Abstract
Purpose:
To determine prognostic factors for poor visual outcomes in diabetic tractional retinal detachments (TRDs) who underwent repair with pars plana vitrectomy (PPV) with 5000 cs silicone oil (SO) tamponade at University Hospital, Newark, NJ between 2006 and 2012.
Methods:
Retrospective series. Patients with prior RD repair were excluded from the study. Patients were grouped according to most recent visual acuity (MRVA). MRVAs of NLP, LP, and HM were considered poor outcomes (group A), and MRVAs of CF and greater were considered good outcomes (group B). Complexity score (CS) was calculated for each TRD. Combined rhegmatogenous and TRDs (rtRDs) were also included.
Results:
Forty eyes with TRD treated with PPV and SO were studied. One eye with dense cataract post-operatively was excluded. Twenty two (56.4%) eyes resulted in poor visual outcome, and 17 (43.6%) resulted in good visual outcome. Mean follow-up time was 21.8 months. LogMAR mean MRVAs groups A and B were 2.43 (20/5383) and 1.20 (20/317), respectively. Pre-op VA in groups A and B were 1.55 and 1.37, respectively. Age, sex, rates of type 1 DM, type 2 DM, hypertension, hypercholesterolemia, dialysis, vitreous hemorrhage, rtRD, pre-op VA, or occurrence of IOP>25 post-operatively did not significantly vary between groups (P>0.05), nor did rates of history of panretinal photocoagulation (PRP). Complication rates were statistically indifferent between the groups for hypotony, cataract, glaucoma, oil migration and emuslification, macular pucker, subretinal oil, recurrent RD, hyphema, fibrosis, and corneal edema (P>0.05). Macular involvement was more prevalent in group A than in group B (0.95 vs 0.71)(P<0.05). Mean CS was higher in group A (6.32) than in group B (5.41)(P<0.05). Duration of DM in group A, 19.33±8.11 years, was higher than that of group B, 10.60±6.36 years (P<0.05). Insulin dependence was more common in group A (0.68) than group B (0.35)(P<0.05). Post-op VA measured at 1 day, 1 wk, 1 mo, 2-6 mo, 1 year, 2 years, and 3 years were all significantly better for group B, with the largest statistical difference found of VA at 1st post-op day (P<0.0005).
Conclusions:
High CS, macular involvement, duration of DM, insulin dependence, poor 1st post-op day VA were all associated with poor prognosis in patients with diabetic TRD repaired with PPV and SO tamponade.