308

Vitrectomy was first introduced by Robert Machemer and plays an important role in the treatment of proliferative diabetic retinopathy (PDR).1 The major indications for vitrectomy in patients with diabetes are a nonclearing vitreous hemorrhage, macula- involving or threatening tractional retinal detachment,

or a combined tractional and rhegmatogenous retinal detachment.2 Although pars plana vitrectomy (PPV) has been one of the major effective treatment meth-ods for complications of PDR after the successful operation, the major vision-threatening complication associated with this procedure is recurrent vitreous hemorrhage.3,4 Brown et al.5 reported the primary cause for reoperation in PDR was recurrent vitreous hemorrhage, with 14% of the rebleeding eyes regress-ing to no light perception. Other reports indicated that of all cases with recurrent vitreous hemorrhage, one-third ultimately require air-fluid exchange or

Current Eye Research, 35(4), 308–313, 2010Copyright © 2010 Informa Healthcare USA, Inc.ISSN: 0271-3683 print/ 1460-2202 onlineDOI: 10.3109/02713680903572491

ORIGINAL ARTICLE

Reasons for and Management of Postvitrectomy Vitreous Hemorrhage in

Proliferative Diabetic Retinopathy

Hua Yan, Jing Cui, Yingjuan Lu, Jinguo Yu, Song Chen, and Yinghai Xu

Department of Ophthalmology, Tianjin Medical University General Hospital, Tianjin, China

ABSTRACT

Objective: To analyze the reasons for postvitrectomy vitreous hemorrhage in proliferative diabetic retinopathy (PDR), and to evaluate the effects of retreatment.Design: Retrospective, nonrandomized, observational case series.Methods: Three hundred and fifteen eyes of 302 consecutive patients underwent primary standard three-port vitrectomy with 20-gauge instruments for complications of PDR from 2000 to 2006. One hundred and forty-two patients were male, and 160 were female. The age ranged from 38 to 72 years with a mean of 56 years. There were 32 eyes which developed postvitrectomy vitreous hemorrhage during follow-up. The mean follow-up was 12 months with a range from 3 to 48 months.Results: Of 315 eyes with PDR and receiving pars plana vitrectomy, 32 eyes had postvitrectomy vitreous hemorrhage. The onset of recurrent vitreous hemorrhage ranged from 1 to 210 days with an average of 51 days. The reasons for postvitrectomy vitreous hemorrhage in PDR mainly included fibrovascular ingrowth at sclerotomy sites (9 eyes), residual or recurrent neovascular membrane on the optic nerve (6 eyes), insufficient retinal photocoagulation (7 eyes), residual and recurrent epireti-nal proliferative membrane (3 eyes), retinal vein occlusion (2 eyes), postoperative low intraocular pressure (2 eyes), and ocular trauma (3 eyes). The visual acuity increased in 31 eyes (96.88%), and decreased in 1 eye (3.12%) after retreatment. The postoperative complications following the treat-ment of recurrent vitreous hemorrhage mainly included posterior synechia of the iris (3 eyes), nucleus sclerosis (18 eyes), and delayed healing of corneal epithelium (3 eyes).Conclusion: Vitrectomy is a safe and effective method for treating PDR. Appropriate and complete analysis of postvitrectomy vitreous hemorrhage can significantly improve the primary treatment effects for PDR.

KEYWORDS: Hemorrhage; Proliferative diabetic retinopathy; Reason; Retreatment; Vitrectomy; Vitreous

Received 16 February 2009; accepted 18 December 2009

Correspondence: Dr. Hua Yan, M.D., Ph.D., Department of Oph-thalmology, Tianjin Medical University General Hospital, No. 154, Anshan Road, Tianjin 300052, China. E-mail: phuayan2000@yahoo.com

16 February 2009

18 December 2009

© 2010 Informa Healthcare USA, Inc.

2010

Current Eye Research

0271-36831460-2202

10.3109/02713680903572491

35

308313

4

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

Reasons for and Management of Postvitrectomy Vitreous Hemorrhagethy 309

© 2010 Informa Healthcare USA, Inc.

vitreous lavage to restore vision.5–10 This complication may slow visual recovery.6,7

The cause of rebleeding after diabetic vitrectomy is often unknown, which, if not treated aggressively, can lead to phthisis bulbi. Sources of rebleeding after vitrectomy for PDR probably include neovasculariza-tion growing from the surgical sclerotomy sites, which is one of the important changes noted in various histopathologic studies, residual or recurrent fibro-vascular membrane on the optic nerve, insufficient retinal photocoagulation, residual or recurrent neo-vascularization on the retina, retinal vein occlusion, postoperative low intraocular pressure, and ocular trauma.11–14 Retinal ischemia is the most important factor contributing to the development of fibrovas-cular tissue in eyes with diabetic retinopathy after vitrectomy. The purpose of our study is to review our diabetic population and analyze the reasons for and management of recurrent vitreous hemorrhage after diabetic vitrectomy.

MATERIALS AND METHODS

Patients

From 2000 to 2006, 315 eyes of 302 consecutive patients underwent primary standard three-port vitrectomy for complications of PDR at our hospital, and these cases were studied for retrospective chart review. One hundred and forty-two patients were male and 160 were female. The age ranged from 38 to 72 years with a mean of 56 years. Excluded were patients with incomplete records and those who had the primary vitrectomy performed elsewhere. The Chinese Diabetic Retinopathy Clinical Staging System was used in this study.15

Methods

Definition of Recurrent Vitreous HemorrhageWe defined recurrent vitreous hemorrhage as vitre-ous hemorrhage that was extensive enough to cause decreased visual acuity and loss of details of retinal vessel under indirect ophthalmoscopy after a postoper-ative period in which the vitreous was clear. Persistent vitreous hemorrhage was not included.

Surgical IndicationThe indication for the first surgery was severe PDR, which included severe unabsorbed vitreous hemorrhage, tractional retinal detachment complicated with macular traction, aggressive fibrovascular prolif-eration, and nonclearing vitreous hemorrhage com-

plicated with cataract. The indication for the second surgery was recurrent vitreous hemorrhage that had not absorbed for at least 1 month.

Pre and Postoperative ExaminationsPre and postoperative examinations mainly included visual acuity, slit-lamp examination, indirect oph-thalmoscopy, intraocular pressure, and B-scan ultra-sound.

Surgical Procedures

All the operations were performed by one surgeon in this study using a 20-gauge vitrectomy system. Informed consent was obtained from each patient. Preoperative, intraoperative, and postoperative data were collected for each patient, which included intraoperative diagnosis, endolaser, cryotherapy, and internal gas or silicone oil temponade. Data regarding time, duration, and treatment of recurrent vitreous hemorrhage, duration of postoperative follow-up, and the additional surgical procedures were also compiled (Table 1).

The eye was anesthetized retrobulbarly and peribulbarly by 2% lidocaine. The exact circum-ferential location of the sclerotomies for the active instrument, light pipe, and infusion line were noted. The cataract was extracted by phacoemulsi-fication if the lens had significant opacity. After the vitreous hemorrhage was resected, the epiretinal proliferative membrane was peeled and panretinal photocoagulation was performed. C3F8 or silicone oil was instilled in the eye if it was needed. The sclerotomies were closed at the end of vitrectomy. Topical steroid, mydriatics, and antibiotic eye drops were administered for all eyes postoperatively. Rou-tine ocular examinations were performed 1, 3, 5, and 7 days after the surgery.16

Management of Recurrent Vitreous HemorrhageIf the recurrent vitreous hemorrhage was not too exten-sive in the operated eye, the two eyes were patched, a semireclining position was adopted, and the Chinese traditional medicine (Sanqi) for absorbing vitreous hemorrhage was taken. If the vitreous hemorrhage was nonclearing, and it was not absorbed significantly dur-ing a period of 1 month of follow-up, a reoperation was recommended. The reason for recurrent hemorrhage was evaluated intraoperatively, and the appropriate management, such as cryotherapy for the fibrovascular ingrowth at the sclerotomy sites, removal of residual vascular membrane, cautery of the vessel stump, or supplementation of retinal photocoagulation, were performed respectively (Table 1).

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

310 Hua Yan et al.

Current Eye Research

A paired t-test was performed to compare the pre-operative intraocular pressure with the postoperative one. A P value of <0.05 was considered statistically significant.

RESULTS

The duration of diabetes within this population was 1.5 to 12 years, with an average of 5 years. Five patients were Type I diabetics, and 297 were Type II diabet-ics. Of 315 eyes that underwent primary vitrectomy, 23 eyes were severe unabsorbed vitreous hemorrhage, 133 eyes had tractional retinal detachment complicated with macular traction, 102 eyes had aggressive fibro-vascular proliferation, and 57 eyes had nonclearing vitreous hemorrhage complicated with cataract. Of 32 eyes with recurrent vitreous hemorrhage after primary diabetic vitrectomy, 5 eyes had diabetic retinopathy stage IV, 8 eyes had stage V, and 19 eyes had stage VI. The duration of vitreous hemorrhage before the initial vitrectomy was 1 to 12 months, with an average of 5 months. The preoperative intraocular pressure ranged from 12 to 17 mmHg with a mean of 14 mmHg, and the postoperative intraocular pressure ranged from 8 to 18 mmHg with a mean of 15 mmHg. There was no significant difference in intraocular pressure before and after the initial vitrectomy (P > 0.05). The average follow-up was 12 months with a range from 3 to 48 months. The visual acuity before the initial vitrectomy ranged from light perception to 0.2.

Recurrent Vitreous Hemorrhage

Of all 315 eyes with PDR and receiving PPV, 32 eyes had postvitrectomy vitreous hemorrhage. The rebleeding rate was 10% in patients with PDR after vitrectomy. Of 32 eyes with postvitrectomy vitreous hemorrhage,

3 eyes had been treated for severe unabsorbed vitreous hemorrhage (3/23, 13%), 19 eyes had been treated for tractional retinal detachment complicated with macu-lar traction (19/133, 14%), 8 eyes had been treated for aggressive fibrovascular proliferation (8/102, 8%), and 2 eyes had been treated for nonclearing vitreous hemorrhage complicated with cataract (2/57, 4%). The onset of recurrent vitreous hemorrhage ranged from 1 to 210 days, with an average of 51 days after the initial vitrectomy.

Reasons for Postvitrectomy Vitreous Hemorrhage (Table 1)

Fibrovascular Ingrowth at Sclerotomy SitesIn this study, 9 eyes had the occurrence of rebleeding and presence of fibrovascular proliferation at the scle-rotomy sites 1.5 to 3 months with a mean of 2 months postoperatively. The fibrovascular ingrowth cases were actually anterior hyaloidal fibrovascular pro-liferation. For those with fibrovascular ingrowth, all cases required another vitreous surgery and additional cryotherapy, and 1 eye eventually required silicone oil infusion.

Residual or Recurrent Neovascular Membrane on the Optic NerveIn this study, 3 eyes had recurrent hemorrhage induced by residual neovascular membrane on the optic nerve and 3 eyes induced by recurrent neovascular mem-brane on the optic nerve 14 to 30 days postoperatively. The presence of neovascular membrane on the optic nerve necessitated reoperation, including membrane peeling and excision of the fibrovascular proliferation with cautery of the stump. The internal tamponade with silicone oil was required in one eye. Finally, one eye had rebleeding after the second vitrectomy surgery, and the patient refused the further treatment.

TABLE 1 Reasons for and management of postvitrectomy vitreous hemorrhage in PDRReasons for recurrent vitreous hemorrhage Eyes (%)

Onset of recurrent vitreous hemorrhage Retreatment Results

Fibrovascular ingrowth at sclerotomy sites

9 (28.13) 2 (1.5–3) months Cryotherapy Recovery

Residual or recurrent neovascular membrane on the optic nerve

6 (18.75) 21 (14–30) days Membrane peeling and cautery of fibrovascular stump

Recurrent vitreous hemorrhage (1 eye)

Insufficient retinal photocoagulation 7 (21.88) 2 (1.5–3) months Air-fluid exchange and photocoagulation

Recovery

Residual and recurrent epiretinal proliferative membrane

3 (9.38) 1.5 (1–2) months Membrane peeling and photocoagulation

Recovery

Retinal vein occlusion 2 (6.25) 4 (2–7) months Air-fluid exchange and photocoagulation

Recovery

Postoperative low intraocular pressure 2 (6.25) 3.5 (3–4) months Silicone oil filling RecoveryOcular trauma 3 (9.38) 2 (1–5) days Conservative treatment RecoveryNote: PDR = Proliferative diabetic retinopathy.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

Reasons for and Management of Postvitrectomy Vitreous Hemorrhagethy 311

© 2010 Informa Healthcare USA, Inc.

Insufficient Retinal PhotocoagulationAlthough all eyes had extensive intraoperative laser photocoagulation of greater than two quadrants, 7 eyes had recurrent hemorrhage induced by insufficient retinal photocoagulation due to delayed supplemental retinal photocoagulation. Insufficient retinal photoco-agulation and irregular retinal leakage on the periphery were proven by fluorescein angiography after air-fluid exchange treatment in all 7 eyes, and supplementary laser photocoagulation was required.

Residual and Recurrent Epiretinal Proliferative MembraneOne eye had rebleeding induced by residual epiretinal proliferation and in 2 eyes rebleeding was induced by retinal reproliferation 1 to 2 months postoperatively. Among the 3 eyes with evidence of proliferative mem-brane and rebleeding, membrane surgery and laser treatment with internal tamponade of C3F8 were suc-cessful in clearing the media.

Retinal Vein OcclusionOne eye had branch retinal vein occlusion, and 1 eye had central retinal vein occlusion 2 to 7 months postoperatively. Air-fluid exchange was successful in clearing the media in 2 cases, and the supple-mental retinal photocoagulation was performed intraoperatively.

Postoperative Low Intraocular PressureTwo eyes had the occurrence of rebleeding and low intraocular pressure by 8 mmHg 3–4 months postop-eratively, and the internal tamponade with silicone oil was required eventually.

Ocular TraumaThree eyes had recurrent vitreous hemorrhage induced by indirect ocular trauma 1–5 days postoperatively, and the rebleeding was cleared with the conservative treatment during a period of 3–4 weeks.

Visual Acuity

Visual acuity increased in 31 eyes (96.88%), and decreased in 1 eye (3.12%) after retreatment. Before the retreatment, the visual acuity was 0.1 in 4 (12.5%) of 32 eyes, between 0.05 and 0.1 in 8 (25%) of 32 eyes, between 0.02 and 0.05 in 16 (50%) of 32 eyes, and worse than 0.02 in 4 (12.5%) of 32 eyes. The visual acuity after the retreatment was 1.0 or better in 4 (12.5%) of 32 eyes, between 0.3 and 1.0 in 24 (75%) of 32 eyes, and worse than 0.3 in 4 (12.5%) of 32 eyes (Figure 1).

Complications

The complications after the second surgery for the treatment of recurrent vitreous hemorrhage mainly included posterior synechia of the iris (3 eyes), nucleus sclerosis (18 eyes), and delayed healing of corneal epi-thelium (3 eyes). Of all the rebleeding cases, only 1 eye had active bleeding at the end of the follow-up.

DISCUSSION

The introduction of the pars plana approach for vit-rectomy has greatly increased the success rate of sur-gical treatment for complicated retinal problems of various origins.17 PDR is a major indication for PPV among the complex vitreoretinopathies. Several stud-ies have shown that patients undergoing vitrectomy for a nonclearing diabetic vitreous hemorrhage have approximately 50–60% chance of achieving 20/100 or better vision.18–20 Among the postoperative complica-tions in treating PDR with vitrectomy, rebleeding is a cause of severe vision decrease. We have reviewed our cases undergoing vitrectomies in a diabetic population to better understand the causes for recurrent vitreous hemorrhage. The rate of rebleeding after primary vit-rectomy for PDR was 10% in our study, and it was

0.00%10.00%20.00%30.00%40.00%50.00%60.00%70.00%80.00%

<0.02 0.02–0.05 0.05–0.1 0.1 0.1–0.3 0.3–1.0 >=1.0

Per

cent

age

Visual acuity

before surgeryafter surgery

FIGURE 1 The changes of visual acuity after the treatment of postvitrectomy vitreous hemorrhage in PDR.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

312 Hua Yan et al.

Current Eye Research

lower than that reported by Yeh.21 The correct manage-ment in recurrent vitreous hemorrhage after vitrectomy should be tailored to each individual patient.

Through direct observation during reoperation, our results and several other studies have confirmed the high correlation of fibrovascular ingrowth at the sclerotomy site and recurrent vitreous hemor-rhage after PPV for PDR.21–23 Although prominent fibrovascular ingrowth at sclerotomy sites after PPV has been noted to be relative to various conditions, the most important factor related to fibrovascular ingrowth is PPV after PDR.22–24 Pülborn et al.25 studied a traumatized eye enucleated after failed vitrectomy for an intraocular foreign body. They noted severe ingrowth of fibrovascular tissue along various planes, all originating from the sclerotomy sites, and attributed a major role to the incomplete removal of vitreous. The ciliary epithelia seem to play an important role in the formation of fibrovascular ingrowth.14 Therefore, the changes occurring at the pars plana incisions are induced by intraocular maneuvers that lead to postop-erative problems.11–13,25 On the other hand, the combina-tion of surgical wounds, vitreous scaffold, angiogenic factor produced by ischemic retina, various growth factors in fresh or old blood, and cytokines released by postoperative inflammation provides a suitable environment for fibrovascular proliferation around the sclerotomy sites. The finding of fibrovascular prolifera-tion at the sclerotomy in eyes with recurrent vitreous hemorrhage will indicate the need of aggressive surgi-cal intervention. Otherwise, an air-fluid exchange or simple observation may suffice. In our study, 32 eyes experienced recurrent hemorrhage after vitrectomy for PDR, and fibrovascular proliferation was found in 9 eyes (28.13%) intraoperatively. The vitreous recurrent rate was relatively lower than that reported by Hersh-berger et al.24

The retained fibrovascular tissue or recurrent fibro-vascular membrane on the retina or optic nerve may be another potential source of recurrent vitreous hemor-rhage. The intraoperative hemorrhage from membrane peeling on the retina or optic nerve could be stanched by elevating the intraocular pressure. However, if the fibrovascular membrane excision and cautery of fibro-vascular stump were not performed completely, the recurrent vitreous hemorrhage might occur as a result of re-establishment of normal intraocular pressures postoperatively.

In some cases, we completely removed the fibrovas-cular tissue, however recurrent bleeding occurred with-out evidence of fibrovascular ingrowth or proliferative membrane. In these cases, we finally confirmed that the intraoperative retinal photocoagulation was insuf-ficient by the examination of fluorescence angiography during follow-up after air-fluid exchange treatment.

The intraoperative panretinal photocoagulation alone may not be intensive enough to inhibit the production of angiogenic factors fully. Therefore, supplementary retinal photocoagulation in the far peripheral area may further reduce the amount of angiogenic factors and minimize the chance of proliferation of fibrovascular tissue. In our study, 7 eyes did not receive supplemen-tary photocoagulation in time leading to recurrent vitreous hemorrhage.

Postoperative retinal vein occlusion is another impor-tant reason for postvitrectomy vitreous hemorrhage in PDR. Although the intraoperative maneuvers, such as membrane peeling, photocoagulation, and cautery of fibrovascular stump, were performed successfully, postoperative retinal vein occlusion could occur because of postoperative unstable blood pressure and abnormal blood biochemistry in diabetics. The reason for recurrent vitreous hemorrhage induced by retinal vein occlusion is the formation of retinal neovascularization. All patients should receive an examination with fundus fluorescein angiography and use of further retinal photocoagulation if necessary when the media is clear postoperatively.

Postoperative low intraocular pressure is a severe complication in patients undergoing vitrectomy for PDR. The possible reasons for postoperative low intraocular pressure are leakages from sclerotomy sites. In cases with low intraocular pressure, the retinal blood vessel is outstretched, and the bleeding may be occurred in the disordered vessel. In our study, 2 eyes with postoperative low intraocular pressure had recur-rent vitreous hemorrhage 3–4 months postvitrectomy for PDR.

The primary reason for poor visual outcome after initial vitrectomy is the presence of a recurrent hemorrhage in this study. Although a poor outcome is possible after the first vitrectomy for diabetic retinopa-thy, we think that retreatment for recurrent hemorrhage can have a better visual outcome or favorable anatomic outcome while maintaining ambulatory vision from our results.

In conclusion, our study is a retrospective one, and contains certain limitations because of incomplete clinical observation and shorter follow-up. It is not possible to exactly identify the reasons for postopera-tive vitreous hemorrhage in all cases, and it may not always be possible to find a definite cause, or a single cause. The reasons for postoperative vitreous hemor-rhage in this study only indicated the main reason in each case. Although there may still be other unknown factors involved in the development of postvitrec-tomy vitreous hemorrhage, this study indicates that the presence of fibrovascular ingrowth at sclerotomy site, residual or recurrent neovascular membrane on the optic nerve, insufficient retinal photocoagulation,

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

Reasons for and Management of Postvitrectomy Vitreous Hemorrhagethy 313

© 2010 Informa Healthcare USA, Inc.

residual or recurrent epiretinal proliferation, retinal vein occlusion, low intraocular pressure, and ocular trauma seem to be the primary reasons for postopera-tive vitreous hemorrhage. In this study, some postop-erative vitreous hemorrhage that were not serious might spontaneously resolve without treatment within 1 month postoperatively, and were not included in the analysis of recurrent vitreous hemorrhage. Based on previous results and our results, we believe if a recur-rent vitreous hemorrhage persists for 1 month, then repeat vitrectomy offers a very good potential for res-toration of visual function.

Declaration of interest: The authors have stated that they do not have a significant financial interest or other relationship with any product manufacturer or pro-vider of services discussed in this article. The authors also do not discuss the use of off-label products, which include unlabeled, unapproved, or investiga-tive products or devices. The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.

REFERENCES

[1] Machemer R, Parel JM, Buettner H. A new concept for vitreous surgery: Instrumentation. Am J Ophthalmol. 1972;74:1022–1033.

[2] Lewis H, Abrams GW, Blumenkranz MS, et al. Vitrectomy for diabetic macular traction and edema associated with posterior hyaloidal traction. Ophthalmology. 1992;99:753–759.

[3] Yang CM. Surgical treatment for diabetic retinopathy: 5-year experience. J Formos Med Assoc. 1998;97:477–484.

[4] Michels RG, Rice TA, Rice EF. Vitrectomy for diabetic vitreous hemorrhage. Am J Ophthalmol. 1983;95:12–21.

[5] Brown GC, Tasman WS, Benson WE, et al. Reopera-tion following diabetic vitrectomy. Arch Ophthalmol. 1992;110:506–510.

[6] Blankenship GW. Management of vitreous cavity hemorrhage following pars plana vitrectomy for diabetic retinopathy. Ophthalmology. 1986;93:39–44.

[7] Blumenkranz MS, Gardner T, Blankenship GW. Fluid-gas exchange and photocoagulation after vitrectomy: Indica-tion, technique, and result. Arch Ophthalmol. 1986;104:291–296.

[8] Han DP, Murphy ML, Mieler WF, et al. Outpatient air-fluid exchange for severe postvitrectomy diabetic vitreous hemorrhage: Long-term results and complications. Retina. 1991;11:309–314.

[9] Schachat AP, Oyakawa KT, Michels TG, et al. Complications of vitreous surgery for diabetic retinopathy. II. Postoperative complications. Ophthalmology. 1983;90:522–530.

[10] Mieler WF, Wolf MD. Management of postvitrectomy dia-betic vitreous hemorrhage. In: Lewis H, Ryan SJ (Eds.). Medical and Surgical Retina: Advances, Controversies, and Management. St. Louis: Mosby, 1994; pp. 330–340.

[11] Buettner H, Machemer R. Histopathologic findings in human eyes after pars plana vitrectomy and lensectomy. Arch Ophthalmol. 1977;95:2029–2033.

[12] Foos RY, Kreiger AE, Nofsinger K. Pathologic study follow-ing vitrectomy for proliferative diabetic retinopathy [case report]. Retina. 1985;5:101–106.

[13] Kreiger AE. Wound complications in pars plana vitrectomy. Retina. 1993;13:335–344.

[14] Koch FH, Kreiger AE, Spitznas M, et al. Pars plana incisions of four patients: Histopathology and electron microscopy. Br J Ophthalmol. 1995;79:486–493.

[15] Ocular Fundus Disease Study Group of Chinese Society of Ophthalmology. Criteria for diabetic retinopathy staging. Chinese J Ocular Fundus Dis. 1985;1:42.

[16] Yan H, Han Q, Xu YH, et al. The effect of vitrectomy on blood sugar in patients with diabetes mellitus. Chinese J Ocular Fundus Dis. 2001;17:236.

[17] Machemer R, Norton EWD. A new concept of vitre-ous surgery. 3. Indications and results. Am J Ophthalmol. 1972;74:1034–1056.

[18] The Diabetic Retinopathy Vitrectomy Study Research Group. Early vitrectomy for severe proliferative diabetic retinopathy in eyes with useful vision: Results of a random-ized trial. Report 3. Ophthalmology. 1988;95:1307–1320.

[19] Benson WE, Brown GC, Tasman W, et al. Complications of vitrectomy for nonclearing vitreous hemorrhage in diabetic patients. Ophthalmic Surg. 1988;19:862–864.

[20] Thompson JT, DeBustros S, Michels RG, et al. Results and prognostic factors in vitrectomy for diabetic vitreous hemor-rhage. Arch Ophthalmol. 1987;105:191–195.

[21] Yeh PT, Yang CM, Yang CH, et al. Cryotherapy of the ante-rior retina and sclerotomy sites in diabetic vitrectomy to prevent recurrent vitreous hemorrhage: An ultrasound biomicroscopy study. Ophthalmology. 2005;112:2095–2102.

[22] Hotta K, Hirakata A, Ohi Y, et al. Ultrasound biomicros-copy for examination of the sclerotomy site in eyes with proliferative diabetic retinopathy after vitrectomy. Retina. 2000;20:52–58.

[23] Bhende M, Agraharam SG, Gopal L, et al. Ultrasound biomicroscopy of sclerotomy sites after pars plana vitrec-tomy for diabetic vitreous hemorrhage. Ophthalmology. 2000;107:1729–1736.

[24] Hershberger VS, Augsburger JJ, Hutchins RK, et al. Fibro-vascular ingrowth at sclerotomy sites in vitrectomized dia-betic eyes with recurrent vitreous hemorrhage. Ultrasound biomicroscopy findings. Ophthalmology. 2004;111:1215–1221.

[25] Pülborn G, Teichmann KD, Teichmann I. Intraocular fibrous proliferation as an incisional complication in pars plana vit-rectomy. Am J Ophthalmol. 1977;83:810–814.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.