Lens-Sparing Vitrectomy for Progressive Tractional Retinal DetachmentsAssociated With Stage 4A Retinopathy of Prematurity

Objective  To describe the results of lens-sparing vitrectomy in infants with progressive,fovea-threatening, tractional retinal detachments associated with stage 4Aretinopathy of prematurity.

Methods  In a retrospective, interventional, consecutive clinical case series,the records of patients with stage 4A retinopathy of prematurity who underwentlens-sparing vitrectomy for progressive retinal detachments were reviewed.Retinal attachment status, reversed or arrested retinal dragging, and visualacuity were assessed after the procedure.

Results  Thirty-two eyes of 29 patients underwent lens-sparing vitrectomy ata mean postconceptional age of 43 weeks. Thirty (94%) of 32 eyes had completeretinal reattachment and arrest or reversal of retinal dragging after 1 lens-sparingvitrectomy. Visual acuity of at least central, steady, and unmaintained wasobserved in 17 (81%) of the 21 eyes in which it was tested and at least central,steady, and maintained vision was measured in 13 (62%) of 21 eyes, with 1eye achieving 20/40 visual acuity.

Conclusions  Lens-sparing vitrectomy is a safe and effective procedure for the treatmentof fovea-threatening retinal detachments in patients with stage 4A retinopathyof prematurity.

Lens-sparing vitrectomy (LSV) has been shown to be an effective andsafe alternative to scleral buckling for management of retinal detachmentsassociated with stage 4A retinopathy of prematurity (ROP).^1-5

To further evaluate the efficacy and safety of LSV, we retrospectivelyanalyzed a consecutive series of patients with stage 4A ROP who had progressive,fovea-threatening, tractional retinal detachments treated by a single surgeon.Herein, we report their anatomic and visual results.

We performed a noncomparative, retrospective, consecutive case seriesof 32 eyes of 29 patients with progressive, fovea-threatening retinal detachmentsassociated with stage 4A ROP. All infants were initially examined and treatedat 2 large retina referral practices by a single surgeon (P.J.F.). No eyewas considered for surgery unless more than 4 temporal clock hours or 6 totalclock hours of nontemporal retinal detachment were seen by indirect ophthalmoscopy.The LSV was performed as previously described.¹

The patients’ records were reviewed for the following data: gestationalage, birth weight, involved eye, lowest zone of ROP involvement, whether theinfant had received preoperative laser ablation or a scleral buckle, the postconceptionalage at the time of LSV, postoperative crystalline lens status, retinal reattachmentstatus, final visual acuity (if available), and length of follow-up.

Main outcome measures included retinal reattachment, reversed or arrestedretinal dragging, and visual acuity.

Thirty-two eyes of 29 patients were analyzed. Sixteen of the infantswere female and 13 were male. Birth weights ranged from 550 to 1420 g, witha mean birth weight of 776 g. The infants’ postconceptional ages atthe time of LSV ranged from 38 to 63 weeks, with a mean of 43 weeks. The infantwho underwent LSV at 63 weeks was an exception resulting from a delayed referral.All eyes had progressive stage 4A disease (2 had undergone a previous scleralbuckling procedure). All eyes had received laser ablation.

Comprehensive patient data are summarized in the Table. Thirty (94%) of 32 eyes had complete retinal reattachmentand arrest or reversal of retinal dragging after 1 LSV. Two eyes had progressiveretinal traction postoperatively and were successfully treated with scleralbuckling. Therefore, retinas were ultimately reattached in all 32 eyes. Sixeyes had zone 1 disease and 26 had zone 2 disease. Specific visual acuitydata were available for 21 eyes. A visual acuity of at least central, steady,and unmaintained was found in 17 (81%) of 21 eyes and at least central, steady,and maintained vision was seen in 13 (62%) of 21 eyes, with 1 eye achieving20/40 visual acuity (via Allen pictures). Mean length of follow-up was 766days (range, 60-2114 days).

Two eyes developed cortical cataracts, one of which developed more thana year after the LSV. These were treated successfully (one with cataract extractionand posterior chamber intraocular lens implant, and the other with lensectomyand aphakic contact lens correction). No eyes had endophthalmitis or developeda rhegmatogenous retinal detachment.

This study demonstrates that LSV is a safe and effective technique forthe management of eyes with progressive, fovea-threatening, tractional retinaldetachments in patients with stage 4A ROP. The most important finding is thehigh rate (94%) of anatomic success in patients with advanced and progressiveROP with more than a 2-year follow-up. Many children in this series were tooyoung for Snellen visual acuity measurement and some were lost to follow-up.However, in 21 eyes we were able to obtain visual acuity data, and they demonstratedan encouraging trend toward development of useful vision with LSV.

Because such a high rate of anatomic success can be achieved with thistechnique, it is important to emphasize the benefits of close comanagementwith a pediatric ophthalmologist for the treatment of significant refractiveerrors and amblyopia to have a successful visual outcomein these visually immature infants. By maintaining the crystalline lens^1,2,6 and effectively reducingretinal traction, LSV offers the infant good visual potential.⁷ Patient22 responded remarkably well to LSV intervention and attained 20/40 visualacuity.

Scleral buckling, which has been the standard of care in these cases,can be helpful, although it has drawbacks. These drawbacks include usuallyinducing significant refractive error and not directly treating vitreous traction,which can allow for further retinal dragging, even though there may be posteriorretinal reattachment.⁸

All eyes operated on in the present series were considered to have progressive,fovea-threatening retinal detachments associated with stage 4A ROP. Althoughsmall, nonprogressive stage 4A detachments may resolve without intervention,all eyes in our study had at least 4 temporal clock hours of a fovea-threatening,rapidly progressive retinal detachment (followed up weekly by 1 examiner),or a total of at least 6 clock hours of progressive nontemporal retinal detachment.It is likely that these eyes would have progressed without some form of intervention.

Although LSV may have a steep learning curve, the anatomic and visualresults are good, with a low complication rate.⁸ Aswith other LSV studies,^1,2,7 nosignificant complications of iatrogenic retinal detachment or endophthalmitiswere encountered. Although 2 cataracts formed after this procedure, one developedwell beyond a year after the LSV and thus was probably not the result of directsurgical trauma. Both cases of cataract were managed effectively.

Compared with a previous series,² whichfound a 90% retinal reattachment rate, our study found a comparable rate of94%. Visual acuity potential was similar between the 2 studies as well. Asboth cohorts of patients become older, we expect to gain further insight withregard to their visual acuity development. The present study, in additionto other studies of LSV,^1,2,7 demonstratesthe effectiveness, safety, and long-term benefit of this technique as a viabletreatment option for selected patients with advanced ROP.

Correspondence: Philip J. Ferrone, MD, LongIsland Vitreoretinal Consultants, 600 Northern Blvd, Suite 216, Great Neck,NY 11021.

Submitted for Publication: January 16, 2004;final revision received June 1, 2004; accepted June 1, 2004.

Financial Disclosure: None.

Previous Presentation: This study was presentedin part at the 2003 Annual Meeting of the Association for Research in Visionand Ophthalmology; May 4, 2003; Ft Lauderdale, Fla.