Real-World Keratoplasty Outcomes
This report sent shockwaves -- well, maybe not shockwaves, but at least ripples -- throughout the cornea world. It pretty much has been taken as "fact" that Descemet-baring DALK visual acuity outcomes were generally the same as those for PK, but graft survival was at least as long (and most likely longer) for DALK than for PK. Visual acuity outcomes of non-Descemet-baring DALK are frequently reported to be worse than those of PK, but the same is not true for graft survival. The perceived "nonsuperiority" in visual acuity results and unproven greater longevity (not to mention increased surgical difficulty and time) of DALK are reflected in the relatively small number of DALKs performed compared with the number of PKs, for appropriate indications.
The EK results were the real shock. People (surgeons included) vote with their feet. If EK truly has worse visual acuity outcomes and greater graft failure rates, why are EKs performed for most grafts for Fuchs dystrophy and PBK? Plus, this trend of more EKs and fewer PKs performed for endothelial dysfunction is only gaining strength.
So, why the great divide between the results of this study and the impressions of most surgeons (in the United States at least, but I believe also around the world), not to mention numerous published case series demonstrating the superiority of EK over PK for endothelial disease?
Coster and colleagues have some ideas. One is that most published studies on EK are from a single surgeon or a small number of experienced surgeons. This is very different from the real world, which is reflected in registries such as the Australian Graft Registry. They also wonder whether "Australian surgeons are inferior to those working elsewhere," but they "consider this to be unlikely."
I do not have a great answer to this question, but other factors can be considered. First, the Australian Graft Registry is voluntary. They didn't estimate the percentage of grafts reported to the registry compared with the total number performed in the country. So, although the numbers in the registry are certainly large, they may not include all grafts or all surgeons. Moreover, visual acuity data were collected at the "last follow-up." Most of us believe that visual acuity recovery after PK can take years. The fact that this study had much longer follow-up for PKs than EKs (EKs were only performed in the last half of the study) will certainly influence the visual acuity data. Longer follow-up data for PKs also increase the time for rejection to occur.
Still, it does make us pause and remember that new is not always better. Registries can be very powerful tools. The American Academy of Ophthalmology has spent a lot of money setting up the IRIS registry, which extracts data directly from electronic health record systems, removing the huge variable of inaccurate and intermittent reporting from the equation. Hopefully future studies will do an even better job of capturing results that allow us to better evaluate long-term graft outcomes in the real world.
Abstract
Viewpoint
This report sent shockwaves -- well, maybe not shockwaves, but at least ripples -- throughout the cornea world. It pretty much has been taken as "fact" that Descemet-baring DALK visual acuity outcomes were generally the same as those for PK, but graft survival was at least as long (and most likely longer) for DALK than for PK. Visual acuity outcomes of non-Descemet-baring DALK are frequently reported to be worse than those of PK, but the same is not true for graft survival. The perceived "nonsuperiority" in visual acuity results and unproven greater longevity (not to mention increased surgical difficulty and time) of DALK are reflected in the relatively small number of DALKs performed compared with the number of PKs, for appropriate indications.
The EK results were the real shock. People (surgeons included) vote with their feet. If EK truly has worse visual acuity outcomes and greater graft failure rates, why are EKs performed for most grafts for Fuchs dystrophy and PBK? Plus, this trend of more EKs and fewer PKs performed for endothelial dysfunction is only gaining strength.
So, why the great divide between the results of this study and the impressions of most surgeons (in the United States at least, but I believe also around the world), not to mention numerous published case series demonstrating the superiority of EK over PK for endothelial disease?
Coster and colleagues have some ideas. One is that most published studies on EK are from a single surgeon or a small number of experienced surgeons. This is very different from the real world, which is reflected in registries such as the Australian Graft Registry. They also wonder whether "Australian surgeons are inferior to those working elsewhere," but they "consider this to be unlikely."
I do not have a great answer to this question, but other factors can be considered. First, the Australian Graft Registry is voluntary. They didn't estimate the percentage of grafts reported to the registry compared with the total number performed in the country. So, although the numbers in the registry are certainly large, they may not include all grafts or all surgeons. Moreover, visual acuity data were collected at the "last follow-up." Most of us believe that visual acuity recovery after PK can take years. The fact that this study had much longer follow-up for PKs than EKs (EKs were only performed in the last half of the study) will certainly influence the visual acuity data. Longer follow-up data for PKs also increase the time for rejection to occur.
Still, it does make us pause and remember that new is not always better. Registries can be very powerful tools. The American Academy of Ophthalmology has spent a lot of money setting up the IRIS registry, which extracts data directly from electronic health record systems, removing the huge variable of inaccurate and intermittent reporting from the equation. Hopefully future studies will do an even better job of capturing results that allow us to better evaluate long-term graft outcomes in the real world.
Abstract
Source...