Computed Tomography Findings in Pathology-Confirmed Rhino-Orbital Mucormycosis During the COVID-19 Era: A...

Rhino-orbital mucormycosis (ROM) emerged as a severe complication during the COVID-19 era, particularly in immunocompromised patients. Computed tomography (CT) plays a critical role in early detection and assessment of disease extent. This study aimed to describe the CT imaging features of pathology-confirmed ROM and to evaluate their associations with orbital involvement, particularly proptosis.

This retrospective, imaging-based study included patients with histopathologically confirmed ROM who underwent noncontrast paranasal sinus CT. CT scans were systematically reviewed for sinus opacification, bone erosion, laterality, orbital involvement, intracranial extension, and proptosis using predefined operational criteria by a single experienced head and neck radiologist. Proptosis was assessed on axial CT images using the interzygomatic line method. Associations between imaging findings and proptosis were assessed using two-sided Fisher exact tests with Benjamini-Hochberg false-discovery-rate (FDR) correction for multiple comparisons; a Firth penalized logistic regression was fitted as an exploratory multivariable analysis because of data separation. Given the retrospective imaging-based design, all analyses are regarded as descriptive and hypothesis-generating.

Of 89 included patients, 47 (52.8%) were male, and 42 (47.2%) were female, with a mean age of 56.25 ± 12.42 years. The most common radiologic finding was opacification of the maxillary sinus (87.6%), followed by the ethmoid sinus (84.3%) and sphenoid sinus (69.7%). In exploratory, unadjusted analyses, ethmoid sinus opacification was more frequent in patients with proptosis than in those without (100.0% vs. 79.4%; two-sided Fisher exact p = 0.034), and patients with sphenoid sinus opacification were older than those without (58.4 ± 10.8 vs. 51.4 ± 14.5 years; Mann-Whitney p = 0.028). After Benjamini-Hochberg correction of the proptosis comparisons for multiple testing, no association remained statistically significant (all q ≥ 0.25); the age comparison was not part of this correction. All findings are therefore descriptive and hypothesis-generating.

CT imaging demonstrates characteristic patterns of sinus, bony, and orbital involvement in ROM. The associations observed between specific CT findings and proptosis were exploratory and did not persist after correction for multiple testing; they should therefore be regarded as hypothesis-generating rather than predictive, while still supporting the value of systematic CT assessment in describing disease extent.