EyeWorld is the official news magazine of the American Society of Cataract & Refractive Surgery.
Issue link: https://digital.eyeworld.org/i/78719
Keep them separated February 2012 by Enette Ngoei EyeWorld Contributing Editor EW CORNEA 49 The eye on the left is one that underwent PK at the 7-year post-op mark; on the right, the same patient's DALK eye at just 18 months post-op, which appears almost identical Source: David D. Verdier, M.D. Experts discuss the effect of the graft-host interface on visual quality in DALK I n select patients, the quality of vision after deep anterior lamellar keratoplasty (DALK) was comparable to that after penetrating keratoplasty (PK) when stromal excision was extended to Descemet's membrane (DM) and inferior to vision after PK when lay- ers of stroma were left adherent to DM, a recent study found. The study, published in Cornea, was a retrospective, nonrandomized, comparative case series where 60 eyes of 60 patients with advanced keratoconus underwent DALK using the big bubble technique. Carried out by researchers from the Unità Operativa di Oculistica, Ospedale Maggiore, Bologna, Italy, the study included 28 patients who had com- plete stromal excision with DM ex- posure [DALK with DM baring (DM-DALK)], whereas in 32 cases, a layer of stroma was left adherent to the recipient DM [DALK without DM baring (pre-DM-DALK)]. This group of patients was compared with 22 eyes of keratoconus patients who underwent PK. The main outcome measures were uncorrected visual acuity, best spectacle-corrected visual acuity, low contrast visual acuity (LCVA), and Pelli-Robson contrast sensitivity. The researchers reported that uncor- rected visual acuity was equal in the three groups, while the median best spectacle-corrected visual acuity was 0.1 logarithm of the minimum angle of resolution (logMAR) in the PK group, compared with 0.06 logMAR in the DM-DALK group (P=0.66) and 0.12 logMAR in the pre-DM-DALK group (P=0.016). The pre-DM-DALK patients exhibited worse LCVA than PK (P=0.029) and DM-DALK patients (P=0.022), whereas PK and DM- DALK patients showed comparable LCVA (P=0.974). Pelli-Robson con- trast sensitivity was equivalent in the PK and DM-DALK groups (P=0.408) and greater in the DM- DALK group than in the pre-DM- DALK group (P=0.038), the study said. The study's results are consistent with the experience of Neda Shamie, M.D., associate pro- fessor of ophthalmology, Doheny Eye Institute, Keck School of Medicine, University of Southern California, Los Angeles. "The concern with any kind of lamellar corneal transplantation is the interface and how the interface can impact light passing through the cornea. When you don't have a stroma-stroma interface with the De- scemet's baring DALK using the big bubble technique for example, then it's essentially the same as perform- ing PK," she said. Historically, before there were techniques to get down to Descemet's membrane, a lamellar dissection that left behind a fair amount of stroma could actually re- sult in interface haze, which would reduce vision, according to Edward Holland, M.D., professor of oph- thalmology, University of Cincin- nati, and director, cornea service, Cincinnati Eye Institute. "Leaving posterior stroma would not allow the patient to get the same visual acuity we see in PK, so techniques such as the big bubble technique, where there's an injection of air to separate stroma from Descemet's, have allowed the visual acuities to be comparable," he said. The great benefit from this would be no endothelial rejection and much less endothelial cell loss compared with PK, Dr. Holland noted. Endothelial cell loss after re- jection is a huge issue because the Cornea Donor Study, which has been going on for 10 years, showed that about 70% of cells are lost in the first 5 years in successful PK, so these grafts won't last a lifetime in a young patient with keratoconus, he said. What if big bubble fails? Surgeons can't always get a perfect separation of stroma and De- scemet's, but if the big bubble tech- nique doesn't work, it's not a failure of the procedure, Dr. Holland said. "You need to adopt some tech- niques to get the dissection as deep as possible, and there are several studies that say somewhere around 50 microns is a magic number. If you get within 50 microns you can still get excellent visual acuities," he said. Similarly, Dr. Shamie said that even the lamellar technique or non-Descemet's baring or near Descemet's baring DALK could even- tually provide excellent vision. "It just requires time," she said. Both in the Descemet's stripping automated endothelial keratoplasty (DSAEK) and deep lamellar endothe- lial keratoplasty procedures, with time and remodeling of that inter- face, visual quality can continue to improve. At a recent meeting, Dr. Shamie and colleagues presented data that showed there was contin- ual improvement of vision in the same eye of DSAEK patients up to 3 years post-op. "There are some changes hap- pening, and I would bet that would be the same in near Descemet's deep anterior lamellar keratoplasty; if you look over time, maybe 2 or 3 years after surgery, it would probably be equal to PK or at least close to it," she said. As to how the remodeling oc- curs, Dr. Shamie said it is still un- known how and why. However, it probably does not occur in 100% of patients, she said. In any case, DALK is far superior to PK in patients with normal en- dothelium, Dr. Holland said. "The classic patient is a patient with a stromal scar or abnormal stroma as we see in keratoconus, and any technique we can use to save the patient's endothelium we should do. That's why DALK has been proven to be the procedure of choice for these patients," he said. EW Editors' note: Drs. Holland and Shamie have no financial interests related to this article. Contact information Holland: 859-331-9000, eholland@holvision.com Shamie: 310-601-3366, nshamie@yahoo.com