Treatment of Olfactory Disorders After SARS – CoViD 2 Virus Infection

Introduction

Even if the pandemic currently seems to have less of an impact on daily life, the long-term effects of coronavirus disease 2019 (COVID-19) remain a challenge in medical practice and therefore a topic of scientific interest. ¹ Impairments of the human sense of smell as a result of a corona virus infection were reported from the beginning of the pandemic.^2-5 Though many patients seem to recover their sense of smell after a few weeks, a certain proportion of them develop a long-lasting olfactory disorder post-COVID-19.^4,6-8 In addition to hyposmia and anosmia, persistent parosmia is reported with striking frequency.^9,10 Less surprising is that olfactory dysfunction is commonly associated with gustatory impairments.^11-13

Even vaccinations cannot completely protect against an infection and thus against possible disorders of the olfactory function. ¹⁴ As of October 2022, there is disagreement about the therapeutic options for post-viral olfactory disorders such as those resulting from Coronavirus Disease 2019. ¹⁵ In our opinion, we are still a long way from a rational therapy recommendation that can claim to be a standard. ¹⁶ The commonly used therapy with a topical nasal corticosteroid (TNC) has an ambiguous grade of recommendation.^17,18 A positive effect of olfactory training (OT) on the recovery from post-infectious olfactory disorders has already been shown in studies.^19-21 This study aims to compare these two therapeutic approaches to the treatment of persistent olfactory dysfunction. Most data on loss of smell rely on the evaluation of orthonasal olfactory performance. An extended assessment also seems advisable for the treatment concept. Therefore, we also aimed to assess retronasal olfactory function over the study period. Based on the current state of the literature, we considered a parallel further control group of patients without any therapy recommendation to be ethically unacceptable.

Material and Methods

Statistical analysis

All analyses were performed using the Statistical Package for the Social Sciences (SPSS Statistics for Windows, Vers. 28, IBM, Armonk, NY, USA). Data are presented as mean ± standard deviation (SD) or median and quartile. A probability value of less than 0.05 was considered statistically significant. Mixed-Design ANOVA was used for a combination of ordinary, variance and covariance analysis with measurement repetitions. To explore the evolution of quantitative variables (T, D, I, TDI and R score) before, under and after the OT, we employed several tests. As confirmed by Lilliefors test and Shapiro–Wilk test (p > 0.05), we can largely assume a normal distribution with a few exceptions. For groups of exactly the same size, we consider the slight deviations from the normal distribution for the ANOVA to be acceptable. ²⁵ Sphericity can be assumed according to the Mauchly test (p = 0.062). There was homogeneity of error variances as assessed by the Levene test (p > 0.05) as well as a homogeneity of the covariances as assessed by the Box test (p = 0.153).

The influence of other factors (age, gender, duration of the olfactory disorder, duration of treatment, baseline value of olfaction) was modeled additionally using a regression analysis among dependent variables. All patients performed the first follow-up and 16 of 20 the second follow-up after baseline. In the three cases (2 in the OT group and one in OT + TNC group), it has been justified to carry forward the last observation (LOCF). Identical values were added for the last visit, assuming that the already significantly improved olfactory ability remained at least the same as the previous visit (Figure 1).

OPEN IN VIEWER

Figure 1.

Evolution of olfactory disorder due to COVID-19: Evolution of olfactory disorder due to COVID-19 in dysosmic patients between December 2020 and July 2021 before, under and after olfactory training with (OT plus TNC, n = 10) or without (OT, n = 10) a concomitant topical nasal corticosteroid therapy (TNC). TDI score = Threshold–Discrimination–Identification score; Colored parts cover anosmic (red), hyposmic (yellow) or normosmic (green) patients according to TDI normative; V1 = Baseline; V2 = Follow-up after 2 months; V3 = final visit after 3 months.

Results

Clinical observations during the study period already indicated that rates of chemosensory dysfunction associated with COVID-19 infection may be decreasing. This was later confirmed in studies and could be explained by the appearance of new subtypes of lower virulence.^4,26 Especially after multiple viral infections, it became more difficult to clearly trace the cause of olfactory disorder to a specific subtype. Our priority in the study design was the best possible comparability of the randomized groups (Table 1).

OPEN IN VIEWER

Table 1.

Characteristics of the treatment groups at screening V1. The clinical-demographic data after randomization suggest a good comparability between the two groups.

			Treatment group
			OT	OT + TNC
Sex	female	Count	8	6
	male	Count	2	4
	Total	Count	10	10
Age	Mean		38	30
	Minimum		21	20
	Maximum		55	46
	Standard Deviation		13	9
TDI Score V1	Mean		26	25
	Minimum		15	11
	Maximum		37	33
	Standard Deviation		6	7
Time of smell disorder in weeks	Mean		21	20
	Minimum		6	4
	Maximum		44	43
	Standard Deviation		15	13

There was no statistically significant interaction between time and group, F(2) = 1.042, p = 0.363, partial η² = 0.055. Effect sizes are also considered by the use of eta-square after Cohen (1988). A value of more than d = 0.2 represents small effects, more than d = 0.5 average effects and more than d = 0.8 large effects. According to Cohen, a moderately interaction effect (eta ² = 0.055, Cohen`s d = 0.5) was found between time and group. While the TDI score steadily increased on average under the combination therapy, the rise under olfactory training alone was initially steeper (Figure 2). After all, a significant main effect for time F(2, 36) = 8.05, p < 0.001, partial η² = 0.309 can still be reported. The undoubtedly effective factor of spontaneous improvement should statistically benefit both arms of therapy equally and is therefore neglected. Age, gender, the duration of the post-viral olfactory disorder and the total duration of therapy did not show any influence on the change in odor (ΔTDI, ΔT, ΔD, ΔI, ΔR). The subjective improvement in retronasal olfactory ability also did not differ significantly according to the treatment arm. We recognize that despite robust statistical analysis, their performance is not sufficient to rule out these factors entirely. These results should therefore not be overestimated, but should be examined more closely with a larger number of cases. However, the better the initial baseline value of olfaction at screening visit (TDI score), the clearer the improvement in smell up to the first follow-up. Side effects of the therapy were not reported in any case, neither due to olfactory training nor as a result of topical steroid application.

OPEN IN VIEWER

Figure 2.

Development of mean TDI Scores: Development of mean TDI Scores over mean 3 months in dysosmic patients due to CoViD-19 under olfactory training (OT) with versus without a concomitant topical nasal corticosteroid therapy (TNC).

Discussion

The treatment of olfactory disorders after a viral infection such as SARS-CoViD 2 is a challenge and a scientifically relevant topic. Although the majority of patients recover completely after a few weeks, a small proportion displays persistent disturbing symptoms such as hyposmia and parosmia.^{4,6,7,9,10,13,27} On the global scale of a pandemic, however, even a small proportion corresponds to a large number of people affected, each of whom is significantly impaired in their quality of life.^4,28 There is no denying that the therapeutic options are not yet satisfactory. ²⁹ The knowledge gained recently could be groundbreaking and serve as a precedent for many other post-viral olfactory disorders that are far less in focus. The crucial thing is that the cause and the temporal connection of a smell and taste disorder can rarely be assigned as precisely as after a proven corona infection. Based on the time interval of enrollment in this study, it is highly probable that the infections were exclusively caused by a non-omicron virus variant. This is certainly one of the strengths of this study. During the infections, recruitment, screening and follow-up examinations of all subjects, the delta variant of the coronavirus SARS-CoV-2 dominated. ³⁰ Although the sense of smell improved significantly in the majority of cases in this comparative study, we expected the effects of the two therapeutic approaches to be more markedly different. The initial advantage with sole olfactory training could be explained by a possibly higher compliance at the beginning of the sole training due to the lack of further drug treatment offers. When the training intensity decreases, the recovery of the sense of smell stagnates. This follow-up observation was reflected in all models, which we consider crucial for our study. Adjunctive therapy ultimately outweighs this short-term benefit. It is also known that compliance is already insufficient with the regular use of pharmaceuticals, and it is even lower with non-drug therapy recommendations such as smell training. Finally, the insufficient distinction in diagnosis between anosmias and parosmias needs to improve.^31,32 The undoubtedly different olfactory disorders should be assessed by extended test methods.

Conclusion

In any case, this study reinforces the recommendation of consistent olfactory training starting as early as possible for patients with dysosmia due to COVID-19. This has proven to be a therapeutically and economically favorable treatment with no side effects for any type of post-infectious olfactory disorder. A therapeutic advantage of a topical nasal corticosteroid was not shown, but cannot be ruled out either. In future investigations with larger cohorts, treatment efficacy could be augmented by established inhalative device-based applicators, which would also improve assessment and compliance. Likewise, smell training could be intensified, monitored and thus standardized using suitable aids. After all, it is long overdue to find a way to better differentiate between hyposmias and parosmias diagnostically. For this purpose, an extended procedure with new subjective and objective olfactometric and imaging tools is being developed.