AIM: To present a novel interpretation of the biexponential nature of chronic endothelial cell loss after penetrating keratoplasty (PK). We hypothesize that the fast component of endothelial cell loss reflects the endothelial cells of graft origin. The slow component might just reflect cell loss of the recipient endothelium. We investigate herein whether this hypothesis is in line with long-term survival in bullous keratopathy (BK: almost no endothelium in the recipient bed) and keratoconus (KK: recipient bed with plenty of endothelium). METHODS: We reviewed endothelial graft failures in PK for BK (n = 88) and KK (n = 87). Patients with immune reactions or a history of glaucoma were excluded. We built a statistical model to predict graft failures from biexponential endothelial cell loss and compared this data to the actual outcomes. RESULTS: After 15 years, the incidence of late endothelial failures was 8% in KK and 33% in BK. The 95% confidence intervals of the simulated outcomes corresponded completely to the actual outcomes during follow-up. CONCLUSIONS: Our novel interpretation of the biexponential model is in line with long-term data of PK for BK and KK. Our findings highlight the importance of the recipient bed endothelial reservoir on the long-term prognosis in PK.
AIM: To present a novel interpretation of the biexponential nature of chronic endothelial cell loss after penetrating keratoplasty (PK). We hypothesize that the fast component of endothelial cell loss reflects the endothelial cells of graft origin. The slow component might just reflect cell loss of the recipient endothelium. We investigate herein whether this hypothesis is in line with long-term survival in bullous keratopathy (BK: almost no endothelium in the recipient bed) and keratoconus (KK: recipient bed with plenty of endothelium). METHODS: We reviewed endothelial graft failures in PK for BK (n = 88) and KK (n = 87). Patients with immune reactions or a history of glaucoma were excluded. We built a statistical model to predict graft failures from biexponential endothelial cell loss and compared this data to the actual outcomes. RESULTS: After 15 years, the incidence of late endothelial failures was 8% in KK and 33% in BK. The 95% confidence intervals of the simulated outcomes corresponded completely to the actual outcomes during follow-up. CONCLUSIONS: Our novel interpretation of the biexponential model is in line with long-term data of PK for BK and KK. Our findings highlight the importance of the recipient bed endothelial reservoir on the long-term prognosis in PK.
Authors: Jonathan H Lass; Beth Ann Benetz; Sanjay V Patel; Loretta B Szczotka-Flynn; Robert O'Brien; Allison R Ayala; Maureen G Maguire; Yassine J Daoud; Mark A Greiner; Sadeer B Hannush; W Barry Lee; Thomas F Mauger; Harry J Menegay; Mark D Mifflin; Michael B Raizman; Jennifer Rose-Nussbaumer; Robert L Schultze; Gregory A Schmidt; Alan Sugar; Mark A Terry; David D Verdier Journal: JAMA Ophthalmol Date: 2019-02-01 Impact factor: 7.389