Keywords
olfaction - olfactory dysfunction - post-infectious olfactory dysfunction - COVID-19
- septorhinoplasty
The recent COVID-19 pandemic has left millions of people with a profound loss of chemosensation
due to the high prevalence of olfactory dysfunction (OD) linked to severe acute respiratory
syndrome coronavirus 2 (SARS-CoV-2) infection.[1]
[2]
[3]
[4] Although a high recovery rate has been observed during the first months,[5] up to 13% of subjects can show persistent COVID-19-related OD (C19OD) at 3 years,[6] with severe impact on quality of life (QoL).[3]
[4]
The olfactory pathophysiology of C19OD is multifactorial. Traditionally, causes of
OD have been classidied according to the anatomical location of the presumed pathology/lesion
and divided as conductive, sensorineural, and central, or a combination of these.
SARS-CoV-2 typically damages the olfactory epithelium (OE), thus creating a sensorineural
loss of smell, and to a lesser extent affects the central primary and secondary olfactory
cortices.[7]
[8] Although C19OD represents a reversible sensorineural olfactory loss in the majority
of cases,[6] the question remains as to why this is not the case for those developing a persistent
loss of sense of smell.
Normal nasal airflow through the olfactory cleft is one of the conditions necessary
for an intact olfactory sense.[9] Current evidence shows that a moderate to severe deviated nasal septum (DNS) results
in decreased olfactory function on the obstructed nasal side (lateralized olfaction)
and olfaction is normalized following septoplasty.[10] OD caused by a structural obstruction is mainly caused by a conductive loss secondary
to a reduction in the access of odorants to the OE and once rectified the sense of
smell returns to normal.[11]
[12]
[13]
[14]
[15] The relationship between nasal airway improvement following septal surgery and improved
olfaction has been consistently demonstrated.[9]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
Functional septorhinoplasty (fSRP), as well as correcting a DNS, can also increase
internal/external nasal valve (INV/ENV) function, which is crucial in regulating airflow
to the olfactory region.[28]
[29] With additional INV and ENV augmentation, there is growing evidence suggesting that
fSRP can improve olfaction to a greater extent than septoplasty alone and implying
that other mechanisms, in addition to the conductive component, are involved in the
smell improvement.[10]
[17]
[25]
[26]
[30] In this regard, Whitcroft et al[10] demonstrated that fSRP can improve olfaction in patients with a combination of conductive
and sensorineural olfactory loss. The authors hypothesized that the observed olfactory
improvement was achieved by an improved OE function caused by an increased nasal airflow
to the olfactory niche.[10]
In the post-COVID-19 era, the unmet need is to find a treatment which could achieve
a meaningful olfactory increase for patients with long-term (>2 years) C19OD. This
noticeable—perceptible for the patient—improvement in the smell function is usually
defined as an increase in the smell scores above the minimal clinically important
difference (MCID).
Currently only few studies have explored treatments for persistent C19OD lasting longer
than 1 year but MCID in olfactory gain has never been achieved.[31] We conducted a pilot study to evaluate olfactory changes in patients with persistent
C19OD undergoing fSRP and compared these to a control group of C19OD patients on OT.
Materials and Methods
Participants were recruited from patients seen in the long-COVID smell clinic at the
Royal National ENT hospital (University College London Hospitals, London, United Kingdom)
between October 2022 and May 2023. Inclusion and exclusion criteria are reported in
[Table 1]. This study was approved by the Hospital Research Ethic Committee (ref. 14/SC/1180)
and was conducted in accordance with the 1996 Declaration of Helsinki. All participants
provided full informed written consent prior to participation.
Table 1
Study inclusion and exclusion criteria
|
Inclusion
|
Exclusion
|
|
Age ≥18
|
Presence of other causes leading/contributing to OD (also confirmed by MRI of the
head/sinuses)a
|
|
Etiology of OD following a polymerase chain reaction—confirmed diagnosis of SARS-CoV-2
infection
|
History of PIOD prior to COVID-19
|
|
OD confirmed at Sniffin' Sticks and longer than 18 months
|
Prior nasal/sinonasal/skull base surgery
|
|
OD failing to improve on conservative treatments, including OT and oral/topical corticosteroids
|
Bleeding disorders
|
|
Aesthetically unacceptable nasal deformity or reduced nasal airflow caused by a confirmed
DNS and/or INV/ENV dysfunction
|
Blood thinners assumption
|
Abbreviations: DNS, deviated nasal septum; ENV, external nasal valve; INV, internal
nasal valve; MRI, magnetic resonance imaging; OD, olfactory dysfunction; OT, olfactory
training; PIOD, post-infectious olfactory dysfunction.
Note: aThese include: congenital olfactory loss, post-traumatic olfactory dysfunction, chronic
rhinosinusitis, neoplasms, previous chemotherapy or radiotherapy to the head and neck,
neurodegenerative diseases.
Subjects satisfying eligibility criteria were offered fSRP. Those refusing it but
willing to take part in the study were asked to continue with OT for the entire study
period and formed the control arm. Subjects in the treatment group were assessed at
baseline (T0), 3 months (T1), and 6 months (T2) from fSRP. Those in the control group, instead, were assessed at T0 and T2 only. During the follow-up period, participants were asked to not start any additional
treatment potentially influencing olfaction. Compliance with OT in the control group
was assessed at T2. fSRP was performed using a standardized external approach involving septoplasty
with nasal bone realignment to increase airway symmetry, and INV and ENV augmentation
using autologous spreader grafts and columellar strut respectively. All operations
were performed by the same team (PJA/ALP) following the same surgical technique.
Sense of smell was evaluated using S'S extended set (Burghart, Medisense) to obtain
the odor threshold (T), discrimination (D), and identification (I) scores.[32] Normosmia was attributed where TDI score was ≥30.75, hyposmia where TDI was >16,
but <30.75, and functional anosmia if TDI was ≤16.[32] The MCID was defined as a clinically significant improvement corresponding to 5.5
points increase in TDI (our primary outcome), 2.5 points for odour threshold, and
3 points for both odour discrimination and identification.[33] Bilateral and unilateral peak nasal inspiratory flow (PNIF) measures were performed
to assess nasal airflow while acoustic rhinometry (AR) was used to obtain unilateral
minimal cross-sectional area (MCA) and nasal volume (NV).[34]
[35]
[36] QoL was assessed using the 36-Item Short Form Health Survey (SF-36). Self-assessment
of olfaction was performed using a visual analogue scale for smell (sVAS—0 represents
“sense of smell absent” and 10 “sense of smell not affected”)[2] whereas sinonasal symptoms were evaluated using the 22-item Sino-Nasal Outcome Test
(SNOT-22).[37] The Nasal Obstruction Symptom Evaluation (NOSE) scale was used to subjectively assess
nasal obstruction.[38] Qualitative olfactory dysfunction (i.e., parosmia/phantosmia) was investigated by
asking the participants if the symptom was present or not at the moment of the examination.
Statistical Analysis
Quantitative variables were summarized using median and interquartile range whereas
qualitative variables were described with frequency and percentage. Comparisons of
measurements between baseline and follow-ups were performed using the Mann-Whitney
test for quantitative variables and the proportion test for dichotomic variables.
Pearson correlation index was used to measure associations between quantitative variables.
p-values were calculated for all tests, and 5% was considered as the critical level
of significance. Sample size was determined using a power analysis of independent
Mann-Whitney test (two-sided) assuming a difference between means at the end of the
study of 5.5 TDI points (MCID)[33] and an equal standard deviation in the two groups of 4 TDI points. Based on that,
a minimum of 10 patients in each group were required to reach a power of 81%, with
an alpha error of 0.05.
Results
This study assessed 104 subjects for eligibility. A total of 25 participants were
selected with 12 forming the treatment group and 13 entering the control group. The
6-month follow-up period was completed by 9 patients in the treatment group and 10
in the control arm (6-month drop-out rate of 25.0 and 23.1%, respectively). No complications
were recorded following fSRP. Demographics and baseline characteristics for the participants,
and comparison between groups, are reported in [Table 2].
Table 2
General characteristics of the treatment and control groups at baseline
|
Treatment group
n = 12
|
Control group
n = 13
|
p-value
|
|
Age, median (P25–P75), yr
|
40.0 (31.5–44.0)
|
49.0 (30.0–54.0)
|
0.66
|
|
Sex, No. (%)
|
|
|
0.77
|
|
Female
|
9 (75.0%)
|
8 (61.5%)
|
|
|
Male
|
3 (25.0%)
|
5 (38.5%)
|
|
|
Length of OD,[a] median (P25–P75), yr
|
2.3 (2.0–2.5)
|
2.4 (1.9–2.8)
|
0.53
|
|
Parosmia, No. (%)
|
10 (83.3%)
|
10 (76.9%)
|
1
|
|
Phantosmia, No. (%)
|
4 (33.3%)
|
2 (15.4%)
|
0.56
|
|
Smoking, No. (%)
|
|
|
0.50
|
|
Ex-smoker
|
1 (100%)
|
0 (0.0%)
|
|
|
Yes
|
0 (0.0%)
|
2 (100%)
|
|
|
No
|
0 (0.0%)
|
0 (0.0%)
|
|
|
Comorbidity, No. (%)
|
|
|
0.33
|
|
None
|
8 (66.7%)
|
9 (69.2%)
|
|
|
Yes
|
4 (33.3%)
|
4 (30.8%)
|
|
|
Hypothyroidism
|
1 (25.0%)
|
1 (25.0%)
|
|
Asthma
|
1 (25.0%)
|
1 (25.0%)
|
|
|
Others
|
4 (100%)
|
3 (75.0%)
|
|
|
Allergic rhinitis, No. (%)
|
2 (16.7%)
|
0 (0.0%)
|
0.94
|
|
Chronic rhinosinusitis, No. (%)
|
0 (0.0%)
|
0 (0.0%)
|
1
|
|
Family history Alzheimer/Parkinson, No. (%)
|
3 (25.0%)
|
1 (7.7%)
|
1
|
|
History of PIOD, No. (%)
|
2 (16.6%)
|
3 (23.1%)
|
1
|
|
History of previous nasal operations, No. (%)
|
0 (0.0%)
|
0 (0.0%)
|
1
|
|
History of head trauma, No. (%)
|
1 (8.3%)
|
0 (0.0%)
|
0.97
|
Abbreviations: OD, olfactory dysfunction; PIOD, post-infectious olfactory dysfunction.
Notes: Statistical difference between groups is also shown. Levels of significance:
*p ≤ 0.05.
a Length of OD is calculated as number of days from the infection date to the day of
enrolment.
Olfactory Scores, Nasal Measurements, and Patient-Reported Outcome Measures (PROMs)
at Baseline
Apart from the median discrimination scores, all S'S subtest scores at baseline were
below normative values when compared to those of an adult population of similar age
group.[32] Similarly, baseline median bilateral and unilateral PNIF as well as AR parameters
were below the reference values for an adult population of similar age group.[35]
[39]
[40] Lower SF-36 scores were found for the health domains role limitations due to physical
health, energy/fatigue, emotional well-being, social functioning, and general health
when compared to normative values for the UK population.[41] Reduced scores were observed for sVAS while raised scores were found for the SNOT-22[42] and NOSE.[43] No statistically significant differences were noted in the olfactory scores, nasal
measurements, and PROMs at baseline between the two groups ([Tables 3] and [4]).
Table 3
Olfactory and nasal measurements, and patient-reported outcome measures (PROMs) at
baseline, 3 months, and 6 months following functional septorhinoplasty for the treatment
group and at baseline and at 6 months for the control group
|
Treatment group
|
Control group
|
|
Baseline (T0)
n = 12
|
3 months (T1)
n = 10
|
6 months (T2)
n = 9
|
Baseline (T0)
n = 13
|
6 months (T2)
n = 10
|
|
Sniffin' Sticks
|
|
|
|
|
|
|
TDI, median (P25–P75)
|
22.3 (20.0–24.8)
|
26.8 (20.0–24.8)
|
30.3 (24.5–30.8)
|
22.0 (18.0–25.0)
|
21.9 (21.1–31.2)
|
|
Threshold, median (P25–P75)
|
1.8 (1.0–3.8)
|
4.6 (1.7–7.0)
|
5.8 (4.0–7.3)
|
4.0 (2.3–4.5)
|
4.4 (2.2–5.4)
|
|
Discrimination, median (P25–P75)
|
10.0 (10.0–11.3)
|
11.5 (10.0–12.0)
|
12.0 (11.0–13.0)
|
9.0 (8.0–10.0)
|
10.5 (8.0–13.0)
|
|
Identification, median (P25–P75)
|
9.0 (8.0–11.3)
|
10.5 (9.0–12.8)
|
12.0 (10.0–13.0)
|
9.0 (7.0–10.0)
|
10.0 (8.3–10.8)
|
|
Normosmics, n (%)
|
0 (0.0%)
|
1 (10.0%)
|
4 (44.4%)
|
0 (0.0%)
|
3 (30.0%)
|
|
Hyposmics, n (%)
|
11 (91.7%)
|
9 (90.0%)
|
5 (55.6%)
|
12 (92.3%)
|
6 (60.0%)
|
|
Anosmics, n (%)
|
1 (8.3%)
|
0 (23.5%)
|
0 (0.0%)
|
1 (7.7%)
|
1 (10.0%)
|
|
Nasal measurements
|
|
|
|
|
|
|
PNIF, median (P25–P75), L/min
|
|
|
|
|
|
|
Bilateral PNIF
|
115.0 (87.5–137.5)
|
137.5 (130.0–157.5)
|
160.0 (125.0–190.0)
|
135.0 (110.0–162.5)
|
|
|
Right PNIF
|
62.5 (50.0–82.5)
|
82.5 (66.3–130.0)
|
110.0 (70.0–120.0)
|
82.5 (66.3–98.8)
|
|
|
Left PNIF
|
60.0 (48.8–77.5)
|
90.0 (81.3–107.5)
|
65.0 (50.0–100.0)
|
100.0 (68.8–116.3)
|
|
|
Acoustic rhinometry, median (P25–P75)
|
|
|
|
|
−
|
|
Right MCA1, cm2
|
0.5 (0.4–0.7)
|
0.5 (0.4–0.7)
|
0.6 (0.4–0.7)
|
0.6 (0.4–0.9)
|
|
|
Right nasal volume (0–5), cm3
|
5.7 (5.2–7.3)
|
8.1 (5.9–10.9)
|
7.9 (5.8–9.4)
|
6.9 (6.2–13.1)
|
|
|
Left MCA1, cm2
|
0.7 (0.4–0.8)
|
0.7 (0.5–0.8)
|
0.6 (0.5–0.8)
|
0.8 (0.6–1.1)
|
|
|
Left nasal volume (0–5), cm3
|
6.6 (5.7–7.4)
|
9.5 (7.6–10.7)
|
8.0 (6.0–9.1)
|
9.0 (6.1–13.1)
|
|
|
PROMs
|
|
|
|
|
−
|
|
SF-36, median (P25–P75), %
|
|
|
|
|
|
|
Physical functioning
|
95.0 (83.8–100)
|
100 (81.3–100)
|
95.0 (90.0–100)
|
100 (87.5–100)
|
100 (91.3–100)
|
|
Role limitations due to physical health
|
62.5 (25.0–100)
|
75.0 (31.3–100)
|
75.0 (50.0–100)
|
100 (75.0–100)
|
100 (100–100)
|
|
Role limitations due to emotional problems
|
100 (33.3–100)
|
100 (66.7–100)
|
100 (100–100)
|
100 (16.7–100)
|
100 (50.0–100)
|
|
Energy/Fatigue
|
50.0 (23.8–73.8)
|
60.0 (37.5–75.0)
|
55.0 (50.0–80.0)
|
45.0 (27.5–62.5)
|
52.5 (38.8–55.5)
|
|
Emotional well-being
|
70.0 (67.0–88.0)
|
72.0 (68.0–88.0)
|
84.0 (76.0–88.0)
|
80.0 (58.0–82.0)
|
72.0 (53.0–90.0)
|
|
Social functioning
|
68.8 (46.9–100)
|
100 (65.6–100)
|
100 (81.3–100)
|
62.5 (50.0–93.8)
|
81.3 (62.5–96.9)
|
|
Pain
|
78.8 (65.0–90.0)
|
90.0 (78.1–97.5)
|
90.0 (73.8–95.0)
|
90.0 (83.8–100)
|
95.0 (71.9–100)
|
|
General health
|
65.0 (45.0–72.5)
|
62.5 (55.0–83.8)
|
75.0 (50.0–85.0)
|
65.0 (40.0–77.5)
|
62.5 (55.0–68.8)
|
|
sVAS, median (P25–P75)
|
4.3 (3.0–5.3)
|
5.0 (4.0–6.4)
|
6.0 (4.5–6.0)
|
4.0 (2.0–5.5)
|
4.0 (3.5–5.0)
|
|
SNOT-22, median (P25–P75)
|
25.0 (14.3–30.0)
|
13.0 (9.3–32.5)
|
11.0 (6.0–15.0)
|
12.0 (10.0–30.5)
|
17.5 (15.3–27.8)
|
|
NOSE, median (P25–P75)
|
25.0 (12.5–45.0)
|
17.5 (7.5–34)
|
10.0 (10.0–15.0)
|
20.0 (10.0–52.5)
|
27.5 (6.5–44)
|
Abbreviations: MCA1, first minimal cross-sectional area; NOSE, Nasal Obstruction and
Septoplasty Effectiveness Scale; PNIF, peak nasal inspiratory flow; PROMs: patient-reported
outcome measures; SF-36, 36-item Short Form Survey; Short-QODNS, short version of
Questionnaire of Olfactory Disorders - Negative Statements; SNOT-22, 22-item SinoNasal
Outcome Test; sVAS, Visual Analogue Scale for sense of smell; TDI, Threshold + Discrimination + Identification.
Table 4
Differences in medians and statistical significance (p-values in brackets)
|
Within group comparisons
|
Between groups
comparisons[a]
|
|
Treatment group
|
Control group
|
|
T0–T1
|
T1–T2
|
T0–T2
|
T0–T2
|
T0
|
T2
|
|ΔT0–T2|
|
|
Sniffin' Sticks
|
|
|
|
|
|
|
|
|
TDI
|
+4.5 (0.15)
|
+3.5 (0.22)
|
+8.0 (0.005)**
|
−0.1 (0.39)
|
+0.3 (0.57)
|
+8.4 (0.19)
|
8.1 (0.06)
|
|
Threshold
|
+2.8 (0.08)
|
+1.2 (0.57)
|
+4.0 (0.01)**
|
+0.4 (0.53)
|
−2.2 (0.11)
|
+1.4 (0.33)
|
3.6 (0.05)*
|
|
Discrimination
|
+1.5 (0.46)
|
+0.5 (0.28)
|
+2.0 (0.05)*
|
+1.5 (0.51)
|
+1.0 (0.07)
|
+1.5 (0.46)
|
0.5 (0.74)
|
|
Identification
|
+1.5 (0.27)
|
+1.5 (0.38)
|
+3.0 (0.04)*
|
+1.0 (0.23)
|
0.0 (0.32)
|
+2.0 (0.08)
|
2.0 (0.07)
|
|
Nasal measurements
|
|
|
|
|
|
|
|
|
PNIF, L/min
|
|
|
|
|
|
|
|
|
Bilateral PNIF
|
+22.5 (0.06)
|
+22.5 (0.84)
|
+45.0 (0.04)*
|
−
|
−20.0 (0.20)
|
−
|
−
|
|
Right PNIF
|
+20.0 (0.11)
|
+27.5 (0.77)
|
+47.5 (0.07)
|
−20.0 (0.13)
|
|
Left PNIF
|
+30.0 (0.03)*
|
−25.0 (0.27)
|
+5.0 (0.40)
|
−40.0 (0.06)
|
|
Acoustic rhinometry
|
|
|
|
|
|
|
|
|
Right MCA1, cm2
|
0.0 (0.62)
|
+0.1 (0.29)
|
+0.1 (0.34)
|
−
|
−0.1 (0.45)
|
−
|
−
|
|
Right nasal volume (0–5), cm3
|
+2.4 (0.07)
|
−0.2 (0.66)
|
+2.2 (0.03)*
|
−1.2 (0.1)
|
|
Left MCA1, cm2
|
0.0 (0.72)
|
−0.1 (0.96)
|
−0.1 (0.84)
|
−0.1 (0.38)
|
|
Left nasal volume (0–5), cm3
|
+2.9 (0.09)
|
−1.5 (0.90)
|
+1.4 (0.11)
|
−2.4 (0.19)
|
|
PROMs
|
|
|
|
|
|
|
|
|
SF-36, %
|
|
|
|
|
|
|
|
|
Physical functioning
|
+5.0 (0.55)
|
−5.0 (0.83)
|
0.0 (0.74)
|
0.0 (0.78)
|
−5.0 (0.63)
|
−5.0 (0.63)
|
0.0 (0.56)
|
|
Role limitations due to physical health
|
+10.0 (0.68)
|
0.0 (0.97)
|
+10.0 (0.61)
|
0.0 (0.74)
|
−37.5 (0.14)
|
−25.0 (0.14)
|
10 (0.32)
|
|
Role limitations due to emotional problems
|
0.0 (0.85)
|
0.0 (0.40)
|
0.0 (0.47)
|
0.0 (0.64)
|
0.0 (0.55)
|
0.0 (0.56)
|
0.0 (0.33)
|
|
Energy/Fatigue
|
+10.0 (0.69)
|
−5.0 (0.62)
|
+5.0 (0.41)
|
+7.5 (0.59)
|
+5.0 (0.77)
|
+2.5 (0.43)
|
2.5 (0.95)
|
|
Emotional well-being
|
+2.0 (0.74)
|
+12.0 (0.44)
|
+14.0 (0.43)
|
−8.0 (1.00)
|
−10.0 (0.67)
|
+12.0 (0.29)
|
22 (0.72)
|
|
Social functioning
|
+31.2 (0.24)
|
0.0 (0.93)
|
+31.2 (0.31)
|
+18.8 (0.73)
|
+6.3 (0.79)
|
+18.7 (0.37)
|
12.4 (0.85)
|
|
Pain
|
+11.2 (0.57)
|
0.0 (0.65)
|
+11.2 (0.74)
|
+5.0 (0.96)
|
−11.2 (0.28)
|
−5.0 (0.42)
|
6.2 (0.27)
|
|
General health
|
−2.5 (0.74)
|
+12.5 (0.97)
|
+10.0 (0.50)
|
−2.5 (1.00)
|
0.0 (1)
|
+12.5 (0.32)
|
12.5 (0.72)
|
|
sVAS
|
+0.7 (0.34)
|
+1.0 (0.84)
|
+1.7 (0.17)
|
0.0 (0.69)
|
+0.3 (0.78)
|
+2.0 (0.25)
|
1.7 (0.82)
|
|
SNOT-22
|
−12.0 (0.27)
|
−2.0 (0.35)
|
−14.0 (0.03)*
|
+5.5 (0.53)
|
+13.0 (0.53)
|
−6.5 (0.13)
|
19.5 (0.41)
|
|
NOSE
|
−7.5 (0.48)
|
−7.5 (0.20)
|
−15.0 (0.05)*
|
+7.5 (0.84)
|
+5.0 (0.93)
|
−17.5 (0.20)
|
22.5 (0.31)
|
Abbreviations: MCA1, first minimal cross-sectional area; NOSE, Nasal Obstruction and
Septoplasty Effectiveness Scale; PNIF, peak nasal inspiratory flow; PROMs, patient-reported
outcome measures; SF-36, 36-item Short Form Survey; Short-QODNS, short version of
Questionnaire of Olfactory Disorders - Negative Statements; SNOT-22, 22-item SinoNasal
Outcome Test; sVAS, Visual Analogue Scale for sense of smell; TDI, Threshold + Discrimination + Identification.
Notes: The sign “ + ” indicates an improvement while the sign “ − ” indicates a worsening
in the median values. Please note that for the intergroup differences the direction
signs have not been used.
Significant p-values in bold. Levels of significance: *p ≤ 0.05, **p ≤ 0.01.
a Comparison made with reference to treatment group (i.e., Treatment group – Control
group).
Within and Between Groups Comparisons at Follow-ups
An improvement in all S'S scores was observed only in the fSRP group both at T1 and T2 but these were statistically significant and all above MCID level (apart from discrimination)
only at T2 ([Fig. 1]
;
[Tables 3] and [4]). A statistically significant improvement at T2 from baseline (T0–T2) was noted only in the treatment group in the bilateral PNIF (p = 0.04) and right NV (p = 0.03), while left PNIF improved significantly only at T1 from baseline (T0–T1, p = 0.03) ([Tables 3] and [4]). A statistically significant reduction in the SNOT-22 and NOSE was demonstrated
at T2 (respectively p = 0.03 and p = 0.05) only in the treatment group ([Tables 3] and [4]). When comparing the gain obtained between T0 and T2 between the two groups, a statistically significant difference was noted for the
threshold (p = 0.05) and a trend toward significance was noted for the TDI (p = 0.06) and the identification (p = 0.07), all in favour of fSRP ([Fig. 2]; [Table 4]).
Fig. 1 Box plots showing changes in TDI (A), threshold (B), discrimination (C), and identification (D) scores for the functional septorhinoplasty (fSRP) group during the study period.
Statistical difference between intervals is also shown. Levels of significance: *p ≤ 0.05, **p ≤ 0.01.TDI, Threshold + Discrimination + Identification.
Fig. 2 Box plots showing changes in TDI (A), threshold (B), discrimination (C), and identification (D) scores between the treatment (cases) and the control groups between T0 (baseline) and T2 (6 months). Statistical difference between intervals is also shown. Levels of significance:
*p ≤ 0.05. TDI, Threshold + Discrimination + Identification.
Correlation Between Olfactory Function and Nasal Measurements
No correlations were found between S'S scores and nasal measurements when considering
all the measurements obtained in the whole population. When we looked at the correlations
between the changes in S'S scores and nasal measurements between T0 and T2 in the fSRP group, we found strong significant correlations between changes in left
PNIF and changes in TDI (r = 0.67; p = 0.05), between changes in total PNIF and changes in discrimination (r = 0.73; p = 0.03) and identification (r = − 0.67; p = 0.05), and between changes in left MCA1 and changes in identification (r = − 0.74; p = 0.03) ([Fig. 3]).
Fig. 3 Correlation matrix showing strength of correlations between changes (Δ) in Sniffin'
Sticks scores and changes (Δ) in nasal airways parameters in the treatment group.
Significant p-values in bold. Levels of significance: *p ≤ 0.05.MCA1, first minimal cross-sectional area; NV, nasal volume; PNIF, peak nasal
inspiratory flow; TDI, Threshold + Discrimination + Identification.
Discussion
Our pilot study shows that fSRP can significantly improve persistent C19OD in patients
who have previously failed other treatment options for post-infectious OD (PIOD).
Patients undergoing fSRP demonstrated a statistically significant improvement in their
olfactory scores at 6 months, above MCID level for all S'S scores (apart from discrimination),[33] while an olfactory improvement was not observed in the control arm (patients on
OT). Importantly, at 6 months, we observed a statistically significant improvement
in threshold gain following fSRP (+3.6 points, p = 0.05). A clear positive trend in TDI gain was also observed but did not reach statistical
significance (+8.1 points, p = 0.06) ([Table 4]). These olfactory gains represent the olfactory improvement secondary to the intervention
minus the control arm, that is, the olfactory benefit obtainable with fSRP when compared
to OT. Several studies confirm that olfactory threshold reflects the peripheral olfactory
apparatus function (i.e., OE).[10]
[44]
[45] This suggests that olfactory benefits following fSRP are primarily related to an
increased peripheral olfactory stimulation (i.e., olfactory sensorineural reactivation
or sensorineural reversibility) implying that ORNs are still present in patients with
persistent C19OD. Our data corroborates previous findings by Whitcroft et al[10] who showed a statistically significant improvement in the mean TDI (+6.5 points,
p = 0.03) in patients with long-term OD undergoing fSRP. Similarly, we demonstrated
TDI improvement in all our fSRP patients, with six of them (66.7%) reaching MCID,[33] when compared to only four (40%) in the control group. Importantly, in the control
arm olfaction further decreased in four patients (40%) while deteroriation was not
demonstrated in any of the fSRP patients. Although fSRP patients demonstrated a noticeable
improvement in their TDI, statistical significance at T2 was not reported with patient-reported olfaction (sVAS) and general QoL scores. However,
in previous studies we found statistically significant correlations between olfactory
scores, sVAS and SF-36,[3]
[4] and we believe these non-significant improvements obtained in the present study
may be related to the small sample size of our cohorts.
The main driving mechanism in the olfactory improvement obtained in the fSRP group
is centered around an increase in nasal airflow as confirmed by a strong significant
correlation between postoperative changes in S'S scores and PNIF/AR ([Fig. 3]). Following fSRP, patients experienced an objective and subjective increase in the
nasal airflow as demonstrated by a significant improvement of bilateral PNIF (p = 0.04) and a decrease in NOSE and SNOT-22 (p = 0.05 and p = 0.03, respectively) at 6 months ([Table 4]). In addition to this prevalent mechanism, a previous functional MRI study showed
that fSRP can lead to structural and functional plasticity of secondary olfactory
cortices, caused by a bottom-up plasticity process.[10] In support of this, we demonstrated a statistically significant improvement of the
identification and discrimination scores, which have been shown to reflect more complex
processing of olfactory information and influenced by cognitive processes.[10]
[44]
[45]
The concept of olfactory improvement following nasal surgery is not new.[9]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26] However, its efficacy in PIOD, and in particular in C19OD, remains unexplored. Consequentely,
in the post-COVID-19 era, in which thousands of people have been left with a debilitating
OD unable to improve on other available options, the potential role of fSRP in improving
OD is gaining increasing attention from many rhinology surgeons. This notion is supported
by recent systematic reviews and meta-analyses showing that fSRP not only constitutes
a safe procedure in terms of long-term olfactory function but can also restore smell.[9]
[30] The majority of studies seem to suggest that an improvement in the nasal airflow
in the olfactory area can lead to improved olfaction by enhancing transport of odor
molecules to the olfactory cleft.[18]
[19]
[23]
[24]
[46]
[47]
[48]
[49]
[50] In particular, a growing body of evidence seems to support the critical role of
the INV in influencing airflow in the olfactory cleft region.[28]
[51]
[52]
[53] Spreader grafts are known to increase the INV angle section[54] and, in fact, a positive association between presence of spreader grafts and olfactory
outcomes has been reported.[21] Anatomical variation of the ENV can also influence direction of the airflow and
play a role in the transportation of odorants to the olfactory cleft.[55]
[56] Our patients underwent bilateral INV augmentation, by using bilateral spreader grafts,
and anatomical variation of the ENV, by means of columellar strut. By increasing the
nasal airflow to the olfactory clefts, growing evidence suggests that this increased
olfactory stimulation, caused by a greater quantity of odorants reaching the olfactory
area, can lead to an improved OE activity. This may contribute to the restoration
of the sensorineural deficit (i.e., OE damage) present in C19OD.[10] This increased peripheral input can then lead to a structural and functional plasticity
of secondary olfactory cortices through a bottom-up plasticity process.[10]
Although nasal airflow improved following fSRP, all patients were enrolled from our
long-COVID smell clinic with a primary diagnosis of persistent C19OD and not nasal
blockage. However, all patients had a mild nasal blockage with average NOSE scores
less than 25 with mildly reduced PNIF/AR scores. Nevertheless, it is important to
delineate the separate preoperative diagnoses of C19OD from long-standing mild nasal
blockage which appear unrelated in causation prior to COVID. All our patients had
reported normal sense of smell prior to COVID-19 (history of OD was an exclusion criteria
to the study).
Despite continuous research efforts, treatments for long-term (>1 year) C19OD today
remain limited and equally have failed to demonstrate a clinically important olfactory
improvement (i.e., above MCID). OT is considered the gold standard treatment for C19OD.[5]
[57]
[58] However, its benefits can abate when OD becomes long-standing and our study, unfortunately,
seems to suggest so. This paints a bleak picture for those untreated patients with
persistent C19OD.
Strengths and Limitations
To the best of our knowledge, this is the first study to have investigated the role
of fSRP, and more widely of nasal airways surgery, in improving sense of smell in
patients with C19OD who have failed previous conservative options for PIOD. Moreover,
it is the only study that looked into new potential treatments to improve olfaction
in patients with a C19OD longer than 2 years, while demonstrating significant olfactory
improvement above MCID. The main limitation of our study is the small sample size
and, although our study was powered enough at baseline and at 3 months, it losts power
at the 6-month follow-up due to patients' dropout.
Conclusion
Our pilot study suggests that fSRP can significantly improve sense of smell in patients
with persistent C19OD lasting more than 2 years with additional significant olfactory
threshold gain when compared to OT. By augmenting the INV angle and optimizing nasal
airflow to the olfactory cleft, fSRP can improve olfaction by increasing transport
of odorants to the OE. This increased stimulation of the olfactory mucosa leads to
a sensorineural improvement of the OE potentially triggering a bottom-up plasticity
process in the central olfactory areas. Nevertheless, further studies on larger populations
are needed to confirm our preliminary findings.
