Tube Erosion Following Insertion of a Glaucoma Drainage Device With a Pericardial Patch Graft | Cardiology | JAMA Ophthalmology | JAMA Network
[Skip to Navigation]
Sign In
Case Reports and Small Case Series
September 1999

Tube Erosion Following Insertion of a Glaucoma Drainage Device With a Pericardial Patch Graft

Arch Ophthalmol. 1999;117(9):1243-1244. doi:

Glaucoma drainage devices are currently used to manage high-risk, complicated, adult and pediatric glaucoma when standard filtration surgery with an antimetabolite is unsuccessful.1 The drainage device consists of a plate and a tube. The tube is directly implanted either into the anterior chamber or through the pars plana in eyes undergoing vitrectomy. Tube coverage is imperative to prevent conjunctival erosion, which would lead to tube exposure and pose a risk for the development of endophthalmitis.

Sclera, dura, fascia lata, and pericardium have been employed to cover the tube and fistula sites.2,3 Raviv et al4 recently published safety data in a study of 44 patients (44 eyes) who have had a pericardial patch graft placed to cover a glaucoma drainage device. In their retrospective study, they reported 5 cases of asymptomatic thinning of the pericardial patch graft without evidence of tube erosion. No cases of infection were reported. The mean ±SD follow-up was 10.2±4.0 months. Herein, we report 2 cases of tube erosion through the conjunctiva following use of commercially prepared pericardial patch grafts occurring 7 and 8 months postoperatively.

Report of Cases

Case 1

A 74-year-old African American woman with a long-standing history of primary open-angle glaucoma in both eyes underwent an uncomplicated insertion of a 350-mm2 Baerveldt glaucoma implant (Pharmacia Upjohn Co, Kalamazoo, Mich) in the right eye with a pericardial patch graft through a fornix-based conjunctival incision. One month later, because of conjunctival retraction with exposure of the tube and fistula site, she underwent exchange of the original pericardial patch graft for a larger pericardial patch graft as well as a conjunctival autograft. The original pericardial patch graft appeared intact but was replaced to provide better coverage of the insertion site.

Three months later the patient was found to have asymptomatic melting of the pericardial patch graft without conjunctival erosion. Four months later the patient returned complaining of foreign body sensation and mild ocular tenderness. On examination, her visual acuity was 20/40+1 OD. Slitlamp examination revealed trace conjunctival hyperemia with a 2-mm area of conjunctival erosion and tube exposure. There was no visible pericardial graft tissue in the subconjunctival space. The cornea was clear. The anterior chamber demonstrated 1+ cells and flare without hypopyon. There were no vitreous cells. Fundus examination findings were unremarkable.

The patient was immediately treated with a topical fluoroquinolone (ciprofloxacin) hourly for presumed early endophthalmitis. The following day she underwent repositioning of the tube, which was covered with a 5 × 5-mm piece of full-thickness donor sclera. The conjunctiva was mobilized from the superonasal quadrant to cover the scleral patch graft.

The patient did well with complete resolution of the anterior chamber reaction. The topical ciprofloxacin antibiotics and prednisolone acetate were tapered. All cultures continued to yield no organisms. Fifteen months after surgery, her best corrected visual acuity was 20/25-1 OD, the overlying conjunctiva was intact, and the scleral patch graft had no observable thinning.

Case 2

An 89-year-old African American woman with a long-standing history of primary open-angle glaucoma in both eyes and multiple ocular surgeries underwent insertion of an Ahmed glaucoma valve (New World Medical Inc, Rancho Cucamonga, Calif) in the left eye with a 4 × 4-mm pericardial patch graft through a fornix-based conjunctival incision. Three months after surgery the pericardial patch graft was no longer visible. Five months later she came to the office complaining of a 4-day history of pain and blurred vision in the left eye. On examination, her visual acuity had decreased from 20/30 to 20/400 OS. Slitlamp examination revealed diffuse conjunctival hyperemia and a full-thickness conjunctival defect overlying the tube 3 mm posterior to the limbus. No pericardial tissue was visible. There were no sutures at the site of the defect. The anterior chamber demonstrated 3+ cells and flare with a 2-mm layering hypopyon. There were 1+ anterior vitreous cells. Fundus examination findings were unremarkable.

She underwent a vitreous tap and intravitreal injection of vancomycin hydrochloride and ceftazidime as well as topical and intravenous administration of vancomycin and ceftazidime for presumed endophthalmitis. Five days later she underwent surgical repair of the conjunctival defect. Intraoperatively, no remnants of the pericardial patch graft were visible. A 5 × 5-mm scleral patch graft was placed over the tube and covered with a conjunctival flap that was mobilized anteriorly and secured to the limbus.

Eight months postoperatively, her visual acuity improved to a baseline acuity of 20/30 OS. The conjunctiva was completely intact, and there was no evidence of scleral patch graft thinning.


The use of cadaveric allografts has been gaining wider acceptance for use in ophthalmic surgery, especially for coverage of glaucoma drainage devices. Pericardial, dura mater, and fascia lata patch grafts have been used. The reported advantages with pericardial patch grafts include uniform size and quality, commercial availability without dependence on an eye bank, potentially lower costs, and a processing method that leads to enhanced immunologic safety and reduced risk of viral transmission.4

The dehydration process leaves the graft cell free and without antigenic stimuli. Tissue sterilization with organic solvents as well as low-dose irradiation leads to the inactivation of potential infectious pathogens, including human immunodeficiency virus and Creutzfeld-Jakob virus.5

Despite the favorable result reported by Raviv et al,4 these 2 cases demonstrate that progressive thinning of the pericardial patch graft may occur in patients without predisposing ocular and systemic factors, such as uveitis, or other systemic immunologic disorders. Furthermore, specific antigen-mediated thinning is unlikely to be a major cause of progressive graft resorption owing to the manner in which the pericardial tissue is processed, leaving the tissue virtually antigen free. Surgical factors, such as exposed sutures, tight conjunctiva, or tube malposition, were not contributory in either case. Both cases of erosion occurred at a site over the tube previously covered with a pericardial patch graft. Moreover, in both cases a fornix-based conjunctival flap with superior and temporal relaxing incisions was used at the time of initial surgery to allow tension-free conjunctival apposition to the limbus. Although a conjunctival autograft was used in the first patient owing to retraction, the original pericardial patch graft was still intact. Additionally, the new pericardial patch graft and conjunctival autograft were placed over the tube without tension. Thus, it is unlikely that graft melting and conjunctival erosion, which occurred 7 months later, were related to excessive tension overlying the tube. Asymptomatic thinning of the graft may eventually lead to tube erosion with subsequent development of intraocular infection. Pericardial patch graft thinning was observed in both cases many months before the acute onset of presumed endophthalmitis.

In addition to these 2 cases, we have observed 7 other eyes with asymptomatic thinning of the pericardial patch graft. Currently, the appropriate mode of action in such cases is unknown, and judgment is reserved to the treating physician on a case-by-case basis.

The long-term safety of pericardial patch grafts for tube coverage is currently unknown. Without a prospective randomized study, the relative safety of pericardial vs other patch graft materials cannot be definitively determined. However, careful clinical observation may help identify potential problems with newer materials.

Corresponding author: Paul J. Lama, MD, Department of Ophthalmology, Division of Glaucoma, University of Medicine and Dentistry, New Jersey Medical School, 90 Bergen St, Newark, NJ 07103 (e-mail:

Minckler  D Perspectives on glaucoma drainage implants.  Ophthalmol Clin North Am. 1995;8383- 391Google Scholar
Krupin  TPodos  SMBecker  BNewkirk  J Valve implants in filtering surgery.  Am J Ophthalmol. 1976;81232- 235Google Scholar
Freedman  J Scleral patch grafts with Molteno setons.  Ophthalmic Surg. 1987;18532- 534Google Scholar
Raviv  TGreenfield  DSLiebmann  JMSidoti  PAIshikawa  HRitch  R Pericardial patch grafts in glaucoma implant surgery.  J Glaucoma. 1998;727- 32Google ScholarCrossref
Sattar  SASpringthorpe  VS Survival and disinfectant inactivation of the human immunodeficiency virus: a critical review.  Rev Infect Dis. 1991;13430- 447Google ScholarCrossref