Effects of glaucoma drugs on ocular hemodynamics in normal tension glaucoma: a randomized trial Title: comparing bimatoprost and latanoprost with dorzolamide Authors:

According to your suggestion, the entire abstract was re-structured. We would kindly like to ask to refer to page 2 in the manuscript for the modified abstract. Background (page 4, first paragraph): “ [...] The effects were compared with the hemodynamic properties of dorzolamide, which has repeatedly been shown to improve ocular blood flow and can therefore be used as a reference compound for the evaluation of hemodynamic effects of antiglaucomatous eye drops [...]” Nevertheless, the dorzolamide group has from our point of view the character of a control group, because we wanted to show with these patients, that we are able to detect changes in ocular hemodynamics by CDI and that we can reproduce previously published findings. The essence in our present paper is the results with latanoprost and particularly bimatoprost, as well as the comparison of both. Thus the introduction and discussion remains focussed on these both compounds. Discussion: You wrote: "While both substances do not improve ocular blood flow to a significant degree, they are both suitable for treatment of normal tension glaucoma". In your results you showed that dorzolamide had a better influence on ocular perfusion. Please discuss the usefulness for NTG treatment in a more neutral way. With a distance of the couple of weeks while the paper was reviewed by BMC-ophthalmology we totally agree with you, that the conclusion needs to be changed. Please find the modified conclusion on page 9. Minor Essential Revisions (such as missing labels on figures, or the wrong use of a term, which the author can be trusted to correct) You wrote: "An altered blood flow velocity due to vasoconstricting activity of bimatoprost reported from Allemann and colleagues is not detectable in humans by CDI measurements". We agree that in-vitro studies can not automatically be extrapolated to humans, even less to glaucoma patients.Nevertheless the fact that both bimatoprost and latanoprost did not significantly improve ocular circulation in these normal tension glaucoma patients, despite a good IOP-lowering effect would rather point in the direction that the drugs themselves are not neutral. If they were neutral, we would rather expect a significant improvement of circulation. As mentioned above, IOP reduction in a range observed in the present study causes a change of ocular perfusion pressure of 5%, resulting in a change of retrobulbar blood flow velocities by also 5%. This effect is too low to be detected by the present study design and probably also by the CDI method. The tendency of an increase of blood flow velocities can be observed in all treated study groups. This topic is addressed in the inserted paragraph on page 7 (bottom) and 8 (top): “In the setting of the present study the absolute IOP reduction in relation to ocular perfusion pressure is low and does affect ocular perfusion pressure only by approximately 5%. Retrobulbar blood flow velocities are directly proportional to ocular perfusion pressure. Subsequently, the effect of such a change in intraocular pressure on retrobulbar blood flow velocities is low, particularly compared to the change seen after administration of dorzolamide. Nevertheless, the tendency of an increase of retrobulbar blood flow velocity is visible in all treated groups.”