Grading of the olfactory cleft. A, Coronal images demonstrate the level of the inferior border of the anterior attachment of the middle turbinate (upper, white line) and the level just below the skull base (lower, white line). B-F, Axial images of the inferior border of the anterior attachment of the middle turbinate (upper row) and first axial image just below the skull base (lower row). B, Grade 0 (no opacification in the olfactory cleft). C, Grade 1 (≤25% opacification of the olfactory cleft). D, Grade 2 (25%-50% opacification). E, Grade 3 (50%-75% opacification). F, Grade 4 (>75% opacification).
Chang H, Lee HJ, Mo J, Lee CH, Kim J. Clinical Implication of the Olfactory Cleft in Patients With Chronic Rhinosinusitis and Olfactory Loss. Arch Otolaryngol Head Neck Surg. 2009;135(10):988-992. doi:10.1001/archotol.127.1.56
To evaluate the relationship between findings via osteomeatal unit computed tomography (OMU CT) of the olfactory cleft and olfactory function in patients with chronic rhinosinusitis (CRS).
Retrospective review of medical records.
Two hundred ten patients with CRS who underwent OMU CT and olfactory function tests were included in this study.
Main Outcome Measures
All the paranasal sinuses were graded via the Lund-Mackay scoring system. The olfactory cleft was graded on a scale of 0 to 4 according to its opacification. Olfactory function was evaluated by the butanol threshold test (BTT) and the 16-odor identification test (OIT).
The radiologic grade of the olfactory cleft was more significantly correlated with olfactory function than the grades of the paranasal sinuses. In patients without allergy, the BTT and OIT scores were inversely correlated with the CT score of the olfactory cleft. However, in patients with allergy, only the BTT score had a negative correlation with the CT score of the olfactory cleft, whereas the OIT score did not. The OIT score showed a significant negative correlation with the opacification of the olfactory cleft in the mild and moderate CRS group only, whereas the BTT score showed a significant negative correlation in all stages of CRS.
The opacification of the olfactory cleft had a negative correlation with the olfactory function scores in patients with CRS. The olfactory cleft findings on OMU CT may give some clues to the olfactory function in patients with CRS.
Olfactory dysfunction is caused by viral upper respiratory tract infection, chronic sinonasal inflammation, or head trauma. However, not many systemic studies have been performed with regard to the implication of chronic rhinosinusitis (CRS) in olfactory dysfunction, even though CRS is one of the most frequent causes of olfactory dysfunction and accounts for 21% to 25% of smell loss.1- 5
Two mechanisms have been suggested to explain CRS-induced olfactory dysfunction. The first mechanism, conductive olfactory loss, can be caused by swollen or hypertrophic nasal mucosa or nasal polyps, which inhibits the transport of odorants to the olfactory receptor neurons. The second mechanism, sensorineural olfactory loss, results from infiltration of inflammatory cells that release inflammatory mediators and may directly or indirectly affect functions or structures of the olfactory neuroepithelial cells.1 The severity of olfactory loss is related to the severity of CRS, and patients with nasal polyps on the superior turbinate, middle turbinate, or cribriform plate have more severe olfactory dysfunction than the patients without nasal polyps.6
Osteomeatal unit computed tomography (OMU CT) is a powerful tool used in the diagnosis of CRS. It is helpful not only for the diagnosis of CRS but also in the assessment of the extent and severity of disease and in determination of the range of operation. Furthermore, OMU CT can provide information about mucosal swelling or obstruction of the olfactory cleft, which is the most important anatomical structure for olfaction. Studies that try to evaluate the relationship between the severity of CRS evaluated by CT and olfactory function have been published, and some reports1 insist that the severity of inflammation shown on CT scans was related to olfactory function. The threshold of olfaction in patients with obliterated olfactory clefts found by CT is higher than that in patients with open olfactory clefts.7 However, a correlation between the severity of obliteration of the olfactory cleft and olfactory function has not yet been found.
Because the olfactory cleft contains olfactory receptor neurons, it is expected that the inflammatory status of the olfactory cleft is related to the degree of olfactory loss. This study evaluates the relationship between the CT findings of the olfactory cleft and olfactory function in patients with CRS.
Patients diagnosed as having CRS by OMU CT at Seoul National University Bundang Hospital from June 1, 2006, through June 30, 2007, were included in this study. Olfactory function tests were performed for all the patients. All patients also underwent endoscopic examination to assess the presence of nasal polyps. The presence of allergic rhinitis was determined by allergic symptoms and a skin prick test or a serum specific IgE test (multiple allergen simultaneous test). Patients with a history of head trauma, acute olfaction loss after upper respiratory tract infection, or central nervous systemic disease were excluded. Patients older than 65 years were not included in the study to exclude senile olfaction loss. Two hundred ten patients (371 nostrils) were included in this study. The nostrils without CRS were not included. The male to female ratio was 67:33, patient age ranged from 10 to 65 years, and the mean age was 41.0 years. This study was approved by the Internal Review Board of Seoul National University Bundang Hospital.
We obtained OMU CT images with 1-mm thickness in the axial plane without contrast enhancement. With 3-dimensional reconstruction, coronal images were also acquired. All the paranasal sinuses were graded on a scale of 0 to 2 via to the Lund-Mackay system.8
The olfactory cleft was graded by its opacification apparent on OMU CT scans. The boundaries of the olfactory cleft were defined as follows. The anterior boundary was the anterior attachment of the middle turbinate, and the posterior boundary was the anterior wall of the sphenoid sinus. The medial boundary was the nasal septum, and the lateral boundary was the middle and superior turbinates. Grading of the olfactory cleft was performed by averaging of the opacified area of the olfactory cleft in 2 axial images. One image was the axial image that included the inferior border of the anterior attachment of the middle turbinate, and the other one was the first axial image just below the skull base. The olfactory cleft was graded on a scale of 0 to 4 via the ratio of the opacified area to the whole area of the olfactory cleft, with 0 indicating no opacification; 1, opacification of the olfactory cleft of 25% or less; 2, 25% to 50% opacification; 3, 50% to 75% opacification; and 4, more than 75% opacification (Figure).
We classified CRS in accordance with severity (mild, moderate, or severe) to identify the significance of severity in relation to the opacification of the olfactory cleft. We defined CRS with a Lund-Mackay score of 4 or higher as mild and CRS with a Lund-Mackay score of 9 or higher as severe. We categorized CRS with a Lund-Mackay score of 5 to 8 as moderate.
For the butanol threshold test (BTT), 100% n-butanol was 3-fold serially diluted with mineral oil to form 13 concentrations. The BTT was performed in each nostril separately. The patients were given 2 polyethylene bottles, one with mineral oil and the other with butanol. The participants were obliged to choose 1 bottle with butanol. The test was repeated until 5 correct choices were made successively. The concentration level was designated as a threshold level.9,10
To test odor identification, a 16-odor identification test (OIT) of the Korean version of Sniffin’ Sticks was used, which is a modification of Sniffin’ Sticks that is well validated and widely used in Korea. The OIT contains 16 odorants familiar to Koreans and was also performed in each nostril separately.11
The correlations between OMU CT findings and the results of olfactory function tests were analyzed statistically and were also reviewed in accordance with the presence of allergic rhinitis and nasal polyps. The multiple linear regression analysis was used to reveal correlations among variables. The Pearson correlation test was also used. All the tests were 2-tailed, and statistical significance was accepted when P<.05.
The mean (SD) BTT score was 7.16 (3.41), and the mean (SD) OIT score was 7.91 (3.39). The mean (SD) OMU CT score of all the sinuses was 4.26 (2.63), and the mean (SD) score of the opacification of the olfactory cleft was 1.16 (1.47).
While we analyzed the correlations between the OMU CT score and BTT score, we found that the opacification of the frontal sinus (β = −.168, P= .003) and the olfactory cleft (β = −.459, P < .001) showed a significant negative correlation with the BTT score. The OIT score also showed a significant negative correlation with the opacification of the frontal sinus (β = −.167, P = .004) and the olfactory cleft (β = −.470, P < . 001) (Table 1).
The implication of the opacification of the olfactory cleft related to the olfactory function was assessed in accordance with the severity of CRS. The opacification score of the olfactory cleft had negative correlations with both the BTT and OIT scores in the mild (BTT: r = −0.368, P < .001; OIT: r = −0.376, P < .001) and moderate CRS groups (BTT: r = −0.467, P < .001; OIT: r = −0.478, P < .001). However, in the severe CRS group, the BTT score was negatively correlated with the olfactory cleft opacification (r = −0.527, P = .002), but the OIT score was not (r = −0.329, P = .06) (Table 2).
Regardless of the presence of nasal polyps, the negative correlations between BTT or OIT scores and the opacification of the olfactory cleft were significant in patients with CRS (Table 3). In patients without allergies, both the BTT and OIT scores were inversely correlated with the CT score of the olfactory cleft (BTT: β = −.490, P < .001, OIT: β = −.512, P < .001). However, in the case of patients with allergy, the BTT score had a negative correlation with the opacification score of the olfactory cleft (β = −.341, P = .03), whereas the OTI score did not (β = −.318, P = .06) (Table 3).
Chronic rhinosinusitis is one of the main causes of olfactory dysfunction. Therefore, patients with olfactory loss owing to CRS can be easily found among patients who visit sinonasal disease clinics. However, otorhinolaryngologists do not pay much attention to the findings of the olfactory cleft and olfactory function of patients during the review of the paranasal sinus findings of OMU CT. One of the goals of this study is to initiate interest in the olfactory function of patients among otorhinolaryngologists.
The relationship between CRS and olfactory impairment has been reported in several studies.1- 3 Osteomeatal unit CT is a valuable method not only for the assessment of the extent and severity of CRS but also for the evaluation of the causes of olfactory dysfunction.12 However, the results of studies on the correlation between the severity of CT findings in CRS patients and the degree of olfactory dysfunction have not been consistent. The severity classification of CRS developed by Min et al13 was related to the BTT threshold and subjective symptoms. In another study,13 a statistically significant correlation was found between the preoperative olfactory dysfunction evaluated with the University of Pennsylvania Smell Identification Test and the severities of CRS classified in accordance with a staging system proposed by Downey et al.14 On the other hand, other authors14,15 have found no correlation between CT findings and olfactory function.
The opacification of the olfactory cleft and that of each paranasal sinus was analyzed in this study. To evaluate the olfactory function, olfactory detection threshold and smell identification were used. It seems that the olfactory threshold test may be more sensitive in the evaluation of conductive olfactory loss, whereas the olfactory identification test may be more sensitive in the assessment of sensorineural olfactory loss. This theory is well supported by the fact that patients with minimal cognitive impairment show dysfunction in olfactory identification only, but not in olfactory detection.16 Because CRS can result in both conductive and sensorineural olfactory dysfunction, it would be more reasonable to assess both identification and detection.
In our study, OMU CT findings of all the paranasal sinuses, except the frontal sinus, did not prove a significant correlation with the results of the olfactory function tests. However, the strongest negative correlation was found with olfactory function and the CT findings of the olfactory cleft. This result was anticipated because the olfactory cleft, which contains olfactory receptor neurons, is the most important structure for olfaction. It is unclear how the frontal sinus affects the olfactory function. However, it is assumed that the mucopus from the frontal sinus is drained to the OMU, which might affect the olfactory cleft. Still, more investigation is needed to prove a more precise pathophysiologic relationship between frontal sinusitis and olfactory function.
Mechanical obstruction of the olfactory cleft is important in CRS patients regardless of the degree and the cause of the obstruction. Partial and complete obstruction affects the capacity to smell.17,18 A study by Pfaar et al19 demonstrated that the mechanical obstruction of the anterior part of the olfactory cleft decreased the orthonasal olfactory function. Some previous studies revealed that OMU CT findings of the olfactory cleft were related to olfactory function. One study4 showed that the olfactory detection threshold had something to do with the patency of the olfactory cleft evaluated by CT. This study compared only the threshold of the BTT score in patients with an open olfactory cleft apparent on a CT scan with that in patients with closed olfactory fissure. Another study1 suggested that the upper portion of the nasal cavity and the posterior half of the middle meatus assessed by CT had a significant correlation with the Sniffin’ Sticks test results and subjective ratings of olfactory dysfunction. These studies support our results, but our study may be more valuable in that the correlation between the CT findings of the olfactory cleft and the results of BTT and OIT was quantitatively assessed.
The opacification of the olfactory cleft on CT was investigated in accordance with the severity of CRS. In mild and moderate CRS, both the BTT and OIT scores were related to the opacification of the olfactory cleft, whereas in severe CRS, only the BTT score showed a negative correlation. The BTT results were mainly affected by conduction of odorants. However, the OIT results were affected by both conduction and sensorineural problems. In severe CRS, the severe inflammation induces inflammatory cytokine production and inflammatory cell recruitment and results in olfactory epithelial dysfunction and sensorineural olfactory loss. Because the CT findings of the olfactory cleft reflect the conduction of odorants and the BTT results reflect conductive olfactory loss better, this might be the reason why the BTT score correlated with the CT finding even in severe CRS. However, in severe CRS, which could cause profound sensorineural olfactory loss, identification test results may not show a good correlation with the CT finding.
Both the BTT and OIT scores correlated with the opacification of the olfactory cleft regardless of the existence of nasal polyps. Although nasal polyps could induce severe obstruction in the olfactory cleft and affect the results of olfactory tests, both the BTT and OIT scores showed a significant correlation in patients with or without nasal polyposis with regard to the correlation with the olfactory cleft opacification. Although the BTT score but not the OIT score correlated with olfactory cleft opacification in patients with allergy, both the BTT and OIT scores were correlative in patients without allergy. These results could also be explained by the similar pathogenesis of olfactory dysfunction in severe CRS. Allergic rhinitis itself could elicit olfactory dysfunction by the induction of changes in the olfactory epithelium.20 This allergic mucosal inflammation causes mucosal swelling and inhibits the conduction of odorants, which results in conductive olfactory dysfunction. The degree of this process can be measured by CT, but olfactory cellular death that causes sensorineural olfactory loss cannot be evaluated by CT. Therefore, sensorineural loss demonstrated by the OIT might not correlate with the CT findings of the olfactory cleft, but conductive loss demonstrated by the BTT might be correlative.
One limitation of our study is that we evaluated only orthonasal olfactory function. Orthonasal olfactory function is more influenced by obstruction of the anterior part of the olfactory cleft than retronasal olfactory function.19 One study21 showed that retronasal olfactory function was less influenced by nasal polyposis than orthonasal olfactory function. Further evaluation of the CT findings of the olfactory cleft, such as retronasal smelling, is required.
In conclusion, CT findings of the olfactory cleft had a statistically significant relationship with olfactory function, such as olfactory detection and identification, in patients with CRS. Otorhinolaryngologists should pay more attention to the olfactory cleft during their review of OMU CT scans of patients with CRS. This process will help in the assessment of olfactory function of patients and the establishment of management plans for problems with their sense of smell.
Correspondence: Jeong-Whun Kim, MD, PhD, Department of Otorhinolaryngology, Seoul National University Bundang Hospital, Gumi-dong, Bundang-gu, Seongnam-si, Gyeonggi-do, 463-707, South Korea (email@example.com).
Submitted for Publication: December 30, 2008; final revision received March 25, 2009; accepted April 23, 2009.
Author Contributions: Dr Kim had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Chang, H. J. Lee, Mo, C. H. Lee, and Kim. Acquisition of data: Chang and H. J. Lee. Analysis and interpretation of data: Chang, H. J. Lee, Mo, C. H. Lee, and Kim. Drafting of the manuscript: Chang. Critical revision of the manuscript for important intellectual content: Chang, H. J. Lee, Mo, C. H. Lee, and Kim. Study supervision: C. H. Lee and Kim.
Financial Disclosure: None reported.
Funding/Support: This article was partly supported by Seoul National University Bundang Hospital grant 02-2008-018 and the grant R11-2000-075-01004-0 from the Engineering Research Center program of the Ministry of Education, Science, and Technology/Korea Science and Engineering Foundation.