Risk factors for palatal and orbital involvement in mucormycosis epidemic–Report of a center in India

Abstract

Background

Mucormycosis of the nose and paranasal sinuses has emerged as an epidemic following COVID-19 pandemic. The management involves surgical debridement, the extent of which depends on the bulk of the disease. Extension to the orbit and palate depends on the involvement of specific sites in the nose and paranasal sinuses. This study intended to identify those sites.

Materials and methods

This was a single-center case-control study. There were 3 groups according to the region involved (Palate, orbit and both). The fourth group with neither involvement was the control. Scoring system was used to estimate the bulk of disease according to the site involved in MRI scan. Odds ratio and chi-square tests were used to study risk and association respectively. ROC curve was obtained for the MRI scores of the cases and controls.

Results

214 patients were studied in all; 44.39% and 61.68% had palate and orbit involvement, respectively. Maxillary sinus roof had significant association and the highest risk for spread of disease into the palate. The risk for the orbit to be involved was increased when there was disease in the cavernous and sphenoid sinus. For the orbit, the site with the significant association was the ethmoid sinus, infratemporal fossa and roof of the maxillary sinus.

Conclusion

Due attention should be given for clearance during surgical debridement to the high-risk sites to prevent recurrences and reduce morbidity and mortality. MRI scores were most helpful to identify the bulk of disease when both palate and orbit were involved.

1. Introduction

Rhino-orbital Mucormycosis cases have emerged as an epidemic within the pandemic of COVID-19 in India and a few other nations. The majority of the patients have involvement of palate, orbit or both adding to the bulk of the disease thereby increasing the morbidity and mortality. The surgeries are extensive and mutilating when they are involved; hence there is a need to determine what leads to palatal or orbital extensions from the nose and paranasal sinuses so that these critical structures might be addressed better during debridement. It is difficult to predict the progressive course of the disease and the extent of involvement of the vital structures in the vicinity of the primary disease that might finally decide the fate of the patient. This spread directly affects the debridement clearance required in the patients.

This study intended to identify the involvement of critical structures that may lead to the extension of disease into the palate/upper alveolus and orbit from the nose and paranasal sinuses and also the patterns of disease presentations of Mucormycosis [MCM].

2. Materials and methods

This was a single center study undertaken for 3 months, in a tertiary referral public hospital during the peak of the epidemic of MCM in South India that commenced during the second wave of the COVID-19 pandemic. Clearance was obtained from the institutional ethics review board. Written informed consent was taken from all the patients. The study included patients of diabetes mellitus and COVID 19 history with and without orbital and or palatal involvement due to MCM of the nose and paranasal sinuses who had undergone surgical debridement for the same and also received Amphotericin B therapy.

The primary outcome was to identify the critical structures responsible for the involvement of the palate/alveolus, orbit or both [Regions]. The patterns of involvement of the various sites in the nose and the paranasal sinus [PNS] were also studied. This was a case-control study with 4 groups:

• a.Case group A[A]: Nose, PNS and palate involved cases [stage 2]
• b.Case group B [B]: Nose, PNS and orbit involved cases [stage 3]
• c.Case group C [C]: Nose, PNS, palate and orbit involved cases [Stage 3] with CNS involvement [Stage 4]
• d.Control group [D]: Nose, PNS or both involved cases [Stage 1]

Every consecutive eligible patient was included in the study. Adult patients in stages 1, 2, 3,4 of MCM of head and neck with diabetes mellitus and post COVID-19 infection (Based on magnetic resonance imaging [MRI], Diagnostic nasal endoscopy [DNE], biopsy and clinical findings) [1], [2]. Orbital involvement was confirmed by restriction of movement, loss or reduction of vision, chemosis, edema of eyelid, proptosis and redness of the eye. Palatal involvement was established by the presence of ulcer, swelling or loose tooth. Diabetes mellitus included precovid type 1, 2 or covid induced diabetes. Inoperable cases, chronic sinusitis without evidence of MCM or pre-existing diabetic retinopathy were excluded. A Scoring system to assess the bulk of disease for every patient was prepared according to the sites involved on the MRI and the net score was calculated [3]. The scoring pattern for the sites was as follows: middle/inferior turbinate, osteomeatal complex, nasal septum, cribriform plate, lateral wall of the maxillary cavity, infratemporal fossa, frontal sinus [Score 1], ethmoids, the floor of the nasal cavity, floor of the maxillary sinus, sphenoid sinus, roof of the maxillary sinus [score 2], pterygopalatine fossa, cavernous sinus [score 3]. The score of cases and controls was compared and analyzed.

2 ∗ 2 contingency tables were prepared separately for every suspected risk factor and the Odds ratio was calculated. Chi-square test was used for the study of the association at 95% confidence intervals. The sensitivity and specificity of the MRI scores in identifying the disease of group A, B, C was calculated. Standard error was calculated [95% confidence intervals] and the ROC curve was obtained by plotting sensitivity against 1-specificity. Stata version 21 was used for statistical analysis.

3. Results

Two hundred and fourteen patients were studied in all that included 170 males and 44 females. The youngest patient was aged 25 years and the oldest 80 years. The mean age was 50.29 ± 11.47. The distribution of the patients according to the age class interval is shown in Fig. 1 . The percentages of the patients in the individual groups and the admission discharge statistics are given in Table 1 . It was found that 44.39% of the patients had palatal involvement and 61.68% had orbital involvement. The most typical symptom and sign were craniofacial pain [Table 2 ] and nasal discharge [Table 3 ], respectively. The findings of the MRI done are given in Table 4 and the commonest site involved was the ethmoid sinus [Table 4]. The risk factor for each site was calculated by odds ratio and then the strength of the association with the palate and orbit [Table 5 ]. The MRI scores of the patients in the various groups is given in class intervals in Table 6 . The area under the ROC curves of groups A, B, C is shown in Fig. 2, Fig. 3, Fig. 4 , respectively. It was found that group C had the maximum area under ROC and hence the MRI scores were the most useful to identify the bulk of disease.

Fig. 1.

Bar chart showing age class interval distribution in the four groups.

Table 1.

Admission discharge statistics of the patients in the 4 groups.

Status of patient	A	B	C	D	Total
	Palate	Orbit	Both	None
Undergoing treatment	28	51	36	25	140 [65.42%]
Discharged from hospital on request	7	7	8	8	30
Discharge against medical advice	1	6	5	4	16
Expired	2 (5.26%)	11 (14.66%)	8 (14.03%)	7 (15.90%)	28 [13.08%]
Total	38 [17.75%]	75 [35.04%]	57 [26.63%]	44 [20.56%]	214

Table 2.

Symptoms in the 4 groups.

Symptoms	A	B	C	D	Total
	Palate	Orbit	Both	None
Nasal obstruction	15	23	20	21	79 (36.91%)
Nasal discharge	13	14	12	11	50 (23.36%)
Craniofacial pain	27	54	36	27	144 (67.28%)
Blurred vision	0	22	11	6	39 (18.22%)
Eye swelling	7	55	41	5	108 (50.46%)
Fever	2	9	4	3	18 (8.41%)
Loose tooth/toothache	23	16	23	7	69 (32.24%)
Disorientation	0	1	2	0	3 (1.40%)
Vomiting	0	1	0	0	1 (0.46%)

Table 3.

Signs in the 4 groups.

Signs	A	B	C	D	Total
	Palate	Orbit	Both	None
Nasal discharge	21	52	33	28	134 (62.61%)
Nasal eschar/slough	15	30	36	22	103 (48.13%)
Eyelid edema	3	57	37	4	101 (47.19%)
Eye congestion	1	27	23	0	51 (23.83%)
Diminished vision	0	26	15	0	41 (19.15%)
Restricted movements of eyeball	0	29	21	0	50 (23.36%)
Proptosis	0	21	15	0	36 (16.82%)
Palatal ulcer	30	0	41	0	71 (33.17%)
Loose or missing teeth	9	2	3	0	14 (6.54%)
Altered sensorium	0	0	6	0	6 (2.80%)

Table 4.

MRI findings suggesting the involvement of various sites in the craniofacial region.

MRI findings of involvement [sites]	A	B	C	D	Total
	Palate	Orbit	Both	None
Middle turbinate	6	20	12	14	52 (24.29%)
Inferior turbinate	7	17	15	10	42 (19.62%)
Septum	2	0	4	0	6 (2.80%)
Lateral wall of maxillary sinus	18	40	26	22	106 (49.53%)
Roof of maxillary sinus	31	48	58	21	158 (73.83%)
Floor of maxillary sinus	33	48	32	25	138 (64.48%)
Floor of nasal cavity	3	5	3	1	12 (5.6%)
Ethmoid sinus	28	66	48	34	176 (82.24%)
Sphenoid sinus	22	56	42	21	141 (65.88%)
Frontal sinus	16	40	31	21	108 (50.46%)
Pterygopalatine fossa	4	7	8	3	22 (10.28%)
Infratemporal fossa	8	24	12	5	49 (22.89%)
Cavernous sinus	0	3	4	0	7 (3.27%)

Table 5.

The odds ratio and the strength of association found between the sites of the craniofacial region and the palate, orbit.

Site involved	Palate	Palate	Orbit	Orbit
	Odds ratio	p-Value	Odds ratio	p-Value
Lateral wall of nasal cavity	0.39	0.001	1.63	0.09
Floor of the maxillary sinus	1.3	0.28	0.63	0.13
Roof of the maxillary sinus	10.74	<0.001	2.35	0.007
Lateral wall of maxillary sinus	0.79	0.41	1.05	0.86
Floor of nasal cavity/septum	2.72	0.05	1.26	0.67
Ethmoid sinus	0.76	0.45	2.04	0.05
Sphenoid sinus	1.12	0.69	2.61	0.001
Frontal sinus	0.93	0.80	1.41	0.22
Pterygopalatine fossa	1.57	0.32	1.37	0.50
Infratemporal fossa	0.82	0.57	1.99	0.05
Cavernous sinus	1.7	0.52	9.86	0.02

Table 6.

MRI scores of the patients in the 4 groups in class intervals.

MRI score	A	B	C	D	Total
	Palate	Orbit	Both	None
1–3	3	2	0	7	12
4–6	9	15	7	14	45
7–9	14	29	26	19	88
10–12	8	21	18	3	50
13–15	4	4	5	1	14
16–18	0	4	1	0	5
Total cases	38	75	57	44	214

Fig. 2.

Area under the ROC curve in palatal involvement only.

Fig. 3.

Area under the ROC curve in orbital involvement only.

Fig. 4.

Area under the ROC curve in both palatal and orbital involvement.

4. Discussion

In the past, several authors have studied a series of cases of MCM over the years [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14]. Also, many cases of MCM have been reported recently (2021) across India and the rest of the world after the COVID-19 pandemic [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31]. However, none has evaluated the risk factors for palatal and orbital involvement against a control. To the best of our knowledge, this is the first paper to assess and report the same. Just as the course of COVID-19 was unpredictable and could become serious in some, so was MCM in this study. Given the high morbidity and mortality of the disease, it is essential to identify the sites in and around the nose and the paranasal sinuses that get involved early and play a lead role in the further extension of the disease to vital organs. There were no apparent reasons or causes for the spread into the palate, orbit or the central nervous system in a few of them. The involvement of these regions additionally required the services of a maxillofacial surgeon, a prosthodontist, an oculoplastic surgeon and a neurosurgeon in our setup. Hence the number of surgeries that the patient had to undergo would increase, thereby increasing the duration of surgery and also the number of times the patient was subjected to general anesthesia.

Diagnosis is confirmed through biopsy and histological analysis of the affected tissues. On microscopy, broad-based ribbon-like non-septate hyphae with irregular right-angled branching are the key diagnostic microscopic features [32]. CT and MRI scans may suggest invasive MCM but they may often be nondiagnostic [11]. If angio-invasion is present, bone erosion may be absent even in the presence of progressive disease. Overall, CT and MRI may be most helpful in assisting surgical planning for, rather than the diagnosis of rhinocerebral MCM [11]. Periorbital cellulitis, extraocular muscle paresis, proptosis and chemosis frequently develop due to disease extension into orbit or cavernous sinus. Spread from the sphenoid sinuses to the adjacent cavernous sinus can result in cranial nerve palsies, thrombosis of the sinus and involvement of the carotid artery [33]. Multicompartmental and extra-sinonasal tissue infarction is possible without overt bone involvement and is caused by the propensity of fungal elements to disseminate from nasal cavity via the perineural and transvascular routes [10], [34], [35]. Angioinvasion with infarction and hematogenous spread are hallmarks of MCM [36], [37]. The fungal hyphae grow along the intimal elastic lamina leading to intimal hyperplasia and intravascular thrombosis, further leading to vessel occlusion [38], [39]. Most regions that proved to be a significant risk factor for either the palate or the orbit were the closely adjoining structures. However, paradoxically the roof of the maxillary sinus has been a high-risk factor for palatal involvement even though it was not the immediate adjacent structure. Besides, in many cases, it was observed during surgical debridement that there were hidden deep pockets of disease with apparently healthy bleeding tissues around them. This could have probably been possible by perineural spread via the infraorbital nerve [12]. A few authors have found the maxillary sinus to be the commonest sinus involved [2], [17], [29], [30]. However, Sharma S et al. found the maxillary sinus involvement only in 52.17% and ethmoid sinuses in 100% of the cases [18]. Likewise, the ethmoid was the commonest sinus involved and the palate was the commonest region in a study by Pippal SK et al. [22] and Kulkarni NS et al. [13]. Hosseini SMS et al. found the involvement of the pterygopalatine fossa in all 10 cases and concluded that it was the main reservoir acting as a conduit for other sites [12]. Mitra et al. [23] and Sen M et al. [25] also found orbit involvement in a more significant number of cases than palate, similar to our study. They opined that the angioinvasive nature and perineural spread were responsible for early optic nerve involvement and orbital apex syndrome. One of the routes of spread from the nasal cavity to the orbit is the nasolacrimal duct that can be noticed on CT soft tissue window [34]. The mortality in our study did not appear to be proportional to the bulk of the disease again, probably due to non-contiguous spread to the brain and cavernous sinus. COVID-19 related lung complications were yet another cause of mortality.

Extension of the disease into the maxillary and ethmoid sinus can lead to orbital involvement through the superior orbital fissure, ophthalmic vein and cribriform plate along the perivascular channels can occur intracranial spread [40]. As observed in Table 4, most of the spread to the orbit and palate depended on what got involved initially in the 3 main sites – ethmoid, the roof of the maxillary sinus and sphenoid sinus. Further, the involvement of other sites only added to the bulk of the disease and MRI score but probably did not play a role in the spread to the orbit and palate.

Even though the epidemic has slowed down considerably now, we still count. The present study could not continue until the end and depicts only a cross-section of the epidemic.

5. Conclusion

The disease in the roof of the maxillary sinus had a significant association and the highest risk to spread into the upper alveolus and the palate. The other sites with increased risk for palatal involvement were the floor of the nasal cavity and the pterygopalatine fossa. Similarly, the risk of involvement in orbit was highest when there was disease in the sphenoid sinus and the cavernous sinus. The other sites with significant association and high risk were the roof of the maxillary sinus, infratemporal fossa and ethmoid sinus. Due attention should be given to the clearance of disease in these high-risk sites during debridement surgery to prevent common recurrences and thus reduce morbidity and mortality. The contiguous spread was not always the rule and skip lesions were possible due to perineural and transvascular spread. MRI scores were most helpful to identify the bulk of disease when both the palate and orbit were involved and least beneficial when the palate alone was involved.