Chapter 238 – Goniotomy and Trabeculotomy
JAMES D. BRANDT
Congenital and pediatric glaucomas are generally regarded as surgical, not medical, disorders because medical therapy alone is rarely effective for these conditions and current techniques of surgery on the anterior chamber angle (goniotomy and trabeculotomy) enjoy a high degree of success. In this chapter the initial surgical approach to the management of congenital and pediatric glaucoma through goniotomy and trabeculotomy is outlined.
Prior to the introduction of goniotomy, congenital glaucoma uniformly resulted in blindness. In the 1930s, up to 7% of children in institutions for the blind suffered from buphthalmos; given the rarity of the disease, it is probable that virtually every bilaterally affected individual was blinded. It, therefore, is no exaggeration to suggest that goniotomy and trabeculotomy have, since their introduction, allowed many individuals to avoid an entire lifetime of blindness.
Attempts to incise the iridocorneal angle across the anterior chamber to treat glaucoma were first described by de Vincentiis in 1893 ; the results were quite poor and the technique fell into disuse. In 1936, Barkan presented a 10-month follow-up of a then unnamed operation in which he incised through the trabecular meshwork into Schlemm’s canal under direct, gonioscopic visualization. In 1938 this new procedure, given the name “goniotomy,” was described in detail, and the requisite instrumentation became available to others. By 1942 the particular usefulness of goniotomy in the treatment of congenital glaucoma was recognized, and the modern era of surgery for congenital glaucoma had begun.
Surgeons soon recognized that a clear cornea and good visualization of angle structures were key ingredients for successful goniotomy; unfortunately, many cases of congenital glaucoma present with cloudy corneas because of markedly elevated intraocular pressure or breaks in Desçemet’s membrane. In such cases, goniotomy becomes little more than a blind procedure. In 1960 Burian and Smith each independently described a new procedure, trabeculotomy ab externo. In this operation, Schlemm’s canal is cannulated from an external approach and a tear is created through the trabecular meshwork into the anterior chamber using a specially designed probe, the trabeculotome. The principal advantage of trabeculotomy is that it can be performed in the presence of an opacified cornea. Further modifications to the procedure and instrumentation by Dannheim and Harms and McPherson have led to a success rate comparable to that of goniotomy. 
PREOPERATIVE EVALUATION AND DIAGNOSTIC APPROACH
The epidemiology and differential diagnosis of primary infantile glaucoma are covered in Chapter 219 . In unilateral cases of congenital glaucoma, in which the usual signs and symptoms of the disease are present, the diagnosis is relatively straightforward. In bilateral cases in young infants, however, the diagnosis may be less certain and the ophthalmologist carries out a brief examination in the office before the child is subjected to a general anesthetic.
In infants younger than 3 months, it is usually possible to perform tonometry using either a handheld Goldmann applanation tonometer or a handheld electronic tonometer, the Tonopen. The Schiøtz tonometer is relatively inaccurate when used to measure intraocular pressure (IOP) of an infant eye and should be used only to gauge a general range of IOP.  The young infant is swaddled and the eyelids are pried apart gently to perform tonometry. If the child is permitted to drink from a bottle after feedings have been withheld for an hour or two, it is usually possible to perform accurate tonometry and a reasonably thorough examination in the office, especially if the use of a lid speculum is avoided. The Tonopen, which has a small (about 2?mm) measuring surface, is well suited to the small, intrapalpebral fissures of an infant and can measure IOPs without the use of undue pressure on the lids. The Tonopen has proved accurate when compared with both cannulation in cadaver eyes and Goldmann tonometry and appears to be reliable in pediatric patients.
The IOP in normal infants is usually in the range 12–15?mmHg (1.6–2.0?kPa). An office measurement of IOP much greater than 20?mmHg (2.7?kPa) in a calm, resting infant should carry a high index of suspicion for glaucoma when other signs and symptoms suggest the disease, as should an asymmetry of >5?mmHg (>0.7?kPa) in suspected unilateral or asymmetric cases. Measurements of IOP while a child cries and resists efforts to hold the eye open are usually not interpretable because the Valsalva maneuver and a squeezed lid may result in a (measured) IOP in the range 30–50?mmHg (4.0–6.7?kPa) even in normal infants.
The optic nerve examination is at the heart of the diagnosis of all forms of glaucoma, and congenital glaucoma, despite its often dramatic anterior segment findings, is no different. In the office, fundus examination is facilitated greatly by using a Koeppe diagnostic infant gonioscopy lens that has no central depression; the lens has a lid retention flange that prevents the infant from squeezing out the contact lens. Once the lens is in place, the infant often calms down sufficiently to permit good visualization of the disc; with dilation, fundus photography to document the appearance of the optic nerve is often possible.
If the diagnosis of glaucoma is suspected strongly on the basis of the office evaluation, an examination under anesthetic and simultaneous surgery on the affected eye or eyes are recommended to the child’s parents. The surgeon must devote ample time to counsel and educate the child’s parents about the diagnostic and surgical plan that is recommended, with emphasis at the outset that although the surgery is usually highly successful, a decade or more of careful follow-up is necessary. Frequent examinations under anesthetic, amblyopia therapy, and other interventions are part of this long follow-up care, particularly in early childhood. Because some of the infantile and pediatric glaucomas are genetic in origin, with subsequent siblings at increased risk for the disease, appropriate counseling by a clinical geneticist familiar with ophthalmic disease is often helpful.
The differential diagnosis of congenital and infantile glaucoma is covered in Chapter 219 . In most cases, the diagnosis is relatively straightforward and examination under anesthetic merely confirms the clinical impression of the office examination. Despite the frequently dramatic clinical presentation of glaucoma in children, every one of the signs and symptoms of pediatric glaucoma may be caused by disorders other than glaucoma. The decision to operate must be based on the presence of elevated IOP and a correspondent glaucomatous optic atrophy (if the fundus can be visualized), not on the basis of corneal findings alone. Unilateral buphthalmos may be mimicked by a microphthalmic fellow eye, and this possibility must always be considered.
ALTERNATIVES TO SURGERY
Uncontrolled congenital and pediatric glaucomas represent clinical conditions that are definitively managed on an urgent, if not emergent, basis; in most cases, examination under anesthetic and (probable) surgery are scheduled within 24–72 hours of the initial diagnosis. When, for logistic or medical reasons, this proves impossible, the IOP may be lowered somewhat using aqueous suppressant medications. Topical ß-adrenergic antagonists are often used safely with dosages given every 12 hours, but because the potential for systemic side effects is increased in a small infant, the lowest possible dose is used initially and the infant is monitored carefully. Timolol has been reported to induce apnea in infants, and outpatient use in an unmonitored setting probably should be avoided in premature infants and neonates whose immature respiratory systems place them at increased risk of apnea and sudden infant death syndrome.
The carbonic anhydrase inhibitor acetazolamide may be administered parenterally or orally, 5–10?mg/kg of body weight every 6 hours. The pharmacist may compound an elixir using the parenteral powder—250?mg/5?ml in a strongly flavored elixir is reasonably palatable, especially when chilled.
Miotics are generally avoided in infants and small children. Not only are these agents rarely effective in congenital glaucoma (a fact noted in 1939 by Anderson in the first English-language monograph on congenital glaucoma), but the side effects of these agents may be dramatic in a small infant. The parasympathetic actions of these drugs (diaphoresis, diarrhea, and intestinal and gastric hypermotility) that occur rarely in adults may be pronounced in infants.
Several new, topical glaucoma medications have come into widespread clinical use. The topical carbonic anhydrase inhibitor dorzolamide is probably equivalent in efficacy to parenteral or oral acetazolamide in infants and avoids the potential electrolyte imbalance that systemic administration may cause. The a2 – agonist agents apraclonidine and brimonidine both suppress aqueous humor production and are potent hypotensive agents with few systemic adverse effects in adults; however, in children, brimonidine has been reported to induce both central nervous system depression and apnea in young infants, prompting the U.S. Food and Drug Administration to relabel the medication as contraindicated in young infants. The prostaglandin F2a analog latanoprost is a potent hypotensive agent that has the significant clinical advantage in pediatric patients of once-daily bedtime dosing. The author’s experience with the newer agents has been a mixed one, with some notable successes and notable failures. Because all topical medications have the potential for significant systemic side effects, they should be administered in close cooperation with the child’s pediatrician.
Medical therapy of congenital glaucoma is not recommended for long-term or even intermediate-term use but is employed only if an examination under anesthetic and surgery must be deferred or as a temporary measure while the effects of surgery are awaited (for example, before a ligated tube-shunt device begins to function). It is far better to proceed promptly with the definitive evaluation and treatment.
General anesthetics have a potent and rapid effect on the IOP in both normal and glaucomatous individuals. For this reason, the anesthesiologist responsible for the general anesthetic in cases of congenital glaucoma must be familiar with the issues involved and be able to work with the surgeon to obtain the most accurate tonometry possible during the crucial initial minutes of administration of the anesthetic.
Virtually all of the inhalational anesthetics are known to depress IOP within minutes of their administration, whereas ketamine HCl slowly elevates the IOP as deeper anesthesia is attained. The benzodiazepines do not appear to have a significant effect on IOP when used in preoperative doses. Midazolam HCl is often used as a preoperative sedative in children, and in many cases the child is sedated sufficiently with this medication to enable a quick IOP measurement to be carried out prior to the administration of an inhalational general anesthetic. Such measurement, with the child resting comfortably but not under the influence of the potent inhalational anesthetics, is probably the most accurate. Methohexital (methohexitone) also does not appear to have a significant effect on IOP, and the author has found that this drug, administered rectally in a dose of 40?mg/kg body weight, reliably induces sedation sufficient to perform much if not all of an examination under anesthetic. If surgery is planned, endotracheal intubation is appropriate, especially when a longer, bilateral case is anticipated. The author prefers an oral U-shaped RAE-style endotracheal tube as it is less likely to be dislodged during the head movements required in bilateral goniotomy. The IOP must be measured prior to the administration of any muscular relaxant—depolarization agents such as succinylcholine tend to raise the IOP, whereas nondepolarization agents lower it. 
Essential data that must be obtained during an initial or follow-up examination under anesthetic are IOP, results of optic nerve and fundus examinations, cycloplegic retinoscopic measurement of axial length and corneal diameter (horizontal and vertical), and gonioscopy results.
Aspects of the examination that require a clear media, such as ophthalmoscopy, retinoscopy, or fundus photography, are performed first, followed by axial length measurement, gonioscopy, and corneal diameter measurement. It is useful to record the data in a standardized manner so that on follow-up examinations the information is quickly at hand to determine disease progression.
Preoperative pupillary dilation is avoided for cases in which the appearances of the iris and anterior chamber angle are important when the initial diagnosis is made, and it is unnecessary for cases in which corneal opacities prevent any view of the fundus. However, ophthalmoscopy, retinoscopy, and fundus photography are important parts of the initial and subsequent evaluations of a child who has glaucoma, and preoperative dilation using a short-acting mydriatic should be considered before the operating room is entered to avoid a needlessly prolonged period under anesthetic while pupillary dilation takes place. If necessary, intracameral acetylcholine is instilled during surgery to constrict the pupil to allow safe intraocular maneuvers.
Fundus photography of the optic nerve allows the surgeon to document the extent of the progression of the disease ( Fig. 238-1 ). It is well known that the optic nerve cupping seen in congenital and juvenile glaucoma may reverse itself to some extent, and this may be used in some cases to gauge the success of therapy. Similarly, axial length measurements using ultrasonic biometry may help to establish the degree of disparity between the two eyes, particularly in unilateral cases. In many cases the disparity decreases after successful therapy,  and thus a constant or increasing disparity between the two eyes may signal a worsening clinical situation. The BioPen handheld ultrasonic biometric ruler is an easily portable device that may be brought to the operation room for intraoperative biometry; because it is similar in size, operation, and appearance to the Tonopen, it may also be used in the office
Figure 238-1 Fundus photography during an examination under anesthetic. When the clarity of the ocular media permits, this greatly facilitates an evaluation of the response to therapy.
Figure 238-2 Koeppe gonioprisms permit the simultaneous comparison of the two eyes, particularly useful in asymmetric cases. In this child, the perceived buphthalmos of the right eye was the result of a large coloboma and microphthalmia of the left eye. Gonioscopy of the right eye revealed a normal anterior chamber angle, and the intraocular pressure and optic nerve were normal.
setting for young children who have become accustomed to IOP measurements using the Tonopen.
Gonioscopy is a mandatory part of the examination under anesthetic. It allows the surgeon to identify the underlying congenital or juvenile glaucoma diagnosis and to plan the appropriate surgical approach. The Koeppe gonioscopic lens has significant advantages over other styles of indirect gonioprisms. Chief among these is the ability to compare simultaneously the gonioscopic appearance of two eyes in the same patient ( Fig. 238-2 ). The surgeon must be familiar with the gonioscopic findings in primary congenital glaucoma as well as those of various secondary glaucomas seen in childhood; photography helps to document the appearance of the anterior chamber angle before and after treatment.
Horizontal and vertical corneal diameters and the presence or absence of Haab’s striae are noted. Enlargement of the corneal diameter over time is worrisome for progressive buphthalmos, and some clinicians feel these measurements to be more sensitive determinants of progression than axial length determinations.  In the author’s experience, corneal diameter measurements may be difficult to make accurately or consistently because of uncertain landmarks at the corneoscleral limbus and are best used in conjunction with other measurements.
Cycloplegic retinoscopy during follow-up examinations (both under anesthetic and in the office) are used to monitor for progressive axial myopia as well as to provide the spectacle correction to be used in conjunction with patching and/or other amblyopia therapy.
Pediatric glaucoma often presents bilaterally, and the surgeon may have to decide whether to perform bilateral simultaneous
Figure 238-3 Goniotomy ab interno head position. The child’s head is positioned to permit easy access to the temporal limbus under a goniotomy prism. Generally, with a slightly tilted operating microscope, the head is tilted approximately 45° away from the surgeon, who is seated at the side of the operating table, with the assistant at the head.
surgery. In older patients, most surgeons avoid bilateral intraocular procedures for fear of a rare but disastrous bilateral endophthalmitis. This fear is heightened in pediatric patients but is counterbalanced by the perhaps equal or greater potential for neurological complications or death associated with multiple general anesthetics in infancy and early childhood. In a series of 410 goniotomies performed under 340 anesthetics, Litinsky et al. reported no ophthalmic complications associated with the decision to perform bilateral surgery but reported a 1.8% incidence of cardiopulmonary arrest; they concluded that simultaneous bilateral surgery was advisable in bilateral cases. This advice remains appropriate despite advances in anesthetic technique.
Where bilateral surgery is indicated, a fresh, sterile preparation, fresh drapes, and newly autoclaved instruments are used for the second eye. It is the author’s preference to perform surgery on the “worse” of the two eyes first, in case the anesthetic is cut short because of cardiopulmonary complications. The “healthier” of the two eyes may withstand the delay better and less disparity of axial length and refraction results from the delay.
Goniotomy has changed little since it was first described, but the procedure has been made easier and safer with the introduction of better instruments, viscoelastic materials, and the operating microscope. Excellent visualization of the anterior chamber angle is absolutely necessary. Mild corneal edema may sometimes be overcome by denudation of epithelium from the cornea, but corneal clarity may worsen during the course of surgery. If corneal edema or corneal opacities prevent adequate visualization of the anterior angle structures, a trabeculotomy ab externo is performed instead.
It is easiest to perform a goniotomy over the nasal 120° of the anterior chamber angle, using a temporal approach, and this is the usual site for an initial goniotomy. The patient’s head is turned away from the surgeon approximately 45° from the vertical and is secured firmly in place ( Fig. 238-3 ). The operating microscope is oriented temporally and is tilted to permit adequate visualization of the nasal anterior chamber angle once the gonioscopic prism has been placed on the eye.
The assistant grasps the eye at the 6 and 12 o’clock positions using toothed forceps to rotate the eye nasally or temporally at will. The surgeon creates a limbal stab incision through the cornea just anterior to the corneoscleral limbus and injects a viscoelastic substance into the anterior chamber to deepen it. If the eye was dilated prior to surgery, a miotic is injected prior to the introduction of the viscoelastic substance.
The Hoskins-Barkan operating gonioprism permits excellent visualization of the anterior chamber angle, allows the assistant to maintain excellent control of the eye, and permits the surgeon to introduce the goniotomy knife through the stab incision ( Fig. 238-4 ). Balanced salt solution is used as the gonioscopic coupling agent.
Various goniotomy knives are available with which to perform a goniotomy; the author prefers a sharp knife with a slight bevel to the handle. Some surgeons prefer irrigation knives and find them helpful to prevent collapse of the anterior chamber; however, when the corneal wound is distorted minimally, collapse of the anterior chamber is not a problem if viscoelastic materials are used.
The stab incision is made just large enough to permit the introduction of the knife. Great care is taken to avoid damage to the crystalline lens as the knife is advanced across the anterior chamber angle ( Figs. 238-5 and 238-6 ). Anterior movement of the lens-iris diaphragm indicates either a stab incision that gapes or an eye that is under external pressure, which permits egress of the anterior chamber fluids. The surgeon determines whether the stab incision has been distorted by the knife handle or is too large; it may be closed partially using a temporary suture.
Figure 238-4 Goniotomy ab interno. The Hoskins-Barkan lens provides a clear, magnified view of the anterior chamber angle while simultaneously allowing the introduction of the goniotomy knife at the limbus.
Figure 238-5 Goniotomy ab interno knife advancement. The goniotomy knife is advanced through a corneal incision across the anterior chamber. Injected viscoelastic substance deepens the chamber and protects both the corneal endothelium and the crystalline lens. The proper location for the goniotomy incision is in the anterior aspect of the middle third of the trabecular meshwork.
The abnormal tissue that obstructs aqueous outflow in infantile glaucoma has an almost translucent appearance and blurs and distorts the angle anatomy of the anterior chamber. Infants generally have light blue or transparent iris stroma, which makes the landmarks somewhat difficult to identify; in eyes that have primary congenital glaucoma, the iris root usually has a scalloped, almost serrated anterior edge, which provides an excellent landmark for the surgeon. The goniotomy knife is used to place an incision at the anterior aspect of the middle third of the trabecular meshwork, behind Schwalbe’s line; the incision is started at a point 180° away from the corneal incision and rotated either clockwise or counterclockwise as appropriate (see Fig. 238-5 ). The assistant rotates the eye as needed to permit the surgeon to extend the goniotomy incision peripherally without distortion of the corneal incision. The knife is then flipped 180° and the goniotomy is completed in the opposite direction.
If the incision is placed too far anteriorly, the goniotomy does not function; if placed too far posteriorly, only the peripheral iris and iris root are incised, which causes a peripheral iridotomy, iridodialysis, or intraocular hemorrhage and makes further surgery difficult at best. The surgeon is able to visualize directly the iris root, which gently falls posteriorly as the abnormal tissue is incised ( Fig. 238-7 ). The goniotomy may be performed in this manner over approximately 120° of the anterior chamber angle and straddle a point 180° away from the corneal stab incision. A small amount of bleeding is common during a goniotomy, but the amount seen is usually less than occurs during a trabeculotomy.
Figure 238-6 Goniotomy ab interno. Operative photograph demonstrating the goniotomy knife advancing across the anterior chamber.
At the conclusion of the procedure, the remaining viscoelastic material is flushed from the eye. The corneal stab incision, although usually self-sealing, may be closed using a single 10-0 nylon suture. Subconjunctival injections of a corticosteroid and antibiotic are usually given.
As with goniotomy, trabeculotomy ab externo has been made easier and safer by the introduction of viscoelastic materials and the operating microscope. The child’s eye is prepared and draped in a manner similar to that for adult eye surgery. If a superior approach is planned, a superior rectus bridle suture or corneal traction sutures may be placed to depress and expose the superior limbal region. The author prefers to perform trabeculotomy at the temporal limbus, leaving the superior conjunctiva untouched for subsequent filtration procedures if required. A lateral canthotomy may be performed to provide adequate exposure in small or premature infants.
A limbal peritomy is performed and a partial-thickness scleral flap (either rectangular or triangular) is created to sit astride the corneoscleral junction ( Fig. 238-8 ). Excessive cautery is avoided, as the deeper layers of the tissues may be distorted and Schlemm’s canal made difficult to identify.
Figure 238-7 Treated and untreated regions of the anterior chamber angle. Gonioscopy demonstrates a region of treated angle (to the left of center) in which the iris insertion has fallen back nicely; the untreated angle (to the right) shows the typical high insertion of the iris.
Figure 238-8 Trabeculotomy ab externo. A partial-thickness scleral flap is developed to straddle the corneoscleral junction. A radial incision that measures about 2?mm is made using a microsurgery blade, with successive layers gradually pared away to unroof Schlemm’s canal. The canal is identified correctly by cannulation using a 6-0 suture of nylon or polypropylene.
Schlemm’s canal lies just below the transition from white scleral fibers to the blue corneal fibers. A radial incision that measures approximately 2?mm in length and straddles the corneoscleral limbus is made in the bed of the scleral flap. Schlemm’s canal is unroofed by a careful dissection layer by layer until the egress of a small amount of aqueous humor is observed without shallowing of the anterior chamber. Shallowing of the chamber and forward movement of the lens-iris diaphragm indicate that the anterior chamber has been violated directly. If this occurs and it becomes impossible to cannulate Schlemm’s canal, the flap is closed, the anterior chamber is deepened using a viscoelastic substance, and another site is developed.
A short length of 6-0 black nylon or blue polypropylene suture material is then used to cannulate Schlemm’s canal. The suture must not be introduced into Schlemm’s canal forcibly, in case a false passage is created. Slight resistance followed by little or no resistance may indicate that the suture has passed into the suprachoroidal space or into the posterior chamber behind the iris and in front of the crystalline lens. If the suture is bent gently, it should snap back into an orientation tangential to the canal; movement of the peripheral iris in this maneuver indicates that the suture has entered the posterior chamber. In general, the suture passes easily if Schlemm’s canal has been cannulated correctly; intraoperative gonioscopy using either a Koeppe lens or a Zeiss four-mirror gonioprism ( Fig. 238-9 ) may be used to verify correct placement.
Once Schlemm’s canal has been cannulated successfully in both clockwise and counterclockwise directions, a trabeculotome that has an appropriate curvature for the particular eye is chosen. The lower probe of the trabeculotome is introduced into Schlemm’s canal and passed along this in a curvilinear fashion; the upper probe is used as a guide to confirm the location of the lower one ( Fig. 238-10 ). This upper probe is crucial to make certain that the lower probe is oriented in such a way that upon its rotation it ruptures Schlemm’s canal internally, passes parallel to the iris, and does not impinge upon the crystalline lens (see Fig. 238-11 ).
A small amount of bleeding and egress of aqueous humor usually occurs at the base of the trabeculotome probe. If the chamber shallows, viscoelastic may be introduced so that the second trabeculotome can be rotated in the opposite direction without risk to the crystalline lens.
Once the trabeculotomy has been performed on either side of the incision, the overlying scleral flap may be closed using 10-0 nylon. The overlying conjunctiva is tacked over the scleral flap
Figure 238-9 Trabeculotomy ab externo, cannulation. Correct cannulation of Schlemm’s canal may be confirmed using a Zeiss four-mirror gonioprism lens intraoperatively.
Figure 238-10 Trabeculotomy ab externo. A, The upper probe of the instrument provides the surgeon with a three-dimensional cue to the orientation of the lower cannulation probe. B, The lower probe of the trabeculotome is inserted into Schlemm’s canal. If the canal has been cannulated properly, there should be little resistance to the passage of the probe.
Figure 238-11 Trabeculotomy ab externo. The trabeculotome is rotated into the anterior chamber. The surgeon must take great care to ensure that the rotation is in an axis that brings the lower trabeculotome probe into the chamber in a plane parallel to that of the iris. A rotation too far anteriorly may strip Descemet’s membrane; a rotation too far posteriorly risks an iridodialysis with extensive bleeding or direct trauma to the crystalline lens.
using a fine, absorbable suture. Subconjunctival injection of corticosteroid and antibiotic is at the discretion of the surgeon.
Trabeculotomy carried out in this manner opens the angle over almost 180° and straddles the entry site. Beck and Lynch described a technique that employs a 6-0 polypropylene suture to cannulate the entire circumference of Schlemm’s canal ( Fig. 238-12 ), a refinement of the technique first described by Smith. This technique has two notable advantages—no metallic instruments are introduced into the anterior chamber, which reduces the potential for damage to the crystalline lens, and the entire anterior chamber angle is treated, which obviates additional angle surgery if the primary surgery fails. Mendicino et al.  performed a retrospective review of 360° trabeculotomy versus goniotomy, reporting that the 360° procedure was at least as successful as multiple goniotomies and overall success was better than that reported in the literature for conventional angle surgery.
Other surgeons advocate the combination of trabeculotomy and trabeculectomy, finding that the combination results in improved results ; this approach has proved particularly useful in the setting of Sturge-Weber syndrome. 
The two most feared complications of glaucoma surgery, endophthalmitis and choroidal hemorrhage, are exceedingly rare in published clinical series of goniotomy or trabeculotomy. Both procedures are remarkably safe when performed correctly, and few complications that threaten sight are associated with either. During both procedures, the crystalline lens may be damaged by
Figure 238-12 Trabeculotomy ab externo. In a modification of the technique of Beck and Lynch, the entire 360° circumference of Schlemm’s canal is cannulated, using both ends of the suture to “cheesewire” the suture into the anterior chamber.
the instruments in the anterior chamber, but this is avoided by careful technique and the use of intracameral miotics and viscoelastic materials. During trabeculotomy, improper axial alignment of the trabeculotome may strip Descemet’s membrane peripherally (which results in localized corneal edema and scarring), cause an iridodialysis with massive intraocular bleeding, or result in direct trauma to the crystalline lens. During both procedures, a small amount of blood is usually released into the anterior chamber; if no viscoelastic substance is retained in the eye, the blood is usually gone by the first or second postoperative day.
In a prospective clinical trial in which bilaterally affected patients underwent trabeculotomy in one eye and goniotomy in the other, Anderson  reported a success rate of 76% after the first trabeculotomy, and 81% after the first goniotomy. A second procedure raised the ultimate success rate to 92% for trabeculotomy and 100% for goniotomy. The study was small, and the differences were not statistically significant. In the patients uncontrolled after the first operation, both eyes usually required a second procedure, which indicates that the underlying disease, rather than the chosen procedure, is the reason for poor control. In Anderson’s study, patients who presented at birth were excluded from analysis; this and many other series confirm the clinical impression that disease presentation at birth, additional anterior segment anomalies such as Peter’s anomaly, aniridia, and gross buphthalmos are all risk factors for the failure of angle surgery and indicate severe, underlying anatomical abnormalities. Various clinical series suggest that success rates approach 100% for both procedures in infants who present with isolated infantile glaucoma during the first year of life.
The need for a second procedure in some patients has raised the possibility that it might be advantageous to perform two simultaneous goniotomies. In a prospective, randomized study of patients with bilateral infantile glaucoma, Catalano et al. found no significant difference between eyes that underwent a single goniotomy at initial surgery and those that underwent two simultaneous procedures.
With its high success rate and low short- and long-term complication rates in infantile glaucoma, it seems apparent that, with rare exceptions, angle surgery is preferred as the initial therapy for the rare disorders discussed here. In cases in which one or two angle surgery procedures have failed to control the disease, trabeculectomy with the adjunctive use of mitomycin seems a reasonable next step.
Several series of primary trabeculectomies in children have been published, with intermediate-term success comparable to that of angle surgery. With the incidence of bleb infections and endophthalmitis estimated to be as high as 3% per year in antimetabolite-augmented blebs,  it is doubtful whether filtration surgery should be considered primary therapy in pediatric patients if a reasonable chance of success exists with angle surgery. In patients who are aphakic or are likely to become so in the near future, a tube-shunt procedure, such as the Molteno or Baerveldt implant, may be of particular help to permit safer contact lens use and reduce the risk of bleb-associated endophthalmitis.
The surgeon who contemplates the surgical management of infantile glaucoma must recognize that the management of this disorder only begins with the surgical procedure. Despite high success rates (if IOP is considered the sole criterion for success) for both procedures, many children affected by the disorders discussed here grow into adulthood with significant visual impairment because of amblyopia, media opacities, and glaucomatous optic neuropathy. The postoperative care of these children lasts a decade or more, with periodic examinations under anesthetic in infancy and early childhood, frequent cycloplegic refractions, amblyopia therapy, and additional surgery as needed. Once IOP is normalized by surgery, amblyopia becomes the problem; thus, clinical series that report only short-term results with no extended follow-up into the years after visual maturation tend to overestimate the true success of treatment. Only if amblyopia is monitored vigilantly and treatment is aggressive can the full promise of goniotomy and trabeculotomy be realized and patients mature into adulthood with normal vision preserved.
1. de Vincentiis C. Incisione del L’angolo irideo Nel Glaucoma. Ann Ottalmol. 1893;22:540.
2. Barkan O. A new operation for chronic glaucoma. Restoration of physiological function by opening Schlemm’s canal under direct magnified vision. Am J Ophthalmol. 1936;19:951–65.
3. Barkan O. Technic of goniotomy. Arch Ophthalmol. 1938;19:217–23.
4. Barkan O. Operation for congenital glaucoma. Am J Ophthalmol. 1942;25:552–68.
5. Burian HM. A case of Marfan’s syndrome with bilateral glaucoma with description of a new type of operation for developmental glaucoma (trabeculotomy ab externo). Am J Ophthalmol. 1960;50:1187.
6. Smith R. A new technique for opening the canal of Schlemm. Br J Ophthalmol. 1960;44:370–3.
7. Dannheim R, Harms H. Tecnik, Erfolge und Wirkungsweise der Trabekulotomie [Technic, success and mode of action of trabeculotomy]. Klin Monatsbl Augenheilkd. 1969;155:630–7.
8. McPherson SD Jr. Results of external trabeculotomy. Am J Ophthalmol. 1973;76: 918–20.
9. Anderson DR. Trabeculotomy compared to goniotomy for glaucoma in children. Ophthalmology. 1983;90:805–6.
10. McPherson SD Jr, Berry DP. Goniotomy vs external trabeculotomy for developmental glaucoma. Am J Ophthalmol. 1983;95:427–31.
11. Bordon AF, Katsumi O, Hirose T. Tonometry in pediatric patients: a comparative study among Tono-pen, Perkins, and Schiotz tonometers. J Pediatr Ophthalmol Strabismus. 1995;32:373–7.
12. Dietlein TS, Jacobi PC, Krieglstein GK. [Clinical discrepancy between Schiotz and Perkins tonometry in juvenile glaucoma]. Klin Monatsbl Augenheilkd. 1996;209:299–303; discussion 298.
13. Pensiero S, Da Pozzo S, Perissutti P, et al. Normal intraocular pressure in children. J Pediatr Ophthalmol Strabismus. 1992;29:79–84.
14. Boger WPD. Timolol in childhood glaucoma. Surv Ophthalmol. 1983;28(Suppl): 259–61.
15. Olson RJ, Bromberg BB, Zimmerman TJ. Apneic spells associated with timolol therapy in a neonate. Am J Ophthalmol. 1979;88:120–2.
16. Anderson JR. Hydrophthalmia or congenital glaucoma—its causes, treatment and outlook. London: Cambridge University Press; 1939:377.
17. Carlsen JO, Zabriskie NA, Kwon YH, et al. Apparent central nervous system depression in infants after the use of topical brimonidine. Am J Ophthalmol. 1999; 128:255–6.
18. Yang CB, Freedman SF, Myers JS, et al. Use of latanoprost in the treatment of glaucoma associated with Sturge-Weber syndrome. Am J Ophthalmol. 1998;126:600–2.
19. Ausinsch B, Munson ES, Levy NS. Intraocular pressures in children with glaucoma during halothane anesthesia. Ann Ophthalmol. 1977;9:1391–4.
20. Mehta S, Dugmore WN, Raichand M. Ketamine in paediatric ophthalmic practice. Anaesthesia. 1972;27:460–3.
21. Ferrari LR, Donlon JV. A comparison of propofol, midazolam, and methohexital for sedation during retrobulbar and peribulbar block. J Clin Anesth. 1992;4: 93–6.
22. Robertson EN, Hull JM, Verbeek AM, Booij LH. A comparison of rocuronium and vecuronium: the pharmacodynamic, cardiovascular and intra-ocular effects. Eur J Anaesthesiol Suppl. 1994;9:116–21.
23. Quigley HA. Childhood glaucoma: results with trabeculotomy and study of reversible cupping. Ophthalmology. 1982;89:219–26.
24. Sampaolesi R, Caruso R. Ocular echometry in the diagnosis of congenital glaucoma. Arch Ophthalmol. 1982;100:574–7.
25. Kiefer G, Schwenn O, Grehn F. Correlation of postoperative axial length growth and intraocular pressure in congenital glaucoma—a retrospective study in trabeculotomy and goniotomy. Graefes Arch Clin Exp Ophthalmol. 2001;239:893–9.
26. Kiskis AA, Markowitz SN, Morin JD. Corneal diameter and axial length in congenital glaucoma. Can J Ophthalmol. 1985;20:93–7.
27. Litinsky SM, Shaffer RN, Hetherington J, Hoskins HD. Operative complications of goniotomy. Trans Am Acad Ophthalmol Otolaryngol. 1977;83:78–9.
28. Beck AD, Lynch MG. 360 Degrees trabeculotomy for primary congenital glaucoma. Arch Ophthalmol. 1995;113:1200–2.
29. Smith R. Nylon filament trabeculotomy. Comparison with the results of conventional drainage operations in glaucoma simplex. Trans Ophthalmol Soc N Z. 1969;21:15–26.
30. Mendicino ME, Lynch MG, Drack A, et al. Long-term surgical and visual outcomes in primary congenital glaucoma: 360 degrees trabeculotomy versus goniotomy. J AAPOS 2000;4:205–10.
31. Mandal AK, Naduvilath TJ, Jayagandan A. Surgical results of combined trabeculotomy-trabeculectomy for developmental glaucoma. Ophthalmology. 1998;105: 974–82.
32. Mandal AK. Primary combined trabeculotomy-trabeculectomy for early-onset glaucoma in Sturge-Weber syndrome. Ophthalmology. 1999;106:1621–7.
33. Catalano RA, King RA, Calhoun JH, Sargent RA. One versus two simultaneous goniotomies as the initial surgical procedure for primary infantile glaucoma. J Pediatr Ophthalmol Strabismus. 1989;26:9–13.
34. Fulcher T, Chan J, Lanigan B, et al. Long-term follow up of primary trabeculectomy for infantile glaucoma. Br J Ophthalmol 1996;80:499–502.
35. Higginbotham EJ, Stevens RK, Musch DC, et al. Bleb-related endophthalmitis after trabeculectomy with mitomycin C. Ophthalmology. 1996;103:650–6.
36. Greenfield DS, Suner IJ, Miller MP, et al. Endophthalmitis after filtering surgery with mitomycin. Arch Ophthalmol. 1996;114:943–9.