Infiltrative keratitis: Experts offer pearls on differential diagnosis, management

Michael D. DePaolis, OD: What are some of the things you look for in a contact lens wearer with a diffuse infiltrative keratitis? How do you determine whether this is contact lens-related, infectious or something else?

David B. Seibel, OD: First, if a patient is still wearing the lenses with bulbar redness and accompanying keratitis, check lens fit including movement to look for peripheral lens alignment around the limbal area — specifically if there is a tight fit, any conjunctival drag or any impediment of vessels in the capillary bed that may lead you to suspect that some corneal waste products and other debris are being retained underneath the contact lens. These types of clinical signs can indicate whether or not you have a contact lens-related etiology.

Dr. DePaolis: So the condition of the lens either increases or reduces your suspicion concerning whether it is lens-related or not?

Joseph P. Shovlin, OD: Whether from a contact lens or a microbial vector, it often looks similar. We also look for any adenopathy in the preauricular and/or submandibular area. By the way, that's not pathognomonic for specifically epidemic keratoconjunctivitis nor adenoviral disease. In general, the natural history of various diseases should be considered as well.

Robert E. Prouty, OD: I agree. General optometrists see acute red eyes but forget to check the preauricular nodes, which is easy to do after taking the history. It gives them a nice sign in the road that says look for a virus here.

Dr. DePaolis: Nonmicrobial lens-related conditions, toxins and hypersensitivity reactions generally don't initiate adenopathy. In most of your early adenoviral keratoconjunctivitis patients is the adenopathy palpable, tender for the patient or both? Does it progess?

Dr. Prouty: I think it is progressive. I use the following guidelines: a week to develop, a week to get worse and a week to get better. Adenoviral patients will present when the second eye presents, not the first one, and that is right at the end of the first week or the beginning of the second. If they present at that time, lymphadenopathy, like any inflammatory response, starts with me being able to palpate without tenderness. That is 1+. Some tenderness is 2+. If it is palpable and they are uncomfortable, 3+. With 4+, I've had patients leave the chair. They might not get to the tender phase, but most adenoviral infections do. The chlamydial patients are rarely tender.

Dr. Seibel: When palpating the preauricular nodes, one thing that helped me was knowing that it feels similar to a hard pea or little bean, and that's one of the signs you pick up before the patient reports tenderness. In those cases you have to make the diagnosis because their ocular symptomatology seems to be more severe than their adenopathy symptomatology.

Dr. DePaolis: Say a contact lens wearer has a diffuse infiltrative keratitis, and the second eye is involved. Palpating the nodes reveals nothing. If you think this is a lens-related entity and not infectious, do you treat those eyes? Do you remove the lenses or do you prescribe a therapeutic and, if so, what?

Dr. Prouty: Do you tickle their placebo effect, basically? Personally, if there is not an infectious etiology, I don't treat. You need thorough patient education. The treatment here is to fix that lens. I teach students that good doctors treat etiologies. General optometric practitioners, out of comfort, throw in an antibiotic, and they frequently can complicate the problem by developing medicamentosa or diffuse keratitis that exacerbates due to cytological changes from the antibiotic. The answer is: take them off everything. Put them on a nonpreserved artificial tear and stay away from the lens and everything else so you can sleuth out what is going on.

Dr. Shovlin: We take a slightly different approach. The first thing is to remove the lens. We often treat patients who have even mild to moderate symptomatology. I think they get better a little faster with anti-inflammatory agents. With that said, though, I can rarely be faulted for not adding steroid. If there is any epithelial defect we will use something bland for antibiotic coverage. However, if there is significant corneal presentation where there is infiltrative keratitis we will use either a combination drug — if there is epithelial involvement — or just straight Flarex [fluorometholone, Alcon] or FML [fluorometholone acetate, Allergan].

Dr. Prouty: What's the etiology? What are you treating?

Dr. Shovlin: We are treating patient symptomatology, the red eye response in particular. I think we are saying the same thing: if you get a florid response, you are going to treat them. It is the mild to moderate presentation you have to labor with. We shift toward treatment on those patients. I think they get better faster and they go back to the contact lens faster if they use a steroid. The inflammatory response is mediated by an array of chemical mediators and cellular events. The use of steroids inhibits the synthesis of these proinflammatory mediators.

Dr. DePaolis: Recent literature suggests that this nebulous entity we call acute red eye may be mediated by certain strains of low virulence, gram-negative organisms. Would you be justified in saying, "I'll throw an antibiotic on there to eradicate low virulence organisms and a steroid to clear the eye up?"

Dr. Prouty: I think we want to be cautious about saying: "If it is red put an antibiotic on that." We have done such an adequate job as educators to scare optometric practitioners away from using a steroid on a red eye that now they fail to go to it when they need to. Again, you need to figure out the etiology and if it is infected, then use antibiotics. If it is infiltrative, edema, hypoxia or contact lens related, steroids are the choice. If we have something cytotoxic, then I get away from everything and wash them with artificial tears.

Dr. Seibel: In addition to clinical etiology, I ask what patient behavior could produce that clinical result. I throw that factor into my prescribing decisions. If I have a patient who I sense is noncompliant, I am less in a hurry to make that patient feel comfortable with more aggressive medications such as corticosteroids. I feel I owe them an understanding of the consequences of noncompliance. For those patients, I may use artificial tears or saline rinses to let them know the eye needs time to recuperate or rehabilitate. Then they understand they have temporarily damaged the eye, and we can then begin to discuss the importance of compliance when they return for follow-up care.

However, I usually have very compliant patients. Then, part of our job is to make the patient feel comfortable. I'm not comfortable applying steroids to a broken epithelium because I feel one of the ocular defense barriers has been broken down. However, if I don't have that I will prescribe a combination medication for patients who need that comfort.

Dr. Prouty: Traditionally, optometrists feel if all else fails, initiate combination therapy. Sure, we don't have to worry about Pseudomonas or the adherent contact lens, but what do antibiotics do for infiltrative keratitis? They may just act as a placebo or simply dilute the toxin, which we do with an artificial tear.

Dr. DePaolis: Say you have deemed that the condition is lens related, not infectious. How do you tease out whether it's hypoxia, mechanical trauma, hypersensitivity or toxicity issues? Do you change everything — refit the lens, change wearing time, change care habits — or is there a more disciplined way?

Dr. Shovlin: Always start with a careful history. The patient may admit to sleeping with lenses and other abusive lens-wearing patterns, but it is difficult to find clinical indicators that suggest one etiology over another. Many times, it is a multifactorial causal relationship. We will often take the shotgun approach. If they have been wearing the lenses too long, we get the lenses off to get more oxygen to the eye. When patients have changed their solutions, you address that. Of course, if it is your own patient, you should be comfortable with the lens fit. But if the patient is from outside the practice, fit is something else to consider. Significant microcysts suggest hypoxia from overwear, but the inflammatory cascade can be from many different factors.

Dr. Seibel: I break the therapy into stages. Priority one is to get the eye quiet, white and healthy again. Priority two is to find out what caused this so we can make a change so this doesn't happen again. My first step in multifactorial situations is to weigh toxicity, infectious, hypersensitivity and autoimmune factors. I discontinue lens wear to make sure the oxygen is there. I go to saline rinses several times a day to flush out and dilute allergens or toxins in the tear film. I see those patients back in a few days to see how they are doing. If there is still inflammation, perhaps I can clean it up with a more assertive drop: a combination drop or a mild steroid.

Dr. Prouty: Starting with education is critical for these patients. However, we can no longer use saline washes with total confidence. This is a cluster of conditions called ocular surface disease, from dry eye to settled exotoxins to contact lens adherence issues.

Dr. DePaolis: Once you've convinced these patients that it's a lens-related problem and all they need is palliative therapy, how long do you keep them out of lenses?

Dr. Prouty: I'm in a referral setting, so if I tell them to stay out of the lenses for 3 weeks, they don't argue.

Dr. Seibel: Unfortunately, it's not a black-and-white decision process. When I first started practicing, I said a contact lens can't go on eyes as long as I see an infiltrate, and I lost a few patients. Now, I take into consideration patient lifestyle. I bend when there is one small infiltrate I can keep watch on, but I want them back in the office sooner to make sure my decision in letting them do this is not sending them down the wrong path.

Dr. Prouty: What about switching to disposable lenses where you can use an antibiotic over the top instead?

Dr. Seibel: I've also prescribed 1-day lenses where the patients could throw them away after their function, where contact lens wear was a must. It becomes a more difficult call in patients who are tied to their contact lenses for their vision, such as in keratoconic or aphakic patients. These are difficult situations at best, and we may have to occasionally bend some of the clinical rules to be more compassionate to the personal needs as well as the ocular needs of our patients.

Dr. Shovlin: It is a tough situation, but sometimes you have to be forceful. I think how long we keep patients out of lenses depends on the response. Patients with a flour-like dusting across the cornea or with deep epithelial involvement may be out of lenses for up to 3 months. But some clear in a few hours. A rough rule of thumb: If you have a red eye with infiltrates, that patient deserves a minimum of a 10-day to 2-week hiatus with appropriate treatment. It goes back to the red eye-white eye analogy: if we have infiltrates that form opacities that persist for a year or longer after adenoviral infection, yet the eye is white with no epithelial involvement, that patient can wear a contact lens, at least on a reduced basis. But if that eye is red, that lens must stay off.

Dr. DePaolis: A lot of these can be due to a subclinical ocular surface disease that needed the right set of circumstances to bring it to fruition. If you do not clean up certain things, are they destined to fail time and time again? What protocol is there for addressing these needs before the patient goes back into lenses?

Dr. Prouty: I see many situations where contact lens experts have done everything they can, and they still throw their hands up. At least in high, dry air in Colorado, many problems are dry eye-related with exotoxin response. Personally, I think human tears are better than any artificial tear on the market, and if we can increase the supply, we can dilute toxins and reduce the responses we got from low-grade bacterial and viral presentations.

For me, we need to clear dry eye and exotoxin responses first, and if we rule out disease components first, our manipulation of contact lens parameters is more successful. If we play with the lenses first and the eye is sick to begin with, we can just exacerbate the problem.

Dr. DePaolis: What's your intuitive approach: lid hygiene, tear supplementation, punctal occlusion?

Dr. Prouty: I do a Schirmer's test even if they have no indication of dry eye. It seems to be the best test for managed care documentation. The biggest failure with Schirmer's is improper administration: Use a drop of proparacaine, wait 5 minutes, dab away the excess, put the strip in and wait 5 minutes with the eyes closed. Then, I do it again. We believe the low finding: You can turn tears on but you can't turn them off. If they're dry, you know that you are dealing with the contact lens parameters.

Now look at the lid components. If meibomianitis is present to a minor degree — which you consider normal in someone else's eye — with a dry eye, you may have a symptomatic patient. I'm not a big expresser, but I do think a warm compress with a downward pull and massage of the lid can be useful. If not, then I do collagen implants to improve tear flow. If their tear film works, I suggest plugging the inferior puncta so we don't have the cesspool stagnation of contact lenses, and then go to work on that lid to see if you can clean up the pathology, including doxycycline or tetracycline. Also, we maximize endogenous tears with new enzymes. Then it's up to the contact lens fitters.

Dr. DePaolis: In considering contact lenses, do you find that you clean up most of these problems by changing solutions, going to higher oxygen materials, going to looser fit, going to more frequent replacement or all of the above?

Dr. Seibel: I try not to overreact unless it is a compelling clinical presentation, such as corneal ulceration. If I have years of solid, quiet, white eye contact lens wear and all of a sudden a patient presents with a mild problem, I try to remember that we were doing something right for a long time and we shouldn't abandon all the successful components of that fit. Once the acute clinical complication is resolved, if the patient's lenses are fresh, fit well and the patient's compliance is good, I may just follow-up with the patient more frequently for a while. Have him or her come back in 3 months instead of 6. Of course, patient education is important to make sure they are compliant: using the right solution, a clean case, etc. However, if a complication presents more than once, then you have to look at what may be broken. Sometimes, I need to change the fit; sometimes I try a different solution.

My greatest frustration in dealing with ocular surface disease is can I rely on patients to put on a warm soak for 5 minutes, take it off, squeeze out the meibomian gland and come back? In the short term, yes. But when I see these patients 6 months later, often we are back where we began because the regimen is too tedious and time-consuming for patients to follow every day. I think many of us don't consider the long-term effectiveness of these regimens.

I invite the development of new techniques and treatment regimens for chronic ocular surface disease because I don't think what we have is completely effective. Many doctors don't recommend meibomian expression and artificial tear substitutes because patients often discontinue their use.

Dr. DePaolis: So you do not make radical changes in a patient who has a good record of wearing a daily wear soft lens?

Dr. Seibel: Yes, as long as the degree of ocular involvement is mild.

Dr. Prouty: We can be our own worst enemy.

Dr. DePaolis: Only if you have a recurrence would you then go into heroic efforts?

Dr. Seibel: If I run into ulceration, I'm more aggressive in getting the patient out of extended wear to look at the fit and increase the oxygen. However, I will tinker on the short term. I might have them start out slow with artificial tear supplements, rinsing the eyes with saline at the beginning or the end of the day and maybe putting them on a reduced wear schedule.

Dr. Shovlin: Always consider infiltrates as harbingers of serious tissue damage. Back to the white eye-red eye analogy, white eye infiltrates are often just messenger cells. A red eye in most cases is a phagocytic response to a microbial antigen or some other vector.

We do pretty much the same thing: we reduce wearing time significantly; if they have an expensive specialty lens, we at least talk about replacing that. We are not afraid of making a switch from reusable to disposable lenses or even a 1-day option.

When all else fails, you must be prepared to move that patient to a rigid lens or no lens at all. As far as screening these patients, if they have significant lid disease, they should not get a contact lens. You need to consider the potential benefits of a rigid contact lens, especially with repeated red eye responses. Educate them properly, and tell them in spite of our best efforts they may be back. Then, we have to move on to special lens treatments.

Dr. DePaolis: Don't statistics say once you have a contact lens acute red eye episode you are three to four times as likely as fellow contact lens wearers to have additional episodes?

Dr. Shovlin: I haven't seen any good controlled studies on it, but clinically speaking that is probably what you will encounter if you don't take the broad-minded approach. You have to be prepared to ferret these things out. Once you have geared up the immune system, the red eye response comes back faster.

Dr. DePaolis: The key is once they have an episode, we can't reverse that, but we can make it better. Patients needs to understand they are in a different risk pool. As you said, you don't want a knee-jerk reaction, but with the greater risk, you have to make more broad-sweeping changes with the lenses to eradicate it.

Dr. Shovlin: And to quote Dr. Linda Casser: "Lid scrubs for every contact lens patient."

Dr. DePaolis: What about the infiltrate you confuse with an infectious ulcer? Despite all the strides we have made, statistics indicate this has the same annual incidence as 10 years ago. How do you differentially diagnose them from a true ulcer? What causes the noninfectious entity?

Dr. Prouty: My favorite article is "Infected vs. sterile corneal infiltrates in contact lens wear" by Ray Stein Sr. [Am J Ophthalmol1988;105:632]. It is a great article, and the top indicators are pain, photophobia, anterior chamber reaction and size. With location, peripheral is sterile and central is infected, just by the nature of avascular cornea. The difference between infected and sterile infiltration is whether it stains or not. I then look at the size of the lesions and the lids.

Dr. Shovlin: That article is great. Pain and discharge are critical. Talk to the patient first, then look with the slit lamp, and you have signal data such as circumferential injection, anterior chamber reaction, size and location of the infiltrate and characteristic staining. However, we all have seen small, peripheral ulcers and early infections with minimal epithelial defect, so there is considerable danger in a cookbook approach. These are not all bacterial, and I've seen intact epithelium with Acanthamoeba and fungal keratitis, so you have to stay away from the laundry list. But if signal data say infection, believe infection and treat it as such: bring the patient back in 24 hours and treat him or her every 1 or 2 hours, even more aggressively if warranted — not four times daily — with a broad-spectrum antibiotic such as a fluoroquinolone. You will not be faulted for treating aggressively, but the opposite is true if you casually treat and it blows up on you.

Dr. Prouty: I agree completely. When these infiltrates are suspicious, I approach it the same way: aggressive antibiotics. I fault on the side of overtreating.

Dr. Seibel: Agreed. I also focus paralesionly, looking just off the side to see if I have additional edema, snow-flakish appearance or infiltration. This is subtle, but it might lead me toward the infectious diagnosis.

Dr. DePaolis: Does everyone agree to treat with fluoroquinolones?

Dr. Prouty: I do because I think we approach suspicious lesions like they are Pseudomonas ulcers until proven otherwise. If it is diffuse superficial punctate keratitis, then avoid fluoroquinolones. Don't use your big guns and risk continued development of resistance. Polytrim [trimethoprim sulfate, polymyxin B sulfate, Allergan] or other medications are fine.

Dr. Seibel: I think optometrists feel more comfortable treating ulcers with fluoroquinolones. They don't feel as if they have to go to fortified tobramycin, which I think scares some optometrists away from prescribing and treating, because working with pharmacists can be difficult. As Dr. Shovlin said, we should hit it hard — really dose these when you think they are infectious. I feel very comfortable seeing a patient every day until I have 100% confidence that it is under control. I can let him or her go for maybe a couple of days after that, but if you see your patient back every day there is not much that can wrong.

Dr. Shovlin: We should use a careful nomenclature to describe these things in our records. When we see aggregate white blood cells in the cornea, we should say it is an infiltrative keratitis presumed to be sterile based on signal data. When we use the term "ulcer" and don't treat that eye appropriately, we are in big trouble. Unfortunately, ulcer in the contact lens world is used synonymously with infection, although that is not accurate. Be concerned, because we all know there are sterile infiltrates, such as chemical burns. These are noninfectious ulcers, so we have to be careful about how we describe these things for medicolegal reasons.

Dr. Prouty: Yes, we want to term these things properly on our charts, even if the CPT code says marginal ulcer.

Dr. DePaolis: A contact lens patient comes in with a red, painful, photophobic eye and a 1½-mm infiltrate around 8 o'clock, 1 mm in from the limbus. It is not escavated but elevated, and there is a little staining and a slight underlying interior chamber reaction. The patient is sending mixed signals. You said we should treat aggressively with fluoroquinolones each hour and then follow up. When do you add steroids?

Dr. Prouty: After about 4 days if I'm confident that sterilization has occurred.

Dr. Shovlin: Your key response is a heaping or mounding effect of the epithelium, and I have done it as early as day 1 in situations like that. Again, you can never really be faulted for not using a steroid.

Dr. Seibel: Documentation is important as you assume more aggressive postures. I'm more conservative. I wait several days until I'm certain I have taken care of any infectious components.

Dr. Shovlin: That's key. Not all opalescent opacities of the peripheral cornea are infectious. My own insecurity is when I'm pretty sure that it's infectious, will I use just a fluoroquinolone? I think with early reading on resistant Staphylococcus and Streptococcus we would add bacitracin, but we use the simple fluoroquinolone approach if that infection is not over the visual axis and not deep and penetrating. That way, we know which medication to stop, because if you use two different medications you are not sure which one is working. The key is to avoid prolonged, casual use of fluoroquinolones to avoid resistance, but when you use them do so every 1 to 2 hours and have that patient back the next day. I think you can feel comfortable without using something with better gram-positive coverage if you use it that frequently.

Dr. DePaolis: With disposable lenses, have we solved a lot of the contact lens ills, yet also embraced molded lenses that don't have great mobility and tend to drape and have become so easy that they encourage noncompliance? Are these infiltrates related to lens solutions or physical fit, or is it to hard to say?

Dr. Shovlin: In many cases, it is multifactorial. I'm not one of those who believes that stagnation of the tear film or retrolens debris is chemotactive for neutrophils. However, oxygen levels have not increased with these materials, yet we prescribe many disposable contact lenses. Solutions play a role in certain patients, and getting away from multipurpose solutions is important even if it just means getting that debris off the lens more efficiently. We need to get oxygen to the eye, whether through different materials or reduced wearing time.

Dr. Seibel: I feel I am looking at a situation of oxygen combined with mechanical lens fit. I tend not to consider solutions as much any more. The materials of soft lenses are exactly the same materials we had for years, yet we haven't made progress on the oxygen supply to the eye. We have taken ourselves out of the hot seat to not be as responsible for the fit of the lens. As long as the patient is happy and the lens is profitable — with the pressures of managed care — the lens may decenter but we don't address some of those issues that cause the problems. It will be interesting in terms of future research. I know one of the questions that Dr. Brien Holden is working on at the Cooperative Research Centre for Eye Research and Technology at the University of New South Wales in Australia [Dr. Seibel: Is this correct (where Brien Holden is)?] is: If the cornea achieves all the oxygen it needs, how much complication will it have? As some of the hyper-Dk materials come to market, we will be able to answer some of these questions.

Dr. Shovlin: Dr. Dwight Cavanaugh's work will show that currently we do not have a soft lens with high enough oxygen flux to take care of the average patient. The key is that to avoid a shut-off apoptosis you need to have a significantly higher amount of oxygen than what is currently being provided.

For Your Information:

• David B. Seibel, OD, can be contacted at (314) 843-5700; fax: (314) 843-1353.
• Joseph P. Shovlin, OD, can be reached at 200 Mifflin Ave., Scranton, PA 18503; (717) 342-3145; fax: (717) 344-1309.
• Robert E. Prouty, OD, can be reached at Omni Eye Specialists, 55 Madison, Ste. 355, Denver, CO 80206; (303) 377-2020; fax: (303) 377-2022, Ext. 11.