We read with interest the paper titled “Sensitivity and Specificity of Spectral-Domain Optical Coherence Tomography in Detecting Idiopathic Polypoidal Choroidal Vasculopathy” by De Salvo and associates. The paper presents excellent data that will help clinicians in developing a high degree of suspicion for the existence of polypoidal choroidal vasculopathy (PCV) based on optical coherence tomography (OCT) features described by the authors.
In this study, though OCT demonstrated a very high (>90%) sensitivity and specificity to detect PCV compared to indocyanine green angiography (ICGA), the authors still report 2 missed cases and 1 misdiagnosed case of PCV using the OCT. Unfortunately, the authors do not provide images for any of these cases that were either missed or misdiagnosed by the OCT. It will be immensely helpful for readers to be able to see such images where ICGA demonstrates PCV but the OCT does not, and vice versa. It would also be prudent to discuss some of the causes of missed diagnosis and misdiagnosis. Additionally, reporting the 95% confidence intervals for sensitivity and specificity using the Wilson score interval may provide an approximate range and better insight into the applicability of the OCT to diagnose PCV with confidence.
Another concern is the coexistence of both age-related choroidal neovascular membrane (CNVM) and PCV in the same eye, known as polypoidal CNVM. This is a poorly described entity with only few anecdotal reports in existing literature and owing to certain overlapping features of CNVM and PCV, such eyes are likely excluded from most clinical studies. We in the Southeast Asian region see such eyes occasionally and would like to know whether the authors encounter such cases in their vast experience with subjects of European descent. We highlight this issue to make readers aware that both PCV and age-related CNVM may coexist in the same eye.
Lastly, 1 of the 4 tomographic features described by the authors to detect PCV is presence of “hyporeflective lumen within hyperreflective lesions adherent to retinal pigment epithelium.” The Spectralis is possibly the best OCT available today and shows choroidal structural changes in great detail. However, being cost prohibitive, the Spectralis may not be widely available, especially in the developing world. In our experience, other spectral-domain OCT machines, especially those without enhanced depth imaging capability, do not enable visualization of the hyporeflective lumens within the PCV lesions in a large proportion of cases. However, readers may still be able to detect PCV in most cases using the other 3 criteria provided by the authors.