Unlock the Surrogates

The importance of adequate monitoring and management of glaucoma needs no reinforcement. Glaucoma-related death of retinal ganglion cells (RGCs) cannot be reversed: early diagnosis is therefore critical, as without timely detection, therapeutic intervention may be too late to prevent permanent vision loss. Yet our options for screening and monitoring the progression of this disease, particularly in its early stages, are profoundly unsatisfactory – and glaucoma remains one of the leading causes of blindness worldwide. How might we resolve this situation? In my view, development of molecular biomarkers – quantifiable surrogates of disease progression and therapeutic response – could transform the management of this most problematic disease.

At present, glaucoma diagnosis and monitoring rely on measurements of IOP, visual field (VF) changes and optic nerve (ON) imaging. But although these approaches are routinely invoked as the basis for treatment decisions and surgical interventions, they are known to be suboptimal (1). In particular:

• Perimetric VF tests are subjective in that they depend on the patient responding to a projected light, and changes in VF testing can take a long time to manifest. They are also reflective of RGC death which is not reversible.
• IOP is not precisely correlated with disease diagnosis or severity, and tonometric IOP measurements can be affected by other factors, such as variations in corneal thickness.
• Assessments based on ON imaging techniques such as OCT require normative databases – which are not yet fully validated, and may introduce errors related to the subjective definition of the rim margin.

Clearly, we need to replace these measures with new surrogates that specifically reflect glaucomatous neurodegeneration. The ideal marker would be present in accessible biological tissues, and would also predict clinical outcomes and treatment effects. Could such molecules exist?

According to our recent findings, they just might. We recently reported (2) that growth differentiation factor 15 (GDF-15) could be a biomarker of RGC death and glaucoma severity. In brief, we tested the effect of axonal injury (rodent optic nerve crush (ONC) model) on a panel of 88 retinal cytokines / growth factor genes, and demonstrated that only one of these genes, gdf-15, had an expression pattern that specifically correlated with RGC death. We also showed GDF-15 increased in the aqueous humor (AH) following ONC, and that these GDF-15 elevations originated in the retinal nerve fiber layer, where RGCs reside. Importantly, gdf-15 expression was unrelated to age and was not upregulated in murine models of photoreceptor death or ocular inflammation; the elevated expression therefore appeared to be specific to axonal injury. We also found increased gdf-15 expression in a murine glaucoma model and elevated GDF-15 protein in aqueous humor samples from human patients with primary open angle glaucoma (POAG). Finally, we demonstrated that higher GDF-15 levels were correlated with increased disease severity, and predicted worse VF test results, in human POAG patients.

Collectively, these preliminary studies suggest that AH levels of GDF-15 could indicate glaucomatous neurodegeneration. Although further studies are needed to investigate the potential of GDF-15 as a biomarker of disease and predictor of therapeutic response, GDF-15 may be one of the strongest candidates yet identified. But there may be more waiting to be found, and identifying quantifiable biomarkers such as this is essential if we are to reliably monitor disease and rationally manage patients, as well as significantly enhance our ability to influence retinal neurodegeneration.