Evaluation of Vitrectomy for Diabetic Macular Edema

Study Design The study is designed as a prospective cohort study. A randomized trial design was considered but rejected after deciding that there was insufficient equipoise on the part of the investigator group to randomize eyes with DME and vitreal traction to surgery or no surgery (thus eyes which potentially may benefit most from vitrectomy would not be included), and there was insufficient information available on the natural course or surgical outcomes of eyes with DME but without significant traction.

A cohort study provides the opportunity to collect data prospectively using a standardized protocol to assess the potential benefits and risks of vitrectomy. The results can be used to determine whether proceeding with a randomized trial has merit and what the design of the trial should be. If a randomized trial is to be conducted, the results plus the cohort study experience can be used to help design the protocol.

Study Objectives

1. To provide information on the following outcomes in eyes with Diabetic Macular Edema (DME) that undergo vitrectomy: visual acuity, retinal thickening, resolution of traction (if present), surgical complications.
2. To identify subgroups in which there appears to be a benefit of vitrectomy and subgroups in which vitrectomy does not appear to be beneficial.
3. To obtain data that can be used to plan a randomized trial.
4. Intervention Vitrectomy performed by the investigator's usual routine.
5. Duration of Follow-Up: Two years
6. Follow-up Visit Schedule Study visits for data collection at 3 and 6 months then 1, and 2 years. Additional visits follow investigator's usual routine.
7. Rationale:

There are at least two avenues of investigation that support the theoretical value of vitrectomy for the treatment of DME, based on (1) vitrectomy for the relief of traction on the macula and (2) vitrectomy to improve oxygenation of the macula leading to decreased permeability with subsequent resolution or decrease in DME.

Vitrectomy to relieve biomechanical traction on the macula has been reported widely. Schepens and coworkers discussed the role of the vitreous and vitreomacular traction in cystoid macular edema in 1984. Nasrallah et al observed in 1988 the resolution of diabetic macular edema in individuals with spontaneous separation of the vitreous gel from the retina. In 1992, Lewis and coworkers reported success with vitrectomy and peeling of a "thickened hyaloid membrane" in eyes with DME that had this anatomical feature. Since this report of a nonrandomized retrospective case series, other authors have prospectively analyzed their series and supported the concept that relief of clear-cut anteroposterior traction, usually in the setting of an epiretinal membrane complex and associated vitreous adherence, may ameliorate macular thickening and edema in DME. Evaluation of these individuals and documentation of pre and postoperative characteristics have been rendered vastly more objective by ocular coherence tomography and the Retina Thickness Analyzer. Series using optical coherence tomography (OCT) to image cases where vitreomacular traction is observed and in some cases treated, has confirmed the clinical impression of mechanical forces at work on the posterior retina and has documented the anatomic improvement with surgery. How and in which cases OCT could refine our ability to diagnose and define clinically important anatomical features or relationships has not been investigated. As Kaiser and coworkers have documented, the OCT findings in the cases that have thus far come to vitrectomy in these situations support a conclusion that the disease process has progressed very far and in many cases the individuals have actual traction retinal detachments in their maculae. These severe cases are the exception in the spectrum of DME: most cases of macular edema have no obvious vitreomacular traction, but this factor has not been investigated adequately with our newer and more sophisticated imaging techniques. It is possible that subclinical traction on the macula exists in a large number of individuals with diabetes, whose internal limiting membranes at the vitreomacular interface often have a thickened, hypercellular appearance and whose vitreous gels, gradually contracting over many years, may exert subclinical but significant traction on the compromised diabetic macular vascular bed.

The other line of reasoning and prior research that supports the possibility that vitrectomy would help DME is that articulated by Steffanson and others indicating that posterior segment oxygenation improves after vitrectomy. Using oxygen sensors on the retinal surface, these investigators have shown that retinal oxygen tensions increase after the vitreous gel is removed and the posterior segment becomes perfused by relatively oxygen-rich aqueous humor. Supporting this conclusion is the additional observation that retinal vessels decrease in caliber after vitrectomy, presumably in response to the improvement in hypoxia, although confounding factors that could contribute to this decrease, such as the addition of endolaser retinal photocoagulation, have not been ruled out. Numerous lines of investigation have elucidated factors producing permeability in retinal blood vessels. One of the most central of these factors is Vascular Endothelial Growth factor (VEGF), formerly known as Vascular Permeability Factor (VPF). VEGF is known to be upregulated by hypoxia, and downregulated by increased oxygenation. The speculated sequence of events in which vitrectomy produces improved oxygenation of the posterior segment, leading to downregulation of VEGF, leading to decreased vasopermeability, resulting in reduced macular thickening, is a plausible one. More rapid clearing of growth factors in the vitrectomized eye has also been postulated as a potential mechanism for this response.

See full protocol at drcr.net for list of references