Greater Occipital Nerve Block: a Diagnostic Test?

Controversies in Neurology

Greater Occipital Nerve Block: A Diagnostic Test?

ften a neurosurgeon or orthopedic surgeon Are nerve blocks diagnostic? requests a diagnostic nerve block to deter- Neural blocks may be useful as an empirical way of Omine, prior to attempting any surgical pro- treating diverse head and neck pains, but such a cedure, whether a specific cervical nerve root is the response is also often used as the criterion for diagno- generator of the patient’s symptoms.1 sis.7,8 But such diagnoses, though clinically useful, are This statement from a contemporary textbook inexact and the procedure may be valid (if proven by reflects the commonly held view that in pain manage- properly designed trials) only as an empirical mode of ment nerve blocks are diagnostic. controlling pain. There is wide variability of headache syndromes Blondi rightly notes that “Occipital nerve blockade, … treated by greater occipital nerve (GON) blockade.1 The often results in a nonspecific regional blockade rather than putative mechanisms by which they might relate to the a specific nerve blockade and might result in a misidenti- JMS Pearce MD, FRCP is an GON are unclear. It seems a priori improbable that such fication of the occipital nerve as the source of pain.” And Emeritus Consultant Neurologist at the Hull Royal Infirmary. His diverse conditions as migraine (with its complex cere- he says: “occipital neuralgia is believed to arise from trau- interests are in Clinical Neurology bral and brainstem mechanisms), cluster headache, ma to or entrapment of the occipital nerve within the neck and the History of Medicine. occipital neuralgia, cervicogenic headache, whiplash or scalp, but the pain may also arise from the C2 spinal syndrome,2 and various tension type headaches, should root, C1–2, or C2–3 zygapophyseal joints or pathologic Correspondence to: either share a common aetiological mechanism or be change within the posterior cranial fossa.” If its source is JMS Pearce, responsive to the same treatment of a peripheral nerve. the nerve roots, how can it be rationally considered to 304 Beverley Road, be a neuralgia of the occipital nerve? Anlaby, Anatomy Despite many published studies, the diagnostic utility East Yorks, HU10 7BG, UK. The GON is composed of the medial fibres from the of employing greater occipital nerve (GON) blockade in Email: [email protected] dorsal ramus of the second cervical nerve. The ventral a variety of headaches and neck strains is unproven. ramus of C2 also contributes to the lesser occipital Many trials contain small numbers. The physician nerve and innervates deeper structures (periosteum of administering the injection in many trials is not blind- the occiput, vertebrae, etc.). In rats, a population of ed to the treatment. Follow-up assessment is commonly neurones of the dorsal horn at C2 shows convergent at about four weeks, too brief a period may have detect- input from both dura and cervical skin and muscle tered significant differences in outcome. The local anaes- ritories, suggesting a functional continuum between the thetic or steroid used, and the doses vary and are com- trigeminal nucleus caudalis and upper cervical seg- monly chosen empirically. Controls are often omitted or ments involved in cranial nociception. GON stimula- poorly matched. And, interpretation is confounded by tion in rats facilitates dural stimulation, implying a cen- subjective criteria of pain relief and marked variation of tral mechanism at the second order neurone.3 C2/C3 techniques. There are therefore, several unresolved blockade is claimed to produce benefit of comparable issues concerning both rationale, claimed benefits, and order to GON blockade and both are said to be effective techniques. in the diagnosis and treatment of cervicogenic Ashkenazi and co-workers report, “The rationale of headache.4 GON blockade for the treatment of headache is based on The literature fails to incriminate specific anatomical the anatomical connections between trigeminal and upper structures as the source of cervicogenic pains. Very sim- cervical sensory fibres at the level of the trigeminal nucle- ilar diagnoses invoke structures such as nerve roots, us caudalis.”9,10 But is this alone a diagnostic foundation individual peripheral nerves, bony structures, and the or a mechanism sufficient to explain such diverse head non-specific cervicogenic pain/headache. For example: pains? “Diagnostic anesthetic blockade for the evaluation of The methodologies of some of the studies are limited cervicogenic headache can be directed to several anatomic by lack of a standardised treatment protocol or by a ret- structures such as the greater occipital nerve (dorsal ramus rospective design. In migraine, for example, improve- C2), lesser occipital nerve, atlanto-occipital joint, ment has been reported after GON blockade, but also atlantoaxial joint, C2 or C3 spinal nerve, third occipital after prophylactic drugs, and injected botulinum A nerve (dorsal ramus C3), zygapophyseal joint(s) or inter- toxin.11 There are claimed to be four major ‘trigger vertebral discs based on the clinical characteristics of the points’ along the course of several peripheral nerves that pain and findings of the physical examination.”5 may cause migraine headaches, which are therefore also The authors state fluoroscopic or interventional MRI- treated by injection.12 Among the peripheral nerves the guided blockade may be necessary, to assure specific greater occipital nerve has become a favourite target of localisation of the pain source; yet they fail to present needle wielders. evidence that such measures do inculpate the actual Structures said to be involved in the pathogenesis of source of pain. The opposing view (which I share) is occipital headache include the aponeurotic attachments that of Silverman: of the trapezius and semispinalis capitis muscles to the “there are no diagnostic imaging techniques of the cer- occipital bone, and entrapment of the GON within vical spine and associated structures that can determine these aponeuroses, causing symptoms of ‘occipital neu- the exact source of pain.”6 ralgia’.13 Whereas cervicogenic headache is a useful clin-

In respect of diagnosis, the current evidence appraised suggests that the use of nerve blocks as the defining or pathogenetic criterion is both unsound and unreliable

ACNR • VOLUME 8 NUMBER 3 • JULY/AUGUST 2008 I 15 Controversies in Neurology

trapezius muscle was penetrated by the Table 1: Variation of anatomical sites for greater occipital nerve. GON in 45% of cases, the semispinalis Study Vertical location of GON (cm) Lateral location of GON (cm) muscle of the head was penetrated in 90% 1. Mosser et al. 3cm below EOP 1.5cm from midline of cases, and the inferior oblique muscle of head in 7.5% of cases. Macroscopic find- 2. Loukas et al. 2cm below EOP 2cm from EOP ings of possible compression were made in 3. Natsis et al. The site where the semispinalis capitis is pierced by the GON 11 cases (27.5%), “indicating that nerve 4. Bovim et al. no anatomical landmark given for injection compression per se may be of minor impor- 5. Becser et al. Along intermastoid line 0.5 to 2.8cm from midline tance since it seems to exist in the absence of 6. Tubbs et al. 2cm above intermastoid line 4cm lateral to EOP headache.” 5. Becser, Bovim and Sjaastad reported EOP = external occipital protuberance. topography shown by dissection and care- ful measurements of 10 embalmed cadavers.18 A great variability in nerve topogra- ical description,14 it is not a diagnosis that trace the normal course of the greater phy was seen interindividually and intrain- accurately inculpates the pathogenic struc- occipital nerve from the semispinalis mus- dividually. The greater occipital nerve tures involved, nor the mechanism of pain. cle penetration to the superior nuchal line. ascended between 5mm and 28mm from The International Headache Society (IHS) Standardised measurements were per- the midline along the intermastoid line. proposed diagnostic criteria15 for cervicogenic formed on 14 specimens to determine the The minor occipital nerve was found headache (11.2.1) are: location of the emergence of the nerve between 32mm and 90mm from the mid- 1. Pain referred from a source in the neck and using the midline and occipital protuber- line along the same landmark. In most perceived in one or more regions of the ance as landmarks. The location of emer- cases, both the GON and the minor occip- head and/or face. gence was determined to be at a point cen- ital nerve pierced the aponeurosis after 2. Clinical, laboratory and/or imaging evi- tered approximately 3cm below the occipi- branching. Thirteen GONs and eight dence of a disorder or lesion within the tal protuberance and 1.5cm lateral to the minor occipital nerves also were embed- cervical spine or soft tissues of the neck, midline.12 ded in this tissue. Twelve of the 20 GONs known to be, or generally accepted as, a 2. Loukas et al.13 examined the course and formed a rich network around the occipi- valid cause of headache. distribution of GON and its relation to the tal artery. Importantly they commented: 3. Evidence that the pain can be attributed to aponeuroses of the trapezius and semi- “anatomic structures with an imminent risk the neck disorder or lesion, based on either spinalis capitis in 100 formalin-fixed adult of causing entrapment were not observed. … clinical signs that implicate a source of cadavers. The greater occipital nerve was results suggest that optimal locations for pain in the neck or abolition of headache located at a mean distance of 3.8cm (range blockade techniques should be reconsidered.” following diagnostic nerve block. 1.5-7.5cm) lateral to a vertical line through 6. Tubbs and colleagues noted the surprising 4. Pain resolving within three months after the external occipital protuberance and the lack of surgical landmarks in the literature successful treatment of causative disorder spinous processes of the cervical vertebrae. for avoiding the cutaneous nerves in this or lesion. It was also located approximately 41% of region. The GON was found to lie at a Interestingly, cervical spondylosis is NOT the distance along the intermastoid line mean distance of 4cm lateral to the EOP. accepted as a valid cause. Taken in turn, these (medial to a mastoid process) and 22% of On all but three sides, a small medial criteria don’t confront these fundamental the distance between the external occipital branch was found that ran medially from issues. protuberance and the mastoid process. The the GON to the 3rd occipital nerve approx- 1. There is no hard evidence, only imprecise location of GON for anaesthesia or any imately 1cm superior to a horizontal line inference that the pain can be attributed to other neurosurgical procedure has been drawn through the EOP. The GON was the neck disorder or lesion. established as one thumb's breadth lateral found to pierce the semispinalis capitis 2. Disorders within the cervical spine or soft to the external occipital protuberance muscle on average 2cm above the inter- tissues of the neck, known to be, or gener- (2cm laterally) and approximately at the mastoid line, and to divide into medial and ally accepted as, a valid cause of headache base of the thumb nail (2cm inferior)13 lateral branches 0.5cm superior to the is an all-embracing and unproven general- This was the first study proposing land- EOP.19 This fits with the observation that it isation that does not indicate the primary marks in relation to anthropometric meas- may be technically difficult to block the pathology. urements. greater occipital nerve without also block- 3. Since there are no conclusive clinical signs 3. Natsis and colleagues reported the course ing the third occipital nerve and some of that prove a source of pain in the neck, this and the diameter of the GON in 40 cadav- the fibres of the semispinalis…20 is spurious; and, abolition of headache fol- ers.17 In three cases, the GON split into two These considerable differences in proposed lowing nerve block is only suggestive, not branches before piercing the trapezius landmarks for the GON are reflected in the diagnostic. Clinical signs or pain relief muscle aponeurosis (TMA) and reunited diversity of proposed and tried treatment after nerve blocks do not constitute attrib- after having passed the TMA, and it sites. Cervical epidural steroid injections are utable proof of causation. pierced the obliquus capitis inferior mus- used in patients with spondylosis,21 but are 4. Pain resolving within three months after cle in another three cases. The GON and largely ineffective.22 Greater and lesser occipi- successful treatment of causative disorder the lesser occipital nerve reunited at the tal nerve blockade may provide temporary, or lesion, illogically and falsely assumes level of the occiput in 80% of the speci- but substantial, pain relief in some cases of that response to treatment is proof of cau- mens. The nerve became wider towards the similar efficacy to blockade of the C2 and C3 sation. Few trials extend to three months periphery. This may be relevant to entrap- nerves.23 Blockade, neurolysis, nerve resection, or beyond. ment of the nerve. In three cases, the GON rhizotomy, and decompressive techniques A further difficulty is that the techniques split into two branches before piercing the have all been employed, each with claims of employed and for which success is claimed TMA. Anaesthetic blockade of the GON success. Bogduk’s team reported24 the efficacy differ from one series to another. for diagnosis and therapy was best aimed of percutaneous radiofrequency medial at the site where the semispinalis capitis is branch neurotomy in the treatment of chron- Anatomical variability pierced by the GON. ic neck pain in 71% of patients, but the pre- An important but neglected factor is the 4. An autopsy study by Bovim et al. on 20 cise nature and anatomical basis of the symp- anatomical variability (Table 1) of the greater cases without known headaches showed a toms are not thereby clarified. What is the occipital nerve.16 marked variation in the relation between basis in the unrelieved substantial 29%? 1. Mosser et al. dissected 20 cadaver heads to the GON and nuchal muscles.16 The Lieppman (1980) described25 164 patients

16 I ACNR • VOLUME 8 NUMBER 3 • JULY/AUGUST 2008 Controversies in Neurology

with comparable symptoms in whom he diag- should therefore be understood as a homoge- uine, validated phenomenon, which may be nosed ‘occipital neuralgia’ and found a high neous but also unspecific pattern of reaction.”28 organically founded in regionally specific cure rate after occipital subcutaneous injec- The placebo effect29 of injections is often changes in brain function; for example, dor- tions of lidocaine; however, he also found a underestimated or neglected. To the scientist, sal-cortical increases and limbic-paralimbic 50% cure rate in patients given occipital sub- a major problem is that placebos influence decreases in glucose metabolism demonstrat- cutaneous injections of saline. A significant patient outcomes after any treatment, includ- ed in a trial of antidepressant vs placebo.33 contributory placebo effect is probable, as in ing surgery. The well-known extent of placebo all pain syndromes. effects in acute head pains was illustrated in Conclusion Harden and coworkers’ trial which concluded: The lack of specificity of greater occipital or Discussion This profound reduction observed after other peripheral nerve, and cervical root Thus, there are problems in accepting GON administration of a placebo prevented accurate blockade in the treatment of diverse and related neural sites subjected to blockades evaluation of the effects of [ketorolac]. The headaches, (including migraine, cluster as a rational as opposed to empirical method: placebo response must be considered in the headache, cervicogenic and occipital pains), 1. Anatomical structures threatening neural design of future trials using intramuscular med- though achieving variable empirical therapeu- entrapment are seldom observed.16 ications in the acute intervention of headache tic success, is a continuing cause of diagnostic 2. Nerve compression per se may be of minor crises. In addition, the use of a standard anal- confusion and wooly thinking. Although vari- importance since it can exist in the absence gesic is necessary to demonstrate both assay sen- ous structures in the neck probably contribute of headache.16 sitivity and magnitude of response to placebo.30 to several patterns of headaches and neck 3. There are many and considerable anatom- Placebo effects and spontaneous remissions pains, they have not to-date been adequately ical variations in the nerve so that consis- can cause apparently good results that are defined. More rigorous studies are needed to tent surgical landmarks in this region are falsely attributed to the efficacy of any treat- identify the several anatomical structure and surprisingly lacking,12,13,16-18 and the injected ment claims.31 As one example, Peres et al. the physiological mechanisms that underlie substance may not have affected the GON. treated 14 cluster headache patients with these ill-defined symptoms. In respect of 4. Clinical techniques of localising the nerve greater occipital nerve block. Four patients treatment, these considerations agree with the are variable and imprecise.18,19 Tenderness (28.5%) had a good response, five (35.7%) a conclusions of Bogduk34 that “The available and evoked pain on palpation are notori- moderate, and five (35.7%) no response. The evidence from the small number of case series ously unreliable. authors concluded GON blockade is a thera- and retrospective studies published in the peer- An apparent response to neural blockade is peutic option for the transitional treatment of reviewed literature is insufficient to conclude often used as one of several criteria for the cluster headache,32 yet their results are entirely that either local injection therapy or surgery is diagnosis of cervicogenic and other consistent with a placebo effect. It is suggested an effective treatment for occipital neuralgia or headaches.26,27 but such a diagnosis, though of that the placebo response may contribute sig- cervicogenic headache.” arguable clinical value, is itself inexact and nificantly to the apparent successes of nerve In respect of diagnosis, the current evi- justifies the procedure only as an empirical blocks, but does not necessarily account for dence appraised suggests that the use of nerve mode of controlling pain. As Pollman and the relief of pain in all cases. Placebo effect blocks as the defining or pathogenetic criteri- colleagues commented, “cervicogenic headache does not imply psychogenesis, but is a gen- on is both unsound and unreliable.

References 1. Warfield CA, Bajwa ZH. Principles and Practice of Pain 13. Loukas M, El-Sedfy A, Tubbs RS, Louis RG Jr, 23. Inan N, Ceyhan A, Inan L, Kavaklioglu O, Alptekin A, Medicine. 2nd edn New York, McGraw Hill, 2004:268-9. Wartmann CH, Curry B, Jordan R. Identification of Unal. C2/C3 nerve blocks and greater occipital nerve 2. Pearce JMS. Headaches in the whiplash syndrome. Spinal greater occipital nerve landmarks for the treatment of block in cervicogenic headache treatment. Funct Neurol. Cord 2001;39:228-33. occipital neuralgia. Folia Morphol (Warsz). 2001;16:239-43. 2006;65:337-42. 3. Bartsch T, Goadsby PJ. Stimulation of the greater occipi- 24. McDonald GJ, Lord SM, Bogduk N. Long-term follow- tal nerve induces increased central excitability of dural 14. Pearce JMS. Cervicogenic Headache : a personal view. up of patients treated with cervical radiofrequency neuro- afferent input. Brain 2002;125:1496-509. Cephalalgia 1995;15:463-9. tomy for chronic neck pain. Neurosurgery. 1999;45:61-7. 4. Inan N, Ceyhan A, Inan L, Kavaklioglu O, Alptekin A, 15. The International Classification of Headache Disorders, 25. Lieppman ME. Occipital neuralgia: The ophthalmic Unal N. C 2/C 3 Nerve Blocks and Greater Occipital 2nd edition Cephalalgia 2004;24suppl 1:1-160. entity and its treatment. Ophthalmology 1980;87:94-5. Nerve Block In Cervicogenic Headache Treatment. 16. Bovim G, Bonamico L, Fredriksen TA, Lindboe CF, 26. Sjaastad O, Fredriksen TA, Pfaffenrath V. Cervicogenic Functional Neurology 2001;16:239-43. Stolt-Nielsen A, Sjaastad O. Topographic variations in headache: Diagnostic criteria. Headache 1998;38:442-5. the peripheral course of the greater occipital nerve. 5. Blondi DM. Cervicogenic Headache: A Review of 27. Bogduk N. The neck and headaches. Neurol Clin N Am. Diagnostic and Treatment Strategies. J. American Autopsy study with clinical correlations. Spine 1991;16:475-8. 2004;22:151-71. Osteopathic Association 2005;105(4)suppl:16-22. 17. Natsis K, Baraliakos X, Appell HJ, Tsikaras P, Gigis I, 28. Pollmann W, Keidel M, Pfaffenrath V. Headache and the 6. Silverman SB. Cervicogenic Headache: Interventional, Koebke J. The course of the greater occipital nerve in the cervical spine: a critical review. Cephalalgia Anesthetic, and Ablative Treatment. Current Pain and suboccipital region: a proposal for setting landmarks for 1997;17:801-16. Headache Reports 2002;6(4):308-14. local anesthesia in patients with occipital neuralgia. J 29. Bouckoms A, Hackett TP. The pain patient. 7. Sjaastad O, Fredriksen TA, Pfaffenrath V. Cervicogenic Clin Anat. 2006;19:332-6. In:Massachusetts General Hospital Handbook of headache: Diagnostic criteria. Headache 1998;38:442-5. 18. Becser N, Bovim G, Sjaastad O. Extracranial nerves in the General Hospital Psychiatry. ed NH Cassem. 3rd edn. 8. Bogduk N. The neck and headaches. Neurol Clin N Am. posterior part of the head. Anatomic variations and their St Lois:Mosby 1991:44-5. 2004;22:151-71. possible clinical significance. Spine 1998:1;23:1435-41. 30. Harden RN, Gracely RH, Carter T, Warner G. The 9. Ashkenazi A, Matro R, Shaw JW, Abbas MA, Silberstein 19. Tubbs RS, Salter EG, Wellons JC, Blount JP, Oakes WJ. Placebo Effect in Acute Headache Management: SD. Greater occipital nerve block using local anaesthetics Landmarks for the identification of the cutaneous nerves Ketorolac, Meperidine, and Saline in the Emergency alone or with triamcinolone for transformed migraine: a of the occiput and nuchal regions. Clin Anat. Department. Headache 1996; The Journal of Head and randomised comparative study. J Neurol Neurosurg 2007;20:235-8. Face Pain 36;352–6. Psychiatry. 2008 Apr;79(4):415-7. Epub 2007 Aug 6. 20. Gawel MJ, Rothbart PJ. Occipital nerve block in the 31. Pearce JMS. The placebo enigma. QJM. 1995;88:215-20. 10. Bartsch T, Goadsby PJ. Stimulation of the greater occipi- management of headache and cervical pain. Cephalalgia tal nerve induces increased central excitability of dural 1992;12:9-13. 32. Peres MFP, Stiles MA, Siow HC, Rozen TD, Young WB, Silberstein SD. Greater occipital nerve blockade for clus- afferent input. Brain 2002;125:1496-1509. 21. Reale C, Turkiewicz AM, Reale CA, Stabile S, ter headache. Cephalalgia 2002;22:520-22. 11. Silberstein S, Mathew N, Saper J, Jenkins S. Botulinum Borgonuovo P, Apponi F. Epidural steroids as a pharma- toxin type A as a migraine preventive treatment. cologic approach. Clin Exp Rheumatol 2000;18(Suppl 33. Mayberg HS, Silva JA, Brannan SK, Tekell JL, Mahurin Headache 2000;40:445-50. 19):S65 -S66. RK, McGinnis S, Jerabek PA. The functional neu- roanatomy of the placebo effect. Am J Psychiatry 12. Mosser SW, Guyuron B, Janis JE, Rohrich RJ. The 22. Barnsley L, Lord SM, Wallis BJ, Bogduk N. Lack of effect anatomy of the greater occipital nerve: implications for of intraarticular corticosteroids for chronic pain in the 2002;159:728-37. the etiology of migraine headaches. Plast Reconstr Surg. cervical zygapophyseal joints. N Engl J Med 34. Bogduk N. The neck and headaches. Neurol Clin N Am. 2004;113:693-7. 1994;330:1047-105. 2004;22:151-71.

ACNR • VOLUME 8 NUMBER 3 • JULY/AUGUST 2008 I 17