Edwin S. Chen

Endothelial Cell Loss after Descemet's Stripping Endothelial Keratoplasty in a Large Prospective Series

PURPOSE To report the donor endothelial cell loss in the first year after Descemets stripping endothelial keratoplasty (DSEK) for the treatment of endothelial dysfunction. DESIGN Prospective noncomparative interventional case series. PARTICIPANTS Eighty eyes of 78 patients with corneal edema. METHODS Eighty eyes with endothelial failure were entered into a prospective study of endothelial keratoplasty (EK). The donor central endothelial cell density (ECD) was recorded postoperatively at 6 months (n = 80) and 12 months (n = 80) and then compared with the preoperative eye bank measurements. The subsets of eyes with the donor prepared manually (DSEK; n = 19) and the donor prepared with a microkeratome (Descemets stripping automated EK [DSAEK]; n = 61) were also evaluated and compared. MAIN OUTCOME MEASURES Preoperative and postoperative central ECDs were prospectively evaluated and the cell loss calculated for each postoperative time point. RESULTS The average and standard deviation ECD at 6 months was 1908+/-354 cells/mm(2), representing a mean cell loss from preoperative donor cell measurements of 34+/-12%. At 12 months, ECD was 1856+/-371 cells/mm(2) (35+/-13% cell loss). The 1% additional cell loss from 6 to 12 months was not significant (P = 0.233). In the subset of DSEK eyes (n = 19), the cell loss from preoperatively to 6 months was 34%, and at 12 months it was 39%. In the subset of DSAEK eyes (n = 61), the cell loss from preoperatively to 6 months was 34%, and at 12 months it was 34%. There was no statistical difference between the cell loss from DSEK and that from DSAEK at 6 months (P = 0.884) or at 12 months (P = 0.224). CONCLUSIONS Descemets stripping EK using our surgical technique has a mean donor endothelial cell loss of 34% at the 6-month postoperative examination, and this average cell loss remains relatively stable up to at least 1 year. We found no difference in cell loss between the DSEK and DSAEK techniques over this 1-year postoperative period.

Descemet-stripping automated endothelial keratoplasty: six-month results in a prospective study of 100 eyes.

Purpose: To report 6-month results in a large, prospective study of Descemet-stripping automated endothelial keratoplasty (DSAEK). Methods: A 5-mm scleral-limbal tunnel approach was created for placement of an automated microkeratome-prepared 8.0-mm endothelial graft after DSAEK in 150 consecutive cases between September 2005 and October 2006. Six-month follow-up data were available on 100 eyes. Intraoperative peripheral scraping was performed to promote adherence of the donor. Preoperative and postoperative visual acuity with and without spectacle correction (BSCVA and UCVA), refractive astigmatism, average topographic keratometry, surface asymmetry index, surface regularity index, and pachymetry were measured prospectively. Results: After DSAEK surgery, average BSCVA improved from 20/86 to 20/38, and average UCVA improved from 20/155 to 20/73, which were both statistically significant (P < 0.05). Excluding 26 eyes with known retinal pathology: 97% of the 74 eyes had a vision of 20/40 or better at 6 months and 14% obtained 20/20 or better. Refractive astigmatism changed an average 0.06 D, and average topographic keratometry changed an average −0.13 D, which were not statistically significant. Surface regularity index and surface asymmetry index improved to normal levels of 0.67 and 1.03, respectively (P < 0.001 and P = 0.002). Pachymetry decreased significantly from 0.70 to 0.66 mm (P = .001). Conclusions: This large prospective study of DSAEK shows that this surgery provides a significant improvement in vision, corneal thickness, and surface regularity. It does not change refractive astigmatism or average topographic keratometry significantly. This newer technique of endothelial keratoplasty yields many of the benefits of its predecessors, deep lamellar endothelial keratoplasty and posterior lamellar keratoplasty, while improving the visual results.

Precut Tissue for Descemet's Stripping Automated Endothelial Keratoplasty: Vision, Astigmatism, and Endothelial Survival

PURPOSE To report 6 and 12 month results using precut tissue for Descemets stripping automated endothelial keratoplasty (DSAEK) and correlate donor characteristics with clinical outcomes. DESIGN Prospective, noncomparative, interventional case series. PARTICIPANTS We reviewed 100 donor corneas precut for 100 eyes of 90 DSAEK patients. METHODS Our first 100 consecutive cases of DSAEK with precut tissue were entered into a prospective protocol. Donor characteristics and the visual, refractive, topographic, and specular microscopy results at 6 and 12 months were analyzed. Correlation analysis comparing donor characteristics with clinical outcomes was performed. MAIN OUTCOME MEASURES Six- and 12-month postoperative best spectacle-corrected visual acuity (BSCVA), refractive astigmatism, topographic keratometry (K), and specular endothelial cell densities (ECD) were measured prospectively and then compared with preoperative values. Donor characteristics analyzed included death to preservation time, death to surgery time, precutting resection to surgery time, and graft thickness. RESULTS Six months after DSAEK surgery, BSCVA improved from 20/83 to 20/38. (P<0.01). In eyes with no known comorbidity (n = 60), 92% had a vision of >/=20/40 at 6 months and 20% obtained > or =20/20. Astigmatism changed an average of 0.09 diopters (D) and K changed by +0.09 D, both of which were not significant and were stable to 12 months. The postoperative mean ECD (n = 65) was 1918 cells/mm(2) at 6 months, and represented a 31% cell loss from preoperatively (P<0.001). The mean ECD (n = 61) was 1990 cells/mm(2) at 12 months, and represented a 29% cell loss from preoperatively (P<0.001) with no significant change from 6 to 12 months (P = 0.172). Improvement of visual acuity from preoperative to postoperative in eyes without comorbidity was not correlated with any donor characteristic. Greater endothelial cell loss correlated with higher preoperative ECD levels (P<0.001) and with a trend toward longer precut resection to surgery times at both 6 months (P = 0.049) and 12 months (P = 0.051). CONCLUSIONS Precut tissue by Eye Banks for use in DSAEK surgery provides an improvement in vision with no significant change in astigmatism. Donor endothelial cell loss from 6 to 12 months is stable and is comparable with reports involving tissue that is cut intraoperatively. FINANCIAL DISCLOSURE(S) Proprietary or commercial disclosures may be found after the references.

Precut Tissue in Descemet's Stripping Automated Endothelial Keratoplasty : Donor Characteristics and Early Postoperative Complications

PURPOSE To describe donor characteristics of eye bank-prepared precut tissue used in Descemets stripping automated endothelial keratoplasty (DSAEK) and report any increase in immediate postoperative complications associated with its use. DESIGN Prospective, noncomparative, interventional case series. PARTICIPANTS One hundred donor corneas deemed appropriate for transplant and 100 eyes undergoing DSAEK for endothelial dysfunction. METHODS Precut donor tissue was evaluated in a prospective study of 100 consecutive cases of DSAEK surgery. Donor characteristics and the immediate postoperative complications of donor tissue dislocation and graft failure with the use of precut tissue were recorded and analyzed. MAIN OUTCOME MEASURES Preoperative donor characteristics (age, time from death to preservation, time from death to implantation, time from cut to implantation, residual stromal bed thickness, pre- and postcut endothelial density), rate of dislocation, and rate of primary graft failure. RESULTS Average donor age was 57.6+/-10.8 years, average time from death to preservation was 9.8+/-3.2 hours, average time from death to implantation was 94.5+/-33.5 hours, and average time from cut to implantation was 26.0+/-17.4 hours. The average residual stromal bed thickness was 169+/-36 microns. The average endothelial cell density (ECD) after cutting was 2709+/-292 cells/mm(2) (n = 100). In the subgroup of donors in whom pre-resection and postresection endothelial cell densities were available (n = 80), the average ECD before cutting was 2743+/-253 cells/mm(2) and the average ECD after cutting was 2644+/-257 cells/mm(2). This average cell loss of 3.7% was statistically significant (P<0.001). There was only 1 dislocation in this entire series of 100 eyes. There were no primary graft failures. CONCLUSION The use of precut tissue in DSAEK had a low rate of early postoperative complications such as graft dislocation (1%) and primary graft failure (0%). A wide range of donor characteristics such as donor age, death to transplantation time, precutting to transplantation time, and donor lenticule thickness resulted in excellent adhesion of the tissue and clear grafts.

Endothelial keratoplasty: the influence of insertion techniques and incision size on donor endothelial survival.

Purpose: To determine the acute endothelial cell damage from trephination and tissue insertion in endothelial keratoplasty (EK) surgery. The influence of insertion technique (forceps insertion vs “pull-through” insertion) of donor tissue and incision size (3 vs 5 mm length) was assessed. Methods: Forty precut 8.-mm-diameter donor posterior buttons were used in this study. Thirty-five buttons were inserted through a limbal incision of either 3 or 5 mm length into the anterior chamber of cadaver eyes and then removed through an open sky technique without further trauma. Five buttons that were trephined but not inserted served as a control group. Vital dye staining and computer digitized planimetry were used to analyze the tissue and quantify the total damaged area over the entire endothelial surface. Five buttons for each of 7 insertion techniques were used. The 8 tissue groups evaluated were as follows: group 1: control group of trephination only, with no insertion; group 2: forceps with folded tissue through 5-mm incision; group 3: suture pull through of nonfolded tissue through a 5-mm incision; group 4: forceps pull through of Busin glide folded tissue through a 5-mm incision; group 5: forceps with folded tissue through a 3-mm incision; group 6: suture pull through with folded tissue through a 3-mm incision; group 7: suture pull through with nonfolded tissue through a 3-mm incision; and group 8: forceps pull through of Busin glide folded tissue through a 3-mm incision. Results: The control group demonstrated 9% ± 2% peripheral cell damage from simple trephination of the tissue but without insertion. In the 5-mm incision surgeries, forceps insertion (group 2) caused 18% ± 3% loss, suture pull-through insertion (group 3) caused 18% ± 2% loss, and Busin glide pull through (group 4) caused 20% ± 5% loss. There were no significant differences in damage between any of the 5-mm incision group techniques (P > 0.99). In the 3-mm incision surgeries, forceps insertion (group 5) caused a 30% ± 3% loss, pull through with folded tissue (group 6) caused 30% ± 5% loss, pull through with nonfolded tissue (group 7) caused 56% ± 4% loss, and Busin glide pull through (group 8) caused a 28%± 5% loss. There was no difference in damage among the 3-mm groups (P > 0.96), with the exception of group 7 where pulling the unfolded tissue through a 3-mm incision was significantly worse than all other techniques (P < 0.001). There was significantly greater cell area damage in the 3-mm groups (36%) than in the 5-mm groups (19%) (P <0.001). Large patterns of striae with cell loss were seen in the 3-mm groups emanating from the peripheral traction site, regardless of whether the traction to pull the tissue through the incision and into the chamber was generated by a suture or cross-chamber forceps. Direct forceps insertion caused circular patterns of injury at the tip compression site regardless of incision size, but this damage was multiplied and exacerbated by insertion through a smaller incision. Conclusions: Smaller size (3 mm) incisions for EK surgery result in greater acute endothelial area damage than larger size (5 mm) incisions. Pull-through insertion techniques through a 5-mm incision seem equivalent in the amount of induced area damage to that of forceps insertion. Compressive injury from the incision appeared less when the tissue was folded than when not folded. Insertion with any technique through a 3-mm incision resulted in larger areas of endothelial damage. All these iatrogenic death zones outside the central endothelial area would be missed clinically by standard early specular microscopy after EK surgery.

An easy and inexpensive method for quantitative analysis of endothelial damage by using vital dye staining and Adobe Photoshop software.

Purpose: We developed a simple, practical, and inexpensive technique to analyze areas of endothelial cell loss and/or damage over the entire corneal area after vital dye staining by using a readily available, off-the-shelf, consumer software program, Adobe Photoshop. The purpose of this article is to convey a method of quantifying areas of cell loss and/or damage. Methods: Descemet-stripping automated endothelial keratoplasty corneal transplant surgery was performed by using 5 precut corneas on a human cadaver eye. Corneas were removed and stained with trypan blue and alizarin red S and subsequently photographed. Quantitative assessment of endothelial damage was performed by using Adobe Photoshop 7.0 software. Results: The average difference for cell area damage for analyses performed by 1 observer twice was 1.41%. For analyses performed by 2 observers, the average difference was 1.71%. Three masked observers were 100% successful in matching the randomized stained corneas to their randomized processed Adobe images. Conclusions: Vital dye staining of corneal endothelial cells can be combined with Adobe Photoshop software to yield a quantitative assessment of areas of acute endothelial cell loss and/or damage. This described technique holds promise for a more consistent and accurate method to evaluate the surgical trauma to the endothelial cell layer in laboratory models. This method of quantitative analysis can probably be generalized to any area of research that involves areas that are differentiated by color or contrast.

Descemet stripping automated endothelial keratoplasty in eyes with previous trabeculectomy and tube shunt procedures: intraoperative and early postoperative complications.

Purpose: To evaluate the intraoperative and early postoperative outcomes of Descemet stripping automated endothelial keratoplasty (DSAEK) in patients with previous glaucoma filtering surgeries. Methods: A retrospective review of all DSAEK surgeries performed at one center comparing complications of DSAEK in eyes with previous glaucoma filtering procedures (study eyes) with a time-matched group of all other DSAEK cases (control eyes). Results: There were 28 study eyes, 19 with previous trabeculectomies and 9 with previous glaucoma drainage devices (GDDs) and 431 control eyes. Study group intraoperative complications included 1 compromised bleb and 1 loss of donor tissue because of traumatic manipulation. One intraoperative complication, a perforation of the donor tissue, occurred in the control group. Venting stab incisions were used more often in study eyes (n = 5; 18%) than in control eyes (n = 12; 4.4%) (P = 0.002). GDD tubes were trimmed in 2 eyes (22%). No intraoperative manipulations were used to occlude the glaucoma filters or tubes. Postoperative complications in the study group included 1 dislocation (3.6%) and 1 decentered graft (3.6%) and 1 eye with loss of pressure control (3.6%), whereas in the control group, there were 10 dislocations (2.3%) and 1 decentered graft (0.2%) (P = 0.267 for dislocations and P = 0.118 for decentered grafts). One episode of pupillary block (0.2%) occurred in the control group, and none occurred in the study group. No primary graft failures occurred in either group. Conclusions: DSAEK surgeries in eyes with previous glaucoma filtering procedures were performed without primary graft failure and with reasonably low dislocation (3.6%) and graft decentration (3.6%) rates. Although the intraoperative complication rate for the study group (7.1%) was higher than the rate for the control group (0.23%), excellent early postoperative outcomes can be achieved when DSAEK is performed in eyes with previous trabeculectomies and GDDs.

Endothelial keratoplasty: improvement of vision after healthy donor tissue exchange.

Objective: To report 3 cases of graft exchange by using a microkeratome-prepared donor tissue in place of a manually prepared donor tissue for inadequate postoperative visual acuity after deep lamellar endothelial keratoplasty and to discuss possible etiologies. Methods: Prospective, observational case series. The patients were 3 consecutive patients who underwent endothelial graft replacement for unsatisfactory vision after initial deep lamellar endothelial keratoplasty. This is a review of clinical findings in 3 cases of endothelial keratoplasty that underwent graft exchange for unacceptable vision after deep lamellar endothelial keratoplasty. Results: Two patients benefited from graft exchange by using a microkeratome-prepared donor in place of a manually prepared donor with improvement in best spectacle-corrected visual acuity and 1 did not because of recipient bed irregularities. Vision improved in this patient with penetrating keratoplasty. Conclusions: Endothelial keratoplasty results in rapid visual recovery and excellent vision. However, fewer eyes achieve 20/20 vision than with full-thickness penetrating keratoplasty. This report shows that some patients with suboptimal vision after endothelial keratoplasty felt to be caused by interface optical problems may benefit from either graft exchange or penetrating keratoplasty.

Complications and Clinical Outcomes of Descemet Stripping Automated Endothelial Keratoplasty With Intraocular Lens Exchange

PURPOSE To evaluate complications and clinical outcomes of Descemet stripping automated endothelial keratoplasty (DSAEK) with intraocular lens (IOL) exchange compared with DSAEK alone. DESIGN Retrospective, interventional case series. METHODS DSAEK was performed in 19 eyes in which the anterior chamber IOL was exchanged for a posterior chamber IOL (study group) and in 188 eyes in which the posterior chamber IOL was left in place (comparison group). The complications of graft dislocations, primary graft failure episodes, and pupillary block were recorded for all eyes. Six-month best spectacle-corrected visual acuity and mean central endothelial cell density were measured prospectively and then compared with preoperative values for all eyes. RESULTS Dislocations occurred in 0 (0%) of 19 eyes in the study group and in 5 (3%) of 188 eyes in the comparison group (P = .47), with 0 primary graft failures and 0 pupillary block episodes in either group. Preoperative mean best spectacle-corrected visual acuity for those eyes without any underlying ocular comorbidities was 20/205 and 20/100 in the study and comparison groups, respectively (P = .18). Mean best spectacle-corrected visual acuity at 6 months improved to 20/48 in the study group and to 20/34 in the comparison group, a statistically significant difference (P = .01). Mean donor cell loss at 6 months was 33% in the study group and 26% in the comparison group (P = .18). CONCLUSIONS Concurrent IOL exchange with DSAEK surgery does not increase the dislocation, primary graft failure, or pupillary block rates in the immediate postoperative period. Donor endothelial cell loss in DSAEK was not increased significantly by IOL exchange. Visual acuity was slightly worse after combined surgery than after DSAEK alone.

Descemet-stripping endothelial keratoplasty : improvement in vision following replacement of a healthy endothelial graft

We describe a case of endothelial graft exchange in a patient who had inadequate vision without a clinically identifiable cause following Descemet-stripping endothelial keratoplasty (DSEK). The initial DSEK was performed with a hand-dissected donor, and the replacement graft was prepared with an automated microkeratome. The best spectacle-corrected visual acuity (BSCVA) improved from 20/50 before replacement to 20/25 after replacement. Although DSEK provides rapid recovery of excellent vision, few patients achieve a BSCVA of 20/20. Some have substandard vision with no clinically identifiable cause. This case demonstrates that replacing a hand-prepared graft with an automated microkeratome-prepared graft may improve vision in patients with no identifiable cause for visual loss.