Greater occipital nerve injection in primary headache
syndromes – prolonged effects from a single injection
S.K. Afridi, K.G. Shields, R. Bhola, P.J. Goadsby*
Headache Group, Institute of Neurology, The National Hospital for Neurology and Neurosurgery, Queen Square, London, UK
Received 2 October 2005; received in revised form 27 December 2005; accepted 17 January 2006
Most patients with primary headache syndromes who have frequent attacks of pain have tenderness in the sub-occipital region.
Injection of the greater occipital nerve (GON) with local anesthetic and corticosteroids has been widely used in clinical practice for
many years, yet there is no clear understanding of its mechanisms of action. Moreover, there is no current gold-standard of practice
regarding GON injections in the management of headache. We audited of our practice to generate hypotheses about the range of
primary headaches that might benefit, to determine response rates to power future studies, and to assess whether we should continue
to do this procedure. Twenty-six of fifty-seven injections in 54 migraineurs yielded a complete or partial response that lasted for the
partial response a median of 30 days. For cluster headache 13 of 22 injections yielded a complete or partial response lasting for a
median of 21 days for the partial response. Tenderness over the GON was strongly predictive of outcome, although local anesthesia
after the injection was not. The presence or absence of medication overuse did not predict outcome. Apart from two patients with a
small patch of alopecia the injection was well tolerated. GON injection is a useful tool in some patients that provides interim relief
while other approaches are explored. It is remarkable that in all conditions in which an effect is observed the response time so much
exceeds the local anesthetic effect that the mechanism of action may well be through changes in brain nociceptive pathways.
? 2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
Keywords: Migraine; Cluster headache; Greater occipital nerve
Patients with primary headaches, such as migraine,
complain of pain affecting both the trigeminally inner-
vated anterior regions of the head and the posterior
region innervated by the C2spinal root. This is thought
to be a consequence of the overlap of processing of noci-
ceptive information at the level of the second order neu-
rons as both trigeminal afferents and C1to C2spinal
afferents have been shown to converge upon the trigem-
inal nucleus caudalis and dorsal horn nuclei of the upper
cervical spinal cord (Kerr, 1961; Goadsby and Hoskin,
1997; Bartsch and Goadsby, 2002). Occipital nerve zone
tenderness has long been associated with headache syn-
dromes and consequently local anesthetic with or with-
out a steroid has been injected into the region in an
attempt to alleviate headache (Hadden, 1940).
The greater occipital nerve (GON), which derives
most of its fibres from the C2dorsal root, is the main
sensory nerve of the occipital area (Bogduk, 1982).
The indications for the use of the GON injection are
not clear. It has been used for a range of headaches
including cervicogenic headache, occipital neuralgia,
migraine and cluster headache (Anthony, 1992, 2000;
Bovim and Sand, 1992; Peres et al., 2002). We wished
to determine the efficacy of the treatment in our practice,
examine any features which may predict a response and
0304-3959/$32.00 ? 2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
*Corresponding author. Tel.: +44 207 829 8749; fax: +44 207 813
E-mail address: firstname.lastname@example.org (P.J. Goadsby).
Pain 122 (2006) 126–129
assess whether there were any systematic complications
that patients should be warned of as part of the process
of informed consent. We had also noted that the positive
clinical effect of GON injection seemed to exceed very
considerably the half-life of the local anesthetic, and
wished to study this issue prospectively. When we noted
local alopecia, we immediately reported that effect sepa-
rately taking the view this was an important finding that
patients and doctors should be aware of (Shields et al.,
2004). Those patients are also included in the total
group in this report.
All patients attending the National Hospital for Neurology
and Neurosurgery, London, for a GON injection over a 16-
month period were assessed. Two clinicians were involved in
administering the injections throughout the period. The injec-
tion was offered to patients with various forms of primary
Chronic Daily Headache, meaning headache on 15 days or
more per month (Welch and Goadsby, 2002), that was relative-
ly treatment-refractory. For our patients this meant failing to
respond to at least three preventive classes for a reasonable tri-
al of treatment using standard approaches (Matharu et al.,
2003; Lance and Goadsby, 2005).
Patients were asked to fill in a headache diary for one week
prior to and four weeks following the injection. On the day of
the injection it was noted whether there was any tenderness in
the region of the greater occipital nerve and what medications
the patients were taking, in particular whether there was any
medication overuse. This is defined by the International Head-
ache Society (Headache Classification Committee of The Inter-
national Headache Society, 2004) as ‘‘the use of a triptan or
ergotamine on 10 days or more per month and/or the use of
analgesics on 15 days or more per month for longer than three
A mixture of 3 ml of 2% lidocaine and 80 mg of methyl-
prednisolone was injected 1–2 cm below the midpoint between
the occipital tubercle and mastoid process unilaterally in all
cases. The injection site was then massaged to spread the solu-
tion. The side of injection was determined by clinical symp-
toms. Sensation to pinprick was noted immediately before
and 20 min following the injection. Patients were followed up
with a telephone call after four weeks and asked to return their
A total of 116 injections were recorded with 101
patients. The breakdown of headache diagnosis (Head-
ache Classification Committee of The International
Headache Society, 2004) is shown with caveat that
migraine indicates definite (1.1) and probable migraine
(1.6; Table 1). The four main groups were migraine
(49%), cluster headache (19%), new daily persistent
headache (14%) and hemicrania continua (9%).
into complete response (pain free), partial response
(reduction in severity or frequency of headache by
>30%), or no response. Patients were informed that the
able. Of the 116 injections, 62 (53%) showed some
response in the form of pain relief. Twenty-six (22%)
resulted in a complete response and 36 (31%) in a partial
response (Table 1).
The mean latency of response was 2 days. The mean
duration of complete response was 20 days and the
median was 7 days (range 1–90 days). The mean dura-
tion of partial response was 45 days and the median
was 20 days (range 3–420 days).
Looking at the migraine group separately, the mean
duration of complete response in the group was 9 days
with a median of 6-day response. The mean and median
duration of partial response was 61 and 30 days,
3.1.2. Cluster headache
In the cluster headache group, the mean duration of
complete response was 17 days with a median of 12
days. The mean and median partial response was 52
and 21 days, respectively.
Response of various headache types in 101 patients to greater occipital nerve injections (n = 116)
Cluster headacheNDPH HC Other
Number of patients
Number of injections
Number with complete response
Number with partial response
Abbreviations: HC, hemicrania continua; NDPH, new daily persistent headache.
Other: SUNCT-3, Chronic Paroxysmal Hemicrania-3, Chronic post-traumatic headache 2, Cervicogenic headache-2, Low CSF volume headache-1.
aMigraine: International Headache Society 1.1 and 1.6 (Headache Classification Committee of The International Headache Society, 2004).
bSUNCT, 1; CPH, 1.
cCervicogenic, 2; post-traumatic, 2; low CSF pressure, 1.
S.K. Afridi et al. / Pain 122 (2006) 126–129
3.2. Predictive factors
Decrease to pin-prick sensation in the distribution
of the GON occurred following 12 of the 23 injec-
tions which led to a complete response and 20 of
the 35 with a partial response. There was no signifi-
cant association between response and presence of
anesthesia in the distribution of the GON (v2= 2,
P = 0.15).
All but one of the patients who had some response
to the injection were found to have moderate or severe
tenderness around the GON prior to injection. There
wasa significant relationship
around the GON region and response (v2= 3.8,
P = 0.05).
Thirty-one of the migraine patients were overusing
analgesics or triptans. Twenty of these had a response
to the injection. There was no significant relationship
between response to injection and medication overuse
(v2= 1.9, P = 0.17).
3.3. Adverse effects
Relatively few adverse effects were reported. One
patient had a vaso-vagal syncopal attack during the pro-
cedure. Three patients reported transient dizziness fol-
lowing the injection lasting less than an hour in two
patients but lasting 2 days in the third patient. Two
cases of alopecia around the injection site were reported
(Shields et al., 2004). A typical headache was triggered
immediately by the injection in three patients: two with
migraine and one with SUNCT. Two patients also felt
that their migraines were worse for one to two weeks fol-
lowing the injection.
response to the injection. It appears that, compared
to the other headache groups, a relatively greater pro-
portion of the cluster headache patients had a complete
response although it must be noted that the number of
cluster patients was much smaller than the number of
migraine patients. The duration of partial response
was greater than the duration of complete response
for both cluster and migraine groups. Remarkably
the time to onset of effect and the duration of effect
for all responders, irrespective of diagnosis, and the
lack of a correlation between local anesthesia and
response, each suggest that any response seen was
not simply dependent on the direct local anesthetic
effect of the injection.
Previous studies have looked at the response to an
injection of steroid and local anesthetic into the occip-
ital nerve region. Saadah and Taylor (1987) performed
injections of 1% lidocaine and 12 mg betamethasone
53% ofour patients demonstrateda
into multiple tender points in the vicinity of the occip-
ital nerves in 112 patients with a variety of headaches:
tension-type, vascular, post-infection and post-traumat-
ic. They noted a response of prolonged relief in 65% of
patients. Anthony looked
(160 mg) injection into 50 patients with ‘‘migraine with
unilateral GON irritation’’ and 86 with occipital neu-
ralgia (Anthony, 1992). Of the migraineurs 88%, and
of the occipital neuralgia patients 87%, had a headache
response for a mean duration of 32 and 31 days,
respectively. Interestingly, Anthony administered an
intramuscular injection of 160 mg methylprednisolone
and a GON injection of lidocaine to 20 migraineurs
and 20 occipital neuralgia patients in order to deter-
mine whether local anesthetic alone or systemic ste-
roids are capable of producing a response. Neither
measure produced a response for longer than 3 days.
In a later study, he found similar results for cervico-
genic headache and described a response in all 20 sub-
jects with chronic cluster headache following injection
of 160 mg methylprednisolone into the occipital nerve
with a mean duration of relief of 32 days (Anthony,
2000). Other studies have also shown a response in
cluster headache (Bigo et al., 1989; Peres et al.,
2002). Ambrosini and colleagues compared local anes-
thetic with saline or long-acting local corticosteroid in
a double-blind study in episodic (n = 16) and chronic
(n = 7) cluster headache. Ninety percent of the cortico-
steroid group and only one of the saline patients had a
reduction in headache frequency (Ambrosini et al.,
2003). These data suggest that the corticosteroid has
an important role at least in patients with cluster
Local anesthetic alone injected into GON was used as
early as 1940 with documentation of headache relief
(Hadden, 1940). More recently, Bovim used lidocaine
alone and looked at immediate pain relief which was
much more significant in cervicogenic headache than
in migraine and tension-type headache (Bovim and
Sand, 1992). Bovim also looked at the response to a sal-
ine injection into the GON in 16 subjects with cervico-
genic headache and found 12 responded but only for
20 min. Caputi demonstrated some response to bupivi-
caine injected into both GON and supraorbital nerves
repeated on alternate days (total of 5–10 injections) in
11 migraineurs (Caputi and Firetto, 1997). It is worth
noting that our group of patients included those with
intractable headache often resistant to prophylactic
4.1. Predictive factors
It is interesting to note that in our patient group
there was no significant association between response
and level of anesthesia following injection. There are
two possible explanations for this. It may be that
S.K. Afridi et al. / Pain 122 (2006) 126–129
due to the slight degree of anatomical variability we
were not injecting exactly into the GON but when
the area was rubbed following the injection the steroid
dissipated sufficiently enough to reach the GON; we
should have detected that during pin-prick testing.
Another possibility which we cannot exclude without
further placebo-controlled studies is that in those with-
out anesthesia the response is a placebo response.
However, consistent with the dissociation between
anesthesia and outcome, the effect times far exceeded
what would be predicted from the effect of the local
anesthetic. Tenderness around the region of the
GON was significantly associated with a positive
response to the injection. This suggests that tenderness
may be useful in selecting out patients who are more
likely to respond. It was also interesting to note that
analgesia or triptan overuse did not appear to have
any affect on response to the GON injection in the
migraine group. This suggests that GON injections
are a potential treatment for patients who overuse
medication and may be particularly helpful during
withdrawal of medication.
It is clear that a placebo-controlled study is
required in order to answer some of the questions
posed by our findings, and confirm and extend those
of Ambrosini and colleagues in cluster headache
(Ambrosini et al., 2003). Our practice might be
improved by the use of a nerve stimulator, as used
by Anthony (2000), to locate the GON which would
take into account the degree of anatomical variation
of the nerve. We could also consider altering the dose
of the steroid component of the injection following
results from a suitable study. It is interesting to note
that Anthony used a steroid dose twice as large as
we do. As mentioned above, we can improve our selec-
tion of patients by assessing tenderness in the region
of the GON. Our results suggest that GON injection
has a place in the management of primary headache
and its use is worthy of exploration. Moreover, the
data converge in a number of ways to suggest that
the effect of the injection is not direct but through
an alteration of nociceptive processing and neuroplas-
tic mechanisms within pathways, such as the trigemi-
nocervical relay (Bartsch and Goadsby, 2005), that
form an important substrate for each of the primary
We thank Dr. Manjit Matharu for his advice during
the planning phase of the audit and Dr. Elisabetta Cit-
tadini for her contribution.
Ambrosini A, Vandenheede M, Rossi P, Aloj F, Sauli E, Buzzi MG,
et al. Suboccipital (GON) injection with long-acting steroids in
cluster headache: a double-blind placebo-controlled study. Ceph-
Anthony M. Headache and the greater occipital nerve. Clin Neurol
Anthony M. Cervicogenic headache: prevalence and response to local
steroid therapy. Clin Exp Rheumatol 2000;18:S59–64.
Bartsch T, Goadsby PJ. Stimulation of the greater occipital nerve
induces increased central excitability of dural afferent input. Brain
Bartsch T, Goadsby PJ. Anatomy and physiology of pain referral in
primary and cervicogenic headache disorders. Headache Curr
Bigo A, Delrieu F, Bousser MG. Treatment of vascular pain of the face
by methylprednisolone injection into the area of the greater
occipital nerve: 16 cases. Rev Neurol (Paris) 1989;145:160–2.
Bogduk N. The clinical anatomy of the cervical dorsal rami. Spine
Bovim G, Sand I. Cervicogenic headache, migraine without aura and
tension-type headache. Diagnostic blockade of the greater occipital
and supraorbital nerves. Pain 1992;51:43–8.
Caputi CA, Firetto V. Therapeutic blockade of greater occipital and
supraorbital nerves in migraine patients. Headache 1997;37:174–9.
Goadsby PJ, Hoskin KL. The distribution of trigeminovascular
afferents in the nonhuman primate brain Macaca nemestrina: a
c-fos immunocytochemical study. J Anat 1997;190:367–75.
Hadden S. Neuralgic headache and facial pain. Arch Neurol Psychiatr
Headache Classification Committee of The International Headache
Society, The international classification of headache disorders (2nd
ed.), Cephalalgia 2004;24:1–160.
Kerr FWL. A mechanism to account for frontal headache in cases of
posterior fossa tumous. J Neurosurg 1961;18:605–9.
Lance JW, Goadsby PJ. Mechanism and management of head-
ache. New York: Elsevier; 2005.
Matharu MS, Boes CJ, Goadsby PJ. Management of trigeminal auto-
nomic cephalalgias and hemicrania continua. Drugs 2003;63:1637–77.
Peres MFP, Stiles MA, Siow HC, Rozen TD, Young WB, Silberstein
SD. Greater occipital nerve blockade for cluster headache. Ceph-
Saadah HA, Taylor FB. Sustained headache syndrome associated with
tender occipital nerve zones. Headache 1987;27:201–5.
Shields KG, Levy MJ, Goadsby PJ. Alopecia and cutaneous atro-
phy following greater occipital nerve infiltration. Neurology
Welch KMA, Goadsby PJ. Chronic daily headache: nosology and
pathophysiology. Curr Opin Neurol 2002;15:287–95.
S.K. Afridi et al. / Pain 122 (2006) 126–129