Epidemiology of COVID-19–Associated Mucormycosis

Abstract

Purpose of the Review

To describe the epidemiology and risk factors for Coronavirus disease-19 (COVID-19)–associated mucormycosis (CAM) based on current published literature.

Recent Findings

COVID-19 is associated with an increased risk of secondary infections. Mucormycosis is an uncommon invasive fungal infection that typically affects people with immunocompromising conditions and uncontrolled diabetes. Treatment of mucormycosis is challenging and is associated with high mortality even with standard care. During the second wave of the COVID 19 pandemic, an abnormally high number of CAM cases were seen particularly in India. Several case series have attempted to describe the risk factors for CAM.

Summary

A common risk profile identified for CAM includes uncontrolled diabetes and treatment with steroids. COVID-19–induced immune dysregulation as well as some unique pandemic specific risk factors may have played a role.

Introduction

The coronavirus disease-19 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-COV-2) is associated with a wide spectrum of disease including severe illness and risk of mortality. COVID-19 increases risk of secondary bacterial and fungal infections [1] due pulmonary tissue damage, immune dysregulation, as well as treatment related factors including invasive interventions or immunosuppressive therapies for COVID-19 [2]. These may worsen the prognosis of disease and increase mortality [3]. Multiple case series and cohort studies have highlighted the emergence of COVID-19–associated fungal infections including COVID-19–associated aspergillosis (CAPA), invasive candidiasis and COVID-19–associated mucormycosis (CAM) [4]. CAM gained widespread attention due to an unusual surge of cases that occurred during the second wave of COVID-19 in India in 2021 [5].

Mucormycosis is an invasive fungal infection associated with high mortality even with standard care [6]. Management is challenging with the need for prolonged antifungal therapy and surgery. This can vary based on location of infection, type of fungus and underlying host condition. Traditional risk factors for invasive disease have included hematological malignancies, organ transplantation, uncontrolled diabetes, neutropenia, use of immunosuppressive therapies including steroids, trauma, and burns [7]. Emerging case series have increased our understanding of the risk factors of CAM. The objective of this review is to describe the epidemiology and risk factors of CAM based on current published literature.

Prevalence and Incidence

Background Prevalence

The global incidence of mucormycosis is unknown in both pre-COVID and COVID periods. Studies vary in terms of populations studied, and by diagnostic criteria used [8, 9]. Case identification may be limited by low diagnostic sensitivity of conventional diagnostic methods. A population-based surveillance study in 1992 from northern California documented prevalence of mucormycosis at 0.17 cases per 100,000 [10]. Increasing temporal number of cases have been reported from hospital series from Belgium (0.14/10,000 patient days, 2009) [11] and Spain (0.3/10,000 hospital admissions, 2015) [12] and a population-based study from France (0.12/100,000, 2006) [13].

In the USA, mucormycosis related hospitalization was estimated to be 0.12/10,000 hospital discharges in the period 2005–2014 based on a retrospective analysis of a hospitalization database [14]. A similar rise in cases is documented in India and Iran from single center case series [15, 16]. Modeling data has indicated a baseline prevalence of 14 cases per 100,000 population in India. This is nearly 70-fold higher than the global prevalence of 0.2/100,000 [17].

In the pre-COVID literature, reports from India [8, 15, 18], Iran [16], and Mexico [19] cite prominent rhino-orbito-cerebral infection (~44–75% cases) as the most common diagnosed syndrome. In India, poorly controlled diabetes seems to be the most common risk factor whereas hematologic malignancy and transplantation seems more common in Europe [20], Australia [21] and the USA [14]. Proportion of pulmonary disease (30–44%) and disseminated disease (15–40%) seems to be higher in reports from developed countries though this may not be estimable from the prior studies due to possible differences in diagnostic ascertainment.

COVID Surge in CAM

While prior estimates of mucormycosis estimate prevalence burden of approximately 171,500 cases for India [22], during the COVID-19 surge, around 41,000 cases of COVID-associated mucormycosis were reported from India alone by the end of June 2021 with around 3200 deaths [23]. Case burden at referral sites seemed to mirror the COVID-19 surge with one multicenter study indicating CAM cases peaking from 0.03% of all hospital admissions at beginning of Jan 2021 to peak of 2.5% of all admitted patients in May 2021 [24]. Another multicenter retrospective study of COVID-19–associated mucormycosis (CAM) among hospitalized COVID-19 patients estimated prevalence at 0.27% [25]. During the COVID-19 surge in India in mid-2021, there was a 2.1-fold increase in number of Mucormycosis cases in 2021 as compared to a pre-COVID period [25]. This very high number of reported cases of CAM during the COVID-19 surge in India suggests that the many traditional risk factors like diabetes, as well as COVID-19–related unique risk factors including perturbation of glycemic control due to SARS CoV2 infection, COVID-19 therapies, as well as evolving diagnostic procedures with wider recognition of the condition and pandemic related factors including access to care aligning in a perfect storm may have played a role in fueling this surge.

Clinical Aspects

Majority of the CAM cases reported have been sinonasal or rhinoorbital (41–74%) or rhinoorbitocerebral (8–37%) [5, 25–29] though pulmonary and disseminated disease is reported more commonly outside India [30] Table 1 includes details of select published data on clinical aspects of CAM. Median age of patients with CAM seems to be around 50–55 years. Demographic and underlying risk factor data show male predominance (70–80%). Diabetes seems to be associated with 66% of cases reported from India vs. 55% of cases from elsewhere [31]. Prevalence of diabetes seems to be anywhere from 55% to up to 100% in reported case series with CAM [25–27, 29, 32–34] and several case series reported new onset diabetes in subjects with no prior history of diabetes [25–28, 35, 36].

Table 1.

Published case control studies, reviews, and select case series with at least 50 patients with CAM

Author	Publication period and details	Number of cases	Median/mean age	Type of CAM	DM prevalence	COVID details	Steroid use	Surgery	Death	Details
Dubey S, Mukherjee D, Sarkar et.al. [90]	Observational, retrospective, study, single center in eastern India, Apr-Jun 2021	55	53 years, male 63.6%	ROCM CNS involvement 34.5%	100%	80% need hospital admission and 60%treated with supplemental oxygen	60%	100%
Ramaswami A, Sahu AK, Kumar A et al. [62]	Observational, retrospective single center study in India, emergency room based, May-Jun 2021	70	44.5 male 60%	ROM/ROCM 17% CNS involvement	70%, 7% diagnosed during COVID illness	Details NA 75.7% recovered from COVID	70% 57% inappropriate use of steroids	100%	23% at 14 days
Rajabi MT, Aghajani A, Rafizadeh SM et al. [91]	Observational, retrospective multicenter study in Iran, Feb–Jul 2021	132	61.6 years, 60% male	ROCM 25% CNS involvement	94.7%	85% admitted to hospital, 32% admitted to ICU	92.4% treated with steroids		9.1% (12.5% males, 3.8% females)	Higher HgA1c, severe vision involvement and brain involvement associated with mortality in multivariate analysis
Paidisetty P, Nagose V, Vaze V et al. [92]	Observational retrospective cross sectional single center study in India, Apr -Jul 2021	51	53 years, male: female 1.83	ROCM 25.5% CNS involvement	67% diabetic, 53% poorly controlled	25.5% received oxygen; 6% had severe disease	67% received steroids	NA	In hospital mortality 6%
Arora S, Narayanan S, Fazzari M et al. [64]	Retrospective multicenter online database of cases India, Nepal Bangladesh Jul 2021–Jun 2022	341 patients	51.7 years male 21%	ROCM 93.5%, Cutaneous 1.2%, gastric 0.9% CNS involvement 15%	84% median A1c 8.8 %		86%. 52% received steroids >10 days	75.7%	24.3% at day 42	Older age, diabetes history, lower BMI and CNS involvement associated with higher risk of mortality Spread beyond sinus disease associated with 77% inc risk of mortality
Rastkar M, SeyedAlinaghi S, Asanjarani B et al. [93]	Retrospective, multicenter observational cross-sectional study in Iran, Mar–Dec 2021	73	57.4 years, male 66%	ROCM	23%	20.5% admitted to ICU for COVID-19	94.5% received steroids	NA	42.5%	ICU admission associated with mortality. Diabetes history or steroid use not a/w mortality
Meher R, Wadhwa V, Kumar V et al. [94]	Observational single center study in India	131	52.7y male, 49.8y female 65% male	ROCM	94% 77% were newly diagnosed diabetics	15% had no known h/o COVID 27% needed oxygen; 5.4% needed ventilatory support.	60%	100%	4.5%
Sebastian SK, Ponnuvelu S, Sharma Y et al. [95]	Retrospective observational cohort, single center, with embedded cross sectional study looking at invasive fungal sinusitis (15 mucor, 15 Aspergillus and 7 mixed cases) Jan 2016-Jul 2021	58 patients 37 COVID associated invasive fungal sinusitis and 21 non COVID cases	52.5 years male 78%	ROCM CNS involvement 40.5%	68% vs 100% in non-COVID mucormycosicases Mean A1c 8.6 vs 11.3 in non COVID cases	62% COVID recovered	78%	73%	32.4%	In combined group, stage 3 and 4 and lack of surgery associated with mortality
Pal R,Singh B, Bhadada SK et al. [63]	Scientific systematic literature review of 30 published case series and reports May 2021	99	52.6 78% male	86% ROC, CNS involvement 28%, pulmonary 10%, disseminated 1% 20% had species diagnosis. 85% Rhizopus species, 10% Rhizopus and 5% Lichtheimia	85% 16% newly diagnosed diabetes	63% with current COVID-19; 37% recovered 67% severe or critical illness	85%	81%	33%	Higher risk with pulmonary mucormycosis vs ROCM alone and with medical management alone
Bhanuprasad K, Manesh A, Devasagayam E et al. [32]	Single center prospective cohort study India with embedded case control Jul 2020-Jun 2021	132 cases (32 controls who were PCR neg for COVID)	51 years, 78% male	ROCM 100%, CNS involvement 29.5%	97% DM, 39.5% newly detected DM Mean HgA1c 10.7	O2 use was 14.4%	55%		In hospital mortality only 9.8%	Steroid use, uncontrolled diabetes associated with higher risk of CAM based on PCR diagnosis
Arora U, Priyadarshi M, Katiyar V et al. [68]	Single center case control study in India Apr-May 2021	152 cases and 200 COVID-19 controls	48.2 years Male 64.2%	ROCM, 14.5% CNS involvement, pulmonary 1.2%, disseminated 0.6% (79% definite diagnosis)	92% 35% were newly diagnosed Mean A1c 10 in CAM vs 7.1 in controls	20% asymptomatic, 47% needed hospitalization, severe disease in 10%. O2 supplementation needed in 35%	66%		NA	Farming, gardening or labor associated with higher risk of CAM, residence in rural area associated with higher risk of CAM, residence or workplace near construction site a/w higher risk of CAM. Zinc use found to be associated with lower risk of CAM In multivariate model, history of diabetes, steroid use, severe COVID, highest blood glucose >200mg/dl, number of COVID-19 swab tests, zinc use, and prolonged use of surgical or cloth mask a/w inc risk of CAM
Chakravarty, Gupta, Tilak et.al. [96]	Single center retrospective case series, India May–July 2021	208 CAM cases	52.8 years, 64% male	98% ROCM	95.7% of whom 42% were newly diagnosed.	39%.5% needed oxygen therapy	65%	90.8%	28% mortality at 3 months. 55% deaths occurred within 15 days of admission. At 6 months 32.7% mortality	Uncontrolled hyperglycemia, SARS-CoV-2 infection, rampant use of antibiotics, zinc supplementation and steroids were some of the risk factors for mucormycosis.
John T, Jacob C Kontoyiannis [97]	Scientific literature search and systematic review of published data Dec 2019–4/2021	41 cases	55 years, 83% male	Sinus 7%, RO41%, CNS involvement 34%, pneumonia 7% Mucorales 68%, Rhizopus 24% Stain diagnosis 98%, culture 44%	94%	NA	88%	80% underwent surgery	In hospital mortality 49%	Review of published CAM cases showed that CAM was typically seen in uncontrolled DM patients and severe or critical Covid-19 infection with mostly ROM or ROCM presentation.
Singh A Singh C, Joshi S et al. [98]	Scientific literature search and systematic review of published data till 5/2021	101 cases 82 cases from India	22-86 79% male	Sinus 89%CNS involvement 22%, pulmonary 7.9%	80%, new onset diabetes 2.2%	NA	76%	Not reported	30.7%	trinity of diabetes, rampant use of corticosteroid in a background of COVID-19 appears to increase mucormycosis.
Kasatwar A, Shukla R, Rathod N et al. [36]	Observational study single district in, India of surgically treated patients with ROCM 3/2021-12/2021	100 cases	49.3 years 70% male	Sinus or Rhinoorbital involvement only	95%, 52% with new onset diabetes	NA	76%	95%	2%	Incidence of 7.1/1000 patients noted in hospitalized patients with COVID-19 with steroid use and diabetes being the main risk factors, with increased GRP78 expression being postulated as underlying mechanism for infection
Tavakolpour S, Irani S, Yekaninejad M et al. [35]	Retrospective observational case control study of single center in Iran 2/2020-9/2021	52 CAM 130 controls	57 years 63% male	100% ROCM Rhinoorbital 11.5%, sinus 34.6% and cerebral involvement 54%	55.8, new hyperglycemia 46%	Severe COVID or need for O2 46%	88.5%	Not reported	21.1 % mortality in CAM group	CAM had higher odds of severe COVID (OR 3.6), prior diabetes (OR 7.37), smoking (OR 4.55), or receipt of high dose /pulse steroid therapy
Dave T, Gopinathan Nair A, Hegde R et al. [33]	Retrospective multicentric observational case series of ROCM, India, 5/2021	58	56 years 76% male	100% ROCM 33% CNS extension	88%	All patients had covid 19 infection, severity not reported	83%	100%	34% at mean 5.6 months follow up	Uncontrolled DM, orbital apex involvement, CNS involvement and steroid use associated with unfavorable outcome. CNS involvement associated with risk of mortality
Rao K, Pooja N [99]	Observational study from ENT clinic, single center, India	180	40-60 years 78% male	ROCM 100% 75.6% smear positive	100%	77.8% of patients were post-COVID 19 infections, severity of COVID-19 not reported	Steroid use is not reported	76%	16%	Most common risk factor was diabetes, patients who underwent medical and surgical management had better outcomes
Choksi T, Agrawal A, Date P et al. [34]	Single center case control study, India 3/2021-5/2021	73 CAM patients	53.5 66% male	ROCM 100%	74%	All patients were post-COVID-19 infection, severity not reported of COVID-19 not reported	98%	45%	36% died within 10 days of admission. Cumulative probability of death 53% at day 21	NIV or mechanical ventilation inc risk of mortality (a OR 8.98, 2.1-38.6) and light perception dec risk of mortality (aOR 0.56, 0.32-0.98) at 40 days
Chavan RP, Ingole SM, Nazir et al. [100]	Single center retrospective study, India May 2021-Jul 2021	75 CAM patients	77 % over 40 years age, 75% male	100% ROCM CNS involvement 6.7%	86.6%	Oxygen requirement 55%	44% received steroids	89% were operated.	25.3%	Study the risk factors associated with Mucormycosis infection, age of > 40, COVID-19 infection and DM were the mostcommon RF for Mucormycosis seen in the study
Sen M, HOnavar S, Sengupta S et al. [29]	Retrospective multicenter observational study with data gathered through ophthalmology providers registry, India 1/2020-5/2021	2826	51.9 71% male	ROCM 100% CNS involvement 21%	78%, uncontrolled 41%	O2 requirement 56%	87% 80% of pts with no supplemental O2 requirement and 67% of ambulatory patients received corticosteroids.	88% FESS/PNS debridement in 56%, exenteration 15% and both in 17%	14%	Cam was mainly seen in middle aged and older males, s/s seen 10-15 days following COVID-19 infection. DM, steroids were consistent RF, empiric amphotericin B and surgery improved mortality and outcomes
Patel A, Agarwal R, Rudramurthy S et al. [25]	9/2021 Multicenter observational study of CAM vs non CAM India ( Sept-Dec 2020	187 CAM pts	56.9 y 80% male	63% RO/23% ROC Pulmonary 8.6%, disseminated 2% Diagnosis — smear 83%. Culture 53% CAM vs. 37% non CAM Path 92 %	60% New DM more frequent (21vs 10%)	Hypoxia 40% vs. 12% nCAM, ICU admission 21% vs. 9% nCAM	78% steroid use	70%	6 weeks 37%, 12 weeks 44% Age, CNS (OR 2.3, 1.3–4.4) and pulm involvement ( OR 3.2, 10.5–10.1) and ICU admission inc risk (OR 2.8, 1.4–5.7)	Prevalence 0.27%. Hypoxemia and inappropriate steroid use associated with development of late CAM (>8 days after COVID diagnosis)
Muthu V, Agarwal R, Rudramurthy S et al. [26]	Jan-Jun 2021 Case control study across 25 hospitals in India	1733 cases and 3911 COVID19 controls	52.7y, 70% male	Sinus 88%, orbital 45%, CNS involvement 15%, pulmonary 7% Microscopy 20%, histopath 10%	80.9%, recent onset DM 17.5% DKA 5.2%	46.2% in CAM vs. 54.5% controls, ICU admission 20.5%	74.8%. Nonhypoxemic pts on steroids was higher in CAM (34.2% vs. 22.3%). Higher doses usage more common in CAM vs controls (61% vs. 49%)	83%	28.6% 6 weeks and 32.2% 12 weeks Age CNS involvement (OR 2.3, 1.6-3.2) a/w higher risk. Surgery and combination medical therapy a/w reduced risk	Zinc usage higher in CAM vs. controls (47.9% vs. 41.3%), Rural residence higher in CAM vs. controls (27.5% vs. 11.8%) Hypoxia, ICU admission and mech ventilation lower in CAM vs. nonCAM Cumulative steroid dose (OR 1.006), Zn supp (OR 2.06, 2.24–3.40), host factors (DM, DKA, renal Tx) and rural residence a/w inc risk CAM
Maini R, Saini N, Bhavana K et al. [28]	Multicenter observational study India, Nepal, Bangladesh, Thailand, and USA 5/2021–9/2021	728 cases	51, 76% male	Sinus 94%, rhinoorbital 57$, ROC 18%, pulmonary 2.5%, GI 1.4%	85% New diagnosis 14%, uncontrolled 48%	Oxygen requirement 58%	76%	65.5% underwent surgery	Death 11.7%, vision loss 34%	Analysis showed confirmation of many characteristics of CAM primarily rhinocerebral with high morbidity and mortality increased representation of patients with diabetes and male sex; a high incidence of hypertension (17.2%), and high steroid use that likely increased the risk for fungal infections.
Anand T, Mukherjee A, Satija A et al. [24]	Jun 2021–Sep 2021, multicenter case control study, India	336 CAM patients and 875 controls	Most pts 45–59, 69% male	Not reported	87.6%	O2 requirement 60%. Higher proportion of CAM needed oxygen	Not reported	Not reported	No report of mortality	Odds of CAM increased with work in dusty environment (aOR 3.24, 1.34–7.82), DM (aOR 31.83–13.9–72.6) and steroid use (aOR 2.71, 1.3–5.4)
Hoenigl M, Seidel D, Carvalho A et al. [30]	Systematic review of published cases of CAM and unpublished data through outreach thorough Fungiscope registry Oct 2019–Apr 2021	80 cases including 29 unpublished cases from 18 countries, 42 cases from India (10 unpublished), USA [8], Pakistan [5], France [4], Iran [4], Mexico [4]	55y, 78% male	ROC 73%, pulmonary 25% Rhizopus sp20%, R arrhizus 14%, Mucor sp. 8%, unknown 49% 98% of pts from India had ROC disease	83%, 69% uncontrolled Diabetes a more predominant risk factor for patients from India than other countries. (95% vs 68%) and 93% of pts with ROC disease had DM	45% severe or critical disease; 80% for pulmonary disease; ICU admission 48%	63%	72% for ROC disease	49%; 37% for ROC disease and 80% for pulmonary disease; Loss of vision 32% OR for mortality 2.44 for CNS involvement vs. ROC without CNS involvement (0.79–7.5); CNS or pulmonary involvement HR 4.19 (1.54–11.48)	Uncontrolled DM (83%) and steroid use (79%) were present in most patients with CAM, rhino-orbital cerebral mucormycosis was the most frequent disease sub type
Patel A, Bakshi H, Shah K et al. [27]	Single center case control study in western India 1/2021–5/2021	64 COVID cases and 205 controls	55.7y, 78% male	Sinus 54.7%, ROC 7.8%, pulmonary 7.8% Microscopy 84.4%, path 92.2%, culture 62.5% Rhizopus arrhizus 57.5%, Rhizopus species 10%, Mucorales 22.5%	DM 75%, new diagnosis of diabetes 34.4%	O2 requirement 53.1%	90.6%	81.3%	4.7% at 12 weeks (high rate of LTFU at 12 weeks- 26%)	Diabetes and new onset diabetes (OR 2.9, 1.4–6.1 and OR 48 for new onset), home isolation (OR 4.8, 2–11.3) and steroid therapy (OR 3.6, 1.2–10.9) higher risk of CAM

The severity of COVID-19 illness in reported cases is variable, though pandemic surges may have limited access to hospital care. In the global review of cases by Hoenigl et al., 45% of CAM patients had severe or critical disease and rate of ICU admission was 48% [30]. This is higher for pulmonary CAM which may be associated with more severe disease. It must be noted that oxygen and critical care access was severely impaired during the COVID-19 surge in India when health resources were severely strained [37]. What does seem a common thread in case series and case control studies is the association of steroid use. Steroids were associated with 55–90% of cases of CAM. Inappropriate use of steroids in patients not on supplemental oxygen (80%), as well as ambulatory patients (67%) was documented in one study [29]. Inappropriate steroid use (higher dose than recommended by the guidelines, or prolonged use>10 days) was associated with development of late CAM (defined as onset of CAM >8 days after COVID-19 diagnosis) in one observational study [25].

The prognosis of mucormycosis is largely determined by agent and host factors but also by timely diagnosis and aggressive management. Independent risk factors for mortality include disseminated disease, pulmonary disease (as compared to rhinoorbital disease), renal disease, and infection with Cunninghamella [6]. Rhinoorbitocerebral disease mortality is higher with delayed diagnosis, CNS involvement, cavernous sinus, or bilateral involvement in addition to host factors [6, 38]. The case fatality for CAM for cases reported from India seems to be lower (36.5%) as compared to cases reported from elsewhere (61.9%) possibly reflecting predominance of rhinoorbitocerebral disease which may have a lower mortality than pulmonary or disseminated disease [31]. In CAM, the need for noninvasive ventilatory care and poor visual acuity at presentation seem to predict mortality and may reflect the severity of disease [34]. In a retrospective multicenter observational case series of rhinoorbital mucormycosis from India, uncontrolled DM, orbital apex involvement, CNS involvement and steroid use were associated with unfavorable outcome which included vision loss at a mean follow up of 5.6 months. CNS involvement was associated with increased risk of mortality [33].

Agent Factors

Mucormycosis is caused by a group of filamentous molds belonging to the order Mucorales which comprises 38 genera associated with human infection [39]. They are a group of fungi with aseptate or sparsely septate broad branching (90°) hyphae and are thermotolerant, ubiquitously found in soil and decaying organic matter. They can reproduce asexually by way of sporangiospores or sexually by formation of zygospores. The nature of spore bearing structure, location of rhizoids and type of zygospores produced and their physiologic characteristics are used to differentiate the species. In a systematic review from 2018, Rhizopus spp., Lichtheimia spp. and Mucor spp. accounted for 75% of all cases [40]. Rhizopus species seems to be the most common, responsible for nearly 60% of mucormycosis cases in humans and accounts for 90% of the rhinoorbitocerebral (ROCM) form [9]. The causative organisms in clinical case series vary by location with infections due to Lichtheimia reported more commonly from Europe, whereas Apophysomyces variabilis was reported more commonly in India in the pre-COVID-19 era [17]. In CAM, Rhizopus species (Rhizopus arrhizus, Rhizopus microporus, and Rhizopus homothallicus) have been reported followed by Apophysomyces variabilis, Mucor species and Lichtheimia corymbifera [5, 27]

Pathophysiology

Patients most commonly get infected via inhalation or ingestion of spores. Direct inoculation can cause cutaneous or contiguous tissue infection. Mucormycosis is associated with angioinvasive infection with tissue destruction. Risk factors for invasive disease include long-term steroid use, prolonged neutropenia in the setting of hematologic malignancies or transplantation. Other risk factors include uncontrolled diabetes, diabetic ketoacidosis, iron overload, and chronic kidney disease. Cutaneous disease occurs in context of injection drug use, wounds sustained in blast injuries or natural disasters, burns, and contaminated surgical wounds [41].

The invasion of host cells is elicited by the spore coat protein homologue CotH3 which is expressed by the spores. This interacts with host receptor glucose regulator protein 78 (GRP 78) on nasal epithelial cells to promote invasion into blood vessels and damage. GRP78 is induced by host factors such as high glucose concentration, high iron concentration, and ketone bodies as seen in patients with diabetic ketoacidosis [7, 42] Serum GRP78 levels have been shown to be increased during COVID-19 pneumonia as compared to healthy controls [43]. Experimental studies have shown that GRP78 can form a complex with the spike protein and ACE2 to act as a host auxiliary factor for SARS-CoV2 entry and infection [44]. Additionally, Mucorales produce the toxin mucoricin, which displays structural and functional features of the plant toxin ricin that can induce inflammation, increased vascular permeability and tissue damage [45]. COVID-19 causes both alveolar epithelial damage and a vascular endotheliaitis [46]. This may lead to exposure of extracellular matrix proteins. Rhizopus spores can adhere to laminin and type IV collagen [47].

On a cellular level, COVID-19 and its management might augment development of CAM. Notably the ability to restrict intracellular iron availability and inhibit fungal persistence and growth is an essential mechanism required to counter infection with Mucorales. Patients with COVID 19 may have an augmented inflammatory state with high levels of cytokines including IL-6 which can cause a hyperferritinic state and stimulate hepcidin synthesis [41]. This could lead to sequestration of intracellular iron in macrophages and possibly increase the risk for invasive mucormycosis.

Diagnostic Evaluation

Diagnosis remains challenging. A high degree of clinical suspicion needs to be present. Findings of nasal or facial pain, swelling, nasal congestion, discharge, dark discoloration of mucosa, or eyelid swelling may guide tissue sampling for sinus or rhinoorbital disease [29]. For a rapid presumptive diagnosis of mucormycosis direct microscopy of KOH wet mounts is an inexpensive yet invaluable method to rapidly give a presumptive diagnosis and to define clear surgical margins for invasive fungal infection intraoperatively, and is strongly recommended, along with histopathology which offers a very important diagnostic tool for demonstration of fungal hyphae typical for mucoromycetes in biopsies of affected tissues [48]. Culture of specimens is important in the diagnosis as it allows identification to the genus and species level, and eventually antifungal susceptibility testing, A positive culture from a sterile site confirms the diagnosis, while a positive culture from a non-sterile site could be due to a contaminant and must be combined with clinical and radiological data to establish a probable diagnosis. The major concern about culture, however, is its low sensitivity, as it can be falsely negative in up to 50% of cases [48, 49]

CAM diagnosis in case series have been by direct microscopy of tissue in majority of reports (>80%) with aseptate hyphae noted in tissue samples [5, 27, 29, 30]. This is similar to non-CAM cases. Tissue biopsies sent for histopathologic identification showed aseptate hyphae in 76–92%. Cultures were positive in 53% as compared to 39% non-CAM in one study [25].

Diagnosis of COVID-associated pulmonary mucormycosis (CAPM) is more challenging and may need more invasive sampling including bronchioalveolar lavage or lung biopsies. This may be challenging in the context of active COVID-19 pneumonia and disruption of clinical and diagnostic services in the context of pandemic surges. This is highlighted in the lower proportion of diagnosis of CAPM during COVID as compared to a pre COVID period in 2 large multicenter studies from India [5, 25].

Molecular testing, with polymerase chain reaction (PCR) using primers targeting the internally transcribed spacer (ITS) or 28S rDNA region performed on blood or bronchioalveolar lavage may be a useful strategy for pulmonary samples and may provide higher sensitivity, specificity, and early diagnosis [48, 50]

Host Factors

High rate of mucormycosis and higher burden of Mucorales in the hospital and outdoor environment had been reported in India [51] during and even before the CAM epidemic and no difference in Mucorales species causing mucormycosis before and during the COVID 19 pandemic hence hypothesizing that COVID 19, Its specific treatment and specific host factors is what may have contributed to the CAM outbreak.

Diabetes Mellitus

The elevated risk of mucormycosis in diabetes mellitus is described prior to COVID-19 [6, 42]. A large majority of cases from India in the pre-COVID-19 period are reported in patients with uncontrolled diabetes (54–76%) [8, 18, 52]. Cases from north of India seem to have a higher prevalence of uncontrolled diabetes with ketosis at diagnosis (90%) as compared to the south (10%) [15].

India and China have the largest population of patients with diabetes mellitus [53]. Prevalence of diabetes may be increasing in countries like India with improved access to care, but more than half of the adults with diabetes remain undiagnosed and more than 75% of diagnosed diabetics have poor glycemic control [54].

COVID-19 has been associated with more severe disease and death with preexisting diabetes. COVID-19 is also associated with dysglycemia, new onset diabetes and severe metabolic complications of diabetes including diabetic ketoacidosis. This is thought to be due to inflammation triggered impaired glucose utilization and/or impaired insulin secretion [55]. Preexisting diabetes is one of the most common risk factors reported with CAM, in India but also seen in other countries like Iran and Mexico with a prevalence of about 74.5% whereas in western countries hematologic malignancies and organ transplant seem to be predominant risk factors [30, 31].

Hyperglycemia and ketosis can affect iron binding by glycosylation of iron binding proteins thus elevating free iron [7] which can be utilized by Mucorales. Acidosis caused by ketosis can increase expression of host receptor protein GRP78 for fungal ligand binding. In addition, acidosis may impair neutrophil macrophage function [41]. This facilitates angioinvasive disease in mucormycosis. Elevated glucose, iron, and ketone bodies in diabetic ketosis also suppress T lymphocyte activity and IFN-Υ [42] thus allowing the fungus to thrive.

Steroid Therapy

The RECOVERY trial showed that dexamethasone significantly reduced mortality in hospitalized patients with COVID-19 who were receiving supplemental oxygen [56]. However, steroids are a double-edged sword. While steroids may impair inflammatory response, high dose or prolonged use may also impair neutrophil and monocyte/macrophage activity by impairing margination, chemotaxis, phagocytosis, oxidative, and non-oxidative functions. These cells constitute the main defense against germinating spores [57].

Steroids also cause lymphopenia and a TH2 predominant response which may contribute to decreased phagocyte effector function. The worsening of glucose metabolism by steroids is well known and steroids may worsen glycemic control in known diabetics or induce hyperglycemia in predisposed patients [41].

A case control study across 25 hospitals in India, the primary exposure variable was the type of treatment offered for COVID 19 with findings of more frequent use of steroid in CAM case patients (71.6%) as well as higher proportion of CAM patients receiving inappropriate use of steroids (in absence of hypoxemia). In addition, patients receiving more than the recommended dose (> 60mg of dexamethasone or equivalent) was also higher among CAM patients than COVID 19 control patients both for hypoxemic (61.3% vs. 49.2% p = 0.0001) and non-hypoxemic (37.7% vs22.6%; p = 0.0001) groups [26].

Kasatwar et al. identified different phenotypes of CAM- people with no prior diagnosis of mucormycosis received aggressive steroid therapy for mild COVID-19 and had CAM. They seemed to have lower levels of hyperglycemia and less severe CAM disease with good prognosis. In contrast, the other group was of people with long standing diabetes with severe COVID who presented with more extensive disease needing more extensive surgery and with poorer outcomes [36]. They also described a series of patients with no steroid use or diabetes with a predilection of jaw involvement.

Antibiotic Therapy

Antibiotic treatment, along with the severe injuries of the epithelium may create a dysbiosis which then provide a permissive environment for developing invasive fungal infections, including mucormycosis [58]. The oral and nasal microbiome is perturbed during COVID-19 infection with abundance of opportunistic pathogens and impaired metabolic pathways [59, 60]. There was widespread antibiotic overuse during COVID-19, especially in India [61]. While high baseline use of antibiotics is common in most series where such data is available [62–64], it may be difficult to draw conclusions about the role of antibiotic therapy in emergence of CAM

Tocilizumab or Other Immunosuppression

Immunosuppressive therapy like baricitinib, tocilizumab in addition to corticosteroids used as part of COVID 19 treatment to reduce symptoms of host immune response and cytokine storm may putatively elevate risk of CAM. However, this may also be a marker of illness severity and the elevated risk of CAM by tocilizumab has not been proven across multiple studies [26–28, 65]. In a large randomized controlled trial in patients who were critically ill with COVID-19, treatment with IL-6 receptor antagonists improved outcomes, including survival; however, it is not known whether this therapeutic strategy of dampening an important cytokine in antifungal immunity could predispose patients to CAM.

Zinc

Zinc may play a role in fungal pathogenesis by binding with virulence factors or by transcriptional regulation of proteins required for infection in yeasts [66]. Fungi have been known to be susceptible to zinc depletion, and zinc chelation along with Posaconazole has been shown to have synergy against some strains of Rhizopus in in vitro studies [67]^. In one multicenter case control study from India, zinc supplementation was associated with an increased risk of CAM [26]. Another case control study however showed zinc to be protective against CAM [68]. While some isolates of Rhizopus arrhizus grew better with zinc enrichment in vitro and had increased fungal biomass in one study, mean serum zinc levels were below reference range and not significantly different between cases of CAM or similar controls without CAM [69]. There was widespread use of zinc supplements to treat COVID-19, given the prevalence of low zinc levels in the background population may be high, and the possible effects of zinc in modulating immune function [70]. However, the effect of zinc on COVID-19 itself is probably limited [71], and its role in pathogenesis of CAM is conjectural at best given the conflicting data.

Environmental Epidemiology

The high baseline prevalence of mucormycosis in India, and differences in species isolated raises the question of the role of differential environmental exposures in reported cases of CAM. The spectrum of species causing mucormycosis in India is wide and Rhizopus microporus and Rhizopus homothallicus as well as Apophysomyces [72–74] are represented in higher proportions. The low nitrogen content of the alkaline soil in India is associated with increased isolation of A. variabilis [73]. In a study from South India, 29% of cases of cutaneous mucormycosis due to Apophysomyces in a hospital were thought to be nosocomial [75]. Majority of nosocomial infections are cutaneous though attribution of nosocomial origin in ROCM may be difficult. In outbreaks of mucormycosis, contaminated medical equipment was thought to be responsible in 41% and contaminated air in 31%, followed by traumatic inoculation, contact or ingestion [76]. Contaminated indoor air and outdoor pollution with dust and particulate matter is a known issue in India and hospital dust has been shown to harbor dematiaceous fungi from environmental sampling data of soil from India [77]. Ecological studies have isolated pathogenic Mucorales from the soil and spores from indoor environments, such as air conditioning equipment and hospital air [51]. In a multicenter case control study of CAM and COVID-19 positive controls from India, working in a dusty environment (farming, gardening, or construction work), as well as longer duration of hospital stay seemed to be associated with elevated risk of CAM [24]. Farming or gardening work was associated with 3.2-fold higher odds of CAM in this study. In another study, residence, or work in an environment where dust exposure was present, as well as residence in a rural area was associated with increased odds of CAM diagnosis [68]. The use of cow-dung to treat COVID-19 or the burning of cow-dung cakes causing spore dispersal and increased risk of CAM was postulated as a risk. However, there was no change in growth of Mucorales in air samples collected before and after cattle dung burning in an aero-mycological study done in India during the COVID-19 period [78].

The hospital environment can also be a potential source especially if dust control or air filtration is suboptimal. In a survey of hospital operating rooms in India over 2 years, air filters of air conditioning units were colonized with fungi in 26%. This was especially high in window mounted units which draw air from the environment [79]. Four cases of mycotic prosthetic valve endocarditis thought to be from contamination of air conditioning equipment in the operating room of a single hospital in India have been reported [80]. Construction or renovation may also disperse fungal spores.

During the surge of cases of COVID-19 in India in mid-2021, there was a shortage of medical oxygen [81], and many improvised oxygen delivery systems were adapted to provide respiratory support for the acutely ill. This included oxygen concentrators which would draw ambient air, compress it, and extract oxygen prior to delivery to patient. It could be hypothesized that if the concentrator or oxygen delivery equipment was contaminated with dust or drawing air from a dusty environment then the risk would be elevated. In the largest case series of CAM, around half of all patients needed oxygen for treatment of COVID-19 [25, 26, 29]; however oxygen requirement was lower in CAM vs. controls in one large multicenter study [26]. The risk of CAM from fluids used for nebulization or humidification is also thought to be low and survival of Mucorales in high oxygen environments may not be optimal.

Cases have been reported in negative pressure rooms which could potentially draw in dust or spores from outside when the opening of doors to such rooms allows airflow disturbances [82, 83]. Strict dust control measures and air handling/filtration and/or maintenance of oxygen systems may not have been feasible in the pandemic surge where health care facilities were overwhelmed due to limited surge capacity.

Pandemic Specific Risk Factors

During COVID-19, many countries including India reported remarkably high incidences of inappropriate medical therapy, including inappropriate steroid use, antibiotic therapy, and home remedies [84, 85]. Treatment protocols varied between states, recommendations often included therapies with low certainty of evidence. This may have led to delay in seeking appropriate medical care. State imposed curfews and lockdowns cause disruptions in non-emergent healthcare access especially among people from low socio-economic status [86]. This may lead to delayed diagnosis and treatment. Early diagnosis and multimodal treatment [29, 34, 87] are important factors that determine outcome. Bhansali et al. noted poor outcome with an increase in mortality by up to 30% with delay in diagnosis and treatment beyond 10 days [88]. Widespread hospital bed shortages, drug shortages [89], and disruptions of support systems may all be unique pandemic specific issues which may worsen the impact of the condition.

Conclusion

The alarming surge in cases of CAM highlighted many areas where current gaps exist in knowledge, evidence, or management of this infection. These include risk mechanisms of secondary infection in COVID-19; factors determining baseline differences in incidence of mucormycosis and factors driving the dramatic rise during the COVID-19 period. It also exposed us to unique pandemic specific factors which may impact diagnosis, access to care, and management of what is a life-threatening infection. Management of COVID-19 should also involve avoiding inappropriate steroid therapy and strict attention to glycemic control. Morbidity and mortality can be reduced by ensuring access to appropriate diagnostic testing and access to specialty care, including surgical intervention when appropriate.

Declarations

Human and Animal Rights and Informed Consent

This article does not contain any studies with human or animal subjects performed by any of the authors.

Conflict of Interest

Dr. Narayanan reports grants from Regeneron Pharmaceuticals, outside the submitted work.

Dr Tayabali and Pothiwalla declare that they have no conflicts of interest.