Reply

TO THE EDITOR:

We thank Drs Hatami and Zare for their letter noting their appreciation and interest in our work, for their insightful comments, and for their interest in advancing neuroprotective strategies for glaucoma. Here, we provide a brief response.

• 1.Regarding retinal nerve fiber layer (RNFL) thickness, we reported that RNFL thickness was significantly increased in treated eyes compared with control eyes. Importantly, this analysis was not based on mean group-level change alone, but rather on the number of patients demonstrating an increase exceeding the 4–5 μm threshold for test–retest variability of the Cirrus OCT platform. Furthermore, because this finding reflected an increase in RNFL thickness, rather than a protection against worsening, a “floor effect” for severe RNFL loss would not be expected to confound this specific analysis. Nevertheless, as the letter notes and as our original manuscript details, these data should be interpreted cautiously given the small sample size, short duration, and lack of placebo control in this phase I study.
• 2.We also examined whether patients who demonstrated RNFL increase correlated with baseline RNFL thickness or Humphrey visual field (HVF) severity. We found that RNFL increases were observed in patients with worse starting baseline thickness values, all less than a baseline of 60 μm (Fig 1A). In contrast, fellow eyes did not demonstrate comparable RNFL increases, regardless of RNFL baseline thickness (Fig 1B). In contrast, we did not observe a relationship between HVF severity and likelihood of RNFL increase; anatomic responders to topical insulin were observed across a broad range of baseline visual field loss (Fig 1C). However, given the small sample size of this phase I trial, these initial observations should not be overinterpreted.
• 3.Regarding the effects of prior trabeculectomy or other ocular surface considerations, the authors raise an important point about the theoretical long-term risk of bleb fibrosis failure that could manifest with heightened signaling of IGF receptors. In our study, 10 patients had a history of prior glaucoma surgery, including 6 patients who underwent trabeculectomy and 4 who had a tube shunt placed. We did not see any effect of topical insulin on IOP, although this was a short trial (Fig 1D). However, we agree that these will be important considerations for patient selection and safety monitoring in any subsequent, longer trial.
• 4.Finally, as noted in our original manuscript, larger studies with full randomization and masking against placebo treatment will be essential to validate these findings and to further assess both efficacy and long-term safety.

Figure 1.

Patient-level data in the topical insulin trial. (A) Relationship between baseline RNFL thickness and change in RNFL thickness in study eyes. In green are the eyes showing an increase in average RNFL thickness after 1 month of topical insulin dosing; some RNFL thinning toward their baseline measurement was detected at month 2, after an additional month of washout off of topical insulin therapy. (B) Relationship between baseline RNFL thickness and change in RNFL thickness in fellow eyes. (C) Change in RNFL thickness at month 1 relative to baseline HVF mean deviation. (D) Change in intraocular pressure in participants who had prior history of incisional glaucoma surgery in the study eye prior to enrollment. Yellow lines represent participants with history of trabeculectomy, and blue lines represent those with history of tube shunt. HVF = Humphrey visual field; RNFL = retinal nerve fiber layer.