Purchase this article with an account.
Daniel Meira-Freitas, Renato Lisboa, Andrew Tatham, Linda Zangwill, Robert Weinreb, Christopher Girkin, Jeffrey Liebmann, Tammy Kuang, Christopher Bowd, Felipe Medeiros; The Combined Structure and Function Index as a Predictor of Glaucoma Development. Invest. Ophthalmol. Vis. Sci. 2013;54(15):1887.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess whether baseline combined structure and function index (CSFI) results are predictive of conversion to glaucoma in patients suspected of having the disease.
The study included 345 glaucoma suspect eyes at baseline followed for an average of 74.8 months in the Diagnostic Innovations in Glaucoma Study (DIGS) and the African Descent and Glaucoma Evaluation Study (ADAGES). All eyes had normal standard automated perimetry (SAP) at baseline. Retinal nerve fiber layer (RNFL) thickness assessment was performed with time domain optical coherence tomography (OCT). Conversion was defined as development of repeatable abnormal SAP or progressive optic disc changes during follow-up. Estimates of the number of retinal ganglion cells were obtained by combining data from SAP and OCT according to a previously described method. The CSFI was calculated as the age-corrected percent loss of retinal ganglion cells. The association between baseline predictive factors and conversion to glaucoma was investigated with Cox proportional hazards models. Multivariable models were adjusted for age, intraocular pressure and central corneal thickness. The c-index and the modified R2 index were used to evaluate and compare predictive abilities of different models.
Fifty-eight (16.8%) eyes converted to glaucoma. Mean baseline CSFI was 14.1% and 0.2% in the converters and nonconverters, respectively. Table 1 shows hazard ratios with 95% confidence interval (CI) for each putative predictive factor for development of conversion. In the multivariable model, higher CSFI was predictive of conversion (adjusted hazard ratio=2.08 per 10% higher; 95%CI: 1.70-2.56; P<0.001). Model including CSFI had superior predictive ability (c-index=0.770; R2=55%) compared to models including standard visual field parameters and OCT average RNFL thickness (c-index from 0.648-0.718; R2 from 15-39% - table 2).
Baseline CSFI values were predictive of conversion to glaucoma and performed significantly better than conventional approaches for risk stratification of glaucoma suspects.
This PDF is available to Subscribers Only