We thank Chiang and Tsai for their interest in our study. They suggest that the 3 cases of immune ring formation described in our study may have been a result of herpes simplex stromal keratitis rather than cytomegalovirus (CMV). We agree that it may be clinically difficult to distinguish between different viral entities; hence, testing for viral DNA is necessary. We adhere to a stringent clinical protocol in our uveitis clinic at the Singapore National Eye Centre in which every patient suspected of having viral uveitis initially undergoes aqueous analysis for herpes simplex virus, varicella zoster virus, and CMV by polymerase chain reaction, as stated in the Methods section of our article. If the results were positive for CMV, the viral load then was determined by real-time polymerase chain reaction. These 3 patients demonstrated positive results only for CMV, and demonstrated negative results for herpes simplex virus and varicella zoster virus.

We agree that both CMV and herpes simplex virus infections respond to valganciclovir. However, it should be emphasized that the valganciclovir gel was given prophylactically in these eyes with immune ring and that the mainstay of treatment was the anti-inflammatory therapy using either topical steroids or nonsteroidal anti-inflammatory drugs.

Today with the advent of confocal microscopy, noninvasive diagnostic tests increasing play an important role in our clinical practice, and we are now using this tool to help us confirm our diagnosis of CMV while ruling out other viruses.