We appreciate the comments of Ambatipudi and associates regarding our recently published work on chromovitrectomy. In this publication, we described a variety of dyes that may be suitable for intraocular surgery. We thank the authors for allowing us to clarify some important issues in this article.

Our work focused exclusively on the ability of new vital dyes to stain intraocular epiretinal membranes (ERMs) and other intraocular tissues. As part of the ongoing process of selecting new vital dyes, we performed selected morphologic analysis in ex vivo porcine as well as human donor eyes. Such data enabled us to differentiate candidate dyes that are highly toxic to intraocular tissues—for example, retina—from other dyes with lesser or no harmful effects. However, as discussed in the article, our published work in The Journal did not aim to evaluate detailed retina toxicity issues. For this purpose, our group used a well-established rabbit model of retinal toxicity, recently published by us as well. This retinal biocompatibility study showed, for instance, that Evans blue was the most toxic dye in comparison with the much safer profile of suitable candidate dyes such as indigo carmine, brilliant blue, or fast green.

Fluorescein is a great dye to stain the vitreous; however, ERMs, the anterior capsule, and internal limiting membranes have a low binding affinity for this dye. Moreover, fluorescein is a very soluble dye that can penetrate various ocular tissues, and therefore, with its use there may be difficulty differentiating stained preretinal structures from the remaining underlying healthy tissue during intraocular surgery.

Ambatipudi and associates suggested the use of a computer software-based test to evaluate dyes in ocular surgery, quoting a work on articular cartilage tissue. Subjective reliable analysis of stained preretinal tissues was performed by our group, based on previous publications in the ophthalmic literature and specifically on the use of vital dyes in ocular surgery. Despite the theoretical more objective data provided by computer software, this approach may be quite artificial and does not reflect the clinical situation during ocular surgery, when the surgeon needs to identify stained membranes using a semisubjective analysis. Such subjective intraoperative evaluation is necessary for an immediate differentiation between the stained membranes and the unstained retinochoroidal background.

ERMs were stained by 16 vital dyes, but because of the marked toxic effects of some dyes on retinal tissue, results of the experiments regarding their affinity for ERMs will be of little clinical importance. For instance, candidate dyes such as toluidine blue showed toxic effects but stained the ERMs in a sufficient manner, whereas other dyes such as eosin were not toxic but also were inadequate for membrane visualization in vitreoretinal surgery.

Indeed, numerous investigations are needed to obtain an effective assessment of safe vital dyes in chromovitrectomy and cataract surgery for the safety of our patients.