Article Text

Download PDFPDF

Overcoming the technical challenges of deep lamellar keratoplasty
  1. M Yamada
  1. Correspondence to: Masakazu Yamada MD, Division for Vision Research, National Institute of Sensory Organs, National Tokyo Medical Center, 2-5-1 Higashigaoka, Meguro-ku, Tokyo 152-8902, Japan;

Statistics from

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.

With continued improvements in surgical technique it may become the procedure of choice

Lamellar keratoplasty was the first form of corneal transplantation attempted, with a history over a century, and has been regarded mainly as a therapeutic technique.1,2 Lamellar grafting offers several advantages over penetrating keratoplasty, including the elimination of allograft rejection and the avoidance of intraocular complications. In addition, more donor cornea can be used in lamellar keratoplasty since the procedure does not require donor endothelium. This is particularly important in countries where donor corneas are scarce. However, the use of lamellar keratoplasty has been limited by difficulties such as irregularity and scarring of tissue interfaces, leading to poor visual outcomes compared with penetrating keratoplasty,1–3 as well as technical difficulties and prolonged operating time. Penetrating keratoplasty has thus been the most common corneal transplantation procedure for visual restoration for many years. Although penetrating keratoplasty has been shown to be effective and safe for most anterior segment pathologies, there are persistent long term risks such as endothelial failure and immunological graft rejection.4

Deep lamellar keratoplasty (DLK) is a logical step in the surgical management of corneal stromal opacification in the setting of functional endothelium.1 In DLK, pathological stroma is excised down to Descemet’s membrane, and offers the promise of better visual outcomes compared with conventional lamellar grafting. Since this procedure was first reported by Arichila in 1985,5 several large case series have described favourable visual results after DLK.3,6 A report of 120 cases by Sugita and Kondo demonstrated that corrected visual acuity improved by 0.09 to 0.6 on average after DLK.6 Anwar and Teichmann reported that 89% of 181 eyes treated by DLK achieved visual acuity of 20/40 or better.3 More recently, Shimazaki and associates performed a randomised prospective trial of DLK versus penetrating keratoplasty,7 showing that visual function after DLK, as measured by corrected visual acuity, contrast visual acuity, the glare test, and corneal topography, was comparable to that achieved in penetrating keratoplasty. It should be noted that DLK was superior to penetrating keratoplasty in measures of operative morbidity such as continuous endothelial cell loss and intraocular complications.7,8 These results suggest that DLK is a safe alternative to penetrating keratoplasty in eyes without endothelial abnormalities. These promising findings regarding DLK underscore the importance of overcoming technical challenges such as achieving thorough stromal tissue excision without perforation of Descemet’s membrane.

Deep lamellar keratoplasty is a logical step in the surgical management of corneal stromal opacification in the setting of functional endothelium

Most corneal surgeons have confronted the technical challenge of deep lamellar dissection and the attendant risk of puncturing Descemet’s membrane during DLK. To facilitate the dissection of stromal tissue while reducing the risk of perforation, Arichila5 and Price9 employed an air injection technique to separate tissue planes. Sugita and Kondo6 and Amayem and Anwar10 used hydrodelamination to separate the deep stromal fibres from Descemet’s membrane. Manche and associates11 used a hyaluronic acid for viscodelamination. These dissection based procedures appear to be useful both in the identification of stromal layers, and in the separation of stromal layers from Descemet’s membrane. An alternative approach involves improving visualisation of the deep stromal layers during surgery. Balestrzzi et al12 used trypan blue to stain and visualise the posterior stromal fibres. Moore and associates13 attempted to use an ocular endoscope to visualise the posterior corneal surface. However, despite all these efforts, perforation of Descemet’s membrane remains common in DLK. Even experienced DLK surgeons have been reported to demonstrate a perforation rate of as high as 25%.2,6 Sugita and Kondo6 also reported that stromal tissue was incompletely removed in one third of cases. Thus, thoroughly baring Descemet’s membrane without perforation has proved to be a difficult, time consuming, and elusive operative goal for most corneal surgeons.

When employing air, fluid, or viscoelastic injection to dissect Descemet’s membrane from the stromal tissue planes, it is ideal to achieve complete separation of these tissue planes with a single injection. When this is achieved, as described by Anwar and Teichmann,3 a large bubble is formed, indicating perfect separation of tissue planes. However, this result is not always achieved, in large part because the proper depth of dissection is difficult to estimate. Melles and associates14 described a technique for the visualisation of the posterior corneal surface by filling the anterior chamber with air. Through a scleral incision, a deep stromal pocket was created across the cornea, using the mirror image of a 30 gauge needle as a reference for dissection depth. Senoo and associates, in this issue of the BJO (p 1597), describe another approach to determining the proper depth of dissection. A sclerocorneal flap, as is employed during trabeculectomy, is made, and direct microscopic visualisation is used to guide dissection of stromal tissues to the region directly overlying Descemet’s membrane. The continued development of such techniques promises to make DLK easier, safer, and less time consuming.

Trends in keratoplasty have been changing over the past decade. Ocular surface reconstruction, consisting of limbal transplantation combined with amniotic membrane transplantation, has enabled us to improve the management of cicatrising diseases.15,16 Posterior lamellar keratoplasty, also referred to as deep lamellar endothelial keratoplasty, was developed for patients with endothelial dysfunction.2,17 These procedures are based on the concept that only the pathological part of the cornea, such as the epithelium or endothelium, should be replaced by donor tissue, leaving the healthy portion of the host cornea intact. DLK is consistent with this paradigm, and can be viewed as a procedure designed to remove pathological stroma from healthy corneas. With continued improvements in surgical technique, including the advance described by Senoo and associates in this issue, DLK may become the procedure of choice for keratoplasty in most eyes without endothelial abnormalities.

Note in Proof

With continued improvements in surgical technique it may become the procedure of choice



  • The author has no proprietary interest in any materials in this manuscript.

Linked Articles