Endogenous Fungal Endophthalmitis following COVID-19 Hospitalization: Evidence from a Year-Long Observational Study

Abstract

Purpose

To describe cases of endogenous fungal endophthalmitis (EFE) post-recovery from or hospitalization for coronavirus disease 2019 (COVID-19).

Methods

This prospective audit involved patients with suspected endophthalmitis referred to a tertiary eye care center over a one-year period. Comprehensive ocular examinations, laboratory studies, and imaging were performed. Confirmed cases of EFE with a recent history of COVID-19 hospitalization ± intensive care unit admission were identified, documented, managed, followed up, and described.

Results

Seven eyes of six patients were reported; 5/6 were male, and the mean age was 55. The mean duration of hospitalization for COVID-19 was approximately 28 days (14–45); the mean time from discharge to onset of visual symptoms was 22 days (0–35). All patients had underlying conditions (5/6 hypertension; 3/6 diabetes mellitus; 2/6 asthma) and had received dexamethasone and remdesivir during their COVID-related hospitalization. All presented with decreased vision, and 4/6 complained of floaters. Baseline visual acuity ranged from light perception (LP) to counting fingers (CF). The fundus was not visible in 3 out of 7 eyes; the other 4 had “creamy-white fluffy lesions” at the posterior pole as well as significant vitritis. Vitreous taps were positive for Candida species in six and Aspergillus species in one eye. Anti-fungal treatment included intravenous amphotericin B followed by oral voriconazole and intravitreal amphotericin B. Three eyes underwent vitrectomy; the systemic health of two patients precluded surgery. One patient (with aspergillosis) died; the others were followed for 7–10 months – the final visual outcome improved from CF to 20/200–20/50 in 4 eyes and worsened (hand motion to LP) or did not change (LP), in two others.

Conclusion

Ophthalmologists should maintain a high index of clinical suspicion for EFE in cases with visual symptoms and a history of recent COVID-19 hospitalization and/or systemic corticosteroid use – even without other well-known risk factors.

Introduction

Endogenous fungal endophthalmitis (EFE) is a sight-threatening condition generally caused by fungemia and hematogenous seeding of fungal pathogens, most commonly Candida and Aspergillus species [1]. EFE sets off at and involves the inner choroid and the retina (chorioretinitis; indicated by fluffy, white chorioretinal lesions) and may then expand into the vitreous cavity (vitritis) and aqueous humorous [2], [3]. EFE, unlike endogenous bacterial endophthalmitis, more commonly presents subacutely, without systemic symptoms of infection – patients may not seek ophthalmic care until a significant visual decline due to vitritis [3]. The most prominent risk factors are indwelling central line, immunosuppression, intravenous drug abuse, and total parenteral nutrition [4].

The coronavirus disease 2019 (COVID-19) pandemic and the widespread use of systemic corticosteroids for hospitalized individuals are speculated to affect several aspects of the ophthalmology practice, given the ocular complications associated with systemic steroids [5]. One such complication may be fungemia and subsequent endogenous endophthalmitis; cases of EFE in patients post-recovery from COVID-19 have recently been reported [6], [7], [8], [9], [10], [11]. We report our one-year experience at a tertiary eye care center; we identified, documented, and described cases of EFE with a recent history of COVID-19 hospitalization.

Methods

The present prospective, single-center audit involved patients with suspected endophthalmitis referred to Labbafinejad University Hospital (a tertiary center affiliated with Shahid Beheshti University of Medical Sciences, Tehran, Iran) within a one-year timeframe – January 2021 to January 2022. Patients with EFE and a history of hospitalization for severe COVID-19 were selected. EFE diagnosis was based on:

• •suggestive symptoms and examination findings, including fluffy white chorioretinal lesions and overlying vitritis;
• •laboratory confirmation by culture or polymerase chain reaction (PCR) test on vitreous samples;
• •lack of a possible exogenous source of infection – i.e., no history of intraocular surgery or trauma within the past year that could have caused the endophthalmitis [3].

All cases with suspected endophthalmitis referred to our center underwent detailed history acquisition and comprehensive ocular examinations, including visual acuity evaluation, indirect ophthalmoscopy, slit lamp examination, and ocular echography. Medical records of patients’ hospitalization due to COVID-19 were obtained and reviewed. The following information was documented/retrieved:

• •demographic data;
• •affected eye;
• •past medical and ocular history;
• •details of COVID-related hospitalization, including duration of hospitalization, admission to intensive care unit (ICU), and treatment regimen;
• •presenting ocular symptoms;
• •time from COVID-hospitalization to ocular symptoms onset;
• •baseline and final visual acuities (VAs);
• •slit lamp and fundus examinations findings;
• •vitreous sample culture and PCR results;
• •details of EFE-related hospitalization including duration and treatment regimens;
• •treatment outcome and follow-up duration.

Color fundus photography, fluorescein angiography, and ocular echography were done as per requirement.

Basic statistical analysis, including calculating the mean, median, and standard deviation, was performed using Microsoft Excel® (Microsoft Corporation). VA values were converted to the logarithm of the minimum angle of resolution (LogMAR) scale. Non-snellen chart-assigned values were converted to the LogMAR scale as follows: “counting fingers (CF): 1.8”, “hand motion (HM): 2.3”, “light perception (LP): 2.8”, and “no light perception (NLP): 3.0”.

Informed consent was signed by patients (or their gaurdians) for collecting and publishing their anonymized data.

Results

Seven eyes with EFE from six patients with recent hospitalization due to severe COVID-19 were identified and met the inclusion criteria. Demographic data of patients, along with their past medical and ocular history, are presented in Table 1 ; their mean age was 55 (range: 35–64), and only one was female (16.6%). Patient No. 5 had bilateral EFE (eyes No. 5 and 6 in Table 2 ). All patients had background conditions, and the most frequently observed ones were hypertension (83%; n  = 5), diabetes mellitus (50%; n  = 3), and asthma (33%, n  = 2). Patient No. 2 had rheumatoid arthritis and was already on immunosuppressants.

Table 1.

Demographic and clinical information of cases.

	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6
Gender	M	M	M	F	M	M
Age	64	35	51	64	53	63
Background conditions	HTN, DM	RA, asthma	HTN, oral candidiasis	Hypothyroidism, DM, HTN, asthma	ESRD, HTN, DM, renal stone	HTN
Primary source of Fungal infection	UC	ETT	UC	UC	UC	UC
Past ocular history	Mild cataract	Dry eye	–	Mild NPDR, macular scar (OS)	–	Glaucoma
Previous ocular surgery	–	PRK	–	–	–	Trabeculectomy
COVID-19 hospitalizationa
Duration of Hospitalization	45 days	30 days	21 days	14 days	14 day	45 days
ICU Admission	Yes	Yes	Yes	No	No	Yes
EFE course and treatment
Duration of Hospitalization	24 days	14 days	12 days	10 days	12 days	9 days
COVID-19 to ocular symptoms interval	35 days	During the COVID-19 course	30 days	21 days	14 days	32 days
Symptoms to diagnosis interval	14 days	3 days	10 days	12 days	3 days	10 days

DM: diabetes mellitus; ESRD: end-stage renal disease; EFE: endogenous fungal endophthalmitis; ETT: endotracheal tube; F: female; HTN: hypertension; IVI: intravitreal; M: male; NPDR: non-proliferative diabetic retinopathy; PO: oral; PRK: photorefractive keratectomy; RA: rheumatoid arthritis; UC: urinary catheter.

^aAll patients had received remdesivir and dexamethasone when hospitalized for COVID-19.

Table 2.

Symptoms, ocular examination findings, treatments, and outcomes.

Eye (Case No.) side	1 (Case 1) OS	2 (Case 2) OD	3 (Case 3) OS	4 (Case 4) OD	5 (Case 5) OD	6 (Case 5) OS	7 (Case 6) OD
Baseline VA	HM	CF1M	CF1M	CF4M	CF4M	CF1M	LP
Final VA	LP	NLP	20/200	20/63	20/50	20/200	LP
Presenting symptoms	↓VA	↓VA, floater	↓VA, red eye	↓VA, floater, red eye	↓VA	↓VA, floater	↓VA, floater, red eye
Vitreous culture	Candida spp.	Aspergillus spp.	No growth	No growth	Candida spp.	Candida spp.	Candida spp.
Vitreous PCR	Candida spp.	Aspergillus spp.	Candida spp.	Candida spp.	Candida spp.	Candida spp.	Candida spp.
SLE findings	AC cells 3+, Flare 1+, hypopyon	AC cells 4+, Flare 2+, fibrin, Hypopyon	AC cells 2+, Flare 1+	AC cells 1+	AC cells 1+, Flare 1+	AC cells 2+, Flare 3+	AC cells 4+, Flare 2+, fibrin
Fundus examination findings	Not visible	Not visible	CWF lesions at posterior pole and vitritis	CWF lesions at posterior pole and vitritis	CWF lesions at posterior pole and vitritis	CWF lesions at posterior pole and vitritis	Not visible
Anti-fungal treatment	AmB (IV) VCZ (PO) AmB (IVI, n = 4)	AmB (IV) VCZ (PO) AmB (IVI, n = 2)	AmB (IV) VCZ (PO) AmB (IVI, n = 3), PPV, PPV + ERMRa	AmB (IV) VCZ (PO) AmB (IVI, n = 2)	AmB (IV) VCZ (PO) AmB (IVI, n = 5)	AmB (IV) VCZ (PO) AmB (IVI, n = 3), PPVb	AmB (IV) VCZ (PO) AmB (IVI, n = 3), PPV
Follow-up duration	10 months	Death	10 months	8 months	8 months	8 months	7 months

AC: anterior chamber; AmB: amphotericin B; CFxM: counting fingers at x meter(s); CWF: creamy-white fluffy lesions; ERMR: epiretinal membrane removal; HM: hand motion; IV: intravenous; IVI: intravitreal; LP: light perception; NLP: no light perception; OD: oculus dexter; OS: oculus sinister; PCR: polymerase chain reaction; PO: oral; PPV: pars plana vitrectomy; SLE: slit lamp examination; Spp.: species; VA: visual acuity; VCZ: voriconazole.

^aDue to incident tractioanl retinal detachment during his hospital stay.

^bDue to incident exogenous fungal endophthalmitis resulting from intravitreal injections.

Regarding information on their COVID-related hospitalization, the length of hospitalization ranged from 2 to 6 weeks (mean ± S.D.: 28 ± 13 days); 67% (n  = 4) had ICU admission. All patients’ treatment regimens consisted of dexamethasone and remdesivir. The time interval between discharge from COVID-related hospitalization to the onset of ocular symptoms ranged from zero (occurring during their hospital stay) to 35 days (mean ± S.D.: 22 ± 12).

Patients presented with severe visual decline (100%; n  = 6), floaters (67%; n  = 4), and red eye (50%; n  = 3). Baseline and final VAs, as well as findings on fundus and slit lamp examinations, are presented in Table 2. Results from imaging studies, including color fundus photography, fluorescein and indocyanine green angiography, and B-scan ultrasonography are presented (Fig. 1 ).

Figure 1.

Color fundus photographs (a–d) of eyes No. 3 (OS), 4 (OD), 5 (OD), and 6 (OS), respectively, showing creamy-white fluffy chorioretinal lesions, vitreous opacities, and marked vitritis; fluorescein and indocyanine green angiography (e), corresponding to eye No. 3 (OS), revealing hypoperfused choroidal regions and large chorioretinal abscess adjacent to the superior arcade; and B-scan ultrasonography (f, g) of eye No. 7 (OD), demonstrating significant vitritis and vitreous opacities.

Intravenous (IV) amphotericin B followed by oral voriconazole were administered empirically for all cases after an initial clinical suspicion of EFE (Table 2). Eyes received multiple intravitreal (IVI) amphotericin B injections (number of injections for each eye presented in Table 2); 3 out of 7 eyes (eyes No. 3, 5, and 6) underwent three-port pars plana vitrectomy (PPV) to salvage their vision. Cases 1 and 2 were not appropriate candidates for surgery, considering their poor health conditions; they did not undergo PPV, despite their severe vitritis.

The mean (± S.D.) baseline and final VAs (LogMAR) were 2.01 (± 0.36) and 1.64 (± 1.08), respectively. The mean (± S.D.) improvement in VA (LogMAR) was calculated to be –0.37 (± 0.89) – when excluding patient No. 2, whose final VA was considered NLP because he passed away due to severe systemic infection, the mean (± S.D.) improvement in VA (LogMAR) was –0.63 (± 0.68).

The identified source of infection was the urinary catheter in four patients and the endotracheal tube in two. Vitreous culture and PCR identified Candida species in 6 eyes (from 5 patients) and Aspergillus species in one eye (case No. 2). All patients were followed up for 7–10 months, except patient No. 2, who passed away due to large aspergilloma in his right lung. Among the survivors (n  = 6), the visual acuity upon the last follow-up visit was unchanged in one eye (No. 7; remained as “light perception” [LP]), worsened in another (No. 1; “hand motion” [HM] to LP), and slightly improved in four (No. 3, 4, 5, and 6) – Table 2.

Discussion

The present study identified and described 7 eyes with EFE post-recovery from, or hospitalized for, severe COVID-19. Several cases of EFE in patients post-recovery from COVID-19 have been reported [6], [7], [8], [9], [10], [11]. A marked male predilection has been observed in the literature; few female patients were reported – e.g., case No. 4 in this study and the one reported by Kaderli et al. [11]. This gender inclination is more likely to reflect the limited number of reported cases given the relatively equal female to male ratios in larger cohorts [12]. The mean age of our patients was 55, similar to previously reported cases of post-COVID-19 EFE [7] and the mean age of 529 EFE cases registered in the National Inpatient Sample (NIS) database (2002–2014) [12] (55 and 54.6, respectively).

The causative organisms in 86% of our infected eyes were Candida species, the most common causative agents in EFE [13]. One eye (patient No. 2) was infected with Aspergillus species; Aspergillus-related EFE is usually more severe, with poorer visual outcomes [1], [14] due to early macular involvement, retinal necrosis, and choroidal damage [15]. Lungs and paranasal sinuses are commonly the original sites of Aspergillus infection before it enters the bloodstream and seeds into the eye [16]. Similarly, the lungs were confirmed as the primary site of infection in patient No. 2 from our study, who developed EFE while still hospitalized for COVID-19 and passed away on his 30th day of hospitalization (Table 1).

All of our cases had recent/concurrent protracted hospitalization for COVID-19, and all had received dexamethasone and remdesivir therapeutic regimens. Hospitalized COVID-19 patients are often exposed to several nosocomial infection risk factors, including prolonged hospital stays, corticosteroid administration, catheterization, and mechanical ventilation [17]. During the pandemic, many hospitals reported higher incidence and mortality rates of candidemia among their patients compared to the pre-COVID-19 era; most cases were hospitalized for COVID-19 [18], [19]. In ICU settings, COVID-19 patients have a 5.1% estimated risk of candidemia – five times that of non-COVID-19 patients (1.1%) [20]. According to the population-based candidemia surveillance by the Centers for Disease Control and Prevention (CDC), COVID-19 patients with candidemia less frequently had classic fungemia risk factors like solid-organ malignancies, chronic liver disease, and recent surgeries. Instead, their predisposing factors were mainly related to the COVID-associated healthcare services they received [21]. In addition to nosocomial risk factors, SARS-CoV-2 infection itself can lead to severe lymphopenia [13], hypothetically hindering one's immunity against fungal pathogens and leading to higher conversion rates of fungemia to hematogenous organ infections.

Screening of patients with Candida bloodstream infection has low practical yield – only about 1% of hospitalized patients with laboratory-confirmed candidemia develop EFE [22], [23]. On the other hand, EFE is a devastating, potentially blinding infection prone to misdiagnosis as non-infectious uveitis [6]. The reasonable approach would be to maintain a high clinical suspicion index of endogenous endophthalmitis when encountering eyes with suggestive features, particularly in patients with risk factors such as recent COVID-19, hospitalization, and systemic corticosteroid use. The outcome may be reasonably well with prompt initiation of therapeutic procedures.

The vitreous sample cultures were negative in 2 of our cases (33.4%); diagnostic yield of vitreous culture – especially if obtained via vitreous tap and not vitrectomy – may be as low as 50%, necessitating prompt administration of systemic anti-fungal agents as soon as a clinical suspicion is made [8], [24]. The visual outcome in our cases depended on the organism, the severity of involvement of the chambers, and the patients’ surgical candidacy.

Conclusion

Protracted hospitalization for COVID-19 and high-dose corticosteroid administration are independent risk factors for opportunistic infectious agents, such as fungi, as well their hematogenous seeding into the eye, giving rise to EFE. Ophthalmologists should maintain a high clinical suspicion index of such entities in cases with visual symptoms who have a history of recent COVID-19 hospitalization and/or systemic corticosteroid use. The importance of timely diagnosis and aggressive intervention in saving eyes with EFE cannot be overemphasized. Patients post-recovery from COVID-19 should be instructed about the visual alarm signs for which they should seek immediate eye care.

Funding

No funding was received for this research.

Disclosure of interest

The authors declare that they have no competing interest.

Availability of data and material

All clinical data pertaining to this study were presented, and no additional data are available.

Authors’ contributions

SF conceptualized and designed this study; SF, MK, and AV conducted the examinations and interventions, and MK acquired and interpreted the data; HN, FE, and SHA drafted the manuscript. All authors have approved the final version of the manuscript.

Consent to participate

All selected patients (or their gaurdians) signed informed consent letters for collection of their data.