30

The Malignant Glaucoma Syndromes

Ian Conner, MD, PhD; Joel S. Schuman, MD, FACS; and David L. Epstein, MD, MMM

The condition of malignant glaucoma,^1,2 sometimes called ciliary block glaucoma^3,4 or aqueous misdirection syndrome, is a dramatic form of secondary angle-closure glaucoma that is sustained despite the presence of a patent peripheral iridectomy. The signature of the condition is that of axial shallowing of the anterior chamber (ie, the crystalline lens or intraocular lens implant has moved forward; Table 30-1). Classically, malignant glaucoma is a postoperative condition encountered after glaucoma filtration surgery, cataract or combined cataract/filtration surgery, or surgical peripheral iridectomy (Figure 30-1). There have also been several reports of its occurrence after laser iridotomy.^5–7

Some have described forms of malignant glaucoma as a preoperative entity.⁸ Although we still have an incomplete understanding of this complicated disorder and we should not be too quick to dismiss new observations, most of these latter observations probably can be explained by other phenomena, such as the normal forward movement of the crystalline lens under the influence of cholinergic drugs,⁹ primary choroidal detachment syndromes resulting in loosening of zonular tension and forward lens movement that accompanies the forward rotation of the ciliary body, undiagnosed crystalline lens subluxation, and certain poorly understood loose lens syndromes. The malignant glaucoma syndromes may share common features (eg, limited vitreous fluid permeability) with other conditions, such as angle-closure glaucoma secondary to central retinal vein occlusion (see Chapter 36) or massive subretinal exudation.¹⁰ However, these latter conditions frequently resolve (although there may be sequelae from the angle-closure episode). In contrast, malignant glaucoma is a sustained progressive process that characteristically results in blindness if not appropriately treated.¹¹ It is incorrect, therefore, to lump all flat chambers into a category of malignant glaucoma. The latter term should only be used to describe a postoperative condition of axial shallowing of the anterior chamber where other causes, such as choroidal detachment and hemorrhage, have been ruled out.

It is because of this characteristic progression to blindness that malignant glaucoma truly deserves to be called malignant. (It may not be wise to use the term malignant in front of patients because of the known connotation with cancer; alternative terms may be used. However, the term malignant glaucoma is well established in the literature, and among clinicians, it seems appropriate to continue to use this and especially to avoid implications of diverse unproven mechanisms implied with use of alternative terminology.)

There is a rich heritage of thought concerning malignant (ciliary block) glaucoma. We owe a great debt to Shaffer,^2–4 Chandler,^1,11–14 Grant,^13,14 and Simmons^14–16 for the evolution of our concepts.

DIFFERENTIAL DIAGNOSIS

It is important to note that not all postoperative angle-closure glaucoma is malignant glaucoma, nor are all postoperative flat chambers associated with elevated intraocular pressure (IOP; Table 30-2). The diagnosis of malignant glaucoma requires at a minimum axial shallowing of the anterior chamber, elevated or normal IOP, the presence of a patent iridectomy or iridotomy, and the absence of suprachoroidal fluid or blood. There may, however, be shallow supraciliary effusions present in malignant glaucoma.

When presented with postoperative angle-closure glaucoma, even with axial shallowing, the first diagnosis to be eliminated is pupillary block. A patent iridectomy or iridotomy (that is not occluded posteriorly) that allows free communication between posterior and anterior chambers must be present. The surgical iridectomy or laser iridotomy must not be too central. If in doubt, the first step is to place another peripheral laser iridotomy. Again, it must be remembered that axial shallowing of the anterior chamber, perhaps initiated by a transient wound leak, can bring the crystalline lens to the pupil and thus lead to pupillary block. (But, as discussed in Chapter 23, pupillary block by itself does not cause axial shallowing of the anterior chamber, only peripheral shallowing.)

In malignant glaucoma, the IOP is typically elevated. However, in the presence of a functional filter or tube shunt, the IOP may appear normal. In contrast, when there is simply overfiltration and secondary choroidal detachment (see Chapter 61) after filtration surgery, the shallow or flat anterior chamber is associated with a very low IOP, usually not in the teens. Thus, although malignant glaucoma is commonly thought of as a flat chamber with high IOP, this concept should be modified in the presence of a functioning filtration bleb.

Bleeding into the suprachoroidal space, which is not uncommon after fistulizing surgery (especially when antifibrotic therapy initially produces postoperative hypotony), actively occupies volume in that space and thus acts as a mass that can rotate the ciliary body forward, release zonular tension, and result in forward lens movement. Delayed postoperative suprachoroidal hemorrhage can thus mimic malignant glaucoma. This condition can be diagnosed by indirect ophthalmoscopy or ultrasound. Such suprachoroidal bleeding usually occurs in a setting of postoperative hypotony accompanied by secondary choroidal effusions, acting to prevent clotting of suprachoroidal blood. Delayed suprachoroidal hemorrhage is a distinct entity, and the term malignant glaucoma should not be used to describe it.

In contrast, the usual postoperative serous choroidal detachment associated with hypotony and overfiltration is a secondary event, which usually does not act as a mass to shallow the anterior chamber. One might simply view this as a secondary response to fill the space in a soft eye. However, there are other conditions where a serous choroidal detachment can act as a mass to shallow the anterior chamber, presenting similarly to a suprachoroidal hemorrhage. Examples of such primary choroidal effusions occur in eyes with nanophthalmos (see Chapter 31), conditions of increased orbital venous pressure such as dural shunt and Sturge-Weber syndrome (see Chapter 47), and active uveal effusion syndromes including the one associated with human immunodeficiency virus (HIV) infection.^17,18 Each of these conditions can cause a secondary angle-closure glaucoma that is distinct from malignant glaucoma and need to be first ruled out in order for a diagnosis of malignant glaucoma to be established.

There are other flat chamber conditions associated with a localized anterior choroidal detachment that appear to be a primary phenomenon. These anterior uveal effusion syndromes have been characterized using ultrasound biomicroscopy and anterior segment optical coherence tomography.¹⁹ Presumably, there is some inflammatory component to these conditions as they seem to respond to corticosteroids and cycloplegics. In the past, these conditions were probably described under categories of ciliary body swelling^20–22 or loose lens syndromes. The literature has reported numerous cases that appeared to be a result of an idiosyncratic systemic drug reaction (see Chapter 37). We obviously have much more to learn about these conditions and other related loose lens conditions, but these nonoperative conditions are not malignant glaucoma and do not involve a primary abnormality in the hyaloid, which we believe to be the underlying and unifying mechanism involved in malignant glaucoma.^23–26

Malignant glaucoma is traditionally associated with smaller, hyperopic eyes. In such eyes that undergo intraocular surgery, the clinician should be on guard for postoperative malignant glaucoma. (It could even be argued that all such eyes should have a surgical peripheral iridectomy at the time of surgery, even if simple cataract surgery alone is performed. There is a strong clinical sense that, in many cases, pupillary block may occur first, followed by malignant glaucoma, and thus this simple surgical iridectomy procedure might reduce the occurrence. Alternatively, perioperative laser iridotomy could also be performed.) However, eyes of any dimension can develop a malignant glaucoma syndrome postoperatively. Some have used the term aqueous misdirection or other names to describe this occurrence, but discrimination in terminology potentially adds to confusion regarding management. As we will describe in this chapter, it is proposed that a similar pathogenic mechanism is involved in malignant glaucoma, whether the eye is hyperopic or not.

UNIFYING CONCEPTS IN PATHOGENESIS

There is evidently much more that we need to learn about this condition that will lead to a clearer understanding of the events involved. One does not wish to oversimplify what is a complex process in eyes with complicated disease and the potential for multiple mechanisms of pathology. Yet, there is a need for some template for our understanding; otherwise, worsening confusion about what to do for the patient can result. For example, lumping suprachoroidal hemorrhage or effusion into a malignant glaucoma category because of lack of clarity about the latter can interfere with appropriate clinical treatment.

What do we know about malignant glaucoma clinically and experimentally? We know that disruption of the anterior hyaloid with the yttrium-aluminum-garnet (YAG) laser by itself can interrupt phakic, aphakic, and pseudophakic forms of the disease.^25–31 We know from Chandler’s observations^11,12 that puncture of the anterior hyaloid through an anterior approach can reverse the condition both in phakic and aphakic eyes. We know that puncture of the hyaloid from a posterior sclerotomy approach,^14,15 as will be described, can cure the condition in phakic malignant glaucoma, if one is far enough anterior. If one remains too far posterior (presumably without disruption of the anterior hyaloid) in performing surgical vitrectomy, the condition of malignant glaucoma can recur.³² Also, although classically the posterior sclerotomy surgical operation was intended to aspirate a posterior pocket of aqueous humor “locked” in the vitreous, the procedure still worked even when liquid vitreous could not be aspirated³² (presumably by the puncture itself disrupting the peripheral anterior hyaloid).

We know that mydriatic-cycloplegic therapy13 can reverse malignant glaucoma in about 50% of affected phakic eyes, but there is risk of recurrence when the therapy is discontinued. Conversely, the use of miotics in postoperative eyes can precipitate the onset of malignant glaucoma.

We know that malignant glaucoma was not rare after surgical iridectomy but is very rare^5–7 after laser iridotomy.

There is also a strong clinical sense that wound leak or overfiltration,³³ even if transient, may predispose an eye to malignant glaucoma, and such events seem to be a common theme in nonhyperopic eyes that develop this condition.

What do we know experimentally?^23,34 At least in enucleated human eyes, we know that malignant glaucoma cannot be produced by simply directing the normal rate of aqueous humor flow (2 to 3 μL/min) entirely into the vitreous.^23,24 In fact, in such circumstances, the anterior hyaloid and vitreous offer insignificant resistance to forward fluid flow. Thus, it would seem that aqueous humor secreted into the posterior chamber normally has free access into (and out of) the vitreous cavity.³⁵ On the other hand, if one limits the amount of available hyaloid for fluid transfer,³⁴ then there is substantial resistance to forward fluid movement from the vitreous, and vitreous volume continues to expand if aqueous humor continues to be directed into the vitreous cavity, and the anterior chamber then shallows. Further, when the anterior chamber is initially decompressed (then reestablished), this process is accentuated.²³

Although the above is true experimentally for the normal aged (enucleated) human eye, in certain nonprimate species such as the enucleated calf eye, the hyaloid normally demonstrates much more resistance to fluid flow.²³ Thus, there clearly are interspecies differences in hyaloid permeability, and it is likely that there are interindividual differences as well. The hyperopic human vitreous may be more viscous and restrictive than that in the myopic eye.

Putting this all together, we propose that the anterior hyaloid, either because of inherent permeability or the amount of available hyaloid surface area for fluid transfer, is involved in the pathogenesis of the malignant glaucoma syndromes. Thus, a maintained increase in total vitreous volume (rather than the presence of sequestered pockets of aqueous humor in the vitreous) is responsible for the axial flattening of the anterior chamber. Expanded vitreous volume can obliterate available peripheral anterior hyaloid for fluid transfer by movement of the hyaloid into apposition with the posterior ciliary body. In this “silent zone” just behind the posterior chamber (and in continuity with it), there is likely great variation in the amount of space between the hyaloid and ciliary body under baseline conditions.³⁶ In the hyperopic eye with a crowded middle segment, the peripheral anterior hyaloid in its normal position is probably quite close to the posterior ciliary body.

Decompression of the anterior chamber as happens with surgical procedures (but not laser iridotomy) might produce movement of the peripheral anterior hyaloid further forward into apposition with the ciliary body. In fact, if one accepts the concept of possible zonular distortion from anterior chamber decompression, one can hypothesize that, with this forward movement, the anterior hyaloid (which is continuous with the posterior zonules³⁷) could be placed into direct apposition with portions of the secreting ciliary processes, even in phakic eyes. Thus, aqueous humor might move directly into the vitreous cavity. This is a hypothesis (and there are other possibilities for the cause of the initial expansion of vitreous volume that is then maintained in an expanded condition from the resulting obliteration of hyaloid surface area), because this process cannot be directly visualized in the phakic eye. On the other hand, this has been directly observed to have occurred in certain aphakic eyes where the hyaloid coated the surface of the ciliary processes.² Presumably, this can also develop in pseudophakic eyes²⁶ with any type of zonular disruption.

The decompression of the anterior chamber that occurs with surgical peripheral iridectomy does not occur with laser iridotomy and results in the movement forward (and peripheral) of the anterior hyaloid, contributing to the above events wherein the anterior hyaloid permeability/surface area is limited. An important clinical corollary of this would seem to be the imperative to minimize this decompression-related shallowing of the anterior chamber both intra- and postoperatively. For example, especially in hyperopic eyes at greater risk, using viscoelastic in the anterior chamber to prevent shallowing intraoperatively might be useful. Using a tighter rather than looser scleral flap with judicious use of suture lysis (one suture at a time) might minimize the chance of postoperative anterior chamber shallowing due to overfiltration that could then lead to a malignant glaucoma process.³³ We believe this approach is beneficial and, in fact, is what we would recommend as a general strategy.

This concept would explain the observed efficacy^25–31 of the YAG hyaloidotomy procedure in phakic, aphakic, and pseudophakic eyes and the above clinical experience with surgical techniques that have indicated the necessity of anterior hyaloid disruption.³² Surgical decompression during predisposing operations could be the initiating event that brings the hyaloid to the ciliary body, which leads to the sustained effect of expanded vitreous volume that shallows the anterior chamber. Mydriatic-cycloplegic therapy might be effective, as originally intended, to tighten the lens-zonular diaphragm¹³ to resist this force, but also by dilation of the ciliary body ring³⁸ movement of the ciliary body outward away from the peripheral anterior hyaloid. Conversely, cholinergic miotics might narrow the ciliary body ring and move the ciliary body closer to the hyaloid.

One also could conceptualize that if it is possible to restore the normal anatomy and the small space normally present between the peripheral hyaloid and the ciliary body in the “silent zone,”³⁶ then without other reinitiating events, such as repeat decompression of the anterior chamber, the vicious circle could be broken. There has been a strong clinical impression that the essential component of any surgical technique for malignant glaucoma is, in fact, to reestablish the normal anatomy and even expand the anterior chamber at the end of the operation.32

Understanding this construct allows one to appropriately diagnose and treat affected patients. For example, an attempt should always be made, at least in the pseudophakic eye, to disrupt the hyaloid (and intervening tissue, such as residual lens capsule) with the YAG laser to allow a channel for free communication between the posterior segment and anterior chamber before proceeding to surgical vitrectomy.²⁶ The critical role of the anterior hyaloid in the pathogenesis seems clear from both clinical and experimental data.

What is missing in this (and any other) construct is an understanding of the initiating event in this whole process (ie, how is vitreous volume initially expanded).

Decompression of the anterior chamber surgically and the resulting movement of the peripheral hyaloid into apposition with the ciliary body could explain the different incidence of malignant glaucoma between surgical iridectomy and laser iridotomy, and there are some consistent experimental data.²³ With zonular distortion that might result from this anterior chamber decompression, it is possible that the peripheral hyaloid might come directly into apposition with the ciliary processes, and aqueous humor could conceivably be directly secreted into the vitreous across the apposed hyaloid. Still, this does not seem quite enough to fully explain the true initiating event.

Perhaps with the sudden onset of pupillary block, which may precede other events, aqueous humor is suddenly forced into the vitreous. That is, some non–steady state event occurs. We do know clinically that, in the classically described surgical procedure for malignant glaucoma, one can inject balanced salt solution over the peripheral iridectomy too vigorously, and vitreous volume can suddenly expand and be maintained with a recurrence of malignant glaucoma. Alternatively, we also know that there is a constant movement of fluid derived from aqueous humor posteriorly through the retina, presumably under the influence of the retinal pigment epithelium pump.^39–44 It is possible that, with the onset of pupillary block and high IOP, this posterior pump is impaired, and vitreous volume suddenly expands.⁴⁵ There is a report of one human eye enucleated shortly after an attack of untreated angle-closure glaucoma, which purportedly did show an expanded vitreous volume (and obliteration of the space between the peripheral hyaloid and ciliary body).⁴⁶

Understanding the initiating event is nevertheless still a missing piece of the puzzle that, if elucidated, might provide important new insights. Yet, there still is good understanding based on both clinical experience and laboratory experimentation on which to successfully treat most patients with this condition, unlike in a former era when malignant glaucoma syndromes were truly blinding.¹¹ Most likely, the initiating event involves some non–steady state phenomenon or anatomical distortion that results in a maintained increase in vitreous volume, which then requires disruption of the anterior hyaloid to reestablish the normal steady state. From a clinical point of view, the (peripheral) anterior hyaloid needs to be disrupted and thus eliminated as a fluid barrier.⁴⁷

If there are sequestered spaces in the vitreous due to inflammation, etc, then the hyaloid potentially needs to be disrupted in multiple locations to permit free communication between the vitreous cavity and the anterior chamber, as is normally the situation.⁴⁸

CLINICAL OCCURRENCE AND THE EXAMINATION OF THE PATIENT

The usual clinical situation is that of a postoperative patient following filtering, combined cataract/filtering, or cataract surgery. Malignant glaucoma can occur any time during the usual postoperative period of a day to weeks, or even occasionally months. Most commonly, there is an indication of the condition on the first postoperative day.

Important times when the possibility of malignant glaucoma should be kept in mind are following suture lysis³³ at the filtration site (where shallowing of the anterior chamber associated with a sudden increase in outflow can subsequently lead to malignant glaucoma). Another time to keep in mind is when routine postoperative cycloplegics are stopped. In this situation, the cycloplegics may have prevented the occurrence of the condition by their therapeutic effects of tightening the lens-zonular diaphragm and dilating the ciliary body ring, and cessation can allow the syndrome to develop.

One begins by carefully examining the filtration bleb or cataract wound. As discussed above, shallowing of the anterior chamber due either to overfiltration or a wound leak (that may have been transient) might initiate the subsequent development of a malignant glaucoma process. More importantly, a shallow chamber from overfiltration or a wound leak needs to be differentiated from malignant glaucoma. In the former situation, the IOP is low, whereas in malignant glaucoma it is usually high, although again it must be emphasized that with a functioning filtration bleb, the IOP might appear within the normal range. In simple overfiltration with a shallow anterior chamber, the IOP is typically in the low single digits. Stated another way, following filtration surgery with an IOP in the low teens, the anterior chamber should not be shallow. Special attention to possible artifacts needs to be considered when measuring the IOP in the presence of a flat chamber.⁴⁹

In a susceptible hyperopic eye, a postoperative shallow anterior chamber due to overfiltration should be more vigorously treated than that in a myopic eye. Such eyes need to be followed very carefully for development of a malignant glaucoma process.

What is the axial depth of the anterior chamber (as well as peripheral depth)? Has there, in fact, been axial shallowing? What is the contour of the iris (is there a convex forward peripheral ballooning of the iris suggestive of pupillary block)? Is the iridectomy or iridotomy patent? Is it far enough peripheral (ie, beyond the lens equator)? Could the iridectomy or iridotomy be sequestered or blocked posteriorly? Has swelling of the crystalline lens from inadvertent violation of the lens capsule occurred (that might be simulating axial anterior chamber shallowing and malignant glaucoma)?

Indirect ophthalmoscopy should be performed, looking for possible choroidal detachment or hemorrhage as alternative diagnoses. In addition, acute central retinal vein occlusion (see Chapter 36) and massive submacular¹⁰ or vitreous hemorrhage could cause a secondary angle-closure glaucoma with shallowing of the axial anterior chamber depth that may mimic malignant glaucoma.

Ultrasound^19,31 may aid in the determination of possible choroidal pathology such as effusion or hemorrhage, which might simulate malignant glaucoma. Ultrasound biomicroscopy can be used to detect the abnormal anterior position of the ciliary processes in malignant glaucoma, and to rule out ciliochoroidal effusion as the cause of axial shallowing. As might be surmised from the above discussion of underlying mechanisms, ultrasound-detected abnormalities in the vitreous are not expected. In fact, except in the case of vitreous hemorrhage producing a pseudomalignant glaucoma syndrome, no vitreous pathology has been consistently observed in malignant glaucoma.

MANAGEMENT OF MALIGNANT GLAUCOMA

The approach to the management of patients with a flat anterior chamber postoperatively is methodical (Table 30-3). Any bleb or wound leak needs to be first addressed. A continued wound leak, once malignant glaucoma has occurred, may confuse the clinical picture because the IOP may not be very high. If there is a leak and no bleb, a large soft contact lens that bridges well over the surgical site can be placed on the eye, and commonly if there is a wound leak, some deepening of the anterior chamber may be observed in 30 minutes. A pressure dressing could also be used in place of a large soft contact lens for a presumed wound leak, but is less predictable (see Chapter 61).

Next, the adequacy of the surgical iridectomy or laser iridotomy needs to be assessed. The iridectomy may be too central or sequestered posteriorly. Is the peripheral iris ballooned forward as if there were pupillary block? If in doubt whether pupillary block exists, another laser iridotomy should be performed, usually at a site remote from the surgical site. Peripheral iridectomy is performed at the time of filtration surgery to prevent pupillary block and also, perhaps even more importantly, to prevent the iris from rolling into the fistula site thus occluding it. If there is any doubt that the original surgical iridectomy might not be accomplishing these goals, then a laser iridotomy at the surgical site would be indicated.

Except in this circumstance, it is usually best to place a laser iridotomy at a site remote from the fistula wound. The laser iridotomy should also be placed so that a YAG hyaloidotomy could be subsequently performed through it, because, often, there is overlying corneal edema and a hazy optical view overlying the original surgical iridectomy site. In the case of a superior wound, one can still place the laser iridotomy superiorly, but away from the original iridectomy and also away from any location where there may have been an adjacent wound leak. It is technically easier in these usually “hot” eyes to perform the laser iridotomy prior to cycloplegic therapy, if this has not already been instituted, while the pupil is still somewhat miotic.

The next step classically (but now after laser iridotomy) would be the use of mydriatic-cycloplegic therapy¹³ in phakic and pseudophakic patients with intact zonules for 5 days prior to surgical vitrectomy. Presently, with the possibility of true cure with a successful YAG hyaloidotomy, and considering the danger of recurrence of malignant glaucoma once cycloplegics are stopped, it is not certain in straightforward cases whether the next step should be mydriatic-cycloplegic therapy or YAG hyaloidotomy. Two points are worth considering here. Often, these are complicated cases with an initially uncertain diagnosis, so if the IOP is controlled medically (see below) and there is no cornea lens touch, then it is not emergent for YAG hyaloidotomy to be performed. In fact, there is a case for quieting the eye first with topical steroids, which might make the procedure easier to perform technically (certainly after corneal edema is improved with IOP lowering). Also, there might be less chance for a recondensation of the hyaloid²⁵ in a quieter eye. On the other hand, the longer the angle is closed, especially in an inflamed eye, the greater the risk of permanent peripheral anterior synechiae. Nevertheless, a reasonable tactic is to lower the IOP medically with aqueous suppressants (beta-blockers, brimonidine, and possibly carbonic anhydrase inhibitors) while simultaneously initiating mydriatic-cycloplegic drug therapy. Osmotics⁵⁰ can also be considered to lower the IOP and to temporarily dehydrate and thus decrease vitreous volume. Of course, if one does too many maneuvers simultaneously, it may be difficult to define exactly what is beneficial. Furthermore, the osmotic effect on vitreous volume may be only temporary.

Stay updated, free articles. Join our Telegram channel

Join