Customize your JAMA Network experience by selecting one or more topics from the list below.
KAREN H.CALHOUNMDRONALD B.KUPPERSMITHMDFrom the Division of Rhinology and Facial Plastic and Reconstructive Surgery, Department of Otorhinolaryngology–Head and Neck Surgery, University of Pennsylvania, Philadelphia.
Limited functional endoscopic sinus surgery has a good chance of curing headache in the patient who has frontal and/or retro-orbital headaches, mild mucosal edema on endoscopy, and a normal sinus computed tomographic (CT) scan.
In considering the relationship of sinus disease and headaches, Stammberger and Wolf1 divide headache patients into 3 groups. One group has headaches specifically linked to sinonasal abnormalities, such as inflammatory disease, neoplasm, barotrauma, or other readily identifiable causes. A second group has headaches unrelated to sinus causes, such as migraines, neuralgias, ophthalmologic problems, and vascular problems. The third group has no clear cause for their headaches, and a sinus origin cannot be clearly identified. This third group poses a challenge for otolaryngologists and is the subject of this report.
Devin M. Cunning, MD
In the late 1800s and early 1900s, reports described patients with headaches that were relieved by a surgical procedure on the nose, sinuses, or both. Sluder2 described a specific pain syndrome related to irritation of the sphenopalatine ganglion in the posterior nasal cavity. Wolff3 demonstrated in the 1940s that mechanical contact as well as mechanical and chemical stimulation within the nasal cavity resulted in referred headaches in 5 healthy volunteers.
More recently, authors have found that a careful endoscopic examination and a scrutinizing evaluation of a "normal" CT scan may reveal undetected subtle abnormalities. Such findings may provide a rationale for medical treatment or surgery on the nose or sinuses for selected patients with headache without a clearly identifiable cause.
Daniel O. Becker, MD
At the onset of this discussion, strict attention must be directed to defining key concepts and definitions. For example, care must be taken to define what is meant by a "normal CT." The advent of endoscopic techniques and CT scans has allowed greater scrutiny of the nasal and sinus cavities for pathological correlates of symptoms. The literature supports a role for directed endoscopic sinus surgery in patients with certain specific findings that are often overlooked on a CT scan, resulting in the label "normal CT" when in fact abnormalities exist. The term directed sinus surgery is preferable to limited sinus surgery here, because the first highlights that the surgery is directed at an identifiable pathologic condition, while the second suggests a nondescript approach with no clear goal in mind.
Stammberger and Wolf1 described their views on the role of endoscopic sinus surgery for the treatment of headaches. They reported that headaches may be of sinus origin even if the patient's history is not suggestive of sinus disease. Small blockages in the ethmoid complex, especially in the key areas of the ethmoid infundibulum or frontal recess, may give rise to headaches. Anatomic variations, including septal spurs and deviations, large agger nasi cells, and anatomic variations of the uncinate process, ethmoid bulla, and middle turbinate, such as concha bullosa, can narrow these critical clefts and predispose to contact of opposing mucosal surfaces, potentially impeding ventilation and drainage. Stammberger and Wolf maintain that, after these underlying causes have been identified, functional sinus surgery directed at this pathology can sometimes provide dramatic relief.
Clinical evidence to support the hypothesis of this article can be found in 4 related studies.4-7 The experience of Kennedy and Loury4 has been that, while in the past the extent of disease was often used as a major criterion in assessing the role of sinusitis in pain, it is now clear that the site of disease is frequently more important. The sinuses appear to be sensitive to pain before the development of chronic mucosal changes. Minor disease in a critical area can give rise to significant symptoms.
Clerico5 described headache relief after surgical or medical treatment of the sinuses in 8 of 10 patients with headaches refractory to conventional antiheadache therapy and without sinus-related symptoms. Four of these patients had x-rays with a radiologic report of "normal," but all of these patients' scans in fact displayed evidence of contact points. These 4 patients had abnormalities including septal spurs contacting a lateral nasal wall structure, enlarged ethmoid bulla, narrowed frontal recess, concha bullosa, and enlarged agger nasi cells. Nine of the 10 patients in this report had findings of abnormalities on sinus CT scan and/or endoscopy, and 8 of these patients improved with either medical (1 patient) or surgical (7 patients) therapy. In 1 patient, inspissated pus was encountered in an obstructed anterior ethmoidal cell. Removal of this material revealed the anterior ethmoid neurovascular bundle coursing along the skull base, suggesting potential neurovascular irritation as a cause in this case.
Clerico et al6 described a series of 19 patients with refractory headaches in the absence of sinus symptoms. They noted that the typical patient in this series was one with chronic, debilitating refractory headaches who had an average of 2.3 prior neurological examinations, and in whom standard medical regimens for headache had failed. A high prevalence of sinonasal abnormalities was identified in these patients, and the authors reported that 79% had experienced decreased severity of pain and frequency after directed sinus surgery.
Parsons and Batra7 reported a reduction in the intensity and frequency of headaches (91% and 85%, respectively) in patients following endoscopic surgery to relieve the contact points identified on CT in a retrospectively reviewed series of 34 patients. Relief of contact points between the septum and lateral nasal wall structures included sites in the superior, middle, and inferior turbinate. Lateralization was performed in all cases without turbinectomy. Notably, all except for 3 of these patients also had sinus complaints.
Parsons and Batra,7 and Clerico et al5,6 feel that the presence of contact points on CT and/or endoscopy is a prerequisite finding. They emphasize, however, that the presence of contact points is not pathognomonic, as multiple patients who have been identified with contact points have no headache pain. Clerico et al5,6,8 feel that it is important in some cases to be able to "prove" the correlation by response to medical therapy or a diagnostic anesthetic block. In one report,5 patients were offered surgery only if they had a clearly positive response (ie, reduction in headache) to intranasal decongestants and anesthetic sprays, or obvious findings on endoscopy and/or CT. Parsons and Batra7 found that the diagnostic blocks were excessively uncomfortable for certain patients and advocated personal physician judgment in evaluating the need for such diagnostic blocks.
The presence of an abnormality on a CT scan is not necessarily indicative of a sinus cause; the findings must be correlated with the symptoms. Calhoun et al9 evaluated CT scans in symptomatic and asymptomatic populations. Patients with sinus disease had significantly more findings of concha bullosa, septal deviation, and disease in the osteomeatal complex, but these findings were also present in a significant number of subjects without sinus disease. Similarly, it has been noted that 24% to 39% of the "normal" asymptomatic population may have mucosal changes consistent with sinusitis10; therefore, a finding on a CT scan may be incidental. With this in mind, it is important to recognize that the normal asymptomatic population will have a percentage of CT scans with incidental mucosal contact points. The surgeon considering sinus surgery for a patient with headaches and "minimal" sinus disease must keep this in mind and make every effort to find a correlation between the patient's findings and symptoms before proceeding with surgery.
Length of follow-up was less than 1 year for most of the patients in the above-mentioned studies. Some may question whether more lengthy follow-up is necessary before a clear statement can be made about the role of surgery in these patients. Unlike the oncology literature, in which a 5-year follow-up is standard, the headache literature is less clear on the appropriate length of follow-up. It would certainly be of interest to know the outcomes of these patients with a more prolonged follow-up.
A biochemical rationale for a sinus origin for some headaches has been proposed.1,5 Substance P, a peptide neurotransmitter believed to play a prominent role in the transmission of pain, has been found in high concentrations in trigeminally derived sensory nerve endings within the nasal mucosa. It has been hypothesized that an intranasal stimulus or irritant could trigger afferent conduction. These impulses are not well localized by higher cortical centers (a well-known phenomenon in referred pain), so the resultant pain could be referred to other sites in the distribution of the ophthalmic and maxillary divisions of the trigeminal nerve. Therefore, a sinonasal irritant/stimulus may cause a headache via a reflex mechanism, with the headache persisting for as long as the irritant stimulus is present.
Further investigation will certainly show the mechanism to be more complex, but clinical experience suggests a role for directed medical and surgical therapy of the nose and sinuses in some headaches.
For those patients whose CT scans are truly normal, and in whom there is no finding to link the headaches to the sinuses, the literature is skeptical of the role of sinus surgery. Again, we must pay close attention to the definition of normal, as a number of findings considered within the range of normal by radiologists may be abnormal in the context of a patient's symptoms.
However, there may be some obstruction or contact point not recognized by CT scan. Indeed, in one study, a large number of symptomatic patients did not demonstrate sinus disease by CT scan.11 Another study found no statistical association between CT findings and patient-based symptoms.12 In this study, using validated CT grading systems and validated health status instruments, no association between CT scan findings and symptom severity could be identified. On the basis of these findings, one might hypothesize the existence of individual patients who might present with headache symptoms in a sinus distribution without findings on a CT scan. In this hypothetical subset of patients, medical and/or surgical therapy could be of potential benefit.
However, experienced surgeons are increasingly reluctant to operate on this smaller subset of patients complaining of headaches in the absence of specific sinus findings (oral communication, D. W. Kennedy, MD, H. Stammberger, MD, March 1999). In the absence of significant abnormality, Kennedy feels that headaches are not sufficiently suggestive of sinusitis. He has little doubt, however, that sinusitis can exacerbate vascular headaches and migraines. Stammberger has also indicated that he is "reluctant" to perform functional endoscopic sinus surgery with a clear CT. He reiterates that if there is a clearly demonstrable finding, surgery can be successful.
Without an identifiable pathologic finding, there is no clear endpoint to surgery. Does simply "opening up the sinuses" (without a clearly defined purpose) somehow relieve pain symptoms in a patient whose only findings upon examination are mild inflammation? Perhaps mild inflammation alone causes pain in some patients, and perhaps surgery could relieve this pain. Numerous surgeons—perhaps all of us—have anecdotal stories in which this appeared to be the case. However, using the more precise definition described above, careful scrutiny reveals an abnormal CT scan in many of these cases.
Directed functional endoscopic sinus surgery has a good chance of curing frontal and/or retro-orbital headache in a patient with specific abnormalities on a CT scan. Yet there are instances when a patient's physical examination and CT scans are labeled normal, but closer scrutiny reveals abnormalities. These patients require careful endoscopy and a detailed, thorough examination of their CT. Findings on examination and on these "normal" CT scans put these patients in a category with a more favorable prognosis. Patients without significant suggestive findings after exhaustive evaluation are in a less favorable group, with a poorer prognosis.
We, too, are reluctant to operate on this latter group of patients. In all cases, we undertake aggressive medical treatment for rhinosinusitis; we require that patients undergo thorough neurologic evaluation, which occasionally is productive; we also require that patients enroll in a pain management clinic; and at times we recommend psychiatric evaluation. However, despite these measures, a few of these patients remain somewhat desperate, so a surgeon may feel compelled to attempt surgery on an individual basis. This is a situation that requires considerable introspection on the part of the surgeon, for while there is a desire to help the patient, there is also the knowledge that the chances for success are unclear at best and the risks of surgery are real.
Accepted for publication June 28, 2000.
Reprints: Daniel G. Becker, MD, Dept of Otorhinolaryngology-Head and Neck Surgery, University of Pennsylvania, 7 Silverstein, 3400 Spruce St, Philadelphia, PA 19104 (e-mail: firstname.lastname@example.org).
Becker DO, Cunning DM. Directed Functional Endoscopic Sinus Surgery and Headaches. Arch Otolaryngol Head Neck Surg. 2000;126(10):1274–1276. doi:10.1001/archotol.126.10.1274
Browse and subscribe to JAMA Network podcasts!
Create a personal account or sign in to: