Anosmia after general anaesthesia: a case report


Correspondence to: Dr Evangelia Tsakiropoulou


Although anaesthetic drugs are included among the aetiological factors of anosmia, limited reports exist of anosmia induced by general anaesthesia. We present the case of a 60-year-old female patient with a 3-month history of altered smell and taste immediately after recovery from general anaesthesia for a urological operation. The anaesthetic drugs used were fentanyl, propofol and sevoflurane. Clinical examination and a computed tomography brain scan did not reveal any pathology. Psychophysical testing showed anosmia and normal taste function. Imaging studies using single photon emission computed tomography of the brain were performed twice: as a baseline examination; and after odour stimulation with phenyl ethyl alcohol. Normal brain activity without reaction to odorous stimuli suggested peripheral dysfunction or stimuli transmission problems. The patient, after four months of olfactory retraining, demonstrated significant improvement. The onset of the dysfunction in relation with the imaging findings may imply that anaesthetics could induce the olfactory dysfunction.

Olfactory dysfunction occurs at a rate of approximately 5% in the general population [1]. Aetiological factors include chronic rhinosinusitis, viral infections, craniofacial trauma, neurological disorders, systemic disorders and drugs [1]. A large number of drugs are reported in the literature as being responsible for olfactory disorders; however this side effect has mainly been observed in animal study populations [2], although a few clinical case reports are present on the subject [3]. For anaesthetic drugs specifically, the issue is still controversial. There are publications regarding the effect of local anaesthesia on olfactory function [4], but only a few on the effect of general anaesthesia [3]. This complication is rare, but when surgery involves the brain, head or neck, there is a known potential risk for olfactory impairment. However, in these cases it is difficult to confirm a cause and effect relationship as surgery, anaesthetics or even both can harm the olfactory system.

The pathophysiology of chemosensory dysfunction after systemically injected or topically applied drugs is largely unclear. Currently, little is known about the sites at which such drugs exert their influences. Thus anaesthetic drugs may affect olfaction at the central nervous system or at a peripheral level via nasal vasodilatation, mucus hypersecretion, and toxicity damaging the olfactory neuroepithelium [5]. Establishing a direct relationship between drug exposure and anosmia is often difficult, and it is frequently hard to determine with certainty the causative role of specific drugs in the development of such a disorder. In this case we report a case of anosmia developing after an operation not involving areas at potential risk for olfactory disturbance; it occurred immediately postoperatively, with findings suggesting general anaesthesia as a causative factor.

Case report

We present the case of a 60-year-old non-smoking female patient who underwent a tension-free vaginal tape procedure for treatment of female urinary incontinence. Following recovery from general anaesthesia she reported acute onset parosmia (distorted smell perception), anosmia and dysgeusia (distorted taste) for which she required referral to the smell and taste outpatient clinic. Regarding the parosmia, the patient reported an unpleasant smell of benzene while eating vegetables.

The original anaesthetic consisted of 100 μg fentanyl and 180 mg propofol administered intravenously for induction. This was followed by the administration of sevoflurane (up to twice the minimum alveolar concentration) for maintenance of anaesthesia.

Her medical history was free from major causes of olfactory dysfunction such as head trauma, chronic rhinosinusitis, recent upper respiratory tract infection, and allergy or neuro degenerative diseases. The patient was under medical treatment for hypothyroidism; however, her blood levels of thyroid hormones were within the normal range both before and after surgery. Her family history was negative for disorders of smell and taste or other neurological conditions.

The patient had a complete otorhinolaryngological examination including nasal endoscopy. The clinical evaluation did not reveal any nasal pathology. The severity of olfactory dysfunction was assessed by means of the validated and reliable ‘Sniffin’ Sticks’ test (Burghart, Wedel, Germany) [6]. Odourants were presented in commercially available felt-tip pens. The examination involves tests for phenyl ethyl alcohol odour thresholds, odour discrimination, and odour identification [6]. The sum of the three sub-tests presents as a composite threshold-discrimination-identification (TDI) score. Based on a multicentre investigation of more than 1000 patients with olfactory loss, patients with a score below 15 are considered functionally anosmic, and those with scores between 15 and 30 as hyposmic [6]. The total TDI score in our patient’s case was nine (anosmia). The individual scores were: threshold 2/16; discrimination 5/16; and identification 2/16.

For the assessment of gustatory function the validated Taste Strips test (Burghart) was applied [7]. This test entailed the use of paper strips impregnated with a taste stimulant, placed on the middle of the tongue following which the patient was asked to close her mouth and choose one of five possible taste experiences (sweet, sour, salty, bitter, no taste). The taste strips were presented randomly in stepwise increasing concentrations. The score obtained was 12/16 indicating normal gustatory function (scores < 8 being considered as loss of taste); as such her taste complaints were attributed to the olfactory dysfunction.

Imaging studies included a computed tomography scan (CT) of the brain and nasal/paranasal cavities. No evidence of intracranial haemorrhage or other central nervous pathologies were identified. The nasal and paranasal cavities were free of pathology with patent olfactory clefts in both sides.

Since CT scans demonstrate only structural abnormalities, we performed a single photon emission computed tomography (SPECT) of the brain in order to obtain functional information regarding cerebral metabolism and perfusion. Our interest was in detecting areas in the brain that could be related to olfactory impairment (e.g. orbito-frontal cortex, temporo parietal region, olfactory bulbs and tracts, cingulated area). A baseline scan was performed following intravenous injection of 20 mCi of 99mTc HMPAO (an agent used to label leucocytes) and the brain activity was recorded. The same procedure was repeated 48 h later, while supra-threshold stimulus of phenyl ethyl alcohol (rose smell) was delivered in both nostrils. Olfactory stimulation was performed during normal breathing for a total of 7 min with odour presentation for 2 s at intervals of 5 s to avoid adaptation effects. Seven minutes following the olfactory stimulus, 20 mCi of 99mTc HMPAO was injected intravenously; image acquisition was performed 60 min after the injection. The evaluation of the post-stimulation values for each segment was expressed as a fraction of the corresponding prestimulation values: [(post-stimulation counts minus prestimulation counts)/prestimulation counts] × 100. A difference of > 25% was considered as significant.

No areas of hypoperfusion were detected in the baseline SPECT scan. The comparison of values between baseline and poststimulation scans showed no significant difference in brain activity (Fig. 1).

Figure 1.

 Single photon emission computed tomography (SPECT) brain scan. (a) The baseline test showed normal brain function in the regions of interest (orbitofrontal, temporoparietal, cingulate areas). The comparison was based on mean pixel value of the occipital lobe (reference area). (b) The postolfactory stimulation SPECT showed similar results in comparison with the baseline test with no significant changes (> 25% counts) in the regions of interest.

The patient followed a scheme of olfactory retraining using four aromas (rose, mint, eucalyptus, lemon) twice a day according to the recommendations of Hummel et al. [8]. At the patient’s follow-up appointment, 4 months later, evaluation of olfactory function showed significant improvement, with a TDI score of 25 (threshold 3/16, discrimination 10/16 and identification 12/16). According to these results the patient was considered as hyposmic. In addition, the patient did not report any further presence of parosmia.


Many drugs produce a temporary or permanent harmful effect on olfaction. The pattern of action for most of the drugs remains unclear despite their well-documented effect on chemical senses. Regarding anaesthetic drugs, while there are reports and experimental studies regarding the influence of topical anaesthetics on olfactory function [4], the impact of drugs used in general anaesthesia has not been determined.

Salvinelli et al. reported the case of permanent olfactory dysfunction after endonasal local anaesthesia with lidocaine 4% due to contact of the anaesthetic with the olfactory epithelium [9]. Welge-Lussen et al. demonstrated a temporary reduction in olfactory function by psychometric testing (an increase in the olfactory threshold and a decrease of discrimination ability) caused by local anaesthesia applied in the middle nasal meatus [4]. When local anaesthetic (4% lidocaine) was applied directly to the olfactory cleft by means of a dropper, the side effects of headaches and transient local inflammation were observed.

To our knowledge, only a few reports exist in the literature regarding the effect of general anaesthesia on olfaction. Where these occur, controversy exists as to the relationship between cause and effect [3, 10, 11].

The significance of the presented case lies in the acute onset of olfactory dysfunction immediately after recovery from anaesthesia, with the surgery being unrelated to the anatomical location of the olfactory system. It is interesting to note that patients who complained of olfactory dysfunction after anaesthesia report its onset sooner compared with patients suffering from olfactory disorders of different aetiology, although the time between onset of dysfunction and presentation to the clinic may be similar [3].

Additionally, the imaging studies of our patient demonstrated no hypo-function or structural changes of the central olfactory system, indicating a potential peripheral type of dysfunction or an olfactory stimuli transmission problem e.g. at the olfactory bulb level.

The course of anosmia in our patient was transient as she presented significant improvement at her last follow up appointment, seven months following the onset of her anosmia, and she was satisfied with her olfactory ability. However, anosmia after general anaesthesia can be permanent as mentioned in previous reports [10, 11].

Sevoflurane (fluoromethyl 2,2,2-trifluoro-1-(trifluoromethyl) ethyl ether) is a volatile liquid for inhalation, with a non-pungent odour. For both the drug and its metabolites, the dose and time of exposure at which toxicity occurs in humans are not known. The actual relationship of sevoflurane to olfactory dysfunction cannot be established with certainty. However, the volatile form of the substance and the fact that it can directly affect the olfactory epithelium make it a potential source of the peripheral type of dysfunction.

In human studies, Salmi et al. demonstrated that sevoflurane and propofol could affect subcortical and cortical γ-aminobutyric acid (GABA) receptor ligand binding [12]. In the human brain, GABA is the main inhibitory neurotransmitter, and its participation has been invoked in deafferentation-induced plasticity. It is active at many areas of the olfactory system such as the cingulated area and synapses of the olfactory bulbs. The cingulate area plays an important role in modulating central nervous system input and demanding cognitive processing and working memory [13]. In a study by Levy and Henkin, it was found that patients with phantosmia had significantly decreased brain GABA levels [13]. Thus any factor that affects the GABA dependant pathways could potentially affect olfactory stimuli transmission to the central structures of the olfactory system.

Regarding propofol, a case of temporary anosmia was reported as a side effect after recovery from general anaesthesia in a man who was also under treatment with other drugs [14]. Besides this report, no further causative explanation was mentioned. An animal study examining the behavioural and electrophysiologic effects of four anaesthetic agents infused into the cerebral ventricles of conscious rats demonstrated depressed olfactory responses following propofol and fentanyl, with propofol producing the strongest effect [2]. Another study in humans assessing the behavioural and cognitive effects of subanaesthetic concentrations of isoflurane and nitrous oxide showed a transient sense of an unpleasant odour [15].

Standard brain and sinus imaging studies (CT scan, MRI) of patients with loss of smell are usually normal except for a relatively small number of cases with nasal pathology, e.g. sinusitis, nasal polyps, or tumours of the nasal cavity, neuroepithelium and olfactory bulbs. To investigate the underlying pathology, functional imaging methods such as functional MRI [16], SPECT [17] and PET [18] have been developed to evaluate olfactory ability. The current available techniques do not provide all the information needed and require further development; however, only functional imaging can investigate these disorders in detail. The precise identification and localisation of the affected areas within the olfactory system is mandatory in order to clarify the nature of the drug-induced olfactory disorders and to evaluate the results of therapy.

The presented case implicates general anaesthetic drugs in the causation of this patient’s olfactory dysfunction. The role of each anaesthetic drug used remains undetermined. Further experimental studies are needed in order to assess the effect of specific anaesthetic drugs on olfaction. Only a high index of clinical suspicion and the detection of similar cases from other clinicians and case reports will help to measure the incidence of these side effects, and to examine whether the problem is sufficient to warrant specific mention in the process of consent for anaesthesia if specific drugs are to be used.


Consent for publication granted by the patient.