Persistent infectious keratitis in a contact lens wearer
In vivo confocal microscopy was used to visualize the cornea to make a diagnosis.
A 21-year-old female nursing student was referred to Tufts New England Eye Center Cornea Service for a persistent infectious keratitis that was not responding to topical antibiotic therapy.
She initially presented to an outside ophthalmologist with 5 days of severe pain, tearing, photophobia and a small infiltrate in the right eye. She was started on Vigamox (moxifloxacin, Alcon) every 2 hours for presumed bacterial keratitis. When symptoms did not improve after 2 days, prednisolone and bacitracin four times daily were added to the regimen. By day 10, her clinical picture had worsened with decreased vision, enlarged infiltrate and prominent conjunctival injection. At this point, her provider obtained corneal cultures and discontinued the steroid drops. On day 13, cultures were still pending, and she was referred to Tufts New England Eye Center for in vivo confocal microscopy (IVCM).
The patient’s ocular history was notable for contact lens use, and she was noted on previous exams to have evidence of lens overwear with faint stromal scars on her superior corneas. She routinely wore lenses while showering and swimming, most recently in a pool containing well water 2 days before the onset of symptoms. Medical, family and social histories were otherwise unremarkable.
On exam, the patient was noted to have a visual acuity of 20/50 in the right eye from a baseline of 20/20. IOPs were normal. She had 2+ conjunctival injection with profuse tearing but no mucus discharge. The cornea was notable for extensive perineuritis and a dense stromal haze most prominent temporally, without associated epithelial defect. There was 1+ cell in the anterior chamber with keratic precipitates on the corneal endothelium. Repeat cultures, slit lamp photos (Figure 1) and IVCM of the cornea (Figure 2) were obtained at this time.
What is your diagnosis?
IVCM showed Acanthamoeba trophozoites within the epithelial layer, sparing the stromal layer. The patient underwent therapeutic limbus-to-limbus epithelial debridement, with samples sent for histologic evaluation and culture. She was started on polyhexamethylene biguanide (PHMB) and chlorhexidine hourly around the clock, atropine twice daily and Vigamox four times daily. Culture returned positive for Acanthamoeba 1 month after the onset of symptoms. Histologic assessment revealed a reactive corneal epithelium with necrosis and Acanthamoeba double-walled cysts. Over the next 4 weeks, the patient underwent two subsequent debridements for persistent epithelial parasitic elements visualized on IVCM.
After 6 weeks of therapy, the perineuritis was resolving and the stromal haze was coalescing into a superotemporal scar. By week 8 of therapy, her visual acuity had improved to 20/30 and repeat IVCM showed no further double-walled cysts. At this time, PHMB and chlorhexidine were decreased to every 3 hours. By week 12, the perineuritis had fully resolved and the stromal scar was stable. Anti-Acanthamoeba medications were slowly tapered over several months.
Acanthamoeba keratitis (AK) is a severe, vision-threatening parasitic infection caused by ubiquitous, free-living protozoa from the Acanthamoeba genus. There are eight species that have been reported to cause keratitis, but the most common are A. castellanii and A. polyphaga. Acanthamoeba are found worldwide in fresh water supplies such as rivers, lakes, wells and swimming pools and in soil, marine sediment and plant material. Organisms have even been isolated from airborne dust particles and air samples.
First described in the United Kingdom in the 1970s, AK was initially considered a rare disease. Its incidence, however, has increased steadily in recent years, partially reflecting a true increase in infection rate but also reflecting better clinician awareness and diagnosis of the disease. Incidence of AK varies by country, with the United States having one of the lower rates at 0.15 per million and the U.K. having the highest at 1.4 per million. The difference in incidence by geographic region is largely explained by domestic water contamination rates and the prevalence of contact lens use. In developed nations, contact lens use remains the major risk factor for AK, with up to 88% of cases occurring in lens wearers. Among contact lens wearers, AK infection is strongly related to exposure to contaminated water (such as showering or swimming in lenses), poor lens hygiene (such as cleansing with tap or well water) and overnight wear. However, the possibility of AK in non-lens wearers must not be overlooked; up to 15% of AK cases occur in non-lens users, usually in association with trauma and exposure to contaminated water or soil. Agricultural workers, people with tank-fed water supplies and those with low socioeconomic status are particularly at risk.
Once introduced to the cornea, Acanthamoeba can be difficult to eradicate due to its ability to encyst. The Acanthamoeba life cycle has two stages: an active trophozoite form and a dormant cyst form. The trophozoite has an amoeboid shape (25 µm to 50 µm), and when infecting the eye, feeds on corneal keratocytes. The cyst (15 µm to 30 µm) is the principal form of the organism and allows it to withstand a hostile environment, such as nutrient deficiency or the toxic effect of anti-amoebic drugs. The double-walled cysts are durable and responsible for the persistent nature of the disease and the need for prolonged therapy to prevent future reactivation once therapy is discontinued.
Prognosis is largely determined by the delay between onset of symptoms and initiation of appropriate therapy. When therapy is initiated quickly, the visual prognosis is typically good, with 90% of patients retaining a visual acuity of 20/40. However, a delay of 3 weeks or more is associated with poor visual outcome. Unfortunately, there is often a delay in diagnosis because the early symptoms of Acanthamoeba infection are variable and nonspecific and the tissue diagnosis is slow. One study found a mean delay to diagnosis among AK patients of 27 days, during which time, 30% received inappropriate therapy with topical antibiotics alone and 50% with topical antibiotics plus steroids.
Given the importance of correct early diagnosis, the clinician must have a high degree of suspicion for AK in the appropriate setting, as well as familiarity with its presenting features. Usually, AK patients present with unilateral pain, photophobia and tearing. The pain is often severe and out of proportion to exam, although lack of pain does not preclude the diagnosis. The exam varies based on the stage of the infection. In early infection, the amoeba resides primarily in the epithelial layer, causing an epitheliopathy with superficial punctate keratopathy, pseudo-dendrites, epithelial infiltrates and limbitis. Perineuritis and perineural infiltrates are classic findings in AK and are present in up to 63% of early cases. As the infection progresses, a ring infiltrate appears in 83% of patients, uveitis in 79% and ulceration in 75%.
Treatment consists of hourly application of topical biguanides (PHMB and chlorhexidine) or diamidines (propamidine and hexamidine). While trophozoites are susceptible to most antimicrobial agents, only biguanides and diamidines can eradicate the Acanthamoeba double-walled cysts. The average duration of medical therapy lasts 5 months to 6 months. Physical removal of cysts with epithelial debridement is a useful adjunctive to medical therapy, and systemic treatment with voriconazole or itraconazole may be helpful to prevent extension to the adjacent structures. If Acanthamoeba penetrates the deep stroma, penetrating keratoplasty may be required to eradicate the disease.
Despite the possible devastating effects of delayed diagnosis, empiric therapy based on exam alone is not generally advisable. Treatment for Acanthamoeba is itself highly toxic and is associated with serious side effects including cataract, iris atrophy, corneal ulceration and corneal melt. Usually, the decision to treat is supported by a second diagnostic method. Unfortunately, traditional tissue diagnostic techniques with culture, biopsy or staining are slow and limited by high false negative rates. Direct microscopy and culture are both highly specific (100%), but only 46% and 75% sensitive, respectively. PCR amplification of Acanthamoeba DNA is rapid and more sensitive (sensitivity 85%, specificity 90%) than culture, but its utility is limited by lack of general availability as well as a higher false positive rate.
In recent years, IVCM has emerged as a rapid, noninvasive diagnostic technique for atypical infectious processes such as Acanthamoeba or fungal keratitis. Multiple studies have shown that IVCM has a sensitivity of 80% for Acanthamoeba and fungal keratitis, another difficult to diagnose process. IVCM can avoid the typical delays in diagnosis and treatment of atypical infectious keratitis, although the sensitivity is highly dependent on the skill and experience of the person acquiring and interpreting the images. IVCM has added utility in monitoring response to treatment because it can be difficult to distinguish persistent infection from the toxic effects of therapy based on exam alone. In the case described above, the decision to taper treatment was based on a combination of improved exam and repeated IVCM without evidence of persistent double-walled cysts.
In summary, AK is a vision-threatening disease that primarily affects contact lens wearers with fresh water exposure. Early recognition and initiation of therapy by the clinician is critical to a good outcome. Often, AK presents a diagnostic challenge because most available diagnostic techniques are slow and have a low sensitivity. IVCM has emerged as a valuable adjunctive tool in the diagnosis of AK that can accelerate correct diagnosis and aid in monitoring response to treatment.
The differential diagnosis for persistent microbial keratitis includes aerobic and anaerobic bacterial, fungal, HSV and protozoal infection. In this case, IVCM (Figure 2 on page 38) was highly suspicious for Acanthamoeba trophozoites and the diagnosis was ultimately confirmed by culture.
- Clarke DW, et al. Microbes Infect. 2006;doi:10.1016/j.micinf.2005.12.009.
- Boggild AK, et al. J Clin Microbiol. 2009;doi:10.1128/JCM.00173-09.
- Chin J, et al. Cont Lens Anterior Eye. 2015;doi:10.1016/j.clae.2014.11.146.
- Cruzat A, et al. Invest Ophthalmol Vis Sci. 2011;doi:10.1167/iovs.10-7048.
- Dart JK, et al. Am J Ophthalmol. 2009;doi:10.1016/j.ajo.2009.06.009.
- Jiang C, et al. Ocul Surf. 2015;doi:10.1016/j.jtos.2015.01.002.
- Khairnar K, et al. Diagn Microbiol Infect Dis. 2011;doi:10.1016/j.diagmicrobio.2011.03.019.
- Kobayashi A, et al. Ophthalmology. 2013;doi:10.1016/j.ophtha.2012.12.038.
- Kumar RL, et al. Semin Ophthalmol. 2010;doi:10.3109/08820538.2010.518516.
- Kurbanyan K, et al. Eye (Lond). 2012;doi:10.1038/eye.2011.270.
- Lau R, et al. J Clin Microbiol. 2015;doi:10.1128/JCM.01607-15.
- Lorenzo-Morales J, et al. Trends Parasitol. 2013;doi:10.1016/j.pt.2013.01.006.
- Müller RT, et al. Ophthalmology. 2015;doi:10.1016/j.ophtha.2015.06.047.
- Parmar DN, et al. Ophthalmology. 2006;doi:10.1016/j.ophtha.2005.12.022.
- Schuster FL, et al. Drug Resist Updat. 2004;doi:10.1016/j.drup.2004.01.002.
- Trabelsi H, et al. Pathol Biol (Paris). 2012;doi:10.1016/j.patbio.2012.03.002.
- Vaddavalli PK, et al. Ophthalmology. 2011;doi:10.1016/j.ophtha.2010.05.018.
- Villani E, et al. Curr Eye Res. 2014;doi:10.3109/02713683.2013.842592.
- For more information:
- Astrid Werner, MD, and Pedram Hamrah, MD, can be reached at New England Eye Center, Tufts University School of Medicine, 750 Washington St., Box 450, Boston, MA 02111; website: www.neec.com.
- Edited by Kristen E. Dunbar, MD, and Kendra Klein, MD, can be reached at New England Eye Center, Tufts University School of Medicine, 750 Washington St., Box 450, Boston, MA 02111; website: www.neec.com.