New loss of smell and taste: Uncommon symptoms in COVID-19 patients on Nord Franche-Comte cluster, France.

BACKGROUND: New loss of smell or taste was not included as common symptoms of COVID-19 until March 2020 when the pandemic started in Western countries. We want to describe the prevalence and features of anosmia and dysgeusia in COVID-19 patients.
METHODS: We retrospectively investigated the clinical features of confirmed cases of COVID-19 in Nord Franche-Comté Hospital, Trevenans, France, between March, 1st and March, 14th 2020. We used SARS-CoV-2 real time RT-PCR in respiratory samples to confirm the cases.
RESULTS: Of 70 patient enrolled, the mean age was 57.0 years and 29 patients (41%) were men. Median Charlson comorbidity index was 1.70(±2.5). Twenty-seven (39%) patients had pneumonia. Fatigue (93% [65]), cough (80% [55]) and fever (77% [54]) were the three main symptoms. Neurologic symptoms were present in more than half of the patients: anosmia (53% [37]) and dysgeusia (48% [34]). The mean duration of anosmia was 7.4 (±5, [1-21]) days, 51% (36/70) recovered before 28 days of evolution. Only one patient with anosmia had not recovered at the end of the follow-up. Patients with anosmia had less often a pneumonia (10/37 vs 17/33, p = 0.036), were less often hospitalized (13/37 vs 20/33, p = 0.033) and needed less often oxygen therapy (6/37 vs 17/33, p = 0.002) than patients without anosmia. There were no statistically differences for viral load between patients with anosmia and patients without anosmia (5.5 [2.0-8.6] vs 5.3 [2.1-8.5] log copies/ml respectively, p = 0.670). The fatality of COVID-19 in our study was 6% with four deaths.
CONCLUSIONS: Anosmia and dysgeusia are present in half of COVID-19 patients. The mean duration of anosmia was 7 days and the outcome seems favorable in less than 28 days.

Summary of the article’s main point

We included 70 patients infected with SARS-CoV-2. Thirty-seven (53%) patients had an anosmia which was associated with dysgeusia in 81% of cases. The mean duration of anosmia was 7 days and 51% (36/70) recovery before 28 days of evolution. Only one patient with anosmia had not recovered at the end of the follow-up.

Introduction

An outbreak of pneumonia began in December 2019 in Wuhan (China) (Zhu et al., 2020), a novel coronavirus was identified as causal agent (Lu et al., 2020), named later the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The 11th of March 2020 the WHO General Director announced that coronavirus disease 2019 (COVID-19) was the first pandemic caused by a coronavirus (WHO, 2020).

Clinical description from COVID-19 outbreak in China reveals that most of the patients (81%) have minor symptoms or an influenza like illness (ILI) or a mild pneumonia and 19% have a severe or critical pneumonia (Wu and McGoogan, 2020). Clinical descriptions show that fever, cough, fatigue and myalgia are usually the main symptoms; the expression of the COVID-19 ILI seems non-specific, no specific symptoms can direct to suspect a case without notion of exposition (Wang et al., 2020a, Huang et al., 2020, Wang et al., 2020b, Wu et al., 2020, Guan et al., 2020, Chen et al., 2020).

A major French cluster of COVID-19 began on March 1st 2020, in Mulhouse city (less than 30 miles from our hospital). After clinical examination of the first patients, we noticed that many cases described a new loss of smell or taste. In our knowledge, anosmia and/or dysgeusia have never been reported in medical literature. Information about these two neurologic symptoms with SARS-CoV-2 infection is scarce.

The concept of anosmia after a viral infection is known as post-infectious/post-viral olfactory loss (POL). Different kind of viruses can induce a POL, including coronaviruses such as HCoV-229E (Suzuki et al., 2007).

In this study we want to describe the prevalence and features of anosmia in COVID-19 patients.

Methods

Less than 30 miles from our hospital (Nord Franche-Comté (NFC) Hospital), Trevenans, a major French cluster of COVID-19 began on March 1st, 2020 in Mulhouse city. We conducted a retrospective observational study between March, 1st and March, 14th 2020, we enrolled all adult patients (≥18 years) with confirmed COVID-19 who were examined at the infectious diseases consultation or hospitalized in the hospital. Pregnant women, children (<18 years) and patients with dementia (who can’t report functional symptoms) were excluded. We stopped the follow-up of the study on March, 24th 2020.

Diagnosis was confirmed by real-time RT-PCR on respiratory samples, mainly nasopharyngeal swabs, sputum, bronchial aspirates or bronchoalveolar lavage fluids. Briefly, viral RNA was extracted using the NucleoSpin® RNA Virus kit (Macherey-Nagel) according to the manufacturers’ instructions, and amplified by RT-PCR protocols developed by the Charité (E gene) (Corman et al., 2020) and the Institut Pasteur (RdRp gene) (Bernard Stoecklin et al., 2020) on LightCycler 480 (Roche). Quantified positive controls were kindly provided by the French National Reference Center for Respiratory Viruses, Institut Pasteur, Paris.

Data required for the study was collected from the medical files of patients: age, sex, comorbidities (Charlson score, HTA, diabetes, cardio-vascular disease defined by –cardiac failure, cardiac arrhythmia, coronary heart disease, stroke, peripheral arterial obstructive disease and thromboembolic disease– chronic obstructive pulmonary disease, asthma, malignancy, immunosuppression defined by –transplantation, cirrhosis, long-term steroids and immunomodulators treatments–), current smoking, health care worker. We noted the presence or not of anosmia and/or dysgeusia, the features of anosmia (date of apparition since the onset symptoms, duration of anosmia). Respondents rated smell and taste as “Good” or “Present” (considered as patients without anosmia/dysgeusia) or “Poor” or “Absent” (considered as patients without anosmia/dysgeusia). We checked whether these symptoms exist before infection. Patients were illegible to differentiate dysgeusia with loss of appetite. We also checked that the patients were not taking any medicine that could cause an alteration of smell or taste. Other signs collected: fever > 38 °C, feeling of fever, highest temperature, fatigue, myalgia, arthralgia, sore throat, headache and its localization (diffuse, frontal, other localization), rhinorrhoea, nasal obstruction, epistaxis, dysgeusia, tinnitus and hearing loss, conjunctival hyperemia, tearing, dry eyes, blurred vision, sneezing, cough, sputum production, haemoptysis, dyspnea, respiratory rate > 22, sat 02 at admission, auscultation with crackling sounds, nausea, vomiting, diarrhea, abdominal pain. We collected also the viral load in respiratory samples and the outcome: hospitalization or not, necessity of oxygen therapy, hospitalized in intensive care unit and death.

A home follow-up is recommended in our national guidelines, for patients who were not hospitalized, until they are asymptomatic for more than 48 h (DICOM_Lisa, 2020). We collected the data from the first contact with the patient at the hospital and during the follow-up. We followed up prospectively each patient until they were asymptomatic. Practically, patients who were not hospitalized were called 7 days (±7 days) after the first symptoms and every week until their recoveries.

In this work, we aim to describe the prevalence and features of anosmia in COVID-19 patients. We divided patients in two groups: patients with anosmia and patients without anosmia in order to compare the characteristics (comorbidities, clinical features and outcome) of patients with anosmia versus (vs) patients without anosmia.

For the statistical analysis, usual descriptive statistics were used. All variables were assessed using a univariate analysis. Continuous variables were expressed as mean and standard deviation (SD) with ANOVA test. Categorical variables were expressed as numbers, percentages, or mean and compared by χ2 test or Fisher's exact test between the two groups (patients with anosmia and patients without anosmia). A p-value < 0.05 was considered significant. We used the SPSS v24.0 software (IBM, Armonk, NY, USA).

Results

Features of the population and neurologic symptoms

70 patients were included in this study. The mean age was 57.0 (±19) years old and 29 (41%) were male. The median Charlson comorbidity index was 1.70 (±2.5). Seven symptoms were present in more than half of the patients: fatigue (93%, n = 65), cough (80%, n = 55), fever (77%, n = 54), headache (73%, n = 51), myalgia (59%, n = 41), arthralgia (54%, n = 38) and anosmia (53%, n = 37) (Table 1 ). Twenty-seven (39%) patients had an auscultation with crackling sounds with a diagnostic of pneumonia. Tinnitus seemed uncommon (10%, n = 7) and hearing loss were scarce (6%, n = 4). The most common neurologic symptom was anosmia (53%, n = 37) followed by dysgeusia (48%, n = 34). The mean duration of anosmia was 7.4 (±5, [1-21]) days. Forty percent (15/37) had a duration ≥ 7 days and 16% (6/37) a duration ≥ 14 days. One patient (1/37) had not recovered at the end of the follow-up (after 28 days). Anosmia began after 4.7 (±1.5, [1-7]) days of symptoms evolution. ‘In 31 cases (84%), COVID-19 patients has both anosmia and dysgeusia and in 21 cases (57%) had both anosmia and rhinorrhoea; however, only 8 cases (22%) of patients with anosmia presented a nasal obstruction’

Table 1

Comparison of the comorbidities, symptoms and outcome of COVID-19 patients with anosmia and patients without anosmia in Nord Franche-Comte Hospital, 2020.

Characteristics				All patients (n = 70)	Patients with anosmia (n = 37)	Patients without anosmia (n = 33)	p-value
Medical history
Age (Y): mean (SD)				57 (±19)	50 (±16)	64 (±20)	0.001
Sex			Female	41 (59%)	25	16	0.106
			Male	29 (41%)	12	17	0.106
Health care worker				22 (31%)	16	6	0.024
Current smoking				10 (14%)	4	6	0.499
Comorbidities	≥1 comorbidity			36 (52%)	13	23	0.004
	HTA			16 (23%)	5	11	0.016
	Cardiovascular diseasea			15 (21%)	5	10	0.087
	Diabetes			10 (14%)	2	8	0.038
	Asthma			7 (10%)	6	1	0.009
	COPDb			4 (6%)	0	4	0.009
	Malignancy			3 (4%)	1	2	0.599
	Immunosupressionc			3 (4%)	1	2	0.599
Charlson comorbidity index: mean (SD)				1.70 (±2.5)	0.70 (±1.5)	2.8 (±2.9)	<0.001
Neurologic symptoms (anosmia/dysgeusia)
Anosmia				37 (53%)	NA	NA	NA
Dysgeusia				34 (48%)	31	3	< 0.001
General symptoms
Number of symptoms (ORL, other): mean (SD)				9.2 (±3.3)	10.7 (±2,4)	7.5 (±3.2)	< 0.001
Fever measured > 38 °C				54 (77%)	28	26	0.757
Feeling of fever				13 (19%)	8	5	0.487
Highest temperature (T°C): mean (SD)				38.7 (±0.9)	38.5 (±0,8)	39 (±1)	0.042
Fatigue				65 (93%)	34	31	1
Respiratory symptoms
Cough				56 (80%)	30	26	0.811
Sputum production				20 (29%)	12	8	0.449
Haemoptysis				6 (9%)	3	3	1
Dyspnea				24 (34%)	11	13	0.395
Rhinolaryngological symptoms
Rhinorrhea				34 (48%)	21	13	0.147
Nasal obstruction				13 (18%)	8	5	0.463
Sore throat				14 (20%)	10	4	0.144
Epistaxis				3 (4%)	3	0	0.242
Tinnitus				7 (10%)	5	2	0.434
Hearing loss				4 (6%)	3	1	0.616
Pain Symptoms
Myalgia				41 (59%)	26	15	0.035
Arthralgia				38 (54%)	24	14	0.060
Headache		All headaches		51 (73%)	30	21	0.116
		Frontal headache		18 (39%)	14	4	0.027
		Diffuse headache		20 (%)	10	10	1
		Other headached		13 (1%)	6	7	0.760
Ocular symptoms
Conjunctival hyperemia				3 (4%)	1	2	0.599
Tearing				4 (6%)	2	2	1
Dry eyes				3 (4%)	2	1	1
Blurred vision				3 (4%)	3	0	0.242
Sneezing				13 (18)	7	6	0.937
Gastro-intestinal symptoms
Nausea				22 (31%)	12	10	1
Vomiting				2 (3%)	0	2	0.219
Diarrhea				28 (40%)	17	11	0.282
Abdominal pain				14 (20%)	10	4	0.144
Physical examination
Respiratory rate > 22/min				15 (21%)	5	10	0.087
Sat 02 at admission (%)				93 (±3.4)	95 (±3.7)	92 (±3.1)	0.068
Auscultation with crackling sounds				27 (39%)	10	17	0.036
Viral load
Viral load in respiratory samples: mean (range)				5.4 [2.1-8.6]	5.5 [2.0-8.6]	5.3 [2.1-8.5]	0.670
Outcome
Hospitalization				33 (47%)	13	20	0.033
Hospitalization in intensive care unit				7 (10%)	4	3	1
Oxygen therapy				23 (33%)	6	17	0.002
Death				4 (6%)	2	2	1

Defined by: cardiac failure, cardiac arrhythmia, coronary heart disease, stroke, peripheral arterial obstructive disease and thromboembolic disease.

Chronic obstructive pulmonary disease.

Defined by: transplantation, cirrhosis, long-term steroids therapy and immunomodulators treatments.

Twelve patients had retro-orbital headache and 1 patient temporal headache.

Comparison between patients with anosmia and patients without anosmia

Patients with anosmia had a lower Charlson comorbidity index than patients without anosmia (0.70 ± 1.5 vs 2.8 ± 2.9, p < 0.001). In the same way, the frequency of HTA, diabetes, cardio-vascular and pulmonary diseases (except for asthma) was higher in the group of patients without anosmia than in the group of patients with anosmia. Patients with anosmia had more often asthma than patients without anosmia (6/37 vs 1/33, p = 0.009).

Patients with anosmia reported a higher number of symptoms than patients without anosmia (10.7 ± 2.4 vs 7.5 ± 3.2; p < 0.001); they reported more myalgia (26/37 vs 15/33, p = 0.035) and frontal headache (14/37 vs 4/33, p = 0.027) than patients without anosmia. No significant differences were found between the two groups about other functional symptoms. Regarding the physical symptoms, patients with anosmia had less often a pneumonia (10/37 vs 17/33, p = 0.036) and the highest measure of body temperature was 0.5 °C lower than patients without anosmia.

Concerning the viral load in respiratory samples there were no statistically differences between patients with anosmia than patients without anosmia (5.5 [2.0-8.6] vs 5.3 [2.1-8.5] log copies/ml respectively, p = 0.670).

Concerning the outcome, patients with anosmia were less often hospitalized (13/37 vs 20/33, p = 0.033) and when they were hospitalized, they needed less often oxygen therapy (6/37 vs 17/33, p = 0.002) than patients without anosmia.

Discussion

Our population have a mean age of 57.0 (±19) years, 59% were female. HTA, diabetes, cardiovascular and pulmonary diseases have a prevalence ≥ 10% as in other studies. The main symptoms of our population are the same than the symptoms described in other studies except for anosmia and dysgeusia which had never been described in COVID-19 patients from Western countries, to our knowledge (Wang et al., 2020a, Huang et al., 2020, Wang et al., 2020b, Wu et al., 2020, Guan et al., 2020, Chen et al., 2020). Only recently, a study published on April 6 conducted by Lechien et al. reported 357 patients with olfactory dysfunction related to COVID-19 but in a European multicentric study (Lechien et al., 2020). The lack of description about neurologic symptoms in COVID-19, especially in Asia is probably a consequence of the potential gravity of the disease. The description of anosmia and dysgeusia may seem accessory, especially when clinicians deal with critically ill patients (Yang et al., 2020). Furthermore, details of symptoms are difficult to obtain when patients are critical. Other assumptions to explain these differences between Asia and Europe are the theoretical possibility of a mutation of SARS-CoV-2 viral genome and genetic variability between ethnic groups which can explain a polymorphic clinical expression.

Concerning the population, patients with anosmia were younger than patients without anosmia, our patients with anosmia have a mean age of 50 years old and are women in 68% of cases. It’s interesting to notice that in the literature, the same age is described for patients with POL with also a proportion of 70% of women (Harris et al., 2006, Lee et al., 2014). Currently, smoking was not associated with anosmia. In the study of Lechien et al. patients with anosmia had a mean age of 37 [± 11.4] years, without cardiovascular comorbidities. However, their population profile which were ambulatory cases who consulted at ENT consultations.

Asthma is significantly associated with anosmia in our study (Six out of seven patients with asthma had an anosmia), Lechien et al. found also an association between anosmia and asthma. In the literature, we know that persistent asthma has an cumulative impact on the loss of smell in patients with nasal polyposis (Alobid et al., 2011) and asthma is recently identified as a factor driving to olfactory loss in patients with chronic rhinosinusitis (Schlosser et al., 2020); however, to our knowledge asthma is not a risk factor of POL. Otherwise, in chronic rhinosinusitis, hyposmia is more related to mucosal inflammation than to nasal obstruction (Gaines, 2010). Maybe some of our asthma patients had a predisposition with mucosal inflammation due to an allergic rhinosinusitis associated with asthma (we didn’t collect the medical history of chronic rhinosinusitis).

Pathogenesis of anosmia related to COVID-19 is unknow. In our study, In 31 cases (84%), COVID-19 patients has both anosmia and dysgeusia; however, only 8 cases (22%) of patients with anosmia presented a nasal obstruction’. Furthermore, there were no statistically significant differences for nasal obstruction between patients with anosmia and patients without anosmia. This lead to suspect another pathogenesis for anosmia than a nasal congestion with nasal obstruction. In addition, anosmia during a viral rhinitis with nasal obstruction resolve in less than 3 days usually (Akerlund et al., 1995), while in our study the mean duration of anosmia was more than 7 days and half of patients had anosmia for more than 7 days. For example, in patients with POL, the main suspected mechanism is a damage of olfactory epithelium: we know that MRI of the olfactory bulb show a reduction of its volume (Rombaux et al., 2006) and biopsy of the olfactory cleft reveal a diminution of the olfactory receptors (Yamagishi et al., 1994). However, central damage during viral infection are also suspected (Kim et al., 2012). Different kind of viruses can induce a POL, including coronaviruses (Chen et al., 2020). However, there is no description of POL induced by one of the three main highly pathogenic coronaviruses for human being. Our patients with anosmia have a strong association with dysgeusia; dysgeusia is also described with patients with POL (Rawal et al., 2016). Anosmia is significatively associated with frontal headache. The frontal localisation of headache could be due to an acute rhinosinusitis with a mucosal inflammation due to SARS-CoV-2. As discussed above, we suppose that mucosal inflammation can participate in the pathogenesis of anosmia. However, in anosmia related to COVID-19 a neurotropism of SARS-CoV-2 should be discuss, assumptions are an invasion of the olfactory receptors or damage of the first cranial nerves in the nasal cavity cell membrane and/or central lesion; as described in POL with other viruses. Furthermore, there are increasing evidence that coronaviruses are not always confined to the respiratory tract and also invade the central nervous system inducing neurological diseases (Anon, 2017). In addition to smell and taste disorders, patients with COVID-19 may present rhinolaryngological symptoms such as tinnitus and hearing loss. In a recent publication, Kilic et al. concluded that sudden sensorineural hearing loss (SSNHL) may be one of the symptoms of COVID-19 (Kilic et al., 2020) and can be considered also as neurologic symptom. Three mechanisms have been implicated in the occurrence of SSNHL associated with viral infections: neuritis caused by viral involvement of the cochlear nerves, cochleitis due to viral involvement of the cochlea and perilymphatic tissues, and the stress response resulting from the cross-reaction of the inner ear antigens to viral infections (Kilic et al., 2020).’

In contrast with functional symptoms, physical symptoms such as pulmonary parenchyma damage and high fever are less often noticed in patients with anosmia than patients without anosmia. It seems that patients with anosmia have more painful functional symptoms but less frequently a pneumonia due to SARS-CoV-2. They are also less frequently hospitalized and they less frequently need oxygen therapy. Our hypothesis is that anosmia concerns younger patients with fewer comorbidities, thus less risk of severe pulmonary damage.

Patients with anosmia don’t have a higher viral load than patients without anosmia. The intensity of viral load in respiratory samples seems not explain the presence or not of anosmia.

Concerning the evolution, only one patient didn’t recover at the end of the follow-up of the study (after a follow-up of 28 days); eighty one percent of our patients recovered before 14 days of evolution. In medical literature, the evolution of POL can be long: a study with 63 patients who had a POL showed that after one year, 80% of patients reported subjective recovery (Lee et al., 2014). In comparison, evolution of acute anosmia linked to COVID-19 seems most of the time favourable in a short time.

A better knowledge of symptoms of COVID-19 is essential for several reasons. Firstly, to help for detection of COVID-19 case, symptoms of anosmia and dysgeusia are uncommon in influenza infection without nasal obstruction (Souty et al., 2019). With a non-specific ILI the presence of anosmia/dysgeusia can lead to suspect a case of COVID-19, especially when we don’t have the notion of exposition in the beginning of an outbreak. In the other hand, to adapt prevention, during an outbreak of SARS-CoV2, consultation for acute anosmia and/or dysgeusia should lead to suspect a case of COVID-19 with the necessity of hygiene measures to protect doctors and other patients. In another study recently published (Zayet et al., 2020) and including 217 outpatients consulting for a suspicion of COVID-19, we concluded that the specificity of anosmia, dysgeusia was respectively of 85% and 84%. Specificity of the combination of anosmia and dysgeusia reached 91% for a positive PCR result (Zayet et al., 2020). The combination of these 2 symptoms had a positive predictive value of 83% for a positive SARS-CoV-2 RT-PCR result (Zayet et al., 2020).

Finally, to adapt treatment, avoid nose cleaning with physiologic serum and contraindicate utilisation of systemic or local corticosteroids. Indeed, uring the outbreak of COVID-19, some patients complained of associated rhinorrhoea with anosmia and may use physiologic serum for cleaning their noses. The existence of a theoretical risk of increased viral dissemination may lead doctors to tell patients to avoid it and, in addition, this prescription is probably useless because there is usually no associated nasal obstruction. The lack of data with intranasal corticosteroids in COVID-19 and the potential risk of using systemic corticosteroids especially in the first stage of this infection may encourage clinicians to contraindicate corticosteroids in this situation.

One of the limitations of our study is the limited number of patients, a bigger study to confirm and support our results would be interesting. Another limitation is the recruitment with a bias of selection of health workers. Our politic of detection of possible cases of COVID-19 is larger for health workers with possibly less symptomatic form; this might explain the fact that we have more health workers in the group of patients with anosmia. However, this is interesting and reveals that probably most of the cases of anosmia may be seen by general practitioners.

Conclusion

Anosmia can be considered as an unknown neurologic symptom in COVID-19. More than half patients with COVID-19 have an anosmia. Eighty four percent of patient presented both smell and taste disorders. The evolution seems favourable in less than 28 days in COVID-19 patients with anosmia.

Declaration of interests

All authors declare no competing interests. We thank all patients involved in the study.

Funding sources

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethics approval and consent to participate

Due to the retrospective nature of the study, the Ethics & Scientific Committee of Nord Franche-Comté Hospital (Unité de recherche clinique de l’hopital Nord Franch-comté : Clinical Research Unit) determined that patient consent was required. Information consent about study participation was officially announced verbally and noted in writing in the patient’s medical record, according to national regulations for retrospective study. All patients data were anonymized prior to the analysis.