We read with interest the recently published article “Graft Failure after Penetrating Keratoplasty in Eyes with Ahmed Valves,” by Hollander and associates. They concluded that corneal graft failure in eyes with Ahmed valves (AVs) was high. They also report that poor intraocular pressure control may be related to graft failure. They found the strongest risk factors for graft failure in this population was a history of previous failed penetrating keratoplasty (PK) and corneal stromal neovascularization, whereas the use of glaucoma medication and a lower intraocular pressure was associated with a significantly decreased risk of graft rejection. We have a few concerns.
It is stated that this was a homogenous group, yet the authors included eyes with primary grafts, multiple grafts, and multiple grafts with repeat AV. The latter 2 groups had more than one risk factor for graft failure; hence as a whole, the study population was not homogenous.
The multiple confounding factors regarding the actual cause of the graft failures in their patient population has affected their statistical analysis. They themselves mention that having not corrected for multiple comparisons, some significant relationships were spurious.
It would have been interesting to know the graft failure rate specifically in eyes that underwent a first AV and subsequent first PK, compared with eyes that underwent a first PK only. In the absence of such a control group, it is difficult to attribute a causative relationship with AV.
Multiple grafts (previously failed grafts), stromal new vessels, and raised IOP are well-reported risk factors for graft failure. The main conclusion of their study was that a history of previously failed PK and corneal stromal neovascularization strongly correlated with graft failure in their study population. From the data presented, the same conclusion would be derived even if the patients had not undergone an AV implant, indicating that the contribution of the AV to graft failure remains indeterminate.