Prognostic Indicators for No Light Perception After Open-Globe Injury: Eye Injury Vitrectomy Study

KANG FENG, YUN TAO HU, AND ZHIZHONG MA To describe ocular characteristics, surgical interventions, and anatomic and visual outcomes of traumatized eyes with no light perception (NLP) following open-globe injury and to investigate prognostic predictors for NLP cases after open-globe injury. DESIGN: Interventional case series study. METHODS: Thirty-three traumatized eyes with NLP were selected from the Eye Injury Vitrectomy Study database, a hospital-based multicenter prospective cohort study. Inclusion criteria were NLP cases following openglobe injury with outcomes of anatomic restoration, phthisis bulbi, or enucleation. Exclusion criteria were cases with missing records, undergoing vitrectomy after injury at nonparticipating hospitals, direct optic head injury, endophthalmitis, and hypotonous or silicone oil sustained eyes. All cases underwent vitreoretinal surgery or enucleation after exploratory surgery and were followed up for at least 6 months. Two outcomes were assessed: favorable outcome (anatomically restored eye globes with light perception [LP] or better vision) and unfavorable outcome (NLP, phthisis bulbi, or enucleation). RESULTS: The following 7 risk factors were signicant between the 2 groups: rupture (P .021); open globe III (P .046); scleral wound >10 mm (P .001); ciliary body damage (P < .001); severe intraocular hemorrhage (P .005); closed funnel retinal detachment or retinal prolapse (P .005); and choroidal damage (P .001). CONCLUSIONS: These 7 risk factors are possible predictors of poor prognosis. Traumatized eyes with NLP can be anatomically restored with LP or better vision if vitreoretinal surgery is attempted, and a favorable anatomic and visual outcome is increased by having a decreased number of these risk factors. (Am J Ophthalmol 2011;152:654 662. 2011 by Elsevier Inc. All rights reserved.)
PURPOSE:

mia, many ophthalmologists usually select primary enucleation for traumatized eyes with NLP.7,8 With the advancement of vitreoretinal surgery, some traumatized eyes with NLP, which would have previously been enucleated, are now saved and recover light-perception (LP) or better vision. When performing exploratory surgery on traumatized NLP cases, clinicians can now make a choice between vitreoretinal surgery and enucleation according to the ocular ndings. However, few published reports describe in detail the intraocular characteristics, surgical interventions, and outcomes of traumatized eyes with NLP following open-globe injury. Further, because of a lack of prognostic indicators, there are no decision-making guidelines that can be used during exploratory surgery in order to make an accurate assessment of prognosis. In this study, we described the ocular characteristics, surgical interventions, and the anatomic and visual outcomes of 33 traumatized NLP cases following open-globe injury and tracked the prognostic indicators for traumatized eyes with NLP vision.

METHODS
CASES IN THIS STUDY WERE SELECTED FROM THE EYE IN-

O
654

PEN-GLOBE INJURY IS ONE OF THE MOST COMMON

worldwide causes of monocular visual impairment and blindness.1 4 Open-globe injury that causes no light perception (NLP) typically carries a dismal prognosis.5,6 Considering the risk of sympathetic ophthal-

Accepted for publication Apr 14, 2011. From the Peking University Eye Center, Peking University Third Hospital, Beijing, Peoples Republic of China. Inquiries to Zhizhong Ma, Peking University Eye Center, Peking University Third Hospital, Huayuan North Street 49, Haidian, Beijing 100191, Peoples Republic of China; e-mail: puh3_yk@bjmu.edu.cn

jury Vitrectomy Study database, which began in January 1997. The Eye Injury Vitrectomy Study is a hospitalbased multicenter prospective cohort study whose purpose is to investigate the epidemiology, intervention of vitreoretinal surgery, and prognosis of severe eye injury. Six tertiary referral hospitals in China have successively participated in it. All injured patients in the database consented to enroll in the Eye Injury Vitrectomy Study. Its inclusion criteria were patients who suffered from severe eye injury and were treated with vitreoretinal surgery, enucleation, or evisceration. Exclusion criteria were patients with an eye injury who did not need vitreoretinal surgical intervention. After patient demographics were obtained at admission, all initial ophthalmic examinations of the injured patients were conducted and reviewed by the chief of staff, and ndings were conrmed by the chief surgeon before surgery. Information including patient age, sex, involved eye, best-corrected visual acuity (BCVA) after injury, and open-globe repair was recorded on a Register of Eye Injury form. Type and zone of injury conform to the recommendations of the United States Eye Injury Registry
RIGHTS RESERVED. 0002-9394/$36.00 doi:10.1016/j.ajo.2011.04.004

2011 BY

ELSEVIER INC. ALL

and the International Society of Ocular Trauma (Birmingham Eye Trauma Terminology [BETT9]) and A System for Classifying Mechanical Injuries of the Eye.10 Cornea, sclera, iris, lens, ciliary body, retina, choroid, intraocular foreign body (IOFB), intraocular hemorrhage, direct injury of optic nerve head, and endophthalmitis were examined and assessed during open-globe repair or exploratory surgery and recorded as baseline information on the Register of Eye Injury form by the chief surgeon. In addition, details of the vitreoretinal surgery and any subsequent surgical procedures and nonsurgical interventions were also recorded. Follow-up information, including the follow-up period, BCVA and intraocular pressure (IOP) at the last visit, tamponade of vitreous cavity, and anatomic outcome was recorded at outpatient follow-up after no less than 6 months. The baseline and follow-up sheets were collected and data were input in the Eye Injury Vitrectomy Study electronic database using Epidata (The EpiData Association, Odense, Denmark). A cross-check for errors was conducted by 2 data entry clerks. If a variable could not be identied or was missing in the record, the data were excluded from this cohort. In this study, the inclusion criteria were NLP cases following open-globe injury. As of December 31, 2009, there were 72 NLP eyes post open-globe injury in the Eye Injury Vitrectomy Study database. These injured eyes did not include cases of missing records, ones that underwent vitreoretinal surgery after injury at nonparticipating hospitals, and those with direct optic head injury, endophthalmitis, and a follow-up period of less than 6 months. The exclusion criteria were hypotonous or silicone oil sustained cases. In total, 33 traumatized eyes with NLP (33 patients) met the studys outcome criteria of anatomically restored eyes, phthisis bulbi, or enucleation.
STUDY TERM DEFINITIONS: NLP. NLP vision was determined by an examination using an indirect ophthalmoscope with the highest-intensity light while the fellow eye was fully occluded. All injured patients who had documented NLP in the emergency room or other nonparticipating hospitals were referred to and examined by at least 2 senior trauma staff ophthalmologists after admission. As a nal step, NLP vision was conrmed by the chief surgeon prior to exploratory surgery.

Lens or iris extrusion. Lens or iris is fully prolapsed out of the globe (missing or under the conjunctiva) at the time of injury, which was conrmed during open-globe repair or exploratory surgery. Intraocular hemorrhage. In some patients, almost all of the vitreous was lost when the open-globe injury occurred, especially following an eyeball rupture. The prolapsed vitreous dragged the whole retinal incarceration into the wound tract, forming a closed funnel, and the hemorrhage accumulated in the subretinal space. When the choroidal laceration occurred, the subretinal and suprachoroidal spaces connected with one another, so the hemorrhage was actually accumulated within the 2 gaps. Under such circumstances, it was quite difcult to discern which types of hemorrhage were in an injured eye globe with severe intraocular tissue disorder. Hence, we considered it more appropriate to refer to this as an intraocular hemorrhage, which was conrmed during globe exploration. In this study, an intraocular hemorrhage included a vitreous hemorrhage, a subretinal hemorrhage, and a suprachoroidal hemorrhage. Severe intraocular hemorrhage was dened as a hemorrhage too dense to allow visualization of the optic disc and identication of the intraocular tissues. Large scleral wound. A large scleral wound was dened as a scleral wound length of 10 mm or more, which was examined during open-globe repair or exploratory surgery. Ciliary body damage. In this study, ciliary body damage included ciliary epithelium detachment, ciliary body detachment, ciliary body defect, ciliary process atrophy, and formation of ciliary membrane, which was conrmed during exploratory surgery. Closed funnel retinal detachment. In open-globe injured eyes, almost all the vitreous prolapsed because of a sudden drop of IOP. The prolapsed vitreous dragged the whole retinal incarceration into the wound tract, forming a closed funnel. Closed funnel retinal detachment was diagnosed during exploratory surgery. Choroidal damage. In this study, choroidal damage included choroidal laceration, choroidal detachment, choroidal rupture, choroidal incarceration, and choroidal defect, which was conrmed during exploratory surgery. Massive suprachoroidal hemorrhage. Dened as a hemorrhage in the suprachoroidal space of sufcient volume to cause extrusion of the intraocular contents of the eye or to force the inner retinal surfaces into apposition due to a sharp drop of IOP after onset of the open-globe injury. 655

Zone of injury. The zone of injury is dened by location of the most posterior full-thickness aspect of the globe opening according to The Ocular Trauma Classication Group:10 open globe I, wound involvement is isolated to the cornea or corneoscleral limbus; open globe II, fullthickness wound involves the sclera no more posterior than 5 mm from the corneoscleral limbus; open globe III, full-thickness wound is posterior to open globe II. VOL. 152, NO. 4 PROGNOSTIC INDICATORS
FOR

OPEN-GLOBE INJURED EYES WITH NLP

656
TABLE 1. Ocular Characteristics, Surgical Interventions, and Visual Outcomes of Traumatized Eyes With No Light Perception After Open-Globe Injury in Favorable Outcome Group
Time to PPV (days) Ciliary Body Damage No Age (years) Type of Injury Zonea Corneal/Scleral Wound (mm) Extrusion of Lens/Iris Retina PVR Choroidal Damage Severe IOH Main Surgical Interventionsb F/U (months) Final BCVA

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

39 19 41 37 26 8 35 27 37 28 36 46 49 37 21 41 41 30

Perforating Penetrating Rupture Rupture Rupture Perforating Perforating Rupture Penetrating Rupture Rupture Rupture Rupture Perforating IOFB Rupture Penetrating Rupture

III II III II II II III III III III III III III II I III II III

3 3 3 3 4 5 5 10 10 11 12 14 27 30 30 42 45 52

Yes / 10 No / 10 Yes / 10 No / 10 No / 10 No / 10 Yes / 10 No / 10 No / 10 No / 10 No / 10 No / 10 No / 10 Yes / 10 Yes / No Yes / 10 Yes / 10 Yes / 10

/ / / / / / / / / / / / / / / / / /

RD F-RD RD RD RD RD RD RD RD F-RD F-RD RD F-RD RD RD RD RD F-RD

Lx, C3F8 Lx, UF, C3F8 Lx, Rx, C3F8 TSD, C3F8 Rx, C3F8 C3F8 Lx, C3F8 Rx, SO MP, Rx, C3F8 Lx, UF, C3F8 UF, TSD, SO C3F8 Lx, MP, Rx, UF, SO MP, Rx, SO MP, Rx, SO TKP, MP, Rx, C3F8 MP, Rx, SO MP, Rx, UF, SO

6 6 29 9 11 7 6 7 14 10 47 7 11 6 7 7 6 6

CF 1 20/160 LP CF 0.2 20/500 HM 20/50 CF 0.5 CF 0.5 20/40 LP 20/160 LP 20/320 CF 0.5 CF 0.5 20/100 LP

AMERICAN JOURNAL
OF

OPHTHALMOLOGY OCTOBER
2011

BCVA best-corrected visual acuity; CF counting ngers; F-RD closed funnel retinal detachment; F/U period of follow-up; HM hand movement; IOFB intraocular foreign body; IOH intraocular hemorrhage; LP light perception; Lx lensectomy; MP membrane peeling; PPV pars plana vitrectomy; PVR proliferative vitreoretinopathy; RD retinal detachment; Rx retinotomy or retinectomy; SO silicone oil; TKP temporary keratoprosthesis; TSD trans-scleral drainage; UF unfold closed funnel retinal detachment. a Zone I: wound involvement is isolated to the cornea or corneoscleral limbus; Zone II: full-thickness wound involves the sclera no more posteriorly than 5 mm from the corneoscleral limbus; Zone III: full-thickness wound is posterior to zone II. b All cases underwent 20-gauge PPV; endolaser and peruorocarbon were used in cases that required a retinal operation.

VOL. 152, NO. 4

No

TABLE 2. Ocular Characteristics, Surgical Interventions, and Anatomic Outcomes of Traumatized Eyes With No Light Perception After Open-Globe Injury in Unfavorable Outcome Group
Time to PPV (days) Corneal / Scleral Wound (mm) Ciliary Body Damage

Age (years)

Type of Injury

Zonea

Extrusion of Lens / Iris

Retina

PVR

Choroidal Damage

Severe IOH

Main Surgical Interventionsb

F/U (months)

Outcome

1 2 3 4 5 6 7 8 9 10 11 12

27 38 51 25 32 41 33 25 22 47 13 23

Rupture Rupture Penetrating Rupture Rupture Rupture Rupture Rupture Rupture Rupture Rupture Rupture

III III III III III III III III III III III III

7 9 10 12 12 13 13 13 14 16 16 18

No / 10 No / 10 Yes / 10 No / 10 Yes / 10 No / 10 No / 10 Yes / 10 No / 10 No / 10 No / 10 Yes / 10

/ / / / / / / / / / / /

All-pro F-RD F-RD All-pro F-RD F-RD F-RD RD All-pro All-pro RD F-RD

PROGNOSTIC INDICATORS
FOR

13

10

Rupture

III

19

Yes / 10

F-RD

14 15

13 21

Rupture Rupture

III II

24 49

Yes / 10 Yes / 10

/ /

RD F-RD

Enucleation Lx, UF, SO Enucleation TKP, SO Lx, UF, SO Enucleation Lx, UF, Rx, SO SB, TSD, C3F8 MP, SO C3F8 Lx, MP, C3F8 SB, UF, MP, Rx, C3F8 MP, UF, Rx, C3F8 TKP, Rx, C3F8 MP, Rx, UF, C3F8

6 7 7 7 9 8 8 79 6

EN PB EN EN PB EN EN AR(NLP) EN PB PB EN

OPEN-GLOBE INJURED EYES WITH NLP 657

EN

7 7

EN AR(NLP)

All-pro all prolapsed; AR anatomic restoration; EN enucleation; F/U period of follow-up; F-RD closed funnel retinal detachment; IOH intraocular hemorrhage; Lx lensectomy; MP membrane peeling; NLP no light perception; PB phthisis bulbi; PPV pars plana vitrectomy; PVR proliferative vitreoretinopathy; RD retinal detachment; Rx retinotomy or retinectomy; SB scleral buckle; SO silicone oil; TKP temporary keratoprosthesis; TSD trans-scleral drainage; UF unfold closed funnel retinal detachment. a Zone II: full-thickness wound involves the sclera no more posteriorly than 5 mm from the corneoscleral limbus; Zone III: full-thickness wound is posterior to zone II. b All cases underwent 20-gauge PPV except for cases 1, 3, and 6; endolaser and peruorocarbon were used in cases that required a retinal operation.

TABLE 3. Seven Prognostic Factors for Outcomes of Open-Globe Injured Eyes With No Light Perception
Count (Column %) Favorable Outcome Unfavorable Outcome

TABLE 4. Outcomes of Open-Globe Injured Eyes With No Light Perception by Assorted Risk Factors
Favorable Outcomeb Risk Factors
a

Unfavorable Outcomec Count Row (%)

Count

Row (%)

Pa

Type of injury Othersb Rupture Zone of injuryc Zone I / Zone II Zone III Scleral wound (mm) None / 10 10 Ciliary body Normal Damage Intraocular hemorrhage Mild / Moderate Severe Closed funnel RD or retinal prolapse No Yes Choroid Normal Damage

8 (44.4) 10 (55.6) 7 (38.9) 11 (61.1) 12 (66.7) 6 (33.3) 12 (66.7) 6 (33.3) 14 (77.8) 4 (22.2)

1 (6.7) 14 (93.3) 1 (6.7) 14 (93.3) 1 (6.7) 14 (93.3) 0 (.0) 15 (100.0) 4 (26.7) 11 (73.3)

.021

.046

0 1 2 3 4 5 6 7

3 3 2 4 2 2 1 1

100.0 100.0 100.0 100.0 66.7 66.7 16.7 11.1

0 0 0 0 1 1 5 8

.0 .0 .0 .0 33.3 33.3 83.3 88.9

.001
a Risk factors include rupture, open globe III, scleral wound 10 mm, ciliary body damage, severe intraocular hemorrhage, closed funnel retinal detachment or retinal prolapse, and choroidal damage. b Anatomically restored eyes with nal vision of light perception or better. c Anatomically restored eyes with nal vision of no light perception, phthisis bulbi, or cases that underwent enucleation.

.001

.005

13 (72.2) 5 (27.8) 10 (55.6) 8 (44.4)

3 (20.0) 12 (80.0) 0 (.0) 15 (100.0)

.005

.001

RD retinal detachment. a P values calculated with Fisher exact test. b Others include penetrating, intraocular foreign body, perforating. c Zone I: wound involvement is isolated to the cornea or corneoscleral limbus; Zone II: full-thickness wound involves the sclera no more posteriorly than 5 mm from the corneoscleral limbus; Zone III: full-thickness wound is posterior to zone II.

Unfavorable outcome. Unfavorable outcome is dened as cases of phthisis bulbi or cases that underwent enucleation. Anatomically restored eyes whose nal BCVA is NLP after 6 months of follow-up are also considered as an unfavorable outcome.
STATISTICAL ANALYSIS:

Anatomically restored eyes. Determined after a period of follow-up no less than 6 months, IOP 8 mm Hg, vitreous cavity lled with aqueous humor, retinal attachment, or only local detachment that does not require surgery. Hypotonous eyes. Determined after a period of follow-up no less than 6 months, IOP 8 mm Hg, vitreous cavity lled with aqueous humor that exhibits the Tyndall effect owing to the presence of damage to the blood-ocular barrier. Silicone oilsustained eyes. Silicone oilsustained eyes refers to eyes injected with silicone oil because of an incomplete treatment, lower IOP (8 mm Hg), unrecovered retinal detachment, or inoperable retinal detachment. Favorable outcome. Favorable outcome is dened as anatomically restored eyes whose nal BCVA is LP or better after 6 months of follow-up. 658 AMERICAN JOURNAL
OF

Because the number of cases was less than 40, the Fisher exact test was used to evaluate the variables including interval of time between injury and vitrectomy, type and zone of injury, corneal and scleral wound, extrusion of iris or lens, ciliary body damage, severe intraocular hemorrhage, closed funnel retinal detachment or retinal prolapse, proliferative vitreoretinopathy (PVR), and choroidal damage according to 2 main outcomes: favorable and unfavorable. The statistical level of signicance was preset at .05 and all the eligible data were analyzed using SPSS version 17.0 statistical software (SPSS Inc, Chicago, Illinois, USA).

RESULTS
DEMOGRAPHICS OF CASES:

A total of 72 cases with NLP after open-globe injury were reviewed during this study period, and 33 injured eyes (33 patients) met the criteria for the evaluation of prognostic indicators. The number of eyes in the favorable outcome group and the unfavorable outcome group was 18 and 15 respectively. The mean patient age was 33.2 years in the favorable outcome group; the youngest patient and the oldest patient were 8 and 49 years old respectively. The mean patient age was 28.1 years in the unfavorable outcome group; the OPHTHALMOLOGY OCTOBER
2011

youngest and the oldest were 10 and 51 years old respectively. The average follow-up period of the favorable outcome group was 11.2 months (ranging from 6 to 47.4 months). The average follow-up period of the unfavorable outcome group was 12.9 months, except for 3 cases that underwent primary enucleation after globe exploration. The shortest follow-up period was 5.4 months for 1 case that was enucleated at a subsequent surgical procedure after a period of follow-up. The longest follow-up time was 78.9 months (Tables 1 and 2).
ANATOMIC AND VISUAL OUTCOMES OF EYES WITH NLP: Of the 33 cases, 20 eyes were anatomically restored

with follow-up visual acuity (VA) ranging from NLP to 20/40; 4 eyes became phthisical after 6 to 79 months; 6 eyes were enucleated at a subsequent surgical procedure after follow-up period of 5 to 8 months; and 3 eyes were enucleated after exploratory surgery. As to visual outcome, 5 eyes recovered vision better than 20/200 (visual impairment); 1 eye retained vision of 20/320 (severe visual impairment); 12 eyes had vision worse than 20/400 (profound visual impairment); and 15 eyes had NLP, enucleation, or phthisis. In the 39 excluded cases, there were 38 silicone oil sustained eyes and 1 hypotonous eye. One eye recovered vision better than 20/200; 1 retained vision of 20/250; 28 had vision worse than 20/400; and 9 had NLP (including 1 hypotonous eye). The proportion of eyes with favorable outcome and useful vision (better than 20/200) in the entire series of 72 cases was 25% (18/72) and 6.9% (5/72) respectively. In this study, the Fisher exact test showed that a rupture (P .021), open globe III (P .046), large scleral wound (P .001), ciliary body damage (P .001), severe intraocular hemorrhage (P .005), closed funnel retinal detachment or retinal prolapse (P .005), and choroidal damage (P .001) predicted an unfavorable anatomic and visual outcome for traumatized eyes with NLP (Table 3). There was a signicant correlation between the probability of a favorable outcome and the number of risk factors when the Spearman rank correlation test was employed (correlation coefcient 0.932, P .001) (Table 4).
PROGNOSTIC INDICATORS:

DISCUSSION
TREATING TRAUMATIZED EYES WITH NLP VISION AFTER

open-globe injury, which implied no further therapeutic intervention previously, is still a challenge for ophthalmologists. During exploratory surgery, clinicians should make a rational choice of surgical interventions according to the ocular characteristics of traumatized NLP cases. Based on current published reports, there are no accepted prognostic VOL. 152, NO. 4 PROGNOSTIC INDICATORS
FOR

indicators for the NLP eyes after open-globe injury. In this study, we have identied 33 cases with veried NLP following open-globe injury that achieved their nal outcomes, and investigated the factors that signicantly predicted prognosis. The exclusion of patients is based on both their preoperative characteristics and their postoperative stable outcomes. The main points of this study are evaluation of whether LP or better vision can be restored and whether anatomic restoration can be achieved in injured eyes with NLP. Silicone oilsustained eyes or hypotonous eyes were excluded for the following reasons: Under normal circumstances, silicone oil was evacuated 3 months after vitreoretinal surgery. In some cases, however, the retina did not attach after 3 months or onset of PVR. Therefore, a rell of silicone oil into the eye globe is necessary for reattaching the retina after membrane peeling and photocoagulation. In addition, silicone oil cannot be evacuated in some silicone oilsustained eyes because the detached retina does not reattach due to extensive brosis or scarring of the retina. Based on the clinical observation and authors experience, the hypotonous eyes status is unstable; that is, it might become phthisical or restore intraocular pressure of 8 mm Hg or more after a long period of follow-up. Considering the poor prognosis and probability of sympathetic ophthalmia, many clinicians advocate primary enucleation for traumatized eyes with NLP. According to Moshfeghis standards,8 open-globe injury with NLP vision is the indication for enucleation. Some scholars11 believe, however, that injured eyes with NLP can be repaired and have a functional outcome. Vitreoretinal surgery can enable injured patients to retain their own eye globes even if LP or better vision cannot be restored. Furthermore, it has been shown that the occurrence of sympathetic ophthalmia is very small, so eyes with NLP can sustain long-term follow-up and be observed without the implementation of enucleation or evisceration.2,1215 In our study, contralateral sympathetic ophthalmia did not occur in any of the patients during the follow-up period. The shortest and longest time interval between injury and exploratory surgery in these cases is 3 and 52 days respectively, which mainly depended on the referral time of patients to the participating hospitals. The delays were almost all attributable to the limitation of economic conditions and medical resources. In China, there is an uneven distribution of medical resources and a poor level of basic hospital treatment in rural areas, and some patients are reluctant to seek medical treatment until the status of their injured eyes is unbearable. Six tertiary referral hospitals participated in this study. These hospitals treat a large number of referred ocular injury patients from north and south China each year, especially patients from remote and rural areas. This study shows there are 7 ocular characteristics that correlate with the nal anatomic and visual outcomes. Rupture, open globe III, and large scleral laceration (10 659

OPEN-GLOBE INJURED EYES WITH NLP

mm) are signicant factors in both the favorable and unfavorable groups. Multivariate analysis in Rofail and associates report3 shows injured eyes with a large laceration (10 mm) and rupture are prone to obtain an unfavorable nal visual acuity. Rahman and associates16 reported that a blunt mechanism of open-globe injury (rupture) on presentation was a signicant risk factor that was associated with eventual enucleation. Research by Matthews and associates7 also indicates that signicant predictive factors for nal visual outcome include the location of the wound and the mechanism of injury. Wounds located posterior to the equator and blunt injury in open globes predict a poor nal visual outcome. The mechanism of a rupture is also an important predictor of outcome in Schmidt and associates classication and regression tree model.17 Furthermore, traumatized eye globes with open globe III usually present serious ocular injury characteristics. Thus, the more posterior the wound extends, the greater the probability of a no-vision outcome. Moreover, a rupture is often accompanied by a large scleral laceration and a prolapse of the intraocular contents following sudden drop of intraocular pressure, which present difculties on wound closure and severely damage the retina and choroid. To summarize, a rupture, open globe III, and a large scleral wound are strong predictors of an eventual unfavorable outcome. In addition, the type and zone of injury and scleral laceration can be examined during open-globe repair, so these ocular characteristics are useful in patient counseling and in clinical decision making regarding further therapeutic interventions. Ciliary body dysfunction attributable to trauma is usually responsible for prolonged or progressive hypotony that could result in irreversible structural and functional changes in the globe, which can lead to phthisis bulbi, a condition where the globe is permanently shrunken and dysfunctional and vision is lost.18 A functional ciliary body is essential for retaining the eye globe, and an injured eye globe could not be anatomically restored if the ciliary body was seriously damaged. In this study, ciliary body damage occurred in all the unfavorable outcome group cases, while it occurred in 6 cases (33.3%) of the favorable outcome group (P .001). So it appears that ciliary body damage after injury is destructive to both visual outcome and anatomic outcome. Based on clinical observation, severe intraocular hemorrhage usually occurs following severe damage of the intraocular contents such as the choroid, retina, and ciliary body. Articles documenting poor prognosis attributable to vitreous hemorrhage are numerous.19 In our study, severe intraocular hemorrhage is also a risk factor of unfavorable outcome, and it did not occur in cases with 3 or fewer risk factors. The authors consider that severe intraocular hemorrhage is not only a predictor of an unfavorable outcome, but also a reection of the seriousness of the ocular damage, which includes severe intraocular tissue injury, such as to the choroid and retina. 660 AMERICAN JOURNAL
OF

In this study, closed funnel retinal detachment or a prolapsed retina is statistically signicant (P .005) between the favorable and unfavorable outcome groups. There are not many published articles on this factor. According to the authors clinical observation and experience, the occurrence of closed funnel retinal detachment is often accompanied by choroidal damage and extensive ciliary body damage in eyes with NLP following openglobe injury, and the prognosis for injured eyes is much worse when they occur simultaneously. There were 2 cases (11.1%; 2/18) in the favorable outcome group and 8 cases (53.3%; 8/15) in the unfavorable outcome group that simultaneously had closed funnel retinal detachment, choroidal damage, and ciliary body damage (P .020). There were 4 cases of retinal prolapse in the unfavorable outcome group: 1 phthisis bulbi occurred after 8 months; 1 injured eye was enucleated after exploratory surgery; and 2 eyes were enucleated after 7 and 8 months respectively. The eyes in these 4 cases were more severely injured than the others, and all 4 cases were complicated by choroidal and ciliary body damage. Articles have been published about suprachoroidal hemorrhage during cataract surgery,20 pars plana vitrectomy,2123 glaucoma surgery,24,25 penetrating keratoplasty,26 and massive spontaneous choroidal hemorrhage.27 There are also publications about choroidal ruptures,28 30 which are mainly attributable to blunt closed-globe injury. Few published articles, however, document the outcome of direct damage to the choroid following open-globe injury. In our study, all cases had direct choroidal damage after open-globe injury except 1 case that had a choroidal rupture following blunt injury (rupture, open globe II). It has been reported that choroidal hemorrhages are primarily caused by the vortex vein tears.31 The authors consider that the choroidal hemorrhage is the main source of dense intraocular hemorrhage in severely injured eyes, for it is usually followed by a massive suprachoroidal hemorrhage. The difference between a massive suprachoroidal hemorrhage caused by trauma and one that occurs during corneal transplant surgery or neovascular glaucoma is that there is a choroidal laceration in open-globe severe eye injuries. In open-globe injured eyes with choroidal lacerations, surgical reattachment of the choroid is very difcult because the choroidal tissue and blood vessels contain abundant collagen and elastic bers, which constrict and make the choroidal wound shorter. Schepens32 considered that persistent choroidal detachment may also be an important factor of hypotony after vitrectomy. In this study, choroidal damage occurred in all cases of the unfavorable group and 8 cases (44.4%; 8/18) of the favorable group (P .001). These data indicate that choroidal damage is the main risk factor for an unfavorable outcome as well as the main source for a dense intraocular hemorrhage of injured eyes. This study shows that a rupture, open globe III, scleral wound 10 mm, severe intraocular hemorrhage, closed funnel retinal detachment or retinal prolapse, ciliary body damage, and choroidal damage are all possible predictors of OPHTHALMOLOGY OCTOBER
2011

an unfavorable outcome, and that they might be the reason for a higher enucleation rate in traumatized eyes with NLP than injured eyes with LP or better vision.6 Traumatized eyes with NLP, however, are not beyond repair, as demonstrated by the acceptable outcomes of NLP cases in this study. In this study, all cases that had no more than 3 of the above risk factors were saved and obtained LP or better vision. In 4 cases (66.7%; 4/6) of injured eyes that had 4 or 5 of the aforementioned risk factors, these eyes obtained a favorable outcome through vitreoretinal surgery. Even 2 injured eyes complicated by 6 to 7 of the risk factors were saved and recovered LP vision by vitreoretinal surgery (Table 1, Cases 11 and 13). The probability of a favorable outcome was increased following a decreased number of risk factors (correlation coefcient -0.932, P .001) (Table 4). One of the primary limitations to this study is that our data are not adaptable to multivariate analysis. The sample pool of 33 cases is too small with respect to the 7 independent variables. Moreover, the number of some independent variables is zero under a certain condition of

the dependent variable. A multivariate model may be developed with larger patient numbers. Applying such analysis may indicate a lack of independently statistical signicance of variables in prognosis, for there is possible interaction between the risk factors. For example, the wound of ruptured eyes often extends to open III zone. In addition, to ensure the accurate results of prognostic indicators, we excluded the 39 eyes (54.2%) that may become phthisical or anatomically restored after a long period. However, the exclusion may overestimate the proportion of eyes with favorable outcome and useful vision. Hence, we report our results based on the entire series of 72 cases, which is relevant and much more realistic, though it may be underestimated. Based on the results of this study, the authors recommend that every effort be made to salvage traumatized eyes with NLP. Vitreoretinal surgery or enucleation should be carefully determined after a comprehensive assessment of the injured ocular tissues during exploratory surgery and review of the 7 risk factors for poor prognosis developed from this study.

PUBLICATION OF THIS ARTICLE WAS FUNDED BY A GRANT FROM THE PEKING UNIVERSITY HEALTH SCIENCE CENTER, 211 Evidence-Based Medicine Discipline Groups, Beijing, China (No.03-9-02). The funding organization had no role in the design or conduct of this research. The authors indicate no nancial conict of interest. Involved in design of the study (Z.Z.M.); conduct of the study (K.F., Z.Z.M., Y.T.H.); collection of the data (Z.Z.M., Y.T.H., K.F.); analysis and interpretation of the data (K.F.); and review and approval of the manuscript (Z.Z.M.). This study was approved by the review board/ethics committee of the Peking University Third Hospital. Informed consent was obtained from every patient.

REFERENCES
1. Sobaci G, Mutlu FM, Bayer A, Karagul S, Yildirim E. Deadly weapon-related open-globe injuries: outcome assessment by the ocular trauma classication system. Am J Ophthalmol 2000;129(1):4753. 2. Savar A, Andreoli MT, Kloek CE, Andreoli CM. Enucleation for open globe injury. Am J Ophthalmol 2009;147(4): 595 600. 3. Rofail M, Lee GA, ORourke P. Prognostic indicators for open globe injury. Clin Experiment Ophthalmol 2006;34(8): 783786. 4. Larque-Daza AB, Peralta-Calvo J, Lopez-Andrade J. Epidemiology of open-globe trauma in the southeast of Spain. Eur J Ophthalmol 2010;20(3):578 583. 5. Salehi-Had H, Andreoli CM, Andreoli MT, Kloek CE, Mukai S. Visual outcomes of vitreoretinal surgery in eyes with severe open-globe injury presenting with no-lightperception vision. Graefes Arch Clin Exp Ophthalmol 2009; 247(4):477 483. 6. Morris R, Kuhn F, Witherspoon CD. Management of the opaque media eye with no light perception. In: Alfaro DV III, Liggett PE, eds. Vitreoretinal Surgery of Injured Eye. Philadelphia: Lippincott-Raven; 1999:113124. 7. Matthews GP, Das A, Brown S. Visual outcome and ocular survival in patients with retinal detachments secondary to open- or closed-globe injuries. Ophthalmic Surg Lasers 1998; 29(1):48 54. 8. Moshfeghi DM, Moshfeghi AA, Finger PT. Enucleation. Surv Ophthalmol 2000;44(4):277301.

9. Kuhn F, Morris R, Witherspoon CD, Heimann K, Jeffers JB, Treister G. A standardized classication of ocular trauma. Graefes Arch Clin Exp Ophthalmol 1996;234(6):399 403. 10. Pieramici DJ, Sternberg P Jr, Aaberg TM Sr, et al. A system for classifying mechanical injuries of the eye (globe). The Ocular Trauma Classication Group. Am J Ophthalmol 1997;123(6):820 831. 11. Kuhn F, Slezakb Z. Damage control surgery in ocular traumatology. Injury 2004;35(7):690 696. 12. Casson RJ, Walker JC, Newland HS. Four-year review of open eye injuries at the Royal Adelaide Hospital. Clin Experiment Ophthalmol 2002;30(1):1518. 13. Du Toit N, Motala MI, Richards J, Murray AD, Maitra S. The risk of sympathetic ophthalmia following evisceration for penetrating eye injuries at Groote Schuur Hospital. Br J Ophthalmol 2008;92(1):61 63. 14. Gurdal C, Erdener U, Irkec M, Orhan M. Incidence of sympathetic ophthalmia after penetrating eye injury and choice of treatment. Ocul Immunol Inamm 2002;10(3):223227. 15. Albert DM, Diaz-Rohena R. A historical review of sympathetic ophthalmia and its epidemiology. Surv Ophthalmol 1989;34(1):114. 16. Rahman I, Maino A, Devadason D, Leatherbarrow B. Open globe injuries: factors predictive of poor outcome. Eye (Lond) 2006;20(12):1336 1341. 17. Schmidt GW, Broman AT, Hindman HB, Grant MP. Vision survival after open globe injury predicted by classication and regression tree analysis. Ophthalmology 2008;115(1):202209. 18. Coleman DJ. Evaluation of ciliary body detachment in hypotony. Retina 1995;15(4):312318.

VOL. 152, NO. 4

PROGNOSTIC INDICATORS

FOR

OPEN-GLOBE INJURED EYES WITH NLP

661

19. Entezari M, Rabei HM, Badalabadi MM, Mohebbi M. Visual outcome and ocular survival in open-globe injuries. Injury 2006;37(7):633 637. 20. Welch JC, Spaeth GL, Benson WE. Massive suprachoroidal hemorrhage. Follow-up and outcome of 30 cases. Ophthalmology 1988;95(9):12021206. 21. Mei H, Xing Y, Yang A, Wang J, Xu Y, Heiligenhaus A. Suprachoroidal hemorrhage during pars plana vitrectomy in traumatized eyes. Retina 2009;29(4):473 476. 22. Speaker MG, Guerriero PN, Met JA, Coad CT, Berger A, Marmor M. A case-control study of risk factors for intraoperative suprachoroidal expulsive hemorrhage. Ophthalmology 1991;98(2):202209; discussion 210. 23. Lakhanpal V, Schocket SS, Elman MJ, Dogra MR. Intraoperative massive suprachoroidal hemorrhage during pars plana vitrectomy. Ophthalmology 1990;97(9):1114 1119. 24. Feretis E, Mourtzoukos S, Mangouritsas G, Kabanarou SA, Inoba K, Xirou T. Secondary management and outcome of massive suprachoroidal hemorrhage. Eur J Ophthalmol 2006; 16(6):835 840. 25. Frenkel RE, Shin DH. Prevention and management of delayed suprachoroidal hemorrhage after ltration surgery. Arch Ophthalmol 1986;104(10):1459 1463.

26. Ingraham HJ, Donnenfeld ED, Perry HD. Massive suprachoroidal hemorrhage in penetrating keratoplasty. Am J Ophthalmol 1989;108(6):670 675. 27. Yang SS, Fu AD, McDonald HR, Johnson RN, Ai E, Jumper JM. Massive spontaneous choroidal hemorrhage. Retina 2003;23(2):139 144. 28. Raman SV, Desai UR, Anderson S, Samuel MA. Visual prognosis in patients with traumatic choroidal rupture. Can J Ophthalmol 2004;39(3):260 266. 29. Gotzaridis EV, Vakalis AN, Sethi CS, Charteris DG. Surgical removal of sequential epiretinal and subretinal neovascular membranes in a patient with traumatic choroidal rupture. Eye (Lond) 2003;17(6):790 791. 30. Ament CS, Zacks DN, Lane AM, et al. Predictors of visual outcome and choroidal neovascular membrane formation after traumatic choroidal rupture. Arch Ophthalmol 2006; 124(7):957966. 31. Wolter JR, Garnkel RA. Ciliochoroidal effusion as precursor of suprachoroidal hemorrhage: a pathologic study. Ophthalmic Surg 1988;19(5):344 349. 32. Schepens C. Retinal Detachment and Allied Diseases. Philadelphia: WB Saunders; 1983:1006.

662

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

OCTOBER

2011

Biosketch
Kang Feng, MD, is currently the Attending Physician of Peking University Third Hospitals Department of Ophthalmology, Beijing, China. Fengs research and clinical interests involve ocular trauma, clinical trials, epidemiology, evidence-based medicine, diabetic eye disease, and neuro-ophthalmology. Dr Feng completed his ophthalmology residency at General Hospital of Peoples Liberation Army in Beijing. Following residency, he has been engaged in the clinical study of ocular trauma and surgical technique at Peking University Third Hospitals Department of Ophthalmology.

VOL. 152, NO. 4

PROGNOSTIC INDICATORS

FOR

OPEN-GLOBE INJURED EYES WITH NLP

662.e1

Biosketch
Zhizhong Ma, MD, is currently the Professor of Ophthalmology and Chairman of Peking University Third Hospitals Department of Ophthalmology, Beijing, China. Mas research and clinical interests involve ocular trauma, clinical trials, and diabetic retinopathy. Currently Dr Ma is Chair of the Chinese Committee of Ocular Trauma, and the China Representative of International Society of Ocular Trauma. He serves on the editorial boards of the Ocular Surgery News Asia-Pacic Edition and American Journal of Ophthalmology-Chinese Edition.

662.e2

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

OCTOBER

2011

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.