Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Anterior lamella keratoplasty (ALK) techniques allow the surgeon to selectively dissect parts of the diseased corneal stroma while sparing the healthy remaining stroma and endothelium.
The stromal dissection has to be deep (residual bed thickness <80 μm) and uniform to provide good visual outcomes comparable with penetrating keratoplasty.
Surgical techniques used to be divided into two classes: predescemetic DALK (pdDALK) and descemetic DALK (dDALK); however, a new classification became necessary after the recent discoveries in corneal anatomy about big bubble (BB) type 1 and type 2.
pdDALK was used to indicate manual dissection techniques, which may be considered challenging and time consuming. dDALK was used to refer to techniques that were thought to expose the Descemet membrane (DM), making the surgery faster and more reliable, like with BB and viscodissection.
It has been recently demonstrated that BB type 1 does not separate DM from stroma but is, in fact, an intrastromal bubble, whereas only BB type 2 truly exposes the DM.
A new and more appropriate classification has been proposed: deep anterior lamellar keratoplasty (DALK) for all the manual dissection techniques that are sufficiently deep, subtotal anterior lamellar keratoplasty (STALK) for all the previous dDALK technique where, in fact, a very thin layer of stroma is left behind together with the DM and the endothelium, and total anterior lamellar keratoplasty (TALK) for the cases were the DM is truly exposed.
The existence of a pre-Descemet layer has not yet been confirmed and is still a matter of debate.
The BB technique is the most common technique. The mastership of other effective techniques is useful in cases were BB fails or is not indicated.
Over the past decade, there has been a changing paradigm in corneal surgery. The focus has shifted away from full-thickness grafts, penetrating keratoplasty (PK), and toward anatomically targeted procedures that avoid the removal of healthy corneal tissue and replace or address only the diseased layer.
Anterior lamellar keratoplasty (ALK), either superficial or deep, involves removal of diseased corneal tissue, leaving behind healthy stroma, endothelium, and Descemet membrane (DM) with great advantages over PK, namely the avoidance of endothelial rejection, longer graft survival, the avoidance of an open sky procedure, and a stronger postoperative wound resistance.
The first ALK techniques involved a time-consuming and difficult free-hand dissection that produced interface irregularities, usually resulting in the loss of vision, or only partial improvements. Given the difficulty of achieving good visual outcomes, PK has been the mainstay of corneal transplantation surgery since the mid-1950s, until recently, no matter which part of the cornea was diseased. ,
Advances in technology and techniques, as well as a greater understanding of corneal physiology and optics, facilitated the resumption of lamellar surgery. The key to obtain a good visual outcome and to reduce the donor-host interface is to perform the deepest possible stromectomy, leaving a smooth and uniform-in-thickness recipient surface. , , Despite these principles having already been described in 1971 by José Barraquer, they were not easy to implement and the expression “lamellar keratoplasty” (LK) was used to refer to manual dissection techniques, regardless of the depth of the stromal removal. , , In 1984, Archila used the term “deep lamellar keratoplasty—DLK” (which became deep anterior lamellar keratoplasty [DALK] in 1999 to distinguish anterior from endothelial LK ) for the first time and, after him, many other authors started to use the acronym DLK to differentiate their techniques from all other LKs that were too superficial, emphasizing the need of a deep dissection ; however, this aim was still difficult to be accomplished with manual techniques, and PK remained the first choice for a few more decades.
It is only with the introduction of the viscodissection (Melles et al. in 1999) , and the “big bubble (BB)” (Anwar and Teichmann in 2002) techniques that surgeons started shifting from PK to DALK worldwide. These procedures use a forceful air or viscoelastic injection into the deep stroma to create bubble formation that detaches the DM from the posterior stroma. The surgery becomes faster and, most importantly, more reliable, allowing the surgeon to be confident to have performed an optimal procedure with a good visual prognosis. , ,
Several studies and reviews have shown how the depth and the smoothness of the stromectomy are important in achieving good visual outcome. The classification between “descemetic DALK (dDALK)” and “predescemetic DALK (pdDALK)” introduced in 2010 by Sarnicola et al. has been fundamental in evaluating and comparing visual outcomes between DALK and PK. dDALK indicates cases where all the stroma was thought to be removed and the DM was macroscopically exposed (mainly the viscodissection and the air BB). pdDALK comprises cases where the stromectomy reached the predescemetic plane, described as a very thin stromal layer of almost constant thickness, without the complete exposure of the DM (largely all the manual dissection techniques) ( Table 120.1 ). No significant difference in postoperative best spectacle-corrected visual acuity (BSCVA) between dDALK and PK has been found, but conflicting opinions have arisen regarding dDALK versus pdDALK. , However, in interpreting the literature, it is important to determine how much stroma can be left at the recipient bed to achieve good postoperative visual acuity. Such data are unfortunately missing in numerous studies. Although there is not unanimous agreement, the vast majority of literature shows that visual recovery after pdDALK is slower but comparable with dDALK with long-term follow-up (2–5 years) when the residual recipient bed thickness is less than 80 μm and homogenous.
New Classification | Surgical Techniques | Old Classification |
---|---|---|
DALK |
|
All the previous pdDALK techniques, with a residual bed thickness <80 μm |
Deep Anterior Lamellar Keratoplasty | ||
STALK |
|
Wrongly identified as dDALK |
Sub-Total Anterior Lamellar Keratoplasty | ||
TALK |
|
The only actual dDALK |
Total Anterior Lamellar Keratoplasty |
While the literature seemed to conform the two terms dDALK and pdDALK, the paper published by Dua in September 2013, entitled “Human Corneal Anatomy Redefined. A Novel Pre-Descemet’s Layer (Dua’s Layer),” challenged these traditional tenets. ,
Despite the criticism arising in literature on the need to redefine the corneal anatomy, Dua’s paper has made a contribution to the knowledge on corneal pathophysiology, elucidating the differences between BB type 1 and type 2, and highlighting that only the BB type 2 really exposes the DM. ,
BB type 1 is the most common type. It can be recognized by its well-circumscribed and white margins, its diameter measures up to 8.5 mm, it starts in the center and enlarges circumferentially toward the periphery, and it is quite resistant ( Fig. 120.1A ; ).
Video 120.1 Big Bubble (BB) Type 1 and Type 2.
BB type 1 can be recognized by its well-circumscribed and white margins, its diameter measuring up to 8.5 mm, and the fact that it starts in the center and enlarges circumferentially toward the periphery.
BB type 2 can be recognized by its thin and clear margins, its larger diameter measuring up to 10.5 mm, the fact that it starts in the periphery and enlarges centrally, and it is often eccentric. Vincenzo Sarnicola, Enrica Sarnicola, Caterina Sarnicola.
This type of bubble, which was thought to be descemetic, is actually intrastromal. The posterior wall of the bubble is made by endothelium, DM, and a very thin layer of posterior residual stroma, as first demonstrated by Jafarinasab in 2010. , ,
BB type 2 is rare and more likely to occur in elderly patients with deep corneal scar and thin corneas. It can be recognized by its thin and clear margins, has a larger diameter measuring up to 10.5 mm, and starts in the periphery and enlarges centrally. Moreover, it is often eccentric and it is very fragile (see Fig. 120.1B and ). ,
The existence of 2 types of BB has been known from 2007 13,17 ; however, both types were wrongly classified as dDALK. In 2011, McKee firstly demonstrated that only BB type 2 really cleaves off the DM from the stroma. ,
Big Bubble type 3 comprises a mixed type of bubbles: BB type 1 and one or more smaller type 2 bubbles.
According to Dua et al., the stroma in the posterior wall of the BB type 1 is not just residual stroma but has some unique characteristics ( Table 120.2 ) that make it an unknown layer in the human posterior cornea that he named Dua’s layer (DL) or pre-Descemet layer (PDL) . ,
|
The assignation of a new corneal layer raised considerable controversy in the scientific community regarding its actual existence, especially considering that few papers that followed have confuted some of the hallmarks of the presumed new layer. Even the terminology assigned to the supposedly new layer created disagreement, because the existence of two different types of bubbles, and the presence of residual stroma in the posterior wall of the BB type 1, had already been described before the DL publication. , ,
What is confirmed from a literature review is reported below :
The existence of two different types of bubbles, first described by Anwar, and then confirmed and histologically investigated by McKee ;
The BB type 1 cleavage plane occurs within the most posterior stroma, which means that a thin layer of stroma remains adherent to the DM and endothelium. , , , This was first published by Jafarinasab et al. and Straiko et al , ;
McKee et al. originally demonstrated that only a type 2 BB cleaves off DM and stroma, and it is not an intra-DM bubble as previously thought , ;
Dua et al. were the first to describe a type 3 bubble (mixed bubble) and to describe the behavior of the different types of bubbles as well as their important surgical and pathophysiologic implications , , ;
The existence of the DL (or pre-Descemet layer) has not been confirmed ;
The “dDALK—pdDALK classification” seems to be outdated.
Regardless of the novelty, or not, of the various types of big-bubble, and the validity of the assignation of a new anatomic layer, the presence of residual posterior stroma in BB type 1 has a great relevance. The clinical and functional implications have been identified and well described by Dua et al. The most relevant are the following:
Type 2 BB is very fragile and easy to break and, therefore, requires extreme caution during surgery;
The maximum diameter of type 1 BB does not depend on the extent of the trephination, and it cannot be enlarged more than a 9.00-mm diameter; a continuous forced insufflation may lead to the bursting of the bubble;
An intraoperative rupture in the case of a type 1 BB is usually reparable, and DM does not roll up on itself (as it would by its nature, during a break in a type 2 BB or in a Descemet membrane endothelial keratoplasty [DMEK] intervention);
The presence of the residual stroma explains the greater resistance to trauma of DALK over PK, the resistance to perforation of the descemetocele, and the recurrences in macular dystrophy after BB type 1 DALK. ,
Despite there being no agreement about the existence of a new corneal layer, it is undeniable that there is the presence of a surgical plane. Only a BB type 2 allows a real separation of stroma and DM, and it should be the only technique to be classified as a dDALK.
BB type 1, which is the most frequent and desirable type of BB, has also been classified until now as dDALK based on its macroscopic surgical appearance, but it is not a true DM exposure and should rather be reclassified as a pdDALK. Although accurate on the histologic level, this latter reclassification would not meet the clinical and surgical characteristics used to distinguish the different DALK techniques and their functional results until now. Moreover, proper comparison of future studies with prior published literature would be difficult. On the other hand, preserving the original classification may be simpler, but the terminology would be inaccurate, histologically imprecise, and likely confusing.
A new nomenclature was proposed by Sarnicola et al. in 2019 to overcome these issues, trying to maintain neutrality upon the diatribe of the anatomic dignity of the DL and its name, and trying to respect both the previous classification and the new findings in microscopic anatomy (see Table 120.1 ) :
DALK —Deep Anterior Lamellar Keratoplasty:
It includes all the deep manual dissection techniques (layer by layer, peeling off, hydrodissection, etc.). These have to achieve a uniform residual bed thickness of less than 80 μm. The manual dissection keratoplasties that leave a significant thicker residual bed should not be classified in this category and could be indicated as ALK (anterior lamellar keratoplasties that are not deep or indeed not deep enough). Essentially, this group includes what was previously intended as pdDALK, and the meaning of the acronym DALK reverts to the “pre-big bubble” and “pre-viscodissection” era.
STALK —Subtotal Anterior Lamellar Keratoplasty:
It includes BB type 1, viscodissection, and airviscobubble (AVB). This group encompasses what was called dDALK (except the type 2 BB) but does not, in fact, truly expose the DM as it was thought.
TALK —Total Anterior Lamellar Keratoplasty:
It includes the type 2 BB, the only previous dDALK that actually exposes the DM.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here