2) Forty-one percent of the cases (18/44) in the current study had vitreo-macular interface abnormalities. The higher preoperative central retinal thickness and poorer presenting visual acuity in this study could have been due to the presence of high number of cases with vitreo-macular interface abnormalities. Also, sub- group analysis of eyes with and without vitreo-retinal interface abnormalities similar to that described by Bonnin et al 4 would have made the study results more robust. 3) Further prospective studies would be needed for better defining the importance of vitrectomy in diabetic macular edema by including the vitrectomy alone, antivegf alone and combination of vitrectomy and antivegf treatment arms. To conclude, the role of vitrectomy in diabetic macular edema needs to be studied further with prospective random- ized control trials. Ramesh Venkatesh, Bharathi Bavaharan, Naresh Kumar Yadav Narayana Nethralaya, Bangalore, India Correspondence to: Ramesh Venkatesh, MS; vramesh80@yahoo.com. T AGGEDH1REFERENCESTAGGEDEND 1. Michalewska Z, Stewart MW, Landers 3rd MB, Bednarski M, Adelman RA, Nawrocki J. Vitrectomy in the management of diabetic macular edema in treatment-na € ıve patients. Can J Ophthalmol. 53(4):402–7. 2. Diabetic Retinopathy Clinical Research Network Writing Committee, Haller JA, Qin H, Apte RS, Beck RR, Bressler NM, Browning DJ, Danis RP, Glassman AR, Googe JM, Kollman C, Lauer AK, Peters MA, Stock- man ME. Vitrectomy outcomes in eyes with diabetic macular edema and vitreomacular traction. Ophthalmology. 117(6):1087–93.e3. 3. Ulrich JN. Pars Plana Vitrectomy with Internal Limiting Membrane Peeling for Nontractional Diabetic Macular Edema. Open Ophthalmol J. 115–10. 4. Bonnin S, Sandali O, Bonnel S, Monin C, El Sanharawi M. VITREC- TOMY WITH INTERNAL LIMITING MEMBRANE PEELING FOR TRACTIONAL AND NONTRACTIONAL DIABETIC MAC- ULAR EDEMA: Long-term Results of a Comparative Study. Retina. 35 (5):921–8. Can J Ophthalmol 2019;54:402–403 0008-4182/17/$-see front matter © 2018 Canadian Ophthalmological Society. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jcjo.2018.09.010 Response to Vitrectomy in diabetic macular edema Dear Editor:—We thank Dr. Tripathy for his interest in our manuscript and we welcome this opportunity to address his questions. Diabetic macular edema (DME) develops in approximately 30% of patients who have had diabetes for more than 20 years and constitutes a major cause of visual impairment worldwide. 1 Identifying treatments that can effectively treat DME is critical to managing the increasing number of affected patients. The main difference between ours and previously published vit- rectomy studies is that we included only treatment-na € ıve eyes. Patients enrolled in our study were neither randomized to treat- ment nor were they consecutively seen in our clinic. Treating sur- geons tended to treat patients with better prognoses (better baseline visual acuity and less ischemia) with anti-vascular endothe- lial growth factor (VEGF) injections, whereas patients with worse prognoses were offered vitrectomy. This helps explain why our patients ’ initial visual acuities were worse than those usually seen in clinical trials. However, we believe that the impressive average visual acuity improvements among our patients suggest a role for vitrectomy in the initial treatment of DME in eyes with poor ini- tial vision. In our study, vitrectomy for DME was both safe and durable with more than 80% of patients experiencing improve- ment in visual acuity. 2 Since ours was a retrospective study, follow-up visits after six months were not scheduled consistently and longer term follow-ups were not always available. For this reason, we chose six months as the primary temporal endpoint. This strategy allows our data to be directly compared to the anti-VEGF tri- als, in which the six-month visual acuity results are universally available and do not differ significantly from those at 12 months. A recent study presented the long-term (mean fol- low-up of 37 months) results of vitrectomy for center-involv- ing DME in previously treated and treatment-na € ıve eyes. The mean visual acuity improved from 20/100 to 20/63 at month twelve (N=53) 3 . The visual acuity numbers in Table 1 are correct but unfor- tunately the wrong number was included in the discussion. In Europe, visual acuities are usually measured on the decimal scale and an error may have occurred when these were con- verted to Snellen fractions. We agree that optic nerve function and diagnostic tests for glaucoma may be of interest in diabetic patients but we did not include optical coherence tomography measurements of the nerve fiber layer in the study protocol. Seki and coworkers found that patients with proliferative diabetic retinopathy and coexisting renal dysfunction are at high risk of developing optic atrophy after vitrectomy 4 and vitrectomy for macular holes or epiretinal membranes may be associated with a decrease in retinal nerve fiber layer thickness. 5 We agree that nerve fiber layer damage may occur during surgery and encourage future researchers to investigate this. Based on the data from ours and many other studies, we believe that a multicenter, randomized, clinical trial compar- ing the efficacy, safety, and cost of vitrectomy versus intravi- treal anti-VEGF therapy is warranted. Zofia Michalewska,* Michael W. Stewart,y Maurice B. Landers IIIz Maciej Bednarski,* Ron A. Adelman,§ Jerzy Nawrocki* Letters to the Editor CAN J OPHTHALMOL—VOL. 54, NO. 3, JUNE 2019 403