We thank Dr Mifflin and associates for their letter, and we commend them on their excellent results in their series of Descemet stripping automated endothelial keratoplasty (DSAEK) under failed penetrating keratoplasty (PK), using a technique that retains recipient bed Descemet membrane and that uses a same size donor. It is obvious that there is more than one way to have successful attachment of donor tissue for this unique cohort of patients.
We wish to point out, however, that the 1 case of dislocation in our series was the result of postoperative hypotony in a severely complex eye, not a routine DSAEK under PK. It was the hypotony that caused the detachment, independent of whether the eye had a prior PK. Removing the factor of hypotony from the series yields a 0% detachment rate, which is what we have achieved now in our current consecutive series of 24 cases of DSAEK under PK. Did Mifflin and associates’ series have a similar situation for their dislocated case, or was it a routine endothelial keratoplasty under PK case?
The point of this is to say that we do not believe that evaluation of the posterior curvature, Descemet membrane stripping, and careful peripheral scraping is the ONLY way to make the tissue attach, but we do believe that these steps decrease the rate of detachment to less than 3%, not only in this setting, but also for our published series of hundreds of routine cases.
We also have observed that thinner donor tissue has a greater propensity to conform to the variable curvature of the recipient bed. It may be that Mifflin and associates were using tissue thinner than prior reports of endothelial keratoplasty under PK and, because of this, had better conformation of their tissue to the posterior protuberances of the PK edge recipient beds, enhancing their adherence rate in this unique setting. In our series, by fitting the tissue between protuberances, we could use donor tissues of any thickness, without worrying about conforming to the recipient bed edges. We recommend that Mifflin and associates perform donor tissue thickness analysis when they submit their longer manuscript for peer review of their study.
Finally, although adherence of the donor tissue is enhanced by the removal of interface fluid, we caution Mifflin and associates about the routine use of venting incisions to accomplish this. There is now a plethora of publications demonstrating the short-term and the long-term liabilities, such as infections, melting, and epithelial downgrowth, that can occur because of these incisions. We have used surface sweeping without venting incisions to evacuate interface fluid for essentially all of our 1300 cases over the past 11 years and have avoided these liabilities, yet retained the lowest dislocation rates in the world.
Once again, we thank Mifflin and associates for their interest in our article, and we look forward to reading their expanded manuscript in the future.