For AMD injections, prevent endophthalmitis with povidone-iodine — and not much else

With the growth in injection treatments for age-related macular degeneration, there is more attention being paid to the rare consequence of endophthalmitis. But there is no consensus on what effectively prevents it.

Ophthalmologists compare the pros and cons of masks vs. drapes vs. no talking rules, gloves or lid speculums, and the merits of antibiotics or displacing the conjunctiva. All the variations “tell you that none of us knows what we’re doing,” George A. Williams, MD, an OSN Retina/Vitreous Board Member, said.

Endophthalmitis is rare and hard to study on a scale large enough to draw definitive conclusions about what prevents it. There are only two things ophthalmologists know for certain: Povidone-iodine works, and nothing will ever be settled by a randomized clinical trial.

George A. Williams, MD, said that he assumes any post-injection inflammation is infectious. Image: Peter Roberts

“If you have an event that has an incidence of one in 1,000, and you want to show a 50% reduction and have a 90% power of doing that, you would need over 35,000 patients to be randomized. Obviously, that is unlikely to happen,” Dr. Williams said.

Areas of consensus

Colin McCannel, MD, an associate professor at Jules Stein Eye Institute in Los Angeles, said he refers to a consensus paper published in 2004 that weighed evidence and drew some conclusions about best practices for injections for AMD. The areas of strong agreement were:

• Use povidone-iodine on the ocular surface, eyelids and eyelashes.
• Use a lid speculum to avoid contaminating the needle with eyelashes or eyelid margin.
• Avoid extensive massage of the eyelids to avoid expression of the meibomian glands.
• Avoid injecting patients who have active eyelid or conjunctival infections.
• Dilate the pupil.
• Use adequate anesthetic.
• Avoid prophylactic paracentesis.

There were also areas of no clear consensus:

• To flush or not to flush the povidone-iodine?
• To drape or not to drape?
• To don or not to don gloves?
• To give or not to give either preoperative or postoperative antibiotics?
• To check or not to check IOP after injection?
• To trust or not to trust patients to competently self-report endophthalmitis?

Procedural protocols vary

According to Dr. McCannel, peer-reviewed literature demonstrates the effectiveness of sterilizing the ocular surface, but nothing is proven to reduce the incidence of endophthalmitis.

Colin McCannel

As with every aspect of injections, physicians use povidone-iodine differently, with the common consensus being that the bactericidal agent should get to the conjunctival fornix and remain in the eye for a few minutes. But some physicians cover just the conjunctiva, lids and lashes with povidone-iodine, while others scrub vigorously or extend the application out to the eyebrows.

The barriers — masks, gloves and drapes — come next. Physician preference seems to be the driving factor when choosing what types of barriers to use.

Because endophthalmitis is linked to oral flora — estimates suggest about one-fourth of cases — and because the majority of the remaining cases are thought to stem from contamination by the patient’s own lashes, some ophthalmologists use drapes, some use masks, and some find both of those cumbersome and simply use the “silent rule” of no talking in the minutes leading up to an injection.

Dr. McCannel has found a body of literature in other medical specialties that looks at strep infection and complications of procedures such as dural puncture, ie, spinal taps and anesthesia. Molecular analyses have linked the cause of infections to the oral flora of the proceduralist in some cases, which is likely the cause in Streptococcus endophthalmitis after intravitreal injections. The Centers for Disease Control and Prevention has issued warning to use masks during dural puncture procedures, which are longer and more invasive procedures, Dr. McCannel said.

“The data suggest that if one just doesn’t talk, that should be good enough to prevent strep endophthalmitis,” Dr. McCannel said. “But the high road might be to wear a mask.”

Bladed lid speculums are often used to keep the lashes out of the way, according to Chirag P. Shah, MD, MPH, a retinal surgeon at Ophthalmic Consultants of Boston. His presentation of nearly 28,000 eyes at the 2010 annual meeting of the American Society of Retina Specialists in Vancouver evaluated variables such as lid speculum use, superior vs. inferior hemisphere of injection, and conjunctival displacement with a cotton tip applicator. None of these techniques affected endophthalmitis rates. Dr. Shah cautioned that this study is underpowered to detect small differences in risk.

As for the needle, Dr. Shah uncaps the needle only in the moment before injection to minimize risk of contamination from the air. Further, he and the patient do not talk during the injection procedure.

It is about creating an aseptic workspace, most physicians agree, not a sterile one.

Contraindications

According to Andrew A. Moshfeghi, MD, MBA, medical director at Bascom Palmer Eye Institute, other risk factors are the usual ones associated with endophthalmitis in general, including an immunocompromised state, for example a kidney transplant recipient who takes immunosuppressive anti-rejection medication or a patient with diabetes whose risk for infection development is greater than for otherwise healthy patients. Although these are not absolute contraindications, they may involve slightly higher risk in general for certain types of microorganisms.

Blepharitis is a manageable risk factor for injections for AMD.

“Pretty much everyone has some stage of blepharitis after a certain age,” Dr. Moshfeghi said. The dandruff-like debris on eyelashes is a “playground for bacteria,” he said, but it can be managed daily and then vigorously cleaned the day of the injection.

Dr. Shah said it is important that patients with blepharitis are managed chronically, often with warm compresses, meibomian gland massage, lid scrubs and possibly oral doxycycline.

“Everyone has a little bit of blepharitis. Even if they have a mild amount, I think it is important to prep the eyelids with povidone-iodine before injecting the anti-VEGF agent,” he said.

A “soft” contraindication is when a patient intends to sit in a hot tub or swim without goggles after the procedure, Dr. Moshfeghi noted.

Clinical recognition

Immediate care is needed to prevent endophthalmitis from progressing, starting with education and letting patients know what early signs to look for. As a retinal specialist, Dr. McCannel said he sees patients who did not make that early call to their physician.

Dr. McCannel said he gives a thorough explanation of the warning signs of endophthalmitis to watch for after each patient’s first injection and then rigorously reminds the patient after each subsequent injection.

Blurry vision is fairly common, but it can be a harbinger of endophthalmitis. When the central vision is blurry, that may relate to AMD or the disease under treatment. If vision is uniformly blurry, including the peripheral vision, “I start to get very concerned,” Dr. McCannel said.

Typically, endophthalmitis manifests within the first 24 to 48 hours after the injection as pain, redness and decreased vision, Dr. Moshfeghi said. Upon examination, the physician may find signs of anterior chamber cell and flare, with or without hypopyon, anterior chamber fibrin and vitreous inflammation.

These patients need immediate access to the physician, and Dr. Williams said he trains his staff about this. The key to diagnosis is a high level of clinical suspicion, he added.

“You want to be certain, if there is an infection, you catch it early,” he said. “[These patients] typically can do well, but like all endophthalmitis cases, there is a wide variability in the presentation; you cannot assume that the classic findings of hypopyon, fibrin and pain are going to be present.”

Dr. Williams said that he assumes any post-injection inflammation is infectious.

“It is impossible to be absolutely certain whether or not post-injection inflammation is infectious. Since the downside of delaying treatment for an infection may be disastrous, I treat any significant post-injection inflammation with intravitreal antibiotics. I am willing to over-treat an occasional eye to avoid missing an early infection,” he said.

Dr. McCannel added, “I err on the side of diagnosing endophthalmitis and treating it. The treatment is very well tolerated. Endophthalmitis is not very well tolerated at all. The bar for making a diagnosis and proceeding with treatment must be set very low. When in doubt, proceed with endophthalmitis treatment to ensure the best possible outcome. Even if you are wrong and it is not endophthalmitis, the treatment is well-tolerated.”

Dr. Shah’s research found that one cannot differentiate clinically between culture-positive and culture-negative endophthalmitis; thus, all cases presenting with signs and symptoms of endophthalmitis should be treated emergently. He outlined a typical course of treatment. He immediately performs a vitreous tap and sends the specimen to the lab for gram stain and cultures. The tap is followed by intravitreal antibiotics, usually vancomycin and ceftazidime. He prescribes topical antibiotics every hour, either fortified vancomycin and tobramycin or a fourth-generation fluoroquinolone. He typically adds topical steroids and cycloplegic drops. For severe cases or for cases that worsen after tap and inject, Dr. Shah recommends vitrectomy with intravitreal antibiotics.

William F. Mieler

“The recovery does take some time. Patients usually start turning the corner during the first 1 or 2 weeks,” he said. His research found that three-fourths of patients with presumed infectious endophthalmitis regained their pre-injection vision within 3 to 6 months.

For the one-fourth of patients who do poorly, they are likely to have either a retinal detachment or optic nerve involvement with a pale, ischemic nerve. These cases are usually due to particularly aggressive bacteria.

“It is a reasonable expectation to regain vision,” William F. Mieler, MD, an OSN Retina/Vitreous Board Member, said. “With other more aggressive organisms like Enterococcus, or certainly Pseudomonas, the chance of losing vision is very real. A lot of times you can tell by the course how effective treatment is going to be. If the patient comes in with a mild inflammation a week or two since the injection, compared with someone who comes in 2 days later and the eye is inflamed, you can guess at what type of organism they have.”

Treatment

If the patient does not present promptly to the ophthalmologist or if the bacterium that causes the endophthalmitis is one of the more severe strains, the condition can deteriorate quickly.

Dr. McCannel said he relies on the Endophthalmitis Vitrectomy Study as a guideline. When the vision upon presentation approaches hand motions or better, he sticks with a tap-and-inject treatment and gets drugs from the pharmacy to start within the hour. If it is worse than hand motions, he will proceed straight to vitrectomy.

“I will often err on the side of vitrectomy in these patients because I know the spectrum of organisms might be a little bit worse,” he said. “Also, if it means waiting 10 hours to get an operating room slot because everything is booked, then I will do a tap and inject in the office and follow it with a vitrectomy that evening or the next morning. Any delay in treatment is not to the eye’s advantage.”

The spectrum of organisms plays a large role in the outcome of restoring vision. It is best to know what one is dealing with, both to help the patient understand the severity of the course and to know how aggressively to treat, Dr. McCannel said.

“There are no good bugs in endophthalmitis, and there are certainly aggressive ones,” he said.

Between 60% and 70% of all endophthalmitis cases are caused by Staphylococcus epidermidis. The major source is the patient’s own skin flora or blepharitis, Dr. McCannel added.

Dr. Williams cited a published series of nine eyes with endophthalmitis, of which five cases stemmed from S. epidermidis.

“The source of the bacteria is almost always the patient’s own flora,” he said. “We would expect these infections to reflect the patient’s flora, which is predominantly gram positive.”

Dr. Mieler said, “No matter how good a person is, no matter how thorough you are with cleansing, most people are going to encounter an infection at some point in time in their career. It is not very often, and in most cases it is treatable. Most patients have a chance of recovering quite well.” – by Ryan DuBosar

Is there a role for pre- and post-procedural antibiotics for AMD injections?

References:

• Aiello LP, Brucker AJ, Chang S, et al. Evolving guidelines for intravitreous injections. Retina. 2004;24(5):S3-S19.
• Bhavsar AR, Googe JM Jr, Stockdale CR, et al; Diabetic Retinopathy Clinical Research Network. Risk of endophthalmitis after intravitreal drug injection when topical antibiotics are not required: the diabetic retinopathy clinical research network laser-ranibizumab-triamcinolone clinical trials. Arch Ophthalmol. 2009;127(12):1581-1583.
• McLure HA, Talboys CA, Yentis SM, Azadian BS. Surgical face masks and downward dispersal of bacteria. Anaesthesia. 1998;53(7):624-6.
• Mezad-Koursh D, Goldstein M, Heilwail G, Zayit-Soudry S, Loewenstein A, Barak A. Clinical characteristics of endophthalmitis after an injection of intravitreal antivascular endothelial growth factor. Retina. 2010;30(7):1051-1057.
• Trautmann M, Lepper PM, Schmitz FJ. Three cases of bacterial meningitis after spinal and epidural anesthesia. Eur J Clin Microbiol Infect Dis. 2002;21(1):43-45.

• Colin A. McCannel, MD, can be reached at Department of Ophthalmology, Jules Stein Eye Institute/UCLA, 100 Stein Plaza, Los Angeles, CA 90095; 310-794-9922; e-mail: cmccannel@jsei.ucla.edu.
• William F. Mieler, MD, can be reached at University of Illinois at Chicago, Department of Ophthalmology and Visual Sciences, 1855 W. Taylor St. M/C 648, Chicago, IL 60612; 773-702-3838; fax: 773-702-8094; e-mail: wmieler@uic.edu.
• Andrew A. Moshfeghi, MD, MBA, can be reached at Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, 7101 Fairway Drive, Palm Beach Gardens, FL 33418; 561-355-8608; fax: 561-355-8601; e-mail: amoshfeghi@med.miami.edu.
• Chirag P. Shah, MD, MPH, can be reached at Ophthalmic Consultants of Boston, 50 Staniford St., Suite 600, Boston, MA 02114; 617-367-4800; e-mail: cpshah@eyeboston.com.
• George A. Williams, MD, can be reached at Associated Retinal Consultants, 3535 W. 13 Mile Road #344, Royal Oak, MI 48073-6769; 248-288-2280; e-mail: gwilliams@beaumont.edu.
• Disclosures: No products or companies are mentioned that would require financial disclosure.