A 78-year-old white female with a longstanding history of type 2 diabetes mellitus and pseudoexfoliation syndrome was referred to the Glaucoma Service at the Wills Eye Institute in Philadelphia for an elevated IOP in her left eye. At the time of her initial evaluation, she complained of poor vision in both eyes that had limited her ability to read and drive for the past 4 years. Her ocular history was significant for uncomplicated cataract surgery in 1999 and recent steroid injections in her right eye for an undetermined condition. Her past medical history was significant for cardiovascular disease necessitating a coronary bypass, hypertension, and hypothyroidism. She stated that the results of her recent fingerstick glucose tests ranged between 130 and 150 mg/dL and that her family history was negative for glaucoma.
The patient was using brimonidine 0.1% (Alphagan P; Allergan, Inc., Irvine, CA) b.i.d. and travoprost 0.004% (Travatan Z; Alcon Laboratories, Inc., Fort Worth, TX) at bedtime in her left eye only. She also instilled a fixed combination of dorzolamide hydrochloride 2%/timolol maleate 0.5% (Cosopt; Merck & Co., Whitehouse Station, NJ) in both eyes b.i.d.
On physical examination, the patient's BCVA was 20/200 OD and hand motions OS. Refraction revealed mild myopic astigmatism in both eyes. Her IOP with applanation tonometry measured 16 mm Hg OD and 48 mm Hg OS.
The patient was unable to perform automated perimetry with her left eye, but Humphrey visual field testing (Carl Zeiss Meditec, Inc., Dublin, CA) of her right eye revealed a loss of central function. An examination of both anterior segments revealed a posterior embryotoxon in her right eye as well as a 3+ relative afferent pupillary defect, a 3+ nuclear sclerotic cataract, and microcystic corneal edema in her left eye. We also noted Krukenberg spindles, pupillary transillumination defects, and pseudoexfoliative material on the pupillary margin and on the anterior lens capsules in both eyes (left greater than right).
Gonioscopy revealed a wide-open angle with a 2+ pigmented trabecular meshwork and no neovascularization bilaterally. An examination of the posterior segment revealed age-related macular degeneration and nonproliferative diabetic retinopathy in both eyes. The vessels, vitreous, and peripheral retina were normal bilaterally.
Funduscopy of the patient's right eye showed a pink, flat optic nerve with sharp margins and a cup-to-disc ratio of 0.2. The left disc was flat and sharp, with central pallor and a cup-to-disc ratio of 0.9. Based on these findings, we determined that the patient's elevated IOP was due to pseudoexfoliation glaucoma.
HOW WOULD YOU PROCEED?
- Would you treat the patient medically or surgically? If surgically, what procedure would you perform?
- Do any aspects of the patient's past medical history contraindicate specific surgical procedures?
- Based on the patient's medical history, what special instructions would you give to prevent postoperative complications?
SURGICAL COURSE
Because we thought the patient's IOP was too high to allow the safe extraction of the cataract from her left eye, we performed an uneventful canaloplasty with the iTrack microcatheter (iScience Interventional, Menlo Park, CA) and a double-threaded 10–0 Prolene suture (Ethicon Inc., Somerville, NJ).
Postoperatively, the patient's IOPs ranged between the high teens and mid-20s with medical management. Approximately 8 months after canaloplasty, however, her IOP spiked to 48 mm Hg OS. Gonioscopy showed a well-positioned Prolene suture in Schlemm's canal. We therefore decided to lower the patient's IOP by implanting an Ahmed Glaucoma Valve (New World Medical, Inc., Rancho Cucamonga, CA) with a Tutoplast patch graft (IOP, Inc., Costa Mesa, CA).
One day postoperatively, the tube was well positioned, and the patient's IOP was 10 mm Hg OS. Six days after the tube's implantation, however, the patient was rushed to the ER at Wills Eye with complaints of severe pain in her left eye after an episode of vomiting. An examination revealed no light perception and an IOP of 35 mm Hg. Ultrasonography confirmed the presence of a suprachoroidal hemorrhage (Figure 1). The patient's IOP declined to 16 mm Hg with ocular hypotensive medicines, and she reported that she was comfortable. The vision in her left eye remained no light perception.
OUTCOME
All patients receive written and verbal instructions to avoid strain after glaucoma surgery. Between the patient's 1-day follow-up and her presentation to the ER, she experienced persistent emesis that led to the development of a suprachoroidal hemorrhage. She had not reported the vomiting to her surgeon, because she had unsuccessfully battled chronic diabetic gastroparesis for some time and did not think she would benefit from additional treatment.
Unfortunately, when we had elicited the patient's past medical history, we did not specifically inquire about diabetic gastroparesis. If we had known that she was prone to vomiting, we would not have used an Ahmed valve. We would have instead chosen a Baerveldt implant (Abbott Medical Optics Inc., Santa Ana, CA) and tied the tube to allow only a small amount of aqueous to flow through the venting slits. This strategy could have avoided the hypotony that can occur in the early postoperative period with a valved shunt.
DISCUSSION
Many of the risk factors for suprachoroidal hemorrhage are preventable but not treatable.1 For example, chronic, poorly controlled hypertension associated with atherosclerosis cannot be easily reversed in patients who require intraocular surgery. Likewise, elevated IOP may sometimes be reduced immediately prior to glaucoma surgery, but uncontrolled IOP is often the reason for the trip to the OR.
We should nevertheless make an effort to identify patients at risk and take steps to prevent suprachoroidal hemorrhages. Preoperatively, aphakic patients and those with very high IOPs should add a carbonic anhydrase inhibitor to their medical regimens to achieve a maximum reduction in pressure.2 For any intraocular surgery in such an eye, surgeons should slowly decompress the eye via a paracentesis and repeatedly burp the tract through the procedure.2,3
Glaucoma specialists may also reduce patients' risk of postoperative suprachoroidal hemorrhage by using modified trabeculectomy techniques. For example, they should avoid large sclerostomies, small scleral flaps, and the use of loose scleral sutures in eyes with reduced structural rigidity (eg, due to prior vitrectomy), because these techniques facilitate the rapid egress of aqueous.4,5 Surgeons should instead use adjustable sutures to achieve a tight closure of a lamellar scleral flap. Hyaluronate sodium may also be used to avoid vitreous loss in aphakic eyes.
To avoid sudden changes in IOP and hypotony after the implantation of a glaucoma drainage device, surgeons should ligate the proximal tube of a nonvalved implant (eg, a Baerveldt device) with a 7–0 910 polygalactin suture. The IOP in the anterior chamber should be restored to a normal level at the end of the procedure, and surgeons should take steps to minimize postoperative inflammation. The use of topical and peribulbar anesthesia versus general anesthesia during filtering procedures may reduce the risk of postoperative vomiting.
Our patient suffered chronic vomiting after surgery that induced a suprachoroidal hemorrhage and an acute loss of vision in her left eye. Her IOP was sufficiently high prior to surgery that a future loss of vision was certain without surgical treatment. Because this patient's gastroparesis was refractory to treatment, we would have had difficulty minimizing this risk factor.
Glaucoma specialists should consider draining a suprachoroidal hemorrhage after 7 to 10 days if the IOP is markedly elevated and the effusion is large (as in kissing choroidals) and has not resolved. By this time, the clot has usually liquefied and is relatively easy to remove by simultaneously infusing balanced salt solution into the anterior segment and creating a full-thickness radial sclerotomy 3 or 4 mm posterior to the limbus.
Ideally, the sclerotomy should be placed over the meridian of the most elevated portion of the effusion, and drainage ports are typically placed in two separate quadrants (commonly diametrically posed). Surgeons usually reserve vitrectomy for cases in which vitreous is incarcerated in the wound in the anterior segment. Likewise, scleral buckling is used only in eyes that have true rhegmatogenous retinal detachment or unresolvable peripheral vitreoretinal traction.
CONCLUSION
The clinical course described herein highlights the importance of asking all diabetic patients prior to ocular surgery if they have a history of chronic gastroparesis.6,7 Postoperative emesis or coughing may raise episcleral venous pressure, and if this force is transmitted into a hypotonous eye, it may cause a suprachoroidal hemorrhage.8 Under other circumstances, we would have considered draining the hemorrhage within the first few hours of the patient's vision loss. We did not pursue this course of action in this case, however, because the patient had no light perception in her left eye and would not have benefitted from additional intervention.
Scott J. Fudemberg, MD, is an instructor of ophthalmology at the Wills Eye Institute and the Thomas Jefferson School of Medicine, both in Philadelphia. He acknowledged no financial interest in the products or companies mentioned herein. Dr. Fudemberg may be reached at sfudemberg@willseye.org.
Parul Ichhpujani, MD, is a clinical research fellow in the Glaucoma Service of Wills Eye Institute in Philadelphia. She acknowledged no financial interest in the products or companies mentioned herein. Dr. Ichhpujani may be reached at parulichhpujani@aol.com.
Marlene R. Moster, MD, is a professor of ophthalmology at the Thomas Jefferson School of Medicine and is an attending surgeon at Wills Eye Institute, both in Philadelphia. She acknowledged no financial interest in the products or companies mentioned herein. Dr. Moster may be reached at (215) 928-3342; moster@willsglaucoma.org.
