Effect of Lubricants on Corneal Thickness After Vitrectomy

Abstract

Purpose

This study examines the impact of corneal surface lubricants used during pars plana vitrectomy (PPV) on corneal edema.

Methods

This prospective, observational clinical study occurred at an academic institution. Participants were individuals over age 18 who had already consented to undergo PPV, without pre-existing corneal pathology. Corneal lubricant was chosen by surgeon. Corneal thickness was measured pre- and post-operatively using pachymetry and anterior segment optical coherence tomography (AS-OCT). Main outcome measure was change in corneal thickness as measured by pachymetry.

Results

41 patients completed study protocol. The 23 subjects in the SHCS group had a significantly smaller increase in corneal thickness as measured by pachymetry compared to the 18 subjects in the HPMC group (29.9 um vs 58.1 um, p-value 0.02). When measured by AS-OCT, the SHCS group had a smaller increase in corneal thickness compared to the HPMC group (0.04 mm vs 0.06 mm, p-value 0.09) but did not reach significance.

Conclusions

SHCS is associated with reduced postoperative increase in corneal pachymetry as compared to HPMC.

Summary Statement

This study examines the impact of corneal surface lubricants used during pars plana vitrectomy (PPV) on corneal edema. The use of 3% sodium hyaluronate and 4% chondroitin sulfate (SHCS) is associated with a significantly smaller increase in corneal thickness when measured by pachymetry when compared to 2.5% hydroxypropylmethylcelluose (HPMC).

Introduction

Corneal clarity during pars plana vitrectomy (PPV) is of paramount importance to maintain visibility during critical intraoperative maneuvers and is often decreased because of corneal epithelial edema.¹ Loss of intraoperative corneal clarity can be overcome with debridement of the corneal epithelium, but this incurs postoperative morbidity such as pain, persistent epithelial defects or keratitis.^1–3,5 Patient and surgical factors that may contribute to increased opacification of the cornea include presence of preoperative systemic disease such as a diabetes, preexisting corneal pathology (eg Fuch’s dystrophy, previous corneal surgery), or intraoperative factors such as length of surgery, use of contact viewing systems, increased intraocular pressure and medication effects.^1,2,5,7–10

The ocular surface is routinely covered with lubricating agents during PPV to prevent desiccation. Limited knowledge has been published on differences between agents that may affect outcomes.^{1,5,6,8,11,12} One study in 2003 indicated a reduced need for corneal debridement with 0.3% hydroxypropyl-methylcellulose (GenTeal gel, Ciba Vision, Duluth, GA, USA) compared to 2.5% methylcellulose (Goniosol, Alcon, Fort Worth, TX, USA), and noted subjective difference in the corneal clarity at surgical conclusion, but robust assessment of corneal status was not performed.¹

Since 2003, no subsequent studies have examined the differences between available corneal surface lubricants in maintaining intraoperative corneal clarity. Yet in the past fifteen years, there have been significant advances in the surgical tools and techniques used in PPV that may impact the effect that corneal surface lubricants have on corneal clarity. Smaller gauge instruments provide the benefits of shorter operative time, less conjunctival scarring, faster visual recovery, improved patient comfort and potential for self-sealing wounds.¹³ New tools, such as wide-angle non-contact viewing systems, precise intraocular pressure control, and endoillumination, have been critical in maintaining efficiency and improving visibility of the entire surgical field.^13–18

Vitrectomy surgery rates are increasing, and account for an increased portion of retinal detachment repairs.¹⁹ As such, understanding the mechanisms that lead to intraoperative corneal opacification will facilitate enhanced surgical outcomes, decreased surgical times, and reduced comorbidities. The study compares the effect of two different lubricants on corneal clarity during PPV by measuring pre- and postoperative metrics of corneal thickness. The authors’ anecdotal intraoperative observations suggested a prolonged retention of a reasonable view with SHCS when compared to HPMC. As such, we hypothesize that the use of SHCS (Viscoat ® Alcon, Fort Worth, TX, USA) will be associated with a smaller increase in corneal thickness after PPV when compared to HPMC (Goniovisc ®, Hub Pharmaceuticals, Rancho Cucamonga, CA, USA).

Materials and Methods

This study was a prospective observational clinical study. To maintain equipoise with study recruitment, surgeons were permitted to maintain their standard practice rather than randomizing subjects to a corneal lubricant. Candidates were recruited from May through August 2017 at the Emory Eye Center who were planning to have vitrectomy. The Emory University Institutional Review Board provided approval for the study. The research study adhered to the tenets of the Declaration of Helsinki. Participants in the study were adult patients over the age of 18 who had already consented to undergo PPV surgery as part of routine clinical decision making. No surgeries were performed exclusively for study purposes. Subjects were excluded from the study if significant corneal pathology, such as Fuch’s Dystrophy, prior radial keratotomy, or corneal scarring was present.

Choice of corneal lubricant was made by the surgeon performing the procedure based on their standard of care. Of the six participating vitreoretinal surgeons, half routinely use SHCS and the other half use HPMC. The composition of SHCS is 3% sodium hyaluronate and 4% chondroitin sulfate. SHCS additionally contains monobasic sodium phosphate monohydrate and dibasic sodium phopshpate anhydrous, along with sodium chloride in water for injection. The pH of SHCS is between 7.0 and 7.4. HPMC contains 2.5% hydroxypropyl methylcellulose as its active ingredient with 0.01% benzalkonium chloride solution as a preservative. Inactive ingredients in HPMC are listed as boric acid, edetate disodium, sodium borate, and water for injection. The pH of HPMC is listed between 6.0 and 7.8.

Corneal thickness was measured pre- and post-operatively using ultrasound pachymetry (Pachette 2, DGH Technology, Exton, Pennsylvania, USA) and anterior segment optical coherence tomography (AS-OCT) (Visante, Carl Zeiss Meditec, Dublin, California, USA). Measurements were performed by the same individual (D.T.S.) for all patients and were taken from the center of the cornea. Pre-operative measurements were collected at the time of consent or on the day of surgery (range: 0–43 days prior to surgery). Post-operative measurements were performed within approximately 24 hours of the conclusion of surgery (range 97–1449 minutes). A clinical assessment of corneal edema was made by the assistant surgeon at the conclusion of the case using a scale of I to IV. Grade I represented a normal cornea. Grade II was minimal corneal haze and Grade III was moderate corneal haze. Grade IV represented severe corneal haze requiring epithelial debridement.

Data collected intra-operatively included the strength of povidone-iodine used for surgical field preparation, use of tetracaine as a corneal anesthetic, duration of surgery, need for corneal debridement, use of contact lens, use of dextrose infusion, and a weighted average of the intra-operative intraocular pressure. Information gathered from the medical record included subject age and sex, indication for surgery, lens status, history of prior scleral buckle or PPV, and history of diabetes.

The comparison of the mean corneal thickness in the SHCS and HPMC groups were analyzed using a Student t-test. A bivariate analysis, using either a t-test or chi-square test as needed, was completed to analyze the relationship between corneal lubricant and various factors. P values less than 0.05 were considered statistically significant. All analyses were completed using SAS 9.4 (SAS, Cary, North Carolina, USA).

Results

47 individuals were consented for the study and 41 completed study protocol. Of the six people who did not complete the study protocol, five had their surgeries rescheduled outside the recruitment period. The other subject declined to complete post-operative measurements due to discomfort.

23 people were in the SHCS group and 18 were in the HPMC group. Baseline demographic characteristics are reported in Table 1. Age, sex, lens status, and history of diabetes were similar between the two groups. However, subjects in the Goniovisc group were more likely to have had a prior PPV (39% vs 9%, p-value of 0.03).

Table 1. Patient Characteristics:

bivariate analysis of the relationship between corneal lubricant and baseline patient features

		Overall	SHCS (n=23)	HPMC (n=18)	P-value
Age		58.4 (11.4), (n=41)	60.3 (11.8), (n=23)	56.0 (10.5), (n=18)	0.2280
Sex	Female	11/41 (27%)	8/23 (35%)	3/18 (17%)	0.2911
	Male	30/41 (73%)	15/23 (65%)	15/18 (83%)
History of Prior PPV		9/41 (22%)	2/23 (9%)	7/18 (39%)	0.0283
Lens Status	Aphakic	1/41 (2.4%)	0/23 (0%)	1/18 (5.6%)	n/a
	Phakic	26/41 (63%)	12/23 (52%)	13/18 (78%)
	Pseudophakic	14/41 (34%)	11/23 (48%)	3/18 (17%)
History of Diabetes		14/41 (34%)	7/23 (30%)	7/18 (39%)	0.5710

The surgical procedures performed were standard three-port pars plana vitrectomies. Table 2 summarizes the indications for PPV among the participants. The leading indication for PPV was retinal detachment (61% and 48% in HPMC and SHCS groups, respectively) followed by macular hole repair (17% in both groups) in both groups. However, the SHCS group had 4 people (17%) undergoing PPV to treat vitreous hemorrhage and 4 to resolve a macular pucker (17%). Two cases (12%) in the HPMC group involved lens manipulation.

Table 2.

Indications for Pars Plana Vitrectomy Surgery

		Overall	SHCS (n=23)	HPMC (n=18)
Indication for PPV	Macular Hole	7/41 (17%)	4/23 (17%)	3/18 (17%)
	Macular Pucker	4/41 (10%)	4/23 (17%)	0/18 (0%)
	Retinal Detachment (RD)	22/41 (54%)	11/23 (48%)	11/18 (61%)
	RD and Lensectomy	1/41 (2.4%)	0/23 (0%)	1/18 (6%)
	Silicone Oil Removal	2/41 (5%)	0/23 (0%)	2/18 (11%)
	Subluxed lens	1/41 (2%)	0/23 (0%)	1/18 (6%)
	Vitreous Hemorrhage	4/41 (10%)	4/23 (17%)	0/18 (0%)

Table 3 describes the surgical features of the two groups. 48% of the subjects in the SHCS had 10% povidone-iodine for surgical field preparation (the other 52% used 5% povidone-iodine) the same whereas all cases in the HPMC group used 10% solution (p-value 0.0003). The operative time between the two groups were similar (average of 65 and 62 minutes, p-value 0.7, for SHCS and HPMC groups, respectively). The difference between the weighted average of the intraoperative intraocular pressure approached but did not reach significance (33.5 mmHg and 38.1 mmHg, p-value 0.1, for SHCS and HPMC groups, respectively). Use of tetracaine, contact lens, dextrose infusion, prior scleral buckle, concomitant scleral buckle, and various agents for vitreous substitute were similar.

Table 3. Surgical Features:

bivariate analysis of the relationship between corneal lubricant and various surgical features

		Overall	SHCS (n=23)	HPMC (n=18)	P-value
Use of Tetracaine		28/41 (68%)	14/23 (61%)	14/18 (78%)	0.2482
Strength of Povidone-Iodine	5%	12/41 (29%)	12/23 (52%)	0/18 (0%)	0.0003
	10%	29/41 (71%)	11/23 (48%)	18/18 (100%)
Use of Contact Lens		12/41 (29%)	7/23 (30%)	5/18 (28%)	0.8528
Dextrose Infusion		10/41 (24%)	4/23 (17%)	6/18 (33%)	0.2889
Scleral Buckle (SB)	None	36/41 (88%)	19/23 (83%)	17/18 (94%)	0.6178
	History of SB	3/18 (7%)	2/23 (9%)	1/18 (6%)
	SB preformed at same time as PPV	2/41 (5%)	2/23 (9%)	0/18 (0%)
Retinal Tamponade	Air	9/41 (22.0%)	6/23 (26.1%)	3/18 (16.7%)	n/a
	C3F8 Gas	9/41 (22.0%)	6/23 (26.1%)	3/18 (16.7%)
	SF6 Gas	18/41 (43.9%)	9/23 (39.1%)	9/18 (50.0%)
	Silicone oil	5/41 (12.2%)	2/23 (8.7%)	3/18 (16.7%)
Weighted Average of Intraoperative Intraocular Pressure (mm Hg)		35.6 (8.0), (n=36)	33.5 (6.6), (n=20)	38.1 (9.2), (n=16)	0.0981
Average Length of Surgery (Minutes)		63.6 (30.6), (n=41)	62.2 (34.0), (n=23)	65.4 (26.4), (n=18)	0.7461

The main outcome variable was the difference in corneal pachymetry between pre- and post-operative measurements. Patients in the SHCS group had a significantly smaller increase in corneal thickness as measured by pachymetry compared to the group using HPMC (29.9 um vs 58.1 um, p-value 0.02). When measured by AS-OCT, the SHCS group had a similar but smaller increase in corneal thickness compared to the HPMC group (40 um vs 60 um, p-value 0.09) but did not reach significance. Due to technical difficulties with the AS-OCT machine, two subjects in the HPMC group did not receive AS-OCT measurements. These results are listed in Table 4 and presented in graphical form in Figures 1 and 2. Timing of the postoperative visit did not seem to impact the results. 20 of 41 eyes were examined at a same day postoperative visit, with consistent demonstration of less increase in corneal thickness in the SHCS group compared to the HPMC group (data not shown).

Table 4. Corneal Outcomes:

Bivariate analysis of the relationship between corneal lubricant and corneal thickness

		Overall	SHCS (n=23)	HPMC (n=18)	P-value
Pachymetry (um)	Pre	553.6 (53.8), (n=41)	559.7 (59.7), (n=23)	545.7 (45.5), (n=18)	0.4144
	Post	595.9 (65.2), (n=41)	589.7 (72.6), (n=23)	603.8 (55.3), (n=18)	0.4980
	Delta	42.3 (36.3), (n=41)	29.9 (24.3), (n=23)	58.1 (43.3), (n=18)	0.0206
AS-OCT (um)	Pre	560 (60), (n=39)	570 (60), (n=23)	550 (40), (n=16)	0.4757
	Post	610 (60), (n=39)	600 (60), (n=23)	610 (50), (n=16)	0.5428
	Delta	50 (40), (n=39)	40 (20), (n=23)	60 (50), (n=16)	0.0870

Figure 1.

Graphical representation of the average change in pre- and post-operative corneal thickness as measured by pachymetry. The error bars represent one standard deviation of the mean.

Figure 2.

Graphical representation of the average change in pre- and post-operative corneal thickness as measured by AS-OCT. The error bars represent one standard deviation of the mean.

Clinical assessment of post-operative corneal clarity was similar between the two groups (Figure 3). Overall, the majority of corneas (68%) were graded as normal at the conclusion of the case. Another 22% were considered to have minimal haze and 10% had clinically moderate haze. No corneas required intraoperative debridement.

Figure 3.

Graphical representation of the corneal clarity grading results by percent of overall cases distributed by corneal lubricant.

Discussion

This study uses corneal thickness as a surrogate for corneal clarity to study the differences between two intra-operative lubricating agents during PPV. Our results are consistent with the hypothesis that SHCS is associated with a significantly smaller increase in corneal thickness after PPV when measured by pachymetry. When measured by AS-OCT, the smaller increase in corneal thickness in the SHCS group approached, but did not reach statistical significance, which is likely due to inadequate sample size. These results are in line with several previous studies that demonstrate a high degree of correlation between central corneal thickness as measured by Visante AS-OCT and ultrasound pachymetry.^20,21

This study builds on a 2003 study that found reduced rates of epithelial debridement in eyes using GenTeal® compared to eyes where Goniosol® was used.¹ In our study, we used two methods, pachymetry and AS-OCT, to measure central corneal thickness in addition to using the endpoints of subjective clarity and need for debridement. Unlike the previous study, none of the corneas in this study required debridement. Prolonged surgical time has been associated with increased rates of corneal epithelial debridement.^2,5 The lack of debridement in the present study may be due to decreased surgical times compared to the previous study (132 to 156 minutes on average compared to 62 to 65 minutes on average). In fact, the average time until debridement noted in the previous study (72 to 168 minutes) was longer than the total average surgery time in the present cases.

It is possible that SHCS may be associated with less corneal swelling after PPV due to differences in chemical composition, osmolarity, or pH¹⁰. Specifically, HPMC contains the presence of preservatives including benzalkonium chloride and edetate disodium (EDTA). Garcia-Valenzuela et al also suggested that the presence of various preservatives may have contributed to increased corneal toxicity and need for debridement in their Goniosol® group.¹ Studies of corneal cell cultures and animal corneas support toxic effects of benzalkonium chloride and EDTA on the cornea by disrupting plasma membranes, inducing DNA strand breaks, and breaking normal connections between cells.^22–25 Sodium hyaluronate based viscoelastic products, such as SHCS, have been shown to have a free radical scavenging effect, which may promote a favorable effect on the cornea.^26,27

The main limitation of this study is the observational design. To maintain ethical recruitment, surgeons were not required to randomize subjects, thus creating potential for selection bias. Different surgeons have variations in their practice such that case complexity, operative times, concentration of povidone-iodine, tetracaine usage, and average intraoperative intraocular pressure are all potential confounders. Indeed, patients in the HPMC group were significantly more likely to have had prior PPV and the use of 10% povidone-iodine in surgical field preparation. A 1988 study examined various risk factors for corneal complications after PPV and found that prior history of PPV only correlated weakly with the development of corneal edema (p-value of 0.07).⁷ Although concentrations of 5 to 10% of povidone-iodine are recommended for disinfection prior to ophthalmic surgery to reduce risk of endophthalmitis, corneal toxicity of povidone-iodine is well-documented in in vivo and in vitro studies.²⁸ Surgeries in the HPMC group were weakly correlated with a higher average intraocular pressure. Elevated intraocular pressure is a known risk factor contributing to corneal epithelial edema by impairing endothelial function.^2,10 Two cases in the HPMC group involved more complex anterior segment procedures, such as lensectomy, which are associated with higher rates of corneal complications.^3,4,7 Tetracaine use did not differ significantly, but has been associated with corneal toxicity and possible increased edema.^29,30

The possibility of measurement error, such as misplacing the pachymeter probe on the cornea, cannot be excluded. The use of only two types of corneal surface lubricants in the present study may limit the generalizability to other practice settings. Furthermore, no individuals with pre-existing corneal pathology were included. We hypothesize that the differences reported from this study would be magnified in eyes with abnormalities such as Fuch’s dystrophy and prior corneal refractive surgery.

Future directions to explore include studying additional corneal lubricating agents in larger patient populations, including eyes with corneal pathology, and using techniques to directly examine corneal clarity rather than corneal thickness as a surrogate.

This study presents valuable insight in improving outcomes associated with vitreoretinal surgery. Although the overall size of effect in this study is relatively small, this is clinically meaningful. Small differences can make a large impact in cases with prolonged operative time or with elevated intraocular pressure. These cases are often of the highest complexity, such as diabetic tractional retinal detachment repair or proliferative vitreoretinopathy membrane peel, and thus, adequate visualization of the posterior segment is critical for success.

In conclusion, understanding the mechanisms that lead to loss of intraoperative corneal transparency will facilitate enhanced surgical outcomes. This study found that the intraoperative use of SHCS was associated with less corneal swelling as measured by pachymetry when compared to HPMC for vitrectomy surgery. These results are particularly relevant in complex cases where factors that lead to increased corneal edema, such as higher intraocular pressures and longer operative times, are difficult to minimize.