In recent years, endothelial lamellar keratoplasty (ELK or DSAEK) has been incorporated into routine practice as a treatment for endothelial pathology in the absence of corneal stromal lesions. The results of the ELK can be studied according to several anatomical or functional aspects. On the other hand, it is interesting to compare the results of this technique with those of other options such as penetrating keratoplasty (PK) or Descemet-endothelial keratoplasty (DEK or DMEK).

GLOBAL RESULTS

Graft survival: Transparent ELK percentages were calculated at 5 years in 92%⁴, at 3 years in 94%⁶, and with follow-ups between 1 month and 3 years in 97.5%⁹.

Functional results: best-corrected visual acuity (BCVA) and induced refractive change. Outstanding means of BCVA published were 0.45 ± 0.39 at 12 months¹, 0.53 ± 0.30 with follow-ups between 1 and 36 months², or 0.63 at 12 months³. The percentage of VA ≥ 0.5 varies between 38 and 100%^4,5. The correlation between graft thickness and BCVA was initially low², but later the same group reported BCVA of 0.80 ± 0.16 at 3 months in 60 cases with the ultrafine ELK technique (99.33 ± 16.97 mm)⁶. ELK induces an average hyperopia of 1 diopter (D), which should be considered in the combined procedures of cataract and ELK. The average induced cylinder is 0.11 D (between -0.4 and +0.6 D).

Complications

• Dislocation of the donor lenticule: It is the most frequent complication and usually occurs in the first 24 hours. Percentages of reformation of the bubble have been published between 0 and 82%^3,4,6,7-9 with an average around 10%. In eyes without capsular support, the frequency doubles⁹.

• Graft rejection: In a 2-year review the percentages range between 0 and 14%⁴. Price et al¹⁰ report a 3-year rejection probability of 9%.

• Primary failure of the graft: Defined as a graft that does not achieve transparency during the first 2 months, its frequency ranges from 0 to 29%^3,4,7,9 with an average of 5%.

• Endothelial cell loss: At 3 years the average percentage of loss is 39-46%^4,6,9. It would be higher if it was necessary to reinject air, with small incisions (3.2 mm) and novice surgeons. Comparative studies of the technique of introduction of the graft do not show significant differences between slider plus forceps and an injector (Endosaver)^7,11.

• Other complications: Intraocular hypertension in the immediate postoperative period due to pupillary blockage by air in the anterior chamber (0.1-9.5%)^4,7,9, epithelial growth in the interphase (0.8-1.6%), infections (0.8-1.5%).

RESULTS ACCORDING TO PATHOLOGIES

Fuchs’ dystrophy is the most frequent indication of ELK in some important series. In general, it has the best results. Graft survival at five years varies from 90% to 96%^12-14, with an immunological rejection rate of 5%, versus 16% in PK due to this indication¹².

Pseudophakic or aphakic corneal edema (Figure 1) leads the causes of secondary endothelial dysfunction and is the most frequent in Asian eyes^14,15. The mean survival of the graft at 1, 2, 3 and 5 years is 94%, 88%, 87% and 76% respectively, significantly lower than in Fuchs' dystrophy (95% at 5 years).

In cases with glaucoma, 5-year survival is especially low in operated eyes, either by trabeculectomy or drainage device (40%), compared to non-operated ones (95%)¹³. In another 20-month study, secondary failure of the graft was also greater in eyes operated on for glaucoma (15.0%) compared to non-operated eyes (3.2%).

The results of the ELK re-grafts seem to be comparable to those of the primary ELK¹⁷. In a study of 113 failed PKs, survival was better in the group treated by ELK (96.2% per year and 86.4% at 5 years) than after a new PK (91.9% and 51.3%). respectively)¹⁸. In cases of PK after CMV endothelitis (with positive aqueous humor PCR), recurrences in the graft occur between 3 and 11 months, while immunological rejection usually occurs between 12 and 18 months¹⁹.

The results in other primary endothelial dysfunctions such as hereditary congenital endothelial dystrophy, posterior polymorphic dystrophy, iridocorneal endothelial syndrome, endothelial failure secondary to phakic lenses or Sato’s radial anterior and posterior keratotomy and in xeroderma pigmentosum, are generally favorable, although they are anecdotal cases or short series.

RESULTS COMPARED WITH PENETRATING KERATOPLASTY

There are no controlled studies comparing ELK and PK. In the available retrospective studies, ELK appears as a procedure with fewer complications and, in the event of occurrence, less serious than in PK^10,20. We find 5 key points:

Economic factors

If one considers the possibility of obtaining pre-cut tissue in the eye bank and obviate the need for the microkeratome, the cost-effectiveness of the ELK would be higher than that of the PK, apart from requiring less postoperative consultations and added procedures²¹.

Factors of the surgical technique

Although mastering ELK requires overcoming a learning curve, a series of advantages inherent in the technique – such as working in a “closed sky”, the possibility of using topical anesthesia, the almost absence of sutures, etc. – make it for many surgeons preferable to PK.

Rejection and survival

Price and Gorovoy¹⁰ presented a non-randomized multicenter trial in 2013, in which graft survival and endothelial loss at 3 years was similar in ELK and PK. The endothelial loss in ELK would be greater than after PK during the first 6 months, but later it stabilizes and in the long run it is lower than in PK. Rejection appears less frequently in ELK and has a better prognosis than in PK²⁵. It is usually asymptomatic, with little clinical and milder symptoms, which is attributed to the prolonged use of corticosteroids in ELK and the absence of complications related to the sutures and healing²⁶. Operating the second eye within the first year after the first did not increase the risk of rejection of the ELK, unlike what is known in PK²⁷.

Complications

Apart from rejection, the main complications of PK result from problems related to penetrating trepanation: need for sutures, scarring and ocular surface problems, such as persistent epithelial defects or neurotrophic ulcers (Figure 2), weakness in trauma, high astigmatism or irregular and late refractive changes. The ELK avoids all these and presents as a specific complication the dislocation of the disc, also related to the learning curve. PK and ELK have in common some complications, such as primary failure, progressive endothelial loss without apparent rejection and infections – with the particularity of the location at the interface in the second –. The risk of massive supra-choroidal hemorrhage is greater in an open globe – as well as its consequences – and the only case reported during an ELK preserved the vision²².

Recovery and visual quality

The published data are clear regarding the superiority of ELK over PK in terms of visual rehabilitation time, although there would be no differences in terms of the final BCVA²³. Post-operative ametropia after PK is usually significant, especially with high astigmatism, compared to the discrete hyperopic change after ELK. The quality of vision, in terms of high-order aberrations, appears superior in ELK²⁴.

RESULTS IN COMPARISON WITH DESCEMETO-ENDOTHELIAL KERATOPLASTY

Despite the abundant literature on ELK, and to a lesser extent, on DEK (DMEK), there are few comparative studies between the two. However, differences have been noted in the following aspects:

Graft detachment

Although it has been suggested that thinner grafts in ELK are more likely to dislocate, this would not be directly applicable to DEK. Although in general there are higher rates of detachment in the DEK, it seems that the main factor is the experience of the surgeon^28,32. When this is experienced, values ​​similar to those of the ELK are reached in some studies³⁰.

Complications and rejection

DEK has shown a higher frequency of air reformations, primary graft failure, and greater tissue losses during its preparation. To limit this last problem, there are eye banks that remit DM-endothelium already dissected^29,34. ELK has shown a lower risk of rejection than PK, but with DEK it would be even lower³⁵.

Endothelial cell loss

As shown by the different studies in each technique, striking falls in endothelial cell density occur in the immediate postoperative period, especially the first 3 months. However, in comparative studies, losses with both techniques are equal at 6 months in 35%^28,30,34.

Visual recovery and refraction

In theory, DEK should have better visual results by not providing a stroma-stroma interface, but the direct apposition of Descemet’s membrane (DM) on the overlying stroma. Studies on ELK have shown smaller and slower visual recoveries than studies on DEK^28,29. Although these differences have been confirmed in comparative studies between the two techniques, VA after ELK continues to improve in the long-term studies (3 years) and approaches the DEK data³⁰. In any case, the results at 1-year favor DEK^28,32,34, and patients who undergo a technique in each eye prefer this technique and would recommend it to their relatives^31,33. DEK generates a minimum hyperopic change, lower than that produced with ELK. Furthermore, with DEK, fewer high-order aberrations are generated³⁶.

Media and training

ELK (DSAEK) requires more means and more expensive than DEK, whether it is invested in material or if the pre-cut tissue is acquired. This makes the second technique applicable in almost anywhere in the world²⁸. In contrast, its results depend to a large extent on the experience of the surgeon, but not in ELK, in which novice surgeons can quickly obtain good results and acceptable complication rates²⁹.