Comparison of Endothelial Cell Loss following the Big Bubble versus the Microbubble Incision Technique during Deep Anterior Lamellar Keratoplasty in Eyes with Keratoconus

Introduction. Deep anterior lamellar keratoplasty (DALK) is now becoming an increasingly popular surgical technique in treating corneal stromal pathologies with healthy endothelium. Several advantages of DALK over penetrating keratoplasty (PKP) have been described such as maintenance of globe integrity, absence of endothelial rejection, and a low rate of chronic endothelial cell loss (ECL). ECL following PKP results in 50% cell loss after 2 years from the estimated graft endothelial cell density (ECD). Although there are several reports confirming that ECL following DALK is similar to the physiologic cell loss 2 years after surgery, few reports discussed the surgically induced ECL due to difficulty in preoperative imaging of ECD. Materials and Methods. &is prospective, interventional study included 20 eyes of 20 patients, who underwent DALK surgery. 11 eyes underwent DALK using the big bubble technique, while 9 eyes underwent the microbubble technique. Postoperative evaluation was done 3 months after surgery and included best spectacle corrected visual acuity (BSCVA), keratometric readings, and refraction measured using an autokeratorefractometer (Topcon KR800, Japan) and endothelial cell density (ECD) using noncontact specular microscopy (Nidek CEM-530, Japan). Results. Regarding postoperative parameters such as postoperative logMAR visual acuity, postoperative mean K, and postoperative K max, there was no statistical difference found between both groups (P � 0.754, P � 0.119, and P � 0.970, respectively). Regarding change in specular endothelial cell density and percent change in the specular endothelial cell density, again there was no statistical difference between both groups with P � 0.057 and P � 0.126, respectively (significance defined as P< 0.05). Conclusion. ECD is not affected by failure of the big bubble to form and continuing DALK via the microbubble technique.


Introduction
Penetrating keratoplasty (PKP) has been the treatment of choice for advanced cases of keratoconus for a long time [1][2][3]. During the past decade, however, because of advancement in surgical techniques, deep anterior lamellar keratoplasty (DALK) has gained popularity in the treatment of keratoconus [3,4].
Comparative studies with PKP have shown overall similar visual outcomes when both techniques are used. Several advantages of DALK over penetrating keratoplasty (PKP) have been described such as maintenance of globe integrity, early suture removal, absence of endothelial rejection, and a low rate of chronic endothelial cell loss (ECL) [5,6].
Anwar and Teichmann proposed a technique, which they called the big bubble technique, where air is injected into the deep stroma with the aim of inducing separation by cleavage between posterior stroma and the DM, allowing the surgeon to gain a safe and direct access to this plane, with the advantages of shortening the surgical time, reducing the risk of perforation, and exposing a smooth, even surface of excellent optical quality [7].
In normal eyes, 0.6% of the endothelial cells are lost every year [15]. Corneal endothelial cell loss after penetrating keratoplasty occurs at a higher than physiologic rate to a cumulative cell loss of 50% or more within the first 10 years.
is suggests that, after the initial surgical trauma, donor endothelial cell survival is compromised in the host ocular environment [16,17].
Deep anterior lamellar keratoplasty has been associated with lower levels of endothelial cell loss compared with penetrating keratoplasty. is may be because of the reduced surgical trauma in DALK compared with penetrating keratoplasty and the absence of endothelial cell rejection in patients undergoing DALK [18,19].
Although there are several reports confirming that ECL following DALK is similar to the physiologic cell loss 2 years after surgery, few reports discussed the surgically induced ECL due to difficulty in preoperative imaging of ECD [20].
We herein report the surgically induced ECL following DALK comparing 2 techniques, the BB and the MB incision techniques.

Patients and Methods
is prospective, interventional study included 20 eyes of 20 patients. e study was conducted in accordance with the tenets of the Declaration of Helsinki and was approved by the Institutional Ethics Committee in Alexandria University.

Exclusion Criteria
(1) Coexistent ocular pathology or previous ocular surgery (2) Corneal scarring involving Descemet's membrane (DM) (3) Advanced disease interfering with specular evaluation in either eye (4) Patients with intraoperative perforation of DM 2.3. Pre-and Postoperative Assessments. Informed consent was obtained from the patients that were included in the study after explanation of the details of the study and of the procedure to be performed.
Preoperative evaluation included the patient's disease history, BCVA measured using the standard Snellen chart and converted to logarithm of the minimum angle of resolution (logMAR) for statistical analysis, slit-lamp microscopy, fundus examination, Pentacam evaluation of the cornea using the Allegro Oculyzer (WaveLight; Alcon, Erlangen, Germany), and ECD using noncontact specular microscopy (Nidek CEM-530, Japan). In eyes with advanced disease interfering with proper image acquisition, the fellow eye is used as a reference to the preoperative ECD.
Postoperative evaluation was done 3 months after surgery and included BCVA, Keratometric readings and refraction measured using an autokeratorefractometer (Topcon KR800, Japan) and ECD.

Surgical
Technique. All surgeries were performed using the BB technique, previously described by Anwar and Teichmann [7] with some modifications from the original technique, such as performing a paracentesis before bubble formation to inject small air bubbles in the anterior chamber and using a Fogla 27-gauge air injection cannula (Bausch and Lomb, Rochester, NY) to create the BB.
Eyes with failed big bubble formation after multiple trials of intrastromal air injection were managed using the MB incision technique previously described by Riss et al. [14].

Statistical Analysis.
Data were fed to the computer and analyzed using IBM SPSS software package version 20.0 [21]. Qualitative data were described using number and percent. Quantitative data were described using range (minimum and maximum), mean, standard deviation, and median. Significance of the obtained results was judged at the 5% level.

Results
is study included 20 eyes of 20 patients, 11 eyes were operated via the BB technique (BB group), while in 9 eyes, the MB incision techniques were used due to failure of BB formation (MB group). e mean patient age was 23.73 ± 6.96 years in the BB group (range 15-40 years) and 26.44 ± 9.41 years in the MB group (range 12-43 years). ere were 3 males and 8 females in the BB group, while 7 males and 2 females were present in the MB group. e difference in age and gender distribution was statistically nonsignificant between the two groups: P � 0.467 and P � 0.07, respectively.
By using Mann-Whitney nonparametric test to compare the mean rank of preoperative logMAR visual acuity 2 Journal of Ophthalmology between the big bubble group (12.50) and the microbubble group (8.06), there was a statistical difference between the mean ranks of the two groups (P � 0.039) (statistical difference considered at P < 0.05). By analysis of the preoperative parameters including preoperative mean keratometric reading (mean K) and preoperative maximum keratometric reading (K max) and comparing their mean ranks in both groups, no statistical difference was found (P � 0.423 and P � 0.879, respectively).
As for postoperative parameters such as postoperative logMAR visual acuity, postoperative mean K, and postoperative K max, there was no statistical difference found between both groups (P � 0.754, p � 0.119, and P � 0.970, respectively) Table 1.
Comparing preoperative and postoperative endothelial cell density (ECD) between both groups showed that there is no statistical difference regarding the mean rank of the preoperative and postoperative ECD between both groups using Mann-Whitney nonparametric test (P � 0.270 and P � 0.732, respectively).
Regarding change in specular endothelial cell density and percent change in the specular endothelial cell density, again there was no statistical difference between both groups with p � 0.057 and P � 0.126, respectively (significance defined as P < 0.05) Table 2.
Testing, if there is a relation between the percent change in specular endothelial cell density (ECD) and percent change in mean K, proved that there is a negative relation between both with correlation coefficient equals −0.450 (P � 0.047) (correlation is significant at the 0.05 level) using spearman correlation test.

Discussion
Sufficient corneal endothelial cell density is required for long-term functional success of any type of keratoplasty. With penetrating keratoplasty, an overall endothelial cell loss has been reported of approximately 33% within the first 2 postoperative years and that the cell density continues to decrease at an accelerated rate up to 20 years after surgery [13,16,[22][23][24].
Following DALK surgery, endothelial cell loss due to allograft rejection decreases substantially. Sugita et al. [13] reported that endothelial cell loss after DALK was 13% at the end of first year. Van Dooren et al. [25] found that ECD showed an 11% decrease during the first six months after DALK, and afterwards, the decrease was 1%-2% per year. ey also found that the decrease in ECD was similar to that of nonoperated healthy corneas [25]. Among the reasons for large decreases in ECD after DALK surgery, perioperative air injection into the anterior chamber and trauma of the recipient endothelium during deep stromal dissection has been proposed [25,26].
No study was done to evaluate the effect of surgical trauma of various techniques of DALK on the ECD.  In the current study, our aim was to see if there is an effect of the technique of DALK surgery (big bubble versus microbubble techniques) on the percent change of the ECD postoperatively.
No statistical difference was seen between the mean rank of percent ECD change of both the big bubble and the microbubble, with a P � 0.126, denoting little effect surgical technique has on the postoperative endothelial cell density.
ere was a negative relation between the percent change of mean K and the percent change of ECD, with r � −0.45; this finding was also mentioned by Salouti et al. [27] with the second-order polynomial regression analysis showing a weak but significant association between the postoperative changes in keratometry and postoperative ECD changes [27].
is study has several limitations: first, the small number of patients, resulting of less power due to the smaller sample size. In addition, we did not study some important covariates, such as intraocular pressure in the regression model of the postoperative endothelial cell loss. Finally, possible inherent inaccuracies in automatic optical measurements of endothelial cell profile might confound the outcomes.
In conclusion, our results show that ECD are not affected by failure of the big bubble to form and continuing DALK via the microbubble technique.

Data Availability
Data used to support the findings of this study are available from the corresponding author upon request. Disclosure e authors alone are responsible for the content and writing of the paper.

Conflicts of Interest
e authors report no conflicts of interest.