- Email: [email protected]
Refractive outcomes of combined phacovitrectomy and delayed cataract surgery in retinal detachment
ORIGINAL ARTICLE
Refractive outcomes of combined phacovitrectomy and delayed cataract surgery in retinal detachment Yong-Kyu Kim, MD,*,† Se Joon Woo, MD, PhD,* Joon Young Hyon, MD, PhD,* Jeeyun Ahn, MD,‡ Kyu Hyung Park, MD, PhD* ABSTRACT ● RÉSUMÉ Objective: To compare the accuracy of refractive outcomes between combined pars plana vitrectomy (PPV) and cataract surgery and delayed cataract surgery after PPV in cases with rhegmatogenous retinal detachment (RD). Design: Retrospective case series. Participants: Thirty-eight eyes underwent combined phacovitrectomy (combined group) and 25 eyes underwent delayed cataract surgery after PPV (delayed group). Methods: RD height was measured using optical coherence tomography. Refractive outcomes were evaluated using mean absolute error (MAE; the difference between final refractive error and target refractive error). Results: Combined group showed significant myopic shift (mean error; –0.40 ⫾ 1.07 vs 0.07 ⫾ 0.56 D, p ¼ 0.028) and large MAE (0.81 ⫾ 0.81 vs 0.48 ⫾ 0.29 D, p ¼ 0.028) compared with delayed group. Multiple logistic regression analysis revealed that only RD height was significantly associated with MAE greater than 2 D after combined surgery (in 100-mm unit, odds ratio 3.23, 95% CI 1.04–10.02, p ¼ 0.042). RD height was also significantly correlated with the difference in axial length (AL) between 2 eyes of the patients (p ¼ 0.006, r ¼ 0.406) and the difference in AL measured at pre- versus post-RD repair in the delayed group (p o 0.001, r ¼ 0.774). Conclusions: Combined phacovitrectomy in patients with rhegmatogenous RD induced significant myopic shift because of underestimation of AL, especially in patients with high RD height. Thus, in cases with high temporal RD or large AL differences between eyes, either delayed cataract surgery or combined cataract surgery using the contralateral AL is recommended. Objet : Comparer la réfraction obtenue après une vitrectomie par la pars plana (VPP) combinée à une chirurgie de la cataracte à la réfraction finale après une chirurgie de la cataracte effectuée après la VPP, chez des patients atteints d’un décollement de rétine rhegmatogène (DR). Nature : Étude de cas rétrospective. Participants : 38 yeux ont été traités par phacovitrectomie combinée (groupe chirurgie combinée) et 25 yeux ont été traités par VPP suivie, plus tard, d’une chirurgie de la cataracte (groupe chirurgie ultérieure). Méthodes : La hauteur du DR a été mesurée par tomographie de cohérence optique. Nous avons évalué la réfraction obtenue en nous basant sur l’erreur moyenne absolue (EMA, écart entre l’erreur de réfraction finale et l’erreur de réfraction prévue). Résultats : Chez le groupe ayant subi la chirurgie combinée, on a constaté myopisation importante (EMA; 0,40 ± 1,07 dioptrie c. 0,07 ± 0,56 dioptrie, p = 0,028) et une grande EMA (0,81 ± 0,81 dioptrie c. 0,48 ± 0,29 dioptrie, p = 0,028) comparativement au groupe chirurgie ultérieure. Une analyse de régression logistique multiple a révélé que seule la hauteur du DR était significativement associée à une EMA supérieure à 2 dioptries après la chirurgie combinée (par unité de 100 µm, OR 3,23, 95 % IC 1,04-10,02, p = 0,042). On constate également une corrélation significative entre la hauteur du DR et la différence de longueur axiale (LA) entre les 2 yeux des patients (p = 0,006, r = 0,406) et entre la différence de LA mesurée avant et après l’opération du DR chez les membres du groupe chirurgie ultérieure (p = o0,001, r = 0,774). Conclusion : La phacovitrectomie combinée pour traiter les patients atteints de DR rhegmatogène a causé une augmentation significative de la myopie en raison d’une sous-estimation de la LA, en particulier chez les patients ayant un DR élevé. Par conséquent, dans les cas de DR temporal élevé ou d’une différence marquée de LA entre les deux yeux, nous recommandons soit une chirurgie ultérieure de la cataracte, soit une chirurgie combinée de la cataracte basée sur la LA controlatérale.
Recently, there have been significant advances in the surgical techniques used for vitreoretinal and cataract surgeries, such as small-gauge sutureless vitrectomy1–3 and a torsional ultrasound modality.4,5 Thus, combined phacoemulsification and pars plana vitrectomy (PPV) has become a common procedure in cases of combined retinal detachment (RD) and cataract.6 According to the U.K. National Ophthalmology Database study, of the 2731 phakic eyes undergoing
PPV as the primary treatment for RD, 320 eyes (11.7%) underwent combined phacovitrectomy. One remarkable finding was that the rate of postvitrectomy cataract surgery increased every year.7 Furthermore, cataract formation is known to be the most common complication of PPV, with a cataract developing in up to 80% of patients within 2 years.8–10 These results imply that it is reasonable to perform combined phacovitrectomy in cases of combined
From the *Department of Ophthalmology, Seoul National University College of Medicine, Seoul National University Bundang Hospital, Seongnam; †Department of Ophthalmology, Hallym University College of Medicine, Kangdong Sacred Heart Hospital, Seoul; and ‡Department of Ophthalmology, Seoul National University College of Medicine, Seoul Metropolitan Government-Seoul National University Boramae Medical Center, Seoul, South Korea
Correspondence to Se Joon Woo, MD, Department of Ophthalmology, Seoul National University Bundang Hospital, 300, Gumi-dong, Bundang-gu, Seongnam City, Gyeonggi Province 463-707, South Korea; [email protected]
Originally received Nov. 7, 2014. Final revision Jun. 5, 2015. Accepted Jul. 8, 2015
360
CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
Can J Ophthalmol 2015;50:360–366 0008-4182/15/$-see front matter & 2015 Canadian Ophthalmological Society. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcjo.2015.07.003
Refractive outcomes after cataract and RD surgery— Kim et al. RD and cataract, because this reduces the number of surgeries required, thus limiting inconvenience to the patient. On the contrary, the accuracy of intraocular lens (IOL) power calculation in phacovitrectomy for RD remains an issue. Jeoung et al.11 evaluated the factors that influence refractive outcomes after phacovitrectomy in various vitreoretinal diseases and found that those patients with preoperative foveal detachment showed a significant postoperative myopic shift with longer postoperative axial length (AL). This suggests that in cases of RD, AL could be underestimated. Rahman et al.12 evaluated the accuracy of IOL power estimation in eyes undergoing phacovitrectomy for RD. Although most cases (75.8%) showed a refractive prediction error within ⫾1.0 D, some (6.3%) showed a large prediction error (4 ⫾2.0 D). In this study, we aimed to compare the accuracy of refractive outcomes between combined phacovitrectomy and delayed cataract surgery after PPV in patients with rhegmatogenous RD and also to evaluate the preoperative clinical factors associated with the inaccurate postoperative refractive outcomes.
METHODS
IOL power calculation
AL and keratometry were measured by ultrasound (CineScan; Quantel Medical, Clermont-Ferrand, France) and an Auto-Refracto-Keratometer (KR-8100; Topcon Corp, Tokyo, Japan), respectively. In 54 of 63 cases (30 cases in the combined group, 24 cases in the delayed group), low-coherence interferometry (IOL Master; Carl Zeiss Meditec AG, Jena, Germany) was also performed for ocular biometric evaluation. Ultrasound measurements were obtained using the immersion technique with the patient in the supine position. A scleral immersion shell was used to support the probe, and normal saline was used as the coupling fluid. Ten AL measurements were taken for each eye; the mean value was used for subsequent calculations. The IOL power calculation was performed using the SRK/ T formula. The final refractive target was defined as the average value derived from ultrasonic and IOL master measurements. However, if IOL master biometry could not be performed, we adopted the IOL power calculation from the ultrasonic biometric measurements. We defined final refractive error as the last measured manifest refractive error at least 1 month after the surgery. Mean error (ME) was calculated as the final refractive error minus the final refractive target in spherical equivalent, and mean absolute error (MAE) was defined as the absolute value of the ME.
Patients
We retrospectively reviewed the medical records of patients who underwent either combined phacovitrectomy (combined group) or delayed cataract surgery after PPV (delayed group) for fovea involving rhegmatogenous RD and cataract at Seoul National University Bundang Hospital from April 1, 2008, to May 30, 2013. This study followed the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Seoul National University Bundang Hospital. Those who underwent corneal refractive surgery such as LASIK or LASEK, scleral buckling surgery, or silicone oil tamponade and those with any other ocular condition that might affect refractive outcomes were excluded. Surgical procedure
For vitrectomy, 23-gauge transconjunctival sutureless vitrectomy (Accurus; Alcon Laboratories, Inc, Fort Worth, Tex.) and gas tamponade (either 18% SF6 or 14% C3F8) was performed. We used perfluorocarbon liquid to stabilize the detached retina, and internal subretinal fluid drainage was performed through the original tear. For cataract surgery, a clear corneal incision, either 2.2 or 2.75 mm, was made at the 10 o’clock position before phacoemulsification (Infiniti System; Alcon Laboratories, Inc). Several acrylic foldable IOLs were used (Alcon AcrySof IQ SN60WF, n ¼ 28; AMO Tecnis ZCB00, n ¼ 27; AMO Sensar AR40e, n ¼ 2; Bausch & Lomb Akreos AO MI-60, n ¼ 6). Target refractive error was determined with consideration of the refractive error of the contralateral eye.
Retinal detachment height measurement
We measured preoperative RD height using the caliper tool in the optical coherence tomography (OCT; Spectralis, Heidelberg Engineering, Heidelberg, Germany) software when available. The RD height was defined as the distance between the outermost part of the detached foveal centre and the innermost part of the retinal pigment epithelium, which was measured perpendicularly. Statistical analysis
We compared ME and MAE between the combined and delayed cataract surgery groups. We also compared clinical characteristics between those with and without large MAE (4 2 D) in the combined group. Using multiple logistic regression analysis, we evaluated factors associated with large MAE (4 2 D). In the delayed group, we compared AL as measured at the time of RD surgery and AL as measured at the time of cataract surgery once the RD had resolved. The correlation between the difference in AL as measured at 2 time points and RD height as measured by OCT was evaluated using Pearson correlation test. All statistical analyses were performed using PASW version 18.0 (SPSS, Inc, Chicago, Ill.), and p values less than 0.05 were considered statistically significant.
RESULTS Thirty-eight eyes of 37 patients who underwent combined phacovitrectomy and 25 eyes of 25 patients who CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
361
Refractive outcomes after cataract and RD surgery— Kim et al. Table 1—Demographics and clinical characteristics of patients who underwent retinal detachment repair with combined and delayed cataract surgery Combined Cataract Operation (n ¼ 38) Age, y Male sex, n (%) IOP, mm Hg Corneal astigmatism, D Axial length, mm Category, n (%) o 22.00 mm 22.00–25.99 mm Z 26.00 mm Difference in measured axial length between eyes, mm† Average K, D Macula off, n (%) RD height, μm RD location, n (%) Superior, superior temporal Inferior, inferior temporal Temporal Combined Mean extent of RD, clock hours Corneal incision size, n (%) 2.75 mm 2.2 mm IOL target calculation method US (SRK/T) Average of US and IOL master (SRK/T) Final refractive target, D IOL power, D Mean error, D Mean absolute error, D Mean follow-up period, mo
Delayed Cataract Operation (n ¼ 25)
60.7 20 10.9 0.97 24.64
⫾ 11.1 (52.6) ⫾ 2.7 (n ¼ 36) ⫾ 0.53 (n ¼ 18) ⫾ 1.92
52.3 12 10.6 1.09 25.01
1 31 6 0.20 43.27 26 313.2
(2.6) (81.6) (15.8) ⫾ 0.48 (n ¼ 36) ⫾ 1.77 (68.4) ⫾ 375.4 (n ¼ 30)
0 19 (76) 6 (24) –0.10 ⫾ 0.78 (n ¼ 23) 43.25 ⫾ 1.39 14 (56) 330.9 ⫾ 428.0 (n ¼ 23)
17 7 9 5 5.5
(44.7) (18.4) (23.7) (13.2) ⫾ 2.2
13 (34.2) 25 (65.8)
13 4 5 3 5.2
⫾ 8.6 (48) ⫾ 3.1 (n ¼ 25) ⫾ 0.87 (n ¼ 25) ⫾ 1.72
(52) (16) (20) (12) ⫾ 1.8
p* 0.001 0.719 0.709 0.613 0.438 0.711
0.075 0.957 0.316 0.873 0.962
0.591 0.434
11 (44) 14 (56) o0.001
18 20 –0.92 19.38 –0.40 0.81 7.2
(47.4) (52.6) ⫾ 0.89 ⫾ 4.23 ⫾ 1.07 ⫾ 0.81 ⫾ 5.0
1 24 –1.90 19.48 0.07 0.48 8.2
(4) (96) ⫾ 1.21 ⫾ 3.72 ⫾ 0.56 ⫾ 0.29 ⫾ 9.5
0.001 0.925 0.028 0.028 0.625
p values less than 0.05 are represented in bold. IOP, intraocular pressure; K, keratometry; RD, retinal detachment; US, ultrasound; IOL, intraocular lens. n
†
Student t test, χ2 test, Fisher’s exact test. Contralateral eye axial length – study eye axial length.
underwent delayed cataract surgery after PPV were ultimately included in this study. Mean age in the combined group was older than that of the delayed group (60.7 ⫾
Fig. 1 — Scatter plot showing the distribution of mean error (ME) for each patient. The combined group showed a greater myopic shift postoperatively compared with the delayed group (ME: combined group vs delayed group, –0.40 ⫾ 1.07 vs 0.07 ⫾ 0.56 D, p ¼ 0.028). However, the overall distribution of ME values for each group is similar, except for 6 patients with extremely large mean absolute error (4 2.0 D) in the combined group (ME excluding 6 patients with large mean absolute error, combined group vs delayed group, –0.01 ⫾ 0.61 vs 0.07 ⫾ 0.56 D, p ¼ 0.628). Bars represent mean (long bars) and SD (short bars) of the data. ME ¼ final refractive error – preoperative refractive target.
362
CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
11.1 vs 52.3 ⫾ 8.6 years, p ¼ 0.001). In the delayed group, cataract surgery was performed at an average of 6 months after RD surgery (6.1 ⫾ 2.1 months, range 3–10 months). The combined group showed a greater myopic shift postoperatively compared with the delayed group (ME, –0.40 ⫾ 1.07 vs 0.07 ⫾ 0.56 D, p ¼ 0.028). The MAE was also significantly larger in the combined group (MAE, 0.81 ⫾ 0.81 vs 0.48 ⫾ 0.29 D, p ¼ 0.028; Table 1). The final postoperative refractive error was measured at an average of 7.6 ⫾ 7.3 months after surgery; for 14 patients (7 each in the combined and delayed groups), the final refractive error was measured 1 month after surgery. There were no IOLrelated complications after surgery, such as postoperative IOL tilt or dislocation, in any patients. In Figure 1, a scatter plot presents the ME distribution for each patient. The combined group showed a larger myopic shift compared with the delayed group. However, we can see that the actual ME distribution is similar in both groups, except for a few patients in the combined group who exhibited extremely large ME (4 2 D) values. There were no significant differences in ME or MAE between groups when the 6 patients with MAE greater than 2 D were excluded from the combined group (ME: combined group [n ¼ 32] vs delayed group [n ¼ 25]: 0.01 ⫾ 0.61 vs 0.07 ⫾ 0.56 D, p ¼ 0.628; MAE: combined group vs delayed group: 0.49 ⫾ 0.35 vs 0.48 ⫾ 0.29 D, p ¼ 0.919).
Refractive outcomes after cataract and RD surgery— Kim et al. Table 2—Demographic and clinical characteristics for patients in the combined phacovitrectomy group with and without large (4 2 D) mean absolute refractive error MAE* r 2.0 D (n ¼ 32) Age, y Male sex, n (%) IOP, mm Hg Corneal astigmatism, D Mean axial length, mm Category, n (%) o22.00 mm 22.00–25.99 mm 426.00 mm o22.00 mm or 426.00 mm Difference in measured axial length between eyes, mm‡ Average K, D Macula off, n (%) RD height, μm RD location, n (%) Superior, superior temporal Inferior, inferior temporal Temporal Combined Mean extent of RD, clock hours Corneal incision size, n (%) 2.75 mm 2.2 mm IOL target calculation method US (SRK/T) Average of US and IOL master (SRK/T) Final refractive target, D IOL power, D Mean error, D MAE, D Mean follow-up period, mo
60.9 20 11.1 0.93 24.56
⫾ 11.2 (62.5) ⫾ 2.7 (n ¼ 30) ⫾ 0.50 (n ¼ 15) ⫾ 1.65
0 28 (87.5) 4 (12.5) 4 (12.5) 0.11 ⫾ 0.43 (n ¼ 31) 43.11 ⫾ 1.80 20 (62.5) 214.6 ⫾ 301.2 (n ¼ 25) 17 5 5 5 5.4
(53.1) (15.6) (15.6) (15.6) ⫾ 2.4
10 (31.3) 22 (68.8)
MAE* 4 2.0 D (n ¼ 6)
p†
59.7 ⫾ 0 10.2 ⫾ 1.17 ⫾ 25.04 ⫾
0.631 0.007 0.357 0.654 0.936 0.031
1 3 2 3 0.75 44.16 6 806.0
11.3 2.5 (n ¼ 6) 0.72 (n ¼ 3) 3.21
(16.7) (50) (33.3) (50) ⫾ 0.49 (n ¼ 5) ⫾ 1.42 (100) ⫾ 336.2 (n ¼ 5)
0 2 (33.3) 4 (66.7) 0 5.8 ⫾ 1.0
0.063 0.008 0.128 0.083 0.002 0.007
0.020 0.230 0.392
3 (50) 3 (50) 0.083
13 19 –0.94 19.80 –0.01 0.49 7.1
(40.6) (59.4) ⫾ 0.87 ⫾ 3.29 ⫾ 0.61 ⫾ 0.35 ⫾ 5.7
5 1 –0.79 17.17 –2.49 2.49 7.7
(83.3) (16.7) ⫾ 1.09 ⫾ 7.65 ⫾ 0.13 ⫾ 0.13 ⫾ 4.0
0.509 0.825 o0.001 o0.001 0.719
P values less than 0.05 are represented in bold. Data presented as mean ⫾ standard deviation, where applicable. MAE, mean absolute error; IOP, intraocular pressure; K, keratometry; RD, retinal detachment; IOL, intraocular lens; US, ultrasound. n MAE ¼ final refractive errorrefractive target . † 2 Mann–Whitney test, χ test, Fisher’s exact test. ‡ Contralateral eye axial length – study eye axial length.
We compared clinical characteristics between those with MAE less than and greater than 2 D in the combined group. Eyes with MAE greater than 2 D were more likely to have shorter AL compared with the contralateral eye (AL difference between eyes: 0.75 ⫾ 0.49 vs 0.11 ⫾ 0.43 mm, p ¼ 0.008), to have greater RD height (806.0 ⫾ 336.2 vs 214.6 ⫾ 301.2 mm, p ¼ 0.002), and to have temporal RD (66.7% vs 15.6%, p ¼ 0.02, Table 2) compared with those with MAE less than 2 D. Eyes with MAE greater than 2 D were also more likely (3/6, 50%) to exhibit an extreme AL (o 22 or 4 26 mm) compared with eyes with MAE less than 2 D (4/32, 12.5%), although the difference was not statistically significant (p ¼ 0.063). We performed multiple logistic regression analysis using the 3 most clinically relevant and statistically significant factors: extreme AL (o 22 or 4 26 mm), temporal RD, and RD height (in 100-mm units). Only RD height was significantly associated with MAE greater than 2 D after combined PPV and cataract surgery in RD patients (in 100-mm unit, odds ratio [OR] 3.23, 95% CI 1.04–10.02, p ¼ 0.042). There were significant correlations between the RD height and the difference in AL between the study and contralateral eyes; the greater the RD height, the greater was the difference in AL between eyes. The correlation was more significant when the AL was measured using the IOL
master (ultrasound, p ¼ 0.013; IOL master, p ¼ 0.006; Fig. 2A, 2B). In the delayed group, AL was measured at 2 time points in 20 of 25 cases (i.e., at the time of RD surgery and at the time of cataract surgery after RD repair). We compared the AL difference between 2 time points with the RD height in the delayed group and found a significant correlation. The difference in AL before and after RD repair increased with an increase in the RD height. This difference was more significant when AL was measured using the IOL master (ultrasound, p ¼ 0.045; IOL master, p o 0.001; Fig. 2C, 2D). The AL difference between pre- and post-RD repair was significantly larger when measured by IOL master than when measured by ultrasound (IOL master vs ultrasound: 0.26 ⫾ 0.35 vs 0.01 ⫾ 0.14 D, p ¼ 0.017). We also compared average keratometry values and intraocular pressure (IOP) before and after RD repair in the delayed group and found no significant difference (average keratometry [K]: pre-RD vs post-RD repair, 42.90 ⫾ 0.29 vs 43.05 ⫾ 0.31 D, p ¼ 0.218; IOP: pre-RD vs post-RD repair, 10.6 ⫾ 0.6 vs 10.7 ⫾ 0.4 mm Hg, p ¼ 0.946). There were no significant correlations between the change in IOP and the change in AL (r ¼ 0.433, p ¼ 0.057) or between the change in IOP and the change in average K values (r ¼ –0.253, p ¼ 0.283) before and after RD repair. CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
363
Refractive outcomes after cataract and RD surgery— Kim et al.
Fig. 2 — A and B, Correlation between retinal detachment (RD) height and axial length (AL) difference between study eye and contralateral eye (y-axis ¼ contralateral eye AL – study eye AL). C and D, Correlation between RD height and AL difference measured at pre- versus post-RD repair in the delayed cataract surgery group (y-axis ¼ post-RD repair AL – pre-RD repair AL). A and C, AL measured by ultrasound (US); B and D, AL measured by low-coherence interferometry (IOL master).
DISCUSSION In this study, we compared the accuracy of refractive outcomes between patients with RD treated with combined phacovitrectomy versus delayed cataract surgery after PPV. The combined group showed significant myopic shift and large MAE compared with the delayed group. Several factors might be connected to the ocular biometric measurement errors in the RD cases presented in this article: (i) errors in AL measurement caused by foveal detachment, (ii) measurement errors in AL and keratometry related to decreased IOP caused by RD, (iii) changes in the refractive index of the vitreous cavity after vitrectomy, and (iv) changes in IOL position after vitrectomy. According to the results of this study, AL measurement error directly associated with foveal detachment seems to be the most important factor associated with inaccurate refractive outcomes. Notably, there were no significant differences in ME or MAE between 2 groups when the
364
CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
6 patients with extremely large MAE (4 2 D) and high RD height were excluded from the combined group. In contrast with the previous studies,13,14 there were no significant changes in IOP as measured pre- versus postRD repair. Furthermore, the changes in IOP did not explain the errors in keratometric and AL measurement. However, the limited number of evaluated patients and variable IOP measurement times may have decreased the statistical power. Most importantly, eyes with RD involving the near total retina were usually operated on using the combined encircling procedure and were excluded from this study. Considering that eyes with extensive RD are associated with a low IOP and hypotony, the lack of differences between preoperative and postoperative IOP can also be explained by the exclusion of such cases. We also excluded cases that involved silicone oil injection, which induces marked changes in refractive error15–17; thus, changes in the refractive index after vitrectomy might be trivial. Previous studies on pseudophakic vitrectomy for RD also showed minimal postoperative refractive changes
Refractive outcomes after cataract and RD surgery— Kim et al. (o ⫾0.5 D in most cases), suggesting minimal effects of vitrectomy on the refractive error in RD cases.18–20 Some authors have suggested that myopic overcorrection can occur after gas tamponade, which may be a result of the forward displacement of an IOL by a gas bubble.21,22 However, in another study, there were no significant anterior chamber depth changes after pseudophakic vitrectomy, regardless of the presence or absence of tamponade.20 We did not measure the IOL position such as anterior chamber depth; thus, we could not evaluate the effect of IOL displacement on refractive outcomes from this study. However, intraocular gas is usually present for less than 1 month; IOL position changes after vitrectomy might not be a permanent or significant factor related to refractive outcome. The comparison of pre- and post-RD repair biometric values from the delayed group revealed significant correlations between RD height and the error in AL measurement. Interestingly, the correlation was more significant when AL was measured by IOL master, and the AL difference between pre- and post-RD repair was significantly larger when measured by IOL master than measured by ultrasound. One possible explanation for this is that both the IOL master and the OCT obtain measurements with the patient in the seated position, whereas ultrasound is performed with the patient in the supine position. The retina might have been flattened in the supine as opposed to the seated position. Although the primary signal used in laser Doppler interferometry is assumed to originate at the retinal pigment epithelium,23,24 in cases of RD, the obscuration of retinal pigment epithelium reflection caused by subretinal fluid might have prevented a precise interpretation.25 In some cases of RD, a good signal may be detected anteriorly by the IOL master, corresponding to the inner limiting membrane.26 Poor fixation during the examination may also result in inaccurate AL measurement using the IOL master.12 Rahman et al.12 showed that, in general, the IOL master is superior to ultrasound for AL measurement. However, a significantly larger ultrasound-measured AL was selected for IOL power calculation in the macula-off group, and subgroup analysis revealed that AL measured using the IOL master was shorter than the corresponding measurement by ultrasound by an average of 0.98 mm.12 Therefore, we believe that for cases of RD, particularly macula-off RD, ultrasound is a more reliable tool to measure AL compared with the IOL master. In this study, of the 6 eyes with MAE greater than 2 D, only 1 had both ultrasound and IOL master data. This eye exhibited an ME of –2.23 D using the ultrasound and IOL master average method; this value would have been –1.99 D if only ultrasound was used. Preoperative OCT might not always be available in the general clinical setting. Our results suggest some clinical cues associated with higher RD and inaccurate measurements of AL. Those with greater MAE (4 2 D) showed a higher rate of temporal RD. Because of the relative position of the optic disc, temporally located RDs are
more likely to induce severe foveal detachment than RD at any other location. Actually, those with temporal RD showed higher RD height compared with those with RD in other quadrants (754.7 ⫾ 375.9 vs 179.9 ⫾ 285.7, p o 0.001). Another important finding is that a large difference in AL between eyes in the same patient is likely accompanied by greater RD height and inaccurate measurements of AL. When these findings are observed before surgery, the IOL power should be selected with caution, and AL might be replaced with that of the contralateral eye. In fact, when we recalculated the accuracy of the IOL power using the contralateral AL for 5 eyes with MAE greater than 2 D and available contralateral AL data, ME and MAE significantly decreased (ME: study eye AL vs contralateral eye AL, –2.43 ⫾ 0.25 vs 0.99 ⫾ 0.93 D; p ¼ 0.043; MAE: study eye AL vs contralateral eye AL 2.43 ⫾ 0.25 vs 1.14 ⫾ 0.69 D; p ¼ 0.043). This study was limited by its retrospective design and small patient numbers. The final refractive errors were measured at different time points for each patient. For 14 patients (7 in each group), final refractive error was measured 1 month after the operation. However, it has been reported that, after cataract surgery with a corneal incision of less than 3 mm, corneal curvature and astigmatic and refractive changes have stabilized by 2 weeks after surgery.27,28 Thus, the short follow-up period in some of these cases is not likely to have significantly affected the results of this study. Further studies with larger numbers of patients and a prospective design might be needed to evaluate the influence of RD on ocular biometric measurement errors. In conclusion, combined phacovitrectomy for rhegmatogenous RD repair induced significant myopic shift because of underestimation of AL, especially in patients with high RD height. To avoid unanticipated myopic shift after a combined operation, we should always check whether there are any discrepancies in AL between 2 eyes of the patients, and ultrasound is a more reliable way to measure AL than IOL master in RD cases. In cases with high temporal RD or large AL difference between eyes, we should be more cautious about IOL power selection, and either delayed cataract surgery or combined cataract surgery using the contralateral AL is recommended.
Disclosure: The authors have no proprietary or commercial interest in any materials discussed in this article. Supported by: This work was supported by a grant from the Joint Research Project, Korea Research Council of Fundamental Science and Technology, Korea (S.J.W.). REFERENCES 1. Fujii GY, De Juan E Jr, Humayun MS, et al. Initial experience using the transconjunctival sutureless vitrectomy system for vitreoretinal surgery. Ophthalmology. 2002;109:1814-20. CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
365
Refractive outcomes after cataract and RD surgery— Kim et al. 2. Eckardt C. Transconjunctival sutureless 23-gauge vitrectomy. Retina. 2005;25:208-11. 3. Fine HF, Iranmanesh R, Iturralde D, Spaide RF. Outcomes of 77 consecutive cases of 23-gauge transconjunctival vitrectomy surgery for posterior segment disease. Ophthalmology. 2007;114:1197-200. 4. Liu Y, Zeng M, Liu X, et al. Torsional mode versus conventional ultrasound mode phacoemulsification: randomized comparative clinical study. J Cataract Refract Surg. 2007;33:287-92. 5. Zeng M, Liu X, Liu Y, et al. Torsional ultrasound modality for hard nucleus phacoemulsification cataract extraction. Br J Ophthalmol. 2008;92:1092-6. 6. Smith M, Raman SV, Pappas G, et al. Phacovitrectomy for primary retinal detachment repair in presbyopes. Retina. 2007;27:462-7. 7. Jackson TL, Donachie PH, Sallam A, et al. United Kingdom National Ophthalmology Database study of vitreoretinal surgery: report 3, retinal detachment. Ophthalmology. 2014;121:643-8. 8. Cherfan GM, Michels RG, de Bustros S, et al. Nuclear sclerotic cataract after vitrectomy for idiopathic epiretinal membranes causing macular pucker. Am J Ophthalmol. 1991;111:434-8. 9. Hsuan JD, Brown NA, Bron AJ, et al. Posterior subcapsular and nuclear cataract after vitrectomy. J Cataract Refract Surg. 2001;27: 437-44. 10. Shousha MA, Yoo SH. Cataract surgery after pars plana vitrectomy. Curr Opin Ophthalmol. 2010;21:45-9. 11. Jeoung JW, Chung H, Yu HG. Factors influencing refractive outcomes after combined phacoemulsification and pars plana vitrectomy: results of a prospective study. J Cataract Refract Surg. 2007;33:108-14. 12. Rahman R, Bong CX, Stephenson J. Accuracy of intraocular lens power estimation in eyes having phacovitrectomy for rhegmatogenous retinal detachment. Retina. 2014;34:1415-20. 13. Solberg T, Ytrehus T, Ringvold A. Hypotony and retinal detachment. Acta Ophthalmol (Copenh). 1986;64:26-32. 14. Cho KH, Park IW, Kwon SI. Changes in postoperative refractive outcomes following combined phacoemulsification and pars plana vitrectomy for rhegmatogenous retinal detachment. Am J Ophthalmol. 2014;158:251-6.e2.
366
CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
15. Stefansson E, Anderson MM Jr, Landers MB 3rd, et al. Refractive changes from use of silicone oil in vitreous surgery. Retina. 1988;8: 20-3. 16. Stefansson E, Tiedeman J. Optics of the eye with air or silicone oil. Retina. 1988;8:10-9. 17. Smith RC, Smith GT, Wong D. Refractive changes in silicone filled eyes. Eye (Lond). 1990;4(pt 1):230-4. 18. Campo RV, Sipperley JO, Sneed SR, et al. Pars plana vitrectomy without scleral buckle for pseudophakic retinal detachments. Ophthalmology. 1999;106:1811-5; discussion 6. 19. Speicher MA, Fu AD, Martin JP, von Fricken MA. Primary vitrectomy alone for repair of retinal detachments following cataract surgery. Retina. 2000;20:459-64. 20. Byrne S, Ng J, Hildreth A, et al. Refractive change following pseudophakic vitrectomy. BMC Ophthalmol. 2008;8:19. 21. Patel D, Rahman R, Kumarasamy M. Accuracy of intraocular lens power estimation in eyes having phacovitrectomy for macular holes. J Cataract Refract Surg. 2007;33:1760-2. 22. Suzuki Y, Sakuraba T, Mizutani H, et al. Postoperative refractive error after simultaneous vitrectomy and cataract surgery. Ophthalmic Surg Lasers. 2000;31:271-5. 23. Hitzenberger CK. Optical measurement of the axial eye length by laser Doppler interferometry. Invest Ophthalmol Vis Sci. 1991;32:616-24. 24. Hitzenberger CK, Drexler W, Dolezal C, et al. Measurement of the axial length of cataract eyes by laser Doppler interferometry. Invest Ophthalmol Vis Sci. 1993;34:1886-93. 25. Keane PA, Mand PS, Liakopoulos S, et al. Accuracy of retinal thickness measurements obtained with Cirrus optical coherence tomography. Br J Ophthalmol. 2009;93:1461-7. 26. Lege BA, Haigis W. Laser interference biometry versus ultrasound biometry in certain clinical conditions. Graefes Arch Clin Exp Ophthalmol. 2004;242:8-12. 27. Oshika T, Tsuboi S. Astigmatic and refractive stabilization after cataract surgery. Ophthalmic Surg. 1995;26:309-15. 28. Masket S, Tennen DG. Astigmatic stabilization of 3.0 mm temporal clear corneal cataract incisions. J Cataract Refract Surg. 1996;22:1451-5.
Refractive outcomes of combined phacovitrectomy and delayed cataract surgery in retinal detachment Yong-Kyu Kim, MD,*,† Se Joon Woo, MD, PhD,* Joon Young Hyon, MD, PhD,* Jeeyun Ahn, MD,‡ Kyu Hyung Park, MD, PhD* ABSTRACT ● RÉSUMÉ Objective: To compare the accuracy of refractive outcomes between combined pars plana vitrectomy (PPV) and cataract surgery and delayed cataract surgery after PPV in cases with rhegmatogenous retinal detachment (RD). Design: Retrospective case series. Participants: Thirty-eight eyes underwent combined phacovitrectomy (combined group) and 25 eyes underwent delayed cataract surgery after PPV (delayed group). Methods: RD height was measured using optical coherence tomography. Refractive outcomes were evaluated using mean absolute error (MAE; the difference between final refractive error and target refractive error). Results: Combined group showed significant myopic shift (mean error; –0.40 ⫾ 1.07 vs 0.07 ⫾ 0.56 D, p ¼ 0.028) and large MAE (0.81 ⫾ 0.81 vs 0.48 ⫾ 0.29 D, p ¼ 0.028) compared with delayed group. Multiple logistic regression analysis revealed that only RD height was significantly associated with MAE greater than 2 D after combined surgery (in 100-mm unit, odds ratio 3.23, 95% CI 1.04–10.02, p ¼ 0.042). RD height was also significantly correlated with the difference in axial length (AL) between 2 eyes of the patients (p ¼ 0.006, r ¼ 0.406) and the difference in AL measured at pre- versus post-RD repair in the delayed group (p o 0.001, r ¼ 0.774). Conclusions: Combined phacovitrectomy in patients with rhegmatogenous RD induced significant myopic shift because of underestimation of AL, especially in patients with high RD height. Thus, in cases with high temporal RD or large AL differences between eyes, either delayed cataract surgery or combined cataract surgery using the contralateral AL is recommended. Objet : Comparer la réfraction obtenue après une vitrectomie par la pars plana (VPP) combinée à une chirurgie de la cataracte à la réfraction finale après une chirurgie de la cataracte effectuée après la VPP, chez des patients atteints d’un décollement de rétine rhegmatogène (DR). Nature : Étude de cas rétrospective. Participants : 38 yeux ont été traités par phacovitrectomie combinée (groupe chirurgie combinée) et 25 yeux ont été traités par VPP suivie, plus tard, d’une chirurgie de la cataracte (groupe chirurgie ultérieure). Méthodes : La hauteur du DR a été mesurée par tomographie de cohérence optique. Nous avons évalué la réfraction obtenue en nous basant sur l’erreur moyenne absolue (EMA, écart entre l’erreur de réfraction finale et l’erreur de réfraction prévue). Résultats : Chez le groupe ayant subi la chirurgie combinée, on a constaté myopisation importante (EMA; 0,40 ± 1,07 dioptrie c. 0,07 ± 0,56 dioptrie, p = 0,028) et une grande EMA (0,81 ± 0,81 dioptrie c. 0,48 ± 0,29 dioptrie, p = 0,028) comparativement au groupe chirurgie ultérieure. Une analyse de régression logistique multiple a révélé que seule la hauteur du DR était significativement associée à une EMA supérieure à 2 dioptries après la chirurgie combinée (par unité de 100 µm, OR 3,23, 95 % IC 1,04-10,02, p = 0,042). On constate également une corrélation significative entre la hauteur du DR et la différence de longueur axiale (LA) entre les 2 yeux des patients (p = 0,006, r = 0,406) et entre la différence de LA mesurée avant et après l’opération du DR chez les membres du groupe chirurgie ultérieure (p = o0,001, r = 0,774). Conclusion : La phacovitrectomie combinée pour traiter les patients atteints de DR rhegmatogène a causé une augmentation significative de la myopie en raison d’une sous-estimation de la LA, en particulier chez les patients ayant un DR élevé. Par conséquent, dans les cas de DR temporal élevé ou d’une différence marquée de LA entre les deux yeux, nous recommandons soit une chirurgie ultérieure de la cataracte, soit une chirurgie combinée de la cataracte basée sur la LA controlatérale.
Recently, there have been significant advances in the surgical techniques used for vitreoretinal and cataract surgeries, such as small-gauge sutureless vitrectomy1–3 and a torsional ultrasound modality.4,5 Thus, combined phacoemulsification and pars plana vitrectomy (PPV) has become a common procedure in cases of combined retinal detachment (RD) and cataract.6 According to the U.K. National Ophthalmology Database study, of the 2731 phakic eyes undergoing
PPV as the primary treatment for RD, 320 eyes (11.7%) underwent combined phacovitrectomy. One remarkable finding was that the rate of postvitrectomy cataract surgery increased every year.7 Furthermore, cataract formation is known to be the most common complication of PPV, with a cataract developing in up to 80% of patients within 2 years.8–10 These results imply that it is reasonable to perform combined phacovitrectomy in cases of combined
From the *Department of Ophthalmology, Seoul National University College of Medicine, Seoul National University Bundang Hospital, Seongnam; †Department of Ophthalmology, Hallym University College of Medicine, Kangdong Sacred Heart Hospital, Seoul; and ‡Department of Ophthalmology, Seoul National University College of Medicine, Seoul Metropolitan Government-Seoul National University Boramae Medical Center, Seoul, South Korea
Correspondence to Se Joon Woo, MD, Department of Ophthalmology, Seoul National University Bundang Hospital, 300, Gumi-dong, Bundang-gu, Seongnam City, Gyeonggi Province 463-707, South Korea; [email protected]
Originally received Nov. 7, 2014. Final revision Jun. 5, 2015. Accepted Jul. 8, 2015
360
CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
Can J Ophthalmol 2015;50:360–366 0008-4182/15/$-see front matter & 2015 Canadian Ophthalmological Society. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcjo.2015.07.003
Refractive outcomes after cataract and RD surgery— Kim et al. RD and cataract, because this reduces the number of surgeries required, thus limiting inconvenience to the patient. On the contrary, the accuracy of intraocular lens (IOL) power calculation in phacovitrectomy for RD remains an issue. Jeoung et al.11 evaluated the factors that influence refractive outcomes after phacovitrectomy in various vitreoretinal diseases and found that those patients with preoperative foveal detachment showed a significant postoperative myopic shift with longer postoperative axial length (AL). This suggests that in cases of RD, AL could be underestimated. Rahman et al.12 evaluated the accuracy of IOL power estimation in eyes undergoing phacovitrectomy for RD. Although most cases (75.8%) showed a refractive prediction error within ⫾1.0 D, some (6.3%) showed a large prediction error (4 ⫾2.0 D). In this study, we aimed to compare the accuracy of refractive outcomes between combined phacovitrectomy and delayed cataract surgery after PPV in patients with rhegmatogenous RD and also to evaluate the preoperative clinical factors associated with the inaccurate postoperative refractive outcomes.
METHODS
IOL power calculation
AL and keratometry were measured by ultrasound (CineScan; Quantel Medical, Clermont-Ferrand, France) and an Auto-Refracto-Keratometer (KR-8100; Topcon Corp, Tokyo, Japan), respectively. In 54 of 63 cases (30 cases in the combined group, 24 cases in the delayed group), low-coherence interferometry (IOL Master; Carl Zeiss Meditec AG, Jena, Germany) was also performed for ocular biometric evaluation. Ultrasound measurements were obtained using the immersion technique with the patient in the supine position. A scleral immersion shell was used to support the probe, and normal saline was used as the coupling fluid. Ten AL measurements were taken for each eye; the mean value was used for subsequent calculations. The IOL power calculation was performed using the SRK/ T formula. The final refractive target was defined as the average value derived from ultrasonic and IOL master measurements. However, if IOL master biometry could not be performed, we adopted the IOL power calculation from the ultrasonic biometric measurements. We defined final refractive error as the last measured manifest refractive error at least 1 month after the surgery. Mean error (ME) was calculated as the final refractive error minus the final refractive target in spherical equivalent, and mean absolute error (MAE) was defined as the absolute value of the ME.
Patients
We retrospectively reviewed the medical records of patients who underwent either combined phacovitrectomy (combined group) or delayed cataract surgery after PPV (delayed group) for fovea involving rhegmatogenous RD and cataract at Seoul National University Bundang Hospital from April 1, 2008, to May 30, 2013. This study followed the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Seoul National University Bundang Hospital. Those who underwent corneal refractive surgery such as LASIK or LASEK, scleral buckling surgery, or silicone oil tamponade and those with any other ocular condition that might affect refractive outcomes were excluded. Surgical procedure
For vitrectomy, 23-gauge transconjunctival sutureless vitrectomy (Accurus; Alcon Laboratories, Inc, Fort Worth, Tex.) and gas tamponade (either 18% SF6 or 14% C3F8) was performed. We used perfluorocarbon liquid to stabilize the detached retina, and internal subretinal fluid drainage was performed through the original tear. For cataract surgery, a clear corneal incision, either 2.2 or 2.75 mm, was made at the 10 o’clock position before phacoemulsification (Infiniti System; Alcon Laboratories, Inc). Several acrylic foldable IOLs were used (Alcon AcrySof IQ SN60WF, n ¼ 28; AMO Tecnis ZCB00, n ¼ 27; AMO Sensar AR40e, n ¼ 2; Bausch & Lomb Akreos AO MI-60, n ¼ 6). Target refractive error was determined with consideration of the refractive error of the contralateral eye.
Retinal detachment height measurement
We measured preoperative RD height using the caliper tool in the optical coherence tomography (OCT; Spectralis, Heidelberg Engineering, Heidelberg, Germany) software when available. The RD height was defined as the distance between the outermost part of the detached foveal centre and the innermost part of the retinal pigment epithelium, which was measured perpendicularly. Statistical analysis
We compared ME and MAE between the combined and delayed cataract surgery groups. We also compared clinical characteristics between those with and without large MAE (4 2 D) in the combined group. Using multiple logistic regression analysis, we evaluated factors associated with large MAE (4 2 D). In the delayed group, we compared AL as measured at the time of RD surgery and AL as measured at the time of cataract surgery once the RD had resolved. The correlation between the difference in AL as measured at 2 time points and RD height as measured by OCT was evaluated using Pearson correlation test. All statistical analyses were performed using PASW version 18.0 (SPSS, Inc, Chicago, Ill.), and p values less than 0.05 were considered statistically significant.
RESULTS Thirty-eight eyes of 37 patients who underwent combined phacovitrectomy and 25 eyes of 25 patients who CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
361
Refractive outcomes after cataract and RD surgery— Kim et al. Table 1—Demographics and clinical characteristics of patients who underwent retinal detachment repair with combined and delayed cataract surgery Combined Cataract Operation (n ¼ 38) Age, y Male sex, n (%) IOP, mm Hg Corneal astigmatism, D Axial length, mm Category, n (%) o 22.00 mm 22.00–25.99 mm Z 26.00 mm Difference in measured axial length between eyes, mm† Average K, D Macula off, n (%) RD height, μm RD location, n (%) Superior, superior temporal Inferior, inferior temporal Temporal Combined Mean extent of RD, clock hours Corneal incision size, n (%) 2.75 mm 2.2 mm IOL target calculation method US (SRK/T) Average of US and IOL master (SRK/T) Final refractive target, D IOL power, D Mean error, D Mean absolute error, D Mean follow-up period, mo
Delayed Cataract Operation (n ¼ 25)
60.7 20 10.9 0.97 24.64
⫾ 11.1 (52.6) ⫾ 2.7 (n ¼ 36) ⫾ 0.53 (n ¼ 18) ⫾ 1.92
52.3 12 10.6 1.09 25.01
1 31 6 0.20 43.27 26 313.2
(2.6) (81.6) (15.8) ⫾ 0.48 (n ¼ 36) ⫾ 1.77 (68.4) ⫾ 375.4 (n ¼ 30)
0 19 (76) 6 (24) –0.10 ⫾ 0.78 (n ¼ 23) 43.25 ⫾ 1.39 14 (56) 330.9 ⫾ 428.0 (n ¼ 23)
17 7 9 5 5.5
(44.7) (18.4) (23.7) (13.2) ⫾ 2.2
13 (34.2) 25 (65.8)
13 4 5 3 5.2
⫾ 8.6 (48) ⫾ 3.1 (n ¼ 25) ⫾ 0.87 (n ¼ 25) ⫾ 1.72
(52) (16) (20) (12) ⫾ 1.8
p* 0.001 0.719 0.709 0.613 0.438 0.711
0.075 0.957 0.316 0.873 0.962
0.591 0.434
11 (44) 14 (56) o0.001
18 20 –0.92 19.38 –0.40 0.81 7.2
(47.4) (52.6) ⫾ 0.89 ⫾ 4.23 ⫾ 1.07 ⫾ 0.81 ⫾ 5.0
1 24 –1.90 19.48 0.07 0.48 8.2
(4) (96) ⫾ 1.21 ⫾ 3.72 ⫾ 0.56 ⫾ 0.29 ⫾ 9.5
0.001 0.925 0.028 0.028 0.625
p values less than 0.05 are represented in bold. IOP, intraocular pressure; K, keratometry; RD, retinal detachment; US, ultrasound; IOL, intraocular lens. n
†
Student t test, χ2 test, Fisher’s exact test. Contralateral eye axial length – study eye axial length.
underwent delayed cataract surgery after PPV were ultimately included in this study. Mean age in the combined group was older than that of the delayed group (60.7 ⫾
Fig. 1 — Scatter plot showing the distribution of mean error (ME) for each patient. The combined group showed a greater myopic shift postoperatively compared with the delayed group (ME: combined group vs delayed group, –0.40 ⫾ 1.07 vs 0.07 ⫾ 0.56 D, p ¼ 0.028). However, the overall distribution of ME values for each group is similar, except for 6 patients with extremely large mean absolute error (4 2.0 D) in the combined group (ME excluding 6 patients with large mean absolute error, combined group vs delayed group, –0.01 ⫾ 0.61 vs 0.07 ⫾ 0.56 D, p ¼ 0.628). Bars represent mean (long bars) and SD (short bars) of the data. ME ¼ final refractive error – preoperative refractive target.
362
CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
11.1 vs 52.3 ⫾ 8.6 years, p ¼ 0.001). In the delayed group, cataract surgery was performed at an average of 6 months after RD surgery (6.1 ⫾ 2.1 months, range 3–10 months). The combined group showed a greater myopic shift postoperatively compared with the delayed group (ME, –0.40 ⫾ 1.07 vs 0.07 ⫾ 0.56 D, p ¼ 0.028). The MAE was also significantly larger in the combined group (MAE, 0.81 ⫾ 0.81 vs 0.48 ⫾ 0.29 D, p ¼ 0.028; Table 1). The final postoperative refractive error was measured at an average of 7.6 ⫾ 7.3 months after surgery; for 14 patients (7 each in the combined and delayed groups), the final refractive error was measured 1 month after surgery. There were no IOLrelated complications after surgery, such as postoperative IOL tilt or dislocation, in any patients. In Figure 1, a scatter plot presents the ME distribution for each patient. The combined group showed a larger myopic shift compared with the delayed group. However, we can see that the actual ME distribution is similar in both groups, except for a few patients in the combined group who exhibited extremely large ME (4 2 D) values. There were no significant differences in ME or MAE between groups when the 6 patients with MAE greater than 2 D were excluded from the combined group (ME: combined group [n ¼ 32] vs delayed group [n ¼ 25]: 0.01 ⫾ 0.61 vs 0.07 ⫾ 0.56 D, p ¼ 0.628; MAE: combined group vs delayed group: 0.49 ⫾ 0.35 vs 0.48 ⫾ 0.29 D, p ¼ 0.919).
Refractive outcomes after cataract and RD surgery— Kim et al. Table 2—Demographic and clinical characteristics for patients in the combined phacovitrectomy group with and without large (4 2 D) mean absolute refractive error MAE* r 2.0 D (n ¼ 32) Age, y Male sex, n (%) IOP, mm Hg Corneal astigmatism, D Mean axial length, mm Category, n (%) o22.00 mm 22.00–25.99 mm 426.00 mm o22.00 mm or 426.00 mm Difference in measured axial length between eyes, mm‡ Average K, D Macula off, n (%) RD height, μm RD location, n (%) Superior, superior temporal Inferior, inferior temporal Temporal Combined Mean extent of RD, clock hours Corneal incision size, n (%) 2.75 mm 2.2 mm IOL target calculation method US (SRK/T) Average of US and IOL master (SRK/T) Final refractive target, D IOL power, D Mean error, D MAE, D Mean follow-up period, mo
60.9 20 11.1 0.93 24.56
⫾ 11.2 (62.5) ⫾ 2.7 (n ¼ 30) ⫾ 0.50 (n ¼ 15) ⫾ 1.65
0 28 (87.5) 4 (12.5) 4 (12.5) 0.11 ⫾ 0.43 (n ¼ 31) 43.11 ⫾ 1.80 20 (62.5) 214.6 ⫾ 301.2 (n ¼ 25) 17 5 5 5 5.4
(53.1) (15.6) (15.6) (15.6) ⫾ 2.4
10 (31.3) 22 (68.8)
MAE* 4 2.0 D (n ¼ 6)
p†
59.7 ⫾ 0 10.2 ⫾ 1.17 ⫾ 25.04 ⫾
0.631 0.007 0.357 0.654 0.936 0.031
1 3 2 3 0.75 44.16 6 806.0
11.3 2.5 (n ¼ 6) 0.72 (n ¼ 3) 3.21
(16.7) (50) (33.3) (50) ⫾ 0.49 (n ¼ 5) ⫾ 1.42 (100) ⫾ 336.2 (n ¼ 5)
0 2 (33.3) 4 (66.7) 0 5.8 ⫾ 1.0
0.063 0.008 0.128 0.083 0.002 0.007
0.020 0.230 0.392
3 (50) 3 (50) 0.083
13 19 –0.94 19.80 –0.01 0.49 7.1
(40.6) (59.4) ⫾ 0.87 ⫾ 3.29 ⫾ 0.61 ⫾ 0.35 ⫾ 5.7
5 1 –0.79 17.17 –2.49 2.49 7.7
(83.3) (16.7) ⫾ 1.09 ⫾ 7.65 ⫾ 0.13 ⫾ 0.13 ⫾ 4.0
0.509 0.825 o0.001 o0.001 0.719
P values less than 0.05 are represented in bold. Data presented as mean ⫾ standard deviation, where applicable. MAE, mean absolute error; IOP, intraocular pressure; K, keratometry; RD, retinal detachment; IOL, intraocular lens; US, ultrasound. n MAE ¼ final refractive errorrefractive target . † 2 Mann–Whitney test, χ test, Fisher’s exact test. ‡ Contralateral eye axial length – study eye axial length.
We compared clinical characteristics between those with MAE less than and greater than 2 D in the combined group. Eyes with MAE greater than 2 D were more likely to have shorter AL compared with the contralateral eye (AL difference between eyes: 0.75 ⫾ 0.49 vs 0.11 ⫾ 0.43 mm, p ¼ 0.008), to have greater RD height (806.0 ⫾ 336.2 vs 214.6 ⫾ 301.2 mm, p ¼ 0.002), and to have temporal RD (66.7% vs 15.6%, p ¼ 0.02, Table 2) compared with those with MAE less than 2 D. Eyes with MAE greater than 2 D were also more likely (3/6, 50%) to exhibit an extreme AL (o 22 or 4 26 mm) compared with eyes with MAE less than 2 D (4/32, 12.5%), although the difference was not statistically significant (p ¼ 0.063). We performed multiple logistic regression analysis using the 3 most clinically relevant and statistically significant factors: extreme AL (o 22 or 4 26 mm), temporal RD, and RD height (in 100-mm units). Only RD height was significantly associated with MAE greater than 2 D after combined PPV and cataract surgery in RD patients (in 100-mm unit, odds ratio [OR] 3.23, 95% CI 1.04–10.02, p ¼ 0.042). There were significant correlations between the RD height and the difference in AL between the study and contralateral eyes; the greater the RD height, the greater was the difference in AL between eyes. The correlation was more significant when the AL was measured using the IOL
master (ultrasound, p ¼ 0.013; IOL master, p ¼ 0.006; Fig. 2A, 2B). In the delayed group, AL was measured at 2 time points in 20 of 25 cases (i.e., at the time of RD surgery and at the time of cataract surgery after RD repair). We compared the AL difference between 2 time points with the RD height in the delayed group and found a significant correlation. The difference in AL before and after RD repair increased with an increase in the RD height. This difference was more significant when AL was measured using the IOL master (ultrasound, p ¼ 0.045; IOL master, p o 0.001; Fig. 2C, 2D). The AL difference between pre- and post-RD repair was significantly larger when measured by IOL master than when measured by ultrasound (IOL master vs ultrasound: 0.26 ⫾ 0.35 vs 0.01 ⫾ 0.14 D, p ¼ 0.017). We also compared average keratometry values and intraocular pressure (IOP) before and after RD repair in the delayed group and found no significant difference (average keratometry [K]: pre-RD vs post-RD repair, 42.90 ⫾ 0.29 vs 43.05 ⫾ 0.31 D, p ¼ 0.218; IOP: pre-RD vs post-RD repair, 10.6 ⫾ 0.6 vs 10.7 ⫾ 0.4 mm Hg, p ¼ 0.946). There were no significant correlations between the change in IOP and the change in AL (r ¼ 0.433, p ¼ 0.057) or between the change in IOP and the change in average K values (r ¼ –0.253, p ¼ 0.283) before and after RD repair. CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
363
Refractive outcomes after cataract and RD surgery— Kim et al.
Fig. 2 — A and B, Correlation between retinal detachment (RD) height and axial length (AL) difference between study eye and contralateral eye (y-axis ¼ contralateral eye AL – study eye AL). C and D, Correlation between RD height and AL difference measured at pre- versus post-RD repair in the delayed cataract surgery group (y-axis ¼ post-RD repair AL – pre-RD repair AL). A and C, AL measured by ultrasound (US); B and D, AL measured by low-coherence interferometry (IOL master).
DISCUSSION In this study, we compared the accuracy of refractive outcomes between patients with RD treated with combined phacovitrectomy versus delayed cataract surgery after PPV. The combined group showed significant myopic shift and large MAE compared with the delayed group. Several factors might be connected to the ocular biometric measurement errors in the RD cases presented in this article: (i) errors in AL measurement caused by foveal detachment, (ii) measurement errors in AL and keratometry related to decreased IOP caused by RD, (iii) changes in the refractive index of the vitreous cavity after vitrectomy, and (iv) changes in IOL position after vitrectomy. According to the results of this study, AL measurement error directly associated with foveal detachment seems to be the most important factor associated with inaccurate refractive outcomes. Notably, there were no significant differences in ME or MAE between 2 groups when the
364
CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
6 patients with extremely large MAE (4 2 D) and high RD height were excluded from the combined group. In contrast with the previous studies,13,14 there were no significant changes in IOP as measured pre- versus postRD repair. Furthermore, the changes in IOP did not explain the errors in keratometric and AL measurement. However, the limited number of evaluated patients and variable IOP measurement times may have decreased the statistical power. Most importantly, eyes with RD involving the near total retina were usually operated on using the combined encircling procedure and were excluded from this study. Considering that eyes with extensive RD are associated with a low IOP and hypotony, the lack of differences between preoperative and postoperative IOP can also be explained by the exclusion of such cases. We also excluded cases that involved silicone oil injection, which induces marked changes in refractive error15–17; thus, changes in the refractive index after vitrectomy might be trivial. Previous studies on pseudophakic vitrectomy for RD also showed minimal postoperative refractive changes
Refractive outcomes after cataract and RD surgery— Kim et al. (o ⫾0.5 D in most cases), suggesting minimal effects of vitrectomy on the refractive error in RD cases.18–20 Some authors have suggested that myopic overcorrection can occur after gas tamponade, which may be a result of the forward displacement of an IOL by a gas bubble.21,22 However, in another study, there were no significant anterior chamber depth changes after pseudophakic vitrectomy, regardless of the presence or absence of tamponade.20 We did not measure the IOL position such as anterior chamber depth; thus, we could not evaluate the effect of IOL displacement on refractive outcomes from this study. However, intraocular gas is usually present for less than 1 month; IOL position changes after vitrectomy might not be a permanent or significant factor related to refractive outcome. The comparison of pre- and post-RD repair biometric values from the delayed group revealed significant correlations between RD height and the error in AL measurement. Interestingly, the correlation was more significant when AL was measured by IOL master, and the AL difference between pre- and post-RD repair was significantly larger when measured by IOL master than measured by ultrasound. One possible explanation for this is that both the IOL master and the OCT obtain measurements with the patient in the seated position, whereas ultrasound is performed with the patient in the supine position. The retina might have been flattened in the supine as opposed to the seated position. Although the primary signal used in laser Doppler interferometry is assumed to originate at the retinal pigment epithelium,23,24 in cases of RD, the obscuration of retinal pigment epithelium reflection caused by subretinal fluid might have prevented a precise interpretation.25 In some cases of RD, a good signal may be detected anteriorly by the IOL master, corresponding to the inner limiting membrane.26 Poor fixation during the examination may also result in inaccurate AL measurement using the IOL master.12 Rahman et al.12 showed that, in general, the IOL master is superior to ultrasound for AL measurement. However, a significantly larger ultrasound-measured AL was selected for IOL power calculation in the macula-off group, and subgroup analysis revealed that AL measured using the IOL master was shorter than the corresponding measurement by ultrasound by an average of 0.98 mm.12 Therefore, we believe that for cases of RD, particularly macula-off RD, ultrasound is a more reliable tool to measure AL compared with the IOL master. In this study, of the 6 eyes with MAE greater than 2 D, only 1 had both ultrasound and IOL master data. This eye exhibited an ME of –2.23 D using the ultrasound and IOL master average method; this value would have been –1.99 D if only ultrasound was used. Preoperative OCT might not always be available in the general clinical setting. Our results suggest some clinical cues associated with higher RD and inaccurate measurements of AL. Those with greater MAE (4 2 D) showed a higher rate of temporal RD. Because of the relative position of the optic disc, temporally located RDs are
more likely to induce severe foveal detachment than RD at any other location. Actually, those with temporal RD showed higher RD height compared with those with RD in other quadrants (754.7 ⫾ 375.9 vs 179.9 ⫾ 285.7, p o 0.001). Another important finding is that a large difference in AL between eyes in the same patient is likely accompanied by greater RD height and inaccurate measurements of AL. When these findings are observed before surgery, the IOL power should be selected with caution, and AL might be replaced with that of the contralateral eye. In fact, when we recalculated the accuracy of the IOL power using the contralateral AL for 5 eyes with MAE greater than 2 D and available contralateral AL data, ME and MAE significantly decreased (ME: study eye AL vs contralateral eye AL, –2.43 ⫾ 0.25 vs 0.99 ⫾ 0.93 D; p ¼ 0.043; MAE: study eye AL vs contralateral eye AL 2.43 ⫾ 0.25 vs 1.14 ⫾ 0.69 D; p ¼ 0.043). This study was limited by its retrospective design and small patient numbers. The final refractive errors were measured at different time points for each patient. For 14 patients (7 in each group), final refractive error was measured 1 month after the operation. However, it has been reported that, after cataract surgery with a corneal incision of less than 3 mm, corneal curvature and astigmatic and refractive changes have stabilized by 2 weeks after surgery.27,28 Thus, the short follow-up period in some of these cases is not likely to have significantly affected the results of this study. Further studies with larger numbers of patients and a prospective design might be needed to evaluate the influence of RD on ocular biometric measurement errors. In conclusion, combined phacovitrectomy for rhegmatogenous RD repair induced significant myopic shift because of underestimation of AL, especially in patients with high RD height. To avoid unanticipated myopic shift after a combined operation, we should always check whether there are any discrepancies in AL between 2 eyes of the patients, and ultrasound is a more reliable way to measure AL than IOL master in RD cases. In cases with high temporal RD or large AL difference between eyes, we should be more cautious about IOL power selection, and either delayed cataract surgery or combined cataract surgery using the contralateral AL is recommended.
Disclosure: The authors have no proprietary or commercial interest in any materials discussed in this article. Supported by: This work was supported by a grant from the Joint Research Project, Korea Research Council of Fundamental Science and Technology, Korea (S.J.W.). REFERENCES 1. Fujii GY, De Juan E Jr, Humayun MS, et al. Initial experience using the transconjunctival sutureless vitrectomy system for vitreoretinal surgery. Ophthalmology. 2002;109:1814-20. CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
365
Refractive outcomes after cataract and RD surgery— Kim et al. 2. Eckardt C. Transconjunctival sutureless 23-gauge vitrectomy. Retina. 2005;25:208-11. 3. Fine HF, Iranmanesh R, Iturralde D, Spaide RF. Outcomes of 77 consecutive cases of 23-gauge transconjunctival vitrectomy surgery for posterior segment disease. Ophthalmology. 2007;114:1197-200. 4. Liu Y, Zeng M, Liu X, et al. Torsional mode versus conventional ultrasound mode phacoemulsification: randomized comparative clinical study. J Cataract Refract Surg. 2007;33:287-92. 5. Zeng M, Liu X, Liu Y, et al. Torsional ultrasound modality for hard nucleus phacoemulsification cataract extraction. Br J Ophthalmol. 2008;92:1092-6. 6. Smith M, Raman SV, Pappas G, et al. Phacovitrectomy for primary retinal detachment repair in presbyopes. Retina. 2007;27:462-7. 7. Jackson TL, Donachie PH, Sallam A, et al. United Kingdom National Ophthalmology Database study of vitreoretinal surgery: report 3, retinal detachment. Ophthalmology. 2014;121:643-8. 8. Cherfan GM, Michels RG, de Bustros S, et al. Nuclear sclerotic cataract after vitrectomy for idiopathic epiretinal membranes causing macular pucker. Am J Ophthalmol. 1991;111:434-8. 9. Hsuan JD, Brown NA, Bron AJ, et al. Posterior subcapsular and nuclear cataract after vitrectomy. J Cataract Refract Surg. 2001;27: 437-44. 10. Shousha MA, Yoo SH. Cataract surgery after pars plana vitrectomy. Curr Opin Ophthalmol. 2010;21:45-9. 11. Jeoung JW, Chung H, Yu HG. Factors influencing refractive outcomes after combined phacoemulsification and pars plana vitrectomy: results of a prospective study. J Cataract Refract Surg. 2007;33:108-14. 12. Rahman R, Bong CX, Stephenson J. Accuracy of intraocular lens power estimation in eyes having phacovitrectomy for rhegmatogenous retinal detachment. Retina. 2014;34:1415-20. 13. Solberg T, Ytrehus T, Ringvold A. Hypotony and retinal detachment. Acta Ophthalmol (Copenh). 1986;64:26-32. 14. Cho KH, Park IW, Kwon SI. Changes in postoperative refractive outcomes following combined phacoemulsification and pars plana vitrectomy for rhegmatogenous retinal detachment. Am J Ophthalmol. 2014;158:251-6.e2.
366
CAN J OPHTHALMOL — VOL. 50, NO. 5, OCTOBER 2015
15. Stefansson E, Anderson MM Jr, Landers MB 3rd, et al. Refractive changes from use of silicone oil in vitreous surgery. Retina. 1988;8: 20-3. 16. Stefansson E, Tiedeman J. Optics of the eye with air or silicone oil. Retina. 1988;8:10-9. 17. Smith RC, Smith GT, Wong D. Refractive changes in silicone filled eyes. Eye (Lond). 1990;4(pt 1):230-4. 18. Campo RV, Sipperley JO, Sneed SR, et al. Pars plana vitrectomy without scleral buckle for pseudophakic retinal detachments. Ophthalmology. 1999;106:1811-5; discussion 6. 19. Speicher MA, Fu AD, Martin JP, von Fricken MA. Primary vitrectomy alone for repair of retinal detachments following cataract surgery. Retina. 2000;20:459-64. 20. Byrne S, Ng J, Hildreth A, et al. Refractive change following pseudophakic vitrectomy. BMC Ophthalmol. 2008;8:19. 21. Patel D, Rahman R, Kumarasamy M. Accuracy of intraocular lens power estimation in eyes having phacovitrectomy for macular holes. J Cataract Refract Surg. 2007;33:1760-2. 22. Suzuki Y, Sakuraba T, Mizutani H, et al. Postoperative refractive error after simultaneous vitrectomy and cataract surgery. Ophthalmic Surg Lasers. 2000;31:271-5. 23. Hitzenberger CK. Optical measurement of the axial eye length by laser Doppler interferometry. Invest Ophthalmol Vis Sci. 1991;32:616-24. 24. Hitzenberger CK, Drexler W, Dolezal C, et al. Measurement of the axial length of cataract eyes by laser Doppler interferometry. Invest Ophthalmol Vis Sci. 1993;34:1886-93. 25. Keane PA, Mand PS, Liakopoulos S, et al. Accuracy of retinal thickness measurements obtained with Cirrus optical coherence tomography. Br J Ophthalmol. 2009;93:1461-7. 26. Lege BA, Haigis W. Laser interference biometry versus ultrasound biometry in certain clinical conditions. Graefes Arch Clin Exp Ophthalmol. 2004;242:8-12. 27. Oshika T, Tsuboi S. Astigmatic and refractive stabilization after cataract surgery. Ophthalmic Surg. 1995;26:309-15. 28. Masket S, Tennen DG. Astigmatic stabilization of 3.0 mm temporal clear corneal cataract incisions. J Cataract Refract Surg. 1996;22:1451-5.