Risk Factors for Recurrent Retinal Detachment

In this article, the authors seek to identify risk factors associated with surgical failure as defined by recurrent detachment after uncomplicated pars plana vitrectomy (PPV) for primary rhegmatogenous retinal detachment (RRD).

This was a retrospective study of 519 eyes of 509 patients with primary RRD, who underwent PPV with at least 3 months of follow-up. At the surgeon’s discretion, surgical methods could include 23- or 25-gauge, use of an encircling scleral buckle, chandelier illumination, triamcinolone, laser or cryopexy, air or sulfur hexafluoride or silicone oil, and combined phacoemulsification if a cataract was present.

The study excluded more complex retinal detachments including those with grade C1 proliferative vitreoretinopathy (PVR) or worse; presence of giant retinal tears; use of perfluorocarbon liquid; and tractional, exudative or traumatic retinal detachments.

The main outcomes were primary success rate and factors associated with surgical failure. Failure was defined by the need for additional surgery to manage recurrent retinal detachment, or the presence of silicone oil at the last follow-up. Secondary outcomes were visual acuity and postoperative complications.

With regard to baseline characteristics, the mean follow-up duration was 9.8 months, with a range of 3 to 66 months. The majority of eyes (or 424 out of 519) were phakic, and of these, 420 of 424, or 99%, had phacoemulsification in combination with vitrectomy whereas only 4 cases had lens-sparing vitrectomy. An encircling band was used in 6% (or 30 of 519 cases), a drainage retinotomy was made in 41% (or 213 of 519 cases), and the majority 98% (or 509 of 519 cases) got SF6 as a tamponade agent, as opposed to only 1% (or 6 cases) getting air and 1% (or 4 cases) getting silicone oil.

The rate of overall surgical success was 93.8%, although the success rate was 96.4% without drainage retinotomy and 90.1% with drainage retinotomy. Similar rates held up when excluding those 30 cases that received an encircling band as well. Univariate analyses revealed that macula-off status, use of drainage retinotomy, and greater extent of retinal detachment (as defined on a scale of 1-4 quadrants of involvement), were all significantly associated with surgical failure.

Multivariate analysis with logistic regression revealed that, in particular, the use of drainage retinotomy was significantly associated with surgical failure. Propensity score matching was performed to eliminate sample size bias and to control for confounding variables, such as inferior main break, poor preoperative best corrected visual acuity, greater extent of retinal detachment, macula-off status, use of encircling band, and use of silicone oil. After propensity score matching, drainage retinotomy was still associated with surgical failure, as it was noted that eyes with a drainage retinotomy had 3.2 greater odds of surgical failure as compared to eyes without drainage retinotomy (P = .027).

In a sub-analysis of 32 surgical failures (or 6.2% of all 519 cases), 44% were due to reopened original breaks, 22% were due to proliferative vitreoretinopathy, 22% were due to new breaks, and 9% were due to a reopened drainage retinotomy. Of note, epiretinal membrane (ERM) formation occurred in 9% (or 49 or 519 eyes), and of these, one third required additional surgery for membrane peeling. Risk factors for postoperative ERM formation were found to include drainage retinotomy and preoperative PVR on multivariate analysis. Other postoperative complications included macular hole formation in 1% of eyes and macular edema in 6% of eyes. Ocular hypertension in 22% of eyes, and of these 116 cases, 23% (or 27 eyes) required longer term topical glaucoma drops, while 4% (or 5 eyes) required glaucoma surgery.

The authors note that the surgical success rate of 93.8% is higher than average in the published literature. They attribute this to the exclusion of more severe cases, which they also recognize as a limitation of the study. They cite several other limitations including selection bias, variability in surgeon skill, and relatively short mean follow-up.

In the setting of conflicting reports in the literature regarding the association between drainage retinotomy and surgical failure, the authors highlight that drainage retinotomy was the only significant risk factor for surgical failure by multivariate analysis using logistic regression and propensity score matching. They propose that the association between drainage retinotomy and surgical failure might be explained by variable severity of retinal detachments requiring drainage retinotomy, as well as by postsurgical inflammation that may result in ERM and PVR. They conclude by suggesting that the avoidance of drainage retinotomy may improve surgical success rate.