Risk Factors for Donor Endothelial Loss in Eye Bank–Prepared Posterior Lamellar Corneal Tissue for Descemet Stripping Automated Endothelial Keratoplasty
Liu, Yu-Chi MD*,†; Alvarez Paraz, Carisa M. MD†; Cajucom-Uy, Howard Yu MD*,†; Agahari, Djoni BDS†; Sethuraman, Selvam BSc†; Tan, Donald T.-H. FRCS*,†,‡,§; Mehta, Jodhbir S. FRCS*,†,¶
*Singapore Eye Research Institute, Singapore, Singapore;
†Singapore National Eye Centre, Singapore, Singapore;
‡Department of Ophthalmology, Yong Yoo Lin School of Medicine, National University of Singapore; Singapore, Singapore;
§Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore, Singapore; and
¶Department of Clinical Sciences, Duke-NUS Graduate Medical School, Singapore, Singapore.
Reprints: Jodhbir S. Mehta, FRCS, Singapore National Eye Centre, 11 Third Hospital Avenue, Singapore 168751, Singapore (e-mail: email@example.com).
The authors have no funding or conflicts of interest to disclose.
Received March 10, 2014
Revision received April 08, 2014
Accepted April 08, 2014
The aim of this study was to investigate donor, tissue, and precut procedure risk factors for endothelial cell density (ECD) loss in posterior lamellar corneal tissue preparation by an eye bank for Descemet stripping automated endothelial keratoplasty.
A total of 259 corneoscleral rims precut by the Singapore Eye Bank from October 2011 to August 2013 were evaluated. Donor characteristics, tissue characteristics, and precut procedure parameters were analyzed.
The mean donor age was 57.18 ± 11.35 years, and the mean cutting transition time was 4.16 ± 0.75 seconds. The mean ECD was 2826 ± 225 and 2787 ± 224 cells per square millimeter before and after precutting, respectively, with an average ECD change of −1.38% ± 3.28%. The precutting procedure failure rate was 1.2%. Mutivariate regression analysis showed that an older donor age, a higher ECD before cutting, and a slower cutting transition speed were significant factors. Corneas with an ECD >2800 cells per square millimeter before precutting, cutting transition time >5.5 seconds, and corneas with donor age >65 years were significantly more likely to have greater than 5% ECD loss after precutting (odds ratio, 6.42, 1.66, and 1.62; 95% confidence interval, 1.44–29.43, 1.45–2.72, and 1.66–5.82, respectively). Donor source, death-to-preservation time (range, 0.67–10.88 hours), death-to-precutting time (range, 0–7 days), and graft thickness (range, 43–232 μm) were not statistically significant factors.
The ECD loss in the precut tissue prepared by the eye bank was very low. The risk factors identified provide better understanding of how to improve the quality and safety profiles when preparing graft tissue for Descemet stripping automated endothelial keratoplasty.
Over the past decade, endothelial keratoplasty (EK) has rapidly increased in popularity as an alternative to conventional penetrating keratoplasty (PK) for the treatment of endothelial diseases. In the United States, 25,025 EK procedures were performed in 2012.1 Among various EK procedures, Descemet stripping automated endothelial keratoplasty (DSAEK) is currently the most popular. It offers many advantages over PK, including closed-globe surgery, lack of graft sutures, faster visual recovery, minimal refractive change, minimal postoperative ocular surface complications, and better retention of corneal strength and integrity.2–5
In DSAEK, posterior lamellar dissection of the donor tissue can be prepared by surgeons immediately before surgery or by trained technicians in eye banks; the latter being termed precut tissue. A main advantage of using precut tissue is that the posterior lamellar corneal tissue can be examined and measured after precutting, providing additional quality assurance to surgeons, that is, slit-lamp biomicroscopy and specular microscopy examinations. It also eliminates the possibility of last-minute cancellation of the surgery resulting from unexpected damage to the tissue due to unsuccessful cutting in the operating theater.6 The use of precut tissue also saves time, workload, and cost for surgeons, especially for those who do not have the equipment to prepare the tissue in an operating theater.6 Furthermore, studies have shown that the use of precut tissue for DSAEK was not associated with increased risk of postoperative complications such as graft dislocation and primary failure,7,8 and eye bank precut tissue provided similar endothelial loss, visual and refractive outcomes, and detachment rates compared with surgeon-dissected tissue.9 Therefore, recently there is an increasing trend toward skilled eye bank personnel performing the posterior lamellar dissection before surgery, and the demands for precut lamellar grafts for DSAEK in eye banks have risen.10 In the United States, more than two thirds of grafts for DSAEK are now precut by eye banks.11 However, in centers outside the United States, for example, in Europe and Asia, precut tissue is not universally available,12 and DSAEK donor tissues are still often cut by surgeons in the operating theater.11 The availability of the precut tissue would make DSAEK accessible to a larger population of corneal surgeons.8,13
The influence of microkeratome-assisted precutting procedures on endothelial cell loss has been reported to be 3.0% to 3.4% evaluated by trypan blue and alizarin red staining,5,14 and 1.0% to 3.7% evaluated by specular microscopy.7,15,16 However, the risk factors associated with donor endothelial loss during microkeratome-assisted precutting have not been well described. Yamazoe et al16 reported that a higher endothelial cell density (ECD) before precutting was significantly associated with higher cell loss with precutting. Nevertheless, other factors regarding donor characteristics, tissue characteristics, and precut procedure parameters have not been extensively analyzed. In this study, we aimed to investigate the association of donor characteristics, tissue characteristics, and precut procedure parameters with donor endothelial loss after precutting.
MATERIALS AND METHODS
A total of 259 corneoscleral rims from 241 donors precut consecutively by the Singapore Eye Bank from October 2011 to August 2013 were evaluated. All the corneoscleral buttons appeared optically clear and were stored in Optisol-GS (Bausch & Lomb, Irvine, CA) at 4°C. Donor characteristics, including age, gender, source, cause of death, and death-to-preservation time, were obtained from the Singapore Eye Bank database17 and overseas eye banks.
Preparation of Posterior Lamellar Corneal Tissue
The central ECD was measured immediately before precutting using a fixed-frame method of cell counting with specular microscopy (EB-10; Konan Medical, Inc, Irvine, CA) by an experienced eye bank technician (D.A.), marking at least 50 to 100 cells per image.18 The manufacturer's recommendations for calibration of magnification were followed. All corneal tissue was prepared by a single trained eye bank individual (H.Y.C.-U.) within 24 hours of DSAEK. The death-to-precutting time was recorded. Within a laminar flow hood, each donor corneaoscleral rim was mounted onto an artificial anterior chamber (AAC) (Moria SA, Antony, France) with a flow of Ringer solution after coating the endothelium with Viscoat (Alcon Laboratories, Ft Worth, TX). The Ringer solution bottle was set at 1.2 m above the AAC. The donor tissue was then held in place by a graft-tightening ring. The irrigation flow was increased to obtain optimal pressure in the AAC as confirmed by a Barraquer tonometer. After epithelial debridement, ultrasound pachymetry was used to measure the tissue thickness. The automated microkeratome was used to cut the donor tissue, and this lamellar cut was performed in one continuous movement, producing an anterior free cap and a residual posterior lamella. Four microkeratome heads were used: the 200-μm and 250-μm heads for thinner donors, and the 300-μm and 350-μm heads for thicker donors. The technician varied the speed of cutting depending on the initial donor thickness, that is, for thicker donors, a slower cutting transition speed was used to obtain thinner posterior lamellar lenticules. The cutting transition time was counted using a stopwatch.
Posterior Lamellar Corneal Tissue Characteristics
The posterior lamellar thickness was immediately measured using repeat ultrasound pachymetry. The depth of actual cutting was calculated as donor thickness before precutting minus the posterior lamellar thickness after precutting. The anterior cap and posterior lamella were then placed back together in the original Optisol GS-filled storage container. The central ECD was subsequently remeasured using a fixed-frame method, marking at least 50 to 100 cells per image, by the same technician, with the same specular microscopy as described above. The change in the ECD was calculated by subtracting the ECD before precutting from the ECD after precutting. The prepared corneoscleral rim was then stored at 4°C until surgery. Complications such as amputated anterior caps due to incomplete cutting or perforated posterior corneal lamellae were recorded as failed precutting procedures.
Statistical analyses were performed using STATA software (version 13, STATACrop, College Station, TX). The Pearson correlation test was used to assess the correlation between ECDs before and after precutting. Differences between groups were assessed with a 2-sided paired t test for continuous variables and with a χ2 test for categorical variables. Multivariate regression analysis was performed for variables that had P values less than 0.1 in the univariate regression analysis to determine the risk factors for donor endothelial loss after precutting. After assessing the collinearity among these factors, a backward stepwise regression analysis was performed for further factor selection. The final model adopted was the most parsimonious one that was believed to adequately explain the data. Multivariate binary logistic regression was further performed to analyze the association between the significant risk factors identified in the multivariate regression analysis and greater than 5% ECD loss after precutting. All data were expressed as mean ± SD. P values less than 0.05 were considered statistically significant.
There were a total of 3 failed precutting procedures consisting of 2 perforated posterior corneal lamellae (both were cut by 350-μm head) and 1 amputated anterior cap (cut by 300-μm head) due to incomplete cutting, representing 1.2% of all attempted procedures. These 3 failed precutting procedures were excluded from further statistical analysis. The remaining 256 donor corneas were procured from 4 eye banks: the Singapore Eye Bank (53 corneas, 20.7%), the Santa Lucia International Eye Bank of Manila (15 corneas, 5.9%), SightLife, Seattle, WA (52 corneas, 20.3%), and the National Eye Bank of Sri Lanka (136 corneas, 53.1%). The mean donor age was 57.18 ± 11.35 years (range, 22–79), and 174 of 256 donor corneas (68.0%) were from male donors. Table 1 summarizes the donor and tissue characteristics, and precut procedure parameters of the precut corneas.
The average percentage of changes in the ECD after precutting was −1.38% ± 3.28% (Table 1 and Fig. 1). The ECD after precutting was strongly positively correlated with the ECD before precutting (r = 0.856; Fig. 2A). The donor and tissue characteristics of local versus overseas precut corneas are shown in Table 2. There were no significant differences in the donor age (P = 0.160), death-to-preservation time (P = 0.071), death-to-precutting time (P = 0.329), ECDs before and after precutting (P = 0.108 and P = 0.395), and percent changes in the ECD after precutting (P = 0.548).
In the univariate regression model, a higher ECD before precutting and a slower cutting transition speed were significantly associated with greater percentage of ECD loss after precutting (P < 0.001 and P = 0.028, respectively; Table 3 and Fig. 2B). Multivariate regression analysis showed that an older donor age, a higher ECD before precutting, and a slower cutting transition speed were significantly associated with greater ECD loss after precutting after adjusting for all other covariates (P = 0.044, P < 0.001, and P = 0.034, respectively; Table 3). Corneas with an ECD >2800 cells per square millimeter before precutting, cutting transition time >5.5 seconds, and corneas with donor age >65 years were significantly more likely to have greater than 5% ECD loss after precutting after adjusting for all other covariates [odds ratio (OR), 6.42, 95% confidence interval (CI), 1.44–29.43, P = 0.025; OR, 1.66, 95% CI, 1.45–2.72, P = 0.039; and OR, 1.62, 95% CI, 1.66–5.82, P = 0.048, respectively]. Among the 256 precut tissues, 6 (2.3%) had unrecognizable and uncountable endothelial cells in the specular microscopy evaluation after precutting and therefore were not released for DSAEK. Compared with the 250 tissues that the endothelial cells were well recognizable by specular microscopy, these 6 corneal tissues had a significantly higher ECD before precutting (2821 ± 114 cells/mm2 vs. 3057 ± 266 cells/mm2, P = 0.019), slower cutting transition speed (4.13 ± 0.46 vs. 5.38 ± 0.53 seconds, P < 0.001), and greater tissue thickness before precutting (432.14 ± 40.14 vs. 466.00 ± 40.85 μm, P = 0.045).
Maintaining a high ECD in a posterior corneal lamella after precutting is clearly important in DSAEK. We found that the mean ECD loss after precutting was 1.38% in our study, which was in agreement with previous studies, ranging 1.0% to 3.7% evaluated by specular microscopy.7,15,16 Compared with the use of endothelial vital staining, the ECDs measured using specular microscopy may be prone to sampling errors because this technique is dependent on extrapolation of data from an assessment of a relatively small area of the cornea.19 However, it is the most widely used tool to analyze corneal endothelium available in eye banks. Furthermore, unlike the highly variable endothelial cell counts measured by specular microscopy before and after precutting in a previously published study,16 our study confirmed that the ECD before and after precutting was strongly correlated (r = 0.856), which indicates consistency in the cell counts. The analysis was performed by the same technician, using the same specular microscope and same method, before and after precutting. In this study, a higher ECD before precutting was significantly associated with greater percentage of ECD loss after precutting, which has been previously noted.16,19 The correlation between the ECDs before and after precutting was positive (r = 0.856). Taken together, it indicates that tissue with a higher ECD before precutting still has a higher ECD after precutting compared with tissue with a lower ECD before precutting, but has a greater percentage of ECD loss after precutting. Moreover, it has been postulated that the substantial elevation of AAC pressure during the microkeratome cutting thins the corneal stroma and flattens any Descemet folds, rendering the visualization and counting of endothelial cells easier after precutting.19
A slower cutting transition speed seemed to be a significant risk factor for greater ECD loss. One possible reason for greater cell loss in slower cutting transition speeds could be that the longer application of oscillating microkeratome blade with oscillation waves may result in greater mechanical damage to endothelial cells. We found that a cutting transition time >5.5 seconds was 1.66 times more likely to have greater than 5% ECD loss after precutting compared with a speed ≤5.5 seconds. Of note, a deeper cut was not significantly associated with ECD loss although the microkeratome blade was closer to the endothelial cells (with the range of residual posterior lamellar thickness being 43–232 μm). Moreover, death-to-preservation time (range, 0.67–10.88 hours), and death-to-precutting time (range, 0–7 days) were not associated with greater ECD loss. In our study, 6 (2.3%) precut tissues had unrecognizable and uncountable endothelial cells in specular microscopy evaluation after precutting, and they were not released for DSAEK because of the uncertainty of the endothelial quality, although they appeared optically clear under slit-lamp immediately after precutting. The availability of specular microscopy examination is one of the major advantages of using precut tissue, because it provides useful information on the ECD and screens the graft quality. We found that these 6 tissues had significantly greater corneal thickness before precutting and longer cutting transition time compared with other tissues.
An older donor age was another risk factor significantly associated with greater ECD loss after precutting after adjusting for all other covariates. Similarly, older donor age has also been reported as a predictive factor for greater ECD loss after PK, after adjusting for baseline ECD.20,21 Morphological changes associated with aging corneas, for example, corneal guttate or advanced glycation end products, can affect the cellular attachment and may explain the greater cell loss.22,23
The ECD change in tissue that has undergone long-distance transportation is an important topic for consideration, especially for centers that import tissue. We did not observe significant differences in the ECD before or after precutting, and in the percentage of ECD loss between local versus overseas corneas (Table 2). In addition, donor source was not associated with the ECD loss after precutting in the regression analysis (Table 3). Shimazaki et al has reported that the endothelial cell loss associated with overseas transportation was acceptable and the surgical outcomes of PK using imported grafts were comparable with those using domestic grafts.16,24
It is hoped that the use of precut tissues prepared by eye banks would make DSAEK more accessible to corneal surgeons. The precutting procedure with a failure rate less than 0.5% was proposed as reliable and robust.19 The overall eye bank precutting failure rate was 1.2% in this study, which was comparable with those published with rates of 1.5% to 1.9%.7,8,19 For the corneoscleral rim that was one of a pair, from the same donor, we reused the microkeratome blade that had been used on the mate of the same donor. In this study, both of the perforated tissues were cut by reused microkeratome blades. However, there were another 37 instances where the blades were used twice but the cuts were uneventful. Last, it has been demonstrated that after 48 hours of precutting and storage, the endothelial cell loss can increase to 11%20,25 and the posterior lamellae swell because of deepithelialization.5,25 Therefore, in our institute, the posterior corneal lamellae were all prepared within 24 hours of DSAEK.
In conclusion, we evaluated the donor, tissue, and precut procedure risk factors for donor endothelial loss after precutting by the eye bank for DSAEK. An older donor age, a higher ECD before precutting, and a slower cutting transition speed were significantly associated with greater ECD loss after precutting. Because DSAEK has evolved to be the most frequently performed EK procedure and the use of precut tissue is becoming increasingly popular,10 it is important to know which factors are associated with greater ECD loss in precutting procedures.
2. Price FW Jr, Price MO. Descemet's stripping with endothelial keratoplasty in 50 eyes: a refractive neutral corneal transplant. J Refract Surg. 2005; 21:339–345.
3. Price FW Jr, Price MO. Descemet's stripping with endothelial keratoplasty in 200 eyes: early challenges and techniques to enhance donor adherence. J Cataract Refract Surg. 2006; 32:411–418.
4. Koenig SB, Covert DJ. Early results of small-incision Descemet's stripping and automated endothelial keratoplasty. Ophthalmology. 2007; 114:221–226.
5. Suwan-Apichon O, Reyes JM, Griffin NB, et al. Microkeratome preparation of lamellar corneal grafts. Eye Contact Lens. 2006; 32:248–249.
6. Terry MA. Precut tissue for descemet stripping automated endothelial keratoplasty: complications are from technique, not tissue. Cornea. 2008; 27:627–629.
7. Chen ES, Terry MA, Shamie N, et al. Precut tissue in Descemet's stripping automated endothelial keratoplasty donor characteristics and early postoperative complications. Ophthalmology. 2008; 115:497–502.
8. Kitzmann AS, Goins KM, Reed C, et al. Eye bank survey of surgeons using precut donor tissue for descemet stripping automated endothelial keratoplasty. Cornea. 2008; 27:634–639.
9. Price MO, Baig KM, Brubaker JW, et al. Randomized, prospective comparison of precut vs surgeon-dissected grafts for descemet stripping automated endothelial keratoplasty. Am J Ophthalmol. 2008; 146:36–41.
10. Ruzza A, Salvalaio G, Bruni A, et al. Banking of donor tissues for descemet stripping automated endothelial keratoplasty. Cornea. 2013; 32:70–75.
11. Campolmi N, Gauthier AS, Montard R, et al. Setting up organ-cultured corneas pre-cutting by a French blood center-eye bank. Acta Ophthalmol. 2013; 91:0
12. Shinton AJ, Tsatsos M, Konstantopoulos A, et al. Impact of graft thickness on visual acuity after Descemet's stripping endothelial keratoplasty. Br J Ophthalmol. 2012; 96:246–249.
13. Terry MA, Shamie N, Chen ES, et al. Precut tissue for Descemet's stripping automated endothelial keratoplasty: vision, astigmatism, and endothelial survival. Ophthalmology. 2009; 116:248–256.
14. Suwan-Apichon O, Reyes JM, Griffin NB, et al. Microkeratome versus femtosecond laser predissection of corneal grafts for anterior and posterior lamellar keratoplasty. Cornea. 2006; 25:966–968.
15. Jones YJ, Goins KM, Sutphin JE, et al. Comparison of the femtosecond laser (IntraLase) versus manual microkeratome (Moria ALTK) in dissection of the donor in endothelial keratoplasty: initial study in eye bank eyes. Cornea. 2008; 27:88–93.
16. Yamazoe K, Yamazoe K, Shinozaki N, et al. Influence of the precutting and overseas transportation of corneal grafts for Descemet stripping automated endothelial keratoplasty on donor endothelial cell loss. Cornea. 2013; 32:741–744.
17. Tan DT, Janardhanan P, Zhou H, et al. Penetrating keratoplasty in Asian eyes: the Singapore Corneal Transplant Study. Ophthalmology. 2008; 115:975–982.e1.
18. Mehta JS, Por YM, Poh R, et al. Comparison of donor insertion techniques for Descemet stripping automated endothelial keratoplasty. Arch Ophthalmol. 2008; 126:1383–1388.
19. Kelliher C, Engler C, Speck C, et al. A comprehensive analysis of eye bank-prepared posterior lamellar corneal tissue for use in endothelial keratoplasty. Cornea. 2009; 28:966–970.
20. Musch DC, Meyer RF, Sugar A. Predictive factors for endothelial cell loss after penetrating keratoplasty. Arch Ophthalmol. 1993; 111:80–83.
21. Cornea Donor Study Investigator Group; Lass JH, Gal RL, et al. Donor age and corneal endothelial cell loss 5 years after successful corneal transplantation. Specular microscopy ancillary study results. Ophthalmology. 2008; 115:627–632.e8.
22. Hillenaar T, van Cleynenbreugel H, Remeijer L. How normal is the transparent cornea? Effects of aging on corneal morphology. Ophthalmology. 2012; 119:241–248.
23. Kaji Y, Amano S, Usui T, et al. Advanced glycation end products in Descemet's membrane and their effect on corneal endothelial cell. Curr Eye Res. 2001; 23:469–477.
24. Shimazaki J, Shinozaki N, Shimmura S, et al. Efficacy and safety of international donor sharing: a single-center, case-controlled study on corneal transplantation. Transplantation. 2004; 78:216–220.
25. Rose L, Briceño CA, Stark WJ, et al. Assessment of eye bank-prepared posterior lamellar corneal tissue for endothelial keratoplasty. Ophthalmology. 2008; 115:279–286.
precut; eye bank; endothelial loss; Descemet stripping automated endothelial keratoplasty
Copyright © 2014 by Lippincott Williams & Wilkins
Highlight selected keywords in the article text.