Editors Selection IGR 22-3

Surgical Treatment: Surgery or Eyedrops first?

Gustavo de Moraes

Comment by Gustavo de Moraes on:

96100 Primary trabeculectomy versus primary glaucoma eye drops for newly diagnosed advanced glaucoma: TAGS RCT, King AJ; Fernie G; Hudson J et al., Health Technol Assess, 2021; 25: 1-158

See also comment(s) by Ricardo Paletta Guedes

Find related abstracts

In this pragmatic randomized clinical trial, King and colleagues investigated the clinical outcomes of patients with advanced glaucoma who were randomized to receive medical intervention versus incisional glaucoma surgery (trabeculectomy) over a period of 24 months. Their primary outcome measure was vision-related quality-of-life as assessed with the NEI-VFQ instrument, whereas the ancillary measures were other quality-of-life instruments, intraocular pressure (IOP), visual fields, and metrics assessing the cost-benefit of each intervention. They found that, in a sample of 453 participants who successfully met criteria for analysis, there was no significant difference in the primary outcome measure, although a significant difference in IOP reduction was noted, where a 42% reduction was seen in the medical treatment and 55% in the trabeculectomy group when comparing baseline versus 24-month IOP measurements. Moreover, safety assessment did not reveal any significant differences between groups during the study and at 24 months of follow-up.

Importantly, the authors observed that the small increase in quality-of-life generated by trabeculectomy was not compensated by its expected additional upfront cost and hence trabeculectomy was deemed unlikely to be cost-effective within the 24-month period of the study. In the cost-benefit analysis, the authors observed that trabeculectomy was associated with an additional cost of £2,687, an additional 0.28 QALYs, and an incremental cost per QALY of £9,670 compared with medical therapy. However, the economic model ‐ which is based upon 24-month data but extrapolates over a lifetime horizon ‐ suggests that trabeculectomy would be more likely to be cost-effective compared with medication in the long run. This difference between 24-month data and lifetime estimates is likely because the cost for trabeculectomy is higher in the beginning but then decreases over time as patients receive fewer subsequent procedures and require less medication. Moreover, when looking at visual field progression, there was no significant difference between groups at 12 or 24 months despite the significant differences in IOP favoring the trabeculectomy group.

The results up to 24 months suggest no clear advantages of one choice versus the other in terms of quality-of-life and cost-benefit ratio decline in the trabeculectomy group during the study although, interestingly, patients on medical therapy were more likely to subjectively refer fear of losing vision. It was also noted that in the subscale analysis patients in the trabeculectomy group were more likely to experience worse "role difficulties" and worse "general vision" as measured with quality-of-life instruments at 4 months compared to medical therapy. This finding, notwithstanding, could have been due to transient changes in vision which are inherent to the first few months after glaucoma surgical interventions. As expected, these differences disappeared at 12 and 24 months suggesting satisfactory recovery for most patients.

Regarding the primary outcome measure of quality of life, one should be reminded that these instruments are based upon binocular vision and are largely influenced by the vision in the better eye. It is therefore plausible that despite having a study adequately powered to detect small differences between groups with the NEI-VFQ, the effects of vision in the fellow eye may have dampened the ability to see significant differences between groups during the study. Therefore, a better estimate of the effects of treatment on visual function may be more interpretable with increased follow-up time by looking at other functional measures such as best corrected visual acuity and visual field indices. It is also worth noting that many of the patients initially randomized to medical therapy ended up receiving trabeculectomy sometime during follow-up (17.5% of patients). Among patients initially randomized medical therapy and who later underwent trabeculectomy, there was no significant difference in the rate of trabeculectomy-related procedures suggesting that there was no detrimental effect on complication rates by postponing surgery in this group of advanced glaucoma patients. Similarly, many of the patients randomized to trabeculectomy also required IOP lowering eyedrops to keep IOP at target (at 4 months, 28% of eyes were using any glaucoma eyedrop).

The results up to 24 months suggest no clear advantages of one choice versus the other in terms of quality-of-life and cost-benefit ratio

In summary, this clinical trial helped better understand the clinical and economic implications of choosing between medical therapy versus trabeculectomy among patients with advanced glaucoma. The results up to 24 months suggest no clear advantages of one choice versus the other in terms of quality-of-life and cost-benefit ratio, although trabeculectomy was more effective in lowering the IOP to target levels. Long-term follow-up looking at IOP and visual field outcomes are likely to provide the evidence needed for better treatment planning on a topic for which little evidence is a currently available to support decision making.

Issue 22-3

Change Issue


WGA Rescources