A, Hematoxylin and eosin (H&E) stain of case 9 demonstrating massive obliteration of the retina, choroidal expansion, fibrosis, subretinal hemorrhage, and an inflammatory focus temporally encompassing the retina and choroid (original magnification ×1). B, Myriad hyphae obvious with H&E or periodic acid–Schiff histology were present within the abscess in case 9 (H&E, original magnification ×40). C, Spore chains (chlamydoconidia) from case 3 (Gomori methenamine silver, original magnification ×40). D, Variety of hyphal forms from case 3 (Gomori methenamine silver, original magnification ×788).
Minckler D, Small KW, Walsh TJ. Clinical and Pathologic Features of Bipolaris Endophthalmitis After Intravitreal Triamcinolone. JAMA Ophthalmol. 2014;132(5):630-632. doi:10.1001/jamaophthalmol.2014.257
Copyright 2014 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
This is a notice to ophthalmologists performing intravitreal injections of triamcinolone or other medications of the risk for endophthalmitis caused by Bipolaris hawaiiensis–contaminated medications.
Histopathologic review of 5 eyes from a series of 17 treated with intravitreal triamcinolone for retinal vascular disease and diabetes mellitus over 1 year from 1 lot prepared by a compounding pharmacy. Endophthalmitis occurred in 14 of 17 eyes, the first presenting a mild vitritis 5 days following injection. The onset of signs and symptoms of endophthalmitis (eg, decreased vision, anterior chamber, or vitreous cells) in the 5 cases presented here varied from 5 to 174 days.
Conclusions and Relevance
Vitreous biopsy with cytospin and periodic acid–Schiff stain for hyphae was the most sensitive method for proving that fungus was present (positive in 3 of 5 of our cases). Five of 5 sample results from the whole series of 14 clinical cases were confirmed as positive for Bipolaris hawaiiensis by genotyping at the Centers for Disease Control and Prevention. Three of the removed eyes received fresh were culture negative; however, all 5 demonstrated organisms with fungal stains. Symptoms of fungal endophthalmitis can be minimal and delayed long beyond the 5 to 7 days expected for acute bacterial endophthalmitis. Also, apparently viable organisms persist in ocular tissues despite prolonged exposure to antifungal agents. This series reminds us that cytological or histological examination of tissue specimens may be necessary to exclude fungal infections that fail to grow in culture.
This article describes the histopathology in 5 eyes, which were from a clinical series of 14 cases of mold-induced endophthalmitis following intravitreal injections of triamcinolone. The organism responsible, Bipolaris hawaiieses, is a rare cause of human ocular infection and its morphology in ocular tissues has not been well described.
We conducted a histopathologic review of 5 eyes from a series of 17 treated with intravitreal triamcinolone for retinal vascular disease and diabetes mellitus over 1 year from 1 lot prepared by a compounding pharmacy (Table 1). This article received University of California, Irvine, institutional review board approval (HS 2012-8716); patient consent was waived.
Three of the eyes (cases 3, 4, and 11 designated by chronology in the series of 17) were clinically and pathologically phthisical, evidenced by gross shrinkage, scleral infolding, and internal disorganization.
Retinal and choroidal architecture were poorly preserved in all eyes including eye 9, which was grossly normal. However, eye 9 demonstrated equitorial granulomas containing numerous fungal elements obvious in periodic acid–Schiff (PAS) and hematoxylin and eosin–stained sections (Figure). Optic nerve and scleral extension were not found in any of our specimens. The choroid in all eyes was markedly disorganized and expanded by inflammatory debris but did not demonstrate organisms. Remnants of retinal pigment epithelium and choriocapillaris persisted in several eyes. In all cases, hyphae were found by both PAS and Gomori methenamine silver staining in inflammatory foci or vitreous remnants (Figure). The results from cultures taken from cases 3, 4, and 11 in the pathology laboratory were negative.
Bipolaris hawaiiensis is a demitaiceous mold and member of a large genus of fungi that includes at least 30 species. It is a well-known saprophyte and plant pathogen.1 Human infections are rare and generally limited to tropical and subtropical regions.1 The most common human infections have occurred in nasal tissues, skin wounds and burns, or lung.2 Allergic sinusitis and onychomycosis are relatively common.2 Endophthalmitis from the same compounding pharmacy was recently reported but without ocular histology.3 Very rarely, central nervous system infections have been described.4 Immunocompromised individuals are at the highest risk for infection by Bipolaris species.
To our knowledge, this is the first reported ocular histology on a series of intraocular infections due to Bipolaris hawaiiensis. The clinical findings of another 9 cases of Bipolaris hawaiiensis endophthalmitis from the same pharmacy (Franck’s Compounding Pharmacy, Ocala, Florida) and same lot number were recently reported by Sheyman et al.3 The same pharmacy a few months earlier had been involved in Brilliant Blue dye fungal contamination. The Centers for Disease Control and Prevention Morbidity and Mortality Weekly Report described 33 cases of fungal endophthalmitis from 7 different states from March to April 2012. Subsequently, triamcinolone was also suspected to have been contaminated in 13 cases. Isolates from 7 confirmed Brilliant Blue dye cases by genetic testing or culture were due to Fusarium incarnotum-equiseti.
Microscopy demonstrated well-preserved hyphal and spore forms except in the eviscerated case 2, in which only fragments of hyphae were found with PAS and Gomori methenamine silver staining. The results from cultures obtained via pars plana vitrectomy were positive for Bipolaris in our cases 2 and 3 and suspect in case 9. DNA sequencing of vitrectomy samples from early clinical cases by the Centers for Disease Control and Prevention documented the species to be identical to those from New York.3 The most rapid and effective proof that fungi were present was PAS staining of cytospin specimens obtained via vitrectomy (Table 2). The morphology of Bipolaris hawaiiensis in our cases was remarkably variable including branching septate hyphae, spores, rings, and chains of spores (catenulate chlamydoconidia) (Figure, C).
Systemic voriconazole (200-300 mg by mouth twice daily), the preferred agent for fungal endophthalmitis, was used in 4 of our 5 cases (Table 2).5 Voriconazole was also injected intravitreally in all 5 cases (Table 2). Each injection of voriconazole was 100 to 200 µg, escalating by 10 µg if injected following pars plana vitrectomy and gas-fluid exchange. Amphotericin B was administered to all 5 of our cases via intravitreal injection at doses of 10, 20, 30, and 40 μg escalating. Case 2 also received intravitreal ceftazidime and vancomycin.
The Bipolaris hawaiieses endophthalmitis cases described in this article convey at least 4 important clinical messages. First, the clinical onset of fungal endophthalmitis can be subtle and painless. Second, in our cases, the onset of symptoms was delayed for between 5 days (case 2) to 64 to 174 days (Table 1), substantially longer than expected for bacterial endophthalmitis. Third, when fungal endophthalmitis is suspected, vitrectomy and PAS staining of cytospin material has the highest likelihood of demonstrating organisms. Fourth, and most importantly, all of our 5 cases reported here received multiple intravitreal doses of voriconazole and 4 of 5 took voriconazole by mouth and still contained apparently viable organisms in ocular tissues. Clearly, appropriate antifungals cannot always kill these organisms.5 Because the worldwide ophthalmic community is doing many thousands of intravitreal injections of anti–vascular endothelial growth factor and steroids per year, especially in diabetic patients, great concern is appropriate for the sterility of compounded drugs.
Corresponding Author: Don Minckler, MD, MS, 1634 Aryana Dr, Encinitas, CA 92024 (firstname.lastname@example.org).
Accepted for Publication: January 21, 2014.
Published Online: March 13, 2014. doi:10.1001/jamaophthalmol.2014.257.
Author Contributions: Drs Minckler and Small had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition of data: Minckler, Small.
Analysis and interpretation of data: All authors.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: Minckler, Walsh.
Administrative, technical, and material support: All authors.
Study supervision: Minckler.
Conflict of Interest Disclosures: None reported.