Most cases of fungal sinusitis are allergic in nature rather than true infections because the etiologic agent is not growing in tissue. The disease process is a consequence of the host immunologic response to the fungus, which leads to marked inflammation, tissue necrosis and severely damaged fungal hyphae. In contrast, when a fungus enters and proliferates within tissue, then invasive infectious sinusitis results. This can have a particularly poor or even fatal outcome in immunocompromised patients. The timely differentiation between allergic and invasive sinusitis has a major impact on management decisions and prognosis.
An 18-year-old Hispanic-American young woman was admitted to the University of Texas Medical Branch in August with aplastic anemia (absolute neutrophil count 0/μL; hematocrit 18%; platelet count 33,000/μL). After 1 week of supportive therapy without signs or symptoms of infection, the patient developed a low grade fever (37–38°C) in addition to a 1-cm furuncle on her proximal left lower extremity. The furuncle was incised and drained, and its contents submitted for routine bacterial and fungal cultures. Empiric vancomycin therapy was begun thereafter, once blood and urine cultures had been obtained. On recovery of Escherichia coli and Acinetobacter baumannii, each in significant numbers, from the leg lesion, antibiotic coverage was changed to imipenem. No fungi were isolated from this lesion, and multiple blood and urine cultures remained sterile.
Over the next several days, the patient continued to spike fevers despite no other identifiable focus of infection. Approximately 2 weeks after hospitalization, however, she began to experience headaches, severe nasal congestion with rhinorrhea and intermittent ear pain. A head computed tomography scan demonstrated diffuse sinus mucosal thickening, most prominent within the right maxillary sinus, without evidence of bone destruction. The findings were suggestive of severe pansinusitis. Sinus endoscopy demonstrated diffuse mucosal erythema and edema with focal necrosis. The patient underwent right maxillary antrostomy, right anterior and posterior ethmoidectomy, right sphenoidectomy and septal resection on hospital day 18. Intraoperative pathology consultation confirmed the presence of invasive, nonpigmented fungal hyphae. Differential diagnoses based on morphology alone included the Zygomycetes, hyaline molds such as Aspergillus and phaeohyphomycosis. Permanent histologic sections of several surgically resected sinonasal sites revealed necrotic tissue containing infiltrative, Gomori methenamine silver-positive, septate hyphae with acute angle branches (Fig. 1A, B). Although the fungal hyphae did not appear pigmented in routine hematoxylin-eosin-stained sections, evidence of limited melanin production was detected using the Fontana-Masson technique. These histologic features were compatible with an invasive fungal sinusitis caused by a dematiaceous mold.1 Tissue submitted for fungal culture yielded a gray-black mold with a dark reverse that was identified as E. rostratum based on its formation of large (∼73- × 15-μm) cylindrical, transversely septate, 6- to 9-celled conidia with protruding hila (Fig. 1C).2
Empiric treatment with amphotericin B (1 mg/kg/d iv) was initiated on postoperative day (POD) 1 based on frozen section results but was subsequently changed to the liposomal formulation (5 mg/kg/d iv) on POD 5 because of its nephrotoxic effects. Voriconazole therapy (400 mg/d iv) was also instituted on POD 5, but elevation of hepatic transaminases prompted its discontinuation 14 days later. Throughout the next several weeks, the patient underwent multiple surgical sinus debridements and received numerous erythrocyte and platelet transfusions in addition to liposomal amphotericin B. On POD 50, itraconazole (400 mg/d iv) was added to the antifungal regimen based on results of mold susceptibility testing performed at the University of Texas Health Science Center San Antonio. Minimal inhibitory concentrations at 48 hours were generated for voriconazole (1 μg/mL), posaconazole (0.06 μg/mL), itraconazole (0.125 μg/mL) and amphotericin B (0.5 μg/mL). Although no interpretive criteria have been established for mold susceptibility testing to date, this particular isolate was inhibited by concentrations of antifungal agents considered to be readily achievable.3 (Dr M. G. Rinaldi, personal communication). Because of the failure of spontaneous bone marrow recovery, the patient also received granulocyte colony-stimulating factor, and after 3 weeks with no response, aggressive immunomodulatory therapy with antithymocyte globulin, prednisone and cyclosporin. Human leukocyte antigen typing of the patient's only sibling and available parent yielded no potential donors for bone marrow transplant.
Throughout her 64-day hospitalization, the patient received a 5-day course of standard amphotericin B (1 mg/kg/day iv), a 61-day course of liposomal amphotericin B (5 mg/kg/d) as well as separated 2-week courses of both voriconazole (400 mg/d iv) and itraconazole (400 mg/d iv). No evidence for dissemination of fungal organisms was ever established, but the patient did eventually develop right orbital proptosis. Magnetic resonance imaging of the head, orbits and sinuses demonstrated bulging of the floor of the orbit and lateral displacement of the right medial rectus muscle as a result of an expansile mass located within the right maxillary sinus. The patient eventually began to refuse further surgical intervention and ultimately succumbed to her aplasia, dying of multiorgan failure in the setting of Staphylococcus bacteremia and severe pulmonary hemorrhage without ever having recovered bone marrow function. An autopsy revealed the persistence of invasive hyphal structures within the sinuses but corroborated the absence of any local extension, including into the orbital fossae, or disseminated fungal disease.
Although fungal sinusitis in a diabetic patient caused by one of the Zygomycetes is quite a recognizable scenario to most, it is an uncommon occurrence. By contrast, a number of similar disease manifestations, which are less well-known but exceedingly more common, occur in the setting of various nonzygomycete fungal pathogen-host immunocompetency status combinations. For instance, the hyaline mold Aspergillus has been increasingly recognized as the causative agent in acute invasive, chronic invasive and allergic noninvasive sinus diseases.4,5 Similarly a number of dematiaceous molds have been implicated in several forms of fungal sinusitis.6,7
Mycotic sinus disease is typically divided into 4 general categories: (1) fungus ball; (2) allergic fungal sinusitis; (3) chronic indolent invasive sinusitis; and (4) acute fulminant invasive sinusitis.1,4,5 Because treatment of these entities may be somewhat disparate, it is prudent that accurate delineation be made on a case by case basis. For instance, authorities have proposed that systemic steroid therapy be instituted after surgical debridement in cases of allergic fungal sinusitis with the rationale that its pathogenesis is probably very similar to allergic bronchopulmonary aspergillosis.1,4 On the other hand, acute fulminant invasive sinusitis, particularly when accompanied by severe immune depression, would likely be approached with a combination of aggressive antifungal therapy and surgery. However, classification of sinusitis subtype might not always be straightforward. Clinically aggressive disease with concomitant bone destruction, for example, may be seen in allergic (ie, noninvasive) fungal sinusitis.1,4,7 Moreover histologic evidence of fungal tissue invasion accompanied by locally destructive disease may be seen in fully immunocompetent patients.8 Such a categorization should therefore be carefully established through the concerted efforts of clinicians, radiologists, surgeons and pathologists.
In a previous report of a 5-year-old with aplastic anemia complicated by E. rostratum invasive sinusitis, Sharkey et al3 detailed a somewhat similar progression as in the present case. Specifically a 3.5-month course of amphotericin B followed by itraconazole therapy prevented recurrence of a “susceptible” Exserohilum isolate (ie, one with low minimal inhibitory concentrations), although a second species of mold (Fusarium) was later recovered and the patient eventually died of bacterial sepsis.3 These findings suggest that the recovery of hematopoietic function in an aplastic patient plays an important role in ultimate outcome. Indeed, Douer et al9 described a patient with chemotherapy-induced neutropenia and invasive Exserohilum sinusitis who recovered completely after amphotericin B therapy and resolution of neutropenia. Because susceptibility data were not reported, however, one can speculate as to the role of white blood cell reconstitution versus antifungal therapy. Another strikingly similar case described by Aquino et al10 provides a particularly illustrative example of the potential severity of disease after dissemination. In that report, a patient with aplastic anemia developed acute E. rostratum sinusitis. Although long term amphotericin B therapy was implemented, surgical debridement was not described, and the patient died on hospital day 35 with severe hemorrhagic fungal bronchopneumonia. As in other fatal cases, the patient's immunohematologic function remained suppressed. Again antifungal susceptibility data were unavailable.
In summary, mycotic sinusitis can present as a chronic, indolent or acute, fulminant disease, the type and/or degree of which may not always be reflected by a patient's immune function. Although the clinical setting (disease severity, history of atopy, immunohematologic status), radiographic features (extent of involvement) and histopathologic findings (tissue invasion versus allergic morphology) can each be suggestive of a particular subtype of disease, ultimate categorization should reflect contributions from practitioners in all of these subspecialties. Finally, based on the present case and review of similar reported cases, it appears that successful treatment of severe invasive fungal sinusitis involving immunocompromised persons can similarly require the joint efforts of infectious disease specialists administering aggressive antifungal regimens, otolaryngologists performing sinus evacuations and surgical debridements as well as hematologist/oncologists assuring that hematologic function is regained as quickly as possible.
1. Torres C, Ro JY, El-Naggar AK, Sim SJ, Weber RS, Ayala AG. Allergic fungal sinusitis. Hum Pathol
2. McGinnis MR, Rinaldi MG, Winn RE. Emerging agents of phaeohyphomycosis: pathogenic species of Bipolaris
and Exserohilum. J Clin Microbiol
3. Sharkey PK, Graybill JR, Rinaldi MG, et al. Itraconazole treatment of phaeohyphomycosis. J Am Acad Dermatol
4. Waxman JE, Spector JG, Sale SR, Katzenstein AL. Allergic Aspergillus
sinusitis: concepts in diagnosis and treatment of a new clinical entity. Laryngoscope
5. Cameli-Rojas V, Mata-Essayag S, de Capriles CH, et al. Aspergillus
species in patients with chronic rhinosinusitis. Mycoses
6. Adam RD, Paquin ML, Petersen EA, et al. Phaeohyphomycosis caused by the fungal genera Bipolaris
and Exserohilum. Medicine (Baltimore)
7. Zieske LA, Kopke RD, Hamill R. Dematiaceous fungal sinusitis. Otolaryngol Head Neck Surg
8. Maskin SL, Fetchick RJ, Leone CR, Sharkey PK, Rinaldi MG. Bipolaris hawaiiensis
-caused phaeohyphomycotic orbitopathy. Ophthalmology
9. Douer D, Goldschmied-Reoven A, Segev, Ben-Bassat I. Human. Exserohilum
infections: report of Exserohilum
nasal infection in a neutropenic patient with acute leukemia and review of the literature. J Med Vet Mycol.
10. Aquino VM, Norvell JM, Krisher K, Mustafa MM. Fatal disseminated infection due to Exserohilum rostratum
in a patient with aplastic anemia: case report and review. Clin Infect Dis