limited sample, prior therapy, delay in presentation, and laboratory studies. Several authors have used these techniques to detect and identify ocular mycobacterial isolates. They afford a more rapid and sensitive method for recovery and identification for both culture-negative and culture-positive samples [29–36].

Management

Clinical Diagnosis

Delay in clinical diagnosis is a common problem in the management of mycobacterial keratitis. The delay in diagnosis contributes to the protracted course and poor patient outcomes. Recognizing the clinical signs of mycobacterial keratitis is crucial in improving the clinical diagnosis. This can be difficult because of the chronic, indolent, progression of the disease, delay in presentation, prior use of steroids and topical antibiotics, mimicry of herpetic, fungal or diffuse lamellar keratitis, and lack of rapid, routine laboratory studies [1, 3, 6, 7, 15].

Clues for patients presenting with postLASIK infections include delayed onset, nonresponsive to topical antibiotics, presence of a white infiltrate in the corneal interface with spread to posterior stromal with or without satellite lesions, and tissue necrosis [1, 3–6, 33, 37–39].

Risk factors for patients with non-LASIK associated mycobacterial keratitis include corneal foreign body, trauma, contact lens wear, penetrating keratoplasty, cataract surgery, corneal suture, radial keratotomy, and chronic steroid use [3, 36]. More than 50% of patients with non-LASIK associated keratitis have an antecedent history of trauma or corneal surgery.

Medical Therapy

Minimal progress in the treatment and management of mycobacterial keratitis has been made since the first case diagnosed by Turner and Stinson. No standard or ideal antibiotic regimen for the prevention or treatment of mycobacterial keratitis exists. A variety of drugs including macrolides (azithromycin, clarithromycin), fluoroquinolones (ciprofloxacin, levofloxacin, gatifloxacin, and moxifloxacin), and aminoglycosides (amikacin, gentamicin, kanamycin, tobramycin) have been used with mixed results [3, 4, 15, 27–29, 40]. Both in vivo and in vitro responses to the commonly used antibiotics are variable and species specific [41] (Table 1.2). In vitro susceptibility results and in vivo response may also differ by source and geographical region [3, 42–47]) Correlation between the two is quite poor and frequently unreliable.

Contributing to this discord is the high lipid content of the organism’s cell wall, poor drug penetration, slow growth rate of the organism, drug toxicity, biofilm formation, and lack of simple, accurate methods for susceptibility testing.

A summary of in vitro results from our Institute (ocular) and the literature (ocular and nonocular) is highlighted in Fig. 1.7 and Table 1.2. Rates for the aminoglycosides

aReferences Brown-Elliot, Wallace, Reddy, Ford, Griffith for ATM

Fig. 1.7 In vitro susceptibility of BPEI mycobacterial keratitis isolates

ranged from 82% to 100%, with M. chelonae and M. fortuitum more susceptible in vitro to amikacin. M. fortuitum was more susceptible to the fluoroquinolones than was M. abscessus or M. chelonae, while the latter two were more susceptible to the macrolides. High-level resistance among rapidly growing mycobacteria isolates from Taiwan (<90% susceptible; M. abscessus – N = 92, M. chelonae – N = 39, and M. fortuitum group – N = 69) was reported for fluoroquinolones, macrolides, and tobramycin. All isolates were susceptible to amikacin. None of the isolates were recovered from keratitis [48]. Both clinical failure and high-level resistance to the fluoroquinolones, including gatifloxacin and moxifloxacin were documented for M. chelonae isolates recovered from a LASIK outbreak [4, 15, 49, 50].

Current recommendations for the management of mycobacteria keratitis include aggressive, topical combination therapy with clarithromycin (10 mg/ml) or azithromycin (2 mg/ml) supplemented with amikacin (50 mg/ml) or a fourth-generation fluoroquinolone [5, 6, 51]. Ford et al. documented clinical failure in 60% of their patients treated with the two-drug regimen of amikacin and clarithromycin, despite being susceptible in vitro [3]. Others documented similar rates of clinical failure with the two-drug regimen [15, 16, 40, 52–54]. The majority of patients (70%) with

Fig. 1.8 A 31-year old woman presented with a central interface infiltrate in her right eye 7 weeks after LASIK (a). Multiple acid-fast bacilli, epithelial cells and polymorphonuclear leukocytes were demonstrated from scrapings taken from the infiltrate (b). The flap was amputated 3 weeks later as new infiltrates appeared (c). Three months after onset, her cornea has a resolving scar with neovascularization, and her vision improved from hand motions to 20/50 (d) (From Solomon et al. [1], with permission)

mycobacterial infections following LASIK were managed with regimens of 3 (49%) or 4 (21%) drugs [6]. Successful medical therapy alone is rare. Epithelial debridement or in the case of LASIK, flap amputation is often required to enhance penetration of the medications and debulk the infection (Fig. 1.8).

Hu and colleagues [46, 55] demonstrated in vitro antagonism when an aminoglycoside was combined with imipenem, ciprofloxacin, and/or clarithromycin. This might account for the protracted course and often need to switch antibiotics to achieve successful eradication. Matoba et al. also found indifferent or antagonist results when combining amikacin with ciprofloxacin to treat mycobacterial keratitis [54]. A high rate of treatment failure in nontuberculous pulmonary disease using the combination of amikacin and clarithromycin has also been documented [5, 41].

Prolonged therapy with amikacin can lead to the emergence of resistant isolates. Similarly, the presence of an inducible erythromycin-resistant gene in all M. fortuitum isolates can lead to treatment failure with prolonged use of the macrolides to treat mycobacterial keratitis. Selection of clarithromycin-resistant isolates has also been documented with prolonged therapy [41, 43, 56].

Addition of a fluoroquinolone (ciprofloxacin, moxifloxacin, or gatifloxacin) to this regimen has improved epithelial healing and final visual outcome [50].