Inflow of ocular surface fluid through clear corneal cataract incisions: a laboratory model☆
Article Outline
Abstract
Purpose
To evaluate the self-sealing properties of standard clear corneal cataract incisions during two events: (1) application of mechanical external pressure, or (2) controlled fluctuation of intraocular pressure (IOP).
Design
Laboratory investigation.
Methods
Eight fresh human donor globes were prepared for Miyake video microscopy. A standard two-plane 3-mm clear corneal incision was created and a 3 × 3-mm sponge soaked with India ink was placed on the wound surface. One globe with a sutured corneal wound served as the control. A transcleral cannula was inserted and connected to a bottle of saline. Intraocular pressure was modified varying the bottle height. External pressure was applied through manual contact on different regions of cornea.
Results
Four of seven eyes demonstrated intraocular presence of ink, three of them after external manipulation and another after varying the IOP.
Conclusion
Self-sealing properties of unsutured clear corneal wounds were compromised in our model. These data may give insight into the possible mechanisms involved in the inflow of extraocular fluid into the eye through clear corneal cataract incisions.
Since the introduction of phacoemulsification, many varieties of surgical techniques have been suggested to allow for a faster and easier procedure. Of all the techniques introduced, perhaps the temporal clear corneal incision is the most popular and widely accepted. It offers several advantages compared with the traditional scleral tunnel incision, including: (1) lack of conjunctival trauma, thereby minimizing discomfort and bleeding, (2) lack of suture-induced astigmatism, and (3) faster visual recovery.1, 2 Because of these advantages over other incision techniques, a progressive increase in the number of surgeons preferring clear corneal incisions has occurred over the past decade.3 In parallel, an increasing incidence of postoperative endophthalmitis after clear corneal cataract surgery has been reported (Jensen, MK. ARVO, 2002). 4, 5, 6, 7, 8, 9 In a meta-analysis study conducted by Powe and associates,10 a review of articles from a period that predates clear corneal incision (1979 to 1991) revealed that the incidence of acute endophthalmitis following cataract extraction was 0.13%. Later studies from different authors compared clear corneal incisions with scleral incisions,4, 6, 8 which revealed a significant increase up to 15 fold in the incidence of endophthalmitis following clear corneal incision. This raises the question whether bacteria might traverse the incisions during the postoperative period, resulting in an increased risk for infection.
Cataract surgeons typically check the clear corneal incision self-sealing properties by reforming the anterior chamber with fluid infusion, and applying some pressure on the cornea to check for wound leakage. Usually, applying pressure directly on the incision itself is avoided. With an adequate positive pressure in the anterior chamber, the posterior wound lip is tightly apposed to the anterior stroma, making fluid leakage unlikely.11 Such a test, however, makes unrealistic assumptions, namely: (1) all eyes will remain well-pressurized during the early postoperative period, (2) the absence of aqueous outflow from the wound correlates with the inability of tear film surface fluid to flow into the eye, possibly contaminating the aqueous humor and predisposing it to infection, and (3) no manual pressure will be applied on the wound during the early postoperative period.
Intraocular pressure (IOP) is reported to fluctuate during the postoperative period, frequently to less than 5 mm Hg.12 Telemetric IOP monitoring devices suggest that large fluctuations in IOP from 15 to 70 mm Hg also occur in response to blinking in rabbits.13 In a human study by Coleman and Trokel,14 voluntary blinking resulted in pressure spikes of 10 mm Hg, while squeezing of the lids resulted in rises of IOP exceeding 80 mm Hg. We recently demonstrated the dynamic morphology of clear corneal incision at varying IOP using optical coherence tomography.11 In that study, we demonstrated that for standard self-sealing clear corneal incisions, normal to high IOP was associated with tight apposition of the wound edges, creating a sealing effect that avoided wound leakage. Conversely, at a low IOP the wound edges tended to gape, starting at the internal aspect of the wound.
Our hypothesis is that fluid from the external surface of the eye may flow across the incision tunnel when the IOP is low or when manual pressure is applied to the eye. Both of these events might likely occur in a standard postoperative course following cataract surgery. We postulate that manual pressure exerted on the incision or elsewhere on the globe, such as from rubbing the eye or even blinking, may compromise the wound integrity, and specifically might lead to imbibition of surface fluid into the anterior chamber. In this study, we evaluated the self-sealing properties of clear corneal incisions in response to IOP fluctuation and direct application of manual pressure to the cornea. To test for fluid inflow into the anterior chamber, as opposed to aqueous outflow, India ink was applied to the corneal surface of eye bank eyes while the anterior chamber was observed using a Miyake microscope.15
Design
This study was a laboratory investigation.
Methods
A Miyake microscope set-up used to view the internal parts of the globe was assembled as previously described.15 To allow for high quality imaging that is required in this project, the modified video system consisted of a 1-chip CCD camera (Sentech STC-630AS, Sensor Technologies America, Inc., Carrollton, Texas), a high-resolution video monitor (Ultrak KM1500CN, Ultrak Inc., Lewisville, Texas), and a DVD video recorder (Philips DVDR 985, Philips Corp., Knoxville, Tennessee).
Eight fresh human globes, not suitable for transplantation, were acquired from Central Florida Lions Eye & Tissue Bank (Tampa, Florida) and kept in a moist chamber. The mean death to preservation time was 5 ± 1.8 hours. Samples were prepared for Miyake microscopy by sectioning the globes at the equator using a razor blade as previously described.15 The vitreous, uvea, and lens from the anterior half of the globe were removed using a pair of forceps to provide an unobstructed view of the internal lip of the incision (Figure 1). The eviscerated, sectioned globe was then positioned at the center of a 2 × 3-inch glass slide (Corning Glass Works, Corning, New York) and sealed in a watertight fashion using cyanoacrylate glue. A 25-gauge needle connected to intravenous tubing attached to a balanced salt solution (BSS) bottle was inserted through the sclera into the vitreous cavity. A second 25-gauge needle attached to a manometer (Digimano 1000, Netech Corp., Hicksville, New York) was inserted into the vitreous cavity to monitor the IOP (Figure 2). Intraocular pressure was modified by changing the infusion bottle height. Once the globes were filled with saline, a test for leakage at the junction of the slide and globe revealed no outflow of fluid from the area of attachment, confirming that there was not an inadvertent influx of ink from an accessory site.
FIGURE 1.
Tissue Preparation. (Left) After the globe is sectioned along the equator, the lens and uvea are removed using a pair of forceps. (Right) A prepared tissue ready for mounting on a glass slide is shown.
FIGURE 2.
Intraocular Pressure Adjustment. Intraocular pressure is adjusted and monitored through infusion ports attached to a balanced salt solution bottle and a digital manometer. A 3 × 3-mm cellulose sponge soaked with India ink is placed over the corneal incision.
After adjusting the BSS bottle height to obtain a baseline IOP from 15 to 18 mm Hg, a 3.0-mm wide, two-plane clear corneal incision was made using a new disposable keratome (Alcon, Forth Worth, Texas) under microscopic visualization (Moeller-Wedel, Wedel, Germany). Incision tunnel lengths were approximately 2.0 to 2.5 mm. The self-sealing properties of the wound were confirmed in each case using the classic maneuvers previously described.
Of the eight globes, one was assigned as the control. A single nylon 10–0 suture was placed in this eye to seal the incision. In all eyes, a 3 × 3-mm cellulose sponge soaked with India ink (Sanford Corp., Bellwood, Illinois) was placed over the incision. Intraocular pressure fluctuation was simulated by the stepwise decrease of the infusion bottle height to lower IOP (from baseline to −5 mm Hg), while the internal opening of the incision was observed through the Miyake monitor for influx of ink. If no visible influx was observed with this maneuver, the BSS bottle was repositioned to obtain a baseline IOP from 15 to 18 mm Hg. Manual pressure, which increased the IOP by 10 to 20 mm Hg, was then applied at four different quadrants of the globe using a blunt instrument to simulate manual external pressure.
Results
The control, sutured eye did not demonstrate ink influx with IOP fluctuation nor with manual pressure. Conversely, of the seven eyes without sutures, three did not show any visible ink influx in response to either IOP fluctuation or manual pressure. In one eye, ink was noted to flow inside when the IOP was initially lowered to a pressure of −5 mm Hg and then slowly repressurized to 5 mm Hg to reform the globe (Video 1). Ink influx was noted as the globe was reforming at 5 mm Hg even without any manual pressure exerted on the globe. Three eyes, at baseline IOP of 15 to 18 mm Hg, demonstrated ink influx after manual pressure was applied to the corneal surface, especially during the brief period following release of pressure (Figure 3), (Video 2). This phenomenon was noted repeatedly, regardless of the area on the globe that the manual pressure was applied, although it was more evident when manual pressure was applied in the quadrant where the incision was located. At no time after release of external manual pressure was a negative IOP recorded, but we cannot exclude the possibility of a transient dramatic undershoot of baseline (premanual pressure) IOP.
Video 1. Miyake view of a globe that demonstrates ink influx through a self-sealing clear corneal incision in response to intraocular pressure fluctuation.
FIGURE 3.
Real-Time Imaging. (Top) Miyake view of the corneal wound (green line). (Center) Manual pressure is applied on the corneoscleral area (black arrow). (Bottom) Influx of ink (white arrows) is noted upon release of pressure.
Video 2. Miyake view of a globe that demonstrates ink influx through a self-sealing clear corneal incision in response to manual pressure.
Discussion
Clear corneal incisions in phacoemulsification have numerous advantages. A recent report showed that 57% of cataract surgeons prefer clear corneal incisions, and 92% of these surgeons favor no-suture closure.3 However, reports over the past several years have raised concerns about the increasing incidence of endophthalmitis and its possible association with clear corneal incision. A number of authors have reported significant increase in endophthalmitis with clear corneal approach compared with scleral tunnel incisions.4, 6, 8
In this study we attempted to simulate the functional dynamics of clear corneal incisions in a cadaver model. To detect the ingress of potentially contaminated extraocular fluid into the eye, we used India ink, an extremely opaque fluid with a particle size similar to bacteria. The biomass of most bacteria is less than 10 μm, varying over more than 10 orders of magnitude, from 0.2 μm to 750 μm in diameter.16, 17 Similarly, Ahlberg and associates18 have reported that the average particle size of India ink is 10 μm. Therefore, when India ink is detected in the anterior chamber, there is the possibility that bacterial particles of similar size may enter the anterior chamber through the incisions.
Clear corneal incisions are well apposed when adequate IOP is achieved.11 Shingleton and associates,12 however, demonstrated that wide variations of IOP are likely to occur in the postoperative period, frequently dropping to less than 5 mm Hg of IOP. At this pressure, we have demonstrated that gaping of the internal aspect of the incision may occur in some cases and that some eyes may indeed have full-thickness gaping along the length of the clear corneal incision.11
Hypotony not only permits extraocular fluid to enter the eye by opening the corneal incision, it also makes the eye more deformable when subject to manual pressure. Our observations support McGowan's suggestion that a suction mechanism might be responsible for the development of endophthalmitis following clear corneal cataract surgery.19 The manual pressure we applied on these eyes produced slight deformities. When the pressure was relieved, the elastic properties of the globe seemed to create a vacuum effect that allowed India ink from the outside to flow into the intraocular space through clear corneal incisions. Other authors have demonstrated that the simple force exerted by the eyelid during a blink is enough to displace the globe in the anteroposterior direction.20 The amount of posterior movement of the globe varies considerably between individuals, ranging between 0.7 and 1.6 mm, perhaps reflecting differences in the correlation between eyelid tightness and orbital fat consistency. Coleman and Trokel14 measured IOP variations from blinking and lid squeezing in a human subject by direct cannulation. Their study revealed that blinking resulted in pressure increases of up to 10 mm Hg, while squeezing of the lids produced prompt elevation to levels up to 110 mm Hg (or a 90-mm Hg increase) followed by 8-mm Hg undershoot after lid opening relative to baseline IOP. In our study, application of manual pressure was a simulation of the force generated by the eyelid during a blink or a hard squeeze and the sudden release of force as would be seen when opening the eyelid in a blink cycle. Postoperative patients who rub their eyes or who apply pressure on their eyelids during application of medication may generate the same or even higher stress to the wound. Our results support the notion that a vacuum effect is created during this brief period immediately after release of external pressure, as evidenced by the ingress of India ink in some eyes with clear corneal incisions.
It is important to stress that our experimental approach did not demonstrate an invariable inflow of ink in all clear corneal incisions. Some eyes were relatively impermeable to the inflow of extraocular fluid under the particular circumstances tested. That might also be the case in a clinical situation, where not all eyes may be exposed to a similar risk. At this time, however, we are not able to determine which variables related to corneal anatomy or wound geometry may predispose to the occurrence of this event. Moreover, the model used may not reflect identical conditions present in an in vivo situation where the endothelial pump is functioning and may contribute to the self-sealing properties of a clear corneal incision.
In summary, this experimental model may provide an insight into the possible mechanisms by which the inflow of extraocular fluid into the eye through clear corneal incisions occurs. Cataract surgeons should be aware of the possible mechanism by which microorganisms may enter the eye, and perhaps have a low threshold for placing a suture in clear corneal incisions.
Acknowledgements
Our thanks to Central Florida Lions Eye & Tissue Bank for the gracious supply of tissues.
References
- . Cataract surgical techniques. Curr Opinion Ophthalmol. 1993;4:39–43
- . Prospective evaluation of early visual and refractive effects with small clear corneal incision for cataract surgery. J Cataract Refract Surg. 1996;22:1456–1460
- . Practice styles and preferences of ASCRS members—2002 survey. J Cataract Refract Surg. 2003;29:1412–1420
- . Endophthalmitis: scleral tunnel vs. clear corneal incision. In: Buzard KA, Friedlander MH, Febbraro JL editor. The blue line incision and refractive phacoemulsification. New Jersey: Slack, Inc; 2001;p. 53–56
- . Effect of prophylactic antibiotics and incision type on the incidence of endophthalmitis after cataract surgery. Can J Ophthalmol. 2000;35:373–378
- Bacterial endophthalmitis after small-incision cataract surgery. Effect of incision placement and intraocular lens type. J Cataract Refract Surg. 2003;29:20–26
- . Infectious endophthalmitis following sutureless cataract surgery. Arch Ophthalmol. 1991;109:1562–1563
- . Endophthalmitis in the western Sydney region (a case-control study). Clin Exp Ophthalmol. 2001;29:400–405
- . Case-control study of endophthalmitis after cataract surgery comparing scleral tunnel and clear corneal wounds. Am J Ophthalmol. 2003;136:300–305
- Synthesis of the literature on visual acuity and complications following cataract extraction with intraocular lens implantation. Cataract Patient Outcome Research Team. Arch Ophthalmol. 1994;112:239–252
- McDonnell PJ, Taban M, Sarayba M, et al. Dynamic morphology of clear corneal cataract incisions. Ophthalmology 2003. Forthcoming
- . Evaluation of intraocular pressure in the immediate period after phacoemulsification. J Cataract Refract Surg. 2001;27:1709–1710
- . Continuous intraocular pressure measurement by telemetry in alpha-chymotrypsin-induced glaucoma model in the rabbit (effects of timolol, dorzolamide, and epinephrine). J Pharmacol Toxicol Methods. 1996;36:223–228
- . Direct-recorded intraocular pressure variations in a human subject. Arch Ophthalmol. 1969;82:637–640
- Preparation and study of human eyes obtained postmortem with the Miyake posterior photographic technique. Ophthalmology. 1990;97:810–816
- . Dense populations of a giant sulfur bacterium in Namibian shelf sediments. Science. 1999;284:493–495
- Patel D. Preparation of cell suspensions. In: Separating Cells. New York: Springer, 2001;1–13
- . A comparison of the apical dye penetration patterns shown by methylene blue and India ink in root-filled teeth. Int Endod J. 1995;28:30–34
- . Mechanism for development of endophthalmitis (letter). J Cataract Refract Surg. 1994;20:111
- . Interactions of eyelids and tears in corneal wetting and the dynamics of the normal human eyeblink. Am J Ophthalmol. 1980;89:507–516
☆ This work was supported by an unrestricted grant from Research to Prevent Blindness, New York, New York, by an unrestricted grant from Allergan Inc., Irvine, California, and by a grant from the Alcon Research Institute, Forth Worth, Texas.
PII: S0002-9394(04)00332-0
doi:10.1016/j.ajo.2004.03.012
© 2004 Elsevier Inc. All rights reserved.
