For decades, surgeons worldwide have worked to improve

Transcription

1 CLINICAL SCIENCE Pachymetry-Guided Intrastromal Air Injection ( Pachy-Bubble ) for Deep Anterior Lamellar Keratoplasty: Results of the First 110 Cases Ramon C. Ghanem, MD, PhD, Ayla Bogoni, MD, and Vinícius C. Ghanem, MD, PhD Purpose: To report intraoperative and 1-year postoperative results of the pachy-bubble technique for deep anterior lamellar keratoplasty (DALK). Methods: This prospective interventional case series included 110 eyes of 107 patients with anterior corneal pathology who underwent DALK, including 78 with keratoconus. Outcome measures included the rate of bubble formation, rate of completing DALK, bubble types, complications, and visual and keratometric parameters. Results: Intrastromal air injection was attempted in 109 eyes, and the air bubble was achieved in 93 eyes (85.3%). Intrastromal 2% methylcellulose injection was attempted in 9 eyes, after unsuccessful air bubble formation, and the viscobubble was achieved in 7 eyes (77.8%). Manual layer-by-layer dissection was performed in 8 eyes. Bubble formation was reached in 100 eyes (90.9%). Overall, 105 eyes (95.5%) achieved DALK. Air bubble occurred as type 1 (white margin) in 96.6% of the cases and as type 2 (clear margin) in 3.4%. There was a statistically significant improvement in all visual and keratometric parameters analyzed. Macroperforations converted to penetrating keratoplasty occurred in 5 eyes (4.5%) and microperforations in 12 (10.9%). There were significantly higher rates of perforation when a bubble was not achieved (P = 0.018) and when it was achieved as type 2 (P = 0.033). Interface haze occurred in 5 eyes (4.5%) and stromal rejection in 11 (10.0%). A short learning curve was observed for air bubble formation. Conclusions: Bubble formation, especially type 1, is the key to decrease the risk of perforation in DALK. The pachy-bubble was safe, effective, and reproducible in promoting DALK with air bubble and viscobubble formation with a short learning curve. Key Words: corneal transplantation, lamellar keratoplasty, Descemet membrane, keratoconus (Cornea 2015;34: ) Received for publication October 19, 2014; revision received January 23, 2015; accepted January 28, Published online ahead of print March 17, From the Cornea Department, Sadalla Amin Ghanem Eye Hospital, Joinville, Santa Catarina, Brazil. The authors have no funding or conflicts of interest to disclose. Reprints: Ramon C. Ghanem, MD, PhD, Sadalla Amin Ghanem Eye Hospital, 35 Camboriu St, Joinville, SC , Brazil ( ramonghanem@gmail.com). Wolters Kluwer Health, Inc. All rights reserved. For decades, surgeons worldwide have worked to improve deep anterior lamellar keratoplasty (DALK) with complete stromal removal. The goal is to replace the affected stroma, preserving healthy Descemet membrane (DM) and endothelium and creating a smooth bed to receive the corneal graft. This procedure combines the excellent visual results of fullthickness penetrating keratoplasty (PK) and avoids its main complication, endothelial rejection. It also decreases the risks of endophthalmitis and expulsive hemorrhage 1 3 and allows the eventual regrafting months or years after the first surgery without compromising the endothelium. However, the major challenge in this surgery is to consistently achieve separation of the posterior stroma from DM or the recently described Dua layer (DL), 4 which can be accomplished by injecting air, viscoelastic, or saline into the corneal stroma. The most commonly adopted method is intrastromal air injection through the trephination groove with a needle, called the Anwar big-bubble technique. 5 When a bubble is achieved, a smooth dissection plane is formed close to DM, which greatly facilitates DALK success. However, to increase the chances of bubble formation, air injections should be made in the deep stroma, close to DM. 6 When air is injected superficially, it spreads laterally inside the corneal lamellae, not forming the big-bubble. Deeper air injections, however, increase the risk of perforation. 5 9 The pachymetry-guided intrastromal air injection ( pachy-bubble ) technique was described in 2012 by Ghanem et al. 10 This technique uses intraoperative corneal pachymetry to guide a diamond knife incision that assists in air injection in the deep stroma using a blunt cannula. The purpose of this article is to report the intraoperative and 1-year postoperative results of the first 110 eyes that underwent DALK with the pachy-bubble technique. PATIENTS AND METHODS This prospective noncomparative interventional case series evaluated the surgical outcomes of 110 eyes of 107 consecutive patients who underwent DALK at Sadalla Amin Ghanem Eye Hospital between October 2009 and August The first 34 consecutive cases of this series were described in the original article defining the technique. 10 A single surgeon (R.C.G.) performed all surgeries using a standardized DALK technique. This study was approved by the institutional review board and was conducted in accordance Cornea Volume 34, Number 6, June

2 Ghanem et al Cornea Volume 34, Number 6, June 2015 with the principles outlined in the Declaration of Helsinki. Informed consents were obtained. Patients were included in the study if they had advanced keratoconus or corneal stromal disease with healthy endothelium and corrected distance visual acuity (CDVA) of 20/40 or worse. Patients with keratoconus who previously had corneal hydrops were excluded. There were 41 women and 66 men, with a mean age of years (range, years). Three patients had bilateral surgery. Preoperative diagnoses included 86 eyes with keratoconus, 9 with corneal ectasia after refractive surgery, 12 with corneal scarring, and 1 case each of granular dystrophy, lattice dystrophy, and Salzmann nodular degeneration. All keratoconus cases were advanced, with a mean apical curvature of D (range, D) on topography maps and a mean central corneal thickness of mm (range, mm). Corneal scarring was present in 24 eyes (27.9%). In 11 eyes (4 with keratoconus), no central ultrasound pachymetry could be obtained. Outcome measures included the rate of completing DALK, rate of air bubble or ophthalmic viscosurgical device (OVD)-bubble (viscobubble) formation, and bubble type. Intraoperative and early postoperative complications were also recorded. Preoperative and postoperative visual (uncorrected distance visual acuity, CDVA), refractive (sphere, cylinder, spherical equivalent), and keratometric (mean, differential, central, and apical keratometry) outcomes were recorded and analyzed. The Medmont E300 topographer (Medmont International, Vermont, Australia) was used to obtain keratometric data. Follow-up examinations were performed 1, 7, and 30 days and 3, 6, 9, 12, 24, and 48 months postoperatively. When further surgical intervention was required, such as refractive surgery on the graft, the results after the procedure were used in the analysis. Intraoperative and early postoperative results were reported for all patients, and only those with a minimum follow-up of 9 months were included in the final visual, refractive, and keratometric analyses. All surgical procedures were performed under peribulbar anesthesia using the pachy-bubble technique as described by Ghanem et al 10 with minor modifications described here. The size of trephination was determined according to the horizontal corneal diameter and the location of the cone. A Hessburg Barron suction trephine (Katena, Denville, NJ) (range, mm) was used for partial-thickness trephination to approximately 60% to 70% of corneal thickness, usually approximately 400 mm. Intraoperative corneal thickness measurements using ultrasound pachymetry (AccuPach VI; Accutome, Malvern, PA) were taken 0.8 mm internally from the trephination groove in the 11 to 1 o clock position (Fig. 1A). In this area, a 2-mm incision was made parallel to the groove with a micrometer diamond knife (Mastel, Rapid City, SD) calibrated to 90% depth of the thinnest measurement (Fig. 1B). A minimum residual stromal bed of 50 mm was maintained with a maximum of 70. The incision was opened with toothed forceps, the deep stroma was exposed, and initial dissection was performed with a Sinskey hook. A tunnel was then created with a Sarnicola DALK spatula (Asico, Westmont, IL) from this initial dissection plane until it reached the central or paracentral cornea (Fig. 1C). Through this tunnel, a Sarnicola 27-gauge cannula (Asico) was used to inject air in the deep stroma (Fig. 1D). Air bubble formation was observed in 2 different patterns, as previously described 11 :awhitishopaque circle delineating the size of the bubble, the white-margin bubble (type 1); or faint well-demarcated translucent DM detachment, the clear-margin bubble (type 2). No cases of bubble type 3 (mixed type) were observed. Surgical videos were reviewed and analyzed to verify bubble types. Subsequently, a limbal paracentesis was created using a 15-degree blade at the 12 o clock position to remove a small amount of aqueous humor and to lower the intraocular pressure. Afterward, anterior keratectomy was performed, and a thin layer of the corneal stromal tissue was left over the air bubble. A small amount of OVD (2% methylcellulose) was injected with an insulin needle inside the bubble to keep the posterior layers away. The small hole created with the needle was then enlarged with blunt-tipped Cohan-Vannas scissors. Next, when needed, the iris spatula was used to dissect the stroma in the periphery of the bubble to reach the trephination borders. Thereafter, the rest of the stromal tissue was completely excised using the Vannas scissors. Fullthickness donor corneal tissues stored in Optisol GS (Chiron Intraoptics, Irvine, CA) corneal storage medium were then used for transplantation. DM and the endothelium were removed with untoothed forceps, and the donor corneal button was punched out with a Hessburg Barron punch trephine. The donor trephine size was chosen according to the recipient s eye axial length. There was no host/donor disparity when the axial length was $23.5 mm. The donor cornea was oversized by 0.3 mm when the axial length was between 23.5 mm and 18 mm and by 0.5 mm when the axial length was shorter than 18.0 mm. In the first 60 cases of this series, after the air injection failed to form the bubble through the pachy-bubble incision, the next step was repeating air injection through another tunnel, at the same depth or slightly deeper. However, this approach was usually unsuccessful in forming the bubble. In the last 50 cases, after 2 or 3 consecutive air injections failed to form the bubble, the next step was to perform anterior keratectomy and to inject OVD in the residual stroma, usually through the pachy-bubble incision, to achieve the viscobubble (Figs. 2A D). Because of its high viscosity, OVD does not easily escape laterally, creating a much higher intrastromal pressure, increasing the chances of bubble formation. When OVD failed to form the bubble after several attempts, manual layer-by-layer dissection was performed to attempt total or subtotal stromal removal. The use of OVD in different ways during DALK was already reported by Sarnicola and Toro, Manche et al, Melles et al, and Shimmura et al When a perforation occurred, both the approximate size and location were recorded and categorized as either a macroperforation (larger than 1 mm) or a microperforation (1 mm or smaller). 16 After surgery, patients were treated as described for conventional DALK. The visual acuity measurements were converted from Snellen values to logarithm of the minimum angle of resolution (logmar) units. The data are presented as the mean 6 SD. Normality was assessed using the Shapiro Wilk test. Paired Wolters Kluwer Health, Inc. All rights reserved.

3 Cornea Volume 34, Number 6, June 2015 Pachy-Bubble Technique for DALK FIGURE 1. Pachy-bubble technique. A, Intraoperative corneal thickness measurements using ultrasound pachymetry are performed 0.8 mm internally from the trephination groove in the 11 to 1 o clock position. B, A 2-mm incision is created with a micrometer-controlled diamond knife, calibrated to 90% depth of the thinnest measurement or at a value of the thinnest corneal thickness at the incision site minus 60 mm. C, The incision is opened with toothed forceps, and a tunnel is created with a DALK spatula in the deep stroma. D, A 27-gauge DALK cannula is used to inject approximately 0.5 ml of air until a mediumto large-size air bubble is formed. analyses were performed using Student t test for keratometric and pachymetric data, which showed a normal distribution. The Wilcoxon and Kruskal Wallis tests were used for variables without a normal distribution. The data were analyzed using Microsoft Excel 2007 (Microsoft Corp, Redmond, WA) and IBM SPSS Statistics 20 software (IBM, Armonk, NY). P, 0.05 was considered statistically significant. RESULTS Rate of Completing DALK DALK was successfully performed in 95.5% of eyes (105 of 110). DALK was completed by air bubble formation in 90 eyes (81.8%), by viscobubble after air failed to form the bubble in 7 eyes (6.4%), and by manual layer-by-layer dissection after air and OVD failed to form the bubble in 8 eyes (7.3%). The mean pachymetry at the incision site was mm (range, mm). The mean incision depth was (range, ), and the mean calculated residual stromal bed was (range, 50 70). Rate of Air Bubble and Viscobubble Formation Overall, bubble formation (ie, air bubble or viscobubble) was achieved in 100 eyes (90.9%), including 3 eyes that were converted to PK during stromal removal. Intrastromal air injection was attempted in 109 eyes, and the air bubble was achieved in 93 eyes (85.3%). All cases except 2 needed only 1 air injection to form the air bubble. OVD injection was attempted in 9 eyes after unsuccessful air bubble formation, and the viscobubble was achieved in 7 eyes (77.8%). In 1 eye with ectasia after thick-flap LASIK, OVD injection was performed in the residual stroma after flap removal without previous air injection, and the viscobubble was achieved. Considering only the keratoconus eyes, DALK was completed in 81 of 86 eyes (93.6%) with a mean pachymetry at the incision site of mm (range, mm). There was no statistically significant difference in preoperative parameters between the eyes that achieved an air bubble and the ones that did not. Bubble Type Air bubble formation occurred as a type 1 bubble (white margin) in 96.6% (85/88) of the cases and as a type 2 bubble (clear margin) in 3.4% (3/88). Five surgical videos were not available for bubble type analysis. Type 2 bubbles were observed in 1 case each of the following disorders: granular dystrophy, keratoconus with central scarring, and herpes simplex stromal scar (Figs. 3A C). No statistically significant differences were observed in preoperative and postoperative variables between these 2 bubble groups, except for the incidence of intraoperative perforations (microperforation and macroperforations). Perforations occurred in type 1 bubbles in 9.4% of cases and in type 2 bubbles in 66.7% of cases (P = 0.033; Fisher exact test) (odds ratio: 19.2, 95% confidence interval: ; z statistic: 2.311; P = 0.021). In the patient with keratoconus and a type 2 bubble, the perforation occurred when injecting air in the anterior chamber during stromal removal, when DM was partially exposed. In this case, because of the large central tear, the case was converted to PK. In the patient with herpes scarring and a type 2 bubble, perforation also occurred during stromal removal but with a blunted-tip scissor touch in DM. The case was managed with gas tamponade. In both cases, DM was very fragile, as already described. 4 Wolters Kluwer Health, Inc. All rights reserved

4 Ghanem et al Cornea Volume 34, Number 6, June 2015 FIGURE 2. Standardized viscobubble technique after pachy-bubble. A, Air is injected through the pachy-bubble incision but fails to form the bubble. B, An anterior keratectomy is performed. C, The DALK cannula, attached to an OVD syringe, is inserted through the remaining of the pachy-bubble incision but fails to form the viscobubble, as OVD is observed refluxing from the incision (arrows). D, Another injection is performed in the midperiphery, achieving the viscobubble. Note the displacement of the anterior chamber air bubble to the periphery (arrow) (small bubble sign). Complications Intraoperative and postoperative complications were divided in to 2 groups: when an air bubble was achieved and when it was not (Table 1). There was a significantly higher rate of perforations when a bubble was not achieved (P = 0.018; Fisher exact test) (odds ratio: 4.5, 95% confidence interval: ; z statistic: 2.463; P = ). Five cases (4.5%), including 1 with a type 2 bubble, were converted to PK because of macroperforations. One early case had macroperforation during tunnel creation with the spatula. 10 The other cases were converted because of macroperforations during stromal removal. Air bubbles were achieved in 3 of these cases. Microperforations occurred in 12 cases (10.9%): 4 during manual layerby-layer dissection after air injection failed to form a bubble, 7 during removal of the residual stroma after air bubble formation, and 1 during the initial incision with the diamond knife. The management of these microperforations included air tamponade in 7 eyes and postoperative injection of 14% perfluoropropane (C 3 F 8 )in5becauseof DM detachment. One patient with an inferior microperforation required 2 C 3 F 8 injections. In 2 cases, intraoperative pachymetric measurements could not be obtained in the usual area of the diamond knife incision. Measurements were obtained in the recipient cornea, allowing a peripheral pachy-bubble technique. Peripheral and paracentral DM folds were observed in 15 cases but were visually insignificant and tended to fade over time. Central folds were not observed in any cases. Interface haze was observed in 5 eyes (4.5%), all graded mild. Haze occurred in 2 eyes after manual dissection, in 2 after viscobubble formation and in only 1 after air bubble formation in a patient with syphilitic interstitial keratitis scarring. In both cases of interface haze after viscobubble formation, residual stroma was observed on optical coherence tomography, 25 mm in1caseand 30 mm in the other. Visual and Keratometric Results All visual and keratometric parameters that were analyzed improved significantly after surgery in 79 eyes with a mean follow-up of months (range, 9 48 months) (Table 2). CDVA was (range, ) preoperatively and improved to (range, ) postoperatively (P, 0.001; related samples Wilcoxon test). The mean K value was D (range, ) preoperatively and decreased to D (range, ) postoperatively (P, 0.001; related samples Wilcoxon test). More eyes in the air bubble group achieved levels of 20/20 and 20/25 CDVA compared with the viscobubble and manual dissection groups, but the difference was not statistically significant. Learning Curve In a comparison of the initial and more recent cases, a short learning curve was observed for air bubble formation. The complications related to residual stromal removal (ie, microperforations and macroperforations) decreased after the 37th case, but the rate of air bubble formation was unchanged. In the last 73 cases, 3 years before this report, only 1 conversion to PK was necessary. A substantial improvement Wolters Kluwer Health, Inc. All rights reserved.

5 Cornea Volume 34, Number 6, June 2015 Pachy-Bubble Technique for DALK TABLE 1. Intraoperative and Postoperative Complications of Pachy-Bubble DALK Complications Air Bubble Achieved (n = 93), % Air Bubble Not Achieved (n = 16), % Intraoperative Microperforation 8 (8.6) 4 (25.0) Conversion to PK 3 (3.2) 2 (12.5) (macroperforation) Postoperative DM folds Paracentral 4 (4.3) 0 (0) Peripheral 9 (9.7) 0 (0) Interface haze 1 (1.1) 3 (18.7) Immunologic rejection Subepithelial infiltrates 7 (7.5) 0 (0) Stromal edema 4 (4.3) 0 (0) Double anterior chamber 4 (4.3) 1 (6.2) Persistent epithelial defect 4 (4.3)* 2 (12.5) (.15 d) Filamentary keratitis 6 (6.4) 1 (6.2) Early resuture 6 (6.4) 0 (0) Transitory ocular hypertension 1 (1.1) 0 (0) High astigmatism leading to graft refractive surgery GRI 2 (2.2) GRI 2 (12.5) GRI + PHACO 2 (2.2) PRK 2 (12.5) GRI + PRK 3 (3.3) PRK 2 (2.2) PIOL 1 (1.1) *Three severely atopic patients. One patient who had chemical burn and limbal stem cell deficiency had a persistent epithelial defect and needed amniotic membrane transplantation. GRI, graft relaxing incision; PIOL, phakic intraocular lens implantation; PRK, photorefractive keratectomy. FIGURE 3. Type 2 bubbles (clear margin) in pachy-bubble DALK. Faint well-demarcated translucent DM detachment (arrows) was observed in 1 case of each of the following disorders: granular dystrophy (A), keratoconus with central scarring (B), and herpes simplex stromal scarring (C). in bubble rate formation was only observed after a standardized approach of injecting OVD in the residual stroma was established for the cases of failed air bubbles (Fig. 4). An air bubble or viscobubble was achieved in all of the last 50 cases. DISCUSSION The pachy-bubble technique represents an effective, reproducible, and easy way of achieving air bubble and viscobubble formation during DALK, the key step for successful surgery with complete stromal removal. When no bubble is achieved, the surgeon has to address manual layerby-layer dissection, which is harder, takes longer, and as shown in this study, caries a higher risk of perforation. The overall rate of air bubble formation in this study (85%) compares favorably to other studies using the big-bubble technique. Fontana et al 7 reported a success rate of 64% in 81 eyes. Bhatt et al 17 achieved big-bubble formation in 25 of 46 patients (54.3%). Kubaloglu et al 18 achieved big-bubble formation in 193 of 234 eyes (82.4%). Muftuoglu et al 19 recently reported a series of 300 eyes using the big-bubble technique with cannula in which only 141 (47%) achieved big-bubble formation after air injection. The authors also nicely described the concept of air viscobubble formation, stating that micro-air-detachments in DM are filled by OVD after it is injected in cases of failed air bubble formation. However, few studies clearly state the rate of bubble formation; in fact, most state only the rate of DALK completion. In our study, considering only the last 50 cases, Wolters Kluwer Health, Inc. All rights reserved

6 Ghanem et al Cornea Volume 34, Number 6, June 2015 TABLE 2. Visual Outcomes After Pachy-Bubble DALK (n = 79)* Air Bubble (n = 64) Viscobubble (n = 7) Manual Dissection (n = 8) P UDVA (logmar) Mean 6 SD Range ( ) ( ) ( ) CDVA (logmar) Mean 6 SD Range ( ) ( ) ( ) $20/20, % $20/25, % $20/40, % *Patients with a minimum follow-up of 9 months were included in this analysis. Independent samples Kruskal Wallis test. UDVA, uncorrected distance visual acuity. when a standardized pachy-bubble/visco-bubble approach was established, a bubble was achieved in all cases, and DALK was completed in all but 1 case in which a type 2 bubble perforated. In this study, a 90% depth incision was performed as a standard, preserving a 50 mm safety limit of the residual stromal bed. As the mean pachymetry at the incision site was mm, the mean residual stromal bed was mm. This distance has historically been a good safety limit, with very good efficacy, and resulted in only 1 case of perforation during diamond knife incision. Because the objective of the pachy-bubble technique is to inject air as close as possible to the DL while keeping a safe distance from it, the authors currently calibrate the knife at a value of the thinnest corneal thickness at the incision site minus 60 mm. This measure may be safer in thin corneas, where we previously used 50 mm as the limit, and it enhances bubble formation rates in thick corneas, where we previously used 70 mm as the limit. The type of air bubble that is formed after air injection during DALK has important anatomical and functional differences. A type 1 bubble (white margin), as recently described by Dua et al, 4 has a stronger posterior wall made of DM and DL and has a popping pressure of 1.5 bar. The DL is an acellular layer made of predominantly type 1 collagen and measures approximately 10 mm. 4 In contrast, a type 2 bubble (clear margin) is made only of DM and has a popping pressure of 0.6 bar. This difference may explain the higher incidence of intraoperative perforations observed in type 2 bubbles in our study. Fortunately, type 1 bubbles were vastly predominant, occurring in 96.6% of the cases. This rate is higher than that of other studies in cadaver eyes 4,20,21 and may be explained by the fact that air injection is always performed in the central cornea with the pachy-bubble technique. When air is injected in the corneal periphery, there is a higher chance of forming type 2 bubbles, as the DL cleavage extends only #9 mm of the central cornea and is strongly adherent thereafter. 4 Perforations were the main complication in our study and were more common when a bubble was not achieved and residual stroma was removed by manual dissection. They also occurred after bubble formation during residual stroma removal but could usually be managed without conversion to PK. Interface haze was uncommon, but unexpectedly occurred in 2 of 7 cases of viscobubble, raising a hypothesis for a different, more superficial, plane of dissection in some cases when OVD is used. This finding was previously observed in some donor corneas in which air and OVD were injected to achieve a bubble. 19 Different studies have reported long learning curves with the Anwar big-bubble technique. 22,23 Smajda et al 22 reported a success rate of 60% in the first 20 cases, improving to 80% in the last 20. Feizi et al 23 reported a success rate of 55% in the first 20 keratoconus cases. Excluding the cases converted to PK, Feizi et al reported a final success rate of 82%. In our study, a short learning curve was observed for air bubble formation, a surgical step that is not dependent on the FIGURE 4. Short learning curve on the rate of bubble formation in pachy-bubble DALK with each consecutive group of 20 procedures. Improvement in the success rate was observed after a viscobubble approach was established for the cases of failed air bubbles Wolters Kluwer Health, Inc. All rights reserved.

7 Cornea Volume 34, Number 6, June 2015 Pachy-Bubble Technique for DALK surgeon s expertise when using the pachy-bubble. However, the conversion rate and the incidence of perforations decreased progressively with surgeon experience. In conclusion, the pachy-bubble technique represents a new way of achieving the air bubble during DALK. It is highly reproducible, easy to learn, has a low risk of perforation, and has shown a high success rate, especially when associated with the viscobubble when air fails to form the bubble. This standardized approach combining both techniques can increase the rate of bubble formation to nearly 100%. REFERENCES 1. Shimazaki J, Shimmura S, Ishioka M, et al. Randomized clinical trial of deep lamellar keratoplasty vs penetrating keratoplasty. Am J Ophthalmol. 2002;134: Watson S, Ramsay A, Dart J, et al. Comparison of deep lamellar keratoplasty and penetrating keratoplasty in patients with keratoconus. Ophthalmology. 2004;111: Sarnicola V, Toro P, Gentile D, et al. Descemetic DALK and predescemetic DALK: outcomes in 236 cases of keratoconus. Cornea. 2010;29: Dua HS, Faraj LA, Said DG, et al. Human corneal anatomy redefined: a novel pre-descemet s layer (Dua s layer). Ophthalmology. 2013;120: Anwar M, Teichmann K. Deep lamellar keratoplasty: surgical techniques for anterior lamellar keratoplasty with and without baring of Descemet s membrane. Cornea. 2002;21: Scorcia V, Busin M, Lucisano A, et al. Anterior segment optical coherence tomography-guided big-bubble technique. Ophthalmology. 2013;120: Fontana L, Parente G, Tassinari G. Clinical outcomes after deep anterior lamellar keratoplasty using the big-bubble technique in patients with keratoconus. Am J Ophthalmol. 2007;143: Fournié P, Malecaze F, Coullet J, et al. Variant of the big bubble technique in deep anterior lamellar keratoplasty. J Cataract Refract Surg. 2007;33: Leccisotti A. Descemet s membrane perforation during deep anterior lamellar keratoplasty: prognosis. J Cataract Refract Surg. 2007;33: Ghanem RC, Ghanem MA. Pachymetry-guided intrastromal air injection ( pachy-bubble ) for deep anterior lamellar keratoplasty. Cornea. 2012; 31: McKee HD, Irion LCD, Carley FM, et al. Residual corneal stroma in bigbubble deep anterior lamellar keratoplasty: a histological study in eyebank corneas. Br J Ophthalmol. 2011;95: Sarnicola V, Toro P. Deep anterior lamellar keratoplasty in herpes simplex corneal opacities. Cornea. 2010;29: Manche EE, Holland GN, Maloney RK. Deep lamellar keratoplasty using viscoelastic dissection. Arch Ophthalmol. 1999;117: Melles GR, Lander F, Rietveld FJ, et al. A new surgical technique for deep stromal, anterior lamellar keratoplasty. Br J Ophthalmol. 1999;83: Shimmura S, Shimazaki J, Omoto M, et al. Deep lamellar keratoplasty (DLKP) in keratoconus patients using viscoadaptive viscoelastics. Cornea. 2005;24: Den S, Shimmura S, Tsubota K, et al. Impact of the descemet membrane perforation on surgical outcomes after deep lamellar keratoplasty. Am J Ophthalmol. 2007;143: Bhatt UK, Fares U, Rahman I, et al. Outcomes of deep anterior lamellar keratoplasty following successful and failed big bubble. Br J Ophthalmol. 2012;96: Kubaloglu A, Sari ES, Unal M, et al. Long-term results of deep anterior lamellar keratoplasty for the treatment of keratoconus. Am J Ophthalmol. 2011;151: Muftuoglu O, Toro P, Hogan RN, et al. Sarnicola air-visco bubble technique in deep anterior lamellar keratoplasty. Cornea. 2013;32: McKee HD, Irion LC, Carley FM, et al. Donor preparation using pneumatic dissection in endothelial keratoplasty: DMEK or DSEK? Cornea. 2012;31: Yoeruek E, Bayyoud T, Hofmann J, et al. Comparison of pneumatic dissection and forceps dissection in descemet membrane endothelial keratoplasty: histological and ultrastructural findings. Cornea. 2012;31: Smadja D, Colin J, Krueger RR, et al. Outcomes of deep anterior lamellar keratoplasty for keratoconus: learning curve and advantages of the big bubble technique. Cornea. 2012;31: Feizi S, Javadi M, Jamali H, et al. Deep anterior lamellar keratoplasty in patients with keratoconus: big-bubble technique. Cornea. 2010;29: Wolters Kluwer Health, Inc. All rights reserved