advertisement
The investigators showed energy and foresight in attempting a controlled neuroprotection trial. Brimonidine has been studied for neuroprotective effects in animals, but human data were lacking. This report leaves some important questions unanswered, due to issues with the design and analysis.
The brimonidine-treated group had about the same worsening rate as EMGT study patients with similar levels of IOP who were untreated. Worsening in the timolol-treated group was dramatically higher than any past estimate for glaucoma patients, treated or untreated. Extrapolating the Kaplan-Meier survival graph to five years, 100% of them would have progressed. Without this anomalously high rate, the study could not have found a difference between groups.
Without this anomalously high rate, the study could not have found a difference between groups
The suggestion that timolol might be detrimental is unlikely given the findings in OHTS and EMGT patients who benefited from beta blocker treatment. It would be good to see the details of the visual field results from patients in this study reviewed in detail by a variety of analytic techniques rather than just the Progressor 3-point criterion. This could test the hypothesis that there was a high false positive rate of progression.
The baseline, untreated IOP in patients was just over 15 and the mean IOP under treatment during the study was around 14 ‐ i.e., there was minimal IOP lowering in either group.
By the start of this study (1998), the Collaborative Normal Tension Glaucoma Study (CNTGS) results had shown significant benefit from 30% IOP lowering in such patients (later corroborated by the EMGT study). A 30% lowering (as in CNTGS) would equal a treated IOP of 11; even a less stringent criterion of 20% lower would equal a mean of 12.5. The standard of practice dictated meaningful IOP-lowering in all participants, not a design in which any pressure under 21 was allowed. Glaucoma neuroprotection trials should include IOP-lowering to a consensus target in all subjects with an additional, randomized test treatment to one group and placebo to the other.
Why didn't timolol and brimonidine lower IOP in these patients? It does so in many 'low-tension' patients and in normals. If non-adherence was a major reason, it further weakens the support for a neuroprotective effect.