October 2018


Presentation spotlight
Atypical mycobacterial interface keratitis after DMEK

by Stefanie Petrou Binder, MD, EyeWorld Contributing Writer

The corneal infiltrates grew along the graft interface. They began as ill-defined whitish spots and became whiter and sharper over time.
Source: Minh Tri Hua, MD


A mycobacteria-contaminated donor cornea can set transplantation surgery into a tailspin

Contamination during the surgical process is always a real possibility. Presenting a case of cornea graft contamination at the 22nd ESCRS Winter Meeting, Minh Tri Hua, MD, UZ Leuven, Belgium, reported on an incident of contamination that originated well outside of the operating room.

Day 1: Uncomplicated DMEK surgery

In March 2017, a 76-year-old male patient underwent Descemet’s membrane endothelial keratoplasty (DMEK) for Fuchs’ endothelial dystrophy. The graft was prepared by the surgeon in the OR, and the surgery was completed without any undo complications, including a thumbs-up on the day 5 postoperative follow-up examination.
The local tissue bank reached out on postoperative day 17 to inform the surgical center that an aerobic culture from the transport medium had turned positive. The foreign eye bank that provided the graft was contacted, and they soon confirmed the contaminated, infected medium. On day 22, the local tissue bank could isolate a nontuberculous atypical mycobacterium from the transport medium.
Nontuberculous mycobacteria are ubiquitous in the environment, found in soil, water, and aquatic animals. They cause opportunistic infections among both immunocompromised and immunocompetent individuals and have a broad spectrum of virulence. A subset of mycobacteria has been characterized as “rapidly growing” and is implicated in infections of catheters, post-LASIK, skin and soft tissue, and the lungs. Rapidly growing mycobacteria are gaining emerging importance in both sporadic infections and outbreak settings.1

Day 25: Interface infiltrates

On postoperative day 25, the patient had an uncorrected visual acuity (UCVA) of 0.6 logMAR. Dr. Hua and his colleagues called the patient back for an examination earlier than scheduled. The patient exhibited no symptoms at this point. A slit lamp exam revealed several corneal infiltrates, which were at first misdiagnosed. The infiltrates were situated between the graft and the corneal stroma. Dr. Hua started the patient on moxifloxacin 4x/day and tobramycin/dexamethasone 4x/day. He did not stop the corticosteroid regimen at that point, although he later would.
Five days later, the infiltrates were seen to be getting worse (postoperative day 30), but visual acuity was as yet unaffected by the infection, with a UCVA of 0.6 logMAR. “The infiltrates were increasing, getting worse and bigger. Optical coherence tomography (OCT) demonstrated the location of the infiltrates perfectly between the graft and the stroma at the interface,” Dr. Hua said.
Azithromycin per os 500 mg/d for 3 days then 250 mg/d was added to the patient’s treatment regimen. Dr. Hua referred the patient to an internist who replaced azithromycin with clarithromycin per os 500 mg 2x/day, which was recommended for 6 months along with amikacin eye drops 8x/day.
The treatment of nontuberculous mycobacteria is lengthy and varies by species. Macrolide-based regimens using agents such as clarithromycin can be costly, induce intolerance, and cause potential toxicity in patients. In patients failing standard therapy, the addition of moxifloxacin to a multi-drug regimen can improve treatment outcomes, although clinical evidence is sparse.2
At this point, Dr. Hua received notice from his hospital’s microbiology department that the organism in question was Mycobacterium chelonae, a nontuberculous mycobacterium classified as rapidly growing. M. chelonae grows best at 30–32 degrees centigrade and may have a long incubation period. Aside from its known association with skin and soft tissue infections, this organism causes catheter and post-surgical infections after implants, transplants, and injections.3
By the second postoperative month, the patient’s UCVA was 0.4 logMAR. After the first month on the prescribed treatment regimen, Dr. Hua noted the beginnings of changes in the appearance of the lesions in the graft/stromal interface. “We saw no new infiltrates, but what was changing was the color of the lesions. They turned bicolor with the central portion becoming whiter and the edges of the lesions becoming sharper. The size of the infiltrates was the same. We kept the patient on amikacin 6x/day, moxifloxacin 4x/day, tobramycin/dexamethasone 1x/day, and clarithromycin per os 500 mg 2x/day. By month 3, UCVA was slightly improved at 0.6 logMAR, but by the time we reached 4 months of treatment, the patient had become irritated at the medication regimen and oral antibiotics,” Dr. Hua said.
Dr. Hua observed that the lesions underwent further changes between the fourth and fifth months of treatment (day 137), with the evolution of the lesions from creamy to whiter and sharper over time.

Day 143: Repeat DMEK surgery

A repeat DMEK was performed on day 143. The graft exchange was carried out in the operating room, under air. “With a scraper we took out the first DMEK graft,” Dr. Hua explained. “The lesions, however, did not come out with the graft; they were still adhering to the stroma. A microbiologist was on standby to check the specimen by culture and staining. With the scraper we manually had to take out the infiltrates and rinsed with balanced salt solution.”
The microbiology report showed the presence of M. chelonae DNA in the infiltrates. The culture was still positive for living mycobacteria in the eye. The staining was also positive for the organism.
After the DMEK exchange, the patient’s UCVA was 0.7. In the first 2 weeks after surgery, he was not given corticosteroid eye drops. He received systemic antibiotics in addition to the previous regimen, which was maintained. Dr. Hua was able to examine the patient shortly before the ESCRS Winter Meeting and was satisfied with the patient’s progress. Antibiotics were stopped by month 5. He said that he would be checking the patient again to exclude any recurrences.


1. De Groote MA, Huitt G. Infections due to rapidly growing mycobacteria. Clin Infect Dis. 2006;42:1756–63.
2. Ryu YJ, et al. Diagnosis and treatment of nontuberculous mycobacterial lung disease: clinicians’ perspectives. Tuberc Respir Dis (Seoul). 2016;79:74–84.
3. Akram SM, Bhimji SS. Mycobacterium chelonae. StatPearls. May 16, 2017.

Editors’ note: Dr. Hua has no financial interests related to his comments.

Contact information

: oogziekten@uzleuven.be

Atypical mycobacterial interface keratitis after DMEK Atypical mycobacterial interface keratitis after DMEK
Ophthalmology News - EyeWorld Magazine
283 110
220 219
True, 10