1 Introduction
Olfactory dysfunction (OD) is a prevalent condition affecting 10%–20% of the general population, with prevalence increasing significantly with age
[1,
2]. Among patients with chronic rhinosinusitis (CRS), 60%–80% experience OD, with higher rates in the subtype with nasal polyps (CRSwNP)
[3]. The COVID-19 pandemic further highlighted this problem, with infection-related OD affecting 40%–80% of patients and stimulating intensive research into olfactory pathogenesis
[4].
The olfactory cleft (OC) is a narrow, paired paramedian channel through which odorant molecules reach the olfactory epithelium
[5]. Its structural integrity and microenvironmental homeostasis are essential for effective olfactory signal transduction. OC abnormalities, including anatomical narrowing, inflammatory infiltration, and microbiota dysbiosis, represent major causes of peripheral OD
[3,
6,
7]. However, most prior research has focused on central neural pathways or neurodegenerative associations
[8], leaving the mechanistic interplay within the OC underexplored. Existing frameworks, such as the microbiota-metabolite-immune (MMI) axis, emphasize microbial initiation of disease but do not account for anatomical obstruction as the triggering event
[9,
10]. This gap is clinically significant: without addressing the structural substrate, anti-inflammatory or antimicrobial therapies may yield only partial, transient improvement. Furthermore, current management strategies remain largely unidimensional, with no integrated approach for simultaneously addressing anatomical repair, immune regulation, and microbial remodeling. The deep location and delicate structure of the OC also render traditional examination methods incapable of fully capturing its local inflammatory and molecular pathological changes.
This review proposes an anatomy-based framework that positions OC obstruction as the initiating event in a self-amplifying cascade. We define OC-related OD as olfactory impairment primarily attributable to pathological processes confined to the OC, encompassing anatomical, inflammatory, and microbial domains, irrespective of the underlying condition (CRS, nasal polyps, respiratory epithelial adenomatoid hamartoma [REAH], or isolated OC stenosis)
[1,
11]. Our objectives are to synthesize current evidence on etiology, evaluate diagnostic and therapeutic advances with Grading of Recommendations Assessment, Development and Evaluation (GRADE)-graded evidence quality, and highlight emerging surgical and regenerative options that address the anatomical root of disease.
2 Methods
A comprehensive literature search was conducted across PubMed/MEDLINE, Web of Science, Embase, the Cochrane Library, CNKI, and Wanfang through July 2026, using Medical Subject Headings (MeSH) and free-text terms including "olfactory cleft," "olfactory dysfunction," "chronic rhinosinusitis," "type 2 inflammation," "nasal microbiota," "biologics," and "olfactory training." Original articles, reviews, meta-analyses, and clinical guidelines in English or Chinese were eligible. Evidence quality was graded using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system: 1A (strong, high quality: meta-analyses and large randomized controlled trials [RCTs]), 2B (moderate: individual RCTs and prospective cohorts), 2C (low: retrospective and case series), and 2D (very low: case reports and expert opinion). Evidence grades are indicated at the end of relevant subsections.
3 Etiology
Normal olfactory function depends on unobstructed ventilation and drainage of the OC, along with the structural and functional integrity of the olfactory epithelium. OD secondary to OC disorders involves three interrelated domains: anatomical obstruction (the initiator), type 2 inflammation (the amplifier), and microbiota dysbiosis (the modulator), as illustrated in Fig. 1.
3.1 Anatomical obstruction
The OC is bounded anteriorly by the middle turbinate attachment, posteriorly by the sphenoid sinus wall, medially by the nasal septum, and laterally by the middle and superior turbinates
[12,
13]. Anatomical abnormalities, including congenital variants and space-occupying lesions, disrupt olfaction by impairing airflow and causing structural compression.
OC narrowing commonly results from aberrant turbinate curvature combined with septal compression
[14]. Topical corticosteroids typically fail to achieve significant recovery in these patients, and surgical intervention yields only modest benefits
[15]. Notably, Imbs et al. found that among 137 pre-septorhinoplasty patients without sinonasal symptoms, 62% exhibited OC stenosis or obstruction on computed tomography (CT), suggesting that anatomical narrowing alone does not invariably cause OD
[16]. Conversely, Besser et al. demonstrated that reversible OC obstruction in 30 healthy volunteers, achieved by stepwise foam application, significantly impaired orthonasal olfaction while retronasal function was partially preserved, establishing a direct causal relationship between structural blockage and olfactory loss
[17]. This paradox, that obstruction is common but OD is not universal, highlights the need for functional rather than purely structural assessment and suggests that individual variability in olfactory reserve and mucosal resilience may modulate the functional impact of anatomical narrowing.
Among space-occupying lesions, nasal polyps and REAH are most relevant. OC polyps cause both mechanical blockage and cytokine-mediated olfactory nerve injury
[18]. Importantly, polyps may contain residual olfactory epithelium, yet endoscopic sinus surgery (ESS) does not compromise postoperative olfactory scores
[19]. REAH, characterized by polypoid proliferation in the OC, requires endoscopic resection; coexisting REAH predicts poorer postoperative olfactory outcomes and higher recurrence rates
[20,
21].
Jankowski et al. established that OC surgery constitutes a distinct surgical subspecialty, complementary to ethmoidal and skull base surgery, given the OC’s role as a site for tumors, cerebrospinal fluid (CSF) rhinorrhea, and OD
[22]. This anatomical specificity underscores the principle that structural OC abnormalities are not merely concurrent findings but the initiating substrate for downstream inflammatory and microbial processes. Anatomical obstruction is the structural seed of the self-amplifying cascade: narrowing impairs mucociliary clearance, fosters pathobiont retention, and facilitates mucosal edema that further narrows the cleft (Fig. 1) (Evidence: 2B).
3.2 Inflammatory damage
Type 2 inflammation is the core amplifier within the tripartite network. In CRS, olfactory impairment affects approximately 80% of patients and is closely linked to eosinophilic infiltration of the olfactory epithelium
[3,
23]. The pathological cascade includes loss of olfactory sensory neurons (OSNs), reduced olfactory bulb volume (OBV), epithelial metaplasia from olfactory to respiratory type, and mucosal edema
[24,
25]. Eosinophilic CRS causes more pronounced olfactory impairment than the non-eosinophilic subtype, driven by interleukin-4 (IL-4), interleukin-5 (IL-5), and interleukin-13 (IL-13), which mediate eosinophil recruitment, activation, and subsequent epithelial damage
[23]. The severity of olfactory loss correlates with tissue eosinophil counts and the intensity of local type 2 cytokine release.
These cytokines mediate OSN injury and barrier breakdown. Eosinophil-derived neurotoxin (EDN) released in the OC reflects the degree of local eosinophilic activity more accurately than peripheral blood counts
[7]. Galectin-10, a marker of intense eosinophil activation identified in OC mucus, further correlates with olfactory loss in CRS, expanding the biomarker panel alongside EDN and IL-5
[25].
Critically, inflammatory edema narrows the OC lumen, creating a feedback loop: obstruction promotes inflammation, and inflammation worsens obstruction. This self-amplifying mechanism explains why anti-inflammatory therapy alone often yields incomplete recovery. If the anatomical bottleneck persists, inflammatory mediators continue to accumulate in the stagnant OC microenvironment, perpetuating epithelial damage even as systemic inflammation resolves. (Evidence: 2B).
3.3 Microbiota dysbiosis
The nasal microbiota maintains olfactory epithelial homeostasis and regulates local immunity. Germ-free mice exhibit impaired olfactory epithelial maturation, confirming the microbiota’s role in olfactory development
[26]. Clinical studies demonstrate significant associations between olfactory function and nasal microbial composition, including taxonomic diversity and the relative abundance of key taxa
[6,
27].
The MMI axis provides a mechanistic bridge: dysbiotic communities enriched in
Staphylococcus aureus form biofilms, impair epithelial barrier integrity, and skew mucosal T-helper responses toward T-helper 2 (Th2) polarization
[10,
28,
29]. Co-infection with
Malassezia sympodialis and bacterial pathobionts elicits distinct sinonasal inflammatory responses in murine sinusitis models, with
Pseudomonas aeruginosa co-infection producing neutrophilic patterns and
S. aureus co-infection producing Th2-skewed responses
[28]. Microbial metabolites, including short-chain fatty acids and other bioactive molecules, further modulate cytokine release within the OC microenvironment
[10,
27]. Health-associated commensals such as
Corynebacterium and
Dolosigranulum appear to be depleted in CRS, suggesting a protective role that is lost during dysbiosis.
However, evidence limitations must be acknowledged. Most microbiota studies sample the middle meatus rather than the OC itself, and no study has directly profiled the OC-specific microbiota in relation to olfactory outcomes. The MMI framework also does not account for anatomical obstruction as the initiating event. The present anatomy-based model addresses this gap by positioning structural narrowing as the trigger that creates the stagnant microenvironment favorable for dysbiosis, thereby integrating the MMI axis into a broader causal hierarchy. (Evidence: 2C).
4 Diagnosis
The diagnosis of OC-related OD requires a multidimensional assessment integrating subjective, psychophysical, endoscopic, imaging, and laboratory approaches. Each modality offers unique advantages and limitations (Table 1), and a comprehensive combination enables etiological stratification and treatment planning (Fig. 2).
4.1 Subjective psychophysical
Subjective assessment tools, including the Visual Analogue Scale (VAS) and Questionnaire of Olfactory Disorders (QOD), are quick and inexpensive but show disease-specific validity. VAS scores correlate strongly with objective measures in CRSwNP but only weakly in post-COVID-19 OD and in patients with normal olfactory function, limiting their utility as standalone tools
[4,
30,
31]. This disease-specific variability means that subjective scores alone are insufficient for reliable diagnosis or monitoring, particularly in mild OD where patients may underestimate or overestimate their impairment.
Psychophysical testing remains the gold standard. The Sniffin’ Sticks (SS) test and University of Pennsylvania Smell Identification Test (UPSIT) evaluate odor threshold, discrimination, and identification, enabling classification into conductive, sensorineural, and mixed OD. A critical limitation is cultural dependence: without adaptation, the SS test misclassifies 44% of healthy individuals as hyposmic, and the UPSIT misclassifies up to 92%
[32]. The culturally adapted Q-Sticks for Chinese populations significantly improves diagnostic accuracy
[33]. Additionally, Juratli et al. demonstrated that OC obstructions can produce dissociated orthonasal and retronasal olfactory deficits, highlighting the incompleteness of single-modality testing
[34]. (Evidence: 2B).
4.2 Endoscopic and imaging
The Olfactory Cleft Endoscopy Scale (OCES) quantifies five OC features (secretion, edema, polyps, crusting, scarring), each scored 0–2
[35]. OCES scores correlate with olfactory function in CRS, particularly postoperatively in CRSwNP
[36]. However, normal OCES does not guarantee intact olfaction, as functional epithelial damage may persist despite morphological normality
[36].
CT is preferred for evaluating OC structural abnormalities. Semi-quantitative OC opacification scoring systems correlate with olfactory function and predict postoperative recovery
[37,
38]. Importantly, OC opacification correlates with olfactory function differentially by CRS subtype: strongly in CRSwNP (conductive mechanism) but not in CRS without nasal polyps (inflammatory mechanism)
[38], and higher scores predict eosinophilic CRSwNP
[39]. Magnetic resonance imaging (MRI) complements CT by quantifying OBV, which reflects central olfactory pathway integrity and predicts recovery potential, reduced OBV correlates with longer disease duration
[40,
41]. Altundag et al. found that OC width measurements on CT differ between COVID-19 anosmia and postviral anosmia, though discriminative value remains limited
[42].
Computational fluid dynamics (CFD) represents an emerging functional tool. Asama et al. demonstrated that CFD can quantify OC airflow even in patients with anatomically normal clefts, potentially identifying functional obstruction invisible to static imaging
[43]. CFD modeling can simulate odorant molecule transport under varying anatomical configurations, offering a mechanistic bridge between structural findings and functional impairment. However, CFD remains investigational, requiring specialized expertise and computational resources not yet available in routine clinical practice. (Evidence: 2B).
4.3 OC biomarkers
OC mucus contains functional proteins reflecting local pathophysiology. Odorant-binding proteins (OBPs) are specifically secreted by the olfactory epithelium, and several inflammation-associated proteins, including lipocalin-1 (LCN1), C-C motif chemokine ligand 2 (CCL2), and granulocyte-macrophage colony-stimulating factor (GM-CSF), correlate with olfactory test scores
[44]. EDN levels in OC mucus are more sensitive than peripheral blood eosinophil counts for assessing local eosinophilic inflammation
[7]. IL-5 exhibits a significant negative correlation with olfactory function across all CRS subtypes, while interleukin-6 (IL-6) and vascular endothelial growth factor A (VEGF-A) show subtype-specific correlations
[45,
46]. Galectin-10 serves as a marker of intense eosinophil activation and predicts olfactory loss
[25].
Despite their promise, OC mucus biomarkers face practical barriers: no standardized sampling device exists, collection requires endoscopic guidance, and processing requires laboratory infrastructure not available in routine clinics. A complete diagnostic assessment involving VAS, psychophysical testing, CT, MRI, and biomarkers may exceed 90 minutes, making comprehensive evaluation impractical in standard outpatient settings. This diagnostic burden underscores the need for a tiered, efficiency-optimized algorithm (Fig. 2) (Evidence: 2B).
5 Treatment
Management of OC-related OD requires a stepwise approach targeting the anatomical, inflammatory, and microbial nodes of the pathogenic cascade. Table 2 summarizes current treatment options with evidence grades.
5.1 Pharmacological and local
Corticosteroids, administered topically or systemically, remain first-line for inflammatory OD. Topical sprays often fail to reach the superior OC due to the cleft’s deep location and narrow geometry, whereas oral corticosteroids demonstrate significant efficacy by achieving systemic distribution
[47,
48]. Nebulization achieves olfactory improvement in 90% of patients, though effects plateau by day 10, suggesting potential glucocorticoid resistance or receptor downregulation with prolonged use
[47]. This delivery challenge is a direct consequence of the anatomical bottleneck: when the OC is stenosed, even nebulized particles may not reach the olfactory epithelium in therapeutic concentrations.
Gelatin sponge packing overcomes the delivery limitation by placing corticosteroid-impregnated sponges directly in the OC, achieving sustained local release. Bardaranfar et al. demonstrated higher complete remission rates at 8 weeks with triamcinolone-impregnated sponges
[49], and Liang et al. showed that mometasone micro-packing achieves in weeks what spray therapy requires months to accomplish
[50].
Balloon dilatation of the OC, validated in cadaveric specimens, improves local drug delivery: dye staining scores increased from 1.34 to 2.66 post-dilatation
[51–
53]. However, clinical efficacy in OC-related OD awaits prospective trials. Steroid-eluting stents provide both structural support and sustained anti-inflammatory effects, with prospective studies confirming improved olfactory VAS scores, reduced IL-5/IL-13, and decreased systemic corticosteroid requirements
[54–
56] (Evidence: 2B).
5.2 Biologics
Four biologics, dupilumab, mepolizumab, tezepelumab, and stapokibart, target type 2 inflammation in CRSwNP
[57,
58]. Dupilumab (anti-interleukin-4 receptor alpha [IL-4Rα]) yields the largest olfactory improvement: a 2025 meta-analysis of biological therapy outcomes confirmed its superiority among type 2-targeted agents, with an olfactory effect size significantly exceeding that of mepolizumab and benralizumab
[59]. Stilo et al. found that olfactory recovery correlates poorly with polyp volume reduction, suggesting anti-inflammatory rather than mechanical decompression mechanisms
[60]. Saccardo et al. reported that approximately 17% of dupilumab-treated patients achieved normal olfactory function at 6 months, compared with only 7% for mepolizumab, which targets IL-5 and shows relatively limited olfactory benefit though it remains useful for patients with severe eosinophilic asthma
[61].
Tezepelumab (anti-thymic stromal lymphopoietin [TSLP]) demonstrates significant improvements in olfactory loss scores, accompanied by reduced polyp scores and surgical needs
[62,
63]. Stapokibart (anti-IL-4Rα, distinct epitope) shows rapid olfactory improvement within one week in a case report of a patient who failed both biologic and surgical treatment
[64,
65]. These agents represent a paradigm shift from symptomatic management to targeted cytokine blockade, though their specific efficacy in OC-isolated disease (as opposed to diffuse CRSwNP) remains to be established. (Evidence: 1A).
5.3 Surgical interventions
Surgical management of OC-related OD encompasses both decompression and, paradoxically, occlusion approaches. ESS remains the standard for OC obstruction caused by polyps or REAH, with postoperative olfactory improvement correlating with preoperative corticosteroid response
[15,
19].
For congenital OC stenosis, Jankowski et al. developed OC dilatation, a transmucosal lateral fracture-dislocation of the turbinate wall to enlarge the cleft and access the olfactory mucosa with minimal trauma
[14]. The technique involves fracturing the lateral wall of each OC, composed of the middle, superior, and supreme turbinates, to gain access to the recess hosting the olfactory mucosa and the cribriform plate roof. This approach directly addresses the anatomical substrate, embodying the anatomy-based principle that structural correction is the prerequisite for functional recovery. It is specifically indicated for dysosmia secondary to constitutional OC stenosis due to abnormal ethmoid development.
In striking contrast, Liu et al. introduced the OC blocking technique for unilateral, persistent peripheral parosmia
[66]. By surgically occluding the anterior and inferior OC openings, this procedure prevents odorants from reaching the epithelium, abolishing parosmia in the reported case. Postoperative endoscopic examination and CT scan confirmed complete obstruction of the targeted OC openings, and the patient’s parosmia resolved without compromising contralateral olfactory function. This approach, blocking rather than opening, reflects a mechanism-based dichotomy: hyposmia from structural obstruction requires dilatation, while parosmia from aberrant peripheral signaling requires blocking. The same anatomical structure thus demands opposite interventions depending on symptom mechanism, a concept not previously synthesized in reviews and one that directly challenges the assumption that unblocking is always the correct surgical strategy (Evidence: 2D).
5.4 Regenerative and adjunctive
Platelet-rich plasma (PRP) injection into the OC represents a shift from anti-inflammatory to regenerative therapy. PRP contains growth factors, including platelet-derived growth factor (PDGF), transforming growth factor-β (TGF-β), vascular endothelial growth factor (VEGF), and epidermal growth factor (EGF), that promote olfactory epithelial regeneration and neural repair. Yan et al. conducted a multi-institutional randomized controlled trial in COVID-19-related OD, demonstrating significant threshold, discrimination, and identification (TDI) score improvement with three intranasal PRP injections
[67]. Lechien et al. reported 66.7% subjective improvement in 33 posttraumatic OD patients
[68]. A systematic review encompassing 10 clinical studies across multiple etiologies confirmed promising outcomes but highlighted small sample sizes and heterogeneous protocols
[69] (Evidence: 2B).
Olfactory training (OT), involving repeated sniffing of standardized odorants, has meta-analytic support for post-infectious OD
[70–
72]and postoperative CRS settings
[73,
74]. For OC-related OD, OT should serve as a universal adjunct alongside etiological treatment, as monotherapy cannot overcome ongoing structural or inflammatory obstruction. Additionally, emerging microbiota-targeted therapies, including probiotic nasal rinsing
[75,
76] and sinonasal microbiota transfer
[77], aim to interrupt the dysbiosis-inflammation loop at its microbial origin. Dupilumab treatment has been shown to shift the nasal microbiota toward a health-associated composition
[78], suggesting that immune-targeted therapy and microbial remodeling are coupled rather than independent (Evidence: 1A for OT, 2C for microbiota therapies).
6 Discussion
This review proposes an anatomy-based framework that positions OC obstruction as the initiating event in a self-amplifying cascade involving type 2 inflammation and microbiota dysbiosis. Unlike the MMI axis, which starts from the microbial end
[10], our model identifies the anatomical substrate as the trigger that creates the stagnant microenvironment favorable for dysbiosis and inflammation. This distinction has direct clinical implications: addressing only the inflammatory or microbial components, while leaving the anatomical bottleneck uncorrected, may explain the refractory nature of OC-related OD, a clinical frustration frequently encountered by practitioners. The framework also predicts that partial intervention at any single node will allow the remaining two domains to sustain and eventually reverse initial gains, which is why combination therapy targeting multiple nodes simultaneously may be necessary to break the self-amplifying cycle.
Tsetsos et al. recently described CRS-related OD as multifactorial, involving conductive, inflammatory, and sensorineural mechanisms
[9], but did not establish a causal hierarchy. Our framework provides this hierarchy: anatomical obstruction is the initiator, inflammation is the amplifier, and dysbiosis is the modulator. The OC dilatation/blocking dichotomy further illustrates how anatomical reasoning can guide surgical decision-making in ways that inflammatory endotyping alone cannot, the same structure requiring opposite interventions depending on whether the symptom is hyposmia or parosmia.
Several limitations should be acknowledged. Most RCTs focus on CRSwNP rather than isolated OC stenosis, and few studies couple OC imaging with mucus biomarkers and microbiota profiles longitudinally. CFD and artificial intelligence (AI)-assisted airflow modeling remain investigational. PRP evidence relies on small samples, and the OC blocking technique has only case-report-level support. The 62% prevalence of OC stenosis in asymptomatic individuals
[16] further complicates the anatomy-based model, suggesting that anatomical abnormality is necessary but not sufficient for OD, and that functional assessment tools, such as CFD or OC mucus biomarkers, are urgently needed to distinguish clinically significant obstruction from incidental anatomical variants. Additionally, most existing RCTs enroll CRSwNP patients as a whole rather than isolating the OC-stenosis subgroup, making it difficult to attribute treatment effects specifically to anatomical correction versus diffuse anti-inflammatory action.
Future directions should prioritize: (1) prospective multicenter cohorts dynamically coupling OC imaging, mucus biomarkers, and microbiota profiles to validate the anatomy-based model; (2) randomized controlled trials of OC-focused interventions (dilatation, PRP, microbiota transfer) with olfactory function as the primary endpoint, and (3) AI-assisted multimodal prediction models integrating endoscopic, radiologic, and molecular data for individualized treatment selection.
The Author(s). This article is published by Higher Education Press at journal.hep.com.cn.
This is an open access article distributed under the Creative Commons Attribution License 4.0 (CC BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.