SUNCT/SUNA and neurovascular
compression: New cases and critical
Valentina Favoni, Daniela Grimaldi, Giulia Pierangeli,
Pietro Cortelli and Sabina Cevoli
Background: Short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing (SUNCT) and
short-lasting unilateral neuralgiform headache with cranial autonomic symptoms (SUNA) are primary headache syn-
dromes. A growing body of literature has focused on brain magnetic resonance imaging (MRI) evidence of neurovascular
compression in these syndromes.
Objective: The objective of this article is to assess whether SUNCT is a subset of SUNA or whether the two are separate
syndromes and clarify the role of neurovascular compression.
Method: We describe three new SUNCT cases with MRI findings of neurovascular compression and critically review
published SUNCT/SUNA cases.
Results: We identified 222 published SUNCT/SUNA cases. Our three patients with neurovascular compression added to
the 34 cases previously described (16.9%). SUNCTand SUNA share the same clinical features and therapeutic options.
At present, there is no available abortive treatment for attacks. Lamotrigine was effective in 64% of patients; topiramate
and gabapentin in about one-third of cases. Of the 34 cases with neurovascular compression, seven responded to drug
therapies, 16 patients underwent microvascular decompression of the trigeminal nerve (MVD) with effectiveness in 75%.
Conclusions: We suggest that SUNCTand SUNA should be considered clinical phenotypes of the same syndrome. Brain
MRI should always be performed with a dedicated view to exclude neurovascular compression. The high percentage of
remission after MVD supports the pathogenetic role of neurovascular compression.
SUNCT, SUNA, autonomic, neurovascular compression, microvascular decompression
Date received: 25 February 2013; revised: 13 May 2013; 20 May 2013; accepted: 21 May 2013
Short-lasting unilateral neuralgiform headache attacks
with conjunctival injection and tearing (SUNCT) is a
primary headache syndrome characterized by unilateral
stabbing or throbbing orbital or temporal pain. There
should be at least 20 attacks, lasting 5–240 seconds, and
both ipsilateral conjunctival injection and lacrimation
(1). Only one of those symptoms or other cranial auto-
nomic signs, along with a longer duration of attacks
(two seconds–10 minutes), characterize short-lasting
unilateral neuralgiform headache with cranial auto-
nomic symptoms (SUNA), which is currently classified
in the Appendix of the International Classification of
Headache Disorders, 2nd edition (ICHD-II) (1). It is
not clear whether SUNCT is a subset of SUNA or
whether the two are separate entities (1–3). In addition,
the growing magnetic resonance imaging (MRI) evi-
dence of neurovascular compression in SUNCT/
SUNA cases, commonly detected in the classic trigem-
inal neuralgia, raises the question of whether it is an
incidental finding or a symptomatic cause (2,3). We
describe three new SUNCT cases with brain MRI find-
ing of neurovascular compression and critically review
IRCCS Institute of Neurological Sciences of Bologna, Headache Centre,
Department of Biomedical and Neuromotor Sciences, University of
Sabina Cevoli, IRCCS Institute of Neurological Sciences of Bologna,
Ospedale Bellaria, Via Altura, 3, 40139 Bologna, Italy.
! International Headache Society 2013
Reprints and permissions:
published literature on SUNCT/SUNA cases, including
We describe three new SUNCT patients with brain
MRI finding of neurovascular compression attending
our headache center between February 2009 and
November 2011. We electronically searched PubMed
using the following keywords: trigeminal, autonomic,
neuralgia, SUNCT, SUNA, lacrimation, tearing, rhi-
norrhea, nasal congestion, ptosis, miosis, mydriasis,
eyelid edema, facial flushing, sweating, neurovascular
(MVD). We screened the English-language literature
to identify SUNCT and SUNA cases, including those
with neurovascular compression. We carefully reviewed
the reference lists of all the papers found. A few
abstracts were included. Secondary SUNCT/SUNA
cases were excluded to have a homogeneous group for
comparison. However, we were not able to exclude
three secondary cases reported in the largest SUNCT/
SUNA series (one SUNA patient with multiple scler-
osis and two SUNCT patients with pituitary lesions)
because they were not reported separately.
Only cases with both conjunctival injection and tear-
ing were classified as SUNCT, according to ICHD-II
criteria. We focused on the following clinical items:
trigeminal involvement and neurovascular compres-
sion, gender, age at onset, laterality of pain, cranial
therapeutic options and surgical treatment. Statistical
analysis was performed using chi square for categorical
data, and unpaired two-tailed t test for continuous vari-
ables. A p value<0.05 was considered significant.
Case 1 (4)
A 55-year-old man with familial thrombocytopenia
complained of headache attacks since the age of 46
years. The headache was characterized by severe
shock-like burning pain occurring over the right orbital
and temporal areas, accompanied by ipsilateral con-
junctival injection, ptosis, lacrimation and rhinorrhea.
The attacks occurred spontaneously or could be trig-
gered by shaving, chewing or touching the right side of
the face. A refractory period after tactile allodynic
elicitations was absent. During the attacks the patient
was restless and irritable. At onset, he had clusters of
attacks with stabs lasting 30–60 seconds that would
occur for several minutes (up to 40 minutes), from
one to six times a day, mainly during daytime, for
nearly a month once a year. Physical and neurological
examinations were normal. He was administered gaba-
pentin (600mg daily) and was rendered asymptomatic.
In the last year before our examination the attacks had
worsened in frequency, occurring daily and nightly
about 15 per hour becoming shorter in duration (sec-
onds). Therapy with pregabalin 300mg/day and verap-
corticosteroids (methylprednisolone 500mg/day) were
then administered for three days without efficacy, thus
they were discontinued together with the other prevent-
ive agents, and indomethacin up to 150mg/day and
carbamazepine (CBZ) up to 1000mg/day were started.
After two days the attacks had dramatically decreased
in frequency. Indomethacin was stopped after one
month when a complete remission was achieved and
maintained with CBZ only. Repeated attempts to
stop CBZ led to a recurrence of attacks. A brain
MRI performed at the onset of symptoms was inter-
preted as normal. A second MRI was repeated after the
worsening of headache and documented an aberrant
loop of the right superior cerebellar artery compressing
the ipsilateral trigeminal nerve (Figure 1). The good
response to pharmacological treatment and the pres-
ence of thrombocytopenia contraindicated surgical
A 53-year-old woman with a medical history of hyper-
tension and glaucoma presented with a three-year his-
tory of headache. She described strictly right-sided
episodes of severe electric shock-like pain, occurring
on periorbital and temporal areas, accompanied by
prominent ipsilateral lacrimation, conjunctival injec-
tion, rhinorrhea, miosis and facial flushing. The
attacks occurred spontaneously or could be triggered
by touching the face, talking, chewing and moving the
eyes. The pain usually lasted from two seconds to
three minutes, occurring daily and nightly more than
100 times daily, for nearly a month once a year. Brain
MRI performed at the onset of symptoms demon-
strated hypo-density of the left caudate nucleus, com-
patible with a lacunar area. Therapy with oxygen,
opioid analgesics, anticonvulsants (lamotrigine and
valproic acid), and nonsteroidal analgesics, including
indomethacin, proved ineffective. The patient came to
our attention in February 2011 because of worsening
headache. Neurologic examination displayed aniso-
coria, allodynia and hyperesthesia over the right tri-
geminal distributions during the attacks. The right
corneal reflex could trigger the attacks. A refractory
period was absent. A diagnosis of SUNCT was made.
A second MRI scan documented right superior cere-
bellar artery compression on the ipsilateral trigeminal
nerve (Figure 2). The patient was administered CBZ
(600mg/day), pregabalin (450mg/day) and indometh-
acin (150mg/day) that maintained her asymptomatic
for three months. Following a head trauma in
September 2011 the symptoms recurred. The patient
was again administered CBZ (600mg/day), pregabalin
(450mg/day) and indomethacin (150mg/day) with
partial improvement. In March 2012, she finally
underwent MVD of the trigeminal nerve, which deter-
mined a complete disappearance of symptoms during
the following 11 months of observation, without any
A 47-year-old man with an unremarkable medical his-
tory complained of headache attacks since the age of 44
years. At onset, the attacks were characterized by
severe shock-like pain occurring on the right orbital
area, ear and nose, accompanied by ipsilateral conjunc-
tival injection, ptosis and lacrimation. Usually the
occurred>20 times a day. The patient was pain free
at nighttime unless he got up. The attacks occurred
spontaneously or could be triggered by moving the
head, washing or touching the right side of the face,
chewing, or wind hitting the face. Initially, the patient
was misdiagnosed as having right chronic sinusitis on
the basis of sinus computed tomography (CT) scan evi-
dence. Sinusotomy using balloon catheter devices was
attempted but after surgery he experienced worsening
headache. Brain MRI documented a neurovascular
compression. The patient underwent first gamma
Figure 2. FISP MPR coronal reconstructed image (a) revealed the trigeminal nerve in its cisternal tract, bilaterally (arrows); 3D-TOF
MR angiography sequence MPR coronal reconstruction (b) and MIP reconstruction (c) showed the contact between right superior
cerebellar artery and the upper surface of the nerve (arrowhead).
FISP: fast imaging with steady-state precession; MPR: multi-planar reconstruction; 3D-TOF MR: three-dimensional time-of-flight
magnetic resonance; MIP: maximum intensity projection.
Figure 1. FSPGR MPR coronal reconstructed image (a) revealed the trigeminal nerve in its cisternal tract, bilaterally (arrows); 3D-
TOF MR angiography sequence MPR coronal reconstruction (b) showed the contact between left superior cerebellar artery and the
upper surface of the nerve (arrowhead).
FSPGR: fast spoiled gradient echo; MPR: multi-planar reconstruction; 3D-TOF MR: three-dimensional time-of-flight magnetic
Favoni et al.
knife treatment without any improvement and one year
later surgery for MVD. During the procedure, the com-
pression of the right superior cerebellar artery and right
peritrigeminal vein on the ipsilateral trigeminal nerve
was identified and a Teflon sponge was interposed to
maintain the separation. After four weeks the attacks
returned with the same features previously described,
but the patient was now experiencing a continuous
burning pain involving all three branches of the right
trigeminal nerve. A trial with CBZ (up to 600mg daily)
offered little benefit and was poorly tolerated. Gamma
knife and sinusotomy were performed again without
any improvement. Therefore, in November 2011, the
patient came to our attention and the neurological
examination showed only dysesthesia within the right
V1-V2-V3 trigeminal areas. Repeat brain MRI with
gadolinium documented greater thickness and contrast
enhancement of the right trigeminal nerve, compatible
with inflammation. The cerebrospinal fluid demon-
strated a normal opening pressure (200mmH20) and
no oligoclonal bands,excluding
system (CNS) inflammation. Intravenous corticoster-
oids were started without lasting benefit. Treatment
with indomethacin (150mg daily) was ineffective.
(50mg daily) and palmitoylethanolamide
daily) seemed to reduce the lacrimation, but not the
frequency or intensity of attacks. Verapamil (240mg
daily) was added without improvement. Oxcarbama-
zepine (1200mg daily) plus lamotrigine provided bene-
fit proximally by 50% on both anesthesia dolorosa and
We found 222 cases in the literature. Of these, 189 had
SUNCT attacks, 31 SUNA (2,3,5–15), two both
SUNCT and SUNA (3,5).
Trigeminal involvement and neurovascular
A total of 201 of 222 (90.5%) published cases of
SUNCT/ SUNA reviewed were investigated by brain
MRI. Thirty-four of 201 cases (16.9%) had MRI evi-
dence of vascular compression of the trigeminal nerve
(2,3,5,7,15–27). Twenty-seven were SUNCT patients,
five were SUNA (3,5,7,15) and two had both SUNCT
and SUNA attacks (3,5). Unfortunately, many reports
do not clarify whether MRI-dedicated views of the tri-
geminal nerve root were performed, so the presence of
neurovascular compression may be underestimated. In
particular, in a series of 24 SUNCT/SUNA cases, neu-
rovascular compression was detected in 15 of 17
patients (88%) who underwent MRI with dedicated
views for the trigeminal nerve (3). By contrast, a vas-
cular compression of the trigeminal nerve was seen in
only three SUNCT patients (7%) in the largest series of
52 SUNCT/SUNA cases (2). The vessels in contact
with the trigeminal nerve were the superior cerebellar
artery (SCA) alone in 17 cases (50%) (3,5,15–
17,19,20,22,23,25–27), the anterior inferior cerebellar
artery (AICA) alone in one patient (3%) (17), and the
vertebral artery in one patient (3%) (21). Both SCA or
AICA and vein were involved in three cases (9%) (3,5).
Bilateral loop was reported in five cases (14.7%) (2,3,5).
Results from the literature review on neurovascular
compression are presented in Table 1.
Gender and age at onset
Considering all patients together, information on sex
and age at onset were available in 215 (96.8%) and
200 (90.1%) cases, respectively. A total of 109
SUNCT patients were male and 74 female (M:F ratio
of 1.5:1). Eleven SUNA cases were male and 19 female
(M:F ratio of 1:1.7). The two cases with both SUNCT
and SUNA attacks were male patients, aged 54 and 58
years, respectively (3,5). The mean age at onset was
47.62?17.11 years considering all patients together:
48.30?17.33 years in SUNCT patients and 42.29?
15.64 years in SUNA syndrome, with no significant
Among SUNCT/SUNA patients with MRI evidence
of neurovascular compression, 19 were male and 12
female (M:F ratio of 1.6:1). Age ranges at onset were
available in 31 cases. The mean age at onset was
Cranial autonomic symptoms
Autonomic features were described in 158 (83.6%) of
189 SUNCT patients and in 26 (83.9%) of 31 SUNA.
Comparing SUNCT and SUNA patients, statistically
significant differences were observed only for conjunc-
tival injection and lacrimation: 158 vs seven (100% vs
26.9%, p<0.001) and 158 vs 12 (100% vs 46.2%,
p<0.001), respectively. Autonomic symptoms were
clearly described in 17 (50%) of 34 SUNCT/SUNA
patients with MRI evidence of a neurovascular com-
pression. No statistically significant differences were
observed comparing patients with and without neuro-
vascular compression. The distribution of cranial auto-
nomic symptoms is shown in Table 2.
Laterality of pain
The side of attacks was available in 205 (92.3 %) of 222
patients. Ninety-two (44.8%) had left-sided attacks, 90
(43.9%) right-sided attacks, seven (3.4%) unilateral
attacks, but the side was not available; 12 (5.9%)
patients had side-shift attacks, four (2%) bilateral
Among SUNCT/SUNA patients with MRI evidence
of a neurovascular compression, 13 (38.2%) patients
had left-sided attacks, 15 (44.1%) had right-sided
attacks, two (5.9%) had unilateral attacks, but the
side was not available; one patient (2.9%) had side-
shift attacks and three (8.8%) had bilateral attacks
Data on triggers factors were reported in 176 (79.3%)
of 222 patients. Triggers such as touching the face,
washing, shaving, eating, chewing, brushing teeth, talk-
ing, coughing or neck movement were identified in 128
(72.7%) of 176 patients. Conversely, 48 (27.3%)
patients had exclusively spontaneous attacks. Of the
34 SUNCT/SUNA patients with neurovascular com-
pression, 26 (76.5%) had triggers, five (14.7%) had
Table 1. Neurovascular compression in SUNCT and SUNA patients.
onset SexDiagnosis Pain side
Ertsey et al. 2000 (16)68M SUNCTR Yes N/ASCA CBZþtiapride
MVD Gardella et al. 2001 (17)43F SUNCTR Yes N/A SCA
Ko ¨seoglu et al. 2005 (18)48M SUNCTL YesN/A AICA None
Lagares et al. 2005 (19) 54F SUNCTL YesNo SCAMVD
Sprenger et al. 2005 (20)47M SUNCTR Yes N/ASCAMVD
Zidverc-Trajkovic et al. 2005 (21)68M SUNCTL YesN/A VAGBP
Cohen et al. 2006 (2)N/A N/ASUNCTU N/AN/AVL N/A
Cohen et al. 2006 (2)N/AN/A SUNCTU N/A N/A Bilateral VLN/A
Cohen et al. 2006 (2)N/A N/A SUNCTBilateralN/AN/A VLN/A
Williams et al. 2008–2010 (3,5) 43F SUNCT BilateralYesNo Bilateral SCANone
Williams et al. 2008–2010 (3,5)53MSUNCTL NoNoAICA and VMVD
Williams et al. 2008–2010 (3,5)46FSUNCTRYesNoSCAMVD
Williams and Broadley 2008 (3)47FSUNCT L/RYesNoBilateral AN/A
Williams and Broadley 2008 (3)17FSUNCTBilateralYesNoBilateral AN/A
Williams et al. 2008–2010 (3,5) 46FSUNAL YesNoAICA and VNone
Williams and Broadley 2008 (3) 19FSUNCTL NoNoAN/A
Williams et al. 2008–2010 (3,5)58M SUNCT/SUNAR NoNo SCA and VNone
Williams et al. 2008–2010 (3,5)19MSUNCTL YesNoSCA MVD
Williams et al. 2008–2010 (3,5) 55MSUNA RYesNoSCA MVD
Williams et al. 2008–2010 (3,5) 65FSUNCTR Yes No SCAMVD
Williams and Broadley 2008 (3)58MSUNALNoNoAN/A
Williams and Broadley 2008 (3)32F SUNCTR YesNoBilateral AN/A
Williams et al. 2008–2010 (3,5)54 MSUNCT/SUNALYesNoSCAMVD
Williams and Broadley 2008 (3)54MSUNCTLYesNoAN/A
Jacob et al. 2008 (7)29M SUNARNoNo VLGBP
Guerreiro et al. 2009 (22)57MSUNCTLYesN/ASCAMVD
Irimia et al. 2010 (23)54MSUNCTLYes NoSCAMVD
Maggioni et al. 2010 (24)40 FSUNCTRYes NoALMT; GBPþamitryptiline
DBSBartsch et al. 2011 (25)67MSUNCTRYesN/A SCA
Chaila et al. 2011 (26)45 MSUNCTL YesN/ASCAMVD
Paliwal et al. 2012 (27) 60MSUNCTR YesYesVLOXAþLMT
LMTPaliwal et al. 2012 (27)55MSUNCTRYesYesSCA
Paliwal et al. 2012 (27)64MSUNCTR YesYes SCA CBZ
Rinaldi et al. 2013 (15)60 FSUNAR YesN/ASCANone
SUNCT: short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing; SUNA: short-lasting unilateral neuralgiform
headache with cranial autonomic symptoms; M: male; F: female; L: left; R: right; U: unilateral; AICA: anterior inferior cerebellar artery; SCA: superior
cerebellar artery; VA: vertebral artery; VL: vascular loop; A: artery; V: vein; CBZ: carbamazepine; MVD: microvascular decompression; LMT:
lamotrigine; GBP: gabapentin; DBS: deep brain stimulation; OXA: oxacarbazepine; N/A: not applicable.
Favoni et al.
only spontaneous attacks and in three cases data were
not available (2).
Data on a refractory period were available in 93
(41.9%) of 222 SUNCT/SUNA cases. No refractory
period was reported in 84 (90.3%) cases. Nine (9.7%)
patients had a preserved refractory period (2,27–30).
Three of them had MRI evidence of neurovascular
SUNCT/SUNA: Therapeutic options
Lamotrigine alone (up to 400mg) was effective in 52 of
81 patients (64.2%) (3,10,12,27,31–42), gabapentine (up
to 3600mg) in 24 of 71 (33.8%) (7,31,43–47), topira-
mate (up to 400mg) in 14 of 43 (32.6%) (32,48,49),
(3,7,12,27,31,50–53), oxcamazepine in one of seven
(14.3%) (54), corticosteroids in seven of 22 (31.8%)
(30,40,46,55,56), verapamil in one of 18 (5.5%) (57),
zonisamide in one of two (58), and lomerazide hydro-
chloride in one of two patients (9).
Indomethacine was effective in only one out of 129
cases (0.78%) (59). Other anti-inflammatory drugs were
completely ineffective. None of the 28 patients who
received triptans achieved benefits. Headache remission
was achieved in 22 of 26 (84.6%) patients treated with
intravenous lidocaine and in nine of 11 (81.8%)
(3,5,10,31,60). A complete response to associations of
two or more drug therapies was reported in single cases
(Table 3). Of the 34 SUNCT/SUNA patients with MRI
evidence of a neurovascular compression, pharmaco-
logical treatment was reported as effective in seven
cases and ineffective in 16 patients (Table 1).
MVD. Sixteen of 34 (47.1%) patients with MRI findings
of a neurovascular compression underwent MVD of
the trigeminal nerve. Twelve were SUNCT patients,
two SUNA and two had both SUNCT and SUNA
attacks (5,17,19,20,22,23,26). Complete pain relief was
achieved in 12 of the 16 patients (75%) lasting up to 32
months after surgery. In addition, three more SUNCT
patients with normal MRI studies were unsuccessfully
treated with MVD (61,62). Two of them had been pre-
viously treated with a trigeminal ablative procedure
(61). Data about MRI findings preceding MVD were
not available in two SUNCT patients. One of them was
asymptomatic for two years following surgery, then the
pain recurred (52). The second one experienced a com-
plete resolution of symptoms after three months of
follow-up (63) (Table 1).
Local blockades. Blockade of the greater occipital nerve
(GON) was performed in 25 patients. Seven SUNCT
patients (28%) achieved a good response: five cases
with lidocaine and depomedrone (31), one with bupiva-
caine (64), and another one with an association of
Table 2. Autonomic features.
SUNCT vs SUNA
SUNCT/SUNA with vs
compression(pts¼158)(pts¼26) (pts¼17) (pts¼167)
n (%)n (%)pn (%)n (%)p
Conjunctival injection 158 (100%) 7 (26.9%)
15 (88.2%)150 (89.8%)0.83
Lacrimation 158 (100%)12 (46.2%) 16 (94.1%) 154 (91.6%)0.91
Rhinorrhea77 (48.7%) 7 (26.9%)0.067 (41.2%) 77 (46.1%)0.89
Nasal obstruction 52 (32.9%)8 (30.8%)0.99 4 (23.5%)56 (33.5%)0.57
Ptosis 47 (29.7%)6 (23.1%)0.64 4 (23.5%)49 (29.3%)0.82
Miosis7 (4.4%) 0 (0%)0.581 (5.9%) 6 (3.6%)0.84
Midryasis2 (1.3%) 0 (0%)0.66 1 (5.9%) 1 (0.6%)0.44
Eyelid edema50 (31.6%) 4 (15.4%)0.15 4 (23.5%)50 (29.9%) 0.78
Facial flushing12 (7.6%)4 (15.4%)0.35 2 (11.8%)14 (8.4%) 0.98
Sweeting 12 (7.6%)3 (11.5%)0.77 2 (11.8%)13 (7.8%)0.92
Salivation1 (0.6%) 0 (0%)0.30 0 (0%) 1 (0.6%)0.16
Others4 (2.5%)3 (11.5%) 0.09 0 (0%)7 (4.2%) 0.85
SUNCT: short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing; SUNA: short-lasting unilateral neuralgiform
headache with cranial autonomic symptoms; pts: patients.
triamcinolone, dexamethasone and lidocaine (65).
Partial improvement within 72 hours had been reported
in another patient treated with a single GON injection
of steroids and local anesthetic (40). Three SUNA
patients experienced improvement by approximately
50% lasting from two to seven weeks after GON injec-
tions (10). Conversely, the treatment was ineffective in
11 patients with SUNCT (10,25,31,37,66,67).
A response to sphenopalatine ganglion and intraorbi-
tal phenol blockades was reported in one SUNCT
patient, with a benefit up to 12 months (13). Another
SUNCT patient experienced
during the 16-month follow-up period following local
opioid blockade of the superior cervical ganglion (68).
However, local blockades of the supraorbital nerve (11
patients), infraorbital nerve (eight patients) and orbicu-
laris oculi muscles (two patients) were performed without
benefits. Blockades of the lacrimal nerve, gasserian gan-
glion (13), stellate ganglion, and the retrobulbar region
were reported as ineffective in single cases (13,67,69). One
patient received local lidocaine injections around the
periorbital region with no improvement (65). Recently
onabotulinumtoxin A was reported to reduce pain inten-
sity and frequency in a single SUNCT case (70).
No local blockades had been used in any of the
SUNCT/SUNA patients with MRI findings of a neu-
Glycerol rhizotomy, gamma knife radiosurgery and percutan-
eous balloon compression. Three of the four patients
(three SUNCT and one SUNA) who underwent gly-
cerol rhizotomy achieved complete pain relief lasting
from seven months to four years. Two of them were
successfully treated twice (13). Conversely, another
Table 3. Drug therapies effective in SUNCT/SUNA cases.
EffectiveIneffective Effective Ineffective
Subcutaneous lidocaine infusions
SUNCT: short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing; SUNA: short-lasting
unilateral neuralgiform headache with cranial autonomic symptoms; pts: patients.
Favoni et al.
SUNCT patient underwent glycerol rhizotomy with
only two weeks of pain reduction (61). Two SUNCT
patients underwent gamma knife targeting both the tri-
geminal nerve and the sphenopalatine ganglion. The
first had complete pain resolution at 39 months
follow-up, without medication (71); the second experi-
enced only rare provoked attacks at four months’
follow-up (72). Another two SUNCT patients under-
went gamma knife radiosurgery targeting only the tri-
geminal nerve. The first had near complete pain relief
for about two months, then the pain gradually recurred
and he suffered from anesthesia dolorosa; the second
had no improvement (61).
Two SUNCT patients showed an apparent benefit
following radiofrequency thermocoagulation of the tri-
geminal nerve that lasted for two and three years,
respectively (37,62). Another three SUNCT patients
were pain free after percutaneous balloon compression
of the gasserian ganglion with benefits ranging from 16
months to ten years of follow-up (13,28,73).
No SUNCT/SUNA patients with MRI findings of a
neurovascular compression underwent destructive tri-
geminal nerve procedures.
Neurostimulation. One drug-resistant SUNCT patient
with MRI evidence of neurovascular compression
underwent successful hypothalamic deep brain stimula-
tion (DBS), although the preoperative oral medication
was not suspended because of an increased frequency of
attacks on drug reduction. The patient had previously
undergone MVD without benefit (25). Another two
patients with intractable SUNCT underwent successful
DBS (66,74). The first patient experienced an 80%
reduction in daily attack frequency without prophylac-
tic therapy after 12 months of stimulation (66). The
second achieved a complete response, but lamotrigine
treatment was added to eliminate sporadic attacks (74).
Seven refractory patients (six SUNCT, one SUNA)
underwent bilateral occipital nerve stimulation (ONS):
At a median follow-up of two years four of them
achieved an excellent improvement, one moderate,
one transient and one no benefit (11).
We describe three new SUNCT patients with brain
MRI evidence of vascular compression of the trigem-
inal nerve. These cases of SUNCT raise questions
about the role played by neurovascular compression
in SUNCT/SUNA manifestations and the best thera-
peutic approach to these patients.
The first patient we describe achieved a complete
remission of symptoms with CBZ, which is the treat-
ment of choice in trigeminal neuralgia (TN). The effi-
cacy of CBZ has already been described in the
treatment of SUNCT, including a case with MRI find-
ings of neurovascular compression (27).
It is also likely that the second patient we report is a
case of SUNCT secondary to trigeminal neurovascular
contact, as suggested by the complete response to MVD
and the ineffectiveness of medical treatment.
The third case we describe had a more atypical and
complicated clinical course. The patient was initially
misdiagnosed as having chronic sinusitis. No medical
treatment was tried before he underwent unsuccessful
gamma knife, MVD, and a second gamma-knife treat-
ment. Following the MVD the patient started experien-
cing continuous burning pain involving all three
branches of the trigeminal nerve, and exacerbation of
SUNCT attacks. The MRI showed greater thickness
and contrast enhancement of the right trigeminal
nerve that may be compatible with a gamma-knife
side effect. We observed only a partial response to an
association of oxcamazepine and lamotrigine, and fail-
ure of other drugs. The clinical course of this patient is
in line with other authors’ hypothesis that a previous
trigeminal ablative procedure may explain the ineffect-
iveness of MVD in some SUNCT patients as occurs in
TN (23). In addition, our case suggests that surgery
should be attempted only after pharmacological treat-
Clinical features and neurovascular compression
We reviewed the available literature to establish
whether SUNCT and SUNA are different entities and
to clarify the pathogenetic role played by the vascular
loop closest to the trigeminal nerve. Since the last
SUNCT syndrome review of medical literature (75),
the number of published cases had rapidly increased
in parallel with the rising recognition of a neurovascu-
lar compression associated with the syndrome, detected
in 16.9% of SUNCT/SUNA patients. Taking all
patients together, our results confirm a mild male pre-
ponderance in SUNCT syndrome, female in SUNA
syndrome, and the presence of trigger factors (2,3,75).
Conversely, age at onset was not significantly different
between SUNCT and SUNA patients. Globally, we
observed no side prevalence. Bilateral and side-shift
attacks are also reported (2,3,48,76). According to the
ICHD-II, all SUNCT patients have both conjunctival
injection and lacrimation. In one-third of SUNA cases
there is conjunctival injection or alternatively lacrima-
tion. The presence of other autonomic features in
SUNA is similar to SUNCT cases. Interestingly, mild
autonomic symptoms are also reported in TN series
and seem to correlate with a worse prognosis after
MVD (77,78), raising the question of whether they
should be considered SUNCT/SUNA cases rather
than TN with autonomic features. A diagnostic clinical
sign able to distinguish SUNCT/SUNA from TN is the
absence of a refractory period. However, as in TN, a
refractory period is described in a few SUNCT syn-
drome cases, including three with neurovascular com-
pression (2,27,29,30). Unfortunately many of the
previous features may be underreported. Similarly,
many subjects were not investigated systematically
with dedicated trigeminal views on MRI. The largest
previous series reported a vessel closest to the trig-
eminal nerve in 7% and 88% of patients (2,3).
The relevanceof neurovascular
the pathophysiology of SUNCT/SUNA is unclear.
However, the involvement of the trigeminal autonomic
reflex in these syndromes is widely acknowledged (79).
A vascular compression of the trigeminal nerve has
been implicated as a peripheral afferent stimulation to
this reflex. The persistent stimulation could lead to cen-
tral disinhibition of the trigeminal autonomic reflex by
hypothalamic activation (5,15). Functional MRI stu-
dies have demonstrated ipsilateral, controlateral and
bilateral hypothalamic activation in SUNCT and
SUNA syndromes (20,31,80). Interestingly, bilateral
hypothalamic activation was reported in a SUNCT
case with MRI finding of neurovascular compression
who underwent successful MDV (20). The high per-
centage (75%) of SUNCT/SUNA cases who remitted
after MVD supported the pathogenetic role of trigem-
inal neurovascular compression. Conversely, doubts
about this assumption are related to the description
of one case with unilateral compression of the trigem-
inal nerve with bilateral pain and two patients with
bilateral loops but with unilateral attacks (2,3). It
could be argued that these findings are coincidental
and not relevant to pain production. In this context,
a simple contact between a vessel and the nerve may
not cause true nerve compression.
Among all cases reviewed, lamotrigine should be con-
sidered the treatment of choice, with efficacy reported
in 64% of patientswhen
Topiramate and gabapentin may be prescribed as an
effective preventive medication in one-third of patients
treated. SUNCT and SUNA patients have also been
reported to respond to CBZ (21%). A variety of other
drugs, alone or in combination, have been tried in
SUNCT/SUNA syndrome, none with reliable effect-
iveness. Among acute treatments, triptans, analgesics
and opioids have not been consistently effective in
SUNCT/SUNA treatment. In particular, indometh-
acin has been generally
Interestingly, intravenous or subcutaneous lidocaine
seems highly effective for short-term prevention (2,3).
The dramatic improvement in SUNCT and SUNA in
response to lidocaine has led some authors to propose
this pharmacological test as a diagnostic clinical tool
to confirm the diagnosis (60). However, there are no
randomized controlled trials because of the rarity of
Trigeminal surgery should be considered only in
selected medically intractable patients, in whom MVD
should be considered the treatment of choice with an
efficacy reported in 75% of subjects with neurovascular
compression. On the contrary, limited evidence sup-
ports benefits after greater occipital nerve procedures,
glycerol rhizotomy, gamma-knife radiosurgery, radio-
frequency thermocoagulation, or percutaneous balloon
compression. To date, few patients with drug-resistant
SUNCT were successfully treated with DBS or ONS,
one of them with MRI findings of neurovascular com-
pression, and further reports are needed.
Our review has some limits. Although every effort was
made to include all patients in the literature presenting
with SUNCT and SUNA attacks, some reported cases
may have been missed, or duplicated, especially in the
case of publications from the same research center. Few
secondary headaches were included. In addition, much
clinical information was reported differently or lacking
so it was difficult to make comparisons, particularly on
Our three new SUNCT patients with MRI findings of
neurovascular compression have been added to the 34
SUNCT/SUNA cases described in the literature. We
suggest that brain MRI with a dedicated view of the
trigeminal nerve should always be performed to exclude
patients. In agreement with Cohen and colleagues (2),
we suggest that SUNCT and SUNA should be con-
sidered clinical phenotypes of the same spectrum of
presentation in the new revision of the international
classification for several reasons. First, we confirm
these headache syndromes share the same clinical char-
acteristics, except for lacrimation and conjunctival
injection. Second, both SUNCT and SUNA attacks
are present in the same patients. Third, neurovascular
compression is detected in both conditions. Finally, the
syndromes respond to the same drug therapies, with
lamotrigine as the first-line treatment. For acute treat-
ment, lidocaine should be considered in selected cases.
Surgical approaches should be attempted in refractory
cases. In particular, as for TN, MVD may represent a
treatment option in patients with MRI evidence of neu-
Favoni et al.
. Short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing (SUNCT)
and short-lasting unilateral neuralgiform headache with cranial autonomic symptoms (SUNA) should be
considered clinical phenotypes of the same syndrome.
. Brain magnetic resonance imaging (MRI) should always be performed with a dedicated view to exclude
. The high percentage of remittance after microvascular decompression (MVD) supports a pathogenetic role
of the neurovascular compression.
The authors thank Dr Francesco Toni (Neuroradiology
Department, Bellaria Hospital, IRCCS Istituto delle Scienze
Neurologiche di Bologna) for reviewing the patients’ MR
images with us, and Anne Collins for English editing.
This research received no specific grant from any funding
agency in the public, commercial, or not-for-profit sectors.
Conflict of interest
1. Headache Classification Committee of The International
Headache Society. The International Classification of
Headache Disorders, 2nd edition. Cephalalgia 2004;
24(Suppl 1): 1–160.
2. Cohen AS, Matharu MS and Goadsby PJ. Short-lasting
unilateral neuralgiform headache attacks with conjunctival
injection and tearing (SUNCT) or cranial autonomic fea-
tures (SUNA)—a prospective clinical study of SUNCT
and SUNA. Brain 2006; 129: 2746–2760.
3. Williams MH and Broadley SA. SUNCT and SUNA:
Clinical features and medical treatment. J Clin Neurosci
2008; 15: 526.
4. Grimaldi D, Batista S, Poisson A, et al. The Almirall
European Headache Awards 2009. J Headache Pain
2010; 11: 207–214.
5. Williams M, Bazina R, Tan L, et al. Microvascular decom-
pression of the trigeminal nerve in the treatment of
SUNCT and SUNA. J Neurol Neurosurg Psychiatry
2010; 81: 992–996.
6. Volcy M, Tepper SJ, Rapoport AM, et al. Short-lasting
unilateral neuralgiform headache attacks with cranial
Cephalalgia 2005; 25: 470–472.
7. Jacob S, Saha A and Rajabally Y. Post-traumatic short-
lasting unilateral headache with cranial autonomic symp-
toms (SUNA). Cephalalgia 2008; 28: 991.
8. Baldacci F, Nuti A, Lucetti C, et al. SUNA syndrome with
seasonal pattern. Headache 2009; 49: 912–914.
9. Tada Y, Ikuta N and Negoro K. Short-lasting unilateral
neuralgiform headache attacks with cranial autonomic
symptoms (SUNA). Intern Med 2009; 48: 2141–2144.
(SUNA)—a case report.
10. Lambru G, Nesbitt A, Shanahan P, et al. Coexistence of
hemiplegic migraine with SUNCT or SUNA: A case
series. Cephalalgia 2012; 32: 258–262.
11. Matharu M, Watkins L and Shanahan P. POH04 Treat-
ment of medically intractable SUNCT and SUNA with
Psychiatry 2010; 81: e51–e52.
12. Leone M, Mea E, Genco S, et al. Coexistence of TACS
and trigeminal neuralgia: Pathophysiological conjectures.
Headache 2006; 46: 1565–1570.
13. Hannerz J and Linderoth B. Neurosurgical treatment of
short-lasting, unilateral, neuralgiform hemicrania with
conjunctival injection and tearing. Br J Neurosurg 2002;
14. Truini A, Barbanti P, Galeotti F, et al. Trigeminal sen-
sory pathway function in patients with SUNCT. Clin
Neurophysiol 2006; 117: 1821–1825.
15. Rinaldi F, Rao R, Venturelli E, et al. Where SUNCT
contacts TN: A case report. Headache. Epub ahead of
print 10 January 2013. DOI: 10.1111/head.12007.
16. Ertsey C, Bozsik G and Afra J. A case of SUNCT syn-
drome with neurovascular compression. Cephalalgia
2000; 20: 325.
17. Gardella L, Viruega A, Rojas H, et al. A case of a patient
with SUNCT syndrome treated with Jannetta procedure.
Cephalalgia 2001; 21: 996–999.
18. Ko ¨ seoglu E, Karaman Y, Ku ¨ cu ¨ k S, et al. SUNCT syn-
drome associated with compression of trigeminal nerve.
Cephalalgia 2005; 25: 473.
19. Lagares A, Go ´ mez PA, Pe ´ rez-Nun ˜ ez A, et al. Short-last-
ing unilateral neuralgiform headache with conjunctival
injection and tearing syndrome treated with microvascu-
lar decompression of the trigeminal nerve: Case report.
Neurosurgery 2005; 56: E413.
20. Sprenger T, Valet M, Platzer S, et al. SUNCT: Bilateral
hypothalamic activation during headache attacks and
resolving of symptoms after trigeminal decompression.
Pain 2005; 113: 422–426.
21. Zidverc-Trajkovic J, Mijajlovic M, Pavlovic AM, et al.
Vertebral artery vascular loop in SUNCT and concomi-
tant trigeminal neuralgia. Case report. Cephalalgia 2005;
22. Guerreiro R, Casimiro M, Lopes D, et al. Video Neuro-
Image: Symptomatic SUNCT syndrome cured after tri-
geminal neurovascular contact surgical decompression.
Neurology 2009; 72: e37.
23. Irimia P, Gonza ´ lez-Redondo R, Domı´nguez PD, et al.
Microvascular decompression may be effective for
J Neurol Neurosurg
refractory SUNCT regardless of symptom duration.
Cephalalgia 2010; 30: 626–630.
24. Maggioni F, Manara R, Mampreso E, et al. Trigeminal
neuralgia and trigeminal-autonomic cephalalgias: A con-
tinuum or simple co-existence? Cephalalgia 2010; 30:
25. Bartsch T, Falk D, Knudsen K, et al. Deep brain stimu-
lation of the posterior hypothalamic area in intractable
short-lasting unilateral neuralgiform headache with con-
junctival injection and tearing (SUNCT). Cephalalgia
2011; 31: 1405–1408.
26. Chaila E, Ali E, Rawluk D, et al. ‘Switching off’ SUNCT
by sudden head movement: A new symptom. J Neurol
2011; 258: 694–695.
27. Paliwal VK, Singh P, Kumar A, et al. Short-lasting uni-
lateral neuralgiform headache with conjunctival injection
and tearing (SUNCT) with preserved refractory period:
Report of three cases. J Headache Pain 2012; 13: 167–169.
28. Morales-Ası´n F, Espada F, Lo ´ pez-Obarrio LA, et al. A
SUNCT case with response to surgical treatment.
Cephalalgia 2000; 20: 67–68.
29. Auer T, Janszky J, Schwarcz A, et al. Attack-related
brainstem activation in a patient with SUNCT syndrome:
An ictal fMRI study. Headache 2009; 49: 909–912.
30. Trauninger A, Alkonyi B, Kova ´ cs N, et al. Methylpred-
nisolone therapy for short-term prevention of SUNCT
syndrome. Cephalalgia 2010; 30: 735–739.
31. Cohen AS. Short-lasting unilateral neuralgiform head-
ache attacks with conjunctival injection and tearing.
Cephalalgia 2007; 27: 824–832.
32. Malik K, Rizvi S and Vaillancourt PD. The SUNCT syn-
drome: Successfully treated with lamotrigine. Pain Med
2002; 3: 167–168.
33. Leone M, Rigamonti A, Usai S, et al. Two new SUNCT
cases responsive to lamotrigine. Cephalalgia 2000; 20:
34. D’Andrea G, Granella F, Ghiotto N, et al. Lamotrigine
in the treatment of SUNCT syndrome. Neurology 2001;
35. Gutierrez-Garcia JM. SUNCT syndrome responsive to
lamotrigine. Headache 2002; 42: 823–825.
36. Chakravarty A and Mukherjee A. SUNCT syndrome
responsive to lamotrigine: Documentation of the first
Indian case. Cephalalgia 2003; 23: 474–475.
37. Piovesan EJ, Siow C, Kowacs PA, et al. Influence of
lamotrigine over the SUNCT syndrome: One patient
follow-up for two years. Arq Neuropsiquiatr 2003; 61:
38. Cosentino G, Fierro B, Puma AR, et al. Different forms
of trigeminal autonomic cephalalgias in the same patient:
Description of a case. J Headache Pain 2010; 11: 281–284.
39. De Cesaris F, Fanciullacci M, Pietrini U, et al. Defining
neuralgiform headache with ipsilateral autonomic symp-
toms: Case report in a headache center. Intern Emerg
Med 2008; 3: 413–414.
40. Irimia P, Ga ´ llego-Pe ´ rez Larraya J and Martinez-Vila E.
Seasonal periodicity in SUNCT syndrome. Cephalalgia
2008; 28: 94–96.
41. Vikelis M, Xifaras M and Mitsikostas DD. SUNCT syn-
drome in the elderly. Cephalalgia 2005; 25: 1091.
42. Vega J, Mauri G and Pascual J. Coexistence of
SUNCT and aura? A case report. Cephalalgia 2012;
43. Etemadifar M, Maghzi AH, Ghasemi M, et al. Efficacy
of gabapentin in the treatment of SUNCT syndrome.
Cephalalgia 2008; 28: 1339–1342.
44. Hunt CH, Dodick DW and Bosch EP. SUNCT
45. Porta-Etessam J, Benito-Leon J, Martinez-Salio A, et al.
Gabapentin in the treatment of SUNCT syndrome.
Headache 2002; 42: 523–524.
46. Graff-Radford SB. SUNCT syndrome responsive to
gabapentin (Neurontin). Cephalalgia 2000; 20: 515.
47. Rocha Filho PA, Rabello GD, Galva ˜ o AC, et al. Gaba-
pentin in the treatment of SUNCT syndrome [in Portu-
guese]. Arq Neuropsiquiatr 2007; 65: 503–505.
48. Rossi P, Cesarino F, Faroni J, et al. SUNCT syndrome
successfully treated with topiramate: Case reports.
Cephalalgia 2003; 23: 998–1000.
49. Kuhn J, Vosskaemper M and Bewermeyer H. SUNCT
syndrome: A possible bilateral case responding to topir-
amate. Neurology 2005; 64: 2159.
50. Putzki N, Nirkko A and Diener HC. Trigeminal auto-
nomic cephalalgias: A case of post-traumatic SUNCT
syndrome? Cephalalgia 2005; 25: 395.
51. Raimondi E and Gardella L. SUNCT syndrome. Two
cases in Argentina. Headache 1998; 38: 369.
52. Schwaag S, Frese A, Husstedt IW, et al. SUNCT syn-
drome: The first German case series. Cephalalgia 2003;
53. Calvo JF, Bruera OC, de Lourdes Figuerola M, et al.
SUNCT syndrome: Clinical and 12-year follow-up case
report. Cephalalgia 2004; 24: 900–902.
54. Dora B. SUNCT syndrome with dramatic response to
oxcarbazepine. Cephalalgia 2006; 26: 1171–1173.
55. de Lourdes Figuerola M, Bruera O, Pozzo MJ, et al.
SUNCT syndrome responding absolutely to steroids in
two cases with different etiologies. J Headache Pain 2009;
56. Maiho ¨ fner C, Speck V, Sperling W, et al. Complete
remission of SUNCT syndrome by intravenous gluco-
corticoid treatment. Neurol Sci. Epub ahead of print 30
December 2012. DOI:10.1007/s10072-012-1282-8.
57. Narbone MC, Gangemi S and Abbate M. A case of
SUNCT syndrome responsive to verapamil. Cephalalgia
2005; 25: 476–478.
58. Ikawa M, Imai N and Manaka S. A case of SUNCT
syndrome responsive to zonisamide. Cephalalgia 2011;
59. Fragoso YD. Black, white and shades of grey: SUNCT or
Neuropsiquiatr 2006; 64: 575–577.
60. Arroyo AM, Dura ´ n XR, Beldarrain MG, et al. Response
to intravenous lidocaine in a patient with SUNCT syn-
drome. Cephalalgia 2010; 30: 110–112.
61. Black DF and Dodick DW. Two cases of medically and
surgically intractable SUNCT: A reason for caution and
an argument for a central mechanism. Cephalalgia 2002;
Favoni et al.
62. Matharu MS, Cohen AS and Goadsby PJ. SUNCT syn- Download full-text
drome responsive to intravenous lidocaine. Cephalalgia
2004; 24: 985.
63. Lenaerts M, Dieerich N and Phuce K. A patient with
SUNCT cured by the Jannetta procedure. Poster presen-
tations, Session IV, 460, 8th congress of the IHS.
Cephalalgia 1997; 17: 460.
64. Porta-Etessam J, Cuadrado ML, Gala ´ n L, et al. Tem-
poral response to bupivacaine bilateral great occipital
block in a patient with SUNCT syndrome. J Headache
Pain 2010; 11: 179.
65. Choi HJ, Choi SK, Lee SH, et al. Whiplash injury-
induced atypical short-lasting unilateral neuralgiform
headache with conjunctival injection and tearing syn-
drome treated by greater occipital nerve block. Clin J
Pain 2012; 28: 342–343.
66. Lyons MK, Dodick DW and Evidente VG. Responsive-
ness of short-lasting unilateral neuralgiform headache
with conjunctival injection and tearing to hypothalamic
deep brain stimulation. J Neurosurg 2009; 110: 279–281.
67. Pareja JA, Kruszewski P and Sjaastad O. SUNCT syn-
drome: Trials of drugs and anesthetic blockades.
Headache 1995; 35: 138.
68. Sabatowski R, Huber M, Meuser T, et al. SUNCT syn-
drome: A treatment option with local opioid blockade of
the superior cervical ganglion? A case report. Cephalalgia
2001; 21: 154.
69. Gantenbein AR and Goadsby PJ. Familial SUNCT.
Cephalalgia 2005; 25: 457.
70. Zabalza RJ. Sustained response to botulinum toxin in
SUNCT syndrome. Cephalalgia 2012; 32: 869–872.
71. Effendi K, Jarjoura S and Mathieu D. SUNCT syndrome
successfully treated by gamma knife radiosurgery: Case
report. Cephalalgia 2011; 31: 870–873.
72. Mathew T, Srinivas D, Aroor S, et al. SUNCT syndrome
treated with gamma knife targeting the trigeminal nerve
and sphenopalatine ganglion. J Headache Pain 2012; 13:
73. Baabor M, Rosas CS and Pe ´ rez-Limonte L. SUNCT
responsive to percutaneous balloon compression of the
gasserian ganglion—10-year follow-up. Headache 2010;
74. Leone M, Franzini A, D’Andrea G, et al. Deep brain
Neurol 2005; 57: 924–927.
75. Matharu MS, Cohen AS, Boes CJ, et al. Short-lasting
unilateral neuralgiform headache with conjunctival injec-
tion and tearing syndrome: A review. Curr Pain Headache
Rep 2003; 7: 308–318.
76. D’Andrea G and Granella F. SUNCT syndrome: The
first case in childhood. Short-lasting unilateral neuralgi-
form headache attacks with conjunctival injection and
tearing. Cephalalgia 2001; 21: 701–702.
77. Rasmussen P. Facial pain. IV. A prospective study of
1052 patients with a view of: Precipitating factors, asso-
ciated symptoms, objective psychiatric and neurological
symptoms. Acta Neurochir (Wien) 1991; 108: 100–109.
78. Simms HN and Honey CR. The importance of auto-
nomic symptoms in trigeminal neuralgia. Clinical article.
J Neurosurg 2011; 115: 210–216.
79. Leone M and Bussone G. Pathophysiology of trigeminal
autonomic cephalalgias. Lancet Neurol 2009; 8: 755–764.
80. May A, Bahra A, Buchel C, et al. Functional magnetic
resonance imaging in spontaneous attacks of SUNCT:
Short-lasting neuralgiform headache with conjunctival
injection and tearing. Ann Neurol 1999; 46: 791–794.