Endogenous Rhodotorula minuta and Candida albicans endophthalmitis in an injecting drug user
- Institute of Ophthalmology, University of Sassari, Sassari, Italy
- Department of Biomedical Sciences
- Section of Experimental and Clinical Microbiology
- University of Sassari, Sassari, Italy
- Accepted for publication 14 February 2001
Editor,—Although still uncommon, endogenous fungal endophthalmitis has increased notably owing to widespread use of immunosuppressive therapy, hyperalimentation, and injecting drug use.Candida is the most common causative organism with isolated cases of Aspergillusreported.1 Rhodotorula species are ubiquitous, airborne, asporogenous yeasts recovered from food, air, soil, water, gastrointestinal tracts, and skin.Rhodotorula spp are usually environmental saprophytes; however, they may seldom cause both systemic and ocular infections, especially in immunocompromised patients.2Reported ocular infections include chronic dacryocystitis,3 keratitis,4 chronic postoperative endophthalmitis,5 and corneal lamellar graft infection.6 We report a case of endogenousRhodotorula minuta andCandida albicans endophthalmitis in an injecting drug user.
A 27 year old man was admitted with a 3 month history of progressive visual loss and intraocular inflammation in his right eye. He had initially been examined by other ophthalmologists who observed two large yellow-white lesions with fluffy borders in the posterior pole, diagnosed toxoplasmic retinochoroiditis, and started treatment with trimethoprim-sulfamethoxazole. After initial improvement, he eventually developed endophthalmitis and was referred to us for further treatment.
His medical history included injecting drug use and chronic hepatitis C virus. On admission, visual acuity was right eye light perception and left eye 20/20. Slit lamp examination of the right eye disclosed conjunctival injection and a 10% hypopyon. Ophthalmoscopic examination showed 4+ vitreous reaction that obscured retinal details. No retinal detachment was visible on B-scan ultrasonography. Left eye examination was unremarkable.
TPHA test and tests for the detection of serum IgM and IgG to cytomegalovirus, hepatitis B virus, HIV, andToxoplasma gondii were negative.
The patient underwent pars plana vitrectomy of the right eye. Hazy visualisation of the fundus disclosed a grey-white necrotic appearing retina. Bacterial and fungal cultures of the vitreous specimen were taken. Amikacin (400 μg) and vancomycin (1 mg) were injected intravitreally. In addition, injecting cefotaxime (500 mg twice daily) and topical tobramycin (0.3%), dexamethasone (0.1%), and atropine (1%) were given. Gram staining of the vitreous specimen revealed the presence of inflammatory cells and Gram positive yeasts, but no bacteria; as a result, oral ketoconazole (400 mg daily) was added to the treatment regimen.
Bacterial cultures were negative. On day 6, fungal cultures showed both red and white colonies, identified as R minuta and C albicans, respectively. Accordingly, injecting cefotaxime was discontinued, intravitreal amphotericin B (5 μg) given, and oral ketoconazole continued for a total of 4 weeks. Once the infection resolved, the retina appeared to be grey-white, ischaemic, and necrotic. Final visual acuity in the right eye was uncertain light perception.
Fungal organisms account for more than one half of cases of endogenous endophthalmitis.1 Endogenous fungal endophthalmitis may be caused by many different species, includingCandida,Aspergillus,Coccidioides,Cryptococcus,Blastomyces, andSporothrix.1 ,7 Patients with endogenous fungal endophthalmitis typically have systemic risk factors, including injecting drug abuse, long term injecting treatment, hyperalimentation, recent systemic surgery or trauma, indwelling bladder catheters, malignancy, immunosuppression, or debilitation. In a recent retrospective study, a history of injecting drug use was obtained in 22% of cases.8 Endogenous fungal endophthalmitis results from the bloodborne spread of organisms to the eye from a site of infection elsewhere in the body or from contaminated catheters or needles. Most cases occur without evidence of an ongoing fungaemia.1
Endogenous fungal endophthalmitis develops slowly as focal or multifocal areas of retinochoroiditis. Granulomatous or non-granulomatous inflammation is observed with keratic precipitates, a hypopyon, and vitritis. Mimicking toxoplasmic retinochoroiditis, posterior pole lesions appear yellow-white with fluffy borders, ranging in size from small cotton wool spots to several disc diameters. The lesions originate in the retina and result in exudation into the vitreous.
To our knowledge, there have been no reports of isolation ofR minuta from eyes with endogenous endophthalmitis. Other authors described a case of chronicRhodotorula endophthalmitis following cataract surgery.5 These types of organisms may be difficult to recover in the vitreous aspirate. We routinely request that the microbiology laboratory keep the cultures for longer than usual to allow uncommon, indolently replicating organisms to grow. As an injecting drug user, our patient most likely developed endogenous endophthalmitis following the bloodborne spread of fungi to the eye from a contaminated needle. Despite treatment with oral and intraocular antifungal agents and vitrectomy, he had poor visual outcome. Although endogenous fungal endophthalmitis generally has a very poor prognosis and early treatment of central lesions seldom salvages useful vision because of damage to central photoreceptors,1 delay in the diagnosis probably contributed to the poor visual outcome in our patient. Our report suggests that R minutashould be considered a possible aetiological agent of endogenous endophthalmitis, especially in patients with history of injecting drug use.