Br J Ophthalmol doi:10.1136/bjophthalmol-2011-301168
  • Clinical science

Deep anterior lamellar keratoplasty using an original manual technique

  1. Giorgio Paganoni
  1. Department of Ophthalmology, Cornea and Ocular Surface Disease, San Raffaele Scientific Institute, Milan, Italy
  1. Correspondence to Dr Paolo Rama, Ospedale San Raffaele, Unità Cornea e Superficie Oculare, Via Olgettina 60, Milano 20132, Italy; rama.paolo{at}
  • Accepted 27 February 2012
  • Published Online First 18 March 2012


Aims To evaluate the clinical findings and visual outcomes of deep anterior lamellar keratoplasty (DALK) using an original manual dissection technique.

Methods 288 eyes (268 patients) with corneal pathologies without endothelial involvement were treated by DALK using an original manual dissection technique guided by a calibrated knife incision based on ultrasonic pachimetry values. Clinical records were examined retrospectively at 2 months, 6 months, 1 year and 2 years. The following outcomes were measured: visual acuity, topographic parameters, endothelial cell density and recipient stromal residue thickness.

Results At the 2-year postoperative follow-up, the mean logarithm of the minimum angle of resolution best spectacle corrected visual acuity (BSCVA) was 0.131±0.087 and topographic astigmatism was 2.87±1.57 diopters. In 12 cases (4.2%) a perforation of Descemet's membrane required conversion of the procedure to penetrating keratoplasty. Mean optical coherence tomography (OCT) residue thickness (measured in 82 eyes with OCT Visante) was 31.63±24.57 μm; lower values of recipient residue thickness were significantly associated with higher BSCVA (Spearman coefficient 0.635, p< 0.001).

Conclusion DALK using a dry manual dissection technique provides visual, refractive and clinical results comparable to other deep lamellar techniques. Eyes with lower values of recipient residue thickness are associated with better visual acuity.

Deep anterior lamellar keratoplasty (DALK) is a corneal graft technique consisting of the transplantation of a partial thickness donor cornea bared of its endothelium onto a recipient bed after full (or partial) removal of recipient stromal tissue. Compared with penetrating keratoplasty (PKP), this technique presents several advantages. First, it is an extraocular procedure and has a lower risk of complications such as the formation of anterior synechia, secondary glaucoma and endophthalmitis.1 Furthermore, there is no risk of endothelial rejection, the major cause of graft failure in PKP, because the host endothelium is not replaced. Consequently, there is no need for long-term immunosuppressive therapy with corticosteroids, decreasing the risk of cataract, glaucoma and infections.1–5 DALK has proved to result in similar visual acuity compared with PKP.1 2 6 The main concern regarding DALK is the technical difficulty regarding the surgical procedure, resulting in a steep learning curve.7 The most feared intraoperative complication is perforation of Descemet's membrane (DM), which occurs more often when the overlying stroma is completely removed. Recent developments in surgical techniques have favoured approaches such as the ‘big bubble’ technique, with the objective of removing all of the corneal stroma to obtain better visual results.8 However, it is not clear whether full stromal removal provides better results than cases in which a small portion of posterior stroma is left in place. Sugita and Kondo2 utilised a manual dissection technique and have shown that leaving a small amount of stroma does not produce differences in visual acuity compared with complete stromal dissection, as long as the deep stroma is healthy, while other authors9 intentionally leave a minimal thickness of stroma in order to reduce the risk of DM rupture.

We adopted an original manual technique in which a deep corneal incision obtained with a calibrated diamond knife is used as a guide to reach a pre-Descemetic plane.

Materials and methods

In this retrospective non-comparative study we analysed the clinical records of a consecutive series of 288 eyes of 268 patients who were treated with a dry dissection DALK technique from March 2003 to December 2008. Ethics committee approval was secured (San Raffaele Hospital Ethics Committee, Milan, Italy) and informed consent was obtained from all subjects. Patients affected by corneal disease requiring a graft without signs of endothelial alterations were recruited for the study. Exclusion criteria were the presence of ocular conditions such as glaucoma, cataract, or retinal diseases and previous ocular surgical procedures. Patients with any form of amblyopia were not included in the analysis of visual acuity.

We performed preoperative and postoperative eye examinations including Snellen uncorrected visual acuity (UCVA), best spectacle-corrected visual acuity (BSCVA), best corrected visual acuity (BCVA), slit-lamp examination, tonometry, fundus examination, corneal topography and endothelial cell specular microscopy. UCVA, BSCVA and BCVA were recorded as the decimal equivalent of a Snellen chart. Values were converted into the logarithm of the minimum angle of resolution (LogMAR) for statistical analysis. BCVA was assessed with a pinhole. Astigmatism was measured using an autorefractometer (NIDEK, Gamagari, Japan) and a corneal topographer (CSO, Firenze, Italy). The spherical equivalent was calculated from the best spectacle correction. Corneal endothelium was photographed after surgery using a non-contact specular microscope (Cellcheck-KONAN Medical Inc., Hyogo, Japan), and endothelial cell density (ECD) was calculated after marking the centre of 35 cells in a standard region of interest. Preoperative central and peripheral corneal thicknesses were measured using an ultrasonic pachimeter (Pacline, Opticon, Rome, Italy). Postoperative residual recipient thickness was measured after surgery using anterior chamber optical coherence tomography (OCT) Visante (Carl Zeiss Meditec, Dublin, California, USA). The residual recipient thickness was measured in the central 3 mm of the cornea, inside the optical zone. Visual function, refraction, topographical astigmatism and ECD were measured at 1, 2, 6, 12 and 24 months after surgery. Results are reported as mean values±SD (fifth; 95th percentiles).

Statistical analysis was performed using SPSS V.17.0. ECD measurements taken at different time intervals were analysed using the Wilcoxon signed-rank test. Stromal residue thickness was compared with mean visual acuity using a Spearman correlation and using the Mann–Whitney test when eyes were divided into two groups using a 20 μm cut-off.

Surgical technique

All patients were operated on by one of two surgeons (PR and GP) at the San Raffaele Hospital, Milan, Italy. Both surgeons were experienced in full-thickness and lamellar graft surgery. Surgical procedures were performed with the patients under general anaesthesia in 145 cases and local anaesthesia in 143 cases. After disinfection with povidone-iodine, the surgeon marked the appropriate graft diameter using a manual trephine and the centre of the cornea was marked in order to position the suction trephine correctly in the following steps. First the surgeon performed a 2 mm corneal incision 1 mm inside the marked diameter at the 12 o'clock position with a diamond knife (model ME425; Janach, Como, Italy) calibrated at a value of the thinnest corneal thickness (measured with ultrasonic pachimetry) minus 20 μm in order to avoid DM perforation. In the following step, a Hanna trephine (Moria, Paris, France) was calibrated to cut at 70% of the thinnest corneal thickness, and the superficial lamella was removed using a crescent knife (BD Medical, Franklin Lakes, New Jersey, USA). Alternatively, a Coronet trephine (Network Medical, Ripon, North Yorkshire, UK) was used. Then, starting from the deep corneal incision, a full diameter peripheral pocket was created using a sharp disc knife (Alcon, Fort Worth, Texas, USA) and Vannas scissors. Consequently, a blunt spatula (model J2401.1a; Janach) was utilised to reach a deep pre-Descemetic plane in the central 7 mm of the cornea. A sclerocorneal button stored in organ culture provided by the Monza Eye Bank (Monza, Italy) was used for the preparation of all the donor corneas. The donor cornea was trephined from the endothelial side with a Hanna trephine with a 0.25 mm greater diameter compared with the recipient bed. Endothelium and DM were removed from the donor button using a blunt spatula. Four interrupted 10.0 nylon sutures (Sharpoint; Angiotech, Vancouver, Canada) secured the graft on the recipient bed. Then a continuous 10.0 nylon running suture was placed. After removing the interrupted suture, a handheld manual keratometer (Duckworth and Kent, Baldock, UK) was used to adjust astigmatism. Ciprofloxacin (0.3%) was instilled topically, betamethasone (1.5 mg) was injected in the subconjunctival space, and the eye was patched at the end of surgery. All patients underwent postoperative prophylaxis with the following scheme: antibiotic topical treatment (ciprofloxacin 0.3%) three times a day for 15 days and topical treatment with dexamethasone (0.15%) three times a day for 15 days, two times a day for the succeeding 15 days and once a day until 2 months after surgery. The main steps of the surgical procedure are summarised in figure 1 and a supplementary video illustrating the surgical procedure is available online.

Figure 1

Surgical procedure. (A) The corneal diameter is marked using a manual trephine. (B) A 2 mm corneal incision 1 mm inside the marked diameter at the 12-o'clock position is performed with a diamond knife calibrated at a value of the thinnest corneal thickness (measured with ultrasonic pachimetry) minus 20 μm in order to avoid perforation of Descemet's membrane. (C) Hanna trephine calibrated to cut at 70% of the corneal thickness. (D) The superficial lamella is removed using a crescent knife (E) Starting from the deep corneal incision, a peripheral pocket is created using a sharp disc knife. (F) A full diameter peripheral pocket is created using a sharp disc knife and Vannas scissors. (G) A blunt spatula is utilised to reach a deep pre-Descemetic plane in the central 7 mm of the cornea. (H) A continuous 10.0 nylon running suture is placed.


During the time covered by this study, 268 patients underwent DALK. Of the 288 eyes, 260 were affected by keratoconus, 12 by post-infective leucomas, nine by leucomas from other causes, four by post-refractive surgery ectasia, two by granular dystrophy and one by reticular dystrophy. Of the patients, 187 were men and 81 were women. The mean age at the time of surgery was 32.49±11.96 years (18; 57). Mean preoperative keratometry was 55.13±6.96 diopters (45.08; 69.84) and mean topographical astigmatism was 6.26±4.27 diopters (1.11; 15.86). During the study, the following number of eyes reached each time point: 210 at 2 months, 214 at 6 months, 206 at 1 year and 196 at 2 years.

In all cases the diameter of the donor punch was 0.25 mm larger than that used for recipient trephination. In most cases, donor diameters were 8.25 mm or 8.5 mm. The minimum trephination diameter was 7.50 mm and the maximum was 8.75 mm. A Hanna trephine was utilised in 281 cases, while a Coronet trephine was used in seven cases.

Intraoperative complications occurred in 22 cases (7.6%). In 12 cases (4.2%) a perforation of DM required conversion of the procedure to PKP. Microperforation of DM occurred in 10 cases (3.5%) and did not influence the surgical outcome; air was injected in the anterior chamber in four cases. The mean time of surgery was 61±8.34 min.

The suture was removed at an average of 8.69±5.54 months (1; 19). At the 2-year postoperative follow-up, mean UCVA was 0.545±0.334 LogMAR (1; 0.069), BSCVA was 0.131±0.087 LogMAR (0.301; 0) and BCVA was 0.097±0.078 LogMAR (0.222; 0). The complete visual acuity results are reported in table 1 and figure 2.

Table 1

Postoperative LogMAR visual acuity. Mean values±SD (fifth; 95th percentiles)

Figure 2

Mean postoperative logarithm of the minimum angle of resolution visual acuity. Error bars: 95% CI.

Two years after surgery, patients had a mean spherical equivalent of −2.70±3.30 diopters (−8.5; 2), refractive astigmatism of −2.22±2.52 diopters (−6; 3.5) and topographic astigmatism of 2.87±1.57 diopters (0.81; 6.10). The complete refraction results are reported in table 2.

Table 2

Postoperative refraction (diopters). Mean values±SD (fifth; 95th percentiles)

Mean ECD was 2630.87±393.25 cells/mm2 (2012.80; 3248.10) at 2 months, 2512.21±356.75 cells/mm2 (1941.70; 3139.85) at 6 months, 2476.92±419.45 cells/mm2 (1755.20; 3180.00) at 1 year and 2411.82±469.38 cells/mm2 (1684.70; 3128.10) at 2 years. Endothelial cell loss was statistically significant only in the interval from 2 to 6 months (p=0.027) thus implying that ECD stabilised at 6 months.

Recipient stromal residue thickness was measured in 82 eyes (due to temporary availability of the instrument); when the stromal residue was not identified, a value of 0 μm was assigned. In our experience with this technique, we have not observed any difficulty in creating a regular stromal bed with a blunt spatula during the surgical procedure; this aspect has also been confirmed by analysis of the regularity of the stromal bed using anterior segment OCT images. Mean OCT residue thickness was 31.63±24.57 μm (0; 70.85). Lower values of recipient residue thickness were significantly associated with higher BSCVA (Spearman coefficient 0.635, p<0.001). The eyes were then assigned to two groups: eyes in group 1 had a residue thickness of 20 μm or less and eyes in group 2 had a residue thickness greater than 20 μm. The mean BSCVA after suture removal was significantly better in group 1 (0.082±0.040 LogMAR) compared with group 2 (0.142±0.061; p<0.001). Figure 3 illustrates three eyes with different residual stromal thicknesses.

Figure 3

Optical coherence tomography Visante images of three patients receiving deep anterior lamellar keratoplasty (12 months after surgery). (A) Stromal residue thickness of 53 μm. Best spectacle corrected visual acuity (BSCVA) was 0.2 logarithm of the minimum angle of resolution (LogMAR). (B) Stromal residue thickness of 26 μm. BSCVA was 0.1 LogMAR. (C) Stromal residue thickness was not measurable. BSCVA was 0.0 LogMAR.

Postoperative complications occurred in 14 cases (4.9%) throughout the 2 years of the study. The most frequent complication was graft rejection: nine cases of stromal rejection (3.1%) and two cases of epithelial rejection (0.7%). There were also single cases of herpes reactivation and cortisone glaucoma. The continuous suture was redistributed in three cases and was replaced by a single bite suture in 16 cases. Arcuate keratectomy to reduce postoperative astigmatism was performed on 11 eyes, five eyes underwent cataract phacoemulsification and one eye underwent photorefractive keratectomy; these subjects were not included in the visual acuity analysis.


Different lamellar graft techniques have been described during the past decade but the most popular at present is the big bubble technique described by Anwar and Teichmann.10 However, the technique can be technically challenging when the air bubble is not formed, requiring completion of the procedure using manual dissection of stromal layers that have been altered by the injection of air.11 The technique seems to show great visual acuity results when big bubble formation is achieved, but data regarding visual acuity when the bubble is not achieved are often contradictory. For example, a recent study12 reported non-significant changes in visual acuity when big bubble formation was achieved or not achieved, while a prospective study by Fontana et al11 clearly showed a significantly lower visual acuity in eyes in which big bubble formation did not occur when compared with eyes with successful bubble formation. More studies are needed to interpret data correctly.

In the current study, our lamellar technique had a success rate of 95.8%, while in 4.2% of cases a macroperforation of DM required intraoperative conversion of the procedure to PKP. Microperforation of DM occurred in 3.5% of cases without influencing the surgical outcome. Our data were in line with the results reported by several studies that indicated various rates of DM perforation and conversion to PKP ranging from 4% to 39.2% and from 0% to 14%, respectively.2 5 7 9–11 13–15 Such variations may depend on differences in indications, surgical techniques and the steep learning curves that characterise lamellar techniques.

A postoperative BSCVA of 20/40 or better was achieved in 96.2% of patients, which is comparable to results reported in other case series.10 12–14 16 Refractive error measured by spherical equivalent, refractive and topographic astigmatism were also similar to values described by other authors.7 9–11 13 14 16 17

Some studies indicated that visual outcomes after DALK are comparable with those after PKP,1 13 while others have described visual acuity results inferior to PKP.17–19 These differences can be attributed to scarring and irregularity at the host–donor interface leading to poorer visual results in DALK. When comparing patients who received DALK or PKP, Ardjomand et al16 reported that visual acuity of the DALK group was similar to the PKP group when the recipient residue thickness was less than 20 μm, while the BSCVA of the DALK group decreased significantly with increasing thickness of the recipient stroma. We decided to use the same 20 μm cut-off to divide our cases into two groups, and demonstrated that BSCVA was significantly better in eyes with a recipient residue thickness of 20 μm or less when compared with eyes that had a residue thickness greater than 20 μm (p<0.001). However, recipient residue was often difficult to measure and there were cases in which residue thickness was less than 20 μm but BSCVA was low, and cases in which residue thickness was over 20 μm with high BSCVA values. Different published studies indicate that there are no differences between groups, with bared DM and pre-Descemet groups showing that the presence of stromal tissue does not influence the final visual outcome.2 9 These conflicting results indicate that a quantitative evaluation of stromal residue thickness may be an important parameter when predicting and evaluating final visual outcome, but other factors that influence the quality of the graft interface probably play an important role. Further studies are needed to understand the mechanisms that determine graft interface alterations and their correlation with final visual outcomes.

ECD was measured at intervals of 2 months, 6 months, 1 year and 2 years after surgery. The loss of endothelial cells was statistically significant during the 2 months to 6 months interval, with a stabilisation of ECD after 6 months. These results suggest that a minimally invasive procedure such as DALK determines an endothelial cell loss a few months after surgery and does not determine important cell loss, as can be observed in eyes that undergo PKP.

Postoperative complications occurred in 14 eyes (4.9%). The most frequent complication was immunological rejection, which occurred in 11 eyes (3.8%): nine cases of stromal rejection and two cases of epithelial rejection. Patients were treated with topical corticosteroids that led to the reversal of rejection in all cases. The rate of immunological rejection observed in this study is comparable to that reported by other studies; ranging from 3% to 8%.7 9–11 13–15 17 Cortisone glaucoma was observed in one case and was treated with topical intraocular pressure-lowering medications. The single case of herpes virus reactivation was resolved with topical and systemic antiviral treatment. In one case, PKP was performed after 2 years as a result of irregularity of the graft interface.

In conclusion, dry dissection DALK can be a valid alternative to other DALK procedures. A reduction in recipient stromal residue thickness determines better final visual outcomes. However, further studies are needed to evaluate the quality of the graft interface and its correlation with final visual acuity results.


  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by San Raffaele Hospital Ethics Committee, Milan, Italy.

  • Provenance and peer review Not commissioned; externally peer reviewed.


Register for free content

Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of BJO.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.