key: cord-0712756-6u6rr5v9
authors: Metwally, Maha Ibrahim; Mobashir, Mohamed; Sweed, Ahmed Hassan; Mahmoud, Sara Mohamed; Hassan, Aya Gamal; ElKashishy, Kamal; Eesa, Mohamed; Elnashar, Ismail; Elmalt, Ashraf; Elsayed, Ahmed Ibrahim; Idris, Shaimaa Khaled; Elshetry, Al Shaimaa Fathi
title: Post COVID-19 Head and Neck Mucormycosis: MR Imaging Spectrum and Staging
date: 2021-12-15
journal: Acad Radiol
DOI: 10.1016/j.acra.2021.12.007
sha: a049817589e6ad44a976a8e17bfa5eb6ac075d24
doc_id: 712756
cord_uid: 6u6rr5v9

OBJECTIVE: : To develop a systematic approach for magnetic resonance imaging (MRI) analysis, imaging spectrum, and classification system for the staging of post-COVID-19 head and neck mucormycosis. METHOD: : The study included 63 post-COVID-19 patients with pathologically proven mucormycosis who underwent head and neck MR imaging. Three independent radiologists assessed the imaging spectrum of mucormycosis, MRI characteristics of sino-nasal mucormycosis, and extra-sinus extension, and submitted a final staging using a systematic approach and a proposed categorization system. A consensus reading was considered the reference imaging standard. The kappa statistics were used to assess the categorization system's diagnostic reliability. RESULTS: : The overall interreader agreement of the MR staging system was very good (k-score= 0.817). MR imaging spectrum involved localized sino-nasal mucormycosis (n=7 patients, 11.1%), sino-nasal mucormycosis with maxillo-facial soft tissue extension (n=28 patients, 44.5 %), sino-nasal mucormycosis with maxillo-facial bony extension (n=7 patients, 11.1%), sino-naso-orbital mucormycosis (n=13 patients, 20.6%), and sino-nasal mucormycosis with cranium or intracranial extension (n= 8 patients, 12.7%). Extra-sinus extension to the orbit and brain did not have significant association with involvement of the posterior ethmoid/sphenoid sinuses and maxillo-facial regions (P >0.05). MRI-based staging involved four stages: stage 1 (n=7, 11.1%); stage 2 (n=35, 55.6%), and stage 3 (n=13, 20.6%), and stage 4 (n=8, 12.7%). Involvement of the bone and MR-based staging were significant predictors of patients’ mortality P=0.012 and 0.033, respectively. CONCLUSIONS: : This study used a diagnostic-reliable staging method to define the imaging spectrum of post-COVID-19 head and neck mucormycosis and identify risk variables for extra-sinus extension.

has predisposed to secondary infections, especially in critically ill patients with a tenfold increase in fungal infections (1) . Mucormycosis, a rare fungal infection, is the most fulminant form of Zygomycosis caused by Mucorales species of the phylum Zygomycota (2) . Literature has established a link between COVID-19 infection, corticosteroid use, immunocompromised states such as diabetes mellitus (DM), and mucormycosis (3) (4) (5) (6) (7) .

Head and neck Mucormycosis starts with sporangiospores inhalation, direct extension to the paranasal sinuses, and hyphae angioinvasion in immunocompromised hosts, resulting in subsequent necrotizing vasculitis, fungal thrombi, and tissue infarctions (8, 9) . The fungus spreads through occlusion of vascular vessels or direct invasion through the superior orbital fissure, cribriform plate, whereas vascular invasion occurs through ophthalmic vessels.

Intracranial involvement is also caused by carotid artery or perineural invasion (10, 11) .

Mucormycosis involvement in the head and neck is divided into four categories: isolated nasal, sino-nasal, rhino-orbital, and rhino-orbital-cerebral mucormycosis (12) .

Mucormycosis diagnosis depends on the clinical features, pathological criteria, and imaging findings that assess the extent of involvement (13) . Early detection and treatment of mucormycosis can save lives, since a one-week delay in treatment raises mortality from 35% to 66%. (14) . Since surgical debridement of devitalized tissue is crucial for the optimal therapy of mucormycosis (15) , radiologists should be aware of the anatomical sites that must be evaluated in order to determine the extent of mucormycosis infection and be familiar with the different imaging patterns.

As an extension of literature discussing COVID-associated mucormycosis (1, 16, 17) , this study aims to define the imaging spectrum and distinct imaging features of head and neck Mucormycosis after COVID-19 infection, as well as propose an imaging-based classification system based on mucormycosis regional involvement and test its reliability.

The study was permitted by our ethics committee (IRB approval no. 7014). Written informed consent was acquired from all patients in our study. The study was compliant with the ethical principles of the declaration of Helsinki. This manuscript was reported in adherence with the Strengthening the Reporting of Observational Studies in Epidemiology.

From June 2021 to September 2021, a cross section study was conducted. All of the patients had a history of COVID-19 infection (which was confirmed by laboratory, clinical, and radiological means) and were referred to our radiology department in (BLINDED) for magnetic resonance imaging (MRI) at the request of their otorhinolaryngologists. Inclusion criteria included: (i) confirmed infection with COVID-19 within the previous three months (ii) clinically suspected mucormycosis patients (e.g., black necrotic turbinate, blood-tinged nasal discharge with ipsilateral facial pain, peri-orbital or facial swelling, skin induration and discoloration, ptosis of the eyelid, and ophthalmoplegia). Exclusion criteria were (i) patients refused surgical interference (n=1) (ii) absent pathological evaluation of the surgical specimens (n=1) (iii) patients who previously had sino-nasal fungus infection or underwent prior endoscopic sinus surgery (n=3) (iv) patients with pathologically proven mixed fungal infection or bacterial sinusitis (n= 6). Our final sample included sixty-three patients.

All patients underwent paranasal MR examination using a 1.5-Tesla MRI system (Achieva-class IIa, Philips Medical Systems) with a standard head coil. The paranasal MRI protocol included According to the total regional expansion, patients were assigned the highest stage.

Surgical debridement was performed with an average lag time of 3.2 days from admission and intervention; during this time, all patients received systemic antifungal medication and underwent MR examination. According to MRI findings, the endoscopic procedure was recommended. During the operation, all necrotic tissues were removed until healthy vascular tissue was visible. Enucleation was performed in cases of intraconal soft tissue extension.

Pathological analysis of surgical specimens validated the diagnosis of mucormycosis. The presence or absence of histological criteria for mucormycosis previously identified in the literature, such as the presence of broad, aseptate, or pauci-septate hyphae with wide-angle branching in the afflicted tissue with/without tissue invasion, were interpreted on specimens (18) .

The Shapiro-Wilk test was used to check and confirm the normal distribution of the data. When applicable, the data of the recruited patients were expressed as mean, standard deviation (SD), and range, or number and percentage. The association between involvement of the posterior ethmoid and sphenoid sinuses, maxillo-facial spaces, and extra-sinus extension to the orbits and brain was determined using Pearson's correlation. Patients' outcomes (independent variable) were dichotomized (died and survived) and multiple logistic regression analysis was used to assess the relationship between multiple predictor variables (demographic, clinical, and imaging characteristics, and staging system) and patients' outcomes in the early post-operative period.

Weighted kappa scores (k-score) were employed to assess the diagnostic reliability of the 

63 consecutive post-COVID-19 patients (22 males and 41 females) were enrolled in the study with a mean age ± SD of 58.3 ± 8.47 years (age range, 40-75). The mean time interval between COVID-19 and mucormycosis infection was three weeks (range, 1-8 weeks). The total number of deaths during the early postoperative period (up to day-8 post-operative) was 11 (17.5%) patients (Table 1) .

The imaging spectrum of mucormycosis was localized sino-nasal mucormycosis (n=7 patients, 11 .1%); sino-nasal mucormycosis with maxillo-facial soft tissue extension (n=28 patients, 44.5 %); sino-nasal mucormycosis with maxillo-facial bony extension (n=7 patients, 11.1%); sinonaso-orbital mucormycosis (n=13 patients, 20.6%); sino-nasal mucormycosis with cranium or intracranial extension (n= 8 patients, 12.7%). Table 2 shows the pattern of sinus involvement, signal intensity, and enhancement characteristics on MR imaging. The most commonly involved paranasal sinus (PNS) was the maxillary sinus (n=63, 100%). 45/63 (71.4%) patients had multiple sinus affection (≥3 sinuses). Unilateral sinus involvement was more common (n=35/63, 55.6%) than bilateral sinus affection. In contrastenhanced T1WI, 90% of the completely opacified maxillary sinuses displayed heterogeneous enhancement manifested with multiple enhanced septa on a background of non-enhanced sinus, giving a -soap bubble appearance‖ sign.

The sites and patterns of extra-sinus extension are summarized in Table 3 . The most frequent site of extra-sinus extension was the pterygopalatine fossa (n= 49, 77.8%).

The overall interreader agreement of the MR staging system was very good (κ = 0.81). Table 4 displays the staging of mucormycosis according to regional involvement in the head and neck.

The prevalence of mucormycosis in stages 1, 2, 3, and 4 mucormycosis was 11.1, 55.6, 20.6, and 12.7%, respectively.

Based on the results of Pearson's correlation, there was no statistically significant association between the involvement of the posterior ethmoid and sphenoid sinuses and extra-sinus extension to the orbit and brain (P= 0.567 and 0.796, respectively). Similarly, there was no significant association between the involvement of the maxillo-facial spaces by mucormycosis and extra-sinus extension to the orbit and brain (P= 0.877 and 0.672, respectively).

Multiple variables were evaluated to determine their predictive value for the patients' outcomes.

Based on multiple logistic regression analysis, MRI-based stage and bone invasion indicated a significant correlation with patients' mortality, P≤0.001. The mortality rate was higher in stage 4 (7 patients) and stage 2b (4 patients 

All patients underwent endoscopic debridement of devitalized fungal infected tissue through modified Denker approach. In addition, six patients underwent inferior maxillectomy, while another one underwent facial resection. Two patients with non-serviceable eyes were treated by orbital exenteration. In patients with partial visual loss, orbital debridement was performed. Operative findings demonstrated a pink or pale, firm, gritty mass with little vascularity infiltrating the underlying tissue.

All of the specimens examined contained hyalinized necrotic tissue with aggregates of inflammatory cells and numerous mucormycosis fungal hyphae. These hyphae infiltrated the tissue and/or bone. There was no evidence of atypia or malignancy..

The current study demonstrated a systematic approach for the analysis of MR imaging in head and neck mucormycosis patients, allowing for a thorough assessment of mucormycosis regional involvement and potential sites of extension in the head and neck region. Also, it revealed different imaging spectra of the disease including, localized sino-nasal mucormycosis, sino-nasal mucormycosis with maxillo-facial soft tissue extension, sino-nasal mucormycosis with maxillo-facial bone extension, sino-naso-orbital mucormycosis, sino-nasal mucormycosis with cranium or intracranial extension. We identified staging system of mucormycosis regional involvement in the head and neck and demonstrated its diagnostic reliability. This classification should improve communication between radiologists and otorhinolaryngologists and standardize the reporting of mucormycosis cases.

In general, head and neck mucormycosis is the most frequent variant of mucormycosis infection that accounts for 50% of the patients (19) . It is more common in diabetics and immunosuppressed patients (20, 21) . In the current study, we focused on assessing head and neck mucormycosis in post-COVID-19 patients. Superinfection and coinfection with COVID-19 are still being investigated to determine whether they are caused by the virus or its management.

(1). High-dose corticosteroids, which are commonly used to treat COVID-19, especially in the presence of DM may result in hyperglycemia and acidosis which increases the risk of mucormycosis by eliciting phagocytes dysfunction (3) . Several research investigations have attempted to establish a relationship between DM and mucormycosis (8, 22) . According to two meta-analyses, DM was the most common predisposing factor for mucormycosis, accounting for 40-64% of cases (23, 24) , and a global rate of 17-88% (25) . In our study, diabetes mellitus was reported in 81% of mucormycosis cases and history of corticosteroid usage was positive in 82.5%.

The black turbinate sign, (non-enhancing, hypointense turbinate), sinus opacification, airfluid level, variable intensity within the sinuses on T1-and T2-weighted images (predominant hypointense on T2), obliteration of the nasopharyngeal planes, preantral fat infiltration, loss of contrast enhancement of the sino-nasal mucosa and extraocular muscles, inflammatory changes in the extraocular muscles and fat, and cerebral leptomeningeal enhancement were all previously documented MR findings of acute fulminant invasive fungal sinusitis (AFIFS) (26, 27 ). In the current study, we concluded that mucormycosis imaging features in post COVID-19 patients did not vary from those previously documented in AFIFS. Except for cerebral leptomeningeal enhancement, all of the previously described MRI features of mucormycosis were detected in varying degrees in the current study. The extra-sinus fat infiltration is caused either by vascular congestion-related edema or fungal infiltration, and it can even happen before bone osteolysis occurs because the fungus spreads primarily through perivascular channels (30, 31). Therefore, in our study, patients with maxillo-facial extension were assigned distinct subcategories in the proposed staging system Nonetheless, there was no statistically significant association between infiltration of the posterior ethmoid and sphenoid sinuses, as well as the maxillo-facial spaces, and extra-sinus extension to the orbit and brain in our study. Finally, we concluded that the reported MRI-based mucormycosis staging had a significant relationship with patient mortality, suggesting that it could be a useful predictor of patient outcome, and recommend further evaluation of its prognostic value in other literature. sample size, which was due to our inclusion and exclusion criteria. Second, we relied on the patients' short-term follow-up since we chose to explore early death rates among mucormycosis patients rather than the frequently measured late (e.g., 90-day) mortality rates. Finally, we did not take into account the pulmonary condition of post-covid-19 patients, which could affect the death rate at different stages.

In conclusion, in post-COVID-19 patients, this study established a systematic approach for appropriate MRI assessment of mucormycosis infection in the head and neck region, as well as distinct imaging spectra of head and neck mucormycosis. Also, an imaging-based diagnosticreliable staging system of mucormycosis was assigned, which could enhance the accuracy of MR reporting and consequently better outcome. Table 2 . MR imaging characteristics of sinonasal mucormycosis Table 3 . Analysis of mucormycosis extra-sinus extension and related MRI findings Table 4 . MR imaging-based classification of mucormycosis staging. 

Sino-orbital mucormycosis in a COVID-19 patient: A case report

Diabetes mellitus as the major risk factor for mucormycosis in Mexico: epidemiology, diagnosis, and outcomes of reported cases

Steroid-induced hyperglycemia: An underdiagnosed problem or clinical inertia? A narrative review

Novel perspectives on mucormycosis: pathophysiology, presentation, and management

Lymphopenia predicts disease severity of COVID-19: a descriptive and predictive study

COVID-19 infection may cause ketosis and ketoacidosis

The endothelial cell receptor GRP78 is required for mucormycosis pathogenesis in diabetic mice

Disease entities in mucormycosis

Central retinal artery occlusion as the presenting manifestation of invasive rhino-orbital-cerebral mucormycosis

Perineural spread of rhinoorbitocerebral mucormycosis caused by Apophysomyces elegans

Silkiss (2020) Rhino-orbital-cerebral mucormycosis (ROCM) and associated cerebritis treated with adjuvant retrobulbar amphotericin B Am

Invasive and non-invasive fungal rhinosinusitis-a review and update of the evidence

Mucormycosis with orbital compartment syndrome in a patient with COVID-19

ECIL-6 guidelines for the treatment of invasive candidiasis, aspergillosis and mucormycosis in leukemia and hematopoietic stem cell transplant patients

Rhino-orbital mucormycosis associated with COVID-19

Imaging features of rhinocerebral mucormycosis: A study of 43 patients

Revised definitions of invasive fungal disease from the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) Consensus Group

Mucormycosis of the head and neck

Novel perspectives on mucormycosis: pathophysiology, presentation, and management

Epidemiology and outcome of zygomycosis: a review of 929 reported cases

Pathogenesis of mucormycosis

The epidemiology and clinical manifestations of mucormycosis: a systematic review and meta-analysis of case reports

A meta-analysis of survival factors in rhino-orbital-cerebral mucormycosis-has anything changed in the past 20 years?

Global epidemiology of mucormycosis

Usefulness of computed tomography and magnetic resonance in fulminant invasive fungal rhinosinusitis

The ‗‗Black Turbinate''sign: an early MR imaging finding of nasal mucormycosis