advertisement
In 2004 the Ocular Hypertension Treatment Study (OHTS) reported that central corneal thickness (CCT) was inversely related to the IOP response after the initial one-eyed therapeutic trial and during the 12 to 60 months of follow-up data then available. They concluded that individuals with thicker corneas have a smaller measured IOP response to ocular hypotensive medications than those with normal or thin corneas. In the present study, Johnson et al. (18) performed a retrospective review of data from individuals evaluated as part of several drug trials at the University of Nebraska. They report similar results to that observed in OHTS, namely that individuals with thick corneas had less IOP-lowering than individuals with thin corneas. They conclude that their data suggest a reduced efficacy for some glaucoma medications in ocular hypertensive individuals with thick corneas.
Is this really efficacy, or simply artifact? Consider the following mind experiment ‐ both a rubber balloon and a basketball of the same diameter are pressurized to 30 PSI. Both the balloon and basketball would feel firm to the touch. If one then dropped the pressure in both spheres to 2 PSI, the basketball would still offer resistance to a pressing finger because of the thickness and stiffness of the basketball material, but the balloon could be easily indented by a finger. Might this be what is going on with both the OHTS and Johnson data? Since there is no accepted algorithm for 'correcting' IOP based on CCT, it is impossible to match starting IOPs and then evaluate IOP-lowering.
It is crucial that clinical trials confirm that the CCTs in their study population are evenly matched between study groupsThe relationship between CCT, IOP and drug response is complex and unlikely to be clarified by studies like this. This study reinforces the need to measure CCT in all IOP-related clinical trials. Because CCT is such an important confounder of tonometry measurements, it is crucial that clinical trials confirm that the CCTs in their study population are evenly matched between study groups. One could easily imagine a situation in which two therapeutic interventions are being compared in a prospective, randomized clinical trial and Group 1 has thinner corneas on average than Group 2. Even if the two therapeutic interventions are equivalent, the therapy used for Group 1 will appear to be more efficacious than for Group 2, leading to an erroneous conclusion based on tonometry artifact.