Partner Agencies

Partnering Agencies21_img.gif19_img.jpg44_img.jpg13_img.gif2_img.jpg83.jpg27_img.jpg31_img.jpg73.jpg63.jpg1_img.png76.jpg50_img.jpg82.png3_img.jpg46_img.jpg15_img.jpg8_img.jpg47_img.gif54.png1_img.gif20_img.gif55.jpg48_img.jpg86.jpg72.jpg49_img.jpg43_img.gif12_img.jpg29_img.gif77.jpg56.jpg81.jpg25_img.jpg65.png32_img.jpg22_img.jpg79.jpg50.jpg52.jpg45_img.jpg37_img.jpg6_img.gif17_img.jpg9_img.jpg16_img.jpg

Treatment For Severe, Advanced, Deep, Large Corneal Ulcers

   I recently wrote this up for the African Vision website and thought I would share with my Global Alliance colleagues. I have never seen this type of severe advanced corneal abscess in the States but they are unfortunately common in the developing world, especially in low-income countries. If you are briefly visiting a developing world country and see one of these dense large corneal ulcers, no matter what else you do,  would strongly consider also doing a Gundersun flap. Some of these recommendations might not be appropriate for North America.

       Unfortunately, we all see patients who present quite late ( > 4-6 weeks ) with a marked, large, dense, deep corneal abscess / ulcer. Many of these patients give a history of some thing going in the eye ( trauma ) --- either some vegetation ( agricultural ) such as a sugar cane leaf or an insect. Often they have self treated themselves with an antibiotic - steroid combination ( chloramphencol / dexamethasone ), another topical antibiotic / agent, an oral agent ( antibiotic ), or perhaps traditional medicine.
     On initial presentation, there is a large corneal abscess/ ulcer involving 70-80% of the corneal area or more. Often the stromal inflammation is so dense, you can not see ( evaluate ) the pupil ( anterior segment ) or tell if there is an hypopyon. Usually there is a large epithelial defect. Some times there is already a descemetocele. In these severe desperate corneal situations, where often it is not possible to obtain a Gram stain or cultures, I have come up with a treatment regiment which I think has saved a few eyes. I assume the ulcer may well not be bacteria but fungus, etc. if it has been cooking for several weeks already without improvement on various antibiotic agent.
     First I apply several times both proparacaine drops ( or tetracaine or amethocaine ) and 5 -10 % provodone-iodine solution. Then I carefully, gently debridge any dead necrotic corneal tissue. Be careful around the descemetocele. Then I soak two Q-tips ( or cotton buds or spear-shaped small sponges ) with povodone-iodine 10% solution. Get the Q-tips really wet ( dripping ). I gently rub the Q-tip onto the entire corneal ulcer for two minutes. If there is a epithelial defect, that's ok --- maybe better. Then I do the same thing with the second cotton bud ( Q-tip) for two more minutes.Really wet. I put in cyclopentolate 1% several times. I usually do not use subconjunctival or peribulbar injections but that is an option.
    Then I prepare what I will give the patient to take home:
      #1. I make up a 50-50  ( half / half ) solution of antibiotic and anesthetic ( ex. chloramphenicol and proparacaine ). Or you can use some other antibiotic such as a fluoroquinolone, Polytrim, etc. or another anesthetic agent such as tetracaine or amethocaine. Chloramphenicol drops is a good antibiotic --- readily available / cheap, less toxic to corneal epithelium, broad spectrum, and little resistance.
      #2.Then put about 10 ml of 10% povodone-iodine in an eye drop bottle for the patient. This is not the time to be timid, use the 10%. I save some of my used up ( empty ) bottles just as in this case when I need to prepare special topical combinations, etc.
       #3. I give the patient Natamycin 5% drops if available ( suspension, shake well ). Amphotericin B  0.25% is rarely available and is irritating.
       #4. Anti-fungal skin or vaginal cream ( fluconazole, itraconazole, miconazole, flucytosine, etc.) can be used after the drops every one hour while awake. It will often say right on the ointment "not to be placed in the eye ". It's ok, you can put in it the eye.  Actually the problem is the pH is wrong so by using the anesthetic drop first, you can use the skin / vaginal preparations in the eye, every one hour while awake.The numbing drops make it so they can tolerate the skin / vaginal cream.
       #5. Oral ketoconazole 200 mg bid with Coke or Pepsi as the absorption is much better with an acidic beverage. Ketoconazole is usually readily available in most hospital formularies but there are other oral anti-fungal agents available. Whatever you have.
       #6. Cyclopentolate 1% bid
       #7. I do not use steroids initially
       #8. Doxycycline 100 mg bid with food ( Do not take with milk / dairy products ). Anti- inflammatory properties and also good for Gram + organisms.     
      So first the patient uses the antibiotic / anesthetic drop, then the provodone - iodine 10% drop, then any other anti-fungal drop you have or can make up, and finally the ointment. The reason I use the numbing drop is all of these drops / ointment can be irritating and the patient quickly reduces his usage. Instead of 12-15 times daily, he's quickly only using them only 4-6 times daily. Often after a few days the patient may refuse most of the drops.
     I only use this regiment in severe, desperate corneal ulcers when you are fighting to save the eye --- "when the wheel has come off".  I do not recommend this regiment in small early ulcers.
    The patient should be told that the treatment will probably take several months not weeks. The patient will obviously have a permanent loss in vision and should be told that at the initial visit. A Gundersun conjunctival flap may be needed eventually. Usually you see no clinical improvement for at least a 7-10 days. For the first week or two, often the inflammation may appear worst as you eliminate ( kill ) any infectious agents. Some times the patient will say the eye is less painful or they are seeing slightly better before you can appreciate clinically any improvement. Fungal keratitis can proceed to endophthalmitis as the fungus can often readily pass through Descemets membrane.
      I have had some success with this regiment and hope it may be of use to you. I would suggest a conjunctival flap ( Gundersun ) in patients who are at extreme risk of perforation. Sooner is often better than later.

 Baxter McLendon MD

Search by continent

GlobalSight Newsletters

Signup to receive our newsletters.
Cron Job Starts