Introduction

Acute invasive fungal rhinosinusitis (AIFRS) is a potentially lethal disease process whose timely diagnosis and management is crucial for the survival of the patient. This entity has many fungi as its causative factor. The commonly isolated fungi are [1, 3] Zygomycota—Mucor, Rhizopus, Rhizomucor, Absidia and other Mucorales, and Ascomycota-aspergillus species.

Patients with haematopoietic malignancies, those receiving bone marrow transplant, status post organ transplant, uncontrolled diabetes mellitus [4] or those on long term steroids are generally prone for this infection. Occasionally immunocompetent individuals are also affected.

Presentation is quite variable with symptoms like nasal obstruction, headache not responding to analgesics, nasal discharge, facial swelling, disturbances in vision, cranial nerve palsies and sepsis (Fig. 1). Mucormycosis [6] can involve rhino-cerebral, pulmonary, gastrointestinal, cutaneous, central nervous system and other miscellaneous sites like myocardium, bones, and kidney. However co-infection of fungi confuses the diagnosis.

Fig. 1
figure 1

Clinical presentation. a Uncontrolled T2DM patient with left facial palsy and nasal crusting. b Same patient in a with palatal ulcer. c CKD patient with orbital cellulitis. d Diabetic patient with non-healing palatal ulcer

Full size image

These fungi are highly angioinvasive and causes tissue infarction, necrosis and thrombosis. This accounts for the blackish eschar and its florid spread. Neutrophils are the main line of defence against the fungi and the patients with neutropenia are particularly affected. Sakeena et al. [2] has proposed guiding treatment on absolute neutrophil counts; but none of our patients, neutropenia was detected at any stage. Instead uncontrolled diabetes mellitus or unmonitored steroid usage was the prime underlying factor. Adequate samples [7] for KOH mounts, fungal culture, biopsy and frozen section of the suspicious tissue guided our treatment. We advocate a combined approach of aggressive antifungal therapy and surgical debridement followed by antifungal nasal douche and serial endoscopic examinations. The treating team involves many physicians like otorhinolaryngologist, nephrologist and ophthalmologist. Considering the success of treatment in our series of patients, we propose a protocol for the treatment of AIFRS.

Methods

In the time frame of September 2015–September 2017, we had a total of 14 patients of AIFRS. Of the total, 8 patients were treated as per proposed protocol (Table 1). Amongst the remaining 6 patients, 4 were excluded as they were lost to follow-up, 1 patient expired before the initiation of treatment and the other patient had a very atypical presentation. The unusual presentation being an isolated non healing zygomatic osteomyelitis which was later histopathologically diagnosed as mucormycosis and amphotericin B was initiated for the same.

Table 1 Patient details and presentation
Full size table

The general mode of treatment was guided by KOH mount, fungal culture and histopathology (Table 2). Mucor is identified by the aseptae, ribbon like hyphae (10–20 µm) and Aspergillous by septate branching hyphae (3–6 µm) (Fig. 2).

Table 2 Diagnostic aids
Full size table
Fig. 2
figure 2

Fungal hyphae noted with inflammatory changes

Full size image

The preoperative imaging studies of the Brain, paranasal sinus and orbit was undertaken depending on the involvement. Medical therapy with liposomal amphotericin B (Lip AmB) was started pre-operatively and all patients underwent endoscopic debridement to decrease the fungal load. With involvement of other sites like palate or zygoma, approaches were selected to debride the respective areas. In all the debridement including hard palate, the unhealthy nonviable tissue was debrided till healthy bleed was noted. Especially in cases of mucor the vascular territory affected was characteristically demarcated by the pale diseased tissue. The intraoral hard palate curettage done was left to heal by secondary intention and postoperative obturators was used in select cases. A minimum of two sittings of thorough debridement were done (Table 3). In case of contradictory results of KOH mount, fungal culture, pathology reports or insufficient response to Lip AmB, the case was considered as co-infection of fungi, and azole group of drugs were started. The non-availability of antifungal sensitivity testing limits evidence for the same. Conventional antifungal nasal douches were done in immediate postoperative period.

Table 3 Surgical management
Full size table

The sensitivity and specificity of the various tests were noted. The overall survival of patients at the end of treatment at an average follow up period of 15 months was noted. Treatment directed outcome was calculated on the number of days of inpatient stay–early being ≤ 45 days and late > 45 days. Chi –square tests were used and p < 0.05 was considered significant.

Results

In the time frame of September2015-September 2017, 8 of the 14 patients who presented with AIFRS were treated as per protocol. The mean age of presentation of the patients was 49.8.

Almost 75% of the patients (n = 6) had multiple comorbidities. About 75% (n = 6) presented with diabetes, 25% (n = 2) had chronic kidney disease, 25% (n = 2) was on long term oral steroids, 25% (n = 2) presented with diabetic ketoacidosis. The average delay in presentation to the hospital was 9 days and the average delay in initiation of treatment was 5.25 days. The varying presentation included persistent headache and facial heaviness, periorbital swelling, restricted movement of eye, palatal perforations, facial swelling, etc. (Table 1).

Preoperative diagnostic battery of tests included KOH mount, fungal culture and histopathological examination. KOH acted as screening test for fungal aetiology; however it was the combination of clinical presentation and KOH results which aided decision to start antifungals. KOH tests were positive for 62% patients, culture were positive in 62% cases and pathologic analysis detected fungi in 100% cases. In 62% cases both KOH and HPE detected fungi (Table 2).

Simultaneous initiation of debridement and medical therapy is crucial. Surgical debridement was done via endonasal, tranzygomatic or transpalatal approach (depending on the site of affliction). Endonasal debridement was done for 25% of the patients while 25% required more than two approaches. Atleast 50% of the patients required two sittings of debridement while 37.5% required only one sitting and 12.5% (n = 1) required three sittings (Table 3). All the patients were trained to do nasal douche with antifungals post operatively for a week. The number of patients with an early duration of inpatient stay (≤ 45 days) was 50% while the other 50% had a delayed stay (> 45 days). Diabetic patients had a significant delay in inpatient stay (p = 0.03) while the presence of hypertension had no relation (p = 0.84).

The patients were managed on liposomal amphotericin B which was started preoperatively or simultaneously in the operation theatre. An average dose of 2.65 gm of Lip AmB was given during the average in patient stay of 47 days. Serial diagnostic endoscopic examination guided the need for second debridement and the total dose of drug. In case of clinical suspicion of co-infection, azole group of drugs were started with considerable clinical improvement. In patient V, isolates from palatal perforation showed aspergillous and nasal mucosa demonstrated mucor and hence the introduction of two classes of drugs significantly helped this patient. Sharana Mahomed et al. [9] also reports successful usage of amphotericin B and azole group of drugs. Co-infection is often a neglected aspect in treatment and limited data are available regarding the same. Patients were regularly monitored for nephrotoxicity and were managed with the help of a nephrologist. Active daily correction of electrolytes and renal parameters prevented the need for dialysis. Anti-fungal nasal douche was done in the immediate post-operative period and its effectiveness is yet to be established in randomised control study. In an average follow-up period of 15.12 months, all of the patients are alive and disease-free (Table 4).

Table 4 Medical management
Full size table

The orbital complications of the patients are as shown in pie chart (Fig. 2). About 50% of patients had preseptal cellulitis, 37% had orbital cellulitis, 25% presented with cavernous sinus thrombosis while 37% had no orbital involvement per se. The orbital complications were managed in consultation with an ophthalmologist conservatively with intravenous antibiotics and topical preparations. Daily visual assessment was done and none of patients required decompression or exenteration for the same.

Discussion

The lack of established protocols for the treatment of invasive fungal sinusitis often leaves question marks when it comes to accurate decision making. The number of cases detected in our institute is on the rise which necessitates the need for a protocol. A sudden hike in cases is noted during the monsoon season. The causative organisms are often present in the upper airways [3] and a predisposing environment triggers the infectious spread. Systematic probing into the probable trigger factors and corrective measures initiated along with medico-surgical treatment is the key.

In our study uncontrolled diabetes mellitus was the most common trigger factor and adequate control of blood sugar is vital [7]. Ketoacidosis is associated with early angioinvasion as the low serum pH decreases the phagocytic activity of neutrophils and other macrophages. Neutrophils are the major line of defence against these fungi. The reported survival rates is quite varied in literature with Kasapoglu et al. [3] reporting 76.5%, Marcus et al. [4] 21%, Justin et al. [5] 46% and Kiran Bala et al. [7] 70.6%.

The presentation varies from patient to patient. The characteristic blackish eschar is a hallmark sign but is present only in 40–50% of cases [6] and one should not wait for the same. The delay in presentation for an Otorhinolaryngology consult and the delay in initiation of treatment (average number of days = 5) can contribute to valuable time lost. The delay was later attributed to the inefficiencies in sample collection and reporting. Educating other departments to suspect the disease entity and efficient sample collection and reporting may be life saving for the patient.

Hence in any suspecting individuals, a battery of tests is done. KOH mounts acts a screening test alone and fungal culture is required, which is time consuming [3, 8]. Fungal culture results are compromised by the reduced viability of zygomycetes hyphae; hence it is isolated only in few patients. Hence pathological evidence of tissue invasion and identification of fungal hyphae was taken into account. This was aided by frozen section or routine HPE. In our study, HPE could identify fungal aetiology in 100% cases. Early identification of aetiology is crucial and thereafter a systematic protocol was followed in all patients (Fig. 3).

Fig. 3
figure 3

Orbital complications

Full size image

All the patients underwent endonasal debridement as the latest studies support improved outcomes with endonasal approach. Kasapoglu et al. [3] in his series has reported better survival rates for patients undergoing endonasal surgery (90%) when compared to open surgery (57.1%). Selected patients underwent debridement via transpalatal or transzygomatic, in addition to endonasal route. The rationale behind the same was to reduce the fungal load and increase drug penetrance. Endoscopic debridement was preferred to radical surgery as the patients were constantly monitored; and equivalent or better results were achieved with less morbidity. However no comparative studies was undertaken simultaneously. An average of 2.6 g of Liposomal amphotericin B was given for all patients. Kiran Bala et al. [7] have reported an improved survival with liposomal amphotericin B (88%) over conventional (66%). AmB is licensed by the US Food and Drug administration [10] as the first line drug for invasive fungal sinusitis and hence was selected as the drug of choice in our study. Posacanazole has been reportedly used as a salvage therapy [1] in cases of refractory mucor and it has been successfully used in patient VIII of our study. Other options like capsofungin has been used by Kazak et al. [11] but the lack of data from randomised controlled studies, adequate response of our patients to AmB and azole groups and the high cost involved limited their usage. The good results of our study may be attributed to the facts that almost all cases were identified in its nascent stage and were primarily sino-nasal with or without orbital/palatal involvement. None of the patients had intracranial extent of the disease. Parikh et al. [12] also reported an overall lower mortality rates compared to other studies and attributed the intracranial spread to be the highest predictor of mortality. Great care was taken to start the medical and surgical treatment simultaneously. Serial DNE and inpatient analysis was meticulously done for at least a month to detect any resurgence of the disease. There was no hesitation in considering a second de Kock no debridement if the mucosa showed unsatisfactory healing. Neutropenia was not detected at any stage of our treatment. Many studies including Cho et al. [13] has attributed neutropenia to increasing mortality rates (Fig. 4).

Fig. 4
figure 4

Management protocol

Full size image

The absence of clinical improvement within 48 h of initiation of treatment should be approached with caution. The possibilities being inadequate debridement, lower dosage of drug, possible co-infection, underlying trigger is not addressed to or the patient may be in sepsis. Serial DNE and imaging guided further plan of action.

Conclusion

A cookie cutter treatment is definitely not applicable to every patient, but a streamlined protocol-wise approach has achieved good results. A high index of suspicion and aggressive medical and surgical treatment is essential. Multiple endoscopic debridement supplemented by site specific debridement have brought good results. A good histopathological and microbiological support is needed to guide treatment.