CASE PRESENTATION
A 39-year-old Southwest Asian male presented for refractive surgery with a previous medical history of seasonal allergies, for which he occasionally took an antihistamine. He had a refraction of -4.50 D sphere OU and underwent uneventful LASIK with the Hansatome microkeratome (Bausch & Lomb, Rochester, NY) and the STAR S3 laser (VISX, Inc., Santa Clara, CA). The patient's postoperative medical regimen included Econopred Plus (Alcon Laboratories, Inc., Forth Worth, TX) and Ciloxan (Alcon Laboratories, Inc.) q.i.d., as well as preservative-free artificial tears q1h to q2h while awake.
On the first postoperative day, the patient presented with a UCVA of 20/40 OU. He was experiencing no pain, photophobia, or foreign body sensation. Examination revealed well-positioned flaps without epithelial defects or striae and 3+ diffuse lamellar keratitis (DLK) bilaterally. I relifted both flaps to treat the DLK and obtained a Gram stain and cultures as a precaution in the event of an infectious etiology. The patient began a regimen of Econopred Plus q1h while awake, Ciloxan q.i.d., and an oral tapering dosage of prednisone, beginning at 80 mg q.d. On the following day, he was free of pain, and his visual acuity was approximately 20/40 OU. The Gram stain revealed no organisms, and the degree of DLK had decreased significantly in both eyes. Applanation measured the patient's IOP at 11 mm Hg OD and 10 mm Hg OS. Confirmatory palpation revealed soft globes bilaterally. I increased the patient's dosage of Econopred Plus to every 30 minutes while awake.
On the third postoperative day, the patient's UCVA deteriorated to 20/400 OU. He still felt no pain, and cultures revealed no growth. His conjunctivae were relatively quiet, and there was no discharge. The corneal examination had changed dramatically, however. Although the diffuse white blood cells in the interface had disappeared, both eyes now had a very dense, confluent, central white opacity with an approximate diameter of 4 mm and intervening, clear “mud cracks.” Each opacity appeared to be at the level of the interface and extended into both the underlying stromal bed and the overlying flap. The periphery of the flaps and corneas was clear, and the anterior chambers exhibited trace cells. The patient's IOP was 5 mm Hg OU by applanation, but peripheral measurements with a tonopen revealed a pressure of 15 mm Hg bilaterally.
HOW WOULD YOU PROCEED?1. Would you admit the patient for IV steroid administration?
2. Relift the flaps a second time?
3. Observe the patient and instruct him to continue taking the topical and oral steroids?
4. Treat the patient for an atypical mycobacterial or fungal infection?
SURGICAL COURSE
The disparity between IOP measurements obtained with applanation versus a peripherally placed tonopen suggested that fluid had collected beneath the flaps. Although I could not rule out an unusual infection at this stage, this young, otherwise healthy male's early presentation argued against the presence of opportunistic organisms. Because of the patient's severe deterioration in visual acuity, I decided to relift his flaps a second time, and I removed milky fluid from beneath each one. I found the epithelium overlying the interface opacities to be extremely friable, and I placed bandage contact lenses in an attempt to prevent sloughing of the epithelium. In addition to continuing treatment with topical and oral steroids, the patient received oral doxycycline 100 mg b.i.d. for its anticollagenase activity. During the following several tense days, the patient's UCVA improved to approximately 20/50 OU. Although the central interface opacities greatly decreased in size and severity, bilateral areas of scarring remained at the level of the interface with scattered, intervening, clearer “mud cracks.” By the fourth postoperative week, the patient had discontinued steroids and achieved a BCVA of approximately 20/30 OU.
OUTCOME
I placed the patient in rigid gas permeable contact lenses approximately 5 weeks postoperatively, and he achieved a BCVA of 20/20 OU. During the next 6 months, his visual acuity steadily improved, and the interface scarring decreased significantly. Central striae remained in both eyes, however. The patient's UCVA improved to 20/40 OU. Manifest refractions of +2.00 -1.00 X 105 OD and +1.75 -1.25 X 107 OS yielded a BCVA of 20/25 OU. Topography demonstrated stromal melting consistent with the hyperopic refraction (Figure 1). The patient has thus far declined any further surgical intervention.
DISCUSSION
This complex case illustrates several points. The patient presented with classic DLK on the first postoperative day but actually developed a second problem on the third day. Graham A. Fraenkel, MD, and colleagues first described central interface opacities after LASIK in 1998.1 They described three patients who experienced relatively painless onsets of acute central stromal opacifications within the first postoperative week—after normal examinations on the first postoperative day. The investigators coined the term central focal interface opacity to describe the condition they observed. They speculated that its etiology might be a toxic substance released into the interface, but the causative factors remained unknown.
Robert K. Maloney, MD, of Los Angeles subsequently described the condition, which he referred to as central toxic keratopathy.2 He differentiated the phenomenon from DLK by noting that central toxic keratopathy is dense and focal, involves the underlying bed and/or overlying flap, does not respond to steroids, and always results in a hyperopic shift owing to tissue melting. He further noted that central toxic keratopathy can occur in the absence of a flap (eg, after PRK). Dr. Maloney also speculated that a toxic etiology might be responsible for the condition, but this theory is as yet undetermined. For that reason, the term central stromal opacity may be more appropriate, as it neither implies a toxic agent nor indicates that the opacity is confined to the interface.
Another fascinating aspect of this case was the disparity between the IOP measurements obtained with applanation versus a peripherally placed tonopen. Investigators have described cases of steroid-induced glaucoma that presented with falsely low applanation pressures due to a collection of fluid in the interface.3 This patient only took steroids for a relatively short period of time, and this case illustrates the point that post-LASIK IOP measured with applanation is suspect whenever there is a possibility of interface fluid. Ophthalmologists should consider taking peripheral IOP measurements with a tonopen in such situations.
Additionally, surgeons should consider central stromal opacities presenting in the postoperative period to be infectious until proven otherwise. In the early postoperative period, bacterial infections predominate, whereas opportunistic organisms such as atypical mycobacteria and fungi occur later, particularly in individuals on corticosteroid therapy. Ophthalmologists should be cautious in concluding that any condition they are following is noninfectious.
The condition after uneventful LASIK described herein is one of the most frightening for both the patient and surgeon. A severe loss of vision during the early stages is typical, and, in this case, the condition was present bilaterally in equal severity. Although there is no known treatment, affected patients will improve significantly if handled properly. The surgeon should resist prescribing unnecessary, long-term topical steroid therapy and must guide the patient through the difficult postoperative period.
Steven J. Dell, MD, is Director of Refractive and Corneal Surgery at Texan Eye Care in Austin, Texas. He holds no financial interest in any of the products or companies mentioned herein. Dr. Dell may be reached at (512)327-7000; sdell@austin.rr.com.1. Fraenkel GE, Cohen PR, Sutton GL, et al. Central focal interface opacity after laser in situ keratomileusis. J Refract Surg. 1998;14:571-576.
2. Piechocki M. Severe corneal lesions after LASIK are not stage 4 DLK. Ocular Surgery News. January 1, 2003. Available at: http://www.osnsupersite.com. Accessed: October 8, 2003.
3. Hamilton DR, Manche EE, Rich LF, Maloney RK. Steroid-induced glaucoma after laser in situ keratomileusis associated with interface fluid. Ophthalmology. 2002;109:659-665.
