Dear Editor,

We thank Drs Serrar, Kodjikian, and Mathis for their interest in our work [1], in which we describe topographical patterns of retinal lesions in primary and secondary forms of multiple evanescent white dot syndrome (MEWDS) using ultrawidefield (UWF) autofluorescence imaging. We also thank you for highlighting your work, in which you quantitatively analyse the area of MEWDS lesions from late-phase 55-degree indocyanine green (ICG) images centred on the macula, which was of great interest to us [2]. We commend your comprehensive identification and phenotyping of a large cohort of patients with secondary MEWDS, given our experience that the mildest cases can often be missed without awareness of this phenomenon.

We agree with most of your comments, but we would like to point out a few additional clinical observations we have made over the years. Although secondary MEWDS was less extensive than primary MEWDS in your cohort, we have seen cases with retina-wide and mostly confluent MEWDS lesions with severe loss of function, associated with angioid streaks [3] or lacquer cracks (unpublished observation). Moreover, although the MEWDS lesions in primary and secondary MEWDS may behave similarly, we would hesitate to comment that the clinical course of both groups of patients are consistently similar. Patients with secondary MEWDS are at higher risk of developing complications simultaneously with the MEWDS episode or shortly afterwards, such as macular neovascularisation, other subretinal fibrovascular proliferation, or chorioretinal atrophy, which may be associated with permanent visual impairment. The frequency of long-term complications such as acute zonal occult outer retinopathy is as yet unknown and will need to be investigated separately for primary and secondary MEWDS.

We agree that a quantitative analysis of MEWDS lesions on UWF imaging would be interesting. Determining topographical distribution and quantitatively analysing this in the absence of UWF imaging is difficult, particularly when MEWDS lesions extend into the far periphery or where areas between central and peripheral MEWDS lesions are spared. In this regard, it is important to note that the extent of detectable MEWDS lesions may depend on the time point of retinal imaging: we have recently documented considerable dynamic changes of MEWDS lesions during the first 3 weeks after symptom onset [4], but it currently remains unclear whether such evolution is common or limited to selected cases. Hence, we propose that future quantitative analyses of the topographic extent of MEWDS should include longitudinal data with standardised intervals at which retinal imaging is performed or — where this is not achievable — should at least take time since onset of symptoms into consideration. Such future studies may improve our understanding of MEWDS and ultimately help patient management and counselling.