Impact of Timolol/Dorzolamide Therapy on Autoregulation in Glaucoma Patients

Patients will be recruited from the glaucoma clinic at the department of Ophthalmology in the Leuven University Hospital. General study setup: Two cohorts of subjects will be included in the study: primary open-angle glaucoma (POAG) and normal tension glaucoma (NTG) patients. At the screening visit, eligible patients will have their topical therapy replaced by timolol 0.5% bid. At week 4, dorzolamide 2% will be added to the existing treatment in a fixed combination therapy (timolol 0.5%+dorzolamide 2% bid). A final visit at week 8. At the screening visit (day 0), the patients will undergo a complete ophthalmic investigation with visual acuity, slit lamp biomicroscopy, tonometry, 90 D fundoscopy, automated perimetry, Heidelberg retinal tomography and color Doppler imaging. Blood pressure will also be measured. Topical monotherapy will be replaced by timolol 0.5% bid. At visit 1 and 2 (timolol - week 4, timolol+dorzolamide week 8, respectively), the patients will undergo ophthalmic investigation with visual acuity, slit lamp biomicroscopy, tonometry, 90 D fundoscopy, blood pressure measurement and color Doppler imaging. Only one eye will be selected for the study (eye with worst glaucomatous damage). Mann-Whitney test will be used in pairwise comparisons. Restricted cubic sp-lines will be used to verify if there is any evidence for nonlinearity in the relation between the RI and PI. Piecewise linear regression models will be used to determine the optimal cutpoint (i.e. the cutpoint yielding the highest likelihood) in all three visits (baseline, timolol and timolol/dorzolamide). Sensitivity analyses will be performed to verify if the result is not due to an (influential) subject with a high RI value. All data will be expressed in mean ± standard deviation. A two sided p-value \<0.05 is considered significant. Sample size calculations were made to address the primary outcome (i.e., a change in the RI cutpoints between baseline and last study visit) in the overall glaucoma population (POAG+NTG patients). Based on our previous results10, setting an α error to 5%, power at 80% and the allowable difference at 10% would require the recruitment of 40 patients. Further post hoc analysis will be made to identify differences between the two study cohorts.