Article Text

Download PDFPDF

“Tuck In” Lamellar Keratoplasty (TILK) for corneal ectasias involving corneal periphery
  1. S Kaushal1,
  2. V Jhanji1,
  3. N Sharma1,
  4. R Tandon1,
  5. J S Titiyal1,
  6. R B Vajpayee1,2
  1. 1
    Rajendra Prasad Centre for Ophthalmic Sciences, All India Institute of Medical Sciences, New Delhi, India
  2. 2
    Centre for Eye Research Australia, University of Melbourne, Melbourne, Australia
  1. R B Vajpayee, Corneal & Cataract Surgery, Centre for Eye Research Australia, University of Melbourne, 32 Gisborne Street, East Melbourne, Victoria 3002, Australia; rasikv{at}


  • Funding: None.

  • Competing interests: None.

View Full Text

Statistics from

Corneal ectatic dystrophies with peripheral thinning are difficult to treat as the routine techniques of contact lens fitting and corneal transplantation often do not achieve optimum visual results. These conditions include keratoglobus with generalised corneal thinning and pellucid marginal degeneration (PMD) with corneal thinning involving the periphery of the cornea within 1–2 mm from the limbus.1 This extreme thinness of the cornea involving the periphery makes the surgical management of these cases challenging. A standard keratoplasty is difficult and usually results in suboptimal optical results.2 Other surgical techniques which have been tried in these cases include large diameter or eccentric penetrating keratoplasty, epikeratoplasty, large-diameter lamellar keratoplasty, corneal wedge resection, crescentic lamellar grafts and corneoscleroplasty.26 We had previously reported a single-stage surgical technique of central lamellar graft with peripheral intralamellar tuck in a patient of keratoglobus.7 In the present study, we report our experience with this technique, which has now been referred to as “Tuck In” Lamellar Keratoplasty (TILK), in 12 cases of corneal ectasias involving both central and peripheral parts of cornea.


Twelve eyes of 12 consecutive patients with central corneal ectasias and peripheral corneal involvement requiring surgical intervention were selected from the Cornea services of our centre. These included eight patients of combined keratoconus and PMD and four patients of keratoglobus. Informed consent was obtained from the subjects, and Institutional Review Board approval was obtained. The cases of keratoconus and pellucid marginal degeneration were diagnosed with the help of clinical history, slit-lamp evaluation, videokeratography findings and ultrasonic pachymetry findings. Rabinowitz criteria were used for the initial diagnosis of keratoconus. The severity of keratoconus and PMD was evaluated based on keratometry values. All patients were contact lens intolerant and had best corrected visual acuity (BCVA) ⩽20/120 with nine patients (75%) having BCVA ⩽20/200. At initial presentation, a detailed ocular examination was performed, which included assessment of visual acuity, refraction, keratometry, nine-point ultrasonic pachymetry (Sonogage, Cleveland, OH), videokeratography (Bausch & Lomb, Rochester, NY), endothelial cell count (Topcon SP 2000P), applanation tonometry and a dilated fundus examination. The intraocular pressure was adjusted according to the corneal thickness. All the refractions were performed by an experienced optometrist both preoperatively and postoperatively, and the best possible subjective acceptance was recorded. Cases with acute corneal hydrops were excluded.

Surgical technique of TILK

Technique for keratoglobus

Preparation of the host bed

All patients were operated on under general anaesthesia. The vertical recti were bridled, the corneal surface thoroughly dried and the pupillary centre marked.

An 8.5 mm Hessburg Barron vacuum trephine (Jedmed Instruments, St. Louis, MO) was used to make an initial partial thickness groove of 180–240 μm, depending on the corneal thickness, and a centripetal dissection was carried out using lamellar dissectors to excise a central anterior stromal disc of 8.5 mm. Subsequently, a centrifugal lamellar dissection was performed using a crescent knife (Alcon Surgical, Fort Worth, TX) to create a peripheral intrastromal pocket. The pocket was made circumferentially in the corneal periphery up to 0.5 mm posterior to the limbus (fig 1).

Figure 1 Schematic diagram depicting TILK with 360° tuck for patients with keratoglobus.
Preparation of graft

A corneoscleral donor button was fixed in an artificial chamber (Moria, Antony, France) and an initial partial thickness incision up to a depth of 300 μm was made using an 8.5 mm Hessburg Barron vacuum trephine (Jedmed Instruments, St. Louis, MO). This was followed by a lamellar dissection peripheral to the trephine cut that was extended up to the limbus circumferentially. The superficial corneal tissue extending from the trephine mark to the limbus was excised leaving a central 8.5 mm full-thickness graft with a peripheral partial thickness flange of about 2.5–3 mm. The tissue thus prepared was punched from the endothelial side with trephines ranging from 11 to 12 mm in diameter. The descemet’s membrane of the donor lenticule was stripped after staining with 0.1 ml of 0.06% Trypan Blue (Visiblue, Shah & Shah, India).

After the donor graft had been prepared, a paracentesis was performed in the host’s eye to decompress the eye, making it easier for the intrastromal tucking of the flange. The flange of the graft was tucked into the peripheral intrastromal pocket of the host created previously, and 16 interrupted sutures with 10–0 nylon were applied to secure the graft using a sharp point needle (Alcon Surgical, Fortworth, TX). The needle of the suture was passed through the partial thickness through the graft and then into the peripheral pocket up to the limbus (fig 1).

Modified technique for keratoconus with PMD

Preparation of the host bed

In the presence of only inferior thinning in cases of PMD with keratoconus, the technique of TILK was modified, and only an inferior 180° peripheral intrastromal pocket was created instead of a circumferential pocket.

Preparation of graft

Similar to the technique used for cases of keratoglobus, an artificial chamber was used to prepare a 12 mm graft consisting of a central 8.5 mm full-thickness part and a peripheral partial thickness flange. The flange was, however, retained only in inferior 180° area, and the rest of it was excised using Vannas scissors (fig 2). The inferior half of the graft with the flange was tucked into the peripheral intrastromal pocket of the host and graft sutured with 16 interrupted sutures with 10–0 nylon. The half flange of the graft was tucked into the peripheral intrastromal pocket of the host and graft sutured with 16 interrupted sutures with 10–0 nylon using a long-point needle (Alcon Surgical, Fortworth, TX). The suture was passed through the partial thickness of the cornea and then into the peripheral pocket up to the limbus in the inferior 180°. In the superior half, an end-to-end suturing was done (fig 2). Postoperatively, all patients received predinisolone acetate 1% eye-drops four times a day, moxifloxacin hydrochloride 0.5% eye-drops three times a day and preservative free artificial tears four times a day.

Figure 2 Schematic diagram depicting TILK with inferior 180° tuck for patients with keratoconus with PMD.

The topical therapy consisting of steroids and antibiotics was tapered over a period of 3 months, and preservative-free artificial tears were continued for 6 months.

Patients were planned for a follow-up ocular examination weekly until 1 month, monthly until 6 months and 6-monthly thereafter (figs 3, 4). At each follow-up visit, a complete ocular examination was performed including slit-lamp biomicroscopy and videokeratography. The parameters recorded were UCVA, BCVA, keratometry, refraction, intraocular pressure and endothelial cell counts. Selective suture removal was done as and when required based on the videokeratography. The parameters were analysed with a Wilcoxon signed rank sum test, and the level of significance was taken as p<0.05.

Figure 3 Slit-lamp photograph of a patient with keratoglobus. (A) Preoperative diffuse illumination. (B) Postoperative diffuse illumination.
Figure 4 Preoperative and postoperative profile view of a patient with keratoglobus. (A) Preoperative. (B) Postoperative.


The mean age of the patients was 37 (SD 14.17) years. There were no intraopertaive complications in any of the cases. Postoperative course was uneventful for all patients except cases 6 and 8, who showed delayed epithelial healing (10 days each). The epithelial defects were measured on slit-lamp examination after staining with 1% flourescein dye.8 9 The mean time for epithelial healing was 5.25 (2.7) days. The mean follow-up period at the last postoperative visit was 1.7 years (range 1–4 years) (table 1).

Table 1 Patient parameters preoperatively and at last follow-up visit

All the grafts were clear at the last follow-up visit. There was no significant corneal interface haze or corneal vascularisation seen in any of the eyes. The mean intraocular pressure at last follow-up was 15.1 (SD 1.97) mm Hg as compared with the preoperative value of 14.5 (2.12) mm Hg (p = 0.435) [table 2].

Table 2 Comparison of mean patient parameters preoperative and at last follow-up

Preoperatively, all patients had BCVA ⩽20/120, with nine patients (75%) having BCVA ⩽20/200 (fig 5). Following TILK, all patients showed an improvement in BCVA. At last follow-up, six patients (50.0%) had BCVA of 20/60 or better, and all patients had BCVA better than 20/80 (fig 5). The surgical procedure was successful in correcting the corneal ectasia as suggested by the decrease in the average keratometry from a preoperative value of 57.54 (SD 6.89) D to 44.2 (3.9) D at 1 week (p = 0.011). This decrease in keratometry was maintained until the last follow-up (46.36 (2.39) D; p = 0.003) (table 2). This was associated with a decrease in the SEQ refractive error from a mean of −7.8 (4.6) D preoperatively to −1.23 (1.88) D at last follow-up (p = 0.007) (table 2). The refractive astigmatism also showed a gradual decline from a mean preoperative value of 5.93 (3.06) D to 3.23 (1.14) D at last follow-up (p = 0.037) (table 2). The mean preoperative endothelial cell count was 2055.56 (451.14) cells/mm2. At last follow-up, the mean postoperative endothelial cell count was 2003.10 (439.89) cells/mm2 (p = 0.008) (table 2).

Figure 5 Bar diagram showing the gain in postoperative best corrected visual acuity in patients at the last follow-up visit after TILK.


Ectatic corneal dystrophies like keratoconus can be successfully managed by routine techniques of corneal transplantation like PK or DALK.1012 On the other hand, conditions like keratoglobus and keratoconus with PMD may not be amenable to these standard surgical modalities. In these cases, routine penetrating keratoplasty is technically difficult and is associated with poor visual results. This is due to the discrepancy between host and donor corneal thickness, and progressive thinning of the diseased corneal periphery after the surgery. Large-diameter penetrating keratoplasty has been tried for keratoglobus and PMD but is associated with a high rate of graft rejection. Speaker et al reported a rejection rate of 100% with the use of large-diameter (⩾9.5 mm) penetrating grafts compared with 36% in smaller grafts (8.5/9 mm).13 These eyes are also prone to developing secondary glaucoma due to damage to the angle structures. Various lamellar surgical procedures have been described for corneal ectatic dystrophies with peripheral corneal involvement. These include epikeratoplasty, large-diameter lamellar keratoplasty, crescentic lamellar keratoplasty, corneoscleroplasty and peripheral tectonic corneoscleral ring.24 1416 These techniques have been known to be associated with problems of epithelial healing, as there may be destruction of limbal stem cells during surgery.

Corneoscleroplasty, which is a combination of a central penetrating graft and a large-diameter peripheral lamellar graft, has been described by Burk and Joussen for a case of keratoglobus.3 The authors used systemic immunosuppression to reduce the risk of graft rejection, which entails an added risk of systemic toxicity and increased costs involved.

Jones and Kirkness have described a two-stage surgical technique for keratoglobus wherein a large tectonic lamellar graft is done followed by central penetrating keratoplasty in the second stage.17

Schanzlin et al have described a new surgical procedure called Intralamellar Keratoplasty (ILK) wherein a femtosecond laser is used to create an intracorneal lamellar channel to insert a donor stromal lenticule.19 The technique has been used in cases of keratoconus and postlasik ectasia with clear central corneas.

In order to provide adequate tectonic support to central as well as peripheral cornea while avoiding damage to limbal stem cells, we devised the technique of TILK. In this technique, the donor lenticule has a full-thickness central part and a peripheral flange of partial thickness posterior stromal tissue.

The central 8.5 mm part of the full thickness graft provides tectonic support to the central ectatic cornea while the thin peripheral flange tucked into the intrastromal pocket integrates into the host and provides tectonic support to the peripheral cornea. Since dissection of the superficial limbal region is avoided, there is no damage to the recipient’s limbal stem cells.

While PMD alone can be suitably treated with a crescentic patch graft, in cases that have a simultaneous presence of keratoconus, a different approach is required. Eccentric penetrating grafts have been tried to include the thinned out area near the inferior limbus. These have been associated with a high rate of graft rejection and high astigmatism.18

Crescentic peripheral lamellar graft with or without a subsequent central penetrating graft has also been tried for surgical correction of these cases.19 20 The success of this procedure has not been established, as there are at present no published large case series. Rasheed and Rabinowitz have reported the use of crescentic lamellar graft with central penetrating keratoplasty as a single-stage procedure in five patients with PMD.20 They have shown improved visual acuity and reduced astigmatism postoperatively which was stable during the study period. Corneal wedge resection with or without relaxing incisions has been routinely used to treat high postkeratoplasty astigmatism. Dubroff, MacLean et al and Cameron have described the use of this technique to treat astigmatism associated with PMD.5 15 16 Since the technique does not provide any tectonic support to the pathological cornea, the effect of the procedure may be short-lived.16

We devised our technique of TILK to treat cases of PMD associated with keratoconus. In this technique, the donor button had a peripheral flange only in the inferior half, and the tucking was done only in the involved inferior cornea. As is evident from the results, this modified technique of TILK was successful in all our cases of keratoconus with PMD.

It is possible that with the passage of time, development of interface haze as seen in most cases of lamellar keratoplasties may reduce vision. In the event of such a complication, a subsequent central penetrating keratoplasty may be performed to achieve optimum visual acuity. In future, the use of newer technologies such as a femtosecond laser for smooth lamellar dissection and harvesting of precise donor lenticules similar to intralamellar keratoplasty6 may help eliminate such a complication.


View Abstract


  • Funding: None.

  • Competing interests: None.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • At a glance
    Harminder S Dua Arun D Singh, Editors-in-Chief