Antifibrotic Agents in Glaucoma Surgery

Patient Groups With a High Risk of Scarring

Patient characteristics associated with a high risk of surgical failure include previous failed trabeculectomies, previous cataract surgery through a conjunctival incision, neovascular glaucoma, inflammatory eye disease, aphakia, neovascular glaucoma, black African ancestry, recent intraocular surgery, young age, and chronic topical medication. Most glaucoma specialists agree that patients with any of these should have higher concentrations of antimetabolite treatment.

The most definitive study of 5-FU injections in high-risk patients was the randomized, prospective National Eye Institute 5-Fluorouracil Filtration Surgery Study, which showed a 51% failure rate after filtration surgery in patients with previous unsuccessful trabeculectomy or in those who had undergone cataract surgery with a conjunctival incision. In the group who received 5 mg 5-FU injections (twice a day for days 1–7, once a day for days 8–14, a total 21 injections), the failure rate was 51% compared with 74% in the placebo group after 5 years. Since then, randomized studies undertaken to compare intraoperative MMC application (0.4–0.5 mg/mL) with 5-FU injections have suggested that in high-risk patients, a single application of MMC provides superior long-term pressure control compared with injections of 5-FU without the risk of keratopathy. Corneal epithelial complications are much more common with injectable 5-FU, but both groups had thin avascular blebs, these being more prominent in the MMC-treated groups. A comparative 3-year study of West African patients found fewer postoperative hypotensive medications were required for the MMC-treated group (0.5 mg/mL for 3 minutes) compared with those who received 5-FU application (50 mg/mL for 5 minutes), although with a total of 68 patients, a statistical difference in success was not reached.

Using a single intraoperative 5-FU application, Egbert et al. carried out a randomized prospective study in a group of West African patients who had a high risk of failure. They showed success rates of 83% in the 5-FU-treated group versus 39% in the control group, with a mean follow-up of 282 days.

The intraoperative MMC regimen, rather than postoperative subconjunctival 5-FU injections, has become the treatment of choice for all patients, and higher concentrations (≥0.4 mg/mL) are considered for high-risk patients. There is increased efficacy, ease of application, and virtual absence of corneal side effects. However, longer periods of follow-up are needed to monitor and determine the development of bleb leaks, hypotony, and endophthalmitis, in view of the thinner and highly avascular blebs sometimes seen with MMC. The incidence of these so-called high-risk blebs has been dramatically reduced, however, because of the introduction of a theory of cystic bleb formation that changed the method of antimetabolite application during glaucoma surgery. Systemic corticosteroids and immunosuppressants are used in particular high-risk groups but are not routine and can have severe side effects if used by an inexperienced physician. The introduction of anti–vascular endothelial growth factor (anti-VEGF) treatment to augment incisional surgery has gained popularity. Although it is unlikely that anti-VEGF agents are as effective in preventing fibrosis after trabeculectomy as MMC, there is evidence that in combination with MMC, low IOP targets are more likely to be achieved. Better bleb morphology and a reduction in bleb vascularization have been observed after intravitreal ranibizumab in combination with intraoperative topical MMC. Subconjunctival bevacizumab has also been used successfully to rescue failing filtering blebs that exhibit neovascularization.

Patient Groups With an Intermediate Or Low Risk of Scarring

In patients with no previous ocular surgery and no significant comorbidity apart from glaucoma, the use of antimetabolites is more controversial. Intraoperative MMC application not only increases the success rate of surgery but also increases the incidence of hypotony, particularly when more stringent surgical measures are not taken to prevent hypotony. Bindlish et al. reported an incidence of late hypotony (IOP <6 mm Hg) in 42% of eyes at some point between 6 months and final follow-up at 5 years, in a series of 123 primary trabeculectomies with MMC. Hypotony occurs more in younger patients, particularly those who have myopia with thin sclera, probably because the sclera is too thin for the scleral flap to achieve an adequate resistance to outflow. Recent surveys of both US and UK glaucoma specialists confirm that the application of MMC is now routine for most subconjunctival incisional surgery with the exception of tube surgery devices.

5-FU injections have been used in lower-risk groups, which include those who have undergone first-time filtration surgery, young patients, and those with normal-tension glaucoma, to achieve lower IOPs and superior success rates. Complications, such as corneal epithelial changes, are more common but generally only in the short term. Cell culture experiments suggest that a single 5-minute application of 5-FU is able to inhibit fibroblasts for several weeks without severe long-term damage and be equivalent to low-dose 5-FU injections. Several studies using single applications of intraoperative 5-FU with short-term follow-up have reported promising results in low-risk patients. Lanigan et al. reported a success rate of 77% in high-risk patients (those with neovascularization, previous failed filtration surgery, aphakia, uveitis, or multiple risk factors), with a 100% success rate in the low-risk groups. Feldman et al. reported an overall success rate of 85% in high-risk patients and a success rate of 92.9% in low-risk patients; no hypotony was seen in the Feldman study. However, both studies were of short duration and had no controls. Therefore failure occurs in higher-risk patients who have a prolonged or aggressive healing response, but in lower-risk patients the success rate is very good (>90%) with no clinically significant hypotony. An improvement in survival was also found for patients in East Africa with a single 5-minute application of low-dose 5-FU (25 mg/mL), interestingly with a much lower failure rate in the control group compared with patients in West Africa. There is evidence that 5-FU trabeculectomy maintains IOP in the low teens compared with nonaugmented surgery.

Comparative studies for primary trabeculectomy have not found one antimetabolite to be clearly superior to another. A trial of 5-FU 50 mg/mL versus MMC 0.2 mg/mL in first-time trabeculectomy (low risk) in the United States did not show any statistically significant differences, to date, in efficacy or side effects of 115 eyes randomized to receive either treatment. These findings were similar to those of WuDunn et al., who used the same doses of 5-FU and MMC. However, the lack of significant differences may be in part because of the postoperative manipulation of the bleb emphasizing the importance of postoperative management.

A single antimetabolite regimen may not be adequate for all patients. The type and dose of drug need to be titrated depending on the individual patient's risk factors and healing response. The authors use a “titratable” regimen that is based on laboratory and clinical data gained from experience using the different single-application agents and concentrations (the authors call this the Moorfields/Florida regimen [“More Flow regimen”]). This regimen evolves constantly, but the present one is summarized in Box 10.28.1 . Postoperative subconjunctival injections of anesthetic/corticosteroid/anti-VEGF agents at the time of surgery and in the clinic with the addition of 5-FU may be used in addition to the application of intraoperative antimetabolites.

Type and Concentration of Intraoperative Agent and Time of Exposure

The optimal concentration of intraoperative MMC has not been established. Chen et al. reported results in high-risk patients since 1981 without failures, but a 66% rate of hypotony with MMC 0.4 mg/mL, a 22% failure rate with no hypotony with MMC 0.2 mg/mL, and a 37% failure rate with no hypotony with MMC 0.1 mg/mL. In a randomized, prospective study of Japanese patients given primary surgery, Kitazawa et al. had a 100% success rate with MMC 0.2 mg/mL (but transient hypotony maculopathy and cataract progression in 18%) and a 64% success rate with MMC 0.02 mg/mL with no hypotony or cataract progression. For intraoperative 5-FU, both 50 mg/mL and the weaker 25 mg/mL have been used, but no direct comparison has been made between the two.

The optimal time of exposure also has not been determined. Megevand et al. retrospectively compared eyes treated with MMC 0.2 mg/mL for 2 or 5 minutes. There was no statistically significant difference in success rate or complications, but hypotony and endophthalmitis still occurred. In another study, MMC was administered intraoperatively at 0.5 mg/mL for 5 minutes or 0.4 mg/mL for 3 minutes in Indian patients. No significant difference occurred in postoperative IOP, hypotony, or postoperative filtration failure rate. However, the group treated with the higher concentration for 5 minutes had a higher incidence of serous choroidal detachment. One patient from each group developed postoperative endophthalmitis during the study period. Shorter applications of 2–3 minutes, compared with 5 minutes, appear to have the same efficacy, but if applications are shortened to less than 2 minutes, suboptimal cellular and tissue absorption may occur. In a study of 5-FU uptake in tissues, the concentrations reached a plateau at 3 minutes. Even shorter application times may result in greater variations in drug delivery.

Changes in the concentration of the agent are more likely to give reproducibly titratable effects compared with variations in exposure time. Therefore to achieve consistent and predictable results, it is probably more important that the individual surgeon becomes accustomed to one or two concentrations and one exposure time. We use MMC 0.2 or 0.5 mg/mL for 3 minutes on the basis of our previous studies.

Type of Sponges

Small variations in technique may have profound effects on the clinical result and complications. The type of sponge may significantly affect the amount of drug delivered. Chen et al. originally used a Gelfoam sponge, but most clinicians use commercially available sponges (e.g., Weck cell, Merocel), which have different retention and drug-releasing capabilities and may be cut to different sizes. Attempts to standardize the dose of drug delivered have been made, and the accurate application of MMC drops to a standard sponge will deliver a known dose. We currently use polyvinyl alcohol sponges (Merocel, Mystic, CT, USA) because these do not disintegrate like methylcellulose and leave fragments in the wound that can cause significant foreign body granulomas. A MMC formulation kit licensed by the US Food and Drug Administration is now available (Mobius Therapeutics, St. Louis, MO, USA).

Larger Areas of Antimetabolite Treatment

On the basis of clinical observation, a hypothesis as to why cystic blebs occur has been put forward ( Fig. 10.28.3 ). By treating larger areas, bleb-related complications can be reduced. This clinical finding has been confirmed experimentally. A specially designed clamp (2-686; Duckworth & Kent, Baldock, UK) for use during either limbal or fornix-based surgery helps protect the conjunctival edge from exposure ( Fig. 10.28.4 ).

Treatment under both the scleral flap and conjunctival flap, particularly in high-risk cases, can be considered. Finally, staining of the antimetabolites with trypan blue has been described permitting the delineation of the treatment area. This principle has been adopted very successfully for current devices with good success rates and diffuse noncystic blebs.