Postoperative Application of Mitomycin for Trabeculectomies

Background  Persistent hypotony is a severe complication following trabeculectomy with intraoperative application of mitomycin.

Objective  To reduce this rate of hypotony by using a lower concentration of mitomycin and applying the mitomycin only topically to the filtering bleb following surgery.

Methods  Patients were enrolled on a consecutive basis and prospectively followed up. Standard trabeculectomies were performed and mitomycin applied postoperatively on the 3 days following surgery (group 1). For comparison, data from previous studies were used for control eyes with intraoperative mitomycin application (group 2) and no mitomycin application (group 3). Preoperative and postoperative data, complications, and the need for further surgical procedures were evaluated.

Results  The study group (group 1) consisted of 22 cases. The mean follow-up was 13.4 and 13.5 months for groups 1 and 2, respectively. Average intraocular pressure values decreased from 33.6 and 31.0 mm Hg (P = .32; t test) to 16.0 and 12.5 mm Hg in the 2 groups (P = .03; t test). The average number of medications decreased from 2.5 and 2.5 to 0.6 and 0.4 (P = .35; t test) in groups 1 and 2, respectively, at the last visit. Hypotony lasting for more than 3 months occurred only in eyes with intraoperative mitomycin application (14/22). Choroidal detachment (3/22) and hypotony maculopathy (2/22) also were only noted in eyes from group 2. In group 3, success rates were much lower.

Conclusions  This is the first clinical study, to our knowledge, to evaluate the efficacy of this new technique of mitomycin application. From the results, it appears that the postoperative application of mitomycin following trabeculectomy is associated with a lower risk of severe and long-standing hypotony. This technique may be promising in eyes at low risk for failure.

PERSISTENT HYPOTONY is a phenomenon frequently associated with the use of mitomycin to enhance the outcome of trabeculectomies.^1,2 This is true for both cases of complicated glaucoma^3,4 and situations in which mitomycin is used for primary procedures.^5-7 An increased outflow of aqueous humor through the filtering site and a toxic effect of mitomycin on the ciliary body epithelium may be important contributing factors for persistent hypotony.⁸ So far, it has been impossible to predict which cases have a higher risk for hypotony, making hypotony impossible to prevent. Factors associated with the application of mitomycin and the amount of mitomycin delivered to the eye include the concentration of mitomycin used, the amount of solution placed on the sponge, the material of the sponge, the duration of application, and whether the sponge is replaced during application.⁷

There is no established rationale for administering mitomycin as a single intraoperative application rather than by another route.⁹ It is not known from a physiologic standpoint whether this way of applying a high concentration of mitomycin is the best way. Although recently it has been shown that Tenon fibroblasts are multidrug resistant,¹⁰ the best time and concentration and the number of applications of mitomycin are still to be determined.

Few variations of the application of mitomycin have been described. Whether it makes a difference if the sponge soaked with mitomycin is placed on the intact sclera or underneath the scleral flap following dissection of the flap is still controversial.^11,12

In a previous study, we examined in an animal model the topical application of mitomycin to the conjunctiva on the 3 days following surgery.¹³ In this report, we present the first clinical results of this new technique of application of mitomycin in a clinical trial. The main advantage of this new technique is that the concentration of mitomycin is much lower. Although, when applied intraoperatively, concentrations of 0.2 to 0.5 mg/mL of mitomycin are frequently used,^14-17 we used the concentration of 0.05 mg/mL for the postoperative application. It may therefore be anticipated that fewer adverse effects and complications are associated with the use of this lower concentration of mitomycin and the different technique of delivery.

In a prospective fashion, consecutive patients with advanced glaucoma and an increased risk of failure of trabeculectomies underwent operation. We aimed to avoid prolonged periods of hypotony for these eyes. Most cases had a diagnosis of secondary glaucoma or end-stage open-angle glaucoma or had already previously undergone trabeculectomies. No eyes with suspected vascular deficiencies or normal tension glaucoma were included, so the intended target pressure following surgery coincided with the usually accepted definitions of surgical failures in glaucoma surgery. This means that an intraocular pressure (IOP) of 21 mm Hg or lower with or without medication was considered clinically sufficient and a complete or qualified success.

Informed consent was obtained from each patient at least 24 hours before surgery. The nature of the specific procedure was explained in detail and alternatives were outlined. In the study group (group 1), mitomycin was applied postoperatively to the conjunctiva overlying the surgical site once on each of the 3 days following trabeculectomy (Figure 1).

Trabeculectomies were performed by 2 surgeons (H.M. and P.C.J.) in the manner originally described by Cairns¹⁸ with slight modifications. A corneal traction suture was placed in the peripheral cornea superiorly to rotate the globe inferiorly. The bulbar conjunctiva was incised 8 to 10 mm posterior to the surgical limbus. After careful dissection toward the limbus and slight cautery of vessels, a half-thickness rectangular scleral flap was prepared. Then, the trabeculectomy was performed, followed by a peripheral iridectomy, and the scleral flap was resutured with 2 single 10-0 nylon sutures. Using a paracentesis, the function of the filter was tested by injecting balanced salt solution into the anterior chamber. Where necessary, additional sutures were put into place. Finally, the conjunctival incision was closed using a single running Vicryl 8-0 suture. At the conclusion of surgery, a corticosteroid and a fortified antibiotic were injected subconjunctivally.

On days 1, 2, and 3 following surgery, the patients received an application of mitomycin on the conjunctiva. For this procedure, patients were placed in a supine position, and topical anesthetic eyedrops were given. The lids were held open using a lid speculum, and a sponge measuring about 5 × 6 mm was placed directly on the filtering site for 5 minutes (Figure 1). The dry sponge (Geuder, Heidelberg, Germany) was soaked with 0.1 mL of mitomycin with a concentration of 0.05 mg/mL. Mitomycin was freshly prepared each time before application. Following removal of the sponge, the whole area was carefully irrigated with 10 mL of balanced salt solution (Figure 1).

For a control group of eyes with intraoperative mitomycin application, we used a historical group of patients who had previously undergone trabeculectomy with mitomycin applied intraoperatively. These results were reported earlier.¹⁹ Of 51 patients, we matched 22 patients in a masked fashion as closely as possible according to their age, sex, diagnosis, and previous ocular surgical procedures.

In these patients, the surgical procedure was performed as described herein, with the exception that a sponge soaked with mitomycin with a concentration of 0.5 mg/mL was applied during surgery before outlining of the scleral flap. The time of application was uniformly 3 minutes. The amount of mitomycin solution given to the dry sponge was 0.1 mL. No postoperative applications of mitomycin were performed in those eyes.

A second historical group had been operated on without the use of mitomycin.²⁰ After receiving a detailed explanation of the risks and benefits of mitomycin use, these patients had decided that they preferred not to receive it. Of 534 patients, 20 controls were matched in a masked fashion as closely as possible according to age, sex, diagnosis, and previous ocular surgical procedures. Because most of the cases were additional trabeculectomies or cases of complicated glaucoma, all of the patients had been offered the use of mitomycin during surgery. Following detailed information, these patients had chosen procedures without the use of antimetabolites.

Following surgery, all patients were treated on an in-patient basis for the first 3 to 5 days and then followed up in our outpatient clinic and/or by their local ophthalmologist. Postoperative medications included a combination of topical corticosteroids and antibiotics given 5 times daily for at least 2 weeks and then slowly tapered. The status of the eyes before, during, and after surgery and all complications during that period were noted. Visual acuities, IOP values, and visual fields using an automated system (Humphrey 30-2, software Peridata 7.0; Zeiss, Germany) were measured on a regular basis.

A significant difference in visual acuity was considered as a doubling of the visual angle equaling a loss of 3 Snellen lines or more. Hypotony was defined as an IOP of less than 10 mm Hg. Although it has been pointed out in many studies that hypotony maculopathy occurs at lower IOP values, it is generally believed that in glaucomatous eyes an IOP of less than 10 mm Hg reflects an abnormality.¹⁹ In addition, such values are usually not present for long periods in eyes undergoing trabeculectomy without antimetabolites.²⁰

For statistical analysis, the 2-way t test for unpaired groups was used. For comparison of success rates, analysis of variance (ANOVA) was used. Surgical success was analyzed with Kaplan-Meier survival analysis.

According to previous clinical studies, success rates were defined as complete surgical success when the IOP was below 22 mm Hg without any medication and qualified surgical success when the IOP was below 22 mm Hg with antiglaucomatous medication. In cases where subsequent surgery was required for IOP control or when the IOP was above 21 mm Hg, the operation was considered a failure. This classification of success does not comply with the definitions used in the Fluorouracil Filtering Surgery Study Group,²¹ in which cases were only considered failures when additional operation was necessary, but does reflect clinical relevance and the criteria used in other studies.^4,19

A total of 22 eyes were operated on with topical postoperative application of mitomycin (group 1) (Table 1). For mitomycin controls, a similar number of eyes receiving intraoperative application of mitomycin were matched (group 2). In addition, a group of 20 eyes of 20 patients who had received no mitomycin was evaluated (group 3). The types of glaucoma included end-stage cases of primary open-angle glaucoma, dysgenetic glaucoma, pigmentary and uveitic glaucoma, and pseudoexfoliation syndrome and cases with chronic angle-closure glaucoma, traumatic glaucoma, iridocorneal endothelial syndrome, neovascular glaucoma, and nanophthalmos. For 7 of 22 cases from group 1 and 10 of 22 cases from group 2, the procedure was an additional trabeculectomy. Previous surgical procedures included argon laser trabeculoplasty, cyclodestructive procedures, peripheral iridectomies, and cataract surgery (Table 1).

Follow-up data were available from all cases in which mitomycin was applied topically (group 1) and ranged from 6 to 28 months (Table 2). For the intraoperative mitomycin group (group 2), the follow-up ranged from 6 to 25 months. The average follow-up for groups 1 and 2 was 13.4 and 13.5 months, respectively.

Average visual acuities were slightly better at the last visit in group 1 compared with the situation before surgery and slightly worse in group 2. These differences were not significant. For variables such as the mean IOP before surgery and at the final visit (both P<.001) and the amount of different medications needed before and after surgery (P<.001), the 2 groups did not differ significantly (Table 2). The mean IOP at the last visit was significantly lower in group 2 compared with group 1 (P = .03). There was a difference between groups in the number of cases in which a significant deterioration of visual acuity occurred, with 0 of 22 cases in group 1 and 3 (14%) of 22 in group 2 (Table 3).

In group 1, an average of 2.5 medications were necessary for IOP control before surgery; this number declined to 0.6 medication at the final visit. In group 2, 2.5 medications were used before surgery and 0.4 at the final visit. This difference at the final visit was not significant (P = .35, t test). In the control group that received no mitomycin, 2.7 medications were used before surgery and 1.7 at the final visit. The reduction of topical therapy was significant within each group.

Surgical procedures following surgery were infrequent, with no significant difference between groups 1 and 2 (Table 4).

The most important complication encountered in both the early and late postoperative phase was hypotony. Hypotony as a transient phenomenon occurred in 8 (36%) of 22 cases in group 1 and 17 (77%) of 22 cases in group 2 (P<.005) (Table 3). Long-standing hypotony developed only in group 2 in 10 (45%) of 22 cases. Hypotony maculopathy resulted in 2 cases (9%) (Table 3). Choroidal detachments as a consequence of hypotony again developed only in group 2. Shallow anterior chambers were also noted more frequently in group 2, whereas larger conjunctival hemorrhages were seen in 2 cases of group 1. The filtering blebs in eyes from group 1 were without large avascular areas and no thinning (Figure 2).

In group 1, no cases of short-term or long-standing corneal erosions or discomfort related to the applications of mitomycin were observed. There was no staining of the conjunctiva following the application. The filtering blebs did not appear avascular as was the case in blebs in group 2.

The need for topical antiglaucomatous therapy during the follow-up period was more frequent in group 1 compared with group 2 (P = .03).

Complete success rates at the final visit were higher in group 2, and the failure rate at the final visit also was higher in that group (Table 5).

Using ANOVA for assessment of survival, no difference was found (P = .64) between the 2 groups. This is also reflected in the Kaplan-Meier graph (Figure 3).

In eyes that received no mitomycin, the success rates were much lower (Table 5). In addition, the average visual acuity loss was higher, with 5.1 lines at an average follow-up of 15.3 months. The average IOP at the final visit was significantly higher than in the mitomycin groups (P<.02; t test). Furthermore, the need for antiglaucomatous medications was higher. The Kaplan-Meier graph comparing no mitomycin and postoperative application of mitomycin is shown in Figure 3. This difference is significant (P<.001; ANOVA).

Since the introduction of mitomycin to glaucoma surgery,¹⁵ several clinical studies have shown the efficacy of mitomycin in improving the outcome of trabeculectomies. This has been demonstrated for cases of both complicated^4,16 and uncomplicated^5,6 glaucoma. Although the concentration of mitomycin used is mostly 0.5 mg/mL,^3-5 other authors used concentrations of 0.2^16,17 or 0.1 mg/mL.¹⁵ The rationale for these different concentrations seems to be clinical experience.

In a recent study, the effect of varying the exposure time of mitomycin to the surrounding tissue was investigated.⁷ In 3 different groups in which no mitomycin, mitomycin for 0.5 to 1 minute, and mitomycin for 3 to 5 minutes were given, it was found that a higher incidence of hypotony was associated with the longest application time of mitomycin. Although the different application times were randomly assigned to the cases, the control group was not matched.

Kitazawa et al⁶ investigated the 2 mitomycin concentrations of 0.2 and 0.02 mg/mL in patients with primary open-angle glaucoma for primary procedures. Although the incidence of complications was similar between the 2 groups, the control of IOP was better in the first. However, since each group consisted only of 11 eyes and the patients did not have secondary glaucoma, it is difficult to determine whether the lower concentration of mitomycin had any effect at all.

An important complication that has been repeatedly commented on in the context of mitomycin surgery is hypotony. The problem of dealing with hypotony starts with the definition. A complete and interesting summary on this topic was recently published.¹ In the present study, we considered hypotony as an IOP below 10 mm Hg measured on at least 2 different occasions. Although such IOP values rarely occur in healthy individuals, especially in elderly patients, such levels are usually not present in eyes with complicated glaucoma. Therefore, this figure may serve as a reasonable cutoff. A second way to look at hypotony may be the incidence of hypotony maculopathy, a functional status that is defined by the clinical pathologic finding and not by a single measured value.

Shields et al²² reported on the outcome of 59 trabeculectomies with mitomycin, which they had tried to separate into high- and low-risk cases. Hypotony was not defined, but at the last visit, 7 (13%) of 52 eyes had an IOP of less than 4 mm Hg. In clinically high-risk cases, mitomycin was applied for 5 minutes, whereas in low-risk cases it was applied for as short as 2 minutes. Interestingly, 4 of 52 patients who had received mitomycin for 2 to 3 minutes only developed hypotony maculopathy. In addition, 4 procedures failed, all of which had exposure times of 3 to 5 minutes. These results suggest that there is no ideal exposure time for mitomycin use in surgery for complex cases of glaucoma, since even a relatively long application of 5 minutes may prove insufficient to prevent filter failure and even a short application may be associated with postoperative hypotony.

Costa et al¹⁷ reported on the results of a series of 169 eyes undergoing trabeculectomy with mitomycin. The concentration of mitomycin used was 0.4 mg/mL, and application times varied from 2.5 to 4 minutes. Hypotony itself was not defined in that study, but hypotony maculopathy was defined as a visual acuity loss of 2 or more Snellen lines and occurred in 5 (3.0%) of 169 cases.

Suner et al²³ gave no data on the incidence of hypotony in their study but had a series of 9 cases during a period of 4 years. Hypotony was defined as an IOP of less than 6 mm Hg. Nuyts et al²⁴ reported on a series of 34 eyes of 32 patients with hypotony maculopathy following surgery with mitomycin. The concentration of mitomycin had been 0.5 mg/mL, which was applied for 5 minutes. Hypotony was defined as an IOP of less than 7 mm Hg. In other studies, the rate of hypotony following trabeculectomy with mitomycin ranges from 4% to 35%.^4,5,14,19

Looking at trabeculectomies in young patients, Jacobi et al²⁵ reported the outcome of surgery performed with and without supplemental mitomycin. Each group consisted of 11 patients, and it was found that, despite similar rates of success and failure in both groups, the incidence of hypotony and hypotony maculopathy was higher in the mitomycin group. In addition, the mean IOP of that group was significantly lower at a follow-up of 5 years.

Explanations for why this hypotony develops still remain controversial. Several studies suggest toxic effects of mitomycin on the ciliary body epithelium. This explanation was substantiated by Gandolfi et al,²⁶ who injected mitomycin subconjunctivally in human eyes with end-stage glaucoma without performing a trabeculectomy. These authors found reduced IOP values afterward. In morphologic studies, damage to the ciliary body epithelium has been demonstrated to occur both in human²⁴ and rabbit eyes.²⁷ In addition, histopathologic abnormalities have been demonstrated in excised filtering blebs of eyes that have been operated on with mitomycin.^22,24,25 Whether these changes really cause excessive leakage of aqueous humor through the conjunctival bleb is unclear. A substantial outflow of aqueous humor would produce a positive Seidel test result when staining the area with fluorescein. This effect has so far not been reported.

The fate of the sometimes large avascular filtering blebs of eyes that have been operated on with mitomycin is still unknown but may bear an increased risk for blebitis and endophthalmitis. In group 1, which was operated on with topical application of mitomycin, no such largely avascular blebs developed (Figure 2).

In the present study, we have described the clinical outcome of trabeculectomies performed with postoperative topical application of mitomycin. The topical application of mitomycin has, to our knowledge, only been described in the context of laser sclerostomies.²⁸ Because this was the first clinical series to evaluate this new technique, we did not randomize the eyes. Therefore, the power of the results of this study may be limited. One major criticism may be that we do not know whether the low concentration and therefore small amount of mitomycin applied to the filtering bleb had an effect at all. However, the cases undergoing surgery had an increased risk of failure; therefore, we speculate that the results in the no mitomycin group would have been worse. In group 2, there were more eyes not requiring antiglaucomatous medications to control the IOP; therefore, one gets the impression that the IOP-lowering effect of intraoperatively administered mitomycin may be larger. Eyes not treated with mitomycin (group 3) had a much worse outcome. This may serve as a hint that the topical application of mitomycin has an effect.

Although the eyes receiving topical mitomycin following surgery had a higher need for antiglaucomatous medications at their final visit, with an average of 13.4 months, the incidence of long-standing hypotony was much lower in that group. The postoperative application of mitomycin is more time-consuming for the surgeon and requires the patient to stay in the hospital or return to the office on the first 3 days following trabeculectomy. In our instance, no additional costs accumulated, because trabeculectomies are usually done on an inpatient basis in our department. Using only topical anesthesia for the application of mitomycin to the filtering bleb, none of the patients complained about pain during the procedure.

The clinical utility of the described postoperative application of mitomycin requires further study. It may be speculated that this application is suitable for low-risk cases or primary procedures. Because in this study the rate of failure was not higher than in the intraoperative mitomycin control group, a randomized study is currently being conducted at our institution.

Accepted for publication June 2, 2000.

Corresponding author: Holger Mietz, MD, Department of Ophthalmology, University of Cologne, 50924 Koeln, Germany (e-mail: h.mietz@uni-koeln.de).