Leon EA, Catalanotto FA, Werning JW. Retronasal and Orthonasal Olfactory Ability After Laryngectomy. Arch Otolaryngol Head Neck Surg. 2007;133(1):32-36. doi:10.1001/archotol.133.1.32
To characterize orthonasal and retronasal olfactory ability in patients who have had a total laryngectomy (TL).
Prospective psychophysical evaluation of orthonasal and retronasal olfactory function.
Academic center outpatient clinic.
Volunteer sample of 36 subjects who underwent laryngectomy 0.5 to 25.0 years after TL (median, 3.5 years) and 36 age-, sex-, and smoking history–matched controls.
Main Outcome Measures
Scores on established psychophysical tests of orthonasal and retronasal olfaction and self-rating scores of smell assessed with a visual analog scale.
Assessment of orthonasal olfactory ability yielded a mean composite score of 4.3 (maximum score, 7) for the TL group and 5.3 for the control group (P = .002). Evaluation of retronasal olfactory ability resulted in a mean score of 11.0 (maximum score, 20) for the TL group vs 14.3 for the control group (P<.001). The mean self-rating scores of smell were 2.9 and 6.6 (maximum score, 10) for the TL and control groups, respectively (P<.001). Self-ratings of smell positively correlated with orthonasal (rS = 0.42; P<.001) and retronasal (rS = 0.50; P<.001) olfactory ability.
Laryngectomy is associated with measurable decreases in olfactory function that are also subjectively perceived. Quantification of decrements in orthonasal and retronasal olfactory function can be used to characterize the severity of hyposmia and to assess the potential for, and efficacy of, olfactory rehabilitation. Although self-assessment with a simple visual analog scale successfully identifies many laryngectomy patients who have objective evidence of olfactory dysfunction, further investigation is necessary to evaluate and compare its validity and reliability with other available survey instruments that purport to measure olfaction.
Smell sensation is frequently altered after total laryngectomy (TL). Severe impairments of olfactory ability have been reported in two thirds of patients who undergo laryngectomy, and some degree of hyposmia was found in the remainder.1,2 Olfactory-evoked potentials are also undetectable in many of these patients.2 The inability to detect smoke or other odorous danger signals can threaten the personal safety of patients after laryngectomy. Moreover, impaired olfactory function adversely impacts their quality of life, which contributes to weight loss and poor nutritional status.3 The ability to sense different flavors requires a functional olfactory epithelium. Retronasal olfaction occurs when odorant molecules from the oral cavity are delivered via the nasopharynx and posterior choanae to the olfactory epithelium in the olfactory cleft. Alternatively, orthonasal olfaction occurs when odorant molecules are delivered to the olfactory epithelium via the nares. Following laryngectomy, patients who score poorly on tests of olfactory function also report decreased gustation.1 To our knowledge, the retronasal olfactory ability of patients who undergo laryngectomy has not previously been objectively measured. This investigation was conducted to characterize the orthonasal and retronasal olfactory ability of patients following TL and to assess the correlation between measured olfactory function and the patient's subjective sense of taste and smell.
Once approval from the institutional review board of the University of Florida, Gainesville, was obtained, subject enrollment began. Patients were eligible for participation if they had undergone TL at Shands Hospital at the University of Florida between 1999 and 2005 or if they had undergone laryngectomy elsewhere and were being actively observed at Shands Hospital. Patients were excluded if their surgery had been performed less than 6 months before enrollment or if they had a documented sinonasal or olfactory disorder, anatomical obstruction of the nasal airway, or a previous allergic reaction to any of the test substances. Current patients from the Shands Hospital outpatient clinics who had not undergone TL were also eligible for participation as part of a control group to evaluate for differences in olfaction. Previous reports have indicated that women score higher on tests of olfaction than men,4,5 increasing age and olfactory ability are inversely related,4- 6 and smoking may6- 8 or may not4,9 negatively affect olfactory ability. Consequently, subjects were chosen for the control group so that the age, sex, and smoking history of the members of both groups were similar, thereby minimizing the impact of confounders. Patients with a smoking history were grouped into the following categories: currently smoking, never smoked, and quit smoking 1 year ago or less, more than 1 through 5 years ago, more than 5 through 20 years ago, or more than 20 years ago. The exclusion criteria for the control group was the same as for the TL group. To detect a 2-point difference of orthonasal olfaction scores and a 3-point difference of retronasal olfaction scores with 80% power at the P<.05 level of significance, it was calculated by t test approximation that a sample size of 36 patients was required for each group. Enrollment in the study was accordingly closed once 36 subjects had been tested in each group. Written informed consent was obtained from each subject.
For each patient in the TL group, the length of time in years between laryngectomy and the testing date was recorded, and any history of radiotherapy was documented. The subjects in both groups were then asked to self-rate their ability to smell on a visual analog scale from 1 to 10. A score of 1 indicated a complete lack of smell, and 10 indicated an excellent sense of smell. Subjects were also asked to rate their ability to taste on a similar scale. Once the self-assessment was completed, all subjects underwent the following tests of olfaction, administered by one of us (E.A.L.) from September to November 2005.
A well-established test of orthonasal olfaction developed at the Connecticut Chemosensory Clinical Research Center (CCCRC) includes a butanol threshold test as well as an odor identification test using common odors.10,11 This test was used with slight modification of the odor identification test to compensate for the inability of patients who have undergone laryngectomy to sample odorants via a “sniff” maneuver.
For each trial, 2 plastic bottles were presented to the subject. One contained water and the other a dilute concentration of butanol. The bottles were of identical appearance and were presented simultaneously. Subjects were instructed to occlude 1 nostril and place the tip of the first bottle immediately beneath the other nostril. The bottle was then squeezed at least twice while the subject focused on smelling the air expelled from the bottle. The second bottle was then sampled in a similar manner, and the subject had to choose which of the bottles contained something other than water. If the choice was incorrect, the next stronger concentration of butanol was presented along with a bottle containing only water. Once the subject identified correctly the same butanol concentration 5 times in a row, the score was recorded for that nostril. The other nostril was then tested separately, and the scores for both nostrils were averaged to arrive at the final score. The strongest butanol concentration (bottle 0) was 4% butanol in deionized water. Each subsequent dilution (bottles 1-9) was a 1:3 dilution with deionized water. Possible scores ranged from 0 to 9, but all scores 7 and higher were scored as 7 per the CCCRC test.10,11
Common household odorants (peanut butter, Ivory soap [Procter & Gamble, Cincinnati, Ohio], mothballs, Vicks VapoRub [Procter & Gamble] chocolate, Johnson & Johnson's Baby Powder [Johnson & Johnson Consumer Products, Skillman, NJ], coffee, and cinnamon) were placed within opaque squeezable bottles. The CCCRC test10,11 uses jars covered with gauze; however, because patients who have undergone laryngectomy are unable to sniff, odors were delivered by squeezing the bottle underneath one nostril, while the other was occluded. Subjects then chose from a printed list containing the correct items as well as an equal number of distractor items. The forced-choice items included the following: Vicks VapoRub, burnt paper, wood shavings, coffee, baby powder, peanut butter, spearmint, cinnamon, Ivory soap, chocolate, mothballs, grape jam, ketchup, black pepper, and rubber. The ability to sense Vicks VapoRub indicates intact trigeminal nerve function. It was easily identified by all subjects and was not included in the final score. Possible scores ranged from 0 to 7 items correctly identified. Scores for both nostrils were averaged to arrive at the final score.
Scores for the butanol threshold test and identification tests were subsequently averaged to arrive at a composite score for orthonasal olfactory ability. As in the CCCRC test, scores were grouped by category as follows: 0 to 1.75, anosmia; 2.00 to 3.75, severe hyposmia; 4.00 to 4.75, moderate hyposmia; 5.00 to 5.75, mild hyposmia; and 6.00 to 7.00, normal.
Different flavored powders were applied to the tongue and then identified using a forced-choice paradigm. This test was closely modeled after a test developed by Heilmann et al,12 which had a reported a test-retest reliability of r = 0.76. This test has been used in studies of patients with nasal polyposis13 and in other more recent studies of retronasal olfaction.14,15 Twenty flavored powders were purchased at a grocery store and included garlic, strawberry, milk, lemon, orange, vanilla, cinnamon, cloves, paprika, curry, butter buds, bread, cocoa, celery, chicken bouillon, grape, raspberry, onion, ginger, and coffee. The test as originally described12 included “muscat,” a wine grape not widely recognized in the United States. Grape flavor was used instead. The original test also included smoked ham, which could be objectionable to some subjects for religious reasons. Chicken bouillon was used instead. Orthonasal olfactory function was tested by Heilmann et al,12 and subjects were grouped into 3 groups: subjects with anosmia, hyposmia, and normal olfactory function. On retronasal testing, the normal group achieved a mean score of 17; subjects with hyposmia had a mean score of 12, and subjects with anosmia had a mean score of 8.5. Retronasal test results in this study were compared with these group mean scores as described by Heilmann et al.12
The testing procedure was as follows: The subject rinsed the mouth with water. He or she was then instructed to close the eyes, occlude the nostrils, and stick the tongue out. The control subjects were warned not to inhale. A small amount of the flavored powder was scooped up into the tip of a disposable straw and deposited on the subject's mid-dorsal tongue. The nostrils were then released, and the subject attempted to taste the powder. If no flavor was sensed, the subject could ask for more powder as necessary. Once a flavor was sensed, the subject chose from a printed list of 4 items for each test: 1 was the correct flavor (bold) and 3 were similar distracters. The forced-choice items included the following:
Mustard, paprika, ginger, curry
Lemon, grapefruit, sour cherry, grape
Pizza, garlic, sauerkraut, bread
Vanilla, milk, coconut, banana
Apple, strawberry, grape, tangerine
Vanilla, cherry, banana, honey
Orange, raspberry, strawberry, cherry
Chives, chicken bouillon, salami, onion
Sage, grape, cocoa, caramel
Chives, parsley, celery, carrots
Butter, bread, fish, milk
Ginger, paprika, curry, mustard
Cinnamon, cocoa, grape, coffee
Garlic salt, chicken bouillon, onion salt, beef soup
Caraway, cloves, dill, anise
Cinnamon, coffee, grape, cocoa
Curry, cheese, cucumber, mustard
Honey, caramel, cocoa, cinnamon
Peach, pineapple, raspberry, white grapes
Chives, garlic, celery, onion.
The subject rinsed out their mouth with water between each tested item. Answers were scored from 0 to 20 correct item identifications. Total testing time including orthonasal and retronasal testing ranged from approximately 45 minutes to 1.5 hours per subject.
Statistical analysis was performed using SAS 9.1 statistical software (SAS Institute, Cary, NC, 1989). The Wilcoxon 2-sample test was used to analyze mean score differences between the TL and control groups on orthonasal and retronasal testing and self-rating scores of smell and taste function. This test was also used to assess for differences between psychophysical test scores of subjects with and without a history of radiation treatment to the head and neck. The Fisher exact test was used to assess for any difference between TL and control groups in terms of smoking history categories and also to assess for differences in category assignment of orthonasal olfactory ability. Spearman correlation matrix was used to relate self-rating of smell and taste ability with psychophysical test scores. Pearson correlation coefficient was used to assess for any correlation between time since laryngectomy and psychophysical test scores. The χ2 test was used to assess for any difference in prevalence of prior radiation treatment between the 2 groups.
Table 1 summarizes the subject characteristics. There was no significant difference in age, sex, or smoking history between the TL and control groups. The mean age was 67.0 years (range, 45-83 years) for the TL group and 66.8 years (range, 43-84 years) for the control group (P = .92). There were 4 women and 32 men in each group. The smoking history of these 2 groups was similar (P = .97). There were more subjects with a history of radiation treatment to the head and neck in the TL group (31/36) than in the control group (16/36) (P<.001).
Results of orthonasal and retronasal olfactory testing are summarized in Table 2. Measurement of orthonasal olfaction showed that patients who had undergone laryngectomy had poorer orthonasal olfactory function than did controls. The mean score (average of butanol threshold and identification scores) was 4.31 of 7 (range, 0.25-6.75) for the TL group and 5.34 (range 1.75-6.75) for the controls, which was statistically significant (P = .002). A score of 4.31 for an individual in the CCCRC scoring system would indicate “moderate hyposmia.” A score of 5.34 would indicate “mild hyposmia.” When the numbers of TL and control subjects that fall into each category of orthonasal function were compared, however, a significant difference was not found (P = .06). These results are summarized in Table 3.
Measurement of retronasal olfaction likewise showed that subjects who had undergone laryngectomy had decreased retronasal olfactory ability compared with controls. The mean score for the TL group was 11.00 of 20 correct (range, 6.00-17.00), whereas the mean score for the control group was 14.33 of 20 correct (range 7.00-19.00) (P<.001). When compared with the scores on retronasal testing reported by Heilmann et al12 for normosmic, hyposmic, and anosmic subjects, scores for the TL group fell between mean reported scores for anosmic and hyposmic subjects, whereas the mean score for the control group fell between mean reported scores for hyposmic and normosmic subjects.12
Finally, the mean self-rating score of smell ability was 2.94 for the TL group and 6.58 for the control group (P<.001). The mean self-rating score of taste was 5.75 for the TL group and 6.86 for the control group (P = .06). Self-rating score of smell correlated with orthonasal (rS = 0.42; P<.001) and retronasal olfactory ability (rS = 0.50; P<.001). Self-rating score of taste did not correlate with orthonasal olfactory ability (rS = 0.17; P = .15,) but did correlate with retronasal olfactory ability (rS = 0.28; P = .02).
No significant correlation was found between time in years since laryngectomy and orthonasal (P = .79) or retronasal (P = .77) psychophysical testing scores. Similarly, history of radiation therapy was not found to influence scores on olfactory testing. Within the TL group, the mean score on orthonasal testing for subjects with a history of radiation treatment was 4.30 of 7; the mean score on orthonasal testing for subjects without a history of radiation treatment was 4.35 (P = .50). Within the control group, the mean score on orthonasal testing for subjects with a history of radiation treatment was 5.31; the mean score on orthonasal testing for subjects without a history of radiation treatment was 5.36 (P = .82).
Olfaction in patients who have undergone laryngectomy remains poorly defined. The present investigation has shown that such individuals have measurable deficits of olfactory function when compared with age-, sex-, and smoking history–matched control subjects. This study provides a unique characterization of patients who have undergone laryngectomy because scores on both orthonasal and retronasal olfactory testing are reported. Laryngectomy results in measurable deficits in olfactory function that in this study correlated with subjective assessments of smell ability.
While the pathophysiologic features of decreased olfaction in patients who have undergone laryngectomy has not been well established, several theories have been proposed. Patients who have undergone laryngectomy obviously lack nasopulmonary airflow and as such cannot “sniff” in the usual sense. It has previously been suggested that complex neural pathways critical to olfaction are interrupted by surgical denervation of the larynx.16 Miani et al17 proposed that degeneration of the olfactory epithelium is due to atrophy, inflammation, or disuse. These investigators noted more severe degeneration of the olfactory epithelium in biopsy samples taken from subjects who underwent laryngectomy compared with controls.17 However, there was a greater number of smokers in the laryngectomy group in this study. Because smoking has been associated with decreased olfactory function,6- 8 the olfactory epithelium degeneration may have resulted from exposure to cigarette smoke and not from laryngectomy. Our results indicate that severe degeneration of the olfactory epithelium in all patients who underwent laryngectomy is unlikely, since 5 of 36 subjects in the TL group achieved scores within the normal range for orthonasal olfaction. The theory of disuse atrophy is also called into question by our results because an increased length of time between laryngectomy and olfactory testing was not associated with poorer olfactory ability. Overall, the TL subjects demonstrated moderate hyposmia rather than anosmia. Consequently, although the ability to smell is impaired, residual olfactory function persists after laryngectomy.
These results are clinically important because patients who underwent laryngectomy who have olfactory deficits identified by self-report and psychophysical testing could potentially benefit from referral for olfactory rehabilitation. Speech pathologists have developed a “nasal airflow-inducing maneuver” that can be taught to patients after laryngectomy. Regular use of the nasal airflow-inducing maneuver has been shown to improve both subjective and objective measures of olfactory function.3,18,19
The tests of orthonasal and retronasal olfaction used in this study can be assembled using easily accessible items and are simple to administer and interpret. However, full testing requires approximately 1 hour, which is a significant time investment in a busy clinical practice. A more rapid screening to identify olfactory deficits after laryngectomy would be preferable. In this study, we found a significant correlation between self-rating scores of smell ability and psychophysical test scores. Potentially, patients could be selected for referral for olfactory rehabilitation based on their responses to a simple visual analog scale of perceived olfactory ability. However, a recent study found no correlation between self-reported ability to smell and measured olfactory ability.20 Further research will be necessary to evaluate the visual analog scale used in this study in terms of its validity and ability to reliably predict olfactory function after laryngectomy.
Although the findings of this investigation provide useful information, there are several limitations. First, insufflation is a nonphysiologic way to deliver odorants to the olfactory cleft. The nonlaryngectomy subjects routinely “sniffed” deeply to sample the odors, whereas the laryngectomy group could not and had to insufflate the odors by squeezing the bottles. Research has shown that the “sniff” is an integral part of odorant delivery to the olfactory epithelium as well as subsequent olfactory neural processing.21 This problem could have been avoided with the use of a device called the “larynx bypass,” which re-creates nasopulmonary flow. As originally described, the bypass is fashioned from a length of plastic tubing connecting the laryngostoma to the mouth, whereby patients can “sniff” using respiratory musculature.22 At present, patients can read on the Internet about how to make and use a similar device from plastic tubing and a baby bottle nipple.23 Such a device was not used in the present study, and practice is required to use the device correctly. Another limitation of this study is that since salivary flow production was not measured, differences in saliva production between the 2 groups could have affected the results. Delivery of odorant molecules via the retronasal route to the olfactory epithelium is more efficient when the powders are dissolved, so radiation-induced xerostomia could have adversely affected odorant delivery in the TL group. Mouth rinsing between each item tested was intended to minimize any difference in salivary flow between subjects, but the unmeasured impact that xerostomia had on retronasal olfaction in this investigation remains unclear.
The impact that sinonasal pathologic features, such as mechanical obstruction of the olfactory cleft, allergic rhinitis, or chronic sinusitis, has on olfaction is not clear because this information was accrued from patient interviews and the medical record. Rhinoscopy was not performed prior to testing, and cases of anatomical nasal obstruction in the 2 groups may have been overlooked. Furthermore, the effect of radiotherapy on olfactory ability cannot be measured in this study, since only 5 patients in the TL group did not receive radiation.
In conclusion, patients who have undergone laryngectomy have decreased olfactory ability and report lower self-rating ability to smell compared with patients without laryngectomy. Awareness is growing among clinicians of the negative impact that disorders of taste and smell can have on quality of life for patients with head and neck cancer. A simple 1 to 10 visual self-rating scale of ability to smell could be used to identify patients who might benefit from olfactory rehabilitation. Moreover, the psychophysical tests of orthonasal and retronasal function described in this study could easily be used by others to study olfactory ability in patients after laryngectomy.
Correspondence: John W. Werning, MD, DMD, Department of Otolaryngology, University of Florida College of Medicine, PO Box 100264, Gainesville, FL 32610-0264 (email@example.com).
Submitted for Publication: May 2, 2006; final revision received July 26, 2006; accepted September 14, 2006.
Author Contributions: Dr Leon 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: Leon, Catalanotto, and Werning. Acquisition of data: Leon and Catalanotto. Analysis and interpretation of data: Leon, Catalanotto, and Werning. Drafting of the manuscript: Leon, Catalanotto, and Werning. Critical revision of the manuscript for important intellectual content: Leon, Catalanotto, and Werning. Obtained funding: Catalanotto. Administrative, technical, and material support: Leon, Catalanotto, and Werning. Study supervision: Leon and Werning.
Financial Disclosure: None reported.
Funding/Support: All funding was from the Department of Otolaryngology, University of Florida.
Role of the Sponsor: The funding source had no role in the design and conduct of the study; in the collection, management, analysis, and interpretation of the data; and in the preparation, review, and approval of the manuscript.
Acknowledgment: We acknowledge the contribution of Linda Bartoshuk, PhD, who provided the original idea for the study.