A rare case of endophthalmitis after PRK surgery

November 1, 2013

The authors previously reported a case of a patient with a history of radial keratotomy (RK) who developed acute postoperative endophthalmitis without keratitis shortly after photorefractive keratectomy (PRK) surgery. In this article, the authors present the case study, highlighting the importance of understanding surgical complications.

Case

An 81-year-old woman underwent elective PRK surgery in her left eye for refractive correction. Her history was significant for uncomplicated eight-incision radial keratotomy (RK) surgery in both eyes 9 years prior and successful cataract surgery in the intervening time. Her preoperative best corrected visual acuity (BCVA) was 20/25 in her left eye. For the PRK surgery, the epithelium was removed with 20% ethanol and no mitomycin C was used; no complications were reported.

On the first postoperative day, no concerning abnormalities were noted; she did have a large epithelial defect, as expected. A bandage contact lens was placed and she was started on topical tobramycin and dexamethasone drops. On the fourth postoperative day, the patient noted blurry vision, redness, and irritation but she did not seek immediate care.

On the sixth postoperative day she presented 20/400 vision and signs and symptoms of acute post-op endophthalmitis, including corneal oedema, 10% hypopyon in the anterior chamber and vitreous cells. She did have a large central epithelial defect, but no ulcer or corneal infiltrate was noted.

Shortly thereafter, the patient was sent to a vitreoretinal surgeon for management of acute postoperative endophthalmitis. Extensive vitreous debris was seen on B-scan ultrasound. A vitreous tap was performed for analysis and culture. Intravitreal injections of vancomycin and ceftazidime were also performed, and topical antibiotic drops given. On regular follow-up, a positive response to the antibiotics was seen. The cultures of the vitreous aspirate yielded methicillin-resistant staphylococcus aureus sensitive to vancomycin.

Over the next days and weeks, the patient's vision improved somewhat and sign and symptoms of endophthalmitis resolved. The hypopyon diminished and the corneal oedema cleared as well. At last follow-up, her BCVA was 20/100.

Discussion

In this case, we show a dreaded complication occurring after a corneal refractive surgery that only very rarely has such serious and vision reducing consequences. It is true that endophthalmitis after refractive surgery is very rare, however a notable exception to this are eye with a history of RK surgery. Endophthalmitis without inadvertent globe penetration or keratitis (i.e., corneal ulcer) has never been reported previously.

A comprehensive meta-analysis of intraocular infection after radial karatotomy surgery yielded 43 cases in total.2 Nearly half presented in the first 2 weeks after surgery and two-thirds were bacterial. In another comprehensive review, the literature was reviewed for serious infections after LASIK refractive surgery; a total of 103 cases were found.3 Half of the cases presented in the first 10 days after surgery and 96% had corneal infiltrates. Three eyes had a history of RK and another eye with a history of RK and PRK was found to have bacterial keratitis following LASIK. Only one case of endophthalmitis was found but this occurred secondary to keratitis. In another published case report, a patient with a history of RK underwent LASIK and developed keratitis that lead to endophthalmitis.4

In the cases previously reported in the literature, all intraocular infections started with keratitis of some sort: corneal ulcer, frank keratitis, or a similar corneal infection. Typically, bacteria is seeded in the cornea and then, if unchecked, can extend into the anterior chamber and then to the vitreous, leading to endophthalmitis.

However in our case, no keratitis was seen. We believe that removal of corneal epithelium for the PRK procedure may have re-opened one or more full thickness RK incisions, allowing bacteria to enter into the anterior chamber. RK incisions typically extend 85–95% of the way through the corneal stroma, but should not penetrate full-thickness. However, it is common for RK incisions to penetrate full thickness and cause micro-perforations. In a study of 466 radial keratotomy procedures, the rate of full thickness micro-perforation was 3.8%.5

Considering the significant number of RK procedures done over the years, lurking full thickness perforations in patients who have had previous RK are not uncommon in a refractive practice. Typically, the epithelium quickly heals over these micro-perforations, safely sealing the inside of the eye from possible bacteria seeding.

LASIK or PRK procedures remove the flap or epithelium and likely re-open these micro-perforations and may sometimes allow any bacteria that may be present to directly enter the anterior chamber and possibly gaining a foothold, causing a intraocular infection without preceding corneal infection or keratitis. In LASIK, the flap is replaced, in some ways sealing the path of entry, but in PRK these micro-perforations are exposed for several days after the procedure as the epithelium heals.

It is important to note, in our case the patient's preoperative BCVA was 20/25 and her postoperative visual outcome was 20/100 after this terrible and rare complication. While advanced refractive procedures are no doubt generally very safe, we must always maintain a healthy respect for complications, including devastating endophthalmitis, as seen in this case.

References

1. P.A. Karth and J.W. Karth, J. Ophthalmic Inflamm. Infect., 2013;3(1):31.

2. S. Jain and D.T. Azar, J. Refract. Surg., 1996;12(1):148–155

3. M.A. Chang, S. Jain and D.T. Azar, Surv. Ophthalmol., 2004;49(3):269–280.

4. J. Levy et al., Can. J. Ophthalmol., 2005;40(2):211–213.

5. S. Leroux les Jardins, I. Bertrand and M. Massin, Refract. Corneal Surg., 1992;8(3):215–216.