SEARCH

SEARCH BY CITATION

Keywords:

  • sinus headache;
  • migraine;
  • computed tomography

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. FINDINGS
  5. COMMENTS
  6. CONCLUSIONS
  7. Acknowledgments
  8. REFERENCES

Objective.— To evaluate the sinus CT scan findings in “sinus headache” migraineurs, and to compare the findings to nonmigraine “sinus headache” patients.

Background.— The majority of patients presenting with “sinus headache” satisfy the International Headache Society (IHS) criteria for migraine headache. Few studies have correlated the rhinologic complaints and computed tomography (CT) findings in these patients.

Methods.— Thirty-five patients with “sinus headache” were evaluated prospectively and referred for CT of the paranasal sinuses. The CT scans were assessed for sinus abnormality (recorded as a Lund–Mackay [L-M] score) and were analyzed for concha bullosa and septal deviation. The findings in the migraine cohort were compared with the nonmigraine “sinus headache” patients.

Findings.— Twenty-six patients (74.3%) satisfied the IHS criteria for migraine. The mean CT scan L-M score did not differ significantly between the migraine (2.07) and nonmigraine cohort (2.66). Five of the migraine group had substantial sinus disease radiographically (with L-M scores of 5 or above). Concha bullosa of at least 1 middle turbinate was more common in the nonmigraine cohort. An analysis of the sidedness of the headaches, sinus disease, concha bullosa, and/or septal deviation is presented.

Conclusions.— The majority of “sinus headache” patients satisfy the IHS criteria for migraine. Surprisingly, these patients often have radiographic sinus disease. This raises the possibility of selection bias in otolaryngology patients, inaccurate diagnosis, or radiographic sinus disease and migraine as comorbid conditions. Positive migraine histories apparently do not obviate the need for a thorough ENT workup, possibly including CT scanning.

Abbreviations:
CT

computed tomography

L-M score

Lund–Mackay score

“Sinus headache” is a patient complaint that can have a variety of underlying causes. Recent studies have shown that the majority of patients with this complaint, at least in a primary care setting, satisfy the International Headache Society (IHS) criteria for migraine headache.1,2 Even in an otolaryngology setting, Perry et al. have shown that over 50% of computed tomography (CT) and endoscopy-negative sinus headache patients referred to a tertiary-level rhinology practice were diagnosed with migraine,3 and Jackson and coworkers found that 49% of referred sinus headache patients satisfied the migraine criteria.4 In this latter study, it was found that these patients were frequently diagnosed with sinusitis or allergic rhinitis as well as migraine. The nature of these comorbidities has not been explained.

Few studies have been performed looking at the sinus headache patient, with or without migraine, and assessing the sinus radiographic findings in these patients. The key question that this study addresses: will screening for migraine allow one to avoid unnecessary CT scanning in the sinus headache patient?

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. FINDINGS
  5. COMMENTS
  6. CONCLUSIONS
  7. Acknowledgments
  8. REFERENCES

Thirty-five consecutive patients complaining of “sinus headache,” referred to an otolaryngology/allergy office were evaluated prospectively, including a thorough headache history focusing on IHS migraine criteria.5 They were then referred for CT scanning of the paranasal sinuses. The CT scans (screening coronal views, 5-mm cuts at 8-mm intervals) were assessed for sinus abnormality (recorded as a Lund–Mackay [L-M] score; see Table 1 for an example of a blank form6) and were analyzed for concha bullosa and septal deviation by an otolaryngologist (MEM) blinded to the headache histories. The findings in the migraine cohort were compared with the nonmigraine “sinus headache” patients. The patients were also questioned about their rhinologic complaints, to assess any correlation between the sidedness of headaches and the radiologic findings, as well as any association with migraine headache. Statistical analysis was carried out using JMP IN version 5.1. One Way Analysis of Variance (T Test) was performed to determine a variance among the means. Comparative bivariate statistics was performed using chi-square analysis. Statistical significance was assumed if P < .05. All patients completed an informed consent document, per local Institutional Review Board requirements.

Table 1.—. Blank Sinus CT Radiologic Staging Form (Lund–Mackay system 1993)
Sinus SystemLeftRight
  • 0 = no abnormalities; 1 = partial opacification; 2 = total opacification.

  • *

    0 = not occluded; 2 = occluded.

  • From: Scadding GK, Lund VJ. Investigative Rhinology. London: Taylor and Francis; 2004:96.

Maxillary (0,1,2)  
Anterior ethmoid (0,1,2)  
Posterior ethmoid (0,1,2)  
Sphenoid (0,1,2)  
Frontal (0,1,2)  
Ostiomeatal complex (0 or 2 only*)  
Total points (L-M score)  

FINDINGS

  1. Top of page
  2. Abstract
  3. METHODS
  4. FINDINGS
  5. COMMENTS
  6. CONCLUSIONS
  7. Acknowledgments
  8. REFERENCES

Twenty-six patients (74.3%) satisfied the IHS criteria for migraine. The mean CT scan L-M score did not differ significantly between the migraine (2.07) and nonmigraine cohort (2.66) (Table 3). Five of the migraine group had substantial sinus disease radiographically (with L-M scores of 5 or above), as did 2 of the nonmigraineurs. Ten of the migraine group had L-M scores of 0, as did 4 of the nonmigraine cohort. Concha bullosa of at least 1 middle turbinate was more common in the nonmigraine cohort. An analysis of the sidedness of the headaches, concha bullosa, and/or septal deviation found no significant correlation between the sidedness of the headache and the presence or absence of these factors on that side (Table 4). This was found to be true in both the migraine and nonmigraine cohort groups. The sidedness of headache did not correlate with an increased mean L-M score on that side (Table 4).

Table 3.—. Comparison of Migraine and Nonmigraine Cohorts
 Migraine N = 26Nonmigraine N = 9P Value
Mean Lund–Mackay score2.072.660.7
Concha bullosa present240.02
Deviated nasal septum830.88
Table 4.—. Correlations of Headache Sidedness
 Headache Left N = 9, 8 MigraineursHeadache Right N = 9, 8 Migraineurs
  • *

    P = .95;

  • **

    P = .65;

  • ***

    P = 1.00 for left-sided headache versus L-M score.

  • 0.59 for right-sided headache versus L-M score.

Concha bullosa right10*
Concha bullosa left01
Deviated septum to right21**
Deviated septum to left22
Mean L-M score right0.671.67***
Mean L-M score left0.672.22***

The majority of patients had substantial rhinologic complaints, but these, again, did not differ substantially between the 2 cohort groups (Table 2).

Table 2.—. Nasal Complaints
Nasal SymptomWhole “Sinus Headache” Group N = 35 (%)Migraine Subgroup N = 27 (%)Nonmigraine Subgroup N = 8 (%)
Congestion778175
Nasal discharge5455.562.5
Itching374412.5
Sneezing374425
Postnasal drip8077.787.5

COMMENTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. FINDINGS
  5. COMMENTS
  6. CONCLUSIONS
  7. Acknowledgments
  8. REFERENCES

In the rhinologic patient, sinus CT scan interpretation is wrought with difficulty. Jones cited a 30% incidence of incidental changes, and supported clinical correlation with radiographic findings.7 Shields et al. reported no correlation between headache, facial pain, and radiographic abnormalities (n = 51).8 No attempt to identify migraine headache patients was reported in this study. Tarabichi found no association between pain severity and mucosal disease (n = 82) in sinus headache patients. Again, migraine was not addressed in this report.9 Kenny et al., similarly, found no correlation between headache, facial pain, and CT disease severity (n = 273).10 Bhattacharyya et al. found no correlation between patient symptoms (SNOT-20) and CT findings, including facial pain.11

Despite the difficulties correlating patient symptoms with CT findings, Anzai et al. reported that CT findings considerably changed management, especially surgical management of these patients.12 Other studies have correlated CT scores with severity of rhinologic symptoms in chronic sinusitis patients.13 Stankiewicz and Chow have presented a reasonable series of recommendations regarding the incorporation of CT scanning in the management of the rhinology patient,14,15 as have recent management algorithms.16 The importance of including CT information in the workup of the chronic sinusitis patient was stressed in a superb recent study by Bhattacharyya.17

In the present study, CT scanning was used as one of the initial steps in evaluating the sinus headache patient. Several authors have supported the use of CT scanning in the rhinology patient with unclear diagnosis.14-17 The CT findings in the present series, however, were often surprising, showing sinus abnormalities in the majority of the migraineurs, with extensive disease (L-M scores >5) in several of the migraine cohorts. The utility of these findings remains to be seen. Radiographic studies, in general, are not considered necessary in the uncomplicated migraine headache patient, at least according to the neurologic literature.18,19

The nasal complaints (congestion, etc.) in this series also did not differ between the 2 groups. Schreiber et al. have shown that “sinus headache” migraineurs frequently have sinonasal complaints, with sinus pressure (84%), sinus pain (82%), and congestion (63%) being particularly common.1 Since the diagnostic criteria for both acute and chronic sinusitis label all of these “major criteria,”20,21 clearly the specificity of these symptoms for the diagnosis of sinusitis is questionable in the “sinus headache” patient. Confounding the issue further are the IHS criteria for the diagnosis of rhinosinusitis-related headache,5 which incorporate these same symptoms. Newer discussions of acute sinusitis diagnosis have focused more on the length of illness, which may be very helpful in the clinical diferentiation of these patients (ie, migraine less than 72 hours, acute bacterial rhinosinusitis a consideration after 5 days of symptoms).22

None of the above addresses the role of radiographic studies in the differentiation of these patients, although the IHS criteria for rhinosinusitis-related headache mention supportive radiographic findings as helping to confirm diagnosis. Since many of the migraine cohort group in this study had “positive” radiographic findings, the interpretation of these findings becomes difficult (see below). Radiographic findings apparently play a supportive role in the diagnosis of acute or chronic sinusitis, which are clinical diagnoses according to recent otolaryngology publications.20,21 Regardless, many otolaryngologists find them useful in the diagnosis and management of these patients.23

One of the secondary issues addressed by this series was the sidedness of headache, and the lack of correlation with findings such as septal spurs and concha bullosa. In this series, no correlation was found between the side of headache (eg, right or left) and the presence of sinus disease (L-M score), septal spurring, or a concha bullosa on that side. This is in agreement with recent studies, which found no correlation between septal spurring and the sidedness of headaches.24,25 The issue of mucosal “contact points” in the etiology of headache remains questionable, although this remains an issue cited anecdotally in the management of these patients. In this series, septal spurs had no correlation with the sidedness of headaches, and are thus questionable as a causative agent for these complaints.

One interesting association found in this small series was that the nonmigraine headache cohort was significantly more likely to have a concha bullosa deformity (pneumatization) of at least 1 middle turbinate. The relevance of this finding, however, is difficult to interpret considering the lack of a correlation between the sidedness of the concha bullosa and the sidedness of the headaches in those patients with unilateral complaints. A larger study may be helpful in resolving any association between concha bullosa and headache.

In summary, this series of patients found no correlation between the radiographic findings in sinus headache sufferers with a positive history for migraine, versus those which do not. Many of the apparent migraine cohorts had substantial radiographic sinus disease. Possible interpretations of these findings:

  • 1
    Selection bias in referred ENT patients. These were all patients referred to an ENT practice for their sinus headache problems. Many of them had prior radiographic studies of their sinuses. Patients with a history of prior radiographic sinus disease may be more likely to be referred to an otolaryngology office. “Sinus headache” migraineurs with normal prior radiographic studies may be less likely to be referred.
  • 2
    Radiographic sinus findings are meaningless, both migraine and sinusitis are clinical diagnoses. This is a very real possibility based on a number of studies (cited above). Still, many otolaryngologists rely on CT scans in the workup of these patients, and the vast majority of them rely on CT for presurgical planning, when sinus surgery is contemplated. Thus, the role of these scans may vary from physician to physician, and patient to patient.
  • 3
    Many of the apparent migraine group really are not migraineurs, using the IHS criteria. The IHS criteria for migraine diagnosis explicitly state that the headaches are “not explained by another cause.” Would the positive CT scans in the migraine cohort group be considered “another cause” to the diagnosing physician? Because chronic sinusitis has never been validated as a cause for headache, this would not necessarily contradict the diagnosis of migraine in these patients. Regardless, the efficacy of sinus surgery in the resolution of headaches in many patients with chronic sinusitis, has been well established.26 Few of these studies, however, screened the patients using IHS migraine criteria.
  • 4
    Sinus disease and migraine may be comorbid conditions. There may be an association with the “sinus headache” presentation of migraine, and the presence of active sinonasal disease. Jackson et al. have reported that 49 of 100 patients referred to an ENT office for sinus headache had previously unrecognized migraine. Among these migraineurs, only 13% had migraine alone, 19 (of the migraineurs) had allergic rhinitis as well, 11 had sinusitis, and 6 had both allergic rhinitis and sinusitis.4 The present series would support the notion that migrainous “sinus headache” and sinonasal disease are often concomitant. The nature of this association, if any, remains obscure.

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. METHODS
  4. FINDINGS
  5. COMMENTS
  6. CONCLUSIONS
  7. Acknowledgments
  8. REFERENCES

Sinus headache patients with positive migraine histories frequently have substantial radiographic sinus disease. These findings did not differ from the nonmigraine sinus headache cohort. The utility of these findings remains unclear, as does the possibility of rhinitis, rhinosinusitis, and migraine headache as comorbid conditions. Regardless, a positive migraine history cannot be used to select patients not to scan.

Acknowledgments

  1. Top of page
  2. Abstract
  3. METHODS
  4. FINDINGS
  5. COMMENTS
  6. CONCLUSIONS
  7. Acknowledgments
  8. REFERENCES

Acknowledgments: This study was presented at the Annual Meeting of the American Rhinologic Society, September 2006, Toronto, Canada. The authors wish to thank Shannon M. Tarr, B.S., Fairview Hospital Department of Surgery, for her assistance in the statistical analysis presented here.

Conflict of Interest: Dr. Mehle is on the Speaker's Bureau for GSK, Schering-Plough, Aventis, Alcon, and Medpointe. Ms. Kremer has no disclosures.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. FINDINGS
  5. COMMENTS
  6. CONCLUSIONS
  7. Acknowledgments
  8. REFERENCES