W. RICHARDGREENMDFrom the Departments of Ophthalmology, University of Berne (Drs Frueh, Imesch, and Böhnke), and the Institute for Medical Microbiology, University Hospital (Drs Dubuis and Bodmer), Berne, Switzerland. None of the authors has a financial interest in the products described in this article.
Nontuberculous mycobacteria rarely cause keratitis. Most reported cases have been described in compromised corneas (ie, after penetrating keratoplasty, refractive surgery, and contact lens wear). The most frequently recovered nontuberculous mycobacteria are the rapid growers, Mycobacterium fortuitum and Mycobacterium chelonae.1,2 Occasional cases of keratitis caused by the slower-growing Mycobacterium gordonae have also been described.3 To our knowledge, this is the first reported case of Mycobacterium szulgai keratitis.
A 40-year-old patient developed a nonhealing peripheral corneal infiltrate in his right eye without a history of trauma. An initial diagnosis of herpetic keratouveitis was made, and he was treated with topical and systemic antiviral medication, topical corticosteroids, and ciprofloxacin. The result of a polymerase chain reaction for herpes simplex was negative.
Five months later, the patient was referred to our service. His visual acuity was 20/50 OD and 20/20 OS. A nasal stromal corneal infiltrate with epithelial ulceration and vascular pannus at its edge was seen in the right eye. Microcystic epithelial edema and an intraocular pressure of 37 mm Hg were noted. Grade 1 anterior chamber inflammation was recorded. A diagnosis of corticosteroid-induced glaucoma and possible herpetic keratouveitis was made, and corneal scrapings were performed.
Coagulase-negative staphylococcus grew in the cultures. The topical treatment included ciprofloxacin hydrochloride (Ciloxan), scopolamine, hydrocortisone with oxytetracycline hydrochloride and polymyxin B sulfate (Terra-Cortril), dorzolamide hydrochloride (Trusopt), latanoprost (Xalatan), and acyclovir ointment (Zovirax). Systemic valacyclovir hydrochloride (Valtrex) and acetazolamide (Diamox) were also given.
Because of worsening of the peripheral ulcer (Figure 1), conjunctival and corneal biopsies were performed 1 month later. The result of the Gram stain was negative, and bacterial cultures remained sterile. However, the auramine O fluorescent and Ziehl-Neelsen stains revealed a few acid-fast bacilli (Figure 2). The histological features showed a granuloma with early central necrosis (Figure 3). Mycobacteria were recovered from cultures (BACTEC 12B; Becton Dickinson, Sparks, Mass) after 14 days of incubation at 37°C. These were subsequently identified as M szulgai by polymerase chain reaction–restriction fragment length polymorphism analysis4 and by conventional methods.5 Minimum inhibitory concentrations were determined using a modification of the microbroth dilution method described by Yajko et al.6 The in vitro susceptibility of M szulgai is as follows:
Slitlamp picture of the left cornea showing a peripheral corneal ulcer and a heavily vascularized nodule.
High-power illustration of acid-fast bacilli (Ziehl-Neelsen, original magnification ×1000).
High-power photomicrograph of the excised conjunctiva shows chronic granulomatous inflammation (hematoxylin-eosin, original magnification ×200).
Topical corticosteroids were discontinued, and the treatment was changed to ciprofloxacin hydrochloride, 5 times daily, and amikacin, 20 mg/mL, 4 times daily. Clarithromycin drops, 20 mg/mL, were prepared according to the method of Helm et al,7 but were not tolerated and had to be discontinued after a few days. Clarithromycin (Klacid), 500 mg twice daily, was given orally for 3 months. Clinically, the inflammation subsided during the following 3 weeks, the epithelial erosion healed, and a vascularized scar formed. Eleven months after the biopsy and 3 months after discontinuation of topical treatment, the best-corrected visual acuity is 20/25 OD and the eye is not inflamed.
To our knowledge, this is the first reported case of keratitis due to M szulgai. The keratitis healed with topical and systemic treatment, and a final visual acuity of 20/25 OD was attained. The patient was immunocompetent and did not wear contact lenses, but had been treated with topical corticosteroids for several months before he was first seen by us. Mycobacterium szulgai is an unusual pathogen that accounts for less than 1% of all cases of nontuberculous mycobacteria infections. Diseases caused by this slow-growing mycobacterium usually involve the lung but may also lead to infections of soft tissue and bone.8
In our case, in vitro susceptibility testing showed rather low minimum inhibitory concentrations to clarithromycin and ciprofloxacin. Because topical clarithromycin was not well tolerated, amikacin was also used. The in vitro susceptibility of mycobacteria to antibiotics is different from that in vivo and may not predict the effectiveness of topical antibiotics. This may explain why the keratitis healed despite the fact that we did not use the optimal drug combination as indicated by the minimum inhibitory concentrations.
This case underscores the importance of performing corneal biopsies to detect unusual pathogens in patients with therapy-resistant keratitis.
Accepted for publication February 11, 2000.
Reprints: Beatrice E. Frueh, MD, Department of Ophthalmology, Inselspital, 3010 Bern, Switzerland (e-mail: firstname.lastname@example.org).
Frueh BE, Dubuis O, Imesch P, Böhnke M, Bodmer T. Mycobacterium szulgai Keratitis. Arch Ophthalmol. 2000;118(8):1123-1124. doi:10.1001/archopht.118.8.1123