Central Retinal Artery Occlusion in Rhino-Orbital-Cerebral Mucormycosis: An Inflammatory-Prothrombotic State

Abstract

Purpose: 

The aim was to evaluate patient profiles of rhino-orbital-cerebral mucormycosis (ROCM) cases with central retinal artery occlusion (CRAO) postcoronavirus disease 2019.

Design: 

A nonrandomized retrospective case-control study.

Methods: 

The ROCM cases presenting with CRAO were compared with a control ROCM group without CRAO at a tertiary care center. Demography, systemic status, clinical features, histopathology, imaging, and blood profile were assessed for any specific risk factors.

Results: 

A total of 12 patients were seen in the CRAO group and 16 in the non-CRAO group. The male-to-female ratio was 3:1 with a mean age of 49.5 years. In the CRAO group, 75% had diabetes mellitus with mean hemoglobin A1c of 9.03%, and 66.7% had received steroid treatment. All cases were histopathologically confirmed positive for mucor. There was a significant difference in mean D-dimer and serum ferritin between the 2 groups, with higher level in the CRAO group. All patients with CRAO had light perception-negative vision, with total ophthalmoplegia and proptosis seen in 66.7% of cases. Four patients had orbital apex involvement, 5 had cavernous sinus involvement, and 8 had intracranial involvement in the CRAO group.

Conclusions: 

Inflammatory markers D-dimer and serum ferritin were significantly associated with CRAO, suggestive of hyperinflammatory and hypercoagulable state. A high index of suspicion should be maintained in cases with elevated markers and prophylactic anticoagulants can be started to prevent CRAO in a subset of patients.

INTRODUCTION

Mucormycosis is a life-threatening opportunistic fungal infection caused by species of the genus mucor of the order mucorales.¹ The recent rise has been attributed to lowering of immunity in coronavirus disease 2019 (COVID-19) patients specifically due to administration of systemic corticosteroids as well as in diabetics. Other listed factors include immunosuppressive drugs, primary or secondary immunodeficiency, hematological malignancies and hematological stem cell transplantation, solid organ malignancies and solid organ transplantation, iron overload, etc.² It is an angioinvasive fungus causing necrotizing vasculitis leading to thrombus formation and infarction.³ It is diagnosed as possible, probable, and proven and staged on the basis of anatomical involvement.⁴ Rhino-orbital-cerebral cases are being reported in high numbers in COVID-19 patients. The infection initially involves sinuses and then spreads contiguously to the orbit and central nervous system. It is rapidly progressive in absence of treatment and has high mortality.⁵

Central retinal artery occlusion (CRAO) in mucormycosis is attributed to infiltration of vessel with angioinvasive fungus, with a reported incidence of 16% to 20%.⁶ It usually presents in advanced stage of orbital involvement with contiguous involvement of orbital apex with associated signs of ophthalmoplegia, ptosis, and proptosis.⁷ The occurrence is unilateral but can be bilateral in cases of extensive rhino-orbital-cerebral mucormycosis (ROCM). The standard treatment includes debridement of necrotic tissue and systemic antifungal therapy.⁸ Exenteration is done in cases with extensive orbital involvement with no light perception vision to limit its spread to brain. Transcutaneous retrobulbar amphotericin-B (TRAMB) injection has been shown to significantly improve the orbital outcome in cases of mucormycosis.⁹ However, it remains to be seen whether any specific systemic factor is associated with the development of CRAO, other than local spread of the disease as it has low incidence. In this study, we evaluated patients of ROCM with and without CRAO to look for systemic factors favoring the development of CRAO.

SUBJECTS AND METHODS

A prospective study of ROCM patients with CRAO treated at a tertiary care center from June 10, 2021, to August 30, 2021, was done after obtaining clearance from the Institutional Ethics Committee (All India Institute of Medical Sciences, ref no: AIIMS/IEC/2021/3553 dated 07/06/2021). The study adhered to the Declaration of Helsinki. A written informed consent was obtained from all the patients enrolled. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were followed for reporting this study.¹⁰ The study consisted of 2 groups: mucormycosis cases with CRAO and mucormycosis cases without CRAO. The CRAO cases were taken from a pool of ROCM cases which were managed at All India Institute of Medical Sciences, and the non-CRAO cases were selected randomly from the same pool. As the patients with CRAO were limited, randomization and matching were not done.

The inclusion criteria were: (1) biopsy-positive ROCM cases, (2) age over 18 years, and (3) consenting to the study. The exclusion criteria were: (1) previously treated ROCM cases, including medical and surgical treatment; (2) with any other comorbidity, excluding diabetes mellitus (DM) and hypertension which could be an independent risk factor for CRAO, such as venous thromboembolism, ischemic heart disease, and prosthetic heart valves; (3) patients on anticoagulants/antiplatelets or a known coagulopathy; and (4) patients who had received retrobulbar amphotericin-B injection.

The diagnosis and staging of ROCM cases were based on the proposed criteria by Honavar⁴ COVID-19 diagnosis consisted of positive reverse transcription polymerase chain reaction (RT-PCR) report in the past or present. CRAO was diagnosed clinically on the basis of visual acuity (no light perception) and on the fundus examination with indirect ophthalmoscope which is characterized by retinal opacity and cherry-red spot. A detailed evaluation of systemic risk factors including previous treatment taken in the form of systemic steroids, supplemental oxygen therapy, medications, diabetic status, and immunosuppressive disease, if any, was noted. A contrast-enhanced magnetic resonance imaging or computed tomography was done in all cases to record the extent of involvement of sinuses, orbit, and central nervous system. Fungal growth from biopsy or culture was noted in all cases. Complete ophthalmic evaluation in terms of visual acuity (bedside), extraocular movements (0 normal to 4 with complete limitation), proptosis (present or absent), pupillary reactions, anterior (bedside) and posterior segment evaluation (dilated fundus examination with indirect ophthalmoscope) was done. Laboratory investigations were recorded. This included complete blood count with erythrocyte sedimentation rate, liver function test (total bilirubin, conjugated bilirubin, aspartate aminotransferase, alanine aminotransferase, serum albumin, serum globulin, alkaline phosphatase, gamma glutamyl transferase), kidney function test (blood urea, serum creatinine, blood urea nitrogen), C-reactive protein, D-dimer, serum ferritin, lipid profile, procalcitonin, viral markers (HIV, hepatitis C virus, hepatitis B surface antigen), and thyroid panel (T3, T4, thyroid stimulating hormone). The data were compared with a similar set of ROCM patients without CRAO. All the patients received intravenous liposomal amphotericin-B along with other antifungals including Posaconazole wherever required. The patients underwent from single to multiple sinus surgeries including sinus debridement, maxillectomy, and others. The patients who had received TRAMB injection were excluded from the study as TRAMB injections can lead to reduction in the orbital mucormycosis, thus altering the stage of the disease or it may have minimal or no effect. This could have led to variations in the characteristics of cases and controls at any time of the study, and therefore, these patients were excluded. The patients underwent computed tomography scan/magnetic resonance imaging weekly. All the patients were admitted patients. Multidisciplinary approach was followed, and ophthalmological evaluation was done daily.

Statistical Analysis

All the data were entered into a Microsoft Excel spreadsheet. Data from patients with and without CRAO were compared for systemic risk factors, laboratory parameters, ocular involvement, and treatment received. The analysis was carried out using Statistical Package for Social Sciences (version 23 for Windows; SPSS Inc., Armonk, NY). For continuous data unpaired t test and for categorical data χ² test were used to calculate the nominal P value. P value <0.05 was considered statistically significant.

RESULTS

A total of 12 patients with CRAO post-ROCM were seen and 16 patients in the reference arm without CRAO. There was a male predominance with male-to-female ratio of 3:1, with a mean age of 49.5 years in the CRAO group, and sex ratio was nonsignificant between the 2 groups (P=0.305). DM was present in 75% (n=9) of CRAO patients with mean hemoglobin A1c (HbA1c) of 9.03 and 50% of non-CRAO cases with mean HbA1c of 7.78, which was not statistically significant (P=0.180). RT-PCR positive COVID-19 history was present in 50% (n=6) of patients in CRAO arm and 87.5% (n=14) in the non-CRAO arm, and the difference was not significant (P=0.077). Eight patients (66.7%) in the CRAO group had received intravenous steroid treatment previously as compared to 9 patients (56.2%) in the non-CRAO group, but the difference was not significant (P=0.576). Three patients received oxygen therapy for COVID-19 in the CRAO group and none in the non-CRAO group. Only 1 patient had received vaccine for COVID-19. Potassium hydroxide mount and periodic acid-Schiff stain for fungal elements were positive in all cases in both groups, and no other fungal element was seen (Fig. 1). There was no significant difference between the 2 groups with respect to DM status and previous treatment received. The mean D-dimer was 2.29 µg/mL in the CRAO group and 0.09 µg/mL in the non-CRAO group (normal D-dimer <0.5 µg). The mean serum ferritin was 667 and 274.17 µg/mL in the CRAO group and non-CRAO group, respectively (normal values for men: 24–336 µg/l, women: 11–307 µg/l). Both values were significantly elevated as compared to reference arm with P values of 0.0423 and 0.001 (95% CI). There was no significant difference in the rest of the laboratory parameters in the 2 groups, including viral markers and thyroid profile (Table 1).

FIGURE 1: Periodic acid-Schiff stain positive fungal profile highlighted by star.

TABLE 1 - Laboratory Findings

	CRAO	Control	P
1. Sex	Male=9, female=3	Male=9, female=7	0.305
2. DM	n=9 (75%)	n=8 (50%)	0.180
3. HbA1c	9.03 (mean)	7.78 (mean)	0.061
4. Steroid	n=8 (66.7%)	n=9 (56.2%)	0.576
5. COVID-19 RT-PCR +	n=6 (50%)	n=14 (87.5%)	0.077
6. D-dimer	2.29 (mean)	0.90 (mean)	0.042
7. Serum ferritin	667.08 (mean)	274.17 (mean)	0.000042
8. Orbital apex	n=4 (33.3%)	n=4 (25%)	0.629
9. Cavernous sinus	n=5 (41.7%)	n=5 (31.2%)	0.569

COVID-19 indicates coronavirus disease 2019; CRAO, central retinal artery occlusion; DM, diabetes mellitus; HbA1c, hemoglobin A1c; RT-PCR, reverse transcription polymerase chain reaction.

All patients with CRAO had no perception of light whereas in non-CRAO group 15 cases had >6/60 vision and 1 had <6/60 vision. Total ophthalmoplegia (Fig. 2) was present in all cases of CRAO as compared to 4 cases in the non-CRAO group, which presented with partial ophthalmoplegia and the rest had no ophthalmoplegia. Eight cases (66.7%) had proptosis in the CRAO group as compared to none in non-CRAO cases (Table 2). On fundus examination, all cases in the CRAO group had typical CRAO picture on indirect ophthalmoscopy whereas cases without CRAO had a normal fundus. Both groups had 4 cases of orbital apex involvement (P=0.629) (Fig. 3) and 5 cases of cavernous sinus involvement (P=0.569), and the comparison was nonsignificant. Intracranial involvement was noted in 8 cases each of the CRAO and non-CRAO groups. All patients in the CRAO group underwent exenteration while none in the non-CRAO group. However, all the patients in the non-CRAO group underwent some surgical procedures in the form of functional endoscopic sinus surgery, maxillectomy, antrostomy, etc., to curb the infection. In exenterated cases, histopathology specimens demonstrated only mucor, and the optic nerve histopathology, which was available only in 4 cases, showed angioinvasion with mucormycosis.

FIGURE 2: Clinical photograph of a patient with ophthalmoplegia.

TABLE 2 - Ocular Findings

	CRAO (n)	Control
VA
PL negative	12	—
<6/60	—	1
>6/60	—	15
EOM
Full	—	12
Partial ophthalmoplegia	—	4
Total ophthalmoplegia	12	—
Fundus
CRAO +	12	—
Proptosis	8	—

CRAO indicates central retinal artery occlusion; EOM, extraocular muscles; PL, perception of light; VA, visual acuity.

FIGURE 3: Contrast-enhanced magnetic resonance image of orbit in a case of rhino-orbital-cerebral mucormycosis acquired using T1, T2, FLAIR, SW, and DW sequences by using GE discovery 750W 3T magnetic resonance imaging, showing peripherally enhancing mucosal thickening in bilateral maxillary, frontal, ethmoid and sphenoid sinuses, extension of the enhancing soft tissue into left orbital apex, severe fat stranding in the left retro-orbital fat, extraocular muscles appear edematous and shows nonenhancement, resultant proptosis on left side, diffusion restriction in left optic nerve. DW indicates diffusion weighted; FLAIR, fluid attenuated inversion recovery; SW, susceptibility weighted.

DISCUSSION

The mode of practice of ophthalmology has been impacted significantly by the COVID-19 pandemic.^11,12 While the pandemic will be over sooner or later, the SARS-CoV-2 infection will still be around. We must be vigilant in terms of knowing, understanding, and managing the orbital and ocular involvements of mucormycosis in COVID-19 patients.

CRAO in mucormycosis presents in advanced stage of the disease with the invasion of central retinal artery leading to necrotizing vasculitis and thrombosis.¹³ The orbital infection occurs contiguously to involvement of paranasal sinuses. It usually occurs with involvement of orbital apex and is associated with loss of light perception, ptosis, proptosis, conjunctival chemosis, and total ophthalmoplegia. It may also be associated with cavernous sinus thrombosis and intracranial spread. Though CRAO in ROCM occurs due to the angioinvasive nature of fungus, it is considered as a rare complication of ROCM with reported incidence of 16% to 20%.⁶ The factors other than extent of disease leading to CRAO in ROCM have not been reported yet in literature. In this study CRAO with ROCM cases were compared to a similar ROCM group without CRAO to evaluate if there are any significant difference between the 2 groups and to find out any other factors associated with the development of CRAO.

In demographic profile, there was no significant difference in age and sex between the 2 groups, with a mean age of 49.5 years in the CRAO group and male preponderance in both groups. It is similar to the data published by Sen and colleagues in COSMIC-1 report, where 2826 ROCM patients were evaluated with a mean age of 51.9 years and a male preponderance.¹³ About 75% cases of CRAO had diabetes whereas the incidence was 50% in the non-CRAO group, and the difference was not significant. Similarly, the mean HbA1c level was 9.03% in CRAO cases and 7.78% in non-CRAO cases, and the difference was not significant. Though statically not significant, there was higher incidence of DM and HbA1c value in the CRAO group as compared to the other group. Many of the patients in both groups received systemic steroids previously for treatment of COVID-19-related symptoms, though at the time of presentation COVID-19 RT-PCR was positive in 50% and 87.5% of cases. There was no significant difference with respect to the use of steroids. Steroid use has been ascribed as the commonest risk factor for mucormycosis as immunosuppression leads to rapid spread of mucor.¹⁴

The patients were also evaluated for serum inflammatory markers as a part of their systemic investigation. The levels of D-dimer and serum ferritin were significantly elevated in CRAO cases as compared to the other group. D-dimer is a fibrin degradation product, present in the blood after a blood clot is degraded by fibrinolysis and signifies thrombotic phenomenon. On the other hand, ferritin is a mediator of immune 111 dysregulation and is proinflammatory.¹⁵ Acharya and colleagues had reported a case of CRAO secondary to COVID-19 disease who had 115 high levels of D-dimer and ferritin. A hypercoagulable and hyperinflammatory state in COVID-19 was hypothesized to be the cause of CRAO.¹⁶ Montesel et al¹⁷ reported a similar case of CRAO in a COVID-19 patient with inflammation and hypercoagulability as risk factors. However, both cases did not have mucormycosis. In our study, both groups had an increased level of D-dimer and ferritin, but it was significantly higher in the CRAO group. As such, in ROCM cases post-COVID-19, a significant hypercoagulable and proinflammatory state along with extensive mucormycosis may predilect the patient for a faster progression to CRAO.

Orbital apex was involved in 33.3% of patients in the CRAO group and 25% in the other group as seen radiologically. There was no significant difference between the 2 groups. Therefore, angioinvasion of central retinal artery or involvement of optic nerve may not be the sole factor in the development of CRAO. In terms of intracranial involvement, there was no significant difference in the involvement of cavernous sinus, though it was slightly higher in the CRAO group. Cavernous sinus involvement signifies the severity of the disease.¹⁸ The involvement of orbital apex and cavernous sinus is lower as reported by Sen et al.¹⁹ However, the difference can be due to small sample size of our study. In comparison of ocular features, all the patients were light perception-negative in the CRAO group, whereas in the other group only 1 patient had vision <6/60 and others had good vision. As vision loss in ROCM mainly occurs due to involvement of optic nerve or central retinal artery, cases otherwise have good vision. All patients with CRAO had total ophthalmoplegia, with proptosis present in 75% of patients, suggesting involvement of orbital apex. No patient had proptosis and only 25% had partial ophthalmoplegia in the non-CRAO group, suggestive of less severity of ocular involvement. A typical fundus CRAO picture of posterior pole retinal opacification with a cherry-red spot at the macula was present in the CRAO group whereas the fundus was clinically normal in the other group.

An increased cytokine level increases ferritin levels leading to increased free iron.²⁰ Free iron is a suitable medium for the growth of mucormycosis. The low concentration of free iron in plasma protects against pathogens by increasing the activity of cyclin dependent kinase inhibitor p21CIP1/WAF1, thus delaying S phase of cell cycle.²⁰ Mucorales species specifically invade the elastic intima of large and small vessels, causing thrombosis.²¹ The thrombotic cascade can lead to elevated D-dimer levels. The COVID-19 patients also have a procoagulant state evident by elevated D-dimer, prothrombin time, activated partial thromboplastin time, fibrinogen, and cytokines.²⁰ Therefore, coexistent elevation of these factors can predispose a patient to CRAO.

Our study had various limitations. First, it is a nonrandomized study. As the pool of inpatient ROCM cases was small, consecutive cases of ROCM with CRAO and without CRAO were included in the study. Second, due to the limited number of patients, the sample size calculation was not done. There is a lack of photographic documentation, fundus fluorescein angiography, and optical coherence tomography of the clinical picture. Additionally, the central retinal artery biopsy was not available in all the cases with CRAO, which could have delineated the pathology better.

CONCLUSIONS

In comparing the 2 groups, the only factors with a significant difference were D-dimer and serum ferritin. These indicate a severe hypercoagulable and hyperinflammatory state in ROCM patients with CRAO. Therefore, along with severity and extent of angioinvasive mucormycosis, a severe prothrombotic and proinflammatory state can be considered as pathogenetic and an early biomarker in the development of CRAO in ROCM cases, and appropriate prophylactic therapy started to prevent this devastating complication in a subset of patients. However, further research is warranted to substantiate the above findings.