Effect of Oral Prednisone on the Prevention and Management of Proliferative Vitreoretinopathy After Open-Globe Injury

Abstract

Purpose: To determine the impact of oral prednisone on the final visual acuity (VA) and prevention of proliferative vitreoretinopathy (PVR) in patients having pars plana vitrectomy (PPV) for globe injuries. Methods: A retrospective chart review was performed of all globe injuries with an initial repair and subsequent PPV between 2009 and 2018. Data included the initial VA, zones of injury, initial closure date, time to secondary intervention (PPV), oral prednisone (1 mg/kg/day) use, the final VA, and enucleation rate. Multivariable regression models were used to assess the impact of oral prednisone use on anatomic and functional outcomes. Results: The mean (±SD) patient age was 46.25 ±18.56 years (range, 13-92); 131 (83.9%) were men. Oral prednisone intake was recorded in 81 patients (52.3%). The prednisone group had significantly more zone 3 involvement (P = .001), worse initial VA (2.28 vs 1.92 logMAR; P = .003), and a greater mean number of surgeries (P = .020) than the no-steroids (control) group but an equivalent final logMAR VA (1.57 vs 1.52; P = .881). The prednisone group had significant VA improvement (P = .025); however, oral prednisone use did not predict the development of PVR (29.23% vs 12.90%; odds ratio [OR], 2.81; 95% CI, 0.89-8.85) or retinal detachment (27.27% vs 29.58%; OR, 0.59; 95% CI, 0.23-1.56). Conclusions: Despite a worse initial clinical presentation, patients who received oral prednisone had significant visual improvement compared with the control group. However, oral prednisone (1 mg/kg/day) use at the time of injury did not decrease the PVR rate.

Introduction

An estimated 2.0 million cases of traumatic eye injuries occur each year in the United States, with approximately 1 million individuals having permanent visual impairment. ¹ Pars plana vitrectomy (PPV) has played a major role in improving the outcomes of ocular trauma with reconstruction of the posterior segment, clearance of vitreous opacities, and prevention of endophthalmitis. ²

The use of oral prednisone in the perioperative period after globe repair is controversial and has been proposed as a measure to reduce the incidence of postsurgical inflammation and proliferative vitreoretinopathy (PVR). The role of adjuvant therapies such as oral prednisone remains a topic of debate. Steroids are known to possess potent antiinflammatory properties that may help alleviate postoperative inflammation and reduce the likelihood of PVR. PVR, a common complication after ocular trauma, involves the formation of fibrous tissue in the vitreous cavity and on the retinal surface, often leading to tractional retinal detachment (RD) and subsequent vision loss. ³ The potential benefits of oral prednisone in this context must be carefully weighed against its potential side effects.

We evaluated the impact of the use of oral prednisone on the final visual acuity (VA) and prevention of PVR after ocular trauma in a cohort of patients who had PPV after globe injury.

Methods

A retrospective chart review was performed of all patients with open-globe and closed-globe injuries hospitalized at the Jones Eye Institute at University of Arkansas for the Medical Sciences (UAMS) who had initial repair or exploration and subsequent PPV between 2009 and 2018. The study protocol was approved by the Institutional Review Board, UAMS, and adhered to the tenets of the Declaration of Helsinki. Data were anonymized and de-identified before analysis.

The same surgeon performed all PPVs, although multiple surgeons were involved in the primary wound closure or globe exploration. All vitrectomies were performed using the 23-gauge Constellation system (Alcon Laboratories Inc). Indications for vitrectomy after ocular trauma included (1) intraocular foreign bodies, (2) suspected or confirmed lens rupture, (3) dense vitreous hemorrhage with a B-scan suspicious for RD, (4) zone 2 or 3 injuries, or (5) endophthalmitis. Patients without a documented initial VA were excluded from the statistical analysis.

Intake of oral prednisone after initial repair of the globe injury was documented from the medication list in the electronic medical records or the physician’s note. Initiation of oral prednisone (1 mg/kg/day) after the primary repair was at the discretion of the surgeon performing the primary repair. Different surgeons had different management protocols for trauma patients when starting oral prednisone. When patients were referred to the retina clinic, the oral treatment regimen was not altered. For patients who were prescribed oral prednisone, the dose was tapered over 4 to 6 weeks.

PVR was defined as full-thickness retinal folds or subretinal bands, either posterior or anterior to the equator, corresponding to PVR grade C. ³ Data elements included the timing of initial globe closure, timing to the first PPV from the initial injury, logMAR best-corrected VA (BCVA) at the last follow-up, rate of enucleation, development of PVR, and type of injury (blunt force, intraocular foreign body, sharp injury, injury from a high-velocity projectile). The final logMAR VA in patients with poor vision was calculated using the methods detailed by Moussa et al. ⁴ Zone 1, zone 2, and zone 3 open-globe injuries were defined, respectively, as wounds limited to the cornea, full-thickness wounds of the sclera up to 5 mm from the corneoscleral limbus, and full-thickness wounds posterior to zone 2. ⁵

A χ² test was used to compare categorical variables and a t test for continuous variables based on demographic, clinical characteristics, and presenting trauma variables. The primary analyses included 2 logistic regression models that assessed the ability of oral prednisone use to predict the development of PVR and RD while adjusting for age, sex, type of injury, VA at presentation, and the zone of injury. All analyses were performed using Stata software (release 14.2, StataCorp LP). Tests with 2-sided P values less than 0.05 were considered statistically significant. All means are ± SD.

Results

This study included 155 eyes. Oral prednisone intake was recorded in 81 patients (52.3%), with 74 (47.7%) patients comprising the no-steroids (control) group. Table 1 shows the patients’ demographics, presenting history, and outcomes. There was no significant difference in age or sex between the oral prednisone group and the control group. The mean age of the entire cohort was 46.25 ± 18.56 years (range, 13-92); 131 patients (83.9%) were men.

Table 1.

Demographics, Presenting History, and Outcomes of Patients With and Without Oral Prednisone Treatment.

Demographic	Oral Prednisone	No Prednisone	P Value a
Patients, n (%)	81 (52.3)	74 (47.7)	—
Sex, n (%)
Female	13 (16.1)	11 (14.9)
Male	68 (83.9)	63 (85.1)	.839
Mean age (y) ± SD	45.5 ± 16.4	47.4 ± 20.6	.506
Injury type, n (%)
Blunt	41 (50.6)	44 (59.5)	—
Sharp	21 (25.9)	13 (17.6)	—
IOFB	13 (16.1)	16 (21.6)	—
Projectile	6 (7.4)	1 (1.35)	.135
Mean initial logMAR VA ± SD	2.28 ± 0.56	1.92 ± 0.87	.003
Mean final VA ± SD	1.57 ± 1.02	1.54 ± 1.02	.881
Mean change in logMAR VA ± SD	−0.73 ± 0.86	−0.41 ± 0.83	.025
Mean final IOP (mm Hg) ± SD	14.31 ± 7.34	14.73 ± 5.89	.713
Zone 1 involvement, n (%)	38 (56.72)	29 (43.28)	.332
Zone 2 involvement, n (%)	43 (59.72)	29 (40.28)	.067
Zone 3 involvement, n (%)	33 (71.74)	13 (28.26)	.001
Mean time to first PPV (days) ± SD	1.69 ± 4.49	5.34 ± 10.46	.005
Mean eye surgeries ± SD	2.51 ± 1.01	2.08 ± 1.06	.020

Abbreviations: IOFB, intraocular foreign body; IOP, intraocular pressure; PPV, pars plana vitrectomy; VA, visual acuity.

^aP values for χ² test for categorical variables and t test for continuous variables.

The mean number of surgeries in each eye was 2.31 ± 1.06 over a mean follow-up of 75.89 weeks. There was a significant difference in the mean number of eye surgeries between the 2 cohorts (P = .020), with patients taking oral prednisone having a greater number of surgeries (Table 1). More patients in the oral prednisone group than in the control group had 2 surgeries (57.35% vs 43.33%) and more than 3 surgeries (36.76% vs 25.00%) after their globe injury (P = .001). The incidence of mild to moderate postoperative vitreous hemorrhage was not collected; however, no vitrectomy was required to clear a postoperative vitreous hemorrhage.

The distribution of injury type was also similar between groups, with most injuries (>50%) being blunt trauma. The mean number of days from the initial injury to the primary surgery was 3.41 ± 8.08.

The mean logMAR VA in the affected eye at presentation was 2.11 ± 0.74. The initial VA was worse in the oral prednisone group than in the control group (2.28 vs 1.92; P = .003), although the final VA was comparable (P = .713). This corresponded to a significantly improved change in VA (final − initial) in the oral prednisone group (−0.73 vs −0.41; P = .025) (Table 1).

In the entire cohort, 28.03% of eyes had a zone 1 injury, 21.21% had a zone 2 injury, 16.67% had a zone 3 injury, 10.61% had a combined zone 1/2 injury, 13.64% had a combined zone 2/3 injury, and 9.85% had all zones impacted (zones 1, 2, and 3). Patients in the oral prednisone group had significantly more zone 3 involvement than patients in the control group (P = .001) (Table 1). There was no significant difference in the rate of PVR between the 2 groups (P = .144). There was a statistically significant difference in the probability distributions between the zone of injury and RD, with more RD cases occurring in eyes with a zone 2 and zone 3 injury (P = .010). The rate of enucleation was 4.76%. The presenting VA was worse in patients who required enucleation (2.77 vs 2.06; P = .014).

Univariate analysis showed that the use of oral prednisone did not significantly affect the rate of enucleation (P = .301) or RD (P = .854). The final intraocular pressure was comparable between the 2 groups (P = .713) (Table 1). Multivariable logistic regression models adjusting for age, sex, type of injury, VA at presentation, and zone of injury found that the use of oral prednisone did not predict the development of PVR (odds ratio [OR], 2.81; 95% CI, 0.89-8.85) or RD (OR, 0.59; 95% CI, 0.23-1.56).

Conclusions

Our study found that use of oral prednisone (1 mg/kg/day) at the time of injury did not seem to have a beneficial effect on the prevention of PVR or RD after open-globe injuries. However, despite having a worse initial VA, more prevalent zone 3 involvement, and a greater number of eye surgeries than the no-steroids (control) group, patients in the oral prednisone group had a statistically significant improvement in VA (P = .025) and an equivalent final VA.

Despite advances in surgical instrumentation and technique, ocular trauma is still a leading cause of visual impairment in the United States. ⁶ At present, the management of open-globe injuries is based on regional or institutional guidelines rather than on standardized care.^6–8 A significant consequence of traumatic open-globe injury is PVR, and its prevention and management remain highly challenging in retina practices, despite the use of both steroidal and nonsteroidal antiinflammatory therapies.^8–10

Multiple studies have assessed the role of triamcinolone in open-globe injuries.^11–15 In a study by Cheema et al, ¹² 24 patients received 4 mg intravitreal triamcinolone after vitreoretinal surgery. The rate of retina reattachment was higher in patients with advanced PVR. In a single-center prospective pilot study that included 40 patients in a randomized control trial, Banerjee et al ¹⁴ evaluated the toxicity and treatment effects of intravitreal triamcinolone after primary closure in eyes with an open-globe injury. They observed a trend toward better outcomes in VA in the adjunct group than in the control group but a comparable percentage of PVR at 6 months. A phase II multicenter randomized control trial of the effectiveness of triamcinolone for the prevention of PVR after vitreoretinal surgeries in 300 patients with an open-globe injury is underway. ¹⁵

Like other types of glucocorticoids, triamcinolone exerts its mechanism of action by inhibiting phospholipase A2 enzyme, which ultimately decreases levels of prostaglandins and leukotrienes by suppressing cyclooxygenase and lipoxygenase activities.^16,17 In addition, the clinical use of triamcinolone in the prevention of PVR is attributed to its antiangiogenic properties. Although the mechanism has not been completely elucidated, Hsiao et al ¹⁸ found that transforming growth factor-β2, a stimulant for the production of proangiogenesis factors, is attenuated in cultured human retinal epithelial cells with the use of triamcinolone. ¹⁸ In a study of triamcinolone’s safety, Veritti et al ¹³ reported that although triamcinolone was associated with some adverse ocular events, such as ocular hypertension, steroid-induced cataracts, and infection, it was an effective drug for various retinal and choroidal diseases. When delivered intravitreally, it was not associated with significant systemic safety risks. ¹³

Agents other than triamcinolone have been used to prevent PVR. Methotrexate is a competitive inhibitor of any enzyme requiring folate as a cofactor. Its clinical use is broad, ranging from being a potent antineoplastic agent at higher dosages to an anti-inflammatory agent at lower dosages. ¹⁹ Methotrexate has also been used as an antifibrogenic agent in treating keloids. Nabai et al ²⁰ found that fibroblasts treated with methotrexate led to decreased levels of secreted and intracellular type-1 collagen compared with a control. ²⁰ Because of these properties (antiproliferation, anti-inflammation, and antifibrosis), methotrexate is potentially an excellent agent for preventing and treating PVR. ¹⁹ In a study with a large sample (639 eyes), Gangaputra et al ²¹ concluded that adding systemic methotrexate to ocular inflammatory disease treatment while tapering off prednisone is well tolerated and has a moderate chance of achieving management goals. ²¹ In a study with a small sample (29 eyes), Sadaka et al ¹⁹ favored the infusion of intravitreal methotrexate for the management of PVR and stated they did not identify any adverse ocular effect of methotrexate. Although rare ocular side effects (eg, maculopathy, vitreous hemorrhage, corneal toxicity) have been reported in the literature, they are often mild and self-limiting.^19–23

The inhibition of cyclooxygenase has also been an area of focus in PVR prevention. Kähler et al ²⁴ reported that acetylsalicylic acid (100 mmol) significantly inhibited the fibroblast growth–promoting activity from intraocular fluid of patients with PVR by inhibiting metabolites in the cyclooxygenase pathway responsible for enhanced intraocular fluid–induced fibroblast proliferation. ²⁴ In a later study, Kähler et al ²⁵ used competitive enzyme immunoassay and measured a rapid, sustained increase in arachidonic acid metabolite levels when cultured fibroblasts were exposed to vitreous fluid extracted from patients with PVR. They identified a significant increase in prostaglandin E2 without significant changes in leukotriene B4 release, prostaglandin I2, and thromboxane A2 release in the control group.

More recently, in an animal study, Tikhonovich et al ²⁶ found that the intravitreal injection of nonsteroidal antiinflammatories was more effective than steroids in suppressing the development of PVR. Specifically, they compared the use of lornoxicam and triamcinolone and found that lornoxicam, but not triamcinolone, significantly reduced membrane formation at the early stages of PVR. Furthermore, lornoxicam prevented choroidal and retinal thickness changes in PVR development. Tikhonovich et al proposed that blocking cyclooxygenase during the development of PVR is more effective and safer than suppressing the entire cascade of arachidonic acid metabolism with steroids. ²⁶

Although we did not observe the prevention of PVR or RD after open-globe injury, given the worse initial presentation of patients in the oral prednisone group and significant improvement in VA compared with the control group, oral prednisone (1 mg/kg/day) did seem to provide a benefit. Banerjee et al ¹⁴ found similar trends, noting a high percentage of VA improvements in the treatment group vs the control group and similar PVR outcomes at 6 months. However, unlike the oral prednisone (1 mg/kg/day) used in our patients, Banerjee et al used both intravitreal triamcinolone injections (0.4 mg/1 mL) intraoperatively and 1-week postoperative oral flurbiprofen (50 mg/twice daily). The significance of these results was not analyzed in their study because, as the authors pointed out, theirs was an exploratory pilot study.

We did not observe significant side effects in patients receiving oral prednisone; however, some patients reported a small weight gain and sleep disturbance during the 4- to 6-week treatment period. Patients with diabetes mellitus were instructed to closely monitor their blood sugar. We also found that a worse VA at presentation was correlated with worse VA at the final follow-up and a higher rate of enucleation. In addition, for more posterior open-globe injuries, there was a trend toward a worse final visual outcome and a higher rate of enucleation. These findings correlate and agree well with the results in the existing literature, with multiple studies reporting that the presenting VA and the wound location both predict the prognosis.^27–30

We recognize the limitations of our single-center nonrandomized retrospective study with its relatively small sample. Inconsistent documentation of oral prednisone intake, with no mechanism to verify patients’ adherence to therapy, was an additional barrier. Also, the dosage may not have been optimal for assessing the beneficial properties of triamcinolone. Future multicenter randomized studies are needed to shed light on the optimal dosage of oral prednisone and assess the use of other agents to prevent PVR after open-globe injury.

In conclusion, the use of oral prednisone (1 mg/kg/day) at the time of injury did not decrease the rate of PVR. However, we observed that compared with patients in the control group, who did not receive steroids, patients receiving oral prednisone had a significant improvement in VA, despite having a worse initial clinical presentation.