 |
Slide 1 |
Corneal Complications after Cataract Surgery
Richard L. Lindstrom, MD · S. Gregory Smith, MD
Complications of the corneal epithelium after cataract surgery and IOL implantation are rare. However, these complications should be identified and treated.
|
Punctate Epithelial Erosions and Keratitis
After corneal surgery, numerous topical medications may be used, including steroid drops, antibiotic drops, and
antiglaucoma medications. If a patient's eye is particularly sensitive to one of these medications or its
vehicle/preservative, punctate epithelial erosions or keratitis (keratitis medicamentosa, Slide 1) may result,
usually 3 to 4 weeks after surgery. Patients with sensitivity will often have decreased vision, a red eye, or
a foreign body sensation.
|
Slit lamp examination may reveal a diffuse injection of the conjunctiva and punctate staining in the central corneal epithelium (Slide 2). To best detect this, a clinician should use the slit lamp with diffuse oblique illumination and fluorescein dye. If the staining is uniform, punctate staining may be difficult to detect but can usually be appreciated if the excess fluorescein dye is irrigated from the eye.
Punctate epithelial erosions and keratitis are best managed by discontinuing the use of toxic topical medications and substituting topical, nonpreserved lubricants. Generally, the keratitis will clear in 2 to 4 weeks. Certain medications are more frequently associated with corneal toxicity, including beta-blockers, aminoglycosides, epinephrine compounds, benzalkonium chloride, thimerosal, and topical steroids.1 Switching to a drop containing a different preservative (e.g., from prednisolone sodium phosphate to prednisolone acetate, or vice versa) may be necessary. It may also be useful to use a nonpreserved formula of prednisolone acetate in a cellulose gum base.
Special care is required in the treatment of patients with severe ocular surface disease, because the use of epithelial toxic medications, especially nonsteroidal anti-inflammatory drugs (NSAIDs), in conjunction with topical steroids can lead to corneal ulceration.
Allergic Conjunctivitis and Dermatitis
Allergic reactions can be
caused by any topical medication. Allergies should be considered
whenever a patient presents with itchy, red eyes and indurated or
erythematous lids (Slide 2). The cornea may also show inferior staining due to lower lid changes. This problem
will resolve 2 to 3 weeks after the offending medication is discontinued; hydrocortisone lotion (0.5% to 1%) may be used to treat the
lids.
Giant Papillary Conjunctivitis
A patient may present with giant papillary conjunctivitis 6 weeks or more after cataract surgery. Giant papillary conjunctivitis may
occur in as many as 25% of the cases in which sutures are not buried. Often, a patient may complain of intermittent blurred vision,
long strings of mucus in the eye, or a foreign body sensation. In this situation, staining will be found in the superior aspect of
the cornea, rather than in the central cornea. To observe this staining, the clinician must raise the eyelid while the patient looks down.
|
If staining is noted, the upper lid should be everted and inspected for the presence of giant papillary conjunctivitis or hypertrophic papillae (Slide 3). The upper lid also may be swollen and tender and the patient may have a slight pseudoptosis on the upper side. Fluorescein will usually reveal either a suture barb protruding through the conjunctiva or an exposed suture that takes up fluorescein stain.
Treatment includes removal of the suture and application of a topical antibiotic, steroids, and lubrication. The lid will become more comfortable within a few days, with complete resolution expected in 3 to 4 weeks.
|
Dry eyes/blepharitis
A patient who had dry eyes, chronic blepharitis, or meibomianitis preoperatively may present with inferior staining of the cornea after
cataract surgery; the eye may be more sensitive to this type of irritation after surgery. Lid hygiene, including frequently applied warm
compresses, lid scrubs, topical lubricants, and oral tetracycline in recalcitrant cases, is indicated (Slide 4).
|
Corneal Melting
A patient who initially presents with punctate corneal staining, which becomes an epithelial defect, may ultimately develop stromal loss.
The majority of these patients do not complain of pain (Slide 5). Yang and Kline reported five cases of stromal loss among approximately
600 patients, which is an incidence rate of 0.01%.2 Gelender reported five cases of stromal loss and Insler and colleagues reported four
cases.3-5 All of the patients described by Yang and Kline and Insler and colleagues had associated collagen vascular disease. All but one
of these patients had rheumatoid arthritis; the other patient had scleroderma. In Gelender's initial report, only one patient had
Stevens-Johnson disease; the other patients were apparently normal. The studies performed to detect collagen vascular diseases after
the ulcer was detected were negative. More recently, several cases believed to be secondary to extended use of topical NSAIDs have been
reported.
The differential diagnosis for corneal ulceration following IOL implantation includes all the usual causes of this problem. Microbial keratitis - including bacterial, fungal, amoebic, and viral etiologies - should be ruled out by appropriate laboratory studies. Rheumatoid arthritis, Wegener granulomatosis, Sjögren syndrome, malnutrition, hypovitaminosis-A, Mooren ulcer, or a marginal ulcer secondary to blepharitis should be considered. Either radiation therapy or an alkali burn may predispose the eye to ulceration. Medication toxicity should also be considered.
Acute bacterial, viral, or fungal ulcers will usually produce a prominent inflammatory reaction, often with a hypopyon, while an eye with melting syndrome will appear quieter. Postoperatively, if a surgeon discovers a corneal ulcer in a patient, the ulcer should be cultured. Drops that are toxic to the epithelium, such as steroids, NSAIDs, antibiotics, beta-blockers, and epinephrine, should be discontinued.
If the eye is relatively quiet and the problem appears to be a corneal melt rather than an infectious ulcer, frequent lubrication with a nonpreserved drop or ointment should be instituted and the patient should be closely observed, particularly if a descemetocele seems imminent. Hospitalization may be necessary. Bandage contact lenses may also be of value, but the patient should be followed closely because a secondary bacterial ulcer may develop.6
Corneal cyanoacrylate gluing is recommended if the cornea appears to be progressively thinning. Healing results in approximately 60% of these cases.7 Spontaneous perforation should be treated with either a lamellar or penetrating keratoplasty. Gelender reported that a conjunctival flap helps stabilize the cornea.5 Tarsorrhaphy and punctal occlusion may also be considered. The proteases produced by polymorphonuclear neutrophil leukocytes (PMNs) and diseased epithelium may contribute to corneal melting. The eye should be treated with nonpreserved lubricating ointments combined with taping the lid shut or tarsorrhaphy unless the eye is inflamed.
|
Dellen
Corneal dellen can form after cataract surgery with IOL implantation, when the upper lid does not contact the corneal epithelial
surface due to a mass (e.g., filtering bleb or pterygium), usually in the conjunctiva. This area will become dry, with no resurfacing
of the tear film. The adjacent portion of the cornea will become dehydrated. This area will often have the appearance of a corneal melt
and may appear to have fluorescein uptake on the surface (Slide 6).
To differentiate between a dellen and a corneal melt, a clinician should place a generous amount of a lubricating ointment in the eye, tape the eye shut for 20 minutes, and examine the eye after the tape is removed. If the problem is a dellen, the eye will have been rehydrated and the cornea will be of a more normal thickness. If the problem is a corneal melt, there will be no change in the eye's characteristics.
A mass or bleb formation of the conjunctiva near the area of corneal melting is significant. In cataract surgery, this bleb usually occurs if a leaking postoperative wound is covered by conjunctiva. In this situation, the eye should be treated with large amounts of lubricating ointments and observed closely. The wound will eventually seal itself off, usually within 7 to 10 days, and the ointment will no longer be necessary.
Cases of prolonged epithelial defects and loss of corneal stromal substance have been observed following the removal of a pterygium during cataract surgery. Lubrication, bandage contact lenses, patching, and tarsorrhaphy may be considered as treatments for this problem. However, these procedures should not be combined.
|
Detached Descemet Membrane
A patient with a detached Descemet membrane is usually asymptomatic, unless the detachment is severe enough to cause corneal edema.
On slit lamp examination, a transparent membrane that is scrolled superiorly at one edge of the wound can often be seen. Sometimes the
Descemet membrane is avulsed and absent. A large Descemet detachment may appear to be a retrocorneal membrane; however, this is contiguous
with the normal Descemet detachment and endothelium in the periphery (Slide 7).
A detached Descemet membrane is rare, but may occur during any manipulation of the cornea, including enlargement of the corneal wound, implantation of an IOL, or passing of instruments into the eye for intraocular manipulations.
The differential diagnosis to consider is a residual piece of anterior capsule that was not removed from the eye during an anterior capsulectomy. Moses reported one such case, with development of bullous keratopathy that was localized over the remaining piece of capsule.8 This complication can be avoided by the careful placement of instruments inside the eye during cataract surgery. If, upon placing an instrument within the eye, the surgeon meets resistance, the instrument should be retracted and wound size or visualization should be improved before the instrument is again placed into the anterior chamber.
Usually, the extent of detachment is small and there are few sequelae. Peripheral anterior synechia may occur if the detached Descemet membrane contacts the iris. The best treatment of a large detachment is repositioning it surgically. A patient with this condition usually presents on the first postoperative day with a significant amount of corneal edema that does not resolve.
According to Vastine and colleagues, to diagnose a detached Descemet membrane, the surgeon may instill topical anhydrous glycerin and look for the detached Descemet membrane's adhering to the cornea along a thin, gray, usually oblique line.9 An inadvertent detachment can usually be reposited with air or a viscoelastic substance placed in the anterior chamber to compress the detached Descemet to its original position. Air appears to be superior in most cases.
If it encompasses one-third to one-half of the corneal surface area, the Descemet membrane can be reattached to the cornea using 10-0 monofilament sutures passed through the cornea. Vastine and colleagues described this surgical repair procedure in detail. The surgeon makes a paracentesis site at the limbus, inferior and temporal to the previous wound; injects air or a viscoelastic substance to force the detached Descemet membrane into its proper position; places the sutures in a reverse fashion, starting at a point posterior to the limbus, passing anteriorly into the anterior chamber, through Descemet and into the cornea stroma; and ties the suture in a way that does not distort the Descemet membrane.
Sutures may not be needed to correct a detached Descemet membrane if it is recognized intraoperatively. The surgeon may make a stab incision in a similar location away from the wound, and inject air or a viscoelastic substance into the eye to force the membrane back to its original position. This technique has been successful, particularly when the detachment has involved not more than one-forth of the corneal area.
Corneal edema may occur after cataract extraction and IOL insertion.
|
Symptoms and Signs
A patient with corneal edema will complain of decreased vision, especially in the morning (morning edema), and a foreign body sensation.
The decreased vision is caused by irregular astigmatism secondary to epithelial edema. Both of these complaints will be more marked with
increasing degrees of epithelial edema. Clinical detection of mild epithelial edema may be difficult. Examining the patient early in the
morning may aid diagnosis, as may fluorescein staining of the cornea (Slide 8).
Pachymetry can confirm a thickened stroma. The normal thickness of the cornea averages 0.53 mm. Corneal thickness of more than 0.65 mm is suggestive of microcystic edema. Thickness of more than 0.7 mm is almost always associated with microcystic edema.
Early corneal edema may resemble a superficial punctate keratitis when there is only epithelial edema. Slit lamp examination may yield minimal findings, yet a dramatic reduction in vision may occur. Often, the diagnosis of corneal epithelial edema can be made only after the cornea progresses to further decompensation, which may take weeks or months.
Incidence
The incidence of endothelial cell loss leading to corneal decompensation following cataract surgery depends on the type of IOL that was
implanted. Iris-supported IOLs have been associated with the greatest rate of corneal edema, followed by anterior chamber lenses. Posterior
chamber lenses have the best results.
|
Preexisting Diseases
The cornea may develop pseudophakic bullous keratopathy after cataract surgery, once the endothelial cell function and the endothelial
cell population have been reduced to a significant level. A cell count of less than 400 per square millimeters, a pachymetry reading of
more than 650 µm, morning edema, or loose epithelium may signal imminent decompensation of the cornea.
In patients without IOLs, early signs of low corneal reserve may include thickening of the cornea, guttata (Slide 9), or pigmented plaques on the endothelium. The surgeon may follow these patients by using pachymetry and, if available, specular microscopy.
Etiologies for this complication include:
|
The preoperative endothelial cell population may be reduced by surgery, which may precipitate postoperative corneal edema. The average cell loss following cataract or implant surgery has varied widely, depending upon the expertise of the surgeon and the use of viscoelastic materials.
If a patient exhibits corneal guttata or has a history of one of these problems, corneal pachymetry and specular microscopy should be performed preoperatively to evaluate the risk of surgery in relation to corneal decompensation. With this information, the physician can better explain to the patient the risks of postoperative corneal decompensation.
If the endothelial reserve is low, the surgeon should perform the procedure with which he or she is most comfortable. My colleagues and I use viscoelastic substances. Implantation of a posterior or anterior chamber IOL is not contraindicated in these situations, because a corneal transplant performed over a posterior or anterior chamber lens allows satisfactory restoration of vision with few intraoperative complications related to the implant.
Normal Postoperative Course
In the early postoperative period, the amount of striae corneal edema may vary from patient to patient, and may seem unrelated to
the difficulty of the procedure performed. If two patients received the same operation, performed with the same degree of technical
skill, only one may develop a marked response with folds in Descemet membrane on the first postoperative day. A preoperative specular
microscopy examination of the corneal endothelium may give the surgeon an idea of what to expect in any given case.
It is difficult for the surgeon to know how a cornea will react until a patient has been observed for 6 to 8 weeks. Frequently, the cornea with 4+ folds in Descemet membrane will clear markedly on the second postoperative day. However, Yang and Kline demonstrated that the presence of marked striae keratopathy on the first postoperative day is directly related to the amount of endothelial cell loss due to the procedure.10
Phacoemulsification cases appear to exhibit greater edema on the first postoperative day, clear rapidly, and, at 3 months, have cell counts similar to planned extracapsular cataract extraction (ECCE) cell loss. The etiology of these findings is not clear.
For 6 to 8 weeks, folds in Descemet may be noted, particularly in patients with diabetes. These folds interfere minimally with vision, and it has been noted that the corneal endothelium is abnormal in function in diabetic patients, even if specular microscopy reveals an adequate count. Striae edema rated at between a trace and 1+ will usually clear 3 to 4 months postoperatively. The physician should wait at least 4 months before passing final judgment on the status of the cornea.
Iritis, increased intraocular pressure (IOP), and other such conditions can affect the degree to which the cornea may swell; the clinician should try to control these factors before making decisions about corneal transplant surgery. Topical steroids should be used frequently to treat patients with corneal edema. IOP should be maintained between 10 and 16. Very low IOP may lengthen the time required for the cornea to reach normal thickness. Antiglaucoma agents should be discontinued if the pressure is less than 9 mm Hg. Hypertonic agents, such as sodium chloride 5% drops or ointment, may be used.
The clinician should examine the eye closely for IOL endothelial touch or a detached Descemet membrane, which demands surgical correction. If specular microscopy is performed before surgery, the clinician can better inform the patient of the potential risks of corneal decompensation, and obtaining more extensive preoperative information may increase the surgeon's confidence in his or her ability to clear the cornea.
IOL Contact Within the Endothelium
A number of authors pointed out that contact between the intraocular lens and the corneal endothelium results in damage to the endothelial cells.11-16 Experiments with animals indicate that central corneal endothelial density stabilizes at three months.17 Human studies support this conclusion.18 Sugar and colleagues confirmed the presence of endothelial cells adhering to IOLs that were removed.19
Kraff and colleagues performed a study in which they noted the degree of intraoperative trauma (cornea-IOL touch) and rechecked the cell count at least 1 month postoperatively. They noted an increase in cell loss in relation to increasing degrees of contact, which appeared to explain the wide range of cell losses that had been reported.20
These reports of high endothelial cell loss led to the development of viscoelastic agents. A variety of coatings and materials were tried, including polyvinylpyrrolidone,21 albumin, serum TC19922 and Mucin,23 before the development of Healon (sodium hyaluronate, Pharmacia & Upjohn),24 which is a polysaccharide that has a repeating unit of sodium glucuronate and N-acetylglucosamine.
Miller and Stegmann initially reported on the beneficial effects of Healon.25 They analyzed a series of patients undergoing intracapsular cataract extraction (ICCE) and Binkhorst IOL implantations, using Healon in one eye and using air and balanced salt solution in the other eye. Both eyes underwent the same procedure. Results showed an 18% cell loss for the eyes in which Healon was used, and a 54% cell loss in the control eyes. These cell counts were taken at one month.26 Miller and Stegmann showed that Healon statistically reduced endothelial cell loss when used in ECCE, compared to balanced salt solution used in ECCE. Cell counts in this study were performed at 8 weeks.27 Miller and Stegmann showed Healon had a protective effect in secondary IOLs, with a 7.1% cell loss at 2 months.28
Methylcellulose was used as a protective agent by Fechner, who found it to be safe and effective. He did not perform preoperative or postoperative specular microscopy among the patients he studied.29,30
Aron-Rosa and colleagues showed methylcellulose to be more effective than air in protecting human corneal endothelium.31 Smith and colleagues found that methylcellulose reduces endothelial cell loss among animals that underwent surgery (ECCE or keratophacoemulsification (KPE) and posterior chamber IOL).32
In their studies, Smith and colleagues showed methylcellulose to be comparable to Healon in human patients undergoing ECCE or KPE and posterior chamber IOL (12.2% cell loss for methylcellulose and 10% cell loss for Healon). Chondroitin sulfate combined with Viscoat (sodium hyaluronate, Alcon) was also used for this purpose. These reports also give evidence of the ability of methylcellulose to protect the endothelium.33-35
Phacoemulsification
Surgeons have believed that the ultrasonic energy from the phacoemulsification probe or some other factor about the instrument,
may cause an increased rate of endothelial cell loss. Sugar and colleagues reported a cell loss rate of 33.8% with eyes that
underwent phacoemulsification, as opposed to a cell loss rate of 14.9% in eyes that underwent ICCE.36 In this study, the
unoperated eye was used as the control; no cell counts were taken preoperatively. Phacoemulsification was carried out in the
anterior chamber. Differences in cell loss rate were statistically significant. Experimental studies by McCarey and colleagues
and by Polack and Sugar in rabbits, showed large areas of endothelial injury, which they felt was due to the probe.37,38
Binder and colleagues showed experimentally that phacoemulsification injured the endothelium in 17 of 21 eyes on which the surgery was performed.39 Sugar and colleagues concluded that increased time of phacoemulsification does not necessarily correlate with increased cell loss in humans.36
Irvine and colleagues and Waltman and Cozean concluded that endothelial cell loss has no correlation to length of ultrasound time, but is the result of intraoperative trauma.40,41 In an experimental model, Beesley and colleagues showed that after 15 minutes of irrigation, aspiration, and phacoemulsification, no increased damage to the endothelium occurred.42 However, Kraff and colleagues showed that the rate of cell loss for posterior chamber phacoemulsification (15.2%) is significantly less than for anterior chamber phacoemulsification (27.3%).43
The results of these data seem to indicate that posterior chamber emulsification is significantly better for the endothelium than anterior emulsification, and that the length of phacoemulsification time is not a factor in corneal endothelial damage unless the nucleus is in contact with the endothelium. This indication of direct trauma on the part of the nucleus is supported by the experimental evidence reported by Beesley and colleagues and the fact the rate of endothelial cell loss is lower in posterior chamber emulsification, where the nucleus is not in contact with the endothelium.
The surgeon may notice linear areas of endothelial swelling after a portion of the nucleus has come in contact with the endothelium. These lines appear to represent an area of direct injury by the nucleus, and generally resolve within 3 to 5 days. Alfonso and colleagues studied these lines in detail; they refer to them as "snail tracks." They noted that 91% of patients who underwent ECCE by one surgeon had these lines. They performed scanning electron microscopy to confirm endothelial injury in cadaver eyes.44
Lines of endothelial swelling seem to be due to injury by the nucleus or instruments, because these lines do not appear in cases in which phacoemulsification is performed completely in the posterior chamber. Phacoemulsification may also result in a thicker cornea during the first few days following ECCE or ICCE, with a similar cell count noted at 3 months.
Irrigation Fluids
Experimentally, Polack and Sugar demonstrated there is no damage to the endothelium after 10 minutes of irrigation.37
In studies of human and rabbit cornea, Kennerdell and colleagues noted no damage to the endothelium after 500 mL of irrigation and
aspiration.45
Irvine and colleagues and Waltman and Cozean showed no correlation of endothelial cell loss to irrigation time or volume.46,41 Beesley and colleagues also showed no injury to the endothelium in an experimental situation with irrigation and aspiration plus phacoemulsification.42 Binder and colleagues showed a greater than 15% cell loss in five of 10 rabbit corneas that underwent 15 minutes of only irrigation and aspiration.39
Mannis and colleagues and McCarey and colleagues presented evidence that forceful flow may injure the endothelium.47,48 O'Grady and colleagues compared a low flow system with a high flow system but did not analyze endothelial cell loss.49 Despite this lack of analysis, it appears that keeping the amount of irrigation to a minimum may be beneficial to the endothelium.
McCarey and colleagues showed that bicarbonate ringers solution damages the endothelium, but Kinsey medium and glutathione bicarbonate ringers (GBR) apparently do not.48
Edelhauser and colleagues showed that corneas perfused with different solutions swell at different rates.50 More swelling occurred with 0.9% isotonic physiologic saline than occurred with lactated ringers solution than occurred with balanced salt solution than occurred with ringer solution with bicarbonate, reduced glutathione, and adenosine (GBR). Because the GBR solution caused no swelling after 6 hours, they pronounced it superior to the other three, on an experimental level.
Araie studied the effects of GBR, commercially prepared glucose glutathione bicarbonate solution (BSS Plus, Alcon), and a citrate-acetate bicarbonate solution on the barrier function of the endothelium in rabbits. He found no differences between BSS Plus and GBR, but did find that citrate-acetate bicarbonate solution allowed greater permeability than the other two.51 To summarize these studies, GBR and BSS Plus appear to cause the least stress to the endothelium.
Air
Olson reported that air in the anterior chamber damages the endothelium. In an experimental study, Beesley and colleagues
noted that an air bubble on the endothelium for 15 minutes during a phacoemulsification procedure resulted in an area of endothelial
cell loss.42 Considering this, as much air as possible should be removed at the end of a procedure, using a viscoelastic agent or
fluid to protect the endothelium.
Chemical Factors
Van Horn and colleagues indicated that thimerosal, in the accepted concentrations for antimicrobial dosage (0.005% to 0.001%),
may cause functional and structural damage to the endothelium. At higher concentrations (e.g., 0.01%), they noted increased corneal
thickness and endothelial cell death within 1 hour.52
Green and colleagues noted that benzalkonium chloride at 0.0001% and cetylpyridinium chloride at 0.01 mL were toxic to rabbit corneal endothelium. They indicated that the use of ophthalmic medications within the eye at concentrations of 0.004% to 0.02% is contraindicated.53
Hull and colleagues studied the use of epinephrine within the eye after reports were filed indicating corneal edema following the use of 1:1000 epinephrine to dilate the pupil.54-57 They compared GBR to commercial epinephrine (1:1000), diluted commercial epinephrine (1:5000), nonpreserved epinephrine bitartrate (1:1000), and a 0.1% solution of sodium bisulfite (the preservative in the epinephrine solutions), in the perfusion of rabbit corneal endothelium. They noted that 1:1000 epinephrine caused severe damage to the endothelium, as did the sodium bisulfite. Neither the diluted (1:5000) epinephrine nor the nonpreserved epinephrine bitartrate caused any damage.
It appears that none of the preservatives commonly used (thimerosal, sodium bisulfite, benzalkonium chloride) should be placed in the eye. If pupillary dilation is necessary, the clinician can use intracardiac epinephrine diluted to 1:5000.
Miotics
Yee and Edelhauser compared Miochol 1% (acetylcholine chloride, CIBA Vision) to Miostat 0.01% (carbachol chloride, Alcon),
using BSS Plus as a control. They found the two to be equally effective in pupillary constriction, but after in vitro perfusion
of human corneas, Miochol caused marked changes in endothelial function and ultrastructure. They recommend the use of Miostat,
as it appears to be less toxic to the endothelium.58
Postoperative Factors: Foreign Particles on the Endothelium
Cortex
Residual pieces of cortical material that are not removed at the time of irrigation and aspiration may occasionally be seen in
the anterior chamber during the first few postoperative days. They will cause some degree of inflammatory reaction, which can be
controlled with topically applied steroid drops. Resorption usually occurs in 4 to 6 weeks, and this condition should not be of
concern unless the eye develops phacoanaphylactic endophthalmitis.
Nucleus
Occasionally, if phacoemulsification is performed, a nuclear particle will be retained in the eye. This happens most frequently if
the posterior capsule has been ruptured. During the first postoperative week, the surgeon may notice a nuclear particle in the inferior
angle. The particle will incite an inflammatory reaction and should be treated with topical steroids. Resorption will take considerably
longer than 4 to 6 weeks, and a localized area of corneal edema may develop. If the edema progresses the particle should be removed
surgically.
In one case, a 1 mm x 1 mm piece of nucleus lodged in the anterior angle and caused corneal edema. It took approximately 6 to 8 weeks for this to be resorbed. The patient's vision improved according to normal expectations, and no further sequelae developed.
Anterior Capsule
Portions of the anterior capsule may become adherent to the corneal endothelium. Moses described a case in which the capsule
remnant caused a localized area of corneal edema. On the first postoperative day there was no visibility of the capsule, due
to the edema. The edema initially subsided. Later, bullous keratopathy developed in the area.59 Moses treated these cases with
hypertonic agents and bandage contact lenses, but advised surgical intervention if a piece of capsule lodged in the visual axis.
The ideal management of a lodged capsule is prevention by making sure that all of the capsule is removed at the time of surgery. Moses noted that when a portion of capsule becomes adherent to the endothelium, irrigating the area in an attempt to remove it is often unsuccessful. He recommended removing these pieces with forceps if they occur in the visual axis. In his studies, the alternative to removal was chronic, progressive endothelial loss.
Cases of lodged capsule should be observed with serial specular microscopy. If it is noted that a foreign body is causing progressive endothelial cell loss, surgical intervention is indicated to prevent total corneal decompensation.
Subluxated IOLs
The corneal endothelium may go into corneal decompensation postoperatively if it is in contact with a foreign body. Katz and Kaufman
demonstrated that any degree of IOL touch can damage the corneal endothelium, including a subluxated intraocular lens (such as an
anterior chamber lens through an iridectomy, which allows the inferior haptic to ride up on the corneal endothelium, a subluxated
iris-supported lens, or a subluxated posterior chamber lens.60
The cornea may also suffer an increased incidence of cell loss caused by sutures that were placed to maintain intraocular lenses. In particular, iris supported lenses are sutured to the iris to prevent subluxation. These sutures may cheesewire through the iris and loosen themselves, thereby touching the corneal endothelium. If the sutures are tied too long, a similar result will occur. Nylon sutures can also undergo degradation. Although the suture may not be in contact with the endothelium at the time of examination, a sufficiently loose suture in the iris should alert the clinician to this possibility. The problem of degrading sutures is managed by obtaining specular microscopy of the eye at 3-month intervals if the physician doubts whether the suture touches the cornea. If the suture does touch the cornea, it can be bent with an argon laser, or removed.
Peripheral Anterior Synechia
Peripheral anterior synechia also may be associated with progressive endothelial cell loss.
Short Anterior Chamber IOLs
A short anterior chamber IOL may rotate or "propeller" in the anterior chamber. Since the patient will have no symptoms until the cornea
decompensates, it is important for the surgeon to note any change in position of the IOL.59 The surgeon should make a drawing in the
patient's chart. If a change occurs, the patient should be monitored by specular microscopy at minimum intervals of 3 months to document
any progressive cell loss. Ideally, if cell loss is documented, the IOL should be exchanged. Generally, these patients have good
postoperative visual acuitiy.61,62
|
Intermittent Touch
Drews first described the syndrome of intermittent touch. The patient presents with ciliary flush, localized corneal edema, and cystoid
macular edema. Drews originally described this complication with iris plane (Medallion IOL) lenses, but it can occur with anterior
chamber lenses (subluxated through iridectomies) and posterior chamber IOLs (Slide 11). Drews warned that the clinician must be
aware of this etiology whenever any of this triad of signs appears.63
The best way to manage intermittent touch is to recognize it and prevent its recurrence. This usually requires the removal of the implant, but the surgeon should first consider a less invasive measure, such as a McCannell suture.
Progressive Endothelial Cell Loss with IOLs
What does the future hold — possible endocapsular cataract surgery? Perhaps, the capsulorrhexis will eventually go the way of zonulysis, and the ophthalmic surgeon will be able to aspirate the cataract through anterior capsule puncture wounds. Upon completion of nuclear and cortical removal, instead of reinflating the capsular bag with viscoelastic, perhaps the surgeon of the future can reinflate it with an injectable lens material. The patient of the future may no longer be required to choose between best-corrected distance and near vision, but, rather, the accommodation will be preserved. One can only speculate about the possible technological developments. But with the introduction of lasers to the armamentarium of cataract techniques, it is certain that doors are being opened to the future.
The Effect of the Blood-Aqueous Barrier
Studies by several authors showed that endothelial cell loss from surgical trauma stabilizes at 3 months. However, in some
eyes there appears to be a progressive endothelial cell loss that may eventually result in corneal decompensation.6,18,64,65
In 1979, Obstbaum and Galin described the corneoretinal inflammatory syndrome. They indicated that the mediators of inflammation cause damage to the endothelial cells, disruption of the blood-ocular barrier, and increased permeability of the macular vasculature.66
Gelender noted endothelial cell loss, CME, and chronic low-grade uveitis in 19 patients with ICCE and iris-supported lenses.67 He reported that the cell loss was due to chronic irritation, which supports Obstbaum and Galin's view of corneoretinal inflammatory syndrome.
Rowland and colleagues demonstrated in a rabbit model that endothelial cells can be damaged by inflammatory mediators; mediators that affect vascular endothelial cells also affect corneal endothelial cells.68 Early iris-supported implants, lens style, and manufacturer seem to be factors in progressive endothelial damage.
Clinicians observing patients with IOLs should check for signs of precipitates on the IOL or endothelium, chronic iritis, and CME. Patients with signs should be observed at intervals of 3 to 6 months, and specular microscopy should be obtained to look for evidence of progressive cell loss. Treatment with topical steroids is recommended. If there is progressive cell loss, removal of the implant should be considered.
The decision to remove the implant should be based on the expected lifetime of the patient, the status of the fellow eye, the acuity of the eye in question, and the endothelial cell count. If the cell count is less than 400/mm2, corneal decompensation will probably occur in the immediate postoperative period. A cell count greater than 700/mm2 will most likely result in clear corneas if the procedure was done with minimal trauma and the use of viscoelastic substances.
It is easier to consider surgery if the eye has a visual acuity of 20/100 rather than 20/30. The status of the fellow eye may allow a margin of security if the visual acuity is good, or it may obviate the need for surgery.
IOL removal is indicated for a patient 50 years of age with progressive cell loss, whereas it might not be necessary for an octogenarian. The clinician may calculate the rate of cell loss per year to determine the age at which the patient's eye will experience a cell count less than 400/mm2. Leisegang and colleagues investigated posterior IOLs.66
Leisegang and colleagues reported a disturbing trend toward endothelial cell loss, after 1 year, with Shearing posterior chamber IOLs. However, no statistical analysis of the rate of loss was done to show that this was significant. Sampling error, technician error, and the fact that there were large standard deviations (13.7%) may or may not be factors in the overall significance of the data or the rate of change.
Southwick and Olson reported no cases of corneal decompensation among a group of Shearing IOL patients at 5 years.69 Given the data in the Leisegang study, with an initial cell loss of 14% and a yearly cell loss of 4%, it would take more than 30 years for the cornea to reach a cell count of 400/mm2, with an initial cell count of 2000/mm2.
The cell count in patients at 5 years in the study by Leisegang and colleagues was 1925/mm2, which is close to the figure of 1947/mm2 reported by Southwick and Olson. The cell loss reported by Leisegang and colleagues for ECCE at 1 year was 9.4%, which progressed to 11.4 at 2 years. This rate of change is 21%. For the ECCE Shearing IOL, there was a 3.8% change starting from 14.0, a rate of change of 27%. The question arises as to whether this represents a statistically significant change or whether there is no difference between the control (21%) and the shearing IOL group (27%).
Another explanation of the results reported by Leisegang and colleagues, if they are significantly different from the controls, might be that chronic blood-aqueous barrier breakdown caused progressive endothelial cell loss, as described by Obstbaum and Galin.66
While studying blood-aqueous barrier breakdown, Miyake noted a significantly higher leakage of fluorescein among eyes that had sulcus-fixated IOLs than among those that had IOLs placed in the capsular bag (P = 0.02).70 He also noted higher leakage in eyes that had anterior chamber IOLs (closed loops), rather than posterior chamber IOLs (P < 0.1 to P < 0.001). He determined that only eyes with posterior chamber IOLs in the bag were equal to aphakic eyes (KPE alone presumes capsule intact). According to Obstbaum and Galin, the Shearing lens is usually sulcus fixated, which implies a greater endothelial cell loss due to low-grade inflammation evidenced by blood-aqueous barrier breakdown.
Sawa and colleagues also studied the blood-aqueous barrier function and its effect on the endothelium. They topically pretreated the eyes with indomethacin, and noted that endothelial function deteriorates (corneal endothelial transfer coefficient) in the first week, and recovers by 3 months.71 In their study, the presence of a posterior chamber implant did not affect the blood-aqueous barrier any more than did ECCE alone. This study, similar to Miyake's, is not as controlled.
The evidence for chronic inflammation causing corneal decompensation over extended periods of time is quite strong. It appears that inflammation does injure the endothelial cells and that the fluorophotometric findings of blood-aqueous barrier breakdown does have significance.
For the management of patients with these lens types, we recommend using daily topical steroids if precipitates develop on the IOL or the cornea, or if there is other evidence of inflammation. Serial specular microscopy is also indicated and capsular bag fixation is recommended.
YAG Laser
As the YAG laser has become more popular and more widely utilized, its effects on the endothelium have been analyzed. Martin and
colleagues found that damage can be caused by focusing the YAG laser too close to the endothelium.72 The higher the energy used, the
farther the laser should be focused from the endothelium to avoid damage. Using the Q-switched laser, Martin and colleagues found
that for 12 µJ of power a distance of 0.75 mm from the endothelium was necessary to avoid endothelial damage. For a 3 µJ pulse, they
found the necessary distance to be less than 0.25 mm.
The surgeon is not likely to approach the endothelium with YAG laser use at either of these distances, except during a peripheral iridectomy or a YAG lysis of vitreous bands. The figures noted by Martin and colleagues should be kept in mind for these situations. Richburg noted the use of the YAG laser to perform anterior capsulotomies.73 Distances are more relevant in such procedures.
A clinician may notice the appearance of retrocorneal membranes in the postoperative period. These membranes may appear as fine gray lines on the posterior corneal surface or may have pigmentation associated with them. The following etiologies should be considered:
Epithelial Ingrowth
Epithelial ingrowth has the poorest prognosis of these complications. This problem may be confirmed by using an Argon laser to
treat the anterior iris stroma using a 500-µm spot in the medium power range. If the white spot immediately occurs with the laser
treatment, epithelial ingrowth is present.
Feder and Krachmer noted the presence of large cells in the anterior chamber associated with a sudden onset of poorly controlled glaucoma in corneal transplant patients with epithelial ingrowth.74
The eye will usually be injected. If steroids are used, the eye may be white with an anterior chamber reaction in excess of the ciliary flush. Anterior chamber paracentesis may also be helpful in confirming the diagnosis histologically.
This type of eye has a poor prognosis for vision. Treatment, as outlined by Stark and colleagues can be applied.75 The surgeon should mark the extent of the epithelial ingrowth, using the laser immediately preoperatively. Having noted the extent of the involvement of the iris, the surgeon should surgically excise the necessary areas. A freeze-thaw-freeze should then be applied to the retrocorneal surface to kill epithelial cells present on the corneal endothelial surface. The iris also can receive the freeze-thaw-freeze procedure before excision.
The freeze-thaw-freeze procedure should be performed using direct application with the vitrectomy endocryopexy probe under air or a viscoelastic substance. Postoperatively, the rate of corneal edema and iritis is high. This problem was more frequent in intracapsular surgery, considering older suture materials and poor wound apposition. It has become a much less frequent problem, because of small incision cataract surgery and improved sutures.
Epithelial cysts also can be seen on the posterior corneal surface, in the anterior chamber, and on the iris surface. Sugar and colleagues reported on a case of five patients with this problem. The patients may present with no symptoms, only the growth of the cyst. If the cyst is growing, treatment should be instituted. Sugar and colleagues treated their patients with Argon laser therapy (1-second burns, 500-µm spot, 750 mW) with resolution in three of five patients. Retreatment was frequent. None of the cysts became an epithelial downgrowth, but these may occur. If epithelial downgrowth occurs, an iridocyclectomy should be performed.
Stromal Overgrowth
Stromal overgrowth is another source of retrocorneal membranes after cataract surgery. In their 1966 rabbit studies, Brown and Kitano
suggested that a substantial amount of endothelial destruction surrounding a wound in the posterior stroma is a prerequisite for the
occurrence of retrocorneal membranes.76
In 1967, Sherrard and Rycroft determined that retrocorneal membranes seen after surgical trauma are due to stromal elements. Working with rabbits, they found that an intact corneal endothelium seems to inhibit the retrocorneal membrane.77
However, in 1979 Silbert and Baum showed conclusively that the retrocorneal membrane in rabbits is derived from the corneal endothelial cells. Sex chromatin counts of the involved cells in the membrane were studied to define their origin.75
However, other studies indicate that fibroblasts present in a membrane are derived from the stroma.78 Based on the work of Silbert and Baum, this entity is further discussed below as endothelial overgrowth.
Corneal Endothelial Overgrowth/Peripheral Anterior Synechiae
Peripheral anterior synechiae (PAS) may occur following any inflammatory condition of the eye associated with a flat chamber.
In the early postoperative period, if the anterior chamber is flat and there is no response to pressure patching, repair of the eye is indicated to avoid PAS. Vitreous loss may be associated with vitreous and iris incarceration in the wound, and may provide a scaffolding for further fibrotic reaction and increasing peripheral anterior synechia. This is generally a late complication and may be improved by surgery in severe cases.
Rowsey and Gaylor noted that the incidence of PAS reported by Lowe and Drews is as much as 35% in ICCE and 43% in ECCE.79 Harris and colleagues noted that corneal endothelial overgrowth and descemetization of the angle is frequently seen in association with peripheral anterior synechiae and, particularly, iris neovascularization.80 Colosi and Yanoff noted a 22% incidence in surgically enucleated globes.81
Harris and colleagues studied 45 eyes. They noted membranes over the trabecular meshwork continuous with the corneal endothelium. They also noted myofibrillar structures within these cells and indicated that the endothelial cells had undergone myoblastic differentiation. They also reported that fibroblasts on the surface of the iris contribute to the development of ectropion, uvea, and angle closure.
McDonnell and colleagues studied 31 eyes in which vitreous was incarcerated in the wound.82 They noted that, in normal limbal wound healing, the posterior aspect of the wound is bridged by endothelium within two weeks of surgery. They also noted that regeneration of Descemet membrane by the endothelial cells could occur as early as 4 months postoperatively. In their study they found fibrous ingrowth in 84% of eyes, iridovitreal synechiae in 87% of eyes, and CME in 19% of eyes. They noted that corneal endothelial cells had migrated onto the surface of the vitreous strand, with the formation of a basement membrane.
They also demonstrated spindle-shaped fibroblastic cells and collagen fibrils in the vitreous strands. These fibrous cells were felt to be of corneal stromal origin. They also noted lymphocytic infiltration of the iris associated with the iridovitreal synechiae, and that 29% of eyes had chronic inflammatory infiltrates.
McDonnell and colleagues felt that the presence of vitreous in the wound blocks the normal healing process, leading to the formation of the vitreous strands, appearing as a "glassy membrane." These strands appear to be a continuing mechanism for progressive endothelial cell loss.
Rowsey and Gaylor described progressive endothelial cell loss, fibrosis metaplasia of the endothelium, and angle cicatrization with glaucoma, when Choyce and iris plane IOLs are implanted in eyes with peripheral anterior synechiae.79 In particular, this cell loss occurs if the iris or vitreous is adherent anterior to Schwalbe's line. They noted that the formation of a glassy membrane composed of endothelial cells (which, they determined on microscopic study, had undergone fibrous metaplasia) caused a progressive distortion and peaking of the pupil. They noted that the iris fixation of an IOL draws the IOL to the angle with resultant endothelial IOL touch and increases cell loss. They determined that if an anterior chamber IOL is in contact with an area of PAS, progressive cell loss will occur as the migrating endothelial cells attempt to grow onto the surface of the IOL.
The best treatment of this problem is not entirely clear. In postsurgical eyes that are uninflamed and have progressive membrane formation, the Argon laser should be used to differentiate this problem from epithelial ingrowth. If progressive peripheral anterior synechiae is seen, removal of inciting such as an IOL is indicated. Control of iritis is also of value. Some surgeons recommend performing a peripheral iridectomy to prevent further PAS formation.
Rowsey and Gaylor recommended the following treatment principles. If PAS is present preoperatively, and the trabecular meshwork is normal, the synechia can be excised and an IOL can be implanted. They recommended that no IOL be placed in an eye where it comes in direct contact with PAS adherent to the cornea because of the progressive endothelial cell loss that will ensue.79
If the PAS progresses and involves the angle structures (trabecular meshwork) the iris must be removed in a sector fashion, or the surgeon may perform a large peripheral iridectomy. This excision of the iris must extend to the normal trabecular meshwork on either side of the iridectomy.
Rowsey and Gaylor noted that if a small area of PAS remains attached to the angle, the PAS will recur and progress circumferentially around the angle. They used automated vitrectomy to perform these procedures, and indicated that tenting of the iris suggests the presence of these membranes, which are difficult to see in aqueous.
It should be noted that these surgical corrections are, in particular, for PAS clinically associated with glassy membranes. If PAS is noted in an eye (e.g., after pupillary block) that was previously normal, the initial treatment should be goniosynechialysis, using the argon laser as described by Simmons, which may be successful in the early postoperative period.83
If there appears to be no glassy membrane formation, the technique described by Campbell and Vela is recommended.84 The surgery proposed by Rowsey and Taylor is an initial treatment for glassy membrane formation and PAS. If a patient has an anterior chamber IOL and PAS, serial specular microscopy should be performed at 3-month intervals, since the mechanisms of endothelial cell loss described by Rowsey and Gaylor and McDonnell and colleagues. may occur subclinically and without inflammation.
Retrocorneal Melanin Pigmentation
Kampik and colleagues performed a very elaborate histopathologic study on 200 eyes, noting that retrocorneal melanin pigmentation is formed by four cell types. A summary of their work is presented.85
Endothelial Cell Phagocystosis
The Krukenberg spindle is an example of endothelial cell phagocytosis. However, it may be present in a more scattered pattern.
Histologically, the endothelial cells engulfed pigment clumps. Kampik and colleagues indicated that this was a nonspecific response
to the release of pigment granules by the iris.
Iris Pigment Epithelial Cells
The appearance of iris pigment epithelial cells is of darker areas of retrocorneal pigment, with sharp and localized outlines.
These cells may frequently appear next to an area of iridocorneal adhesion (iris-to-wound PAS). Histologically, the corneal
endothelium is absent where these plaques are located. The researchers noted retrocorneal new collagen formation in a large
number of cases. These cells also can be seen as isolated islands, or plaques, on the posterior surface of the cornea. If the
clinician notes such plaques, he or she should realize that the endothelium is not healthy, and should perform specular microscopy.
Pigmented Macrophages
Pigmented macrophages were seen in combination with the other cell types and are determined to be nonspecific to the release of pigment
granules by the iris. Clinically, they are not easily identified.
Iris Stromal Melanocytes The researchers described iris melanocytes as forming a golden-colored plaque on the endothelial surface. Unlike the previous cell types, they are able to migrate over the anterior chamber angle as well as from points of iridocorneal synechiae. These cells can cause secondary glaucoma. The authors said that these cells are stimulated by surgical or accidental trauma and by inflammation. Again, treatment of this problem is difficult, and the prognosis is poor. Control of the inflammation and removal of inciting factors is indicated.
