Mycobacterial keratitis has been described after ocular trauma and various types of penetrating ocular surgery1 as well as in contact lens users.2 The most common bacterial subtypes are the nontuberculous, rapid-growing mycobacteria, including Mycobacterium chelonae and Mycobacterium fortuitum.3 This group of mycobacteria is ubiquitous in soil and water.4
The diagnosis and treatment of mycobacterial keratitis after laser in situ keratomileusis (LASIK) is often delayed when it is confused with other entities including diffuse lamellar keratitis. A delayed diagnosis contributes to the poor clinical outcomes reported in the literature.4,5 In a series of 24 cases of mycobacterial keratitis after various corneal procedures in south Florida, the average time to diagnosis was 4 months.3 A cluster of 7 cases of M chelonae keratitis after hyperopic LASIK were diagnosed 13 to 21 days postoperatively.5
Of the largest series of post-LASIK infectious keratitis described to date (15 cases) (C.L. Karp, Department of Ophthalmology, Bascom Palmer Eye Institute [BPEI], personal communication, 2000), 6 were Mycobacteria species followed by Staphylococcus aureus,4 fungi,3 acanthamoeba,1 and Pseudomonas.1 Five of the 6 cases with mycobacterial involvement resulted in flap amputation; none had penetrating keratoplasty (PKP). The findings in this series underscore the importance of considering atypical pathogens such as mycobacteria, fungi, and acanthamoeba in cases of post-LASIK infectious keratitis. The south Florida BPEI communication plus other series5,6 reporting post-LASIK infectious keratitis found a correlation between early diagnosis and improved visual outcomes.
The standard workup for a patient suspected of having mycobacterial infection includes acid-fast staining and plating on Lowenstein-Jensen culture media. However, given the rarity of mycobacterial isolates in routine infectious keratitis, most clinicians typically use Lowenstein-Jensen medium only if there is strong clinical suspicion of an atypical infection. Patients who have LASIK, however, are typically treated preoperatively or perioperatively with a broad-spectrum topical antibiotic such as a fluoroquinolone. Therefore, unusual organisms including mycobacteria, acanthamoeba, nocardia, and fungi are not adequately covered and must be considered in the differential diagnosis for any postoperative infiltrative lesion. For this reason, Lowenstein-Jensen media and acid-fast staining should be routinely used to culture post-LASIK infections.4
A healthy 51-year-old woman had uneventful myopic LASIK in the right eye at an ambulatory surgical center (ASC). Preoperatively, the best corrected visual acuity was 20/20 in both eyes. The patient had no previous ophthalmic surgery, trauma, or eye disease. She had not worn contact lenses for 2 years. Her medical and family histories were unremarkable. Pachymetry and topography were normal in both eyes.
The patient was pretreated for 3 days before surgery with ofloxacin 0.3% (Ocuflox®) 4 times a day and lid scrubs. An additional drop of ofloxacin 0.3% was instilled at the beginning and at the completion of the case. The lids were prepped using povidone–iodine, with care to avoid contact with the ocular surface and lashes. The flap was created with an Amadeus microkeratome (Allergan, Inc.), a 180 μm head, and an 8.5 mm suction ring. The Visx S3 excimer laser was used to perform the ablation. There were no intraoperative complications, and the flap was in good position before discharge. Postoperative drops included ofloxacin 0.3% and fluorometholone 0.1% (FML®), both 4 times daily.
On the first postoperative day, the uncorrected visual acuity (UCVA) was 20/20 in the right eye. The surgeon noted some central irregular epithelium but maintained the patient on the same ophthalmic drops. Follow-up was scheduled for 1 week.
On day 3, the patient reported a slight change in visual acuity in the operated eye that persisted until she presented to our group on day 10. The patient denied trauma to or getting water in either eye. However, she was exposed to dust while gardening. On examination, her UCVA was 20/20 in the right eye. The eye was white and quiet and the flap well apposed. However, a white focal infiltrate measuring less than 1.0 mm was seen in the anterior stroma at the edge of the pupil margin (Figure 1). The lesion was discrete with minimal cellular infiltrate, and the epithelium was intact. The anterior chamber was quiet. In reviewing the examination from the first postoperative day, the presence of unusual epithelium was noted in the same location as the white infiltrate.
At this time, the patient was taken to the laser suite to relift the flap and obtain cultures. The flap was reflected, and the white infiltrative lesion was seen on the undersurface of the flap rather than in the stromal bed. Both interface surfaces were scraped for cultures and copiously irrigated with a balanced salt solution and ofloxacin 0.3%. A pledget of ofloxacin 0.3% was placed in the interface, with the flap draping over it for approximately 30 seconds. Irrigation with a chilled balanced salt solution was again performed in the interface, and the flap was then refloated. A Johnston applanator was used to express extra fluid from underneath the flap; this was followed by 3 minutes of drying time. A bandage contact lens soaked in ofloxacin 0.3% was placed on the eye. The patient was directed to use ofloxacin 0.3% every hour for the first 24 hours, and the fluorometholone 0.1% was tapered to twice daily for 1 day and then discontinued.
Scrapings of the lesion were plated on blood and chocolate agar, thioglycollate broth, viral transport medium, and Sabouraud dextrose agar. Although the surgeon initially suspected the possibility of mycobacterial involvement, no Lowenstein-Jensen culture media was available at the ASC. Therefore, 2 glass slides were submitted, 1 for gram stain and the other for acid-fast staining, including the auramine-rhodamine fluorochrome method and the Kinyoun acid-fast stain. All specimens were submitted to the microbiology laboratory at the Medical University of South Carolina.
The gram stain showed rare leukocytes. The auramine-rhodamine fluorescent acid-fast stain was performed first and read as 1+ acid-fast bacilli (ie, 1 to 9 per 100 microscope fields in 1 sweep). The same slide was stained by the Kinyoun method, which verified the presence of acid-fast bacilli. An attempt was made to replate portions of the Sabouraud slant onto Lowenstein-Jensen media, but without success. Therefore, no culture-proven diagnosis was made. All other cultures were negative.
Within 1 day of relifting, the topical antibiotic regimen was tailored to treat atypical mycobacteria. The patient was immediately started on topical ciprofloxacin 0.3% and fortified amikacin 2.5%, alternating every 30 minutes during the day and every hour at night. The next day, topical clarithromycin (40 mg/mL) and fortified tobramycin (15 mg/mL) were added to the regimen, with all 4 drops administered every hour. This regimen was maintained for 3 weeks and then tapered to every 2 hours for another 4 weeks. All drops were discontinued by 8 weeks. The patient experienced moderate discomfort associated with the clarithromycin drops; however, it was not sufficient to discontinue the drops, and no surface toxicity or epithelial deposits were noted. No oral antibiotics were used.
Photographs taken on day 24 (Figure 2) showed resolution of the infiltrate. In this case, no sensitivities were available to further focus the antibiotic treatment. Artificial tears were added over the subsequent 3 weeks as the antibiotic drops were tapered. A low dose of fluorometholone 0.1% was restarted after the first 2 weeks of intensive treatment to decrease the inflammatory reaction. At the most recent follow-up visit at 6 months, no opacity was noted and the UCVA was 20/15.
We found 24 cases of mycobacterial infections after LASIK reported in the literature (Table 1). Nine (38%) proceeded to flap amputation, and 3 (13%) resulted in PKP. Delayed and/or ineffective treatment of mycobacteria may have contributed to the poor outcomes in these difficult cases.
Mycobacteria are aerobic, non-spore-forming, acid-fast, nonmotile bacilli. Members of the subgroup nontuberculous mycobacteria (NTM) are otherwise known as atypical mycobacteria. Of these, the “rapidly growing” species (including M fortuitum and M chelonae) are the most commonly isolated organisms in postsurgical keratitis.3Mycobacterium abscessus, a subspecies of M chelonae, has been described in post-LASIK infectious keratitis, as has Mycobacterium mucogenicum, a member of the M chelonae-like organisms. Mycobacterium tuberculosis often takes weeks to grow, whereas these rapidly growing mycobacteria are identifiable in the laboratory within 3 to 7 days.
Although the NTM species are ubiquitous, being found in soil, water, food, and animals, their source in post-LASIK keratitis is unknown. They must originate intraoperatively via instrumentation or postoperatively via environmental exposure. Only rarely do NTM cause disease. Mycobacterium fortuitum is known to cause cellulitis and osteomyelitis in traumatic and surgical wound infections.15Mycobacterium chelonae is most commonly found in immunosuppressed patients.16
A definitive diagnosis of mycobacteria requires growth of the organism in culture. However, in cases in which no appropriate transport media is available, smears may be used to give a rapid indication of the possibility of mycobacterial involvement. The sensitivity of the direct acid-fast smears (ie, Kinyoun, Ziehl-Neelsen) on pulmonary specimens is reported to be between 22% and 80%.16 In general, fluorochrome stains are more sensitive than acid-fast stains, presumably because organisms are more easily observed as a result of their fluorescence. Specificity to both methods is high.16 Positive fluorochrome smears may be confirmed with the Kinyoun stain on the same slide. Somoskovi et al.17 found that auramine-rhodamine fluorochrome is the most sensitive method to detect NTM in pulmonary specimens.
The therapy of choice for mycobacterial keratitis is amikacin and clarithromycin. Other potentially useful agents include topical tobramycin, the fluoroquinolones ciprofloxacin and ofloxacin,18,19 and the macrolide azithromycin.20 Topical clarithromycin, a macrolide antibiotic that has good tissue penetration and a long half-life, was efficacious in rabbit models of keratitis.4,5,21 Gross and coauthors21 showed that clarithromycin (20 mg/mL and 40 mg/mL) applied topically every 2 hours to rabbit corneas achieved therapeutic tissue concentrations of drug greater than the minimum concentration that will inhibit 90% of NTM after 12 hours of application. A large series of 24 cases of NTM keratitis in south Florida3 showed that, in general, the organisms were sensitive to amikacin and clarithromycin but resistant to the fluoroquinolones.
A study comparing topical amikacin (25 mg/mL), imipenem (25 mg/mL), ciprofloxacin (3 mg/mL), clarithromycin (20 mg/mL), amikacin plus ciprofloxacin, amikacin plus imipenem, and amikacin plus clarithromycin in the treatment of M chelonae keratitis in rabbit models19 found no significant difference in efficacy among the treatment groups (N = 90). In addition, viable M chelonae organisms were found in all eyes after 2 weeks of treatment. In a prospective study of M chelonae and M fortuitum keratitis, Hu and Luh18 found that only 2 of 11 patients responded to amikacin. Of the 9 patients refractory to amikacin treatment, 4 responded to topical ciprofloxacin (M fortuitum, n = 3; M chelonae, n = 1).
Given the variability in resistance patterns from these studies and the poor visual outcomes in the literature, we believe that multidrug coverage is warranted until drug sensitivities are known. Our initial therapy is hourly alternations of amikacin, clarithromycin, and ciprofloxacin. In addition, experience suggests there is a hazard in continuing topical steroids in the presence of mycobacterial keratitis.3,4 Therefore, steroids should be tapered and cautiously reintroduced only after prolonged multidrug treatment for the mycobacteria.4
Early identification and aggressive topical therapy resulted in resolution of the mycobacterial keratitis in our case; however, flap amputation may be necessary in cases in which the diagnosis is delayed and the infection is severe or worsening. Excising or lifting the flap may aid drug delivery to the organism. It may also alter the growth environment of the mycobacteria. Both factors may contribute to the resolution of the infection that is frequently observed once the flap is excised.
The literature indicates that early diagnosis and appropriate aggressive treatment are key to achieving a successful visual outcome in cases of post-LASIK infectious keratitis. Physicians must be on the alert for atypical organisms such as mycobacteria, and a routine workup should include evaluation of these organisms. Early flap lifting and lesion scraping using a wide choice of culture media including Lowenstein-Jensen as well as acid-fast and/or fluorochrome stains are essential.
We believe treatment should include a multidrug topical regimen until sensitivities are made available. This regimen should include, but may not be limited to, amikacin and clarithromycin. Physicians must ensure that compliance with frequent dosing is observed. Promptly tapering topical steroids also appears to be important in obtaining a good visual outcome.
The present case illustrates the successful diagnosis and treatment of a mycobacterial infection after uneventful LASIK. The excellent visual outcome in our patient 6 months posttreatment may be attributable to the early suspicion of mycobacterial involvement, with immediate confirmation using standard acid-fast and fluorochrome staining techniques followed by early tapering of topical steroids and frequent dosing with a 4-drug regimen.
Luanna Bartholomew, PhD, assisted with the report.
1. Robin JB, Beatty RF, Dunn S, et al. Mycobacterium chelonei keratitis after radial keratotomy. Am J Ophthalmol 1986; 102:72-79
2. Broadway DC, Kerr-Muir MG, Eykyn SJ, Pambakian H. Mycobacterium chelonei keratitis: a case report and review of previously reported cases. Eye 1994; 8:134-142
3. Ford JG, Huang AJW, Pflugfelder SC, et al. Nontuberculous mycobacterial keratitis in south Florida. Ophthalmology 1998; 105:1652-1658
4. Garg P, Bansal AK, Sharma S, Vemuganti GK. Bilateral infectious keratitis after laser in situ keratomileusis; a case report and review of the literature. Ophthalmology 2001; 108:121-125
5. Kouyoumdjian GA, Forstot SL, Durairaj VD, Damiano RE. Infectious keratitis after laser refractive surgery. Ophthalmology 2001; 108:1266-1268
6. Solomon A, Karp CL, Miller D, et al. Mycobacterium interface keratitis after laser in situ keratomileusis. Ophthalmology 2001; 108:2201-2208
7. Chandra NS, Torres MF, Winthrop KL, et al. Cluster of Mycobacterium chelonae keratitis cases following laser in-situ keratomileusis. Am J Ophthalmol 2001; 132:819-830
8. Reviglio V, Rodriguez ML, Picotti GS, et al. Mycobacterium chelonae keratitis following laser in situ keratomileusis. J Refract Surg 1998; 14:357-360
9. Chung MS, Goldstein MH, Driebe WT Jr, Schwartz BH. Mycobacterium chelonae keratitis after laser in situ keratomileusis successfully treated with medical therapy and flap removal. Am J Ophthalmol 2000; 129:382-384
10. Gelender H, Carter HL, Bowman B, et al. Mycobacterium keratitis after laser in situ keratomileusis. J Refract Surg 2000; 16:191-195
11. Seo KY, Lee JB, Lee K, et al. Non-tuberculosis mycobacterial keratitis at the interface after laser in situ keratomileusis. J Refract Surg 2002; 18:81-85
12. Brancato R, Carones F, Venturi E, et al. Mycobacterium chelonae keratitis after excimer laser photorefractive keratectomy. Arch Ophthalmol 1997; 115:1316-1318
13. Suresh PS, Campbell I, Herzig S, Rootman DS. Mycobacterium keratitis following hyperopic laser in situ keratomileusis. Can J Ophthalmol 2001; 36:272-274
14. Becero F, Maestre JR, Buezas V, et al. Queratitis por Mycobacterium chelonae tras cirgia refractiva con LASIK [letter]. Enferm Infect Microbiol Clin 2002; 20:44-45
15. Weitzul S, Eichhorn PJ, Pandya AG. Nontuberculous mycobacterial infections of the skin. Dermatol Clin 2000; 18:359-377, xi–xii
16. Murray PR, ed in chief. Manual of Clinical Microbiology, 7th ed. Washington, DC, ASM Press, 1999
17. Somoskovi A, Hotaling JE, Fitzgerald M, et al. Lessons from a proficiency testing event for acid-fast microscopy. Chest 2001; 120:250-257
18. Hu F-R, Luh K-T. Topical ciprofloxacin for treating nontuberculous mycobacterial keratitis. Ophthalmology 1998; 105:269-272
19. Hu F-R, Wang I-J. Comparison of topical antibiotics for treating Mycobacterium chelonae keratitis in a rabbit model. Curr Eye Res 1998; 17:478-482
20. Tabbara KF, Al-Kharashi SA, Al-Mansouri SM, et al. Ocular levels of azithromycin. Arch Ophthalmol 1998; 116:1625-1628
21. Gross RH, Holland GN, Elias SJ, Tuz R. Corneal pharmacokinetics of topical clarithromycin. Invest Ophthalmol Vis Sci 1995; 36:965-968