Sir,

We would like to thank Dulku1 and de Klerk and Moriarty2 for their comments on our article.3 Both authors have made pertinent and valuable comments.

We agree with Dulku1 that lack of interval censoring will introduce a right-sided bias and the success rates may be overestimated. This is applicable to all studies on glaucoma surgery, including the recently published tube versus trabeculectomy (TVT) trial.4 This right-sided bias in survival outcomes may become more pronounced with later failures as patients are usually longer, about every 6 months. To mitigate this bias, 95% confidence intervals (CIs) for outcomes of survival analyses should be considered. We did not provide CI data in the article. However, figure 2 illustrates the complete success rates with 95% CIs. At 3 years after surgery, complete success rates with 95% CI were 70–85%.3

de Klerk and Moriarty2 have raised an important point on the routine use of MMC with glaucoma filtering surgery in Caucasian patients. A randomized controlled trial in the USA has failed to show a benefit of MMC over 5-FU in primary trabeculectomy in the long term.5 The evidence for routine use of MMC in primary trabeculectomy is tenuous. It is important to note that in our study most patients were at a higher risk to failure than in their cohort, where 13% had previous cataract surgery. In our study, all patients had previous intraocular surgery and more than half had a previous failed glaucoma surgery. In fact, our patients had a higher number of surgeries per eye than in the aforementioned TVT trial. Our findings suggest that pseudophakia, unlike for trabeculectomy, is not a risk factor for failure of DS. Medication-free success rates at 3 years were 82% for eyes with previous cataract surgery, 71% with previous trabeculectomy and 60% for eyes with both trabeculectomy and cataract surgery. Interestingly, all eyes with delayed hypotony in our study had previous glaucoma surgery. This may be because of aggressive postoperative management in these eyes, such as early laser goniopuncture and needle revision with MMC.

We continuously audit outcomes of surgery performed in our department. Primary DS procedures are now augmented with subconjunctival bevacizumab. After 2 years, we were unable to find any difference in IOP outcomes between bevacizumab and MMC-augmented DS. A long-term audit on combined phacoemulsification and DS showed no significant benefit of MMC supplementation. MMC should be reserved for eyes, which have a high risk for subconjunctival fibrosis. In the endeavor to achieve IOPs in the low teens or even lower, patients are often exposed to sight-threatening complications.