Invasive Fungal Sinusitis Presenting as Acute Posterior Ischemic Optic Neuropathy

ABSTRACT

Invasive fungal sinusitis causes painful orbital apex syndrome with ophthalmoplegia and visual loss; the mechanism is unclear. We report an immunocompromised patient with invasive fungal sinusitis in whom the visual loss was due to posterior ischaemic optic neuropathy, shown on diffusion-weighted MRI, presumably from fungal invasion of small meningeal-based arteries at the orbital apex. After intensive antifungal drugs, orbital exenteration and immune reconstitution, the patient survived, but we were uncertain if the exenteration helped. We suggest that evidence of acute posterior ischaemic optic neuropathy should be a contra-indication to the need for orbital exenteration in invasive fungal sinusitis.

Introduction

Invasive fungal sinusitis (IFS), is a potentially lethal disease, often presenting as an acute or subacute, painful orbital apex syndrome, with monocular visual loss and ophthalmoplegia. The precise mechanism of the ophthalmoplegia and visual loss is not clear and neither is the role of radical surgery–orbital exenteration. We report the case of a patient, immunocompromised after bone-marrow transplantation, who developed IFS while taking antifungal prophylaxis. In this patient the mechanism of the visual loss was posterior ischaemic optic neuropathy (PION)–shown on diffusion-weighted MRI, we presume from fungal invasion of small meningeal-based arteries at the orbital apex. While the patient survived after intensive antifungal drug treatment plus orbital exenteration and finally immune reconstitution, we were not certain, if the exenteration was really necessary.

Case history

A 34 year old man with relapsed Philadelphia chromosome positive, acute lymphoblastic leukaemia received double cord-blood transplantation, after conditioning with cyclophosphamide and total body irradiation. During this period, he was given antifungal prophylaxis with itraconazole 400mg daily. A detailed chronology of events afterwards follows:

Day 4. Febrile neutropenia noted. CT of the brain/orbit/sinuses showed opacification of the ethmoid and sphenoid air cells; CT chest clear. Empirical intravenous antibiotics were started: Cubicin (daptomycin) and Tazocin (pipercillin plus tazobactam).

Day 8. Patient noted pain and swelling around his right eye. Examination showed conjunctival injection, chemosis, and mild proptosis. Vision was 6/6 in each eye without a relative afferent pupillary defect but with a mild, generalised limitation of right eye movement, without ptosis. Repeat CT now showed a breach of the right medial orbital wall with fat-stranding and thickening of medial rectus, and also a breach of the right anterior cranial fossa floor with a low density area of the adjacent brain, indicating cerebritis. Antifungal therapy was now changed to liposomal amphotericin 10mg/kg/day to cover for possible rhino-cerebral mucormycosis. Endoscopic examination of the paranasal sinuses revealed no eschar and no fungal organisms were identified on histopathology of the ethmoid mucosa and bone biopsy specimens. Culture failed to isolate any significant pathogens; however PCR testing was positive for Aspergillus species.

Day 12. Patient noted to have a near complete right ophthalmoplegia. Vision in the right eye was now down to 6/9 with a relative afferent pupillary defect. Right eye intraocular pressure was 40mmHg; left was 16mmHg. An emergency canthotomy/cantholysis was performed at the bedside for presumed orbital compartment syndrome.

Day 13. Patient suddenly loss all right eye vision, down to no-light-perception. MRI of orbits and brain showed extensive opacification of the right ethmoid sinus and of both sphenoid sinuses and extensive fat-stranding, swelling and enhancement of all right eye extraocular muscles and an increasing area of cerebritis and dural enhancement, in and below the right frontal lobe. (Figure 1a, Figure 1b). DWI showed high signal intensity of the right posterior optic nerve, confirmed on ADC map representing abnormal restricted diffusion most likely due to ischaemia/infarction (Figure 2a). Immediate endoscopic medial orbital wall decompression performed. Angio-invasive fungal infection confirmed on histopathology of the nasal mucosa; Aspergillus fumigatus was eventually cultured from this specimen (Figure 3a).

Figure 1a.

Contrast enhanced coronal T1 sequence (Day 13) demonstrates a communication between the right posterior ethmoid air cells and both the orbit and the anterior cranial fossa: fluid collection with enhancing margin extends from the sinuses to the inferior frontal lobe. The appearances are in keeping with an abscess. Note the herniation of the right medial rectus muscle through the surgical defect in the medial orbital wall. There is extensive pan-sinus opacification. There is no evidence of compression of the optic nerve at the orbital apex. ON=optic nerve; MR=medial rectus.

Figure 1b.

More anteriorly the thickened right medial rectus muscle extends medially into the right ethmoid air cells. SR=superior rectus; SO=superior oblique; MR=medial rectus; NS=nasal septum.

Figure 2a.

Diffusion weighted image (DWI) before orbital exenteration (Day 13) shows diffusion restriction (confirmed on ADC map) in the right optic nerve at the orbital apex and in the optic canal.

Figure 3a.

Ethmoid mucosa from FESS performed on 7 August shows angio-invasive fungal inflammation. Branching septate fungal hyphae are seen within infarcted vessels (arrow). PCR and culture were both positive in this specimen for Aspergillus fumigatus. DPAS special stain.

Day 17. Emergency exenteration of the right orbit, septectomy and skull base resection was performed by a team comprising an oculoplastic surgeon (RG) and an ENT-skull base surgeon (AA). Histopathology of the exenteration specimen showed no abnormality of the eye, the optic nerve, the central retinal artery or vein. The slides were eventually examined by three anatomical pathologists (RG, SC, MR) and some degenerate fungal hyphae found in the deep orbital tissue -the tissue had been fixed only after 48 hours, (Figure 3b). Aspergillus PCR was positive on both sinus and orbital tissue but culture was negative.

Figure 3b.

Deep orbital apex tissue. A few degenerate fungal hyphae were identified. The specimen was PCR positive but culture negative for Aspergillus. Preservation artefact is present as the specimen submitted first to microbiology for culture and PCR and to Anatomical Pathology only 48 hours later.

Day 22. Following confirmation of Aspergillus fumigatus cultured from the specimen taken on Day 13, antifungal therapy converted to combination therapy with anidulafungin (100mg daily) plus voriconazole (300mg bd). Repeat DWI MRI still showed high-intensity signal in the residual optic nerve posterior to the limit of the exenteration (Figure 2b). Two weeks later anidulafungin stopped but voriconazole was continued.

Figure 2b.

Diffusion weighted image (DWI) after orbital exenteration (Day 22) shows showing persistent diffusion restriction in the residual right optic nerve.

Following immune reconstitution, evidenced by recovery of the total white cell count, the patient developed headache. MRI now showed that the right frontal lobe abscess was surrounded by increased oedema, so it was evacuated at craniotomy. Histopathology showed angio-invasive fungal hyphae but the tissue remained culture negative despite extended incubation. The patient made a gradual recovery and was discharged from hospital on voriconazole 300 mg bd, 108 days after the bone-marrow transplant. He remains fungus-free and leukemia-free 2 years later.

Discussion

Our patient, febrile and neutropenic from aggressive treatment for acute leukemia, already on antifungal prophylaxis with oral itraconazole, developed a painful swollen, red eye, then ophthalmoplegia, initially without loss of vision. He was found on CT to have paranasal sinusitis extending into the anterior cranial fossa and orbit and was started on intravenous amphotericin. Four days later he suddenly lost all right eye vision. MRI confirmed orbital and intracranial invasion and showed posterior optic nerve infarction without evidence of cavernous sinus invasion or carotid artery occlusion. Right medial orbit decompression biopsy showed angio-invasive Aspergillus, but the orbital exenteration specimen showed a normal globe, optic nerve and central artery and only scanty fungus. Specimens from drainage of right frontal abscess also showed Aspergillus.

Since in the exenteration specimen, devitalized fungal hyphae were identified only in the tissue from the orbital apex, we propose that our patient went blind not from direct neural invasion of the optic nerve by fungus but from acute posterior ischaemic optic neuropathy (PION), perhaps from fungal angio-invasion of small meningeal arteries supplying the optic nerve in and around the optic foramen.^1,2

PION is a difficult clinical diagnosis³ that can, in some cases, be convincingly validated by showing restricted diffusion along a segment of the optic nerve^4–13 – as in this case. Restricted diffusion along the optic nerve can also occur in PION from septic cavernous sinus thrombosis^14,15, in traumatic¹⁶ and in lymphomatous optic neuropathy¹⁷, in optic neuritis^10,18,19 and at the optic disc in papilledema.²⁰

Recognised causes of PION are hypotension, sometimes from surgical bleeding, giant cell arteritis²¹ and invasive fungal sinusitis (IFS).^22–25 Optic neuropathy as a result of direct fungal invasion by Aspergilllus fumigatus has been reported in IFS²⁶ (from this institution 40 years ago), and also from PION due to fungal embolization in endocarditis.¹ These observations are of clinical importance. The reason we offered our patient, after an agonising discussion, such a mutilating operation was that, we were convinced that the fungus had already invaded his orbital apex and was heading for his cavernous sinus and carotid artery, with likely fatal consequences.²⁷ While it is possible that this was what was about to happen, there was little evidence of actual fungal invasion of the orbit. So it is more likely that antifungal drugs saved this patient from such an unhappy fate rather than the surgery. There is no agreement on which patients with IFS should have exenteration, debridement or just antifungal drugs^28–31–even within the ophthalmology²⁸ and otorhinolaryngology²⁹ departments of the same university. One of our previous patients with IFS due to mucormycosis also survived–one of the few who did, after exenteration³²; in that case there was fungus in the exenteration specimen, but in only 3/10 samples. The present case unfortunately raises more questions about IFS than it answers. One of these is why do IFS patients develop ophthalmoplegia? Simply from congestion of the extraocular muscles, apparent on MRI (Figure 1b), from intra-cavernous perineural invasion^33,34 inapparent on MRI or perhaps from invasion of small meningeal arteries near the orbital apex.²

Culture-based fungal diagnostics are insensitive and slow.³⁵ Antigen-based (galactomannan) and molecular (PCR) testing is better, especially in patients with haematologic malignancies³⁶, although the sensitivity of galactomannan and PCR testing is reduced with antifungal prophylaxis, less so with itraconazole.³⁷ Although combination therapy with voriconazole plus anidulafungin has not been shown to be better than voriconazole alone³⁸, we decided to use both in this patient due to the rapid progression of his disease despite exenteration. Following bone marrow engraftment an immune reconstitution inflammatory syndrome (IRIS) ³⁹ occurred with new headache, fever and increasing edema around the frontal lobe abscess. Such paradoxical worsening of symptoms and infection is the hallmark of IRIS and corresponds to the recovering immune system reacting to an opportunistic pathogen. The role of steroids in the treatment of IRIS remains controversial. This patient’s immunosuppression (with corticosteroids) was increased but with a poor response: ongoing headache and fever, which resolved only when the brain abscess was evacuated. With immune reconstitution, engraftment and symptomatic improvement, voriconazole alone was considered sufficient anti-fungal therapy.

This question remains: “why exenterate IFS?” The answer is, presumably, to try to extirpate a focus of devitalized, necrotic orbital tissue, supposedly teeming with fungi, in order prevent intracranial and intravascular spread of infection. However, in this patient while there was already intracranial spread, the exenterated orbital tissue was virtually normal and contained almost no fungi. The patient’s neuroophthalmic deficits–the optic neuropathy and ophthalmoplegia, were not due to massive involvement of the orbital tissues, but presumably due fungal invasion and occlusion of small dural arteries at the orbital apex supplying the optic and ocular motor nerves. In other words there was really no point in exenteration. Perhaps the clue that we ignored was the PION which indicated that while there was invasion of small orbital apex arteries, there was no disease throughout the orbit and no point in an exenteration. Next time we should know better.

Acknowledgments

Thanks to Mr Michael J. Todd for helping with the figures.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article.