Photographic Monitoring of Herpes Simplex Virus Keratitis During Anti-inflammatory Treatment

In the course of the Herpetic Eye Disease Study, we validated digital photomicrography and computer-assisted image analysis for evaluating the severity of stromal keratitis and endotheliitis due to herpes simplex virus.¹ We have now conducted a nested prospective cohort study to investigate how corneal imaging can track the geometric metamorphosis of herpetic keratitis among 62 patients during the systematic administration of a topical corticosteroid and antiviral agent.

Individuals with herpes simplex virus stromal keratitis or endotheliitis gave informed consent under protocols approved by institutional review boards and were assigned to a Herpetic Eye Disease Study treatment regimen of prednisolone sodium phosphate, 1%, tapered from 8 times per day to once daily over 5 weeks and trifluridine, 1%, 4 times per day for 3 weeks and then twice daily.^2,3 Standardized corneal photographs were obtained at baseline, and 62 patients had repeated photography taken a mode of 35 days (range, 32-38 days) later. Diapositives archived at the Herpetic Eye Disease Study Photography Reading Center were later scanned, converted to gray-scale equivalents, and calibrated to linear and luminance scales.¹ Interactive image processing by one of us (B.M.M.), who was masked to slitlamp biomicroscopic measurements, estimated paired morphometric, cartographic, and densitometric values for area (in millimeters squared), shape factor (4π × area/perimeter²), location (polar coordinates on a corneal template), and relative intensity (average pixel-based gray level) of corneal inflammation and opacification at baseline and at 5 weeks.

The area of corneal opacification contracted significantly (P < .001) during 5 weeks of topical treatment with prednisolone and trifluridine. Inflammatory signs resolved with the prescribed treatment schedule in 43 eyes, while 19 eyes had lingering corneal inflammation that decreased in area by a median of 38% (interquartile range, 15%-49%) using image planimetry. The median shape factor of 0.68 (interquartile range, 0.53-0.79) of the zone of stromal infiltration and edema at baseline did not significantly differ (P = .61) from that of 0.70 (interquartile range, 0.54-0.83) 5 weeks later, at which time the geometric center of the corneal opacity remained within a median distance of 0.8 mm (interquartile range, 0.5-1.1 mm) of its initial position. Neither a larger (>20 mm²) initial area of stromal keratitis (P = .48) nor the presence of iritis (P = .89) at baseline was associated with the relative severity of residual corneal opacification measured on corneal photographs captured 5 weeks after enrollment. However, eyes with an initially whiter intensity (>81 gray-value units) of inflammation and edema were more likely (P < .001) to end up with a denser opacity. On completing 5 weeks of treatment, eyes having visual acuity worse than 20/100 on a modified Bailey-Lovie chart averaged a whiter opacity than those with better visual outcome (P = .03).

Slitlamp photography is able to monitor dynamic alterations of corneal disease.⁴ In managing herpes simplex virus keratouveitis and endotheliitis, the examiner strives to adjudicate treatment responses during dosage adjustment of corticosteroids and antivirals.⁵ We found that digitized photographs can supplement the clinical follow-up of patients with herpetic keratitis and could potentially contribute to therapeutic decision making.

Photoanalysis demonstrated how the disciform contour of stromal inflammation and edema fades and shrinks with treatment while retaining an ellipsoidal shape centered at its initial topographic position. Image processing also confirmed that a greater intensity of stromal inflammation predisposes to a whiter corneal opacity that in turn contributes to poorer vision.

New modalities in documenting conditions of the anterior segment are leading to improved representation and quantitative interpretation of ocular disorders.⁶ The integration of bioimaging and other ophthalmic metadata into a comprehensive electronic record offers the prospect of enriching patient management and facilitating teleconsultation in corneal practice.

Correspondence: Dr Wilhelmus, Sid W. Richardson Ocular Microbiology Laboratory, Department of Ophthalmology, Baylor College of Medicine, 6565 Fannin St, Houston, TX 77030 (kirkw@bcm.edu).

Author Contributions: Dr Wilhelmus had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Group Information: A list of the Herpetic Eye Disease Study Group members was published in Wilhelmus KR, Mitchell BM, Dawson CR, et al; Herpetic Eye Disease Study Group. Slitlamp biomicroscopy and photographic image analysis of herpes simplex virus stromal keratitis. Arch Ophthalmol. 2009;127(2):161-166.

Financial Disclosure: None reported.

Funding/Support: This work was supported by grants EY07479, EY07480, EY07482, EY07483, EY07486, EY07487, EY07488, EY07489, and EY07496 from the National Institutes of Health, by an unrestricted grant from Research to Prevent Blindness, Inc, and by the Sid W. Richardson Foundation.

Trial Registration: clinicaltrials.gov Identifier: NCT00000138

Additional Contributions: This study involved the collaboration and guidance of Chandler R. Dawson, MD, Bruce A. Barron, MD, Herbert E. Kaufman, MD, Joel Sugar, MD, Robert A. Hyndiuk, MD, Peter R. Laibson, MD, R. Doyle Stulting, MD, PhD, and Penny A. Asbell, MD.