hallmarks of cluster headache, SUNCT, but also
trigeminal neuralgia.
Case report and methods
Patient
A 68-year-old caucasian male construction engineer
with a 12 years history of TAC and without history
of migraine was imaged. He presented with recur-
rent strictly left sided stabbing head-pains, being
most severe between November and March, but
occurring throughout the whole year with a fre-
quency of 4–100 attacks per day. The pain maxi-
mum was located retro- and peri-orbitally with
ipsilateral conjunctival injection and lacrimation as
well as ipsilateral nasal rhinorrhoea and hypersali-
vation. During the attacks swelling of the eyelids
and left sided facial sweating was evident. Usually,
the attacks occurred in the second part of the night
or in the early morning. These late night episodes
started mostly spontaneously, but at other daytimes
multiple triggers were also able to provoke the
same pain episodes. Potential triggers were cold
wind on the left eye, light touch to the left facial
side and gaping, but also chewing and speaking.
Beginning and cessation of the headache were
always abrupt. The patient described two similar
pain manifestations, which differed only in their
duration. On the one hand short episodes with a
maximum duration of 1 minute, therefore resem-
bling SUNCT attacks, and on the other hand longer
episodes of up to 20 min being more compatible
with paroxysmal hemicrania or cluster headache.
Whereas predominantly the longer attacks occurred
during the months before and after the fMRI scan-
ning, the patient’s clinical presentation had then
shifted to the shorter attacks thereafter. The
response to medication was frustrating. Oxygen
administration results in a slight to medium relieve
of symptoms, which is usually only seen in cluster
headache. All other medications which were inves-
tigated in the course of the long-lasting illness, such
as indomethacin (up to 150 mg/day per os), corti-
sone, lithium, carbamazepine, amitriptyline, ergota-
mine, metamizol, tramadol and acupuncture were
all without effect on the pain.
The neurological examination was completely
normal and cranial CT and MRI were repeatedly
unremarkable.
During the fMRI scanning, the patient provoked 3
of the above described pain episodes by touching of
a buccal triggerpoint with his tongue. The first with
a length of 4 min, the second lasting 1.5 min and the
third 5 min.
Methods
Functional magnetic resonance imaging was per-
formed on a 1.5 Tesla Philips Gyroscan NT scanner.
The patients head was positioned comfortably inside
a receive-only birdcage head coil, supplied with ear
plugs, heavily padded and secured with a strap
across the forehead in order to minimize head
motion. Echo Planar Imaging was used for the acqui-
sition of the functional data. Acquisition parameters
were TR: 6000 ms; TE: 50 ms; flip angle: 90∞; Matrix:
128 ¥ 128; FOV: 230 ¥ 230 mm; 25 slices (thickness
5 mm). The resulting voxel size was
1.8 ¥ 1.8 ¥ 5.0 mm. 300 functional images were
acquired with 101 images during the pain. The
© Blackwell Publishing Ltd Cephalalgia, 2004, 24, 753–757
 Hypothalamic activation in trigeminal autonomic cephalgia 755

patient indicated the beginning and end of the pain

attacks by pressing buttons on a nonmagnetic
button-box.
Preprocessing and statistical analysis of the fMRI
data was conducted using SPM99 (8). FMRI data
series were realigned and resliced with sinc inter-
polation to correct for motion artifacts. Subse-
quently, the image series were transformed into
standard stereotactic space to allow comparisons
of stereotactic coordinates with published data on
idiopathic headache syndromes. The resampled
voxel volume of the normalized images was
2 ¥ 2 ¥ 2 mm. Data were then smoothed with an
isotropic Gaussian kernel of 4 mm full-width at
half maximum to reduce high frequency noise.
Condition-specific effects (pain vs no pain) were
estimated in SPM with the General Linear Model
using a boxcar approach convolved with the hae-
modynamic response function (8). A statistical
parametric map was generated as t-contrast at a
threshold of P < 0.001. Activation was considered
significant at a threshold of 0.05 corrected for mul-
tiple comparisons.
Results
During the scanning, no head motion greater than
1 mm was detected with and without pain. When
comparing the pain attacks with the pain free state
using fMRI, we detected significant cerebral activa-
tion increases (BOLD signal increases) in the hypo-
thalamic grey matter. This hypothalamic activation
extended laterally from a medial peak (Fig. 1). Addi-
tionally, with a less conservative threshold of
P < 0.001 uncorrected, trends of activation were
observed in the cingulate cortex, insula, temporal
cortex, and frontal cortex (Table 1).
Discussion
Previous neuroimaging studies have shown signifi-
cant activation of multiple brain areas in patients
with cluster headache (5, 9). Although most of these
areas have also been shown to be involved in general
pain processing by studies using experimental pain
stimulation paradigms such as tonic heat (10) or
laser pain (11), ipsilateral posterior hypothalamic

a b

Figure 1 Statistical comparison of an fMRI study of three triggered trigeminal autonomic headache attacks and rest (no pain)
in one patient. The pain induced activations are shown as statistical parametric maps of significant BOLD increases (P < 0.001
uncorrected for descriptive purposes) superimposed on a T1-weighted anatomical reference MR image in (a) coronal, (b) sagittal
and (c) axial planes. The images are displayed in radiological convention.

© Blackwell Publishing Ltd Cephalalgia, 2004, 24, 753–757

756 T Sprenger et al.
Table 1 BOLD-signal increases during an attack of trigeminal autonomic cephalgia compared with the pain-free state, P < 0.001
uncorrected. Coordinates are displayed in radiological convention
Brodmann
Region areas
Ipsilat. posterior hypothalamus
Middle frontal gyrus 9 53
Inferior frontal gyrus 45
Superior frontal gyrus 8 4
Insula
Transverse temporal gyrus 41
Cingulate cortex 23
***Significant after correction for multiple comparisons.
activation was exclusively observed in TAC. Based
on clinical grounds, a role of the hypothalamus had
already before been proposed in the pathology of
cluster headache. Especially the seasonal pattern of
bout periods and the neuroendocrine changes (12)
led to this assumption.
A new and very convincing hint is the clinical
observation of continuous long-term electrostimula-
tion of the ipsilateral infero-posterior hypothalamus
leading to pain reduction or even complete pain ces-
sation in patients with severe chronic cluster head-
ache (13, 14).
In other forms of TAC, not as much evidence for
a central or hypothalamic pathological participation
is available, but at least in SUNCT, one case study
showed activation of a hypothalamic area nearby
(laterally) the one reported in cluster headache (7).
Despite the lack of imaging studies in paroxysmal
hemicrania, similarities in the pathophysiology of
all TAC’s, namely a hypothalamic involvement, are
likely, as they share many clinical features.
Before the emergence of studies pointing towards
a central mechanism in TAC, a mainly vascular
pathogenesis was assumed. Today, a trigemino-
vascular activation in the course of a TAC attack is
still unquestioned, but this parasympathic reflex is
thought to be mediated secondary to a central pain
triggering (9). Therefore, these headaches are now
referred to as neurovascular headaches.
We presented a case with TAC evidencing clinical
features of SUNCT, paroxysmal hemicrania, cluster
headache, but also trigeminal neuralgia. Whereas
the patients age and sometimes short duration of
attacks fit the definition of SUNCT, the frequent
occurrence during the night, the response to oxygen
administration and the sometimes longer attacks
point more towards the diagnosis of cluster head-
aches or paroxysmal hemicrania. The very high fre-
quency, and possibility to trigger attacks by stimuli
Tailarach coordinates
Z score of peak
x y z activation
6 -8 0 5.31***
23 32 3.92
55 18 8 3.89
30 46 3.82
-26 20 6 3.39
40 -27 11 3.36
-4 -18 32 3.32
in the distribution of the trigeminal nerve are best
explained by trigeminal neuralgia. The lack of clini-
cal improvement by indomethacin medication
argues against chronic paroxysmal hemicrania.
Although the patient described two different types
of headaches, in terms of their duration, we think,
that because of the otherwise completely identical
clinical presentation and intensity of the head-pain,
a differential classification of the shorter vs longer
attacks the patient described, to different subtypes
of TAC is not justified. Rather, we think that a patho-
physiologically identical pain may be sometimes
abortive in this patient’s case. Taking into account
these considerations, one possible description for the
patients pain would be as an ‘atypically duration
SUNCT’.
However, our fMRI results show headache
induced activation of the hypothalamic grey matter
in a position almost identical to the one shown in
cluster headache. This medially localized activation
peak extends laterally and overlaps with the one
previously reported in SUNCT (7). Therefore, it may
be speculated that the mixed fMRI activation pattern
reflects the curious clinical presentation our patient
shows, where no clear classification is applicable.
These results suggest that there is sometimes not
only considerable clinical overlap between certain
of the primary cranial pain syndromes (15), but
that these overlaps may actually be substantiated by
imaging techniques. Although this case sheds more
light on the assumed pathophysiology of TAC, it
needs to be emphasized that there is a strong need
to investigate more and clinically clearly defined
TAC syndromes to substantiate considerations of the
pathogenesis of TAC as well as to explain differences
in phenotype.
In conclusion, our new data provide more evi-
dence for a central origin of TAC. The hypothalamus
seems to be a key region not only in cluster
© Blackwell Publishing Ltd Cephalalgia, 2004, 24, 753–757
 Hypothalamic activation in trigeminal autonomic cephalgia
headache, but also in other forms of TAC. Although
a subclassification within the TAC group is desir-
able, as therapeutic responses to medication differ
within this group of primary headaches, it is not
feasible in all patients. Therefore, we strongly sup-
port the integration of the term trigeminal auto-
nomic cephalgia into the now published revision of
the classification system of the International Head-
ache Society (16).
Acknowledgements
We would like to thank Dr Arne May for his very helpful
advice during the preparation of this manuscript. This study
was supported by the ‘Irmgard und Gerhard Schulz Fond’
and the SFB 391 C9.
References
1 Goadsby PJ, Lipton RB. A review of paroxysmal hemicra-
nias, SUNCT, syndrome and other short-lasting headaches
with autonomic features, including new cases. Brain 1997;
120:193–209.
2 Headache Classification Committee of the International
Headache Society. Classification and diagnostic criteria for
headache disorders, cranial neuralgias and facial pain.
Cephalalgia 1988; 8 Suppl 7:1–96.
3 Sjaastad O, Saunte C, Salvesen R, Fredriksen TA, Seim A,
Roe OD, et al. Shortlasting unilateral neuralgiform head-
ache attacks with conjunctival injection, tearing, sweating,
and rhinorrhea. Cephalalgia 1989; 9:147–56.
4 Goadsby PJ, Duckworth JW. Effect of stimulation of
trigeminal ganglion on regional cerebral blood flow in
cats. Am J Physiol 1987; 253:R270–4.
5 May A, Bahra A, Buchel C, Frackowiak RS, Goadsby PJ.
Hypothalamic activation in cluster headache attacks. Lan-
cet 1998; 352:275–8.
6 May A, Ashburner J, Buchel C, McGonigle DJ, Friston KJ,
Frackowiak RS, et al. Correlation between structural and
© Blackwell Publishing Ltd Cephalalgia, 2004, 24, 753–757
757
functional changes in brain in an idiopathic headache syn-
drome. Nat Med 1999; 5:836–8.
7 May A, Bahra A, Buchel C, Turner R, Goadsby PJ. Func-
tional magnetic resonance imaging in spontaneous attacks
of SUNCT. short-lasting neuralgiform headache with con-
junctival injection and tearing. Ann Neurol 1999; 46:791–4.
8 Friston KJ, Holmes AP, Worsley KJ, Poline JP, Frith CD,
Frackowiak RSJ. Statistical Parametric Maps in functional
imaging: a general linear approach. Human Brain Map-
ping 1995; 2:189–210.
9 May A, Bahra A, Buchel C, Frackowiak RS, Goadsby PJ.
PET and MRA findings in cluster headache and MRA in
experimental pain. Neurology 2000; 55:1328–35.
10 Tolle TR, Kaufmann T, Siessmeier T, Lautenbacher S, Ber-
thele A, Munz F, et al. Region-specific encoding of sensory
and affective components of pain in the human brain: a
positron emission tomography correlation analysis. Ann
Neurol 1999; 45:40–7.
11 Bornhovd K, Quante M, Glauche V, Bromm B, Weiller C,
Buchel C. Painful stimuli evoke different stimulus-
response functions in the amygdala, prefrontal, insula and
somatosensory cortex: a single-trial fMRI study. Brain
2002; 125:1326–36.
12 Pringsheim T. Cluster headache: Evidence for a disorder
of circadian rhythm and hypothalamic function. Can J
Neurol Sci 2002; 29:33–40.
13 Leone M, Franzini A, Bussone G. Stereotactic stimulation
of posterior hypothalamic gray matter in a patient with
intractable cluster headache. N Engl J Med 2001; 345:1428–
9.
14 Franzini A, Ferroli P, Leone M, Broggi G. Stimulation of
the posterior hypothalamus for treatment of chronic
intractable cluster headaches: first reported series. Neuro-
surgery 2003; 52:1095–101.
15 Peatfield R, Bahra A, Goadsby PJ. Trigeminal-autonomic
cephalgias (TACs). Cephalalgia 1998; 18:358.
16 Headache Classification Committee of the International
Headache Society. The International Classification of
Headache Disorders, 2nd edn. Cephalalgia 2004; 24 Suppl
1:1–160.