Endoscopy-Negative, Computed Tomography-Negative Facial Pain in a Nasal Clinic

The LaryngoscopeLippincott Williams & Wilkins, Inc., Philadelphia© 2001 The American Laryngological,Rhinological and Otological Society, Inc.

Endoscopy-Negative, ComputedTomography-Negative Facial Pain in aNasal Clinic

Brian West, FRCS; Nick S. Jones, MD, FRCS

Objectives: To establish the etiology of facial painin individuals attending the nasal clinic of the De-partment of Otorhinolaryngology, University Hospi-tal, Nottingham, U.K., with normal nasal endoscopyand computed tomography of the paranasal sinuses.Study Design: A retrospective analysis of a cohort of973 patients consecutively presenting to the nasalclinic with symptoms of rhinosinusitis and/or facialpain. Methods: We reviewed the case notes of 973 con-secutive patients who presented to the nasal clinicwith either symptoms of rhinosinusitis or facial pain,and in particular 101 who had facial pain without anyobjective evidence of nasal disease as detected by na-sal endoscopy or computed tomography. The diagno-sis was based on the outcome and response to treat-ment after a mean of 2 years 2 months. Results: Onehundred one patients had pain as a predominantsymptom with normal nasal endoscopy and computedtomography of the paranasal sinuses. None of thesepatients responded to either medical or surgicaltreatment for what some workers have hypothesizedcould be “occult” sinonasal disease. A neurologicaldiagnosis was made in 99 patients. Eighty patientsreceived successful medical treatment for “neurolog-ical” diagnoses, 8 patients experienced spontaneousresolution of their symptoms, 7 failed to respond toany treatment modality, 2 were lost to follow-up, and2 refused any treatment. Conclusion: The majority ofpatients presenting to a rhinologic clinic with facialpain and no objective evidence of sinus disease, asdetected by endoscopy and computed tomography,responded well to neurological treatment and surgi-cal intervention was unnecessary. These patientsshould receive a trial of medical therapy, such as low-dose amitriptyline for 6 weeks in the first instance,before any surgical intervention is considered. KeyWords: Facial pain, CT, endoscopy, nasal, sinusitis.

Laryngoscope, 111:581–586, 2001

INTRODUCTIONFacial pain is a common reason for referral to an

otorhinolaryngology clinic. Patients’ expectations are of-ten tempered by a prior diagnosis of “sinusitis” by them-selves or their primary care physician as the cause of theirfacial pain. Apart from sinus infection as part of the dif-ferential diagnosis of facial pain, some workers have hy-pothesized other rhinological causes in the etiology of fa-cial pain. In 1908 Sluder first described “sphenopalatineneuralgia” as a cause of an ipsilateral, boring, burningfacial pain beginning along the lateral side of the nose andin the eye, forehead, orbit, temporal, and mastoid regions,constant or paroxysmal, often occurring several times in24 hours and associated with lacrimation, rhinorrhea, in-jected conjunctiva, and sometimes involving the cheek.1

He said that cocainization of the sphenopalatine ganglionstopped the pain and postulated that involvement of thesphenopalatine ganglion in an inflammatory processwould lead to referred pain to the face. Since then, thesymptom complex, which he described, has been found tobe the result of a cluster headache.

Sluder also described a different type of frontal painthat he attributed to “vacuum” headaches, which couldproduce ocular symptoms. These symptoms were not as-sociated with pus or a contact point, and the pain wasrelieved by applying astringents in the area of the middlemeatus. In resistant cases removal of the middle turbinatehelped, but the pain returned in 2 to 3 years.2 In 1943McAuliffe described cases in which stimulation of variouspoints within the nasal cavity and paranasal sinusescaused referred facial pain.3 These findings have beenused to argue for theories which state that mucosal con-tact points within the nasal cavity can result in facialpain, although these studies did not describe contactpoint-induced facial pain.4,5 Stammberger and Wolf pos-tulated that variations in the anatomy of the nasal cavityresult in mucus stasis, infection, and ultimately facialpain.5 They also stated that mucosal contact points mightresult in the release of a neurotransmitter peptide calledsubstance P, a recognized neurotransmitter in nociceptivefibers. Other authors have embraced these concepts toexplain how anatomical variants, such as a concha bul-losa6 and pneumatized superior turbinate,7 might produce

From the Department of Otorhinolaryngology, University Hospital,Nottingham, U.K.

Editor’s Note: This Manuscript was accepted for publication January25, 2001.

Send Correspondence to Professor Nick S. Jones, Department ofOtorhinolaryngology, University Hospital, Nottingham NG7 2UH, U.K.E-mail: [email protected]

Laryngoscope 111: April 2001 West and Jones: Endoscopy-Negative, CT-Negative Facial Pain

581

facial pain. Cook et al.8 advocated operating on patientswith facial pain with no computed tomography (CT) evi-dence of sinus pathology. However, the rationale behindthis was not adequately explained; although 12 of the 18patients who underwent surgery in their series had asignificant reduction in their pain severity, no patientachieved complete elimination of their symptoms.

The evidence in support of these rhinological theoriesof facial pain is based around either anatomical theories,reports that are largely anecdotal with no scientific expla-nation or proof by objective means, or a controlled study. Itis possible that most of the series reported in the litera-ture represent coincidental anatomical variations in pa-tients with facial pain. They do not prove that the ana-tomical variations or contact points cause facial pain. Thatsurgery temporarily can help a proportion of patients withcontact points may be influenced by the effect of cognitivedissonance,9 or possibly an alteration to the neurologicalpathways to and in the caudal nucleus of the trigeminalnerve.10,11 Our experience has been that after a follow-upperiod of over 1 year, patients’ pain returns in the majorityof these patients. We present our experience of examiningpatients presenting to a rhinology clinic, and describethose patients with facial pain, a normal examination, anda normal CT scan, and the results of their treatment andfollow-up.

METHODSThe case notes of 973 patients who presented with symp-

toms of rhinosinusitis or facial pain were reviewed retrospec-tively. The patients who had normal nasal endoscopy and a nor-mal CT scan are presented in detail. A normal CT was one inwhich no mucosal thickening or obstruction of the sinus ostia wasdetected. Note was made of their history, clinical findings, inves-tigations, diagnosis, and treatment prior to being seen in thenasal clinic of the Department of Otorhinolaryngology, Univer-sity Hospital, Nottingham, U.K.. The study examined the rele-

vant clinical findings in light of the diagnosis made after treat-ment and follow-up. The mean follow-up for patients with painwas 2 years 2 months; for the aforementioned reasons, this is animportant aspect of a study on facial pain.

RESULTSOf the 973 consecutive patients presenting to this

clinic, 564 (58%) had no pain and 409 (42%) experiencedfacial and/or head pain or pressure. Patients having facialpain, headache, or symptoms of rhinosinusitis were in-cluded. Patients were excluded if they had catarrh or post-nasal drip as their only symptom, nasal deformity, septaldeviation without rhinosinusitis, nose bleeds, rhinitis medi-camentosa, benign and malignant tumors, valve collapse,olfactory dysfunction without rhinosinusitis, and nasal gran-ulomatous disorders without pain and vestibulitis.

A breakdown of the primary diagnoses in the wholecohort of 973 patients after a mean of 2 years 2 months isshown in Table I. The 17 patients in the “others” categoryconsisted of 5 patients with trigeminal neuralgia, 2 withChurg-Strauss syndrome, 2 with postherpetic neuralgia, 2with Wegener’s granulomatosis, 1 with a thrombosed su-pratrochlear hemangioma, 1 with chronic fatigue syn-drome, 1 with a pseudotumor of the orbit, 1 with skullbase meningioma, and 2 with dental abscesses.

Of the 973 patients, 679 (70%) had evidence of si-nonasal disease as detected by rhinoscopy, endoscopy, orCT, and 294 (30%) had no evidence of sinonasal disease.Of the 679 patients with evidence of sinonasal disease,including allergic and idiopathic rhinitis, only 119 (18%)patients also had pain. This group of 119 patients com-prised 76 patients with pain attributable to sinonasaldisease and 43 patients with facial pain that was found tobe incidental. These 43 patients failed to respond to eithermedical or surgical treatment of their nasal disease, butthen responded to medical treatment directed at a neuro-logical cause of their facial pain, and they responded to theappropriate medication. Therefore, of the 697 patientswith sinonasal disease, only 11% had pain attributable tonasal disease. The remaining 7% had a neurological causewith incidental nasal symptoms.

Of the 409 patients with facial pain, 101 (25%) hadnormal nasal endoscopy and CT scan findings. The termi-nology used to classify facial pain is that of the Interna-tional Headache Society,12 with the exception of midfacialsegment pain, which will be described. Of the 101 patientswith normal endoscopy and CT, the most common diagno-sis was midfacial segment pain (n 5 35). This is a form oftension-type headache, which has all the same qualities ofthat pain but it affects the face and may involve thenasion, under the bridge of the nose, either side of thenose, the periorbital region, retro-orbitally, or across thecheeks. Often the forehead is also affected. It is describedas a dull ache, a feeling of pressure, or tightness. It can bechronic or episodic, and the skin and soft tissues over theforehead or cheek may be sensitive to touch. The majorityof patients with this condition respond to low-dose ami-triptyline, but may require up to 6 weeks of 10 mg, andoccasionally 20 mg, before it works. It should then becontinued for 6 months before attempting to discontinue;however, many patients need to restart the medication as

TABLE I.The Primary Diagnoses in 973 Consecutive Patients Attending aRhinology Clinic After a Mean Follow-up of 2 Years 2 Months.

Diagnosis No.

Allergic rhinitis 240

Nasal polyposis 217

Midfacial segment pain 107

Infection 98

Idiopathic rhinitis 67

Tension-type headache 66

Migraine 51

Atypical facial pain 35

Cluster headache 23

TMJ/Myofascial pain 22

Others (see text) 17

No diagnosis 14

Aspergillosis 8

Barotrauma 5

Paroxysmal hemicrania 3

See text for inclusion and exclusion criteria.TMJ 5 temporomandibular joint.


582

the pain returns. If this fails, then relief may be obtainedfrom either carbamazepine or propranolol; if they do notrespond to these, then sodium valproate or gabapentinshould be prescribed.

The next most common diagnosis was atypical facialpain (n 5 30). This is described as a deep, ill-defined pain;no organic cause for the pain can be found, and it oftencrosses recognized neurological dermatomes. It is oftenunilateral. It is more common in women over the age of 40and it may alter in its location. Psychological factors playa role in most patients’ facial pain, but in atypical facialpain they often play a major role. It responds to eitheramitriptyline or carbamazepine, and it may take up to 6weeks for these drugs to have an effect.

Next in frequency was tension-type headache (n 516). It is described as dull, a feeling of pressure, or a tightfeeling. It is usually frontal, but there can be an occipitalor temporal component, and it can be chronic or episodic.It most often responds to amitriptyline, but propranolol,sodium valproate, or a change in lifestyle may result inthe successful relief of symptoms. Migraine was the nextmost common diagnosis (n 5 10). This can be either classicor common. Classic migraine is less common, although itis easier to recognize because there is not only nausea, butan aura and photopsia. The patient has a sharp, severepulsatile pain. This variant is three times more commonin women and there is often a family history. Classicmigraine responds to aspirin, if given early, or one of thetriptans. If episodes are frequent, pizotifen, propranolol,amitriptyline, nifedipine, or sodium valproate can begiven for prophylaxis. The common variant of migraine isalso described as sharp, severe, and pulsatile in natureand is often accompanied by nausea. There is, however, nopremonition or photopsia, but there is often a family his-tory. Common migraine responds to the same treatmentas classic migraine. Five patients were diagnosed as hav-ing cluster headache. This is a severe unilateral painoccurring in the frontal, ocular, cheek, and temporal re-gions. It often wakes the patient from sleep and can beaccompanied by unilateral nasal congestion, rhinorrhea,and sometimes lacrimation and injection of the eye andcheek. It occurs more commonly in men aged 30 to 40years. It responds in the acute phase to triptans or ergot-amine. Recurrent clusters may respond to prophylacticpizotifen, amitriptyline, sodium valproate, methysergide,or steroids; and resistant cases may gain benefit fromverapamil and lithium carbonate. Three patients had tem-poromandibular joint (TMJ) dysfunction and were re-ferred to a maxillofacial surgeon. One patient had trigem-inal neuralgia (TGN). TGN is an agonizing, lancinatingpain. There is often a trigger point, and it is more commonin the maxillary and mandibular divisions of the trigem-inal nerve. It occurs more commonly in women over 40years of age. It responds to either carbamazepine or phe-nytoin. This patient responded to carbamazepine. Onepatient had chronic paroxysmal hemicrania. This is anexcruciating pain occurring in women at any time of nightor day. It can affect the frontal, ocular, cheek, or temporalregions and lasts 30 minutes to 3 hours. The patient canexperience several episodes in 24 hours; and nasal conges-tion, lacrimation, and facial flushing can all be features.

Chronic paroxysmal hemicrania usually responds to indo-methacin; failing this, propranolol can be an effectivetreatment. This patient responded to amitriptyline.

Of the patients with midfacial segment pain (n 5 35),20 responded to amitriptyline, 6 responded to propranolol,1 responded to sumatriptan and propranolol, 2 settledspontaneously, 1 improved after a change in lifestyle, 2 tohomeopathy, and 1 was lost to follow-up. Of the 29 pa-tients with atypical facial pain, 23 responded to amitrip-tyline, 1 settled on carbamazepine, 2 improved spontane-ously enough to be tolerated, 2 failed to respond to any ofthe aforementioned drugs as well as gabapentin, 1 refusedtreatment, and 1 was lost to follow-up. Six of the 15patients with tension-type headaches responded to ami-triptyline, 1 patient responded to carbamazepine, 1 re-sponded to propranolol, 1 to nifedipine, and 1 to ibuprofen.Five of the 10 patients with migraine responded to imig-ran, 2 responded to amitriptyline, 2 responded to pizo-tifen, and 1 responded to propranolol. Three patients withcluster headaches responded to triptans, 1 to ergotamine,and 1 responded to pizotifen.

Seven patients failed to respond to treatment. Thesecomprised 2 with midfacial segment pain, 2 patients withtension-type headaches, 2 with atypical facial pain, and 1patient with TMJ dysfunction. In a further 8 patients, thesymptoms resolved without treatment. This group com-prised 2 patients with atypical facial pain, 2 with tension-type headache, 1 with midfacial segment pain, and in 2patients a diagnosis was not made. Two patients were lostto follow-up, 1 with atypical facial pain and 1 with midfa-cial segment pain. Two patients refused to take any med-ication, 1 with atypical facial pain and 1 with tension-typeheadaches. In two patients, no diagnosis was made. Thefirst was a 45-year-old woman who had episodic left peri-orbital swelling associated with pain, but CT was normaland her pain settled spontaneously on each occasion. Theother was a 23-year-old man with left frontal and retro-orbital pain who had a normal CT; his pain settled afterseveral months once a range of medical treatment, whichfailed to help, had been stopped. One differential diagno-sis was demyelinating disease, but his MRI and lumbarpuncture were normal.

Twenty-one patients in the subgroup had undergoneoperative procedures to treat their pain before being seenin the authors’ clinic with no long-term success, and in 5patients their pain was worse after surgery. Several pa-tients had a transient improvement in their symptomsafter surgery, only for their pain to return between 2weeks and 18 months after surgery. Eight patients hadundergone previous endoscopic sinus surgery and 4 re-ported that their pain had preceded surgery. Two patientshad had a submucous resection (SMR). Three patients hadundergone intranasal antrostomies (INA). Three patientshad had a SMR and endoscopic sinus surgery. Two pa-tients had undergone a SMR and bilateral antral wash-outs. One patient had had a Caldwell-Luc procedure. Onepatient had had a Caldwell-Luc procedure and later en-doscopic sinus surgery. One patient had had a septoplastyand intranasal antrostomies. In the “postsurgical” group,17 of the 21 patients responded to medical neurologicaltreatment, i.e., amitriptyline, triptans, and so on. One


583

patient with tension-type headaches and previous INAresolved spontaneously. The 3 remaining patients, 1 withtension-type headaches, 1 with atypical facial pain, and 1with tension-type facial pain, did not respond to a range ofmedication, including amitriptyline, propranolol, carbam-azepine, sodium valproate, triptans, and gabapentin.

In comparison with the remaining 308 patients withfacial or head pain, 175 had an abnormal endoscopy and143 an abnormal CT scan. Of the 175 patients with anabnormal endoscopy, and after a mean of 2 years 2months, 76 were thought to have pain attributable tosinonasal pathology (see Table II). Of the 76 patientswhose pain was related to sinus disease, 24 had purulentnasal polyposis, 32 purulent rhinosinusitis, 14 with nasalpolyposis (180 patients had polyps in the whole cohort), 2Churg-Strauss syndrome, and 2 Wegener’s granulomato-sis. Forty-one patients had coincidental seasonal allergicrhinitis, 18 had perennial allergic rhinitis, and 40 hadnonspecific endoscopic signs such as unilateral edema ofthe anterior end of the middle turbinate without pus inthe middle meatus. One hundred forty-three patients hadCT changes, which varied from minor mucosal thickeningin an ethmoidal cell to no aeration of any of the paranasalsinuses. It was only in the group of 76 patients, alreadymentioned, that the patient’s facial pain was attributed tosinonasal pathology. In the remaining 67 patients, theabnormal endoscopic or CT changes were thought to becoincidental, as these patients did not respond to medicalor surgical treatment directed at sinonasal disease, butdid respond to neurological treatment.

Therefore, of 101 patients with normal nasal endos-copy and CT scans, a neurological diagnosis of the cause oftheir facial pain was made in 99 and in 2 no diagnosis wasmade. Medical “neurological” treatment resulted in reso-lution in 80 patients and in a further 8 patients their painresolved spontaneously, 2 were lost to follow-up, 2 refusedtreatment, and 7 did not respond to any treatment.

A contact point, defined by persistent mucosalblanching, which remained after topical decongestion, was

found in 4% of patients with no facial pain and 4% ofpatients with facial pain. In the 18 patients with facialpain, 9 of the contact points were the result of a septalspur touching the lateral wall and 9 were the result of themiddle turbinate contacting the septum. Of the 4 withunilateral symptoms, 2 had a contact point on the otherside. After a mean follow-up of 2 years 2 months, 11 of the18 patients responded to treatment for tension-type head-aches, midfacial segment pain, migraine, or cluster head-aches, and 3 patients were better after surgery for coex-isting purulent nasal disease; of the 4 patients who had aseptal spur causing a contact point, one remainedsymptom-free at 2 years but the others’ pain returned at 3to 12 months postoperatively and then went on to respondto low-dose amitriptyline.

DISCUSSIONAnalysis of this cohort of patients with normal nasal

endoscopy and CT showed that these patients have painfrom a range of neurological causes. This group compriseda significant proportion of the patients seen in this clinicwith facial pain. In previous reports, much has been madeof nasal findings of mucosal contact points, enlarged mid-dle turbinate, and concha bullosa. However, the paperswhich advance theories to explain the association of thesefindings and facial pain have major methodological flaws,and therefore these explanations cannot be accepted asscientific fact.

Several authors have used the application of localanesthetic, nasal decongestants, or both to mucosal con-tact points as a diagnostic test to predict whether surgerywould be beneficial in these patients.6,7,13–21 Only one ofthese papers21 mentions the use of a control to validatethis method as a meaningful diagnostic test. One paperadvocated the topical application of cocaine to contactpoints as having prognostic value in selecting patientssuitable for surgery, but also offered middle turbinectomyto patients even if this test was negative.17 They claimedrelief of headache in a number of these patients who had

TABLE II.Diagnoses in 409 Patients Who Had a Headache, Facial Pain, or Pressure After a Mean of 2 Years

2 Months.

DiagnosisNormal Endoscopyand CT (n 5 101)

Abnormal Endoscopy(n 5 175)

Midfacial segment pain (see text for description) 35 31

Atypical facial pain 29 3

Tension-type headache 15 23

Migraine 10 18

Cluster headache 5 7

Temporomandibular joint dysfunction 3 10

“Others” (see text for detail) 1 4

Paroxysmal hemicrania 1 1

No diagnosis 2 6

Purulent sinusitis 0 32

Purlent and polyposis 0 24

Polyposis 0 14

Barotrauma 0 2

CT 5 computed tomography.


584

a negative cocaine test, but then excluded them from thisstudy.

Reviewing the publications in these areas reveals adisparate group made up of series of case reports. Thesemay represent patients with coincidental facial pain andan incidental nasal finding, e.g., a contact point, a conchabullosa, or a superior turbinate. The selection of patientsin these reports is not clear and the patients reported mayhave been “cherry-picked” to meet the authors’ preconcep-tions, and ignored those with contact points that had nopain and those without contact points that had pain. Sev-eral papers have suggested that CT scan of the paranasalsinuses can be useful in diagnosing “occult” sinus dis-ease.8,15,22–24 However, the value of CT scan of the para-nasal sinuses is doubtful in the diagnosis of sinus disease.Jones et al.25 showed that anatomical variants as seen inthe CT scans are as prevalent among asymptomatic con-trol subjects as patients with confirmed rhinosinusitis.They quote, for instance, concha bullosa in 23% of controlsubjects as compared with 18% of patients with rhinosi-nusitis, ostiomeatal narrowing in 41% of control subjectsas compared with 35% of patients with disease, and septaldeviation in 24% in both groups. These findings are con-firmed by Bhattacharyya et al.26

Cook et al.8 stated that in a selected group of patientswith a normal CT scan and nasal endoscopy, endoscopicsinus surgery can help symptoms of facial pain. This pa-per has had a significant impact on the practice of rhinol-ogy, but the evidence for this approach is at best limited.The authors stated that all patients in this group, com-prising 18 patients, underwent comprehensive medicalmanagement before surgery. They had not, however, con-sidered the possibility of a non-rhinological origin for thepain, such as is advocated in this report. The measuresused to assess outcome were not validated. The most tell-ing statement was that complete elimination of symptomswas not accomplished in any patient. If sinus disease werethe cause of these symptoms, then surgery would be ex-pected to result in a more complete resolution of theirsymptoms, at least between upper respiratory tract infec-tions. We recognize that acute sinusitis can occur withcomplete resolution of mucosal and CT changes, and pa-tients with a genuine infective episode have severe painand pyrexia. If their symptoms are less clear, we askpatients to attend when they are symptomatic to haveendoscopy and CT if necessary to help confirm or refutethe diagnosis.

We accept that to validate a diagnosis of facial pain isdifficult, because there is no diagnostic test. The result oftreatment can be influenced by the placebo effect, cogni-tive dissonance, spontaneous resolution of disease, and analteration in the inhibitory effect of supraspinal fi-bers.11,27 However, the majority of their effects do notpersist after a year and therefore a mean follow-up of 2years 2 months was chosen in this study. Failure of sur-gery does not necessarily mean that the patient’s pain isneurogenic in origin. However, if the ostia of the sinusesare patent and the mucosa healthy, and the patient thengoes on to respond to neurogenic medication, it makes itmore probable that sinonasal disease or an anatomic vari-ation is not the cause. The analgesic effects of amitripty-

line mainly occur over doses of 75 mg and within a fewdays, whereas the effects of amitriptyline in midfacialsegment pain act after 4 to 6 weeks and at a much lowerdose.

We hypothesize that in patients who do undergo sur-gery, as had been reported in other series which reportsome improvement after the removal of a contact point oropening sinus ostia when there is no objective evidence ofdisease,7,8,13–21 that the neurological stimulus of surgerymay be responsible for neurological changes or neuroplas-ticity which result in a transient alteration in the input tothe caudal nucleus of trigeminal which helps a proportionof these patients.11,28 Unfortunately, surgery only helpstemporarily in a minority of patients with tension-typeheadaches or midfacial segment pain, and it can maketheir pain worse in approximately 25% of them. In themajority of patients, it provides no relief. Follow-up re-veals that in the vast majority of these patients, painreturns after surgery.

The vascular–supraspinal–myogenic model has beenevoked to explain the genesis of tension-type headaches.10

They postulated that there is convergence of somaticovis-ceral afferent nerves from blood vessels, both intra- andextracranial, and pericranial muscles in the trigeminalcaudal nucleus. At this point, the sensory information canbe modulated by supraspinal afferents, for instance, inresponse to stress and other psychological factors. Painresults from nociceptive activation and sensitization fromnon-nociceptive stimuli. The type of pain a patient expe-riences depends on the balance of inputs from these threesources. We think there is a similar entity, namely mid-facial segment pain, which has all the characteristics oftension-type headaches, although its distribution can bethe nasion and/or under the eyebrows, behind the eyes,the cheeks and either side of the nasal bones, or under thenasal bridge. Midfacial segment pain often overlaps withtension-type headaches and did so in this series in 82% ofpatients with midfacial segment pain. Thus, tension-typeheadaches and midfacial segment pain may result fromstrong myofascial input, weak vascular input, and an in-adequate modulation from supraspinal fibers. If the vas-cular input were greater, then migraine would be expectedto also be present.

This report, in the light of recent work on the patho-genesis of facial pain and headache,29,30 and clinical stud-ies which have reported up to 38% of patients havingpersistent facial pain despite the lack of evidence of si-nonasal pathology after endoscopic sinus surgery,31 high-lights the need for surgeons to consider the neurologicalcauses of facial pain. The majority of patients presentingwith facial pain and no objective evidence of sinus diseaseon endoscopy and CT respond well to neurological treat-ment, and surgical intervention is often unnecessary. Themost common cause of non-rhinologic facial pain is mid-facial segment pain, a pain similar to tension-type head-aches in all but location. This should be considered in thedifferential diagnosis of all patients presenting with facialpain, especially if no objective evidence of rhinosinusitiscan be found.


585

CONCLUSIONA fundamental problem appears to be in the seem-

ingly indiscriminate labeling of all facial pain as “sinus-itis” by many primary care physicians and the lay public.Anecdotal reports in the otorhinolaryngologic literaturehave perpetuated this and have led to much unnecessaryand ineffective surgery. The aim of this article is to high-light this problem, and alert clinicians to the possibility ofother clinical entities in the diagnosis of facial pain. Themajority of patients presenting to the rhinologist withfacial pain, and no subjective or objective evidence of sinusdisease, can receive effective neurological treatment,avoiding surgical intervention.

BIBLIOGRAPHY1. Sluder G. The role of the sphenopalatine ganglion in nasal

headaches. New York Medical Journal 1908;87:989–990.2. Sluder G. Headaches and Eye Disorders of Nasal Origin.

London: Henry Kimpton, 1919:57–85.3. McAuliffe GW, Goodell H, Wolff HG. Experimental Studies on

Headache: Pain From the Nasal and Paranasal Structures,vol 23. New York: Association for Research in Nervous andMental Disease, 1943:185–208.

4. Wolf HG. Headache and Other Head Pain. New York: OxfordUniversity Press, 1948.

5. Stammberger H, Wolf G. Headaches and sinus disease: theendoscopic approach. Ann Otol Rhinol Laryngol 1988;143:3–23.

6. Blaugrund SM. The nasal septum and concha bullosa. Oto-laryngol Clin North Am 1989;22:291–306.

7. Clerico DM. Pneumatized superior turbinate as a cause ofreferred migraine headache. Laryngoscope 1996;106:874–879.

8. Cook PR, Nishioka GJ, Davis WE, McKinsey JP. Functionalendoscopic sinus surgery in patients with normal com-puted tomography scans. Otolaryngol Head Neck Surg1994;110:505–509.

9. Homer J, Sheard CE, Jones NS. Cognitive dissonance, theplacebo effect and the evaluation of surgical results. ClinOtolaryngol 2000;25:195–199.

10. Olesen J. Clinical and pathophysiological observations in mi-graine and tension-type headache explained by integrationof vascular, supraspinal and myofascial inputs. Pain 1991;46:125–132.

11. Seesle BJ. Acute and chronic craniofacial pain: brainstemmechanisms of nocioceptive transmission and neuroplas-ticity, and their clinical correlates. Crit Rev Oral Biol Med2000;11:57–91.

12. The Headache Classification Committee of the InternationalHeadache Society. Classification and diagnostic criteria for

headache disorders, cranial neuralgia and facial pain.Cephalalgia 1988;8:1–96.

13. Gerbe RW, Fry TL, Fisher ND. Headache of nasal spur origin:an easily diagnosed and surgically correctable cause offacial pain. Headache 1984;24:329–330.

14. Goldsmith AJ, Zahtz GD, Stegnjajic A, Shikowitz M. Middleturbinate headache syndrome. Am J Rhinol 1993;7:17–23.

15. Clerico DM. Sinus headaches reconsidered: referred cephal-gia of rhinologic origin masquerading as refractory pri-mary headache. Headache 1995;35:185–192.

16. Clerico DM, Evan K, Montgomery L, Lanza DC, Grabo D.Endoscopic sinonasal surgery in the management of pri-mary headaches. Rhinology 1997;35:98–102.

17. Morgenstein KM, Krieger MK. Experiences in middle turbi-nectomy. Laryngoscope 1980;90:1596–1603.

18. Ramadan HH. Nonsurgical versus endoscopic sinonasal sur-gery for rhinogenic headache. Am J Rhinol 1999;13:455–457.

19. Puig CM, Driscoll CLW, Kern EB. Sluder’s sphenopalatineganglion neuralgia—treatment with 88% phenol. Am JRhinol 1998;12:113–118.

20. Chow JC. Rhinologic headaches. Otolaryngol Head Neck Surg1994;111:211–218.

21. Ryan RE, Kern EB. Rhinologic causes of facial pain andheadache. Headache 1979;18:44–50.

22. Rosbe KW, Jones KR. Usefulness of patient symptoms andnasal endoscopy in the diagnosis of chronic sinusitis. Am JRhinol 1998;12:167–171.

23. Acquadro MA, Salman SD, Joseph MP. Analysis of pain andendoscopic sinus surgery for sinusitis. Ann Otol RhinolLaryngol 1997;106:305–309.

24. Brown JA. Sinusitis and headache. J S C Med Assoc 1991;87:502–504.

25. Jones NS, Strobl A, Holland I. A study of the CT findings in100 patients with rhinosinusitis and 100 controls. ClinOtolaryngol 1997;22:47–51.

26. Bhattachayya T, Piccirillo J, Wippold FJ. Relationship be-tween patient-based descriptions of sinusitis and parana-sal sinus computed tomographic findings. Arch Otolaryn-gol Head Neck Surg 1997;123:1189–1192.

27. Homer J, Jones NS, Sheard C. Cognitive dissonance, theplacebo effect and the evaluation of surgical results. ClinOtolaryngol 2000;25:195–199.

28. Ren K, Dubner R. Central nervous system plasticity andpersistent pain. J Orofacial Pain 1999;13:155–163.

29. Graff-Radford SB. Facial pain. Curr Opin Neurol 2000;13:291–296.

30. Bendtsen L. Central sensitization in tension-type headache—possible pathophysiological mechanisms. Cephalalgia2000;20:486–508.

31. Tarabichi M. Characteristics of sinus-related pain. Otolaryn-gol Head Neck Surg 2000;122:842–847.


586

Documents

Endoscopy-Negative, Computed Tomography-Negative Facial Pain in a Nasal Clinic