Anterior segment optical coherence tomography (AS-OCT) is a valuable tool in corneal transplantation surgery. In deep anterior lamellar keratoplasty (DALK), AS-OCT gives valuable information on preoperative decision process. AS-OCT image of the recipient cornea helps the surgeon to decide whether to proceed with a big bubble DALK or a manual dissection. Preoperative AS-OCT imaging of recipient stroma can also predict Descemet membrane perforation during big bubble induction. Intraoperative AS-OCT assures reaching the proper depth in order, and postoperatively, a surgeon can objectively assess the Descemet membrane and donor association and define complications such as a double anterior chamber. Before deciding for a Descemet membrane endothelial keratoplasty (DMEK), preoperative stromal scars can be identified with an AS-OCT and might affect the decision of the surgical procedure (DMEK or penetrating keratoplasty). After DMEK, graft detachment can be visualized by AS-OCT and treated accordingly.
- anterior segment optical coherence tomography
- lamellar keratoplasty
- deep anterior lamellar keratoplasty
- Descemet stripping automated endothelial keratoplasty
- Descemet membrane endothelial keratoplasty
Since lamellar keratoplasty has developed for replacing only a diseased corneal layer, new advantages and challenges showed up [1, 2, 3, 4]. Changing only the pathologic tissue brings closed globe surgery, which is safer than conventional penetrating keratoplasty . It also helps to fasten rehabilitation after surgery. The anterior layer of corneal pathologies may be treated by deep anterior lamellar keratoplasty (DALK), and endothelial dysfunctions can be managed by posterior lamellar procedures such as Descemet stripping automated endothelial keratoplasty (DSAEK) and Descemet membrane endothelial keratoplasty (DMEK) [6, 7, 8]. Each technique requires specific experience and tricks and has their own difficulties and complications. Beside these challenges, all these procedures provide safer surgery, faster rehabilitation, better functional, and anatomical outcomes .
Anterior segment optical coherence tomography (AS-OCT) is a non-contact imaging system that obtains high-resolution images of cornea and anterior chamber [9, 10]. It allows visualizing the shape and depth of corneal lesions such as scarring, measuring central corneal thickness, and lesion size [11, 12]. It can be used before, during, and after lamellar keratoplasty to evaluate the cornea . In this chapter, we discuss the role of AS-OCT on the basis of different corneal lamellar procedures such as DALK, DSAEK, and DMEK.
2. OCT in deep anterior lamellar keratoplasty
DALK has been introduced as an alternative to penetrating keratoplasty (PK) for the diseases that affect anterior layers of cornea [14, 15]. Preserving the endothelium in DALK obtains reduced graft rejections [16, 17]. Many techniques have been described to separate corneal stroma and Descemet’s membrane (DM). Peeling stroma up to near DM is called manual dissection [18, 19]. This procedure can be used as the first choice in some situations such as deep scarring, or the second choice intraoperatively when it cannot be progressed with other techniques [20, 21, 22]. Other option is the separation of stroma and DM with injection of air, fluid, or viscoelastic into deep stroma [14, 23, 24, 25].
Anterior lamellar keratoplasty was described first by Gasset in keratoconus patients which was named as conectomy and included transplantation of DM stripped full thickness graft . Archila et al. introduced deep lamellar keratoplasty, using intrastromal air to separate stroma from DM . After that, different techniques have been described, and today, the most frequently used procedure is “Big Bubble” that Anwar and Teichmann reported . The technique includes 27-gauge needle with air-filled syringe inserted into deep stroma for 3–4 mm centrally and air injection to separate DM from stroma. It has been shown that outcomes of DALK were similar to PK with lower rejection or failure rates. However, Big Bubble can be obtained in only 56–82% of eyes [28, 29, 30].
For the improvement of these surgical techniques, new and unique complications beside advantages have emerged. The main purpose during separation of stroma from DM is to obtain a non-damaged DM . But perforation or tears in DM occur in 4–20% of cases which is the most important complication during surgery [28, 29, 30, 32]. The risk of perforation is higher if proper big bubble formation is not achieved . Unsuccessful big bubble formation is more frequently seen in patients with keratoconus and deep corneal scar involving DM [34, 35, 36]. Microperforations can be managed with air injection to the anterior chamber and careful manual dissection, but most of the macroperforations need to be converted to PK [14, 37]. The other complication is pseudoanterior chamber . It can occur as double or triple anterior chamber [39, 40]. It usually develops after perforation in DM . Remnant viscoelastic material may also cause double anterior chamber . Pseudoanterior chamber may resolve spontaneously when it is shallow but larger chambers should need surgical intervention .
The role of AS-OCT in DALK begins at a preoperative assessment. Patient selection and the right indication for surgery can be decided with many variables including AS-OCT findings. AS-OCT images provide central corneal thickness, anterior chamber biometry, angle and iris measurements, and lens thickness . These findings help planning trephination depth during surgery, intraocular lens power assessment if needed combined with phacoemulsification and intraocular lens implantation, detecting narrow angle before surgery . AS-OCT can also visualize corneal pathologies such as scarring in central or periphery, corneal degenerations or dystrophies and cysts. The importance of the imaging is to inform the surgeon about the location, size, and depth of the corneal lesion.
Big bubble procedure in DALK requires well-adjusted trephination depth and centrally inserted air cannula without damaging DM. This technique needs experience and has a steep learning curve. One of the difficulties is reaching the intended depth of the cornea with cannula and advancement to the central stroma. To overcome this challenge, preoperative central and peripheral corneal thickness measurement with AS-OCT is sensible. Busin et al. reported a modified big bubble technique, which needs AS-OCT . They evaluated the thickness of trephination size with AS-OCT and aimed to reach a depth within 50 μm from the internal corneal surface. Air cannula was advanced only 1 mm into the cornea and air injected. They reported a successful pneumatic dissection in 85% of patients with this procedure. AS-OCT was very useful to show the depth of trephination site, and perforation during trephination occurred in only 2.3% (two patients) of patients. Moreover, AS-OCT has superiority over topography with Scheimpflug camera while planning the depth of trephination for DALK. Riss et al. measured corneal thickness with Scheimpflug camera, planned the depth of trephination to 90% of the thinnest pachymetric value, and faced perforation during trephination in 30.1% of patients . Intraoperative AS-OCT seems to increase the choice of DALK rather than PK with easing lack of experience and surgical difficulties.
It is known that corneal stromal scar has effect on the success rate of DALK. Big bubble can be achieved in 56–82% of keratoconus patients without a corneal stromal scar. The success rate varies from keratoconus grade, central corneal thickness, and trephination size in these patients [28, 29, 46]. Corneal stromal scar is a very important risk factor in keratoconus patients and other corneal diseases influencing DALK success rate. Ozmen et al. reported that big bubble formation was achieved in 63.2% of patients and DALK was completed in 91% of patients . The most important factor affecting DALK and big bubble success was the ratio of corneal scar depth and central corneal thickness (scar depth/central corneal thickness) which was calculated with AS-OCT, a cut-off point was given as >53% for scar depth/central corneal thickness, which has an acceptable sensitivity (100%) and specificity (67%). AS-OCT has been evaluated as crucial for measuring this risk factor and planning the right operation for patients. Corneal scar in keratoconus patients may be caused by hydrops. DALK has a relative contraindication for patients with stromal scar . But considering rejection rates after PK, DALK can be an option when carried with great attention in patients with stromal scar. To avoid macroperforation during DALK, preoperative assessment and planning should be proper (Figure 1). Nanavaty et al. reported that all of their patients had successful DALK and none required PK even though microperforation occurred in 60% of patients, with the help of AS-OCT at preoperative evaluation .
Recent improvements in surgical devices allow us to use AS-OCT at intraoperative evaluation. Several studies point out that visual acuity after DALK is better when stromal dissection is carried to the DM or when there is minimal residual stroma [20, 49, 50, 51]. AS-OCT provides intraoperative assessment of the cornea with high resolution. AS-OCT can be used to measure the depth of air cannula while using big bubble technique and may increase the rate of a successful big bubble formation. Scorcia et al. reported the first study in 2013 using intraoperative AS-OCT to visualize air cannula depth and big bubble formation . It was emphasized that if the cannula reached the optimal depth, big bubble could be obtained more probably, but bubble achievement reduced when air was injected superficially. Successful big bubble in 90% of patients was obtained when the cannula reached within 100 μm from the DM. De Benito-Llopis et al. reported similar results and indicated that AS-OCT is also useful when big bubble could not be achieved and manual dissection was performed . AS-OCT images can be used to evaluate a residual stromal bed during the manual dissection of corneal stroma.
Real-time intraoperative AS-OCT device provides imaging of all surgical steps continuously. Steven et al. reported a case series with using intraoperative AS-OCT. They succeeded monitoring big bubble formation in two of six patients . They performed a micro-bubble incision technique in a patient that they could not obtain big bubble and documented with AS-OCT. Half of the patients was complicated with DM rupture. They indicated that the accurate measurement of trephination depth and cannula advancement into the deep central stroma is possible with real-time intraoperative AS-OCT. Sharma et al. published a case report of a patient who has undergone DALK. The procedure was complicated with DM detachment . They used continuous intraoperative AS-OCT while intracameral sulfur hexafluoride injection and visualized re-attachment of DM with real-time images. Chaniyara et al. reported a complicated DALK case with repair of graft dehiscence with descemetopexy under the guidance of continuous intraoperative AS-OCT . The use of real-time intraoperative AS-OCT seems to increase.
Intraoperative AS-OCT may obtain the depth and central advancement of the cannula while performing surgery. Image distortion caused by metallic objects is the major disadvantage of intraoperative AS-OCT . The depth of the cannula could be assessed with only the image of the tunnel that is created by the cannula. With the manufacturing of plastic needles instead of metallic ones, one may overcome this drawback. Another disadvantage of intraoperative AS-OCT is caused by the devices which are mounted on the operation microscope separately. These devices have a handicap that they need to interrupt the surgery to take AS-OCT images. This disadvantage causes a delay in surgery. New devices that could obtain real-time images that are integrated into the operating microscope are being developed.
Pseudoanterior chamber due to DM detachment is frequently associated with microperforations during surgery, and management includes observation, intracameral gas injection, or reoperation (Figure 2). Slit-lamp biomicroscopy is enough to detect clinically significant detachment, but minimal double anterior chamber may not be seen by examination. The decision of treatment type and follow-up requires documentation of the patient. AS-OCT is very helpful to detect DM detachment and follow-up of the patient . It allows to measure the size of pseudoanterior chamber initially and after the treatment . AS-OCT can also visualize Dua’s layer in case of triple anterior chamber occurrence .
AS-OCT can evaluate other postoperative complications of DALK. Bahadir et al. reported a candida interface keratitis after DALK, showing the keratitis site at AS-OCT . Mukhopadhyay et al. reported a rhinosporidiosis keratitis case and they performed a repeat DALK procedure successfully after determining the site of the infiltrate in corneal stroma with AS-OCT . Muller et al. published a case report with granular corneal dystrophy type-1 and reported that although they achieved pneumatic dissection to predescemetic layer during DALK, AS-OCT showed granular opacities between DM and graft . It is thought that granular deposition occurs at predescemetic layer in granular corneal dystrophy type-1 and PK could be a better alternative than DALK. Costa et al. reported an iris cyst seen after DALK which was documented with AS-OCT and anterior segment photography .
3. OCT in posterior lamellar keratoplasty
Posterior lamellar keratoplasty has emerged as an alternative to PK to treat corneal endothelial pathology [63, 64]. Evolution of the posterior lamellar keratoplasty procedures began with Melles et al., who reported deep lamellar endothelial keratoplasty (DLEK) that includes sutureless transplantation of posterior lamellae of cornea from a scleral incision . Following DLEK, Descemet’s stripping endothelial keratoplasty (DSEK) and DSAEK were developed [3, 66]. Corneal graft is prepared with automated microkeratome in DSAEK and it has been used worldwide . Recently, DMEK is introduced as the latest posterior lamellar keratoplasty technique by Melles et al. . Today, DSAEK and DMEK are the most popular posterior lamellar keratoplasty procedures .
DSAEK and DMEK are completely different interventions and both have different maneuvers . DSAEK needs a 4- or a 5-mm limbal or corneo-scleral incision to insert the donor lamella. Descemet stripping of the host cornea is performed over an 8-mm diameter circle with a reverse Sinskey hook. Preparation of donor cornea is performed with automated microkeratome. Donor button consists of a thin stroma and Descemet’s endothelial complex. Femtosecond laser has also been used recently for the preparation of the DSAEK graft. After trephination of donor lamella, it is inserted into the anterior chamber. Graft is attached to the host cornea with an air bubble. By contrast, DMEK provides transplantation of the endothelium with DM layer with minimal or absent stroma . DM is stripped from the posterior stroma . There are two methods of stripping: the first described procedure consists of DM stripping after corneal button trephination and the other technique consists of scoring edge of DM and stripping away from the stroma nearly half way to the center for 360° before trephination. Recently, DM dissection can be performed by an automated microkeratome and called as Descemet’s membrane automated endothelial keratoplasty (DMAEK) [72, 73].
Several studies and meta-analysis reported that DMEK has superiority over DSAEK, considering visual outcome and patient satisfaction but DSAEK is used more frequently because of the steeper learning curve of DMEK [1, 74, 75, 76, 77]. DMEK and DSAEK have a similar complication profile [78, 79, 80]. The most common early complications following posterior lamellar keratoplasty are graft dislocation and primary graft failure . Other complications are graft rejection, endothelial cell loss, iatrogenic pupillary block glaucoma, keratitis, and endophthalmitis.
AS-OCT is a useful tool for posterior lamellar keratoplasty. It is valuable for preoperative assessment, intraoperative maneuvers, and postoperative follow-up. The preoperative duration of corneal stromal edema is found to be an important factor for visual outcome in patients who underwent endothelial keratoplasty . It is known that the long duration of corneal stromal edema causes an increase in fibroblastic activity, irreversible fibrotic changes, and corneal scarring [83, 84, 85]. PK can be chosen in patients who have bullous keratopathy more than 12 months, instead of endothelial keratoplasty [82, 86]. AS-OCT may detect these stromal changes and obtain objective data for the decision of a planned surgery (Figure 3). AS-OCT images may help to decide which patients with stromal edema more than 12 months will benefit from endothelial surgery. AS-OCT can also obtain fine images of corneal epithelium. Long-standing bullous keratopathy may also cause subepithelial haze and fibrosis . Agarwal et al. reported that endothelial keratoplasty with epithelial debridement provides a better visual outcome in patients with chronic stromal edema . AS-OCT was used to show corneal layers before and after endothelial keratoplasty which indicates that subepithelial haze and opacification of cornea decreased with surgery (Figure 4).
DMEK needs special maneuvers to unfold the graft and attach to the host. Attachment of the graft in both DMEK and DSAEK was performed with an air bubble, and there should be no interface fluid. Intraoperative AS-OCT can evaluate the graft orientation and obtain images of interface. Ide et al. reported the successful use of intraoperative AS-OCT in six consecutive patients showing no interface fluid at the end of the DSAEK . Knecht et al. reported that their intraoperative AS-OCT study with six cases underwent DSAEK and indicated that detectable interface space does not mean a failure because the fluid may regress 1 day after surgery . They also reported that detectable interface fluid may disappear with vent incisions by the guidance of intraoperative AS-OCT. In addition to that, PIONEER study published two separate results from DSAEK surgery with intraoperative AS-OCT. The first study indicated that transient interface fluid at the end of the surgery and first postoperative day is more likely to develop textural interface opacity . The second study reported that a larger residual interface fluid volume, area, and thickness at the end of surgery are associated with early-graft detachment . These results emphasize that minimal or no interface fluid should be left at the end of the DSAEK surgery.
Initial studies were performed with handheld AS-OCT devices, and imaging caused interruption of the surgery. Initial publications were followed by studies with real-time intraoperative AS-OCT which is integrated with the operating microscope [93, 94, 95, 96]. Real-time imaging provided more comfortable surgery with simultaneous imaging. Besides evaluating interface fluid during surgery, intraoperative AS-OCT has additional benefits. Intraoperative AS-OCT assists in obtaining images of the nearly opaque corneas caused by stromal edema which could not be assessed well with an operation microscope . This helps the surgeon to manipulate the graft easily in the operation.
DMEK is a relatively new technique than DSAEK, and studies on DMEK with intraoperative AS-OCT are fewer. Steven et al. published a study with 26 patients who underwent DMEK and reported that the usage of intraoperative AS-OCT enhances the graft visibility and surgeon’s orientation . Saad et al. reported that intraoperative AS-OCT enables a faster graft positioning with less manipulation in DMEK . DISCOVER study indicated that intraoperative AS-OCT is very useful to confirm graft orientation and to reduce the iatrogenic graft failure .
Postoperative follow-up for graft dehiscence and graft failure is crucial in posterior lamellar keratoplasty. The fluid between the cornea and the graft can be seen with a slit-lamp biomicroscopy but Tarnawska and Wylegala reported that half of the interface fluid cannot be seen with examination which was detected by AS-OCT . Early detection of interface fluid and graft detachment provides early intervention. In addition to that, postoperative measurement of graft thickness is an important factor for graft failure in DSAEK patients . Eyes with thick grafts are more prone to graft failure (Figure 5). It is important to provide improvement in visual acuity after endothelial keratoplasty but fewer patients achieve excellent quality of vision. Many reasons were suggested including anterior stromal changes, graft-related problems, induced high-order aberrations, and interface-related problems. Turnbull et al. commented that the evolving anterior segment imaging could increase our understanding of the reason to the solution of these difficulties . Suh et al. reported that the evaluation of epithelial ingrowth in DSAEK patient can be possible with AS-OCT . Kymionis et al. reported a case with residual DM after DSAEK that was shown with AS-OCT . Lopez and Melles et al. published a study that described rebubbling techniques in DMEK, and they showed detached and folded DM with AS-OCT . The assistance of AS-OCT for the assessment and management of complications in DMEK was studied in many publications [104, 105, 106, 107].
AS-OCT is a valuable tool for assessing lamellar keratoplasty in all steps of the surgery. The use of AS-OCT starts from patient selection to postoperative late complications. Both anterior and posterior lamellar procedures need proper patient selection to obtain desired outcomes. Imaging the graft-host relationship with AS-OCT allows proper assessment of interface and helps surgeon to perform fast and successful surgeries. Some complications such as graft detachment can be visualized with AS-OCT postoperatively, more accurate than a conventional slit-lamp examination. AS-OCT is currently being used widely in corneal surgeries, and its role in lamellar keratoplasty seems to be increasing.
The authors thank the technicians of anterior segment imaging, Gazi University Medical School Department of Ophthalmology (Aynur Kartal, Cengiz Aksel, Nilgun Taşçı, Suat Avcı, Şenay Çay, Zafer Yıldırım), for the excellent photographs that have been used in this chapter.
Conflict of interest
None of the authors have any financial interest in any material or method mentioned in this chapter.