Incidence of cataract extraction after diabetic vitrectomy

To determine the frequency of visually significant cataracts after vitrectomy for complications of diabetic retinopathy.

We studied 40 patients and 56 concurrent control patients in a retrospective, consecutive, comparative case series in an institutional setting. Entry criteria included phakic patients with a clear lens or mild lens opacity undergoing anatomically successful diabetic vitrectomy, without lens removal at the time of vitrectomy, without intraocular gas or silicone oil use, and with at least 1 year of postoperative follow-up examination information. Two comparative phakic control groups with the diagnosis of macular hole or epiretinal membrane were selected, also with follow-up examination information of at least 1 year postoperatively. The occurrence of cataract extraction was the principal outcome measure. Its validity as a measure of cataract formation was evaluated by ascertainment of improved visual acuity after cataract extraction. A secondary endpoint analysis included eyes that needed cataract extraction at the final follow-up examination.

For the primary analyses (clear lens preoperatively), there were 26 patients in the study group, 38 in the macular hole control group, and 18 in the epiretinal membrane control group. The cumulative cataract extraction rates at 2 years were 15%, 66%, and 53% respectively. By using multivariate survival analysis, the patient age was an important factor, with a younger age associated with a lower rate of progression to nuclear sclerosis. After controlling for age, the difference in these three groups was still statistically significant. In the four patients with diabetes undergoing cataract surgery, the visual acuity improved at least two lines in only one of the eyes and did not change in three eyes. After expanding the outcome measures to include study patients with mild lens opacities at baseline or those judged to be in need of cataract surgery at the final follow-up examination, there was still a strongly statistically significant difference between the three groups.

The rate of cataract extraction after vitrectomy in patients with diabetes is lower than in patients without diabetes undergoing vitrectomy and suggests a lower rate of cataract formation. This inference should be considered when attributing subnormal vision in a patient who has had a diabetic vitrectomy to a cataract. This is especially significant because the risk ratio in patients with diabetes in general and in patients with a previous vitrectomy is likely less favorable compared with the general population.