Invasive cavernous sinus aspergillosis: an orbitocranial approach for diagnosis and management

Sino-orbital aspergillosis classically involves the paranasal sinuses before invading the orbit. Early treatment is typically indicated given the aggressive nature of this disease. However, challenges arise given its rarity and ambiguous presentation both clinically and radiographically.^{,  ,  ,}  We report an immunocompetent patient with invasive cavernous sinus aspergillosis presenting with limited sinus involvement and illustrate the critical role of biopsy in diagnosis and therapy.

An 85-year-old male with type 2 diabetes mellitus (hemoglobin A1c: 8.5%) presented afebrile with severe eye pain, diplopia, and decreased vision in the left eye. The patient was presumptively diagnosed with Tolosa-Hunt syndrome 3 months ago at an outside institution and was treated with high-dose intravenous (IV) methylprednisone. His vision loss and pain improved; however, symptoms recurred after discontinuing treatment. On examination, visual acuity was no light perception on the left with complete dyschromatopsia. Fundus examination showed temporal pallor of the optic nerve head in the left eye with an ipsilateral afferent pupillary defect on pupillary examination. Ocular motility showed limited abduction and infraduction to 15 degrees and limited supraduction to 30 degrees. Hertel exophthalmometry showed 5 mm of relative proptosis with complete ptosis and mild resistance to retropulsion. Neurologic examination revealed reduced sensation on all branches of the trigeminal nerve, greatest in the ophthalmic division. White blood cell and platelet counts were normal on laboratory testing. Compared with scans from 3 months ago, magnetic resonance imaging (MRI) of the brain/orbit (Fig. 1) after glucocorticoids showed interval worsening of a poorly defined, infiltrative mass involving the left orbital apex, optic canal, anterior clinoid process, and anterior cavernous sinus with concomitant involvement of the carotid artery. Lateral rectus thickening and increased T2 signal intensity of the optic nerve were noted (Fig. 1B). The paranasal sinuses appeared clear of disease without arterial or dural venous sinus thrombosis on magnetic resonance angiography. Given the unknown diagnosis, biopsy was performed using an orbitocranial lateral orbitotomy with a bone flap to explore the cavernous sinus (Video 1, available online). Tissue samples were collected intraoperatively, formalin-fixed, and submitted to pathology for microscopic analysis. Samples of the left orbital apex, lateral rectus, periosteum, apical bone, and a whitish fluid expressed upon opening the annulus were obtained for culture and cytology. A white, necrotic mass (1.2 cm × 0.3 cm × 0.3 cm) was removed from cavernous sinus and sent for pathology. The patient tolerated the procedure well and described immediate postoperative relief of eye pain. There was no interval improvement of vision. Pathology results for the left orbital apex, lateral rectus, periosteum, and apical bone showed benign tissue with mild inflammation. Fungal and bacterial cultures, along with acid-fast staining of the purulent liquid, were negative. Biopsy samples of the necrotic mass and surrounding cavernous sinus revealed pauci-septate and morphologically broad hyphae, branching at wide angles with invasion of the underlying cavernous sinus tissue. The patient was diagnosed with mucormycosis and was started on a 2-week course of IV liposomal amphotericin B (5 mg/kg for 4 days followed by a 10-day course of 10 mg/kg) and caspofungin (70 mg loading dose followed by a 50 mg maintenance dose daily). Blood and biopsy cultures were negative for bacterial and fungal growth. Given these ambiguous findings, infectious disease was consulted and cavernous sinus tissue was sent for polymerase chain reaction (PCR) testing, which detected Aspergillus fumigatus. Amphotericin was discontinued and voriconazole 200 mg PO q12h was added. Repeat MRI showed stable postsurgical changes with no significant interval expansion of fungal invasion. The patient did not have deterioration in ptosis, alignment, or facial sensation after surgical intervention. The patient was discharged on a life-long regimen of voriconazole 200 mg PO q12h and caspofungin 50 mg IV daily and is currently stable 6 months after treatment without recurrence of ocular symptoms.

We report a case of an immunocompetent patient who presented with nonspecific ocular and neuro-imaging findings, and was initially diagnosed with a nonspecific, steroid-responsive inflammatory condition. Given a worsening clinical picture, the patient underwent orbitocranial biopsy of the cavernous sinus and was diagnosed with invasive cavernous sinus aspergillosis by PCR.

The mechanism for intra-orbital fungal invasion remains poorly understood. Postulated theories include osseous, vascular, lymphatic, and perineural spread.^,  Typically, paranasal sinus involvement precedes or is concurrent with orbital disease. In our case, sinus involvement was hardly appreciable. This may have been suggestive of an alternate route for fungal transmission, although no other source was identified. Alternatively, sinus involvement could be extremely subtle, for instance, involving an Onodi cell, which was visible on the patient's MRI (Fig. 1D). Onodi cells, pneumatized posterior ethmoid cells, can play a significant role in sino-orbital optic neuropathy given their close anatomical association with the optic nerve. Aspergillus infection of an Onodi cell has been previously reported. However, these cases were either localized to an isolated mucocele or exhibited extensive paranasal sinus involvement, most commonly the sphenoid sinus. Interestingly, our patient presented without widespread paranasal sinus involvement. In our case, the Onodi cell likely served as a portal of fungal entry for orbital and cavernous sinus invasion. This suggests that the site of fungal entry may not be obvious and clinical suspicion for invasive aspergillosis may be reduced.

Invasive aspergillosis can be mistaken for a variety of conditions ranging from inflammation to malignancy.^{,,  ,}  Diagnosis often requires repeated neuroimaging and biopsy as seen in cases where aspergillosis was not diagnosed until autopsy.^{,  ,} 

Histology serves a critical diagnostic role for aspergillosis with characteristic septated and acute-angled hyphae observed on microscopy. In our case, however, the hyphae appeared pauci-septate and morphologically broad, branching at wide angles on microscopy, suggesting mucormycosis. This challenge in histologically differentiating aspergillosis and mucormycosis has been similarly reported by Arkin et al. Although culture is considered the gold standard for diagnosis, the blood, purulent fluid, and tissue samples did not yield fungal growth, and PCR was needed for definitive diagnosis.

Invasive aspergillosis masquerading as Tolosa-Hunt syndrome has been similarly observed by Marcet et al, whereby aspergillosis was diagnosed only after the patient's death. Evidence for the therapeutic efficacy of steroids for aspergillosis is lacking and remains heavily debated. Our patient's unresponsiveness to steroids further supports the notion that, although corticosteroids can provide temporary symptomatic relief, visual prognosis may worsen.^{,  ,} 

Biopsy may be indicated in such cases with nonspecific, acute optic neuritis. Previous reports have suggested fine needle aspiration biopsy as an initial diagnostic modality. However, utility of this approach has been limited to cases with significant sinus involvement. Given the atypical presentation of our case with respect to limited sinus involvement, a surgical approach was indicated. We performed a lateral orbitotomy extending intracranially to the middle cranial fossa to directly biopsy the region without a craniotomy, minimizing surgical morbidity. Given that the biopsy of the orbital apex tissue was negative for hyphae, sampling of this region alone may likely have lowered our suspicion of a fungal infection. Entry to the cavernous sinus in our case was of diagnostic value and necessary for initiating an antifungal regimen. In addition to direct tissue sampling, this surgical approach also provided a therapeutic decompression of the cavernous sinus and resolution of his pain. Because our patient presented with painful ophthalmoplegia secondary to fungal invasion of the cavernous sinus, removal of the necrotic mass within this region most likely explains our patient's immediate postoperative relief of eye pain.

Several practical implications can be deduced from this case. Initially, aspergillosis can mimic the presentation of orbital apex inflammatory syndrome, including temporary response to steroid therapy and minimal sinus findings. Orbitocranial aspergillosis should not be excluded despite the absence of fever, thrombophlebitis, and meningitis signs or symptoms. Biopsy may be necessary for definitive diagnosis, and, if so, a direct orbitocranial approach carries certain advantages over open craniotomy techniques. Our patient was initially started on IV amphotericin and was transitioned to voriconazole given its greater efficacy and fewer side effects. It should be mentioned that topical amphotericin B in conjunction with the IV formulation may be beneficial as various groups have shown increased drug delivery while successfully avoiding exenteration in selected patients. Aspergillus infection is typically associated with a greenish or yellow fluid discharge. Intriguingly, our case of Aspergillus presented with a purulent white discharge, which has similarly been observed with mucormycosis. These findings suggest that fluid color may not be mutually exclusive across pathogens, and differentiation by fluid appearance should be considered with caution. Finally, ancillary testing for speciation such as PCR may be required for selection of appropriate medical therapy.