Pars Plana Vitrectomy Versus Combined Pars Plana Vitrectomy Scleral Buckle for Secondary Repair of Retinal Detachment

CLINICAL SCIENCE Pars Plana Vitrectomy Versus Combined Pars Plana Vitrectomy Scleral Buckle for Secondary Repair of Retinal Detachment Ryan B. Rush, MD; Matthew P. Simunovic, MB, BChir, PhD; Saumil Sheth, MD; Assaf Kratz, MD; Alex P. Hunyor, MB, BS INTRODUCTION BACKGROUND AND OBJECTIVE: To investigate the optimal technique for repairing recurrent rhegmatogenous retinal detachments. PATIENTS AND METHODS: A two-year retrospective review of recurrent rhegmatogenous retinal detachments by 23-gauge pars plana vitrectomy (PPV) or combined 23-gauge PPV with encircling scleral buckling was performed. The primary outcome was anatomical success. The secondary outcome was the likelihood of achieving a final best corrected visual acuity of 6/12 or better at 6-month follow-up. RESULTS: Anatomical success was achieved in 65.2% (95% CI, 53.4% to 75.4%) of the PPV group versus 74.3% (95% CI, 57.9% to 85.8%) of the PPV scleral buckling group with one additional procedure (not statistically significant). There was no significant difference in the likelihood of achieving a final acuity of at least 6/12 between groups at 6-month follow-up. CONCLUSION: The results of our study do not demonstrate a superiority of method of primary repair, or of one method of secondary repair, over another. [Ophthalmic Surg Lasers Imaging Retina. 2013;44:xxx-xxx.] Rhegmatogenous retinal detachment can be successfully repaired after a single vitreoretinal procedure in approximately 80% to 90% of cases. The most common reason for failure of primary repair is believed to be the development of proliferative vitreoretinopathy (PVR), which accounts for the failure of 7% to 10% of primary repairs and an increased proportion of secondary procedures. Although high anatomical success rates are reported using various surgical retinal reattachment techniques for both primary and secondary repair, permanent anatomical failure ultimately results in 3% to 5% of cases. 1,2 Not much is known about the influence of primary surgery on final outcome in recurrent retinal detachment; even less is known regarding the merits of different techniques of repairing recurrent detachment. A recent case series reported by Mansouri et al in which 48 of 286 eyes experienced recurrent detachment suggested that those undergoing scleral buckling (SB) as the primary procedure require a statistically significant lower number of reoperations in order to achieve anatomical success when compared to patients undergoing pars plana vitrectomy (PPV) or combined procedures. 3 Furthermore, the study authors assert that patients undergoing SB are more likely to achieve a better final visual outcome, with a lower incidence of cataract formation and less requirement for silicone oil as the tamponading agent. Quiram et al conducted a retrospective study of 56 patients with PVR re-detachment who underwent PPV with inferior retinectomy. 4 Their results suggested improved outcomes with silicone oil tam- From the Sydney Eye Hospital, Sydney Medical School, University of Sydney, 8 Macquarie Street, Sydney NSW 2000, Australia. Originally submitted May 16, 2012. Accepted for publication February 4, 2013. The authors have no financial or proprietary interest in the materials presented herein. Address author correspondence to Ryan Rush, MD, 5407 Nova Scotia Court, Amarillo, TX 79119; 806-351-1870; fax: 806-351-1690; email: ryanbradfordrush21@hotmail.com. doi: 10.3928/23258160-20130604-02 XX/XX 2013 Vol. 44, No. X 1

compare success rates of PPV and combined PPV and SB in recurrent retinal detachments following failed primary repair procedures. Figure. Summary statistics for anatomical success rate by number of retinal detachment repair operations: one (medium gray), two (light gray), or three (dark gray). ponade and radical anterior vitreous base dissection with retinectomy. In contrast, addition of a buckle or buckle revision was not associated with an improved chance of anatomical success. Tseng et al performed a retrospective analysis of 81 cases of recurrent detachment with PVR and similarly found that silicone oil tamponade in combination with relaxing retinotomy was associated with an increased incidence of success. 5 These findings are in keeping with the observation that reattachment surgery with the aid of either silicone oil or perfluoropropane is associated with improved chances of anatomical success. 6-8 Because the optimal secondary repair procedure following failed primary rhegmatogenous retinal detachment (RRD) repair is unclear, vitreoretinal surgeons tend to select their method of repair based on personal preference and the clinical features of the recurrent detachment, which can vary from a single missed break with localized detachment to multiple breaks, total detachment, and advanced PVR. Although the success rate of buckle revision may be as high as 78.6%, 9 the current trend is to treat recurrent retinal detachments with PVR by PPV with intraoperative peeling of membranes combined with tamponading agents. 10 Other options for secondary repair include combined PPV with SB encirclement or revision, PPV combined with buckle removal, 11 and pneumatic retinopexy. 12 The aim of this paper is to MATERIALS AND METHODS The institutional review board approved this retrospective, nonrandomized, comparative chart review of secondary RRD procedures performed at Sydney Eye Hospital between January 2007 and January 2009. All patients with recurrent RRDs following a failed primary repair procedure within this time period regardless of phakic status, number or location of breaks, refractive error, presence of PVR, or macula status were included. Eight vitreoretinal surgeons performed or supervised vitreoretinal fellows in all surgical procedures included. Patients were excluded from the study if there was no documented follow-up of at least 6 months, preexisting macular disease prior to the primary repair procedure, or if the primary repair technique did not involve PPV, SB placement, or a combination of PPV and SB placement. Patients who underwent a lensectomy or pneumatic retinopexy in combination with PPV or PPV and SB were excluded. Surgical Technique Patients who underwent secondary repair by PPV had a standard three-port 23-gauge PPV with or without vitreous base shaving using a wide-angle viewing system according to surgeon preference. Vitreous cutting rates of 2,500 cuts per minute with low vacuum settings were applied using the Accurus surgical system (Alcon, Fort Worth, TX). If PVR was evident, intraoperative membrane peeling was performed with or without the use of adjunctive membrane-staining dyes and/or perfluoro-n-octane assistance. An air fluid exchange was performed and subretinal fluid was removed either with a posterior drainage retinotomy or by anterior drainage with or without the assistance of perfluoro-n-octane. Retinectomies were performed at the surgeon s discretion. All retinal breaks were treated with cryotherapy and/or endolaser photocoagulation. The vitreous was substituted with either a premixed variable expansile/nonexpansile concentration of sulfur hexafluoride or perfluoropropane, perfluoro-n-octane, or silicone oil 1,300 centistokes depending on surgeon preference. The sclerotomies were sutured at the surgeon s discretion. Lensectomies were not performed during any of the procedures. Patients who underwent both SB placement and PPV had the encircling buckle element placed prior to the PPV portion of the case. A No. 42 encircling band was sutured onto the sclera according to the sur- 2 Ophthalmic Surgery, Lasers & Imaging Retina Healio.com/OSLIRetina

geon s preference. The PPV portion of the case proceeded according to the steps above. Data Collection and Analysis The data were collected by chart review of consecutive patients presenting with recurrent RRD following failed primary repair procedures. The preoperative data collected included the patient s best corrected visual acuity (BCVA), age, phakic status (pseudophakic, phakic, or aphakic), myopic status (ie, greater than or equal to 6 diopters of myopia historically), presence of PVR grade B or higher (according to the classification of Machemer et al 13 ), presence of missed breaks, and macula status (detached/off or attached/on) prior to undergoing secondary repair for a recurrent detachment. The intraoperative data collected included the type of secondary procedure performed, the type of vitreous substitute utilized, and whether a retinectomy was performed during the secondary repair procedure. The postoperative data collected included whether the retina re-detached, the length of follow-up, the last documented BCVA, and the development of an epiretinal membrane (ERM), macular hole, cataract formation/progression requiring surgery, cystoid macular edema lasting 6 or more months, or a postoperative increase in intraocular pressure requiring either filtration surgery or one or more ocular antihypertensive medications for more than 3 months. Subjects were considered to have an ERM only if the clinician documented the ERM as a visually significant exam finding in the postoperative period. The retina was considered reattached only if 100% of the remaining retina was anatomically in place. The primary outcome of this study was the anatomical reattachment rate for each RRD repair procedure following a failed primary repair surgery. The secondary outcome was the proportion of patients achieving a final acuity of 6/12 or better. Exact 95% confidence intervals (CIs) for anatomic success rates were computed. For the purpose of analysis, Snellen visual acuities were converted to logmar values developed by Ferris et al. 14 LogMAR values of +3.0 and +2.0 were assigned for visual acuity of hand movement and counting fingers, respectively, according to Holladay s methods. 15 Additionally, we assigned a logmar value of +4.0 for an acuity of light perception. We also analyzed our data to explore other variables influence on outcomes, including presence of PVR, location of break (inferior vs other), number of breaks (single vs multiple), and phakic status. Finally, we assessed the number of reoperations required to anatomically reduce secondary retinal detachment in those undergoing SB, PPV, and combined SB and TABLE 1 Comparison of Clinical Variables by Method of Secondary Detachment Repair Variable PPV SB/PPV Pseudophakia 27 of 69 20 of 35 0.148 Macula off 48 of 69 22 of 35 0.513 Inferior break(s) 27 of 69 17 of 35 0.532 Multiple breaks 45 of 69 29 of 35 0.119 Presence of PVR* 50 of 69 30 of 35 0.239 Primary SB 18 of 69 0 of 35 Total = 18 Primary PPV 41 of 69 34 of 35 Total = 75 Primary SB/PPV 10 of 69 1 of 35 Total = 11 * Presence of PVR prior to secondary repair procedure. All probability calculations are Fisher exact calculations (two-tailed). The final three rows compare primary procedure to choice of secondary procedure. PVR = proliferative vitreoretinopathy; SB = scleral buckling; PPV = pars plana vitrectomy. PPV as the initial procedure as well as factors influencing the choice of procedure. Outcomes were assessed using the VassarStats and AcaStat statistical software packages. In our analysis, P.05 was the threshold for statistical significance. RESULTS A total of 565 case notes of patients undergoing retinal detachment surgery were reviewed, from which 104 patients met the criteria for study inclusion; of the latter, 69 underwent PPV and 35 underwent combined PPV and SB procedures. The average age (mean ± SD) of patients included in the study was 61.4 ± 14.4 years, and the average time from primary repair to subsequent re-detachment was 95.1 ± 148.5 days. The average duration of follow-up was 23.6 ± 1.67 months (23.6 ± 1.5 months for secondary PPV and 23.7 ± 2.0 months for secondary PPV and SB). A total of 14 patients developed ERMs, four developed macular holes, four required cataract surgery, and two underwent filtration surgery for elevated intraocular pressure. No patients had cystoid macular edema lasting 6 months or more. Overall rates of anatomical success in secondary repair (68.3%; 95% CI, 58.8% to 76.4%) were significantly lower than for primary repair (83.5%; 95% CI, 80.4% to 86.1%; Fisher exact P <.001). The average number of reoperations when analyzed by primary repair method was 1.61 (95% CI, 1.22 to 2.00) for SB, 1.35 (95% CI, 1.22 to 1.48) for PPV and 1.36 (95% CI 0.91 to 1.82) for combined PPV and SB (Figure). This difference was not found to achieve statistical significance (ANOVA P =.28, df = 2). Vitreoretinal fellows were XX/XX 2013 Vol. 44, No. X 3

TABLE 2 Rates of Reattachment by Clinical Characteristics Category Successful Unsuccessful Primary operation SB 10 8 PPV 53 22 SB/PPV 8 3 Secondary operation PPV 45 24 SB/PPV 26 9 Univariate involved in all reported cases, and no statistical differences in anatomical outcomes were found between the different vitreoretinal surgeons. We analyzed factors influencing the surgeon s choice of reattachment technique (Table 1). Patients Multivariate P =.2667 P =.6564 P =.3495 P =.5212 Lens status P =.1744 P =.3930 Phakic 35 21 Pseudophakic 36 12 Tamponade P =.6980 P =.9505 SF6 23 12 C3F8 17 8 Oil 27 11 Octane 4 2 Retinectomy P =.6854 P =.6959 No 68 31 Yes 3 2 Tear number P =.3545 P =.4498 Single 23 10 Multiple 48 21 Tear location P =.9869 P =.7932 Non-inferior 41 20 Inferior 31 13 PVR P =.1343 P =.3168 Absent 14 11 Present 57 22 Myopia P =.5258 P =.8227 Absent 51 21 Present 20 11 SB = scleral buckling; PPV = pars plana vitrectomy; PVR = proliferative vitreoretinopathy. Probabilities are for univariate and multivariate logistic regression. who underwent primary SB were more likely to have PPV than PPV with SB revision. Patients undergoing primary combined SB and PPV were more likely to have repeat PPV than PPV with SB revision. Finally, patients who previously underwent primary PPV were evenly divided between those who underwent secondary PPV and those who underwent combined SB and PPV. Analysis of primary procedure (SB, PPV, or combined SB and PPV) versus secondary procedure (PPV or combined SB and PPV) suggests that these differences in secondary repair selection are statistically significant (x 2 = 16.7, df = 2, P <.001). Choice of secondary detachment repair method was not associated with break number, location, phakic status, or macula status. Regarding primary outcome, 65.2% (95% CI, 53.5% to 75.4%) of PPV and 74.3% (95% CI, 57.9% to 85.8%) of combined PPV and SB procedures were successful in the treatment of re-detachment; no statistically significant difference was found between these methods (Table 2). Similarly, there was no difference in anatomical outcomes between primary repair methods. There was no association between anatomical outcomes and mode of tamponade, number of breaks, location of breaks, lens status, myopic status, or whether retinectomy was performed (Table 2). Macula-on detachments were associated on univariate analysis with an increased likelihood of achieving a BCVA of 6/12 or better, as was a better presenting BCVA. Only the association between presenting visual acuity and final visual acuity remained statistically significant on multivariate analysis. There was no association between the likelihood of achieving a final acuity of 6/12 or better and primary or secondary repair method, mode of tamponade, number of breaks, location of breaks, lens status, 4 Ophthalmic Surgery, Lasers & Imaging Retina Healio.com/OSLIRetina

myopic status, or whether retinectomy was performed (Table 3). DISCUSSION Recurrent retinal detachment presents a surgical dilemma in terms of appropriate management and results in a significantly lower rate of anatomical success when compared to primary repair, yet comparatively little has been written about the optimal means of treatment. 3,9 Part of the problem in determining the optimal management lies in the fact that the number of potentially confounding variables is large, including the anatomical and pathological variables of the initial pathology, as well as the methods and nuances of initial and subsequent management. This study is limited by the high number of potentially confounding variables relative to the number of patients and is further hampered by the fact that it is a retrospective, observational study. While combined SB and PPV was associated with a higher rate of success than was PPV alone, the difference did not reach statistical significance in our group. Similarly, there was no difference between methods in terms of likelihood of achieving a BCVA of 6/12 or better. Unlike Mansouri et al, we found no evidence to suggest that failed initial SB is associated with fewer reattachment surgeries when compared to PPV and combined SB and PPV. 3 Furthermore, we did not find any evidence to suggest that initial SB provides a superior visual outcome to either PPV or combined PPV and SB on either univariate or multivariate analysis. This latter finding contrasts with the findings of Mansouri et al, who suggested that SB is associated with a superior final visual acuity. Our analysis did not find any association between break number or location and failure rates. Furthermore, there was no association between tamponading agent and success rates; specifically, there was no evidence to support the assertion made by Quiram et al 3 and Tseng et al 4 that tamponade of PVR detachment with silicone oil is associated with increased incidence of successful RD repair. Unsurprisingly, there was an association between the likelihood of achieving 6/12 or better BCVA and mac- TABLE 3 Presenting and Treatment Variables and Visual Acuity Category VA < 6/12 VA 6/12 Univariate Multivariate Primary operation P =.6924 P =.4539 SB 13 2 PPV 57 16 SB/PPV 9 2 Secondary operation P =.9705 P =.5044 PPV 51 13 SB/PPV 28 7 Lens status P =.8002 P =.5731 Phakic 42 10 Pseudophakic 37 10 Tamponade P =.5238 P =.7781 SF6 23 9 C3F8 21 3 Oil 31 6 Octane 4 2 Retinectomy P =.9418 P =.9433 No 75 20 Yes 4 0 Tear number P =.0748 P =.1089 Single 26 4 Multiple 49 16 Tear location P =.5082 P =.3347 Inferior 46 10 Other 33` 10 PVR P =.5510 P =.7180 Absent 15 5 Present 64 15 Macular status P =.0185 P =.3011 On 21 11 Off 58 9 Presenting acuity CV CV P =.0127 P =.0087 Myopia P =.6909 P =.4627 Absent 55 13 Present 24 7 See study text for analysis of visual acuity data. VA = visual acuity; SB = scleral buckling; PPV = pars plana vitrectomy; PVR = proliferative vitreoretinopathy; CV = continuous variable. XX/XX 2013 Vol. 44, No. X 5

ula-on detachment and better presenting BCVA on univariate analysis. Our study is limited in the comparison of postoperative tamponading agents, given that the number of cases with retained perfluoro-noctane was very low. Our outcome analysis is also limited in regard to retinectomy because so few were performed in this series. In summary, secondary retinal detachment repair affords a lower chance of success when compared to primary repair; little evidence is available to aid the surgeon in deciding which treatment modality may be optimal. The surgical approach to secondary repair may be dictated by the method of primary repair, characteristics of the detachment, or other ocular features, as well as the surgeon s underlying preferences. Our retrospective study does not suggest a clear superiority of any of the two methods studied, although there was a trend toward superior anatomical outcomes for combined PPV and SB over PPV. Future multicenter, randomized, controlled studies, along the lines of the SPR study, 1 with a sample size large enough to provide adequate power to assess a difference of 9.1% (n = 399 for a power of 0.80 and a cut-off of 0.05 for statistical significance) should address these issues. REFERENCES 1. Heimann H, Hellmich M, Bornfeld N, Bartz-Schmidt KU, Hilgers RD, Foerster MH. View 2: the case for primary vitrectomy. Br J Ophthalmol. 2003;87(6):784-787. 2. Ahmadieh H, Moradian S, Faghihi H, et al. Anatomic and visual outcomes of scleral buckling versus primary vitrectomy in pseudophakic and aphakic retinal detachment: six-month follow-up results of a single operation--report no. 1. Ophthalmology. 2005;112(8):1421-1429. 3. Mansouri A, Almony A, Shah GK, Blinder KJ, Sharma S. Recurrent retinal detachment: does initial treatment matter? Br J Ophthalmol. 2010;94(10):1344-1347. 4. Quiram PA, Gonzales CR, Hu W, et al. Outcomes of vitrectomy with inferior retinectomy in patients with recurrent rhegmatogenous retinal detachments and proliferative vitreoretinopathy. Ophthalmology. 2006;113(11):2041-2047. 5. Tseng JJ, Barile GR, Schiff WM, Akar Y, Vidne-Hay O, Chang S. Influence of relaxing retinotomy on surgical outcomes in proliferative vitreoretinopathy. Am J Ophthalmol. 2005;140(4):628-636. 6. Schwartz SG, Flynn HW Jr, Lee WH, Ssemanda E, Ervin AM. Tamponade in surgery for retinal detachment associated with proliferative vitreoretinopathy. Cochrane Database Syst Rev. 2009(4):CD006126. 7. Vitrectomy with silicone oil or perfluoropropane gas in eyes with severe proliferative vitreoretinopathy: results of a randomized clinical trial. Silicone Study Report 2. Arch Ophthalmol. 1992;110(6):780-792. 8. Vitrectomy with silicone oil or sulfur hexafluoride gas in eyes with severe proliferative vitreoretinopathy: results of a randomized clinical trial. Silicone Study Report 1. Arch Ophthalmol. 1992;110(6):770-779. 9. Sharma T, Challa JK, Ravishankar KV, Murugesan R. Scleral buckling for retinal detachment. Predictors for anatomic failure. Retina. 1994;14(4):338-343. 10. Holekamp NM, Grand MG. Vitrectomy for the management of recurrent retinal detachments. Curr Opin Ophthalmol. 1997;8(3):44-49. 11. Rossi T, Boccassini B, Iossa M, Lesnoni G, Mutolo PA. Scleral buckle removal associated with pars plana vitrectomy for recurrent retinal detachment. Eur J Ophthalmol. 2009;19(6):1050-1054. 12. Sharma T, Badrinath SS, Mukesh BN, et al. A multivariate analysis of anatomic success of recurrent retinal detachment treated with pneumatic retinopexy. Ophthalmology. 1997;104(12):2014-2017. 13. Machemer R, Aaberg TM, Freeman HM, Irvine AR, Lean JS, Michels RM. An updated classification of retinal detachment with proliferative vitreoretinopathy. Am J Ophthalmol. 1991;112(2):159-165. 14. Ferris FL 3rd, Kassoff A, Bresnick GH, Bailey I. New visual acuity charts for clinical research. Am J Ophthalmol. 1982;94(1):91-96. 15. Holladay JT. Proper method for calculating average visual acuity. J Refract Surg. 1997;13(4):388-391. 6 Ophthalmic Surgery, Lasers & Imaging Retina Healio.com/OSLIRetina