What do we know about diffuse lamellar keratitis?

Although the syndrome has several names, diffuse lamellar keratitis is the most accurate, researchers say.

ORLANDO, U.S.A. — To date there are only four reports in the peer reviewed literature that specifically describe diffuse lamellar keratitis (DLK). The syndrome has also been referred to as nonspecific diffuse interlamellar keratitis, LASIK [laser in situ keratomileusis] interface keratitis, Sands of the Sahara syndrome, central focal interface opacification after LASIK and interface inflammation after LASIK.

“Diffuse lamellar keratitis is the name that has taken precedence,” Michael A. Lawless, MD, said during a presentation here at the American Academy of Ophthalmology meeting. “The clinical description is that it usually presents 1 to 6 days after LASIK. It’s confined to the interface. Inflammation usually starts peripherally and moves centrally. The conjunctiva is not inflamed and the anterior chamber is quiet.”

According to Dr. Lawless, the syndrome is difficult to pinpoint with an exact diagnosis due to the syndrome’s variability. Additionally, there are some very mild forms of the syndrome that go undiagnosed. Treatment strategies vary amongst the range of DLK stages. Various doctors have proposed anti-endotoxin strategies in efforts to avoid and manage DLK in order to preserve postoperative vision.

The source of the problem

“There are too many causes to actually list,” Dr. Lawless said. “And when there are too many to list, no one really knows what the actual cause is and that there are probably multiple causes.”

The scope of DLK varies, with some high-volume LASIK surgeons never having seen a case, while others experience focal outbreaks of DLK. Louis E. Probst, MD, has suggested an incidence of about 1 in 500. Causes may include povidone-iodine prep, balanced salt solution, metallic debris, contaminants from the eyelids and meibomian gland secretions, talc from surgical gloves, laser thermal effect, secondary inflammation after corneal epithelial abrasions, contaminants on the instruments, topical medications such as local anesthetic, bacterial cell wall exotoxin and endotoxin hypersensitivity and lubricant from the keratome. But according to Dr. Lawless, the bacterial cell wall hypothesis is the most likely explanation.

Bacterial cell wall hypothesis

The unproved theory of the bacterial cell wall implies that the microkeratome or irrigating cannula becomes contaminated from the surgical environment by bacteria. Gram-negative bacteria are the culprit with endotoxin-related DLK. Gram-positive bacteria do not liberate endotoxin on their “death.” Gram-negative bacteria can grow on improperly cleaned instruments prior to sterilization or more likely from colonization of the autoclave reservoir or internal tubing. When instruments are cleaned and left overnight, according to Dr. Lawless, bacteria can feed upon the trace proteins, and the bacterial count multiplies. Even when the instruments are autoclaved the next morning prior to surgery, the autoclave process sterilizes the bacteria. But the bacteria will then release heat-stable lipopolysaccharides (LPS). The LPS are not detoxified by standard autoclaving at 137°C for 10 minutes. Thermal detoxification requires approximately 250°C for 2 hours. LPS can cause an inflammatory reaction that will digest the stromal collagen in a relatively short time, researchers say. LPS is linked to Lipid A, which is the toxic part of LPS that causes an inflammatory reaction that will digest collagen very quickly, Dr. Lawless said.

Bacteria management strategies

“There are many agents that will bind LPS,” Dr. Lawless said. “The only one that’s available to us, as ophthalmologists, to use topically is polymyxin.” Polymyxin is a polypeptide antibiotic that will bind and neutralize most biological activities of Lipid A. According to Dr. Lawless, polymyxin has not been used for the treatment of DLK and it will not be useful for subsequent problems from DLK, but may be useful in the early stages.

John F. Doane, MD, Eric J. Linebarger, MD, and Dr. Probst have created their own sterilization protocols or techniques for managing outbreaks of DLK (See accompanying article).

Dr. Doane suggested that once DLK is encountered, all surgical instruments and the autoclave involved need to be tested for endotoxin levels. Dr. Doane stated it is more important to distinguish isolated cases of DLK (likely due to any one particular etiology) from multiple cases on a single day (an “outbreak”) that is likely due to gram-negative endotoxin from autoclave colonization with gram-negative bacteria. Dr. Probst has used Dr. Doane’s protocol since January 1999. Dr. Probst also uses disposable instrumentation when possible and instructs all patients to use topical corticosteroid drops every hour on the second postoperative day to suppress any potential cases of DLK. However, regardless of this thorough attention to detail, DLK occurs in Dr. Probst’s hands in 1 in 500 cases. This suggests that further measures are needed, and that the bacterial cell wall theory is possibly not the only explanation for the outbreaks.

Most patients respond to rigorous topical corticosteroids if DLK is identified early, within 24 to 48 hours after surgery. According to Dr. Lawless, it has been found that just a 12-hour gap can make a difference. The sooner the treatment starts, the better, but some patients’ DLK will progress regardless of how intense the topical corticosteroids are and how early treatment is initiated. Other patients will not show signs of DLK until the third or fourth postoperative day. Clumping and aggregation of the infiltrate centrally and a hyperopic shift is a warning sign.

Prompt relifting of the flap at the first sign of central cellular migration, with cleaning and irrigation of the flap bed, can successfully manage most of the severe cases. Dr. Lawless said the key is to recognize the problem early and advocated the frequent and intensive use of topical corticosteroids to prevent progression to the severe grades of DLK. Lifting the flap prior to stromal edema and secondary necrosis is important. And lastly, those patients who do not present until after the first 24 or 48 hours are most likely going to require lifting and irrigation after a very short course of intense topical corticosteroids.

Dr. Lawless recommends, when an outbreak of DLK occurs, to do a limulus anti-LPS assay to test for endotoxin. The limulus anti-LPS is an anticoagulant that inhibits the portion of the horseshoe crab coagulation system.

Increased awareness of DLK has led to early recognition of the condition. According to Dr. Lawless, one should always have a high suspicion of bacterial infection after LASIK. This way early cases can be identified before they become symptomatic. Dr. Lawless said it is not necessary to treat every patient with hourly topical corticosteroids, as you would be overtreating the majority of LASIK patients.

For Your Information:

Michael A. Lawless, MD, can be reached at The Eye Institute, 270 Victoria Ave., Level 3, Chatswood, NSW 2067, Australia; +(61) 8-8922-8146; fax: +(61) 2-9410-3000; e-mail: mlawless@theeyeinstitute.com.au. Dr. Lawless has no direct financial interest in any of the products mentioned in this article, nor is he a paid consultant for any companies mentioned.