The patient underwent B-scan ultrasonography of the left eye, which excluded scleral thinning or associated orbital involvement. Soon thereafter he was taken to the operating room for excision of the involved area of conjunctiva and lower lid nodule, as well as exploration of the anterior orbit and inferior rectus muscle. A vascularized, infiltrative process that involved the conjunctiva and Tenon capsule and extended into the capsule of the inferior rectus muscle was noted. The underlying sclera appeared grossly normal and was not resected. The process spared the intraconal space. The affected areas were excised and debrided and the eyelid lesion was excised externally. Reconstruction was not done to maximize the effectiveness of topical therapy. Cultures for aerobic, anaerobic, fungal and acid fast bacilli (AFB) were obtained. Formalin-fixed tissue was submitted for histopathologic evaluation.

Postoperatively, the patient was empirically placed on bacitracin/polymyxin B ointment, fortified cefazolin and tobramycin drops, and oral clarithromycin. On postoperative day two, histopathology demonstrated foci of granulomatous inflammation encompassing microabscesses in the conjunctiva and skin. Special stains disclosed numerous AFB within vacuoles within the inflamed conjunctival tissue (See Figures 2 & 3). As a result, fortified amikacin and gatifloxacin drops were administered. Two weeks later, one of three anaerobic cultures was positive for “very light AFB,” and one of three AFB cultures was positive for atypical mycobacteria without speciation. Fungal cultures were negative. A reference laboratory identified the pathogen as Mycobacteria chelonae, resistant only to ciprofloxacin and doxycycline. Over the next two weeks, the patient’s clinical appearance improved dramatically. Oral and topical treatment has continued 12 weeks postoperatively, with no evidence of recurrent infection.

Although it remains relatively rare, non-tuberculous or “atypical” mycobacteria (NTM) have become increasingly significant ocular pathogens during the past few decades. Ramana S. Moorthy and colleagues reviewed clinically significant NTM and compiled more than 200 case reports of ocular and adnexal NTM infections.¹ NTM phenotypically are classified into four categories, known as Runyon Groups, based on their ability to produce pigment and their rate of growth in culture. Group IV, “rapid growers” that do not produce pigment, cause most ocular and adnexal NTM infections and include M. chelonae, M. abscessus and M. fortuitum.

Reflecting their rarity, there are no clear guidelines for treatment of ocular and ocular adnexal infections caused by NTM. Only two cases of conjunctival NTM infection have been reported previously in the literature, and both were associated with prior ocular surgery. The first was a case of conjunctivitis and scleritis caused by M. chelonae following vitrectomy.² The infected eye ultimately was enucleated despite four weeks of intravenous amikacin, imipenem and vancomycin in addition to oral clarithromycin and topical amikacin. The second case was an isolated conjunctivitis culture-positive for M. abscessus following scleral tunnel incision and phacoemulsification.³ The patient recovered after 15 weeks of therapy with oral clarithromycin, ethambutol, rifampicin, and ciprofloxacin and topical ciprofloxacin. In both cases, diagnosis was significantly delayed because initial conjunctival swab cultures were negative, necessitating conjunctival excision for biopsy and culture. The organisms in both patients also were resistant to initial topical antibiotics, reflecting NTM’s frequent resistance to clarithromycin and amikacin. Based on this admittedly limited anecdotal data, at least one month of systemic intravenous antibiotics (amikacin, imipenem or vancomycin) and topical amikacin are recommended therapy for isolated NTM conjunctivitis.¹

When considering the pathogenesis of our patient’s infection, it should be noted that, in contrast to the aforementioned cases of NTM conjunctivitis, the postop infection reported here was not localized to the region of initial surgical incisions. It is possible that the patient’s conjunctiva may have been colonized with mycobacteria prior to surgery, or he could have acquired the infection postoperatively via contaminated water.⁴ Given a number of reports regarding bacterial contamination of multiple use vials,^5,6,7 it is also conceivable that the offending pathogen was iatrogenically introduced when peribulbar anesthetic was administered intraoperatively and tracked along anatomic planes within the conjunctiva and Tenon fascia. However, no similar postop infections developed in other patients treated in the same operating room who received injections from the same vials of anesthetic. Although intraoperative use of mitomycin for pterygium excision has been linked to an increased risk of postoperative infections,⁸ it is unlikely to have played a role in this case because the mitomycin was applied nasally and the infection arose inferotemporally. Even more confounding is the involvement of the patient’s isolated left lower lid nodule, given its distance from the site of surgical manipulation.  

The key to timely, accurate diagnosis of NTM is a high level of clinical suspicion preoperatively, which assures that appropriate microbiological cultures and special histopathological stains for AFB and fungi will be performed. In this case, the indolent, relatively painless course with the appearance of multiple nodules alerted the treating physicians of the possibility of AFB infection.  REVIEW