Coats病的手术治疗

目的总结手术治疗Coats病的效果。方法对Coats病伴不同程度渗出性视网膜脱离的患者16例17只眼行巩膜外冷凝术和玻璃体手术治疗，手术后随访时间4.25～62.25个月, 平均随访时间13.10个月。结果手术治疗后8只眼视网膜完全复位(无硅油充填)，视网膜复位率47%；1只眼在硅油充填的情况下视网膜在位；8只眼手术后视网膜未能完全复位。5只眼手术后视力提高，2只眼视力稳定，7只眼视力下降。手术并发症有一过性渗出性视网膜脱离加重，视网膜前局限性增生，白内障形成，继发青光眼和玻璃体积血。结论Coats 病伴视网膜脱离经手术治疗后大多数病例视网膜可复位，部分患者视力提高。(中华眼底病杂志,2005,21:145-147)     

玻璃体视网膜手术治疗先天性视网膜劈裂及其并发症的疗效观察

目的 观察玻璃体视网膜手术治疗先天性视网膜劈裂(XLRS)及其并发视网膜脱离和(或)玻璃体积血的疗效.方法 回顾分析接受玻璃体视网膜手术治疗的XLRS并发视网膜脱离和(或)玻璃体积血患者21例27只眼的临床资料.所有患眼眼底及光相干断层扫描(OCT)检查均发现黄斑微囊样劈裂病变合并周边部视网膜劈裂.平均视力0.11±0.09,黄斑劈裂平均面积为(1.09±0.56) mm2.12只眼并发孔源性视网膜脱离,5只眼并发牵拉性视网膜脱离,6只眼并发玻璃体积血,4只眼同时合并视网膜脱离和玻璃体积血.均行经扁平部三通道闭合式玻璃体切割手术.根据情况行内界膜剥离,眼内激光光凝,C3 F8或硅油填充.手术后随访9～122个月,平均随访时间51个月.观察视力以及视网膜解剖结构改善情况.结果 末次随访视力提高者20只眼,占74.1％;维持不变者7只眼,占25.9％.平均视力提高至0.26±0.15.与治疗前平均视力比较,差异具有统计学意义(t=-6.320,P=0.000).27只眼视网膜解剖结构复位良好,视网膜平伏.OCT检查显示,黄斑劈裂平均面积(0.29±0.21) mm2,较治疗前黄斑劈裂平均面积显著缩小(t=10.358,P=0.000);黄斑微囊样病变得到明显的改善.随访期间4只眼出现并发症,占14.8％.其中,2只眼分别在手术后6、8个月并发增生性玻璃体视网膜病变伴牵拉性视网膜脱离;1只眼在手术后4个月出现并发性白内障;1只眼在手术后15个月因新发视网膜裂孔而发生玻璃体积血.给予再次手术治疗后,4只眼视网膜复位良好.结论 玻璃体视网膜手术能有效提高XLRS患者视力,恢复视网膜解剖结构,获得良好的治疗效果.     

伴脉络膜损伤的外伤性视网膜脱离玻璃体手术预后观察

目的 评价玻璃体手术治疗合并脉络膜损伤的外伤性视网膜脱离的效果。 方法 对1995年至2005年间经玻璃体手术治疗1075只外伤眼进行回顾性分析，外伤性视网膜脱离合并浆液性脉络膜脱离、出血性脉络膜脱离（含外伤性脉络膜分离）或视网膜下出血在内的不同类型脉络膜损伤共41例41只眼（3.8％），采用闭合式玻璃体手术进行治疗，统计比较手术预后。 结果 视网膜复位38只眼（92.7％），最终视力大于0.1者10只眼（24.4％）；手术后视力提高者共29只眼（70.7％），其中合并视网膜下出血组14只眼（87.5％,14/16），合并浆液性脉络膜脱离组12只眼（75.0%, 12/16），合并出血型脉络膜脱离组3只眼（33.3%,3/9），三组患眼的视力提高率χ2＝8.394，P＝0.015，P<0.05。最终黑朦6只眼，均为出血型脉络膜脱离者。17只低眼压眼中8只眼（47.1％）需持续硅油填充，出血型脉络膜脱离5只眼（55.6 ％，5/9）。 结论 合并脉络膜损伤的外伤性视网膜脱离玻璃体手术处理得当可获得较好的结果，其中合并视网膜下出血者手术预后较好；合并出血型脉络膜脱离眼预后较浆液型明显差，但也并非眼球摘除适应证。严重外伤性脉络膜分离眼预后较差常 为低眼压需长期硅油填充。 (中华眼底病杂志，2006，22：295-298）     

玻璃体切除术治疗牵拉性视网膜脱离34例

目的：探讨玻璃体切除术治疗牵拉性视网膜脱离的方法和疗效.方法：各种病因导致的牵拉性视网膜脱离34例36眼,行玻璃体切除术治疗,术后观察视力、视网膜复位情况及手术并发症等,随访3-6（平均3.8）mo.结果：术后视网膜复位,视力有不同程度的提高者32眼（89%）;术后3mo内视网膜再脱离4眼,其中2眼经再次手术后复位,2眼眼球萎缩未再手术.手术并发症主要有术中牵拉性裂孔、术中及术后玻璃体积血、术后高眼压、视网膜再脱离等.结论：玻璃体切除术是治疗牵拉性视网膜脱离的有效方法,尤其对于存在广泛固定牵拉、视网膜大范围脱离、玻璃体积血、合并严重的增殖性玻璃体视网膜病变的患者是唯一有效的手术方法.     

玻璃体手术治疗视网膜静脉周围炎并发症的疗效

目的 观察视网膜静脉周围炎合并玻璃体积血和(或)牵引性视网膜脱离行玻璃体切割术的疗效及并发症。 方法 回顾性分析1989～2001年行玻璃体切割术治疗合并玻璃体积血和(或)牵引性视网膜脱离的视网膜静脉周围炎患者69例的临床和随访资料。随访时间为手术后6个月～12年，平均随访时间45个月。 结果 (1)手术后视力较手术前显著提高。(2)11只眼有手术中并发症，占14.3%，其中医源性裂孔7只眼，手术中出血3只眼，晶状体损伤1只眼。(3)手术后1个月内20只眼有并发症，占260%，其中玻璃体再积血14只眼，一过性高眼压6只眼，视网膜脱离5只眼，前房出血2只眼，前房机化膜形成1只眼。(4)远期并发症主要为白内障(9只眼)及黄斑病变(6只眼)。 结论 玻璃体切割术联合眼内激光光凝、剥膜等是治疗视网膜静脉周围炎合并玻璃体积血和(或)牵引性视网膜脱离的有效手段。 (中华眼底病杂志, 2002, 18: 215-217)     

进展型增生性糖尿病视网膜病变玻璃体 切割手术的价值与意义

目的了解有不同并发症的进展型增生性糖尿病视网膜病变眼进行玻璃体手术的结果。方法将患有Ⅰ、Ⅱ型糖尿病进展型增生性糖尿病视网膜病变的314只眼分为玻璃体积血合并局限牵拉性视网膜脱离组;广泛纤维血管膜合并牵拉性视网膜脱离组;牵拉孔源混合性视网膜脱离组;玻璃体积血视网膜脱离合并老年性白内障行玻璃体手术联合白内障摘除及人工晶状体植入组,分别进行回顾性分析。结果玻璃体积血合并局限牵拉性视网膜脱离组中Ⅰ、Ⅱ型糖尿病患 者手术后获得0.1以上视力的分别占39.4%和66.7%,广泛纤维血管膜合并牵拉性视网膜脱离组中Ⅰ、Ⅱ型糖尿病患者手术后获得0.1以上视力的分别占31.6%和51.6%,牵拉孔源混合性视网膜脱离组手术后获得0.1以上视力者占31.6%,玻璃体积血视网膜脱离合并老年性白内障行玻璃体手术联合白内障摘除及人工晶状体植入组手术后获得0.1以上视力者占62.5%。首要的术中 并发症是医源性视网膜裂孔,术后视力丧失的主要原因包括新生血管性青光眼、视网膜脱离和视网膜中央动脉阻塞。结论玻璃体切割手术联合全视网膜光凝术,能有效地改善进展性糖尿病视网膜病变患者的视力。(中华眼底病杂志,2001,17:171-174)     

早产儿视网膜病变全视网膜脱离的玻璃体视网膜手术治疗预后

目的：评价玻璃体视网膜手术治疗早产儿视网膜病变(ROP)视网膜全脱离的临床效果。方法:回顾分析接受玻璃体视网膜手术治疗的56例5期（视网膜全脱离）ROP患儿73只眼的临床资料。手术时患儿年龄为3～84个月，平均年龄(13.02±14.64)个月，出生胎龄为25～36周，平均胎龄(29.5±2.22)周；出生体重为900～2500 g，平均体重(1400.19 ±300.05) g；吸氧时间为2～90 d，平均吸氧时间为(20.53±18.91) d。手术包括开放玻璃体切割手术和闭合晶状体切除、玻璃体切割手术。视网膜解剖复位的标准以黄斑复位为视网膜复位成功，视网膜复位黄斑脱离为部分成功，视网膜未复位为手术失败。结果:手术后10只眼视网膜复位成功，占13.69%；20只眼视网膜复位部分成功，占27.39%；43只眼视网膜复位失败 ，占58.9%。手术后视力为手动者9只眼，占12.33%；条栅视力＞0.004者6只眼,占8.21% 。结论:ROP 视网膜全脱离的玻璃体视网膜手术治疗手术难度较大，手术后仅少数患儿视网膜复位成功 ，部分患儿视功能有一定恢复。     

玻璃体切除术治疗合并脉络膜脱离的孔源性视网膜脱离18例

目的：探讨玻璃体切除术治疗孔源性视网膜脱离并发脉络膜脱离的方法和疗效。方法：18例18眼并发脉络膜脱离的孔源性视网膜脱离,行玻璃体切除术治疗,术后观察视力、视网膜复位情况及手术并发症等,随访2～6(平均3．3)mo。结果：在18眼中13眼(72％)术后视网膜复位,大部分视力均有不同程度的提高;5眼视网膜未复位,其中3眼经再手术后复位,2眼眼球萎缩,未再手术。手术并发症主要有术后葡萄膜炎、玻璃体积血、术后高眼压等。结论：及时的玻璃体手术治疗合并脉络膜脱离的孔源性视网膜脱离,大多数视网膜能够获得复位,部分恢复视功能。     

糖尿病视网膜病变玻璃体切除术后玻璃体出血的临床分析

目的 探讨糖尿病视网膜病变(DR)玻璃体切割手术后玻璃体积血的原因，处理措施以及对预后的影响。 方法 回顾性分析98例DRⅣ期患者122只眼行玻璃体手术治疗后发生玻璃体积血25只眼的临床资料。 结果 玻璃体切割手术后发生玻璃体积血占本组玻璃体切割手术患者的20.5%。积血发生在手术后1周内者8只眼，1周至1个月者6只眼，1个月以上者11只眼。25只眼中C₃ F₈填充眼占31.1%,硅油填充眼占6.1%；空气填充眼占33.3%；灌注液填充眼占26.3%。视网膜周边部新生血管增生9只眼。3只硅油填充眼中2只眼积血自行吸收，1只眼局部形成视网膜前膜，在硅油取出同时行前膜剥除；22只非硅油填充眼中6只眼积血自行吸收；2只眼积血加重，但未及时处理，1只眼发生新生血管性青光眼，1只眼广泛玻璃体视网膜增生脱离，视力无光感；14只眼观察2周积血无吸收后进行了再次手术治疗，12只眼1次手术处理后未再积血。随访结束时，视力无光感者3只眼，手动者2只眼，数指～0.1者10只眼，0.3及以下者4只眼，0.3以上者6只眼。 结论 DR玻璃体切割手术后发生玻璃体积血的患者多数有周边部新生血管增生，经过及时手术治疗，预后较好。 (中华眼底病杂志,2007,23:241-243)     

玻璃体切除术治疗合并视网膜脱离的重症眼外伤疗效分析

目的评价玻璃体手术救治合并视网膜脱离的重症眼外伤的疗效。方法对近3年来我院收治的39例(39只眼)合并有玻璃体出血、外伤性视网膜脱离的重症机械性眼外伤患者的病史资料和随访情况作回顾性研究分析报告。结果39只眼中最后随访时36只眼(36/39,92.3%)视网膜成功复位,其中32只眼(32/39,82.1%)首次手术后痊愈。15只眼巨大裂孔中14只眼(14/15,93.3%)视网膜成功复位。最后随访时34只眼(34/39,87.2%)视力有明显提高,手术前后视力比较,差异显著(χ2=29.632,P<0.01)。结论玻璃体视网膜手术救治合并视网膜脱离的重症眼外伤有良好的疗效,各期良好的手术处理均对最后视功能的恢复有重要作用。     

玻璃体积血的病因分析及手术治疗的疗效

目的：探讨玻璃体积血的病因及行玻璃体切割术治疗的临床疗效。

方法：对162例173眼玻璃体积血患者进行病因分析,采用玻璃体切割术治疗。

结果：术后诊断：173眼中增殖期糖尿病视网膜病变83眼(48.0%),视网膜分支静脉阻塞24眼(13.9%),Eales病13眼(7.5%),视网膜中央静脉阻塞10眼(5.8%),外伤性玻璃体积血9眼(5.2%),视网膜裂孔8眼(4.6%),视网膜脱离9眼(5.2%),增生性玻璃体视网膜病变7眼(4.0%),单纯玻璃体积血4眼(2.3%),视网膜大动脉瘤2眼(1.2%),息肉样脉络膜血管病变2眼(1.2%),脉络膜视网膜炎1眼(0.6%),年龄相关性黄斑病变1眼(0.6%)。不同年龄病因分布不同,术后随访3～15mo,术前与术后视力相比,41眼(23.7%)视力不变,115眼(66.5%)视力提高,17眼(9.8%)视力下降。术后视力与术前相比差异具有统计学意义( P<0.05)。

结论：增殖期糖尿病视网膜病变、视网膜分支静脉阻塞、Eales 病是导致玻璃体积血的主要原因。玻璃体切割联合术手术并发症少,能在一定程度上提高患者视力,是治疗玻璃体积血安全有效的方法。     

玻璃体切割手术治疗Eales病严重并发症的随访观察

目的 观察玻璃体切割手术(PPV)治疗Eales病严重并发症的效果.方法 回顾性分析接受首次PPV治疗的Eales病患者27例30只眼的临床资料.患者中,男性20例,女性7例;年龄15～54岁,平均年龄30.7岁.最佳矫正视力(BCVA)为眼前手动～0.5.根据检查结果将患者分为玻璃体积血组、玻璃体增生机化组、局部视网膜脱离组和广泛视网膜脱离组,分别为3、14、7、6只眼.手术方式为标准PPV,必要时联合巩膜扣带手术、晶状体切除手术、眼内激光光凝、巩膜外冷冻、电凝、增牛膜剥离切断、视网膜切开或切除、气液交换、玻璃体腔注射曲安奈德、膨胀性气体及硅油填充.30只眼分别接受1～8次手术,平均手术次数2.4次.手术后随访观察6个月～10年.对比观察治疗前后BCVA、视网膜复位以及并发症发生情况.结果 末次随访时,BCVA光感～1.5.其中,BCVA≥0.1者24只眼,占80.0%;0.03者1只眼,占3.3%;数指者1只眼,占3.3%;手动者3只眼,占10.0%;光感者1只眼,占3.3%.BCVA提高者22只眼,占73.3%;不变者2只眼,占6.7%;下降者6只眼,占20.0%.手术前后BCVA比较,差异有统计学意义(t=5.132,P＜0.01).广泛视网膜脱离组BCVA较其他3组低,差异均有统计学意义(F=4.570,P均＜0.05);单纯玻璃体积血组视力预后较好,但与玻璃体机化增生组和视网膜局部脱离组比较,差异无统计学意义(P＞0.05).所有患眼PPV手术后第1天视网膜完全复位.末次随访时,无硅油填充视网膜在位24只眼;患者拒绝手术,局部视网膜脱离1只眼;硅油依赖眼5只眼.出现并发症16只眼,占53.3%.结论 玻璃体视网膜手术是治疗Eales病严重并发症的有效手段,手术前存在广泛视网膜脱离的患眼手术后视力预后较差.

Abstract：

Objective To observe the clinical efficacy of vitrectomy on the serious complications of Eales disease. Methods The clinical data of 30 eyes of 27 patients (20 males and 7 females) with Eales disease who underwent vitrectomy were retrospectively analyzed. The age was ranged from 15 to 54 years old, with a mean of 30.7 years. The best corrected visual acuity (BCVA) was ranged from hand movement to 0.5. The patients were divided into the vitreous hemorrhage group (3 eyes ), proliferative vitreoretinopathy group (14 eyes), local retinal detachment group (7 eyes), and wide retinal detachment group (6 eyes) according to the results of examinations. The standard pars plana vitrectomy (PPV) were performed and scleral buckling, lensectomy, endolaser, transscleral cryotherapy/cautery, membrane removal, retinotomy, fluid-air exchange, intravitreal injection of triamcinolone, gas/oil tamponade can be combined if necessary. Those eyes underwent 1 to 8 times (with a mean of 2.4 times) of surgery. The follow-up was ranged from 6 months to 10 years. The BCVA, retinal reattachment, complications before and after surgery was comparatively analyzed. Results At the end of the follow-up, the BCVA was ranged from light perception to 1.5. The BCVA was ＞0.1 in 24 eyes (80.0%) ,0.03 in 1 eye (3.3%), counting finger in 1 eye (3.3%), hand moving in 3 eyes (10.0%) and light perception in 1 eye (3.3%). The BCVA improved in 22 eyes (73.3%), stable in 2 eyes (6.7%) and decreased in 6 eyes (20.0%). The differences are statistically significant between pre- and postoperative BCVA (t=5.132, P＜0.01). The BCVA of wide retinal detachment group was less than other 3 groups (F=4.570, P＜0.05); while the BCVA of vitreous hemorrhage group, proliferative vitreoretinopathy group and local retinal detachment group was the same (P＞0.05). Complete retinal reattachment was achieved in all eyes at the next day after PPV. At the end of the follow-up, retina reattached in 24 eyes without silicone oil tamponade, local retinal detachment occurred in 1 eye (the patient refused further surgery) and silicone oil tamponade-dependant retinal reattachment 5eyes. During the follow-up, 16 eyes (53.3 %) had developed some complications. Conclusions Vitrectomy is an effective way to cure serious complications of Eales disease. The BACV prognosis of patients with wide retinal detachment is poor.     

Vitrektomie bei fortgeschrittenem Morbus Coats

BACKGROUND: Coats disease is a retinal vasculopathy of unknown cause. Untreated cases usually lead to an exudative retinal detachment and rubeosis iridis with secondary glaucoma. Photocoagulation and/or cryotherapy are generally the first interventions in treating the disease. Pars plana vitrectomy may be indicated in cases of vitreous hemorrhage or retinal detachment. METHODS: We performed pars plana vitrectomy in 9 eyes with Coats disease between 1992 and 1999. A retinal detachment was present in 3 cases, and three showed a vitreous hemorrhage. In two cases surgery was indicated because of paramacular localization of the pathological vessels with associated exudations. RESULTS: The two cases with paramacular involvement showed improvement in visual acuity of eight lines. In the remaining cases visual acuity remained within two lines compared to the initial visual acuity. All eyes except one could be saved. CONCLUSION: Pars plana vitrectomy is a useful option in treating advanced Coats disease, especially in cases associated with vitreous hemorrhage or retinal traction.     

Vitrectomy for Complicated Eales Disease

Purpose: We determined the efficacy of pars plana vitrectomy in a series of patients with complicated Eales disease.Methods: Clinical records were reviewed on 47 consecutive patients (49 eyes) who underwent vitrectomy for persistent vitreous hemorrhage (PVH) and/or traction retinal detachment (TRD) with neovascular membranes (NVM). All cases were divided into 4 subgroups of eyes with PVH (without obvious NVM, n = 9), NVD (NVM from the optic disc, n = 10), NVP (peripheral NVM, n = 11), or ENV (extensive NVM and logn-standing TRD,n- 17).Results: The study included 8 women and 39 men with a mean age of 27. 3 years (range, 16 to 50 years). The visual acuity was light perception in 15 eyes, hand motions in 18 eyes, counting fingers in 6 eyes and 0. 02 to 0. 1 in 10 eyes before vitrectomy. Visual acuity improved postoperatively in all eyes with PVH, NVD of NVP with a mean visual acuity of 0. 32 (range,0. 05 to 1. 0,n = 28) except 3 eyes unchanged There was no significant difference ( P > 0. 05 ) in the patien     

Studies on vitrectomy cases associated with complicated branch retinal vein occlusion

PURPOSE: To study the preoperative factors leading to vitrectomy, and to demonstrate the postoperative conditions causing the decreased visual acuity that results from branch retinal vein occlusion (BRVO). METHODS: In 113 patients (114 eyes) with BRVO, the following data were analyzed: age, general complications, distribution of occluding vessels, location of retinal breaks, classification of vitreoretinal pathology, and the number of cases, period from onset of BRVO to vitreous hemorrhage and from vitreous hemorrhage to vitrectomy, number of operations, relationship between posterior vitreous detachment (PVD) and number of operations, preoperative photocoagulations, pre- and postoperative visual acuity, and cases with poor visual outcome. RESULTS: The visual prognosis was much better in cases with vitreous hemorrhage only than in those with proliferative membrane and retinal detachment (P =.0023). Repeated surgeries were needed in the cases where there was only partial PVD (P =.0029). Macular disorders and optic nerve atrophy were the main causes of postoperative visual acuity < 0.1. CONCLUSIONS: Early vitrectomy before development of vitreo retinal proliferation and retinal detachment, especially in cases where there is only partial PVD, seems to be essential for case management and treatment to attain better visual acuity for the patient.     

Long-term results of vitreous surgery for proliferative diabetic retinopathy

In order to study long-term anatomical and functional results the authors evaluated the data from 260 patients who underwent pars plana vitrectomy for proliferative diabetic retinopathy. Indications for surgery were: vitreous hemorrhage, 68 eyes (26.2%); vitreous hemorrhage & tractional retinal detachment, 84 eyes (32.3%); tractional retinal detachment, 82 eyes (31.5%); and combined tractional-rhegmatogenous retinal detachment, 26 eyes (10%). In 118 eyes vitreoretinal surgery was combined with silicone-oil tamponade.The retina was completely attached posterior to a scleral buckle in 251 eyes (96%) at the time of the last examination. After a follow-up period of at least 12 months in a group of patients with vitreous hemorrhage, visual acuity improved in 88% of the eyes. Visual acuity was better than 0.5 in 31% of eyes. In group of eyes with nonresorbing vitreous hemorrhage & tractional retinal detachment visual acuity improved in 52% of eyes. Visual acuity improved in 76% of eyes with tractional retinal detachment and in 81% of eyes with combined tractional & rhegmatogenous retinal detachment. When comparing the latest postoperative visual acuity to visual acuity after three months postoperatively, visual acuity was unchanged in 88%, in 10% it became worse and in 3 cases (1%) became better. In the postoperative period, recurrent vitreous hemorrhage occurred in 33 (13%) eyes, reproliferation in 12 eyes.Cataract developed in 45 of 168 phakic eyes. If postoperative visual acuity before cataract formation was good, extracapsular cataract extraction with posterior chamber intraocular lens implantation was performed. Otherwise simple intra or extracapsular cataract extraction was performed. In 19 cases cataract operation was performed together with silicone oil extraction. Neovascular glaucoma developed postoperatively in 15 eyes (6%). Retinal detachment occurred postoperatively in 21 eyes (8%). In 15 eyes the retina was successfully reattached after additional operations.     

Vitrectomy for nondiabetic vitreous hemorrhage. Not associated with vascular disease

Forty-nine consecutive eyes with nonclearing vitreous hemorrhage not associated with retinal or choroidal vascular disease underwent vitrectomy. Etiologies included vitreous hemorrhage during anterior segment surgery (22 eyes), blunt trauma (8 eyes), retinal tears with and without retinal detachment (8 eyes), Terson's syndrome (2 eyes), avulsed retinal vessel (1 eye), and idiopathic cases (8 eyes). The final visual acuity improved in 48 eyes (98%). Follow-up was 6-91 months (mean, 20 months). Of the 49 eyes, 40 eyes (82%) had a best postoperative visual acuity of 20/100 or better, 31 eyes (63%) had visual acuity of 20/40 or better, and 12 eyes (24%) had visual acuity of 20/20. The major complications included intraoperative iatrogenic retinal breaks (5 eyes), postoperative progressive cataract (7 eyes), late retinal detachment (4 eyes) and recurrent vitreous hemorrhage (2 eyes). The major complication associated with later visual loss was progressive cataract.     

Different outcome among eyes with proliferative diabetic retinopathy indicated for vitrectomy

PURPOSE: The postoperative outcome was evaluated in each group of surgical indications of vitreous surgery for proliferative diabetic retinopathy (PDR), to investigate the factors responsible for postoperative visual prognosis. METHODS: Primary vitrectomy was performed in 119 eyes of 92 patients with PDR. Average postoperative follow-up period was 19 months. The indications for vitrectomy included vitrous hemorrhage in 58 eyes, macular tractional retinal detachment in 17 eyes, extramacular tractional retinal detachment in 10 eyes, macular heterotopia in 11 eyes, and progressive fibrovascular proliferation in the posterior fundus in 23 eyes. RESULTS: The visual acuity finally improved by 2 lines or more in 91 eyes (77%), remained unchanged in 10 eyes (8 %), and decreased by 2 lines or more in 18 eyes (15%). Final postoperative visual acuity was significantly better in cases of vitreous hemorrhage or progressive fibrovascular proliferation in the posterior fundus than in others. Preoperative rubeosis iridis and macular tractional retinal detachment were probably responsible for the final visual impairment, and intraocular tamponade affected the difference in visual prognosis between the groups of surgical indication. Multivariate analysis in all cases revealed that factors influencing visual outcome were preoperative rubeosis iridis and anemia. CONCLUSION: Rubeosis iridis and macular tractional retinal detachment were prognostic factors of the surgery. Vitrectomy for PDR may be effective in improving postoperative visual acuity if performed in the early stage of progressive fibrovascular proliferation in the posterior fundus after sufficient retinal photocoagulation.