key: cord-0791382-8qsegzf7 authors: Kumar GG, Sharath; Deepalam, Saikant; Siddiqui, Ata; Adiga, Chaitra P.; Kumar, Savith; Shivalingappa, Shivakumar Swamy; Acharya, Ullas V.; Goolahally, Lakshmikanth N.; Sharma, Saksham; Andrew, Dhilip; Hosmani, Pradeep; Nair, Satish; Medikeri, Gaurav; Rao, Ravi Mohan; Agadi, Jagadish B.; Kumar, Sujit; Adoor, Gurucharan; Sharma, Suryanarayana; Hegde, Raghuraj; Saini, Jitender; Kulanthaivelu, Karthik title: COVID 19 associated Rhino-orbito-cerebral Mucormycosis – A multi-institutional retrospective study of imaging patterns date: 2022-03-04 journal: World Neurosurg DOI: 10.1016/j.wneu.2022.02.107 sha: 7c564b4e5ad6642869d06c52b1039639d8ba3afa doc_id: 791382 cord_uid: 8qsegzf7 BACKGROUND: Mucormycosis infection of the maxillofacial region and brain has been associated with coronavirus disease 2019 (COVID-19) infection. Mucormycosis was relatively a rare infection before COVID-19, and imaging findings are not very well described. MATERIALS AND METHODS: A retrospective imaging study of 101 patients diagnosed with COVID-19-associated mucormycosis by histopathology and/or culture was performed. All patients underwent computed tomography and/or magnetic resonance imaging based on the clinical condition of the patient and on consensus decision by the team of treating physicians. A simple 3-stage classification system based on imaging findings was adopted. RESULTS: One hundred one cases were included in the final analysis (mean age = 55.1 years; male/female ratio = 67:34). The affected patients had diabetes in 94% of the instances (n = 95), 80.1% (n = 81) received steroids), whereas 59.4% (n = 60) patients received supplemental oxygen. The majority underwent surgical intervention, whereas in 6 cases, patients were treated with antibiotic regimens. Sixty subjects improved following therapy, whereas 18 eventually succumbed to the illness. We noted a significant positive correlation between the imaging stage and outcomes. No association was seen between other clinical parameters and final clinical outcomes. Salient imaging findings include lack of normal sinonasal mucosal enhancement, perisinus inflammation, ischemic optic neuropathy, perineural spread, pachymeningeal enhancement, and presence of strokes. CONCLUSIONS: We describe the imaging findings in the largest cohort of patients with rhino-orbito-cerebral mucormycosis in the context of the current COVID-19 pandemic. A simplified staging system described here is helpful for standardized reporting and carries prognostic information. As India battles a new wave of COVID19 pandemic, Rhino-orbito-cerebral Mucormycosis Clinical features include nasal stuffiness, epistaxis, nasal discharge, swelling of the face, facial and/or orbital pain, worsening headache, proptosis, sudden loss of vision, facial paresthesia, sudden ptosis, diplopia, facial palsy, fever, paralysis, and focal seizures. Early detection and treatment are crucial to improving outcome and is highlighted by worse clinical outcomes in subjects with higher imaging stages. CT in ROCM has a relatively lesser role as compared to the MRI which has overall better soft tissue resolution to assess disease extension. Contrast-enhanced Magnetic resonance imaging (CE-MRI) plays a vital role in both diagnosis and prognostication. Because of superior spatial resolution and soft-tissue contrast, MRI is the preferred imaging modality to evaluate intra-orbital J o u r n a l P r e -p r o o f extension, skull base extension, meningeal involvement, brain parenchymal involvement, perineural and angioinvasion. Recently, Mazzai et al elaborated a pictorial review of Mucormycosis from onset to vascular complications, putting forth a three-stage grading system of involvement 3 . Briefly, the three stages in progressive increments of involvement include sinonasal, orbital, and intracranial involvement. From this multi-institutional study, we intend to highlight the radiological imaging patterns of ROCM that is associated withCOVID-19. As a secondary endpoint we aim to seek possible associations with other clinico-demographic variables. This is a retrospective imaging study of 102 patients who were diagnosed with COVID 19 associated with Mucormycosis by histopathology and/or culture. Institutional approval was obtained for the collection of imaging and basic clinical data. All patients underwent CT and/or MRI with or without contrast, based on the clinical condition of the patient and on consensus decision by the team of treating physicians. Imaging was evaluated and staged by two experienced neuroradiologists independently. The imaging protocol is discussed in Table 1 . Patients either had active COVID19 infection or had recently recovered from COVID19 infection with an interval period of 3 to 30 days. Cases were pooled from four different tertiary care centers in the city of Bangalore-India, during COVID19 second wave in the month of May 2021. Apart J o u r n a l P r e -p r o o f from imaging, other essential clinical details were also collected, including the status of diabetes, administration of steroids or other immune modulators such as Tocilizumab, supplementary oxygen, the interval between COVID19 and diagnosis of Mucormycosis, treatment and outcome. A simple three staging classification system was adopted. The proposed staging system was adapted from Mazzai et al and modified for detailed analysis (Table 2) Wallis non-parametric test revealed no significant association of steroid usage (p=0.53), diabetes (p=0.74), oxygen supplementation (p=0.12) with higher imaging-stages, neither was an association discernible between the aforementioned variables and clinical outcomes. As hypothesized, there was a significant correlation of higher imaging stages with poor clinical outcomes (p=0.0003). In this study, we present a series of patients who developed ROCM in the background of COVID 19 infection. The results of the study can be summarized thus-Disease severity ranged from isolated involvement of sinuses to extensive involvement of the brain. Disease severity on imaging showed a correlation with the clinical outcomes. The purported disease modifying variables, viz diabetes, use of steroids that have been identified in prior works were noted with a similar preponderance in our study 4 . The mean age of the patients was relatively high with male preponderance. The demographic characteristics that we report in terms of the age group and gender distribution of the subject population of our study are similar to that noted in the recent past in the Indian subcontinent 5 . No other significant association was noted between any of the clinical J o u r n a l P r e -p r o o f variables and imaging stages and clinical outcomes. Most patients in the series were managed surgically and eighteen patients eventually succumbed to the illness. The imaging features which correlates to the clinical outcomes can be elucidated by the three-stage system adapted in the study for prudent clinical decision-making. It is to be noted that this system also takes into account the pathogenetic mechanisms involved in the progression. In stage 1, the disease is characterized by the involvement of the nasal cavity and paranasal sinuses ( Table 3) . Early signs of invasive sinusitis are periantral loss of fat planes and lack of enhancement involving turbinates, nasal septum and palate. Lack of normal turbinate enhancement on contrast study is called 'black turbinate' sign and indicates tissue infarction 6 . It needs to be distinguished from benign turbinate hypertrophy which shows mucosal enhancement in immediate scan followed by progressive complete enhancement on delayed images 7 . Non-enhancing mucosal thickening with diffusion restriction is another important early indicator of the disease. Fungal elements within the paranasal sinuses are seen as T2 hypointensity due to iron and other minerals contrary to the hyperintense signal on bland mucosal thickening 8 . The pterygopalatine fossa (PPF) has extensive connections with the deep face and sinuses and its involvement facilitates extension to the deep neck tissues, orbit and brain. Secondary cutaneous mucormycosis shows inflammatory changes in the facial soft tissue as seen in most of our cases. Subtle signs of bony erosion may be seen on CT but during the early disease stage, it may be completely normal. In Stage 2, the orbital compartment is involved with the spread of disease to extraocular muscles and retroorbital fat with the formation of subperiosteal collection in some patients. Narrowing of the posterior globe angle to less than 130 degrees is called globe tenting or 'Guitar pick sign' and it indicates raised intraorbital tension caused by the ongoing retrobulbar inflammation and consequent orbital compartment syndrome 9 . This is not specific for ROCM but indicates poor J o u r n a l P r e -p r o o f visual prognosis 10 . As the disease progress, other findings like optic nerve sheath thickening, involvement of orbital fissures, orbital apex and enlarged superior ophthalmic vein with or without thrombosis may be noted. In case of vision loss, diffusion restriction of the optic nerve may be seen which indicates optic nerve ischemia, and it is another indicator of poor outcome. Skull base and central nervous system involvement suggest Stage 3 disease. Focal pachymeningeal thickening and enhancement mainly involving anterior and middle cranial fossa along the anteroinferior temporal convexity and lateral wall of the cavernous sinus and tentorium cerebelli may be seen. Skull base osteomyelitis is seen as bone marrow oedema and bone erosions. Perineural invasion seen in a significant number of cases in our study cohort may be an underrecognized entity. Extension to the nerves via vasa nervorum is the possibility and more recent studies have shown high affinity of mucor to epithelial growth factor receptor (EGFR) and extracellular matrices in basement membranes, specifically laminin and type 4 collagen and both are abundant in peripheral nerves 11 . Perineural invasion is also determined by the nerve microenvironment and neurotrophic elements secreted along the nerves 12 . Perineural spread is seen as enlargement and variable enhancement of the cranial nerves, especially the trigeminal nerve either in the cisternal portion or within the foramen or canal 13 . The disease may also spread directly into the anterior cranial fossa through the cribriform plate. Meckel's cave involvement is demonstrated distinctly on T2WIas hypointense or dirty signal intensity. Isolated sixth cranial nerve palsy in the absence of other cranial neuropathies suggests a predominant skull base infective process 14 . The higher (43.54 %) composition of Stage 3 that we note in this study may likely pertain to referral bias or may be due to the smaller number of patients included in the previous studies due to the relative rarity of the mucor infection. The higher likelihood of poor outcomes associated with intracranial extension implies that meticulous efforts are needed to identify subtle J o u r n a l P r e -p r o o f signs of the intracranial extension during the early course of the disease which may enable timely initiation of appropriate treatment. We contextualize the spectrum of the imaging findings observed in this study of COVID-19 associated ROCM from a pathophysiological perspective. Cerebritis or brain abscess is a severe complication and occurs either due to the direct spread from the adjacent paranasal sinuses or via a hematogenous route. Fungal abscesses show diffusion restriction along the wall with intracavitary projections and sparing the core of the lesion ( Figure 6F ) 3 . As regards stroke, free iron in the blood and tissue plays a major role in vascular invasion 15 . Endothelial injury by SARS COV2 infection along with the dysregulated immune response may also be contributing to the arterial involvement resulting in both ischemic strokes as well as aneurysms and haemorrhage 16 . Internal carotid artery invasion is possible either by direct invasion or through retrograde spread from the ophthalmic artery. Posterior circulation stroke occurs due to basilar system involvement ( Figure 6 ) possibly by retrograde perineural spread or direct invasion after skull base osteomyelitis. Mycotic aneurysms are seen as variable-sized irregular outpunching on TOF MRA source images. DWI helps in detecting early ischemia/ infarction. Venous involvement leads to thrombosis or thrombophlebitis has also been observed. Cavernous sinus involvement may result in thrombosis and/or ophthalmoplegia 16, 17 . Our study has several limitations. The retrospective nature of this work and the potential referral bias may have influenced the relatively higher percentages of patients assigned under imaging stage 3. Also, since this is a recent and ongoing scenario, we do not have longer-term follow-ups on our patient cohort. To conclude, we describe the imaging findings of ROCM associated with COVID19 by employing an adapted simplified (3-stage) staging system that can be readily employed in day-to-day practice and help standardize and improve communication between radiologists and clinicians. Disease severity ranges from isolated involvement of sinuses to extensive involvement of the brain. The clinical outcomes of COVID-19 patients with ROCM progressively scale alongside the graded severity on imaging. This is a rapidly evolving infection with high morbidity and mortality; hence an early diagnosis of ROCM is crucial with imaging playing a key role in the staging of the disease process and assessing the involvement of deeper structures that may not be evident clinically. Whilst both CT and MRI are useful, CE-MRI is the investigation of choice and helps in mapping out the disease extent, particularly with regards to deep facial, orbital and intracranial spread. • 2A -Involvement of medial and/ or inferior orbital compartment only. • 2B -Diffuse unilateral orbital involvement with or without optic nerve, nasolacrimal duct and vascular involvement. • 2C -Bilateral orbital involvement. • 3A -Involvement of pachymeninges, cribriform plate, cavernous sinus/ Meckel's cave. • 3B -Vascular involvement (infarct/ bleed)/ perineural spread and skull base involvement. • 3C -Leptomeningitis, cerebritis or abscess formationfocal or diffuse involvement. F: Coronal T1W fat saturated post contrast image shows enhancement along the right infraorbital nerve within the infraorbital foramen suggesting early perineural spread of the disease (arrow). Note the right maxillary sinus mucosal thickening and adjoining orbital fat stranding. He was treated and discharged 5 days back from the hospital with moderate COVID19 pneumonia and CT severity score of 13/25. He was a known hypertensive and diabetic on oral hypoglycemics agents for the past 12 years and briefly treated with supplementary oxygen and steroids for 12 days during hospitalization. J o u r n a l P r e -p r o o f Figure 6 : Top row -Key imaging features of complicated stage 3 disease in a 36-year-old man, having recovered from COVID19 22 days back. He was treated and discharged from hospital with moderate COVID19 pneumonia and CT severity score of 10/25. He had FESS and orbital exenteration following which he was doing well for three days and on day four was found in altered sensorium and succumbed to the illness next day. Epidemiology and clinical manifestations of mucormycosis Rhino-Orbital Mucormycosis Associated With COVID-19 Imaging features of rhinocerebral mucormycosis: from onset to vascular complications COVID-19 and orbital mucormycosis Mucormycosis in COVID-19: A systematic review of cases reported worldwide and in India Sign: An Early MR Imaging Finding of Nasal Mucormycosis The Black Turbinate Sign, A Potential Diagnostic Pitfall: Evaluation of the Normal Enhancement Patterns of the Nasal Turbinates Angioinvasive rhinocerebral mucormycosis with complete unilateral thrombosis of internal carotid artery-case report and review of literature Guitar pick sign" on MRI Globe tenting: a sign of increased orbital tension Pathogenicity patterns of mucormycosis: epidemiology, interaction with immune cells and virulence factors Rhinocerebral mucormycosis: Pathology revisited with emphasis on perineural spread Perineural spread of rhino-orbitocerebral mucormycosis caused by Apophysomyces elegans Imaging of mucormycosis skull base osteomyelitis Imaging of cerebrovascular complications of infection Mucormycosis of the Central Nervous System Carotid Artery Occlusion by Rhinoorbitocerebral Mucormycosis. Case Reports in Surgery Bottom row -Key imaging features of complicated stage 3 disease in a 46-year-old man, recovered from COVID19 7 days back and presented with holocranial headache, left hemi-facial pain and stuttering right hemiparesis progressing to a locked-in state in next 12 hours. He succumbed to illness the following day. He was initially treated and discharged from the hospital with moderate COVID19 pneumonia and CT severity score of 14/25.