A Novel Olfactory Cleft CT Staging System Reveals a Tri-Zone Anatomical Model and Biphasic Progression in Central Compartment Atopic Disease.

BACKGROUND: Central compartment atopic disease (CCAD) is an early stage of type 2 chronic rhinosinusitis (CRS), characterized by type 2 inflammation and frequent olfactory dysfunction, primarily due to olfactory cleft (OC) obstruction. OBJECTIVE: To develop a novel OC CT scoring system for evaluating multi-sectional obstruction patterns in CCAD, identify key regions of OC obstruction, and examine their associations with olfactory function and eosinophilic inflammation. METHODS: A prospective cross-sectional study enrolled 257 adults with primary bilateral CRS. Participants were divided into a development cohort (n=177) and a validation cohort (n=80), with 75 allergic rhinitis patients as controls. A longitudinal retrospective cohort of 35 CCAD patients with paired CT scans (interscan interval: 6 months-3 years) was included for temporal validation of the OC opacification progression pattern. Patients with CRS were classified as CCAD or non-CCAD according to the European Position Paper on Rhinosinusitis and Nasal Polyps 2020 (EPOS2020). The Multi-section Olfactory Cleft Opacity Staging System (MOCOS) was developed to assess apical (AO) and middle-lower (MLO) opacification by evaluating six coronal sections. These sections were generated by subdividing both the anterior (A) and posterior (P) OC regions into front (F), middle (M), and posterior (P) segments, creating three anterior (AF, AM, and AP) and three posterior (PF, PM, and PP) segments. Olfactory function was assessed using Sniffin' Sticks. Peripheral blood was analyzed for inflammatory endotypes. Multivariate logistic regression was used to identify risk factors for CCAD, and a random forest algorithm was used to assess the relative importance of variables associated with olfactory dysfunction. RESULTS: Patients with CCAD showed significantly higher scores for PP-MLO (p=0.006), PP-AO (p=0.025), PF-MLO (p<0.001), PM-MLO (p=0.002), and AP-MLO (p=0.044) compared to non-CCAD patients. Moreover, univariate regression analysis identified all these parameters as potential risk factors for CCAD. These specific OC subsites, delineating the middle-lower posterior OC, were designated as the "disease signature zone." Multivariate analysis identified PP-MLO (OR=5.17) and blood eosinophil counts (BEC) (OR=7.90) as independent predictors of CCAD. A model integrating these factors achieved high accuracy (AUC=0.907, Sensitivity=93.5%, Specificity=86.8%), which was validated in a separate cohort with 75% accuracy. The seven OC subsites (AF-AO, AF-MLO, AM-AO, AM-MLO, AP-AO, PF-AO, and PM-AO) were collectively designated as the "olfactory functional zone," given that their opacification was strongly correlated with threshold-discrimination-identification scores (R=-0.686 to -0.863, all p<0.01) and their feature importance was affirmed by the random forest model. Opacification progression from the disease signature zone to the olfactory functional zone drives both the onset and progressive worsening of olfactory dysfunction. A biphasic pattern, characterized by superior extension in the posterior OC from the superior meatus and sphenoethmoidal recess and inferior extension in the anterior OC from its apical portion, was longitudinally validated in a retrospective cohort of CCAD patients with paired CT scans. BEC positively correlated with AF-AO (R=0.527, p=0.036), AF-MLO (R=0.500, p=0.049), and AM-MLO (R=0.524, p=0.037). Consequently, these three subsites were collectively termed the "eosinophilic inflammation zone." CONCLUSIONS: Based on the MOCOS, a tri-zone OC model is established in CCAD, revealing a biphasic opacification progression that drives olfactory dysfunction onset and progression, thereby refining disease classification and elucidating its underlying pathogenesis to inform targeted therapy.