Endothelial Keratoplasty - CAM 90322

Description
Endothelial keratoplasty also referred to as posterior lamellar keratoplasty, is a form of corneal transplantation in which the diseased inner layer of the cornea, the endothelium, is replaced with healthy donor tissue. Specific techniques include Descemet stripping endothelial keratoplasty, Descemet stripping automated endothelial keratoplasty, Descemet membrane endothelial keratoplasty, and Descemet membrane automated endothelial keratoplasty. Endothelial keratoplasty, and particularly Descemet stripping endothelial keratoplasty, Descemet stripping automated endothelial keratoplasty, Descemet membrane endothelial keratoplasty, and Descemet membrane automated endothelial keratoplasty, are becoming standard procedures. Femtosecond laser-assisted endothelial keratoplasty and femtosecond and excimer laser-assisted endothelial keratoplasty have also been reported as alternatives to prepare the donor endothelium.

For individuals who have endothelial disease of the cornea who receive Descemet stripping endothelial keratoplasty or Descemet stripping automated endothelial keratoplasty, the evidence includes a number of cohort studies, a randomized controlled trial (RCT), and systematic reviews. Relevant outcomes are change in disease status, morbid events, and functional outcomes. The available literature has indicated that these procedures improve visual outcomes and reduce serious complications associated with penetrating keratoplasty. Specifically, visual recovery occurs much earlier. Because endothelial keratoplasty maintains an intact globe without a sutured donor cornea, astigmatism or the risk of severe, sight-threatening complications such as expulsive suprachoroidal hemorrhage and postoperative catastrophic wound failure are eliminated. The Descemet Endothelial Thickness Comparison Trial (DETECT) RCT reported improved visual acuity outcomes with Descemet membrane endothelial keratoplasty compared to ultra-thin Descemet stripping automated endothelial keratoplasty. The evidence is sufficient to determine that the technology results in a meaningful improvement in the net health outcome.

For individuals who have endothelial disease of the cornea who receive Descemet membrane endothelial keratoplasty or Descemet membrane automated endothelial keratoplasty, the evidence includes a number of cohort studies and systematic reviews. Relevant outcomes are change in disease status, morbid events, and functional outcomes. Evidence from the cohort studies and meta-analyses has consistently shown that the use of Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty procedures improve visual acuity. When compared with Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty, Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty showed significantly greater improvements in visual acuity, both in the short term and through 1 year of follow-up. The evidence is sufficient to determine that the technology results in a meaningful improvement in the net health outcome.

For individuals who have endothelial disease of the cornea who receive femtosecond laser-assisted endothelial keratoplasty and femtosecond and excimer laser-assisted endothelial keratoplasty, the evidence includes a multicenter RCT comparing femtosecond laser-assisted endothelial keratoplasty with penetrating keratoplasty and an RCT comparing femtosecond-prepared Descemet stripping automated endothelial keratoplasty to microkeratome-prepared Descemet membrane automated endothelial keratoplasty. Relevant outcomes are change in disease status, morbid events, and functional outcomes. Mean best-corrected visual acuity was worse after femtosecond laser-assisted endothelial keratoplasty than after penetrating keratoplasty, and endothelial cell loss was higher with femtosecond laser-assisted endothelial keratoplasty. With the exception of dislocation and need for repositioning of the femtosecond laser-assisted endothelial keratoplasty, the percentage of complications was similar between groups. Complications in the femtosecond laser-assisted endothelial keratoplasty group were due to pupillary block, graft failure, epithelial ingrowth, and elevated intraocular pressure, whereas complications in the penetrating keratoplasty group were related to sutures and elevated intraocular pressure. Worsened visual acuity and a 100% graft dislocation rate was reported for femtosecond-prepared Descemet stripping automated endothelial keratoplasty compared to 0% in manually-prepared Descemet stripping automated endothelial keratoplasty. The evidence is insufficient to determine the effects of the technology on health outcomes.

Clinical input obtained in 2009 supported Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty as the standard of care for endothelial failure, based on improved outcomes compared with penetrating keratoplasty.

Clinical input obtained in 2013 on evolving techniques for endothelial keratoplasty uniformly considered Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty to be medically necessary procedures.

Most input obtained in 2013 considered femtosecond laser-assisted endothelial keratoplasty and femtosecond and excimer laser-assisted endothelial keratoplasty to be investigational.

Input obtained in 2013 was mixed on the exclusion of patients with anterior corneal disease. Additional indications suggested by the reviewers were added as medically necessary. 

Background
Corneal Disease
The cornea, a clear, dome-shaped membrane that covers the front of the eye, is a key refractive element for vision. Layers of the cornea consist of the epithelium (outermost layer); Bowman layer; the stroma, which comprises approximately 90% of the cornea; Descemet membrane; and the endothelium. The endothelium removes fluid from and limits fluid into the stroma, thereby maintaining the ordered arrangement of collagen and preserving the cornea’s transparency. Diseases that affect the endothelial layer include Fuchs endothelial dystrophy, aphakic and pseudophakic bullous keratopathy (corneal edema following cataract extraction), and failure or rejection of a previous corneal transplant.

Treatment
The established surgical treatment for corneal disease is penetrating keratoplasty, which involves the creation of a large central opening through the cornea and then filling the opening with full-thickness donor cornea that is sutured in place. Visual recovery after penetrating keratoplasty may take 1 year or more due to slow wound healing of the avascular full-thickness incision, and the procedure frequently results in irregular astigmatism due to sutures and the full-thickness vertical corneal wound. Penetrating keratoplasty is associated with an increased risk of wound dehiscence, endophthalmitis, and total visual loss after relatively minor trauma for years after the index procedure. There is also the risk of severe, sight-threatening complications such as expulsive suprachoroidal hemorrhage, in which the ocular contents are expelled during the operative procedure, as well as postoperative catastrophic wound failure.

A number of related techniques have been, or are being, developed to selectively replace the diseased endothelial layer. One of the first endothelial keratoplasty techniques was termed deep lamellar endothelial keratoplasty, which used a smaller incision than penetrating keratoplasty, allowed more rapid visual rehabilitation, and reduced postoperative irregular astigmatism and suture complications. Modified endothelial keratoplasty techniques include endothelial lamellar keratoplasty, endokeratoplasty, posterior corneal grafting, and microkeratome-assisted posterior keratoplasty. Most frequently used at this time are Descemet stripping endothelial keratoplasty, which uses hand-dissected donor tissue, and Descemet stripping automated endothelial keratoplasty, which uses an automated microkeratome to assist in donor tissue dissection. These techniques include donor stroma along with the endothelium and Descemet membrane, which results in a thickened stromal layer after transplantation. If the donor tissue comprises Descemet membrane and endothelium alone, the technique is known as Descemet membrane endothelial keratoplasty. By eliminating the stroma on the donor tissue and possibly reducing stromal interface haze, Descemet membrane endothelial keratoplasty is considered a potential improvement over Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty. A variation of Descemet membrane endothelial keratoplasty is Descemet membrane automated endothelial keratoplasty. Descemet membrane automated endothelial keratoplasty contains a stromal rim of tissue at the periphery of the Descemet membrane endothelial keratoplasty graft to improve adherence and improve handling of the donor tissue. A laser may also be used for stripping in a procedure called femtosecond laser-assisted endothelial keratoplasty and femtosecond and excimer laser-assisted endothelial keratoplasty.

Endothelial keratoplasty involves removal of the diseased host endothelium and Descemet membrane with special instruments through a small peripheral incision. A donor tissue button is prepared from the corneoscleral tissue after removing the anterior donor corneal stroma by hand (eg, Descemet stripping endothelial keratoplasty) or with the assistance of an automated microkeratome (e.g., Descemet stripping automated endothelial keratoplasty) or laser (femtosecond laser-assisted endothelial keratoplasty or femtosecond and excimer laser-assisted endothelial keratoplasty). Donor tissue preparation may be performed by the surgeon in the operating room or by the eye bank and then transported to the operating room for final punch out of the donor tissue button. For minimal endothelial damage, the donor tissue must be carefully positioned in the anterior chamber. An air bubble is frequently used to center the donor tissue and facilitate adhesion between the stromal side of the donor lenticule and the host posterior corneal stroma. Repositioning of the donor tissue with the application of another air bubble may be required in the first week if the donor tissue dislocates. The small corneal incision is closed with 1 or more sutures, and steroids or immune-suppressants may be provided topically or orally to reduce the potential for graft rejection. Visual recovery following endothelial keratoplasty is typically 4 to 8 weeks.

Eye Bank Association of America statistics have shown the number of endothelial keratoplasty cases in the United States increased from 30710 in 2015 to 32221 in 2016.1 The Eye Bank Association of America estimated that, as of 2016, nearly 40% of corneal transplants performed in the United States were endothelial grafts. As with any new surgical technique, questions have been posed about long-term efficacy and risk of complications. Endothelial keratoplasty-specific complications include graft dislocations, endothelial cell loss, and rate of failed grafts. Long-term complications include increased intraocular pressure, graft rejection, and late endothelial failure.

Regulatory Status
Endothelial keratoplasty is a surgical procedure and, as such, is not subject to regulation by the U.S. Food and Drug Administration (FDA). Several microkeratomes have been cleared for marketing by the FDA through the 510(k) process. 

Related Policies
90301 Keratoprosthesis

90318 Optical Coherence Tomography of the Anterior Eye Segment 

Policy
Endothelial keratoplasty (Descemet’s stripping endothelial keratoplasty (DSEK)), Descemet’s stripping automated endothelial keratoplasty (DSAEK), Descemet’s membrane endothelial keratoplasty (DMEK), or Descemet’s membrane automated endothelial keratoplasty (DMAEK) may be considered MEDICALLY NECESSARY for the treatment of endothelial dysfunction, including but not limited to:

  • ruptures in Descemet’s membrane,
  • endothelial dystrophy,
  • aphakic, and pseudophakic bullous keratopathy,
  • iridocorneal endothelial (ICE) syndrome,
  • corneal edema attributed to endothelial failure,
  • and failure or rejection of a previous corneal transplant.

Femtosecond laser-assisted corneal endothelial keratoplasty (FLEK) or femtosecond and excimer lasers-assisted endothelial keratoplasty (FELEK) are investigational and /or unproven and are therefore considered NOT MEDICALLY NECESSARY.

Endothelial keratoplasty is NOT MEDICALLY NECESSARY when endothelial dysfunction is not the primary cause of decreased corneal clarity.

Policy Guidelines
Endothelial keratoplasty should not be used in place of penetrating keratoplasty for conditions with concurrent endothelial disease and anterior corneal disease. These situations would include concurrent anterior corneal dystrophies, anterior corneal scars from trauma or prior infection, and ectasia after previous laser vision correction surgery. Clinical input has suggested that there may be cases where anterior corneal disease should not be an exclusion, particularly if endothelial disease is the primary cause of the decrease in vision. Endothelial keratoplasty should be performed by surgeons adequately trained and experienced in the specific techniques and devices used.

Coding
Please see the Codes table for details. 

Benefit Application
Blue Card/National Account Issues
State or federal mandates (e.g., FEP) may dictate that all FDA-approved devices, drugs, or biologics may not be considered investigational, and thus these devices may be assessed only on the basis of their medical necessity.

Rationale
Evidence reviews assess the clinical evidence to determine whether the use of technology improves the net health outcome. Broadly defined, health outcomes are the length of life, quality of life, and ability to function - including benefits and harms. Every clinical condition has specific outcomes that are important to patients and managing the course of that condition. Validated outcome measures are necessary to ascertain whether a condition improves or worsens; and whether the magnitude of that change is clinically significant. The net health outcome is a balance of benefits and harms.

To assess whether the evidence is sufficient to draw conclusions about the net health outcome of technology, 2 domains are examined: the relevance, and quality and credibility. To be relevant, studies must represent one or more intended clinical use of the technology in the intended population and compare an effective and appropriate alternative at a comparable intensity. For some conditions, the alternative will be supportive care or surveillance. The quality and credibility of the evidence depend on study design and conduct, minimizing bias and confounding that can generate incorrect findings. The randomized controlled trial (RCT) is preferred to assess efficacy; however, in some circumstances, nonrandomized studies may be adequate. RCTs are rarely large enough or long enough to capture less common adverse events and long-term effects. Other types of studies can be used for these purposes and to assess generalizability to broader clinical populations and settings of clinical practice.

Comparative Studies
Dunker et al. (2020) published the results of a prospective, multicenter RCT comparing the efficacy of Descemet membrane endothelial keratoplasty (n = 29) versus ultrathin Descemet stripping automated endothelial keratoplasty (n = 25) in patients with Fuchs endothelial corneal dystrophy.2, Fifty-four patients were enrolled from 6 corneal centers in the Netherlands. There was no significant difference in best spectacle-corrected visual acuity (BSCVA) at 3 (P = 0.15), 6 (P = 0.20), or 12 months post-surgery (P = 0.06), between study arms. However, the percentage of eyes achieving 20/25 Snellen vision was significantly higher with Descemet membrane endothelial keratoplasty at 12 months (P = 0.02).

Woo et al. (2019) published the results of a retrospective comparative cohort study comparing long-term graft survival outcomes and complications of patients enrolled in the Singapore Corneal Transplant Registry.3 Patients with Fuchs endothelial corneal dystrophy and bullous keratopathy underwent Descemet membrane endothelial keratoplasty (121 eyes), Descemet stripping automated endothelial keratoplasty (423 eyes), or penetrating keratoplasty (405 eyes). Descemet membrane endothelial keratoplasty demonstrated better graft survival compared to Descemet stripping automated endothelial keratoplasty or penetrating keratoplasty in both Fuchs endothelial corneal dystrophy and bullous keratopathy. Overall cumulative graft survival was 97.4%, 78.4%, and 54.6% (p<0.001) in Descemet membrane endothelial keratoplasty, Descemet stripping automated endothelial keratoplasty, and penetrating keratoplasty groups, respectively. In eyes with Fuchs endothelial corneal dystrophy, the graft survival was 98.7%, 96.2%, and 73.5% (p=0.009) in Descemet membrane endothelial keratoplasty, Descemet stripping automated endothelial keratoplasty, and penetrating keratoplasty groups, respectively. In eyes with bullous keratopathy, the graft survival was 94.7%, 65.1%, and 47.0% (p=0.001) in Descemet membrane endothelial keratoplasty, Descemet stripping automated endothelial keratoplasty, and penetrating keratoplasty groups, respectively. Graft rejection was lowest in eyes undergoing Descemet membrane endothelial keratoplasty (1.7% vs. Descemet stripping automated endothelial keratoplasty 5.0% vs. penetrating keratoplasty 14.1%; p=0.001).

Descemet Stripping Endothelial Keratoplasty and Descemet Stripping Automated Endothelial Keratoplasty
Clinical Context and Therapy Purpose
The purpose of Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as penetrating keratoplasty, in patients with endothelial disease of the cornea.

The question addressed in this evidence review is: Does the use of Descemet stripping endothelial keratoplasty or Descemet stripping automated endothelial keratoplasty improve the net health outcome for patients with endothelial disease of the cornea?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals with endothelial disease of the cornea. Diseases that affect the endothelial layer include Fuchs endothelial corneal dystrophy, aphakic and pseudophakic bullous keratopathy (corneal edema following cataract extraction), and failure or rejection of a previous corneal transplant.

Interventions
The therapy being considered is Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty.

Comparators
Comparators of interest include penetrating keratoplasty and Descemet membrane endothelial keratoplasty.

Outcomes
The general outcomes of interest are change in disease status, morbid events, and functional outcomes. Relevant outcome measures include visual acuity, endothelial cell densities, patient satisfaction or quality-of-life, and complications including graft rejection, graft dislocation, and need for rebubble procedures.

Follow-up generally occurs through 1-2 years post-surgery.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

  1. To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs;
  2. In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
  3. To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
  4. Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
In 2009, the American Academy of Ophthalmology performed a review of the safety and efficacy of Descemet stripping automated endothelial keratoplasty, identifying a level I study (RCT of precut vs. surgeon dissected) along with 9 level II (well-designed observational studies) and 21 level III studies (mostly retrospective case series).4 Although more than 2,000 eyes treated with Descemet stripping automated endothelial keratoplasty were reported in different publications, most were reported by the same research group with some overlap in patients. The main results of this review are as follows:

Descemet stripping automated endothelial keratoplasty-induced hyperopia ranged from 0.7 to 1.5 diopters (D), with minimal induction of astigmatism (range, -0.4 to 0.6 diopters).

The reporting of visual acuity was not standardized in studies reviewed. The average best-corrected visual acuity ranged from 20/34 to 20/66, and the percentage of patients seeing 20/40 or better ranged from 38% to 100%.

The most common complication from Descemet stripping automated endothelial keratoplasty was posterior graft dislocation (mean, 14%; range, 0%-82%), with a lack of adhesion of the donor posterior lenticule to the recipient stroma, typically occurring within the first week. It was noted that this percentage might have been skewed by multiple publications from a single research group with low complication rates. Graft dislocation required additional surgical procedures (rebubble procedures) but did not lead to sight-threatening vision loss in the articles reviewed.

Endothelial graft rejection occurred in a mean of 10% of patients (range, 0%-45%); most were reversed with topical or oral immunosuppression, with some cases progressing to graft failure. Primary graft failure, defined as unhealthy tissue that has not cleared within 2 months, occurred in a mean of 5% of patients (range, 0%-29%). Iatrogenic glaucoma occurred in a mean of 3% of patients (range, 0%-15%) due to a pupil block induced from the air bubble in the immediate postoperative period or delayed glaucoma from topical corticosteroid adverse events.

Mean endothelial cell loss, which provides an estimate of long-term graft survival, was 37% at 6 months and 41% at 12 months. These percentages of cell loss were reported to be similar to those observed with penetrating keratoplasty.

Reviewers concluded that Descemet stripping automated endothelial keratoplasty appeared to be at least equivalent to penetrating keratoplasty regarding safety, efficacy, surgical risks, and complication rates, although long-term results were not yet available. The evidence also indicated that endothelial keratoplasty is superior to penetrating keratoplasty regarding refractive stability, postoperative refractive outcomes, wound- and suture-related complications, and risk of intraoperative choroidal hemorrhage. The reduction in serious and occasionally catastrophic adverse events associated with penetrating keratoplasty has led to the rapid adoption of endothelial keratoplasty for treatment of corneal endothelial failure.

A Cochrane review of Descemet stripping automated endothelial keratoplasty compared to Descemet membrane endothelial keratoplasty for corneal endothelial failure was published in 2018.5 The literature search identified 4 nonrandomized trials including 72 adult participants (144 eyes) who received Descemet stripping automated endothelial keratoplasty in the first eye followed by Descemet membrane endothelial keratoplasty in the fellow eye published between 2011 and 2015. All participants met criteria for Fuchs endothelial dystrophy and endothelial failure requiring a corneal transplant. Studies reported outcomes at various time points, including 6, 12, and 6-24 months. At 1 year post-procedure, Descemet membrane endothelial keratoplasty resulted in better best-corrected visual acuity compared to Descemet stripping automated endothelial keratoplasty (mean difference, -0.14; 95% confidence interval [CI], -0.18 to -0.10 Logarithm of the Minimum Angle of Resolution [logMar] ; low-certainty evidence). Two studies reported that Descemet membrane endothelial keratoplasty provided a higher cell density at 1 year. Graft dislocations requiring rebubbling were more common using Descemet membrane endothelial keratoplasty, although this difference could not be precisely estimated (relative risk [RR], 5.40; 95% CI, 1.51 to 19.3; very low-certainty evidence). The paired, contralateral eye studies in which Descemet stripping automated endothelial keratoplasty in 1 eye preceded Descemet membrane endothelial keratoplasty in the fellow eye for all patients was found to be at high-risk for bias due to potential unknown confounding factors.

Marques et al. (2019) conducted a meta-analysis of Descemet membrane endothelial keratoplasty compared to Descemet stripping automated endothelial keratoplasty for Fuchs endothelial dystrophy.6 A literature search through August 2017 identified 10 retrospective studies of moderate methodological quality (n=947 eyes; 646 Descemet membrane endothelial keratoplasty). The primary outcome consisted of the mean difference in best-corrected visual acuity at 3, 6, and 12 months post-procedure. Secondary outcomes included rates of graft failure, rejection, rebubbling, endothelial cell density, subjective visual outcomes, and patient satisfaction. Best-corrected visual acuity was improved with Descemet membrane endothelial keratoplasty at all time points compared to Descemet stripping automated endothelial keratoplasty (12 months: 0.16 logMAR vs. 0.30 logMAR; p<0.001). Descemet membrane endothelial keratoplasty had a 60% reduced rate of rejection (RR, 0.4; 95% CI, 0.24 to 0.67; p=0.0005) but required more rebubblings (RR, 2.48; 95% CI, 1.32 to 4.64; p=0.005). Descemet membrane endothelial keratoplasty had an increased number of primary graft failures and less endothelial cell density loss; however, these differences did not reach statistical significance. More patients reported being satisfied after Descemet membrane endothelial keratoplasty (odds ratio [OR], 10.29; 95% CI, 3.55 to 29.80; p<0.0001).

Randomized Controlled Trials
Chamberlain et al. (2018) compared clinical outcomes of ultrathin-Descemet stripping automated endothelial keratoplasty with Descemet membrane endothelial keratoplasty in patients with damaged or diseased endothelium from Fuchs endothelial dystrophy or pseudophakic bullous keratopathy in the Descemet Endothelial Thickness Comparison Trial (DETECT).7 The primary outcome measure was best spectacle-corrected visual acuity at 6 months. Secondary outcomes included 3- and 12-month best spectacle-corrected visual acuity, endothelial cell counts, and complications. The study included 50 eyes from 38 patients with 25 eyes randomized to each treatment arm. Compared to ultrathin Descemet stripping automated endothelial keratoplasty, Descemet membrane endothelial keratoplasty had superior visual acuity results. Best spectacle-corrected visual acuity was 1.5 lines better at 3 months (95% CI, 2.5 to 0.6 lines better; p=0.002), 1.8 lines better at 6 months (95% CI, 2.8 to 1.0 lines better; p<0.001), and 1.4 lines better at 12 months (95% CI, 2.2 to 0.7 lines better; p<0.001). Average endothelial cell counts were 1855 cells/mm2 in and 2070 cells/mm2 in ultrathin Descemet stripping automated endothelial keratoplasty at 12 months (p=0.051). Intraoperative and postoperative complications rates were not statistically different between groups. Duggan et al (2019) reported an update on corneal higher-order aberrations after ultrathin Descemet stripping automated endothelial keratoplasty versus Descemet membrane endothelial keratoplasty in DETECT.8 In patients receiving Descemet membrane endothelial keratoplasty, the posterior corneal surface had significantly fewer coma aberrations (p≤0.003) and total higher-order aberrations (p≤0.001) at 3, 6, and 12 months post-surgery compared to ultrathin Descemet stripping automated endothelial keratoplasty. Descemet membrane endothelial keratoplasty was found to decrease whereas ultrathin Descemet stripping automated endothelial keratoplasty was found to increase posterior corneal higher-order aberrations compared with presurgical values, potentially accounting for the better visual acuity observed with Descemet membrane endothelial keratoplasty. Hirabayashi et al (2020) reported on an update of corneal light scatter outcomes as measured by densitometry in DETECT.9 Both Descemet membrane endothelial keratoplasty and ultrathin Descemet stripping automated endothelial keratoplasty were found to improve the degree of corneal light scatter after surgery, with no differences between groups observed at 12 months post-surgery.

Observational Studies
Fuest et al. (2017) compared 5-year visual acuity outcomes in patients receiving Descemet stripping automated endothelial keratoplasty (n=423) or penetrating keratoplasty (n=405) in the Singapore Cornea Transplant Registry.10 Mean age of patients was 67 years. The Descemet stripping automated endothelial keratoplasty group had a higher percentage of Chinese patients, a higher percentage of patients with Fuchs endothelial dystrophy, and a lower percentage of patients with bullous keratopathy than the penetrating keratoplasty group. Controlling for preoperative best spectacle-corrected visual acuity, which differed significantly between groups, patients receiving Descemet stripping automated endothelial keratoplasty experienced significantly better vision through 3 years of follow-up than patients undergoing penetrating keratoplasty. Four- and 5-year follow-up measures showed similar best spectacle-corrected visual acuity among both treatment groups. Subgroup analyses by Fuchs endothelial dystrophy and bullous keratopathy showed similar patterns of significantly better vision through the first 3 years of follow-up in patients receiving Descemet stripping automated endothelial keratoplasty than in patients receiving penetrating keratoplasty.

Heinzelmann et al. (2016) reported on 2-year outcomes in patients who underwent endothelial keratoplasty or penetrating keratoplasty for Fuchs endothelial dystrophy or bullous keratopathy.11 The study included 89 eyes undergoing Descemet stripping automated endothelial keratoplasty and 329 eyes undergoing penetrating keratoplasty. The postoperative visual improvement was faster after endothelial keratoplasty than after penetrating keratoplasty. For example, among patients with Fuchs endothelial dystrophy, 50% of patients achieved a best-corrected visual acuity of Snellen 6/12 or more 18 months after Descemet stripping automated endothelial keratoplasty versus more than 24 months after penetrating keratoplasty. Endothelial cell loss was similar after endothelial keratoplasty and penetrating keratoplasty in the early postoperative period. However, after an early decrease, endothelial cell loss stabilized in patients who received endothelial keratoplasty whereas the decrease continued in those who had penetrating keratoplasty. Among patients with Fuchs endothelial dystrophy, there was a slightly increased risk of late endothelial failure in the first 2 years with endothelial keratoplasty than with penetrating keratoplasty. Graft failure was reported to be lower among patients with bullous keratopathy compared with patients with Fuchs endothelial dystrophy (numbers not reported).

Longer-term outcomes have been reported in several studies. Five-year outcomes from a prospective study conducted at the Mayo Clinic were published by Wacker et al (2016).12 The study included 45 participants (52 eyes) with Fuchs endothelial dystrophy who underwent Descemet stripping endothelial keratoplasty. Five-year follow-up was available for 34 (65%) eyes. Mean high-contrast best spectacle-corrected visual acuity was 20/56 Snellen equivalent presurgery and decreased to 20/25 Snellen equivalent at 60 months. The difference in high-contrast best spectacle-corrected visual acuity at 5 years versus pre-surgery was statistically significant (p<0.001). Similarly, the proportion of patients with best spectacle-corrected visual acuity of 20/25 Snellen equivalent or better increased from 26% at 1 year post-surgery to 56% at 5 years (p<0.001). There were 6 graft failures during the study period (4 failed to clear after surgery, 2 failed during follow-up). All patients with graft failures were regrafted.

Previously, 3-year outcomes after Descemet stripping automated endothelial keratoplasty were reported by an eye institute.13 This retrospective analysis (2012) included 108 patients who underwent Descemet stripping automated endothelial keratoplasty for Fuchs endothelial dystrophy or pseudophakic bullous keratopathy and had no other ocular comorbidities. Best spectacle-corrected visual acuity was measured at 6 months and 1, 2, and 3 years. Best spectacle-corrected visual acuity after Descemet stripping automated endothelial keratoplasty improved over 3 years of follow-up. The percentage of patients who reached a best spectacle-corrected visual acuity of 20/20 or greater was 0.9% at baseline, 11.1% at 6 months, 13.9% at 1 year, 34.3% at 2 years, and 47.2% at 3 years. Ninety-eight percent of patients reached a best spectacle-corrected visual acuity of 20/40 or greater by 3 years.

Table 1. Summary of Key Nonrandomized Trial Characteristics

Study Study Type Country Dates Participant n eyes DSAEK, N PK DMEK Follow-Up
Fuest et al. (2017)10 Prospective Singapore 1991-2011 Total N=828 423 N=405 NR 5 yrs
Heinzelman et al. (2016)11 Cohort Germany 2011-2014 Total N=868 89 N=329 N=450 2 yrs
Wacker et al. (2016)12 Prospective U.S. 2006-2010 Total N=52 34 NR NR 5 yrs (n=34, 65%)
Li et al. (2012)13 Retrospective U.S. 2005-2007 Total N=207 108 NR NR 3 yrs

DMEK: Descemet membrane endothelial keratoplasty; DSAEK: Descemet stripping automated endothelial keratoplasty; NR: not reported; PK=penetrating keratoplasty.

Table 2. Summary of Key Nonrandomized Trial Results

Study BSCVA SE   Cylinder    
Fuest et al. (2017)10 at 5-yrs (n); mean(SD) at 5-yrs (n); mean (SD)   at 5-yrs (n); mean (SD)    
Total (n=89);0.62(0.6); p=0.037 (n=62); -1.7(2.7); P=0.017   (N=62); -3.1(2.1); P<0.001    
DSAEK (N=25);0.46(0.5); p=0.037 (N=18); -0.8(1.7); P=0.017   (N=18); -1.6(1.1); P<0.001    
PK (N=25); 0.63(0.6); p=0.037 (N=44); -2.1(2.9); P=0.017   (N=44); -3.75(2.1); P=0.001    
Study % of BSCVA of Snellen 6/7.5 or better at 24-months   Chronic endothelial cell loss > 500 cells/mm 2 at 15 mos   Chronic endothelial cell loss > 500 cells/mm 2 at 24 mos  
Heinzelmann et al. (2016)11            
FED DMEK 53%   95%   NR  
FED DSAEK 15%   93%      
FED PK 10%   99%   NR  
BK DMEK NR   NR   NR  
BK DSAEK NR   NR   NR  
BK PK NR   NR   90%  
Study Mean high-contrast BSCVA presurgery Mean high-contrast BSCVA at 5-yrs        
Wacker et al. (2016)12            
FECD DSEK 20/56 20/25        
Study % of eyes achieving a % of eyes achieving a   % of eyes achieving a   % of eyes achieving a
  BSCVA of 20/40 at 3-years BSCVA of 20/30 at 3-yrs   BSCVA of 20/25 at 3-years   BSCVA of 20/20 at 3-yrs
Li et al. (2012)13            
FED+BK DSAEK 98.1% (N=106) 90.7% (N=98)   70.4% (N=76)   47.2% (N=51)

N:eyes; DSAEK:Descemet stripping automated endothelial keratoplasty; BK:bullous keratopathy; NR; not reported; PK:penetrating keratoplasty; BSCVA: best spectacle-corrected visual acuity; SD: standard deviation;
SE:spherical equivalent; DMEK:Descemet membrane endothelial keratoplasty; FED/FECS: Fuchs’ endothelial corneal
dystrophy; DSEK: Descemet stripping endothelial keratoplasty.

Section Summary: Descemet Stripping Endothelial Keratoplasty and Descemet Stripping Automated Endothelial Keratoplasty
Evidence for the use of Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty consists of a systematic review and several large observational studies with follow-up extending from 2 to 5 years. The review and the studies showed that patients undergoing Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty experience greater improvements in visual acuity than patients undergoing penetrating keratoplasty. Also, patients undergoing Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty experienced significantly fewer serious adverse events than patients undergoing penetrating keratoplasty.

Descemet Membrane Endothelial Keratoplasty and Descemet Membrane Automated Endothelial Keratoplasty
Clinical Context and Therapy Purpose
The purpose of Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as penetrating keratoplasty, in patients with endothelial disease of the cornea.

The question addressed in this evidence review is: Does the use of Descemet membrane endothelial keratoplasty or Descemet membrane automated endothelial keratoplasty improve the net health outcome for patients with endothelial disease of the cornea?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals with endothelial disease of the cornea. Diseases that affect the endothelial layer include Fuchs endothelial dystrophy, aphakic and pseudophakic bullous keratopathy (corneal edema following cataract extraction), and failure or rejection of a previous corneal transplant.

Interventions
The therapy being considered is Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty. It has been suggested that by eliminating the stroma on the donor tissue, Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty may reduce stromal interface haze and provide better visual acuity outcomes than Descemet stripping endothelial keratoplasty or Descemet stripping automated endothelial keratoplasty.14,15

Comparators
Comparators of interest include penetrating keratoplasty.

Outcomes
The general outcomes of interest are change in disease status, morbid events, and functional outcomes. Relevant outcome measures include visual acuity, endothelial cell densities, patient satisfaction or quality-of-life, and complications including graft rejection, graft dislocation, and need for rebubble procedures.

Follow-up generally occurs through 1-2 years post-surgery.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

  1. To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs;
  2. In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
  3. To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
  4. Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
The American Academy of Ophthalmology conducted a systematic review of the safety and outcomes of Descemet membrane endothelial keratoplasty and investigated whether Descemet membrane endothelial keratoplasty offered any advantages over Descemet stripping endothelial keratoplasty (Deng et al. [2018]).16 The literature search, conducted through May 2017, identified 47 studies for inclusion. Quality was assessed using a scale from the Oxford Centre for Evidence-Based Medicine. Two studies were rated level I evidence (well-designed and well-conducted RCTs), 15 studies were level II (well-designed case-control or cohort studies or RCTs with methodologic deficits), and 30 studies were level III (case series, case reports, or poor-quality cohort or case-control). Mean length of follow-up among the studies ranged from 5 to 68 months. A best spectacle-corrected visual acuity of 20/25 was achieved by 33% to 67% of patients (5 studies). A best spectacle-corrected visual acuity of 20/20 was achieved by 29% to 32% (3 studies) at 3 months postsurgery and by 17% to 67% at 6 months postsurgery. Seven studies, 6 of which were rated as level II evidence, directly compared Descemet membrane endothelial keratoplasty with Descemet stripping endothelial keratoplasty and all 7 showed a faster visual recovery and a better visual outcome after Descemet membrane endothelial keratoplasty compared with Descemet stripping endothelial keratoplasty. The rate of endothelial cell loss, graft failure, and intraoperative and postoperative complications was similar between Descemet membrane endothelial keratoplasty and Descemet stripping endothelial keratoplasty.

Singh et al. (2017) conducted a systematic review and meta-analysis of studies comparing Descemet membrane endothelial keratoplasty with Descemet stripping endothelial keratoplasty or Descemet stripping automated endothelial keratoplasty.17 The literature search, conducted through May 2016, identified 9 studies for inclusion in the qualitative analysis and 7 studies for inclusion in the meta-analysis. A quality assessment of studies was not presented. Meta-analyses of 343 eyes showed that the 6-month mean difference in best spectacle-corrected visual acuity was significantly better in patients undergoing Descemet membrane endothelial keratoplasty than in patients undergoing Descemet stripping endothelial keratoplasty (-0.13; 95% CI, -0.16 to -0.09). The 6-month mean difference in endothelial cell density (n=348) did not differ significantly between groups (76.8; 95% CI, -79.8 to 233.4), though the interpretation of this result is limited due to high heterogeneity. A higher rate of air injection/rebubbling was reported among patients in the Descemet membrane endothelial keratoplasty group compared with the Descemet stripping endothelial keratoplasty group.

Pavlovic et al. (2017) conducted a meta-analysis of 11 studies comparing Descemet membrane endothelial keratoplasty (n=350) with Descemet stripping automated endothelial keratoplasty (n=373).18 The date of the literature search and quality assessment methods were not reported. The mean difference in best spectacle-corrected visual acuity did not differ significantly at the 3-month follow-up (-0.12; 95% CI, -0.28 to 0.04), but was significantly better in the Descemet membrane endothelial keratoplasty group than in the Descemet stripping automated endothelial keratoplasty group at both the 6-month (-0.12; 95% CI, -0.15 to -0.10) and at the 6-month and beyond follow-ups (-0.13; 95% CI, -0.17 to -0.09). There were no statistical differences in endothelial cell loss between the 2 procedures at 6 (mean difference, 0.2; 95% CI, -5.6 to 6.1) or 12 months (3.6; 95% CI, -3.7 to 10.9). There were more graft rejections reported among patients in the Descemet stripping automated endothelial keratoplasty group compared with those in the Descemet membrane endothelial keratoplasty group, but the difference was not significant (OR, 2.7; 95% CI, 0.6 to 11.9). There were more graft failures reported in the Descemet membrane endothelial keratoplasty group compared with the Descemet stripping automated endothelial keratoplasty group, but this difference, too, was not significant (OR, 2.8; 95% CI, 0.7 to 10.6).

Li et al. (2017) conducted a systematic review and meta-analysis comparing Descemet membrane endothelial keratoplasty and Descemet stripping endothelial keratoplasty.19 The literature search, conducted through January 2017, identified 19 studies for inclusion: 15 retrospective control studies, a prospective nonrandomized case series, and 3 for which the study designs could not be determined from the meeting abstracts. A modified version of the Newcastle-Ottawa Scale was used to assess the quality of the studies. Eight items relating to selection, comparability, and outcome were assessed, and if a study received a score greater than 6, it was considered relatively high quality. Two studies had a score of 7, 8 studies had a score of 6, 3 studies had a score of 5, and 6 studies had a score of 4. A total of 2,378 eyes were included in the studies, 1,124 receiving Descemet membrane endothelial keratoplasty and 1,254 receiving Descemet stripping endothelial keratoplasty. Meta-analyses of 13 studies showed an overall mean difference in best spectacle-corrected visual acuity that was significantly improved in the Descemet membrane endothelial keratoplasty group compared with the Descemet stripping endothelial keratoplasty group (-0.15; 95% CI, -0.19 to -0.11). This significant mean difference in best spectacle-corrected visual acuity was seen at the 3-, 6-, and 12-month follow-ups. Meta-analyses, which included 354 Descemet membrane endothelial keratoplasty and 313 Descemet stripping endothelial keratoplasty eyes (total n=667), showed no significant difference in endothelial cell density between groups (mean difference, 14.9; 95% CI, -181.5 to 211.3). The most common complication in both procedures was partial or total graft detachment, with significantly more occurrences in the Descemet membrane endothelial keratoplasty group than in the Descemet stripping endothelial keratoplasty group (OR, 4.6; 95% CI, 2.4 to 8.6).

Table 3. SR & M-A Characteristics

Study Dates Trials N (Eyes) Intervention N (Range) Design Duration
Deng et al. (2018)16 NR-05/2017 47 9046; patients with corneal endothelial dysfunction DMEK 9046 (25-905) RCT; case-control and cohort; case series, case reports 5.3-68 mos
Singh et al. (2017)17 NR-05/2016 9 586 DMEK, DSAEK 586 (20-155) NR NR
Pavlovic et al. (2017)18 NR 11 723 DMEK (n=350); DSAEK (n=373) NR NR NR
Li et al. (2017)19 NR-01/2017 19 2378 DMEK; DSEK 2378 (20-739)   3.1-22.55

DMEK: Descemet membrane endothelial keratoplasty; DSEK: Descemet stripping endothelial keratoplasty; DSAEK: Descemet stripping automated endothelial keratoplasty; M-A: meta-analysis; NR: not reported; RCT: randomized controlled trial; SR: systematic review.

Table 4. SR & M-A Results

Study Mean BCVA at 6 mos after DMEK Mean endothelial cell loss at time Change in SE Minimal Induced Astigmatism
Deng et al. (2018)16        
Total N*=9046 Range: 20/21 to 20/31 33% (range, 25%-47%) (6-mos) +0.43 diopters (D; range, -1.17 to +1.2 D) +0.03 D (range, -0.03 to +1.11 D)
Study BCVA at 6-mos ECD at 6-mos Graft Detachment overall Graft Rejection
Singh et al. (2017)17        
After DMEK, mean; SD, p-value 0.161; 0.129; P<0.0001; N=184 1855; 442 P=0.708 NR NR
After DSAEK, mean; SD, p-value 0.293; 0.153 P<0.0001; N=159 1872; 429 P=0.708 NR NR
Pooled mean difference (CI, SD) -0.13 (95% CI, 0.16 to 0.09); N=343 Could not be interpreted due to high statistical heterogeneity NR NR
Pavlovic et al. (2017)18 Not available ECL* at 6-mos Not available Not available
Mean difference between DSAEK and DMEK group -0.12; 95% CI, -0.15 to -0.10 0.2; 95% CI, -5.6 to 6.1 Not available Not available
Li et al. (2017)19 N=108 N=108 N=108 N=108
Comparison between DMEK and DSEK (MD[95%CI] % weight) -0.13 (-0.17, -0.08) 51.29 25.59 (-183.15, 234.32) p=0.810 4.56 (2.43, 8.58) -0.04 (-0.08, - 0.002)
Pooled mean difference (CI, SD) overall -0.15 ( -0.19 to -0.11) P<0.001) 14.88 (-181.5 to 211.27) P=0.882 NR NR

*N=eyes
BCVA: best-corrected visual acuity; CI: confidence interval; DMEK: Descemet membrane endothelial keratoplasty; DSEK: Descemet stripping endothelial keratoplasty; DSAEK: Descemet stripping automated endothelial keratoplasty; ECD: endothelial corneal dystrophy; ECL: endothelial cell loss; M-A: meta-analysis; MD: mean difference; NR: not reported; SD: standard deviation; SE: spherical equivalent; SR: systematic review.

Observational Studies
Oellerich et al. (2017) reported on 6-month outcomes of a large cohort of patients undergoing Descemet membrane endothelial keratoplasty by 55 surgeons from 23 countries.20 Outcomes of interest were best spectacle-corrected visual acuity, a decrease in endothelial cell density, and complications. Subgroup analyses were conducted by number of procedures performed by the surgeon (1 to 24 [39%], 25 to 99 [38%], and ≥100 [23%]). In the total population, 91% of patients achieved best spectacle-corrected visual acuity improvement, with 5% experiencing no change and 5% experiencing deterioration in visual acuity. Subgroup analyses showed that the proportion of patients achieving best spectacle-corrected visual acuity improvement did not differ significantly between patients whose surgeons had performed 100 or more procedures and those whose surgeons had performed fewer than 25 procedures. Nine percent of patients experienced intraoperative complications, with the rate decreasing significantly as the surgeon performed more procedures. The most frequent postoperative complication was partial graft detachment (27%), which also decreased significantly with surgeon experience. Rates of other postoperative complications such as graft failure, cataract, and glaucoma did not differ based on surgeon experience.

Tourtas et al. (2012) conducted a retrospective comparison of 38 consecutive patients/eyes that underwent Descemet membrane endothelial keratoplasty and 35 consecutive patients/eyes who had undergone Descemet stripping automated endothelial keratoplasty.21 Only patients with Fuchs endothelial dystrophy or pseudophakic bullous keratopathy were included. After Descemet membrane endothelial keratoplasty, 82% of eyes required rebubbling. After Descemet stripping automated endothelial keratoplasty, 20% of eyes required rebubbling. Best spectacle-corrected visual acuity in both groups was comparable at baseline (Descemet membrane endothelial keratoplasty=0.70 logMAR; Descemet stripping automated endothelial keratoplasty=0.75 logMAR). At 6-month follow-up, mean visual acuity improved to 0.17 logMAR after Descemet membrane endothelial keratoplasty and 0.36 logMAR after Descemet stripping automated endothelial keratoplasty. This difference was statistically significant. At 6 months following surgery, 95% of Descemet membrane endothelial keratoplasty treated eyes reached a visual acuity of 20/40 or better, and 43% of Descemet stripping automated endothelial keratoplasty treated eyes reached a visual acuity of 20/40 or better. Endothelial cell density decreased by a similar amount after both procedures (41% after Descemet membrane endothelial keratoplasty, 39% after Descemet stripping automated endothelial keratoplasty).

Van Dijk et al. (2013) reported on outcomes of their first 300 consecutive eyes treated with Descemet membrane endothelial keratoplasty.22 Indications for Descemet membrane endothelial keratoplasty were Fuchs endothelial dystrophy, pseudophakic bullous keratopathy, failed penetrating keratoplasty, or failed endothelial keratoplasty. Of the 142 eyes evaluated for visual outcomes at 6 months, 79% reached a best spectacle-corrected visual acuity of 20/25 or more, and 46% reached a best spectacle-corrected visual acuity of 20/20 or more. Endothelial cell density measurements at 6 months were available in 251 eyes. Average cell density was 1674 cells/mm2, representing a decrease of 34.6% from preoperative donor cell density. The major postoperative complication in this series was graft detachment requiring rebubbling or regraft, which occurred in 10.3% of eyes. Allograft rejection occurred in 3 eyes (1%), and intraocular pressure was increased in 20 (6.7%) eyes. Except for 3 early cases that may have been prematurely regrafted, all but 1 eye with an attached graft cleared in 1 to 12 weeks.

A 2009 review of cases from another group in Europe suggested that a greater number of patients achieve 20/25 vision or better with Descemet membrane endothelial keratoplasty.23 Of the first 50 consecutive eyes, 10 (20%) required a secondary Descemet stripping endothelial keratoplasty for failed Descemet membrane endothelial keratoplasty. For the remaining 40 eyes, 95% had a best spectacle-corrected visual acuity of 20/40 or better, and 75% had a best spectacle-corrected visual acuity of 20/25 or better. Donor detachments and primary graft failure with Descemet membrane endothelial keratoplasty were problematic. In 2011, this group reported on the surgical learning curve for Descemet membrane endothelial keratoplasty, with their first 135 consecutive cases retrospectively divided into 3 subgroups of 45 eyes each.24 Graft detachment was the most common complication, which decreased with surgeon experience. In their first 45 cases, a complete or partial graft detachment occurred in 20% of cases, compared with 13.3% in the second group and 4.4% in the third group. Clinical outcomes in eyes with normal visual potential and a functional graft (n=110) were similar across the 3 groups, with an average endothelial cell density of 1747 cells/mm2 and 73% of cases achieving a best spectacle-corrected visual acuity of 20/25 or better at 6 months.

A North American group reported on 3-month outcomes from a prospective consecutive series of 60 cases of Descemet membrane endothelial keratoplasty in 2009, and in 2011, they reported on 1-year outcomes from these 60 cases plus an additional 76 cases of Descemet membrane endothelial keratoplasty.25,26 Preoperative best spectacle-corrected visual acuity averaged 20/65 (range of 20/20 to counting fingers). Sixteen eyes were lost to follow-up, and 12 (8.8%) grafts had failed. For the 108 grafts examined and found to be clear at 1 year, 98% achieved a best spectacle-corrected visual acuity of 20/30 or better. Endothelial cell loss was 31% at 3 months and 36% at 1 year. Although visual acuity outcomes appeared to be improved over a Descemet stripping automated endothelial keratoplasty series from the same investigators, preparation of the donor tissue and attachment of the endothelial graft were more challenging. A 2012 cohort study by this group found reduced transplant rejection with Descemet membrane endothelial keratoplasty.27 One (0.7%) of 141 patients in the Descemet membrane endothelial keratoplasty group had a documented episode of rejection compared with 54 (9%) of 598 in the Descemet stripping endothelial keratoplasty group and 5 (17%) of 30 in the penetrating keratoplasty group.

The same group also reported on a prospective consecutive series (2011) of their initial 40 cases (36 patients) of Descemet membrane automated endothelial keratoplasty (microkeratome dissection and a stromal ring).28 Indications for endothelial keratoplasty were Fuchs endothelial dystrophy (87.5%), pseudophakic bullous keratopathy (7.5%), and failed endothelial keratoplasty (5%). Air was reinjected in 10 (25%) eyes to promote graft attachment; 2 (5%) grafts failed to clear and were successfully regrafted. Compared with a median best spectacle-corrected visual acuity of 20/40 at baseline (range, 20/25 to 20/400), median best spectacle-corrected visual acuity at 1 month was 20/30 (range, 20/15 to 20/50). At 6 months, 48% of eyes had 20/20 vision or better, and 100% had 20/40 or better. Mean endothelial cell loss at 6 months relative to baseline donor cell density was 31%.

Table 5. Summary of Key Observational Study Characteristics

Study Study Design Country Dates Participants Treatment 1 Treatment 2 Follow-Up
Oellerich et al. (2017)20 Retrospective cohort Europe, Asia, Africa, North America, South America, Australia Aug 2008-July 2015 Patient age [mean SD (range)] (n=2448); 69.8 +/- 11.0 (16-99); 37% male, 57.9% female, 5.2% not specified; 74.4% FED, 16.8% BK; 7.6% failed transplant, 0.9% other; 0.3% not specified DMEK (n=2448) NR 6-mos
Van Dijk et al. (2013)22 Prospective Netherlands NR Patient age (n=248 patients), [mean +/- SD (range), female/male]; 67 +/- 13 (30-93), 166/134; FED=272; BK=17 patients; Failed DSEK/PK=9/1 patients DMEK (N=300) NR 6-mos
Tourtas et al. (2012)21 Retrospective cohort Germany Aug 2009-Dec 2009; DSAEK: Aug 2008-Mar 2009 Patient age [mean +/- SD (range), female/male] (n=73) DMEK: 68.3 +/- 9 (42-85), 16/22; DSAEK: 68.1 +/- 11 (48-87), 20/15 DMEK (N=38) DSAEK (N=35) 6-mos
Ham et al. (2009)23 Prospective case Netherlands NR Patient s with FED; 23 men, 27 women; ages (range) 41-88 yrs; N=50 eyes DMEK (N=40) DMEK followed by DSEK as a back-up procedure in the event of DMEK graft failure (N=10) 6-mos
Dapena et al. (2011)24 Retrospective Netherlands Feb 2005-Dec 2010 118 patients with FED, 49 male, 69 female; ages (range) 33-93 yrs (N=135) DMEK (N=135) NR 6-mos
Price et al. (2009)25 Prospective U.S. Feb 2009-Oct 2009 58 patients with FED, PK, or failed previous graft; mean age +/- SD (yrs)=68 +/- 9.9 (48-85); female/male=34/26 (N=60) DMEK (N=60) NR 3-mos
Guerra et al. (2011)26 Prospective U.S. Feb 2009-Oct 2009 Patients (n=112 with FED, PK, or failed previous graft; mean age +/- SD (yrs)= 78 +/- 10.36 (48.12-89.99); female/male=72/40 (N=136) DMEK (N=136) NR 1-y
McCauley et al. (2011)28 Prospective U.S. NR 36 patients (n=40 eyes) treated with DMAEK. Mean patient age 69 y (range: 48-88yrs); 53% female DMAEK (n=40) NR 6-mos
Anshu (2012)27 Comparative U.S. Feb 2009-Oct 2009 Patients undergoing DMEK compared retrospectively with matched cohort undergoing DSEK (598) and PK (n=30), treated at same center, with similar demographics, follow-up, duration, indications for surgery DMEK/DSEK PK 2-yrs

BK: bullous keratopathy; DMAEK: Descemet membrane automated endothelial keratoplasty; DMEK: Descemet membrane endothelial keratoplasty; DSAEK: Descemet stripping automated endothelial keratoplasty; DSEK: Descemet stripping endothelial keratoplasty; FED/FECD: Fuchs endothelial corneal dystrophy; NR: not reported; PK: penetrating keratoplasty; SD: standard deviation.
N=eyes except where indicated otherwise.

Table 6. Summary of Key Observational Study Results

Study BCVA preoperative BCVA 6 mos FU ECD preoperative mean +/- SD (cells/mm 2) ECD 6 mos FU mean +/-SD (cells/mm 2) Postoperative complications
Oellerich et al. (2017)20 N=2430 N=1959 N=1956 N=1405 N=2363
DMEK N (%) ≥ 20/25 Snellen = 46.17 (1.9%) N (%) ≥ 20/25 Snellen = 889 (45.4%) 2635 +/- 294 1575+/- 489 647 (27.4%) [for all types of post-operative complications]
Van Dijk et al. (2013)22 N=221 N=221 N=251 N=251 N=300
DMEK N (%) ≥ 20/25 Snellen = 16 (7%) N (%) ≥ 20/25 Snellen = 175 (79%) NR 1674 +/- 518 31 (10%) for most frequent complication, (partial) graft detachment
(N total=300)          
Tourtas et al. (2012)21 N=73 N=73 N=73 total N=73 N=73
DMEK (n=38) Mean +/- SD; 0.70 +/- 0.48 logMAR Mean +/- SD; 0.17 +/- 0.12 logMAR (n=38) 2575 +/- 260 1520 +/- 299 31 (82%) required air injections for partial dehiscence of the EDM
DSAEK (n=35) N +/- SD; 0.75 +/- 0.32 logMAR N +/- SD; 0.36 +/- 0.15 logMAR (n=35) 2502 +/- 220 1532 +/- 495 7 (20%) required air injections for partial dehiscence of the EDM
Ham et al. (2009)23 N=50 N=47 N=47 N=43  
Pooled (N=50) NR N (%) ≥ 20/25 Snellen = 47 (66%) 2623 2623 All complications, N=14 (28%)
      +/-193 (n=47) +/- 193 (n=43)  
DMEK only (N=40) NR N (%) ≥ 20/25 Snellen = 30 (75%) 2618 1876 +/- 522 (n=35) NR
      +/-201 (n=40)    
Dapena et al. (2011)24 N=135 N=110 N=135 174 +/- 527 (n=106) Primary graft failure (2.2%, 3/135)
DMEK N=135 NR N (%) ≥ 20/25 Snellen = 80 (73%) NR    
Price et al. (2009)25 N=60 N=57 at 3-mos N=60 N=57 at 3-mos NR
DMEK Median preoperative BSCVA (N=52)=20/50 N (%) ≥ 20/25 Snellen=36 (63%), 3010 +/- 200 (range, 2520-3430) 30% +/- 20% (range, 2.7%-78%) NR
Study BSCVA BSCVA FU [time] ECD pre-operative (mean +/- SD, cells/mm 2) ECD 6m FU (mean +/- SD, cells/mm 2) Donor Tissue Loss (N=corneas)
Guerra et al. (2011)26 N=108        
DMEK 0.51+/- 0.44 logMar of the minimum angle of resolution units (20/65; range, 20/20 - 20/2000) 1-year: 0.07 1 +/- 0.09 logMar of the minimum angle of resolution units (20/24; range, 20/15 - 20/40); p<0.001 2980+/-252 (2514-3706) *at 1-year 1911+/-593 (range, 347-2976) at 1-year N=6 (4.2%)
McCauley et al. (2011)28 Median pre-op BSCVA was 20/40 (range: 20/25-20/400) 6-mo: median BSCVA was 20/25 (range: 20/15-20/40); 48%? 20/20; 74%? 20/25; 93%? 20/30; all ≥ 20/40 The median donor ECD=3140 cells/mm 2 (range: 2695-4630 cells/mm2) 6m FU, median ECD was 2121 cells/mm2 (range: 1204-4268 cells/mm 2 centimeter, n=30) Not statistically significant
DMAEK (N=40) Median OS n (%) n (%) n (%) NR
Study Probability of Rejection % at 1-y Probability of Rejection % at 2-yrs Eyes still followed without rejection (n) at 1-y Eyes still followed without rejection (n) at 2-yrs  
Anshu (2012)27          
DMEK (n=141) N=769 N=769 N=349 N=125

-

DSEK (N=598) 1 1 80 35  
PK (N=30) 8 12 246 79  

BCVA: best-corrected visual acuity; CI: confidence interval; DMEK: Descemet membrane endothelial keratoplasty; DSAEK: Descemet stripping automated endothelial keratoplasty; DSEK: Descemet stripping endothelial keratoplasty; EDM: endothelium-Descemet's membrane; ECD: endothelial corneal dystrophy; FU: follow-up; NR: not reported; OS: overall survival; PK: penetrating keratoplasty; BSCVA: best spectacle-corrected visual acuity; SD: standard deviation.

Section Summary: Descemet Membrane Endothelial Keratoplasty and Descemet Membrane Automated Endothelial Keratoplasty
Evidence for the use of Descemet membrane endothelial keratoplasty or Descemet membrane automated endothelial keratoplasty consists of several systematic reviews with overlapping studies, and several observational studies, some of which had no comparators and some of which compared Descemet membrane endothelial keratoplasty or Descemet membrane automated endothelial keratoplasty with Descemet stripping endothelial keratoplasty or Descemet stripping automated endothelial keratoplasty. Analyses in the individual studies and the meta-analyses consistently showed that patients receiving Descemet membrane endothelial keratoplasty or Descemet membrane automated endothelial keratoplasty experienced significantly better visual acuity outcomes postprocedure than patients receiving Descemet stripping endothelial keratoplasty or Descemet stripping automated endothelial keratoplasty, both short-term and through 1 year of follow-up. A large cohort study showed that intraoperative complications decreased as surgeon experience increased. Some studies reported similar complication rates between the procedures, some reported more complications with Descemet membrane endothelial keratoplasty than Descemet stripping endothelial keratoplasty, though the complications were not considered severe.

Femtosecond Laser-Assisted Endothelial Keratoplasty and Femtosecond and Excimer Laser-Assisted Endothelial Keratoplasty
Clinical Context and Therapy Purpose
The purpose of femtosecond laser-assisted endothelial keratoplasty and femtosecond and excimer laser-assisted endothelial keratoplasty is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as penetrating keratoplasty, in patients with endothelial disease of the cornea.

The question addressed in this evidence review is: Does the use of femtosecond laser-assisted endothelial keratoplasty or femtosecond and excimer laser-assisted endothelial keratoplasty improve the net health outcome for patients with endothelial disease of the cornea?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals with endothelial disease of the cornea. Diseases that affect the endothelial layer include Fuchs endothelial dystrophy, aphakic and pseudophakic bullous keratopathy (corneal edema following cataract extraction), and failure or rejection of a previous corneal transplant.

Interventions
The therapy being considered is femtosecond laser-assisted endothelial keratoplasty and femtosecond and excimer laser-assisted endothelial keratoplasty. Variations of femtosecond laser-assisted endothelial keratoplasty include femtosecond laser-assisted Descemet stripping automated endothelial keratoplasty.

Comparators
Comparators of interest include penetrating keratoplasty, microkeratome-prepared Descemet stripping automated endothelial keratoplasty, and manual Descemet membrane endothelial keratoplasty.

Outcomes
The general outcomes of interest are change in disease status, morbid events, and functional outcomes. Relevant outcome measures include visual acuity, endothelial cell densities, patient satisfaction or quality-of-life, and complications including graft rejection, graft dislocation, and need for rebubble procedures. Follow-up generally occurs through 1-2 years post-surgery.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

  1. To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs;
  2. In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
  3. To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
  4. Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Randomized Controlled Trials 
Ivarsen et al. (2018) conducted an RCT of ultrathin Descemet stripping automated endothelial keratoplasty or femtosecond-prepared Descemet stripping automated endothelial keratoplasty using the Ziemer LDV Z8 femtosecond laser.29 Outcome measures were planned after 1, 3, 6, 12 and 24 months with visual acuity, refraction, Scheimpflug tomography, whole eye scatter measurement and anterior optical coherence tomography. However, graft dislocation occurred in all patients randomized to femtosecond-prepared Descemet stripping automated endothelial keratoplasty which was managed with rebubbling. No patients with ultrathin Descemet stripping automated endothelial keratoplasty experienced graft dislocation. Additionally, all patients treated with femtosecond-prepared Descemet stripping automated endothelial keratoplasty had significantly poorer clinical outcomes compared with ultrathin Descemet stripping automated endothelial keratoplasty patients. After 3 months, visual acuity was scored as approximately 2.5 times worse. The optical scatter index was also significantly greater in patients receiving femtosecond-prepared Descemet stripping automated endothelial keratoplasty compared to ultrathin Descemet stripping automated endothelial keratoplasty at 3 months (12 [SD, 3; r ange, 8 to 16] vs. 5 [SD, 3; r ange, 2 to 9]). While the planned enrollment was set at 80, after 1 month only 6 patients were treated with femtosecond-prepared Descemet stripping automated endothelial keratoplasty and 5 patients received ultrathin Descemet stripping automated endothelial keratoplasty. Due to the large differences in observed clinical outcomes, no further patients were recruited and the study was suspended. Cheng et al. (2009) conducted a multicenter randomized trial in Europe that compared femtosecond laser-assisted endothelial keratoplasty with penetrating keratoplasty.30 Eighty patients with Fuchs endothelial dystrophy, bullous keratopathy, or posterior polymorphous dystrophy, and a best spectacle-corrected visual acuity less than 20/50 were included in the trial. In the femtosecond laser-assisted endothelial keratoplasty group, 4 of the 40 eyes did not receive treatment due to significant preoperative events and were excluded from the analysis. Eight (22%) of 36 eyes failed, and 2 patients were lost to follow-up due to death in the femtosecond laser-assisted endothelial keratoplasty group. One patient was lost to follow-up in the penetrating keratoplasty group due to health issues. At 12 months postoperatively, refractive astigmatism was lower in the femtosecond laser-assisted endothelial keratoplasty group (86%) than in the penetrating keratoplasty group (51%, with astigmatism of ≥3 D) ); however, there was a greater hyperopic shift in the femtosecond laser-assisted endothelial keratoplasty group than in the penetrating keratoplasty group. Mean best spectacle-corrected visual acuity was better following penetrating keratoplasty than femtosecond laser-assisted endothelial keratoplasty at the 3-, 6-, and 12-month follow-ups. There was greater endothelial cell loss in the femtosecond laser-assisted endothelial keratoplasty group (65%) than in the penetrating keratoplasty group (23%). With the exception of dislocation and need to reposition the femtosecond laser-assisted endothelial keratoplasty grafts in 28% of eyes, the percentage of complications was similar between groups. Complications in the femtosecond laser-assisted endothelial keratoplasty group were due to pupillary block, graft failure, epithelial ingrowth, and elevated intraocular pressure, whereas complications in the penetrating keratoplasty group were related to the sutures and elevated intraocular pressure.

Nonrandomized Studies
Sorkin et a.l (2019) reported 3-year outcomes of a retrospective, interventional study comparing femtosecond laser-assisted Descemet membrane endothelial keratoplasty with manual Descemet membrane endothelial keratoplasty in patients with Fuchs endothelial corneal dystrophy.31 Sixteen eyes of 15 patients were evaluated in the femtosecond-prepared Descemet membrane endothelial keratoplasty group for an average follow-up up 33.0 ± 9.0 months and 45 eyes of 40 patients were evaluated in the manual Descemet membrane endothelial keratoplasty group for an average follow-up of 32.0 ± 7.0 months. Best spectacle-corrected visual acuity was not statistically different at 1, 2, and 3 years post-surgery (p=0.849, p=0.465, and p=0.936, respectively). Rates of significant graft detachment were significantly higher in the manual Descemet membrane endothelial keratoplasty group than in the femtosecond prepared Descemet membrane endothelial keratoplasty group (35.6% vs. 6.25%; p=0.027). Rebubbling rates were also significantly higher in the manual Descemet membrane endothelial keratoplasty group (33.3% vs. 6.25%; p=0.047). Endothelial cell loss rates were significantly lower in the femtosecond prepared Descemet membrane endothelial keratoplasty group at 1 year (26.8% vs. 36.5%; p=0.042) and 2 years (30.5% vs. 42.3%; p=0.008), however, this trend was lost at 3 years (37% vs. 47.5%; p=0.057).The primary graft failure rate was 0% in femtosecond prepared Descemet membrane endothelial keratoplasty compared to 8.9% in manual Descemet membrane endothelial keratoplasty (p=0.565). While study authors speculate that the higher detachment and rebubbling rate in manual Descemet membrane endothelial keratoplasty may be related to retained Descemet tags and islands, this study is limited by its retrospective nature and nonrandomized design and cannot account for potential baseline differences in patient anatomy.32 Hosny et al. (2017) reported on results from a case series on 20 eyes (19 patients) that underwent a femtosecond prepared Descemet stripping automated endothelial keratoplasty.33 After 3 months of follow-up, patients experienced significant improvements in corneal thickness, measured by anterior segment optical coherence tomography. Visual acuity significantly improved each month of the 3-month follow-up, with the largest improvement seen in the first month postprocedure. Complications specific to the femtosecond laser-assisted procedure were thickness disparities causing protrusion of the posterior disc (n=6) and air trapping in the interface (n=2). The former complication was corrected by modifying procedure parameters, and the latter was corrected by venting of the air bubble. 

In a small retrospective cohort study, Vetter et al. (2013) found a reduction in visual acuity when the endothelial transplant was prepared with a laser (femtosecond laser-assisted endothelial keratoplasty=0.48 logMAR; n=8) compared with a microtome (Descemet stripping automated endothelial keratoplasty=0.33 logMAR; n=14).34 There was also greater surface irregularity with femtosecond laser-assisted endothelial keratoplasty. 

Femtosecond and excimer laser-assisted endothelial keratoplasty was also reported in a small case series (N=3) by Trinh et al. (2013).35

Section Summary: Femtosecond Laser-Assisted Endothelial Keratoplasty and Femtosecond and Excimer Laser-Assisted Endothelial Keratoplasty
Evidence for femtosecond laser-assisted endothelial keratoplasty consists of 3 small observational studies and 2 RCTs. One observational study showed improvements following the procedure, though there was no comparison group and the other showed worse outcomes with the laser compared with Descemet stripping automated endothelial keratoplasty. One RCT indicated that patients undergoing penetrating keratoplasty experienced better outcomes than patients in the femtosecond laser-assisted endothelial keratoplasty group after 1 year of follow-up. Complication rates were similar between groups. Another RCT reported better clinical outcomes and no instances of graft dislocation with microkeratome-prepared Descemet stripping automated endothelial keratoplasty compared to femtosecond prepared Descemet stripping automated endothelial keratoplasty. 

Evidence for the use of femtosecond and excimer laser-assisted endothelial keratoplasty consists of a single small case series described in a letter publication.

Summary of Evidence
For individuals who have endothelial disease of the cornea who receive Descemet stripping endothelial keratoplasty or Descemet stripping automated endothelial keratoplasty, the evidence includes a number of cohort studies, a randomized controlled trial (RCT), and systematic reviews. Relevant outcomes are change in disease status, morbid events, and functional outcomes. The available literature has indicated that these procedures improve visual outcomes and reduce serious complications associated with penetrating keratoplasty. Specifically, visual recovery occurs much earlier. Because endothelial keratoplasty maintains an intact globe without a sutured donor cornea, astigmatism or the risk of severe, sight-threatening complications such as expulsive suprachoroidal hemorrhage and postoperative catastrophic wound failure are eliminated. The Descemet Endothelial Thickness Comparison Trial (DETECT) RCT reported improved visual acuity outcomes with Descemet membrane endothelial keratoplasty compared to ultra-thin Descemet stripping automated endothelial keratoplasty. The evidence is sufficient to determine that the technology results in an improvement in the net health outcome. 

For individuals who have endothelial disease of the cornea who receive Descemet membrane endothelial keratoplasty or Descemet membrane automated endothelial keratoplasty, the evidence includes a number of cohort studies and systematic reviews. Relevant outcomes are change in disease status, morbid events, and functional outcomes. Evidence from the cohort studies and meta-analyses has consistently shown that the use of Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty procedures improve visual acuity. When compared with Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty, Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty showed significantly greater improvements in visual acuity, both in the short term and through 1 year of follow-up. The evidence is sufficient to determine that the technology results in an improvement in the net health outcome. 

For individuals who have endothelial disease of the cornea who receive femtosecond laser-assisted endothelial keratoplasty and femtosecond and excimer laser-assisted endothelial keratoplasty, the evidence includes a multicenter RCT comparing femtosecond laser-assisted endothelial keratoplasty with penetrating keratoplasty and an RCT comparing femtosecond-prepared Descemet stripping automated endothelial keratoplasty to microkeratome-prepared Descemet membrane automated endothelial keratoplasty. Relevant outcomes are change in disease status, morbid events, and functional outcomes. Mean best-corrected visual acuity was worse after femtosecond laser-assisted endothelial keratoplasty than after penetrating keratoplasty, and endothelial cell loss was higher with femtosecond laser-assisted endothelial keratoplasty. With the exception of dislocation and need for repositioning of the femtosecond laser-assisted endothelial keratoplasty, the percentage of complications was similar between groups. Complications in the femtosecond laser-assisted endothelial keratoplasty group were due to pupillary block, graft failure, epithelial ingrowth, and elevated intraocular pressure, whereas complications in the penetrating keratoplasty group were related to sutures and elevated intraocular pressure. Worsened visual acuity and a 100% graft dislocation rate was reported for femtosecond-prepared Descemet stripping automated endothelial keratoplasty compared to 0% in manually-prepared Descemet stripping automated endothelial keratoplasty. The evidence is insufficient to determine that the technology results in an improvement in the net health outcome. 

The purpose of the following information is to provide reference material. Inclusion does not imply endorsement or alignment with the evidence review conclusions.

Clinical Input From Physician Specialty Societies and Academic Medical Centers
While the various physician specialty societies and academic medical centers may collaborate with and make recommendations during this process, through the provision of appropriate reviewers, input received does not represent an endorsement or position statement by the physician specialty societies or academic medical centers, unless otherwise noted.

2013 Input
Clinical input was sought to help determine whether the use of endothelial keratoplasty for individuals with endothelial disease of the cornea would provide a clinically meaningful improvement in net health outcome and whether the use is consistent with generally accepted medical practice. In response to requests, clinical input was received from 3 specialty society-level response(s) and 3 academic medical centers.

For individuals who have endothelial disease of the cornea who receive Descemet membrane endothelial keratoplasty and Descemet membrane automated endothelial keratoplasty, clinical input supports this use provides a clinically meaningful improvement in net health outcome and indicates this use is consistent with generally accepted medical practice.

For individuals who have endothelial disease of the cornea who receive femtosecond laser-assisted endothelial keratoplasty and femtosecond and excimer laser-assisted endothelial keratoplasty, clinical input does not support a clinically meaningful improvement in net health outcome and does not indicate this use is consistent with generally accepted medical practice.

2009 Input
Clinical input was sought to help determine whether the use of endothelial keratoplasty for individuals with endothelial disease of the cornea would provide a clinically meaningful improvement in net health outcome and whether the use is consistent with generally accepted medical practice. In response to requests, clinical input was received from 3 specialty society-level response(s) and 2 academic medical centers.

For individuals who have endothelial disease of the cornea who receive Descemet stripping endothelial keratoplasty and Descemet stripping automated endothelial keratoplasty, clinical input supports this use provides a clinically meaningful improvement in net health outcome and indicates this use is consistent with generally accepted medical practice.

Practice Guidelines and Position Statements
Guidelines or position statements will be considered for inclusion in ‘Supplemental Information' if they were issued by, or jointly by, a U.S. professional society, an international society with U.S. representation, or National Institute for Health and Care Excellence (NICE). Priority will be given to guidelines that are informed by a systematic review, include strength of evidence ratings, and include a description of management of conflict of interest.

American Academy of Ophthalmology
In 2009, the American Academy of Ophthalmology (AAO) published a position paper on endothelial keratoplasty, stating that the optical advantages, speed of visual rehabilitation, and lower risk of catastrophic wound failure have driven the adoption of endothelial keratoplasty as the standard of care for patients with endothelial failure and otherwise healthy corneas. The 2009 AAO position paper was based in large part on an AAO comprehensive review of the literature on Descemet stripping automated endothelial keratoplasty.4 AAO concluded that “the evidence reviewed suggests Descemet stripping automated endothelial keratoplasty appears safe and efficacious for the treatment of endothelial diseases of the cornea. Evidence from retrospective and prospective Descemet stripping automated endothelial keratoplasty reports described a variety of complications from the procedure, but these complications do not appear to be permanently sight threatening or detrimental to the ultimate vision recovery in the majority of cases. Long-term data on endothelial cell survival and the risk of late endothelial rejection cannot be determined with this review.” “Descemet stripping automated endothelial keratoplasty should not be used in lieu of penetrating keratoplasty for conditions with concurrent endothelial disease and anterior corneal disease. These situations would include concurrent anterior corneal dystrophies, anterior corneal scars from trauma or prior infection, and ectasia after previous laser vision correction surgery.”

National Institute for Health and Care Excellence
In 2009, the National Institute for Health and Care Excellence released guidance on corneal endothelial transplantation.36 Additional data reviewed from the United Kingdom Transplant Register showed lower graft survival rates after endothelial keratoplasty than after penetrating keratoplasty; however, the difference in graft survival between the 2 procedures was noted to be narrowing with increased experience in endothelial keratoplasty use. The guidance concluded that “current evidence on the safety and efficacy of corneal endothelial transplantation (also known as endothelial keratoplasty is adequate to support the use of this procedure.” The guidance noted that techniques for this procedure continue to evolve, and thorough data collection should continue to allow future review of outcomes.

U.S. Preventive Services Task Force Recommendations
Not applicable.

Ongoing and Unpublished Clinical Trials
Some currently unpublished trials that might influence this review are listed in Table 7.

Table 7. Summary of Key Trials

NCT No. Trial Name Planned Enrollment Completion Date
Ongoing      
NCT00543660 Corneal Transplantation by DMEK - is it Really Better than DSAEK? 1000 Mar 2018
(unknown)
NCT00521898 Prospective Clinical Study on Descemet Membrane Endothelial Keratoplasty 100 Feb 2020
(unknown)
NCT03619434 Pilot Study of Femtolaser Assisted Keratoplasty Versus Conventional Keratoplasty 30 Dec 2021
(recruiting)
NCT02373137 Descemet Endothelial Thickness Comparison Trial (DETECT) 38 May 202 3
(ongoing)
NCT02470793 Technique and Results In Endothelial Keratoplasty (TREK) 100 Dec 2025
(recruiting)
Unpublished      
NCT00800111 Open-enrollment, Prospective Study of Endothelial Keratoplasty Outcomes 2593 Feb 2018
(completed)
NCT02793310 Corneal Transplantation by DMEK - is it Really Better Than DSAEK? 54 Feb 2019
(completed)

NCT: national clinical trial; DMEK: Descemet membrane endothelial keratoplasty; DSAEK: Descemet stripping automated endothelial keratoplasty.  

References 

  1. Eye Bank Association of America. 2019 Eye Banking Statistical Report. 2019; https://restoresight.org/wp-content/uploads/2020/04/2019-EBAA-Stat-Report-FINAL.pdf. Accessed February 9, 2021.
  2. Dunker SL, Dickman MM, Wisse RPL, et al. Descemet Membrane Endothelial Keratoplasty versus Ultrathin Descemet Stripping Automated Endothelial Keratoplasty: A Multicenter Randomized Controlled Clinical Trial. Ophthalmology. Sep 2020; 127(9): 1152-1159. PMID 32386811
  3. Woo JH, Ang M, Htoon HM, et al. Descemet Membrane Endothelial Keratoplasty Versus Descemet Stripping Automated Endothelial Keratoplasty and Penetrating Keratoplasty. Am J Ophthalmol. Nov 2019; 207: 288-303. PMID 31228467
  4. Lee WB, Jacobs DS, Musch DC, et al. Descemet's stripping endothelial keratoplasty: safety and outcomes: a report by the American Academy of Ophthalmology. Ophthalmology. Sep 2009; 116(9): 1818-30. PMID 19643492
  5. Stuart AJ, Romano V, Virgili G, et al. Descemet's membrane endothelial keratoplasty (DMEK) versus Descemet's stripping automated endothelial keratoplasty (DSAEK) for corneal endothelial failure. Cochrane Database Syst Rev. Jun 25 2018; 6: CD012097. PMID 29940078
  6. Marques RE, Guerra PS, Sousa DC, et al. DMEK versus DSAEK for Fuchs' endothelial dystrophy: A meta-analysis. Eur J Ophthalmol. Jan 2019; 29(1): 15-22. PMID 29661044
  7. Chamberlain W, Lin CC, Austin A, et al. Descemet Endothelial Thickness Comparison Trial: A Randomized Trial Comparing Ultrathin Descemet Stripping Automated Endothelial Keratoplasty with Descemet Membrane Endothelial Keratoplasty. Ophthalmology. Jan 2019; 126(1): 19-26. PMID 29945801
  8. Duggan MJ, Rose-Nussbaumer J, Lin CC, et al. Corneal Higher-Order Aberrations in Descemet Membrane Endothelial Keratoplasty versus Ultrathin DSAEK in the Descemet Endothelial Thickness Comparison Trial: A Randomized Clinical Trial. Ophthalmology. Jul 2019; 126(7): 946-957. PMID 30776384
  9. Hirabayashi KE, Chamberlain W, Rose-Nussbaumer J, et al. Corneal Light Scatter After Ultrathin Descemet Stripping Automated Endothelial Keratoplasty Versus Descemet Membrane Endothelial Keratoplasty in Descemet Endothelial Thickness Comparison Trial: A Randomized Controlled Trial. Cornea. Jun 2020; 39(6): 691-696. PMID 31939923
  10. Fuest M, Ang M, Htoon HM, et al. Long-term Visual Outcomes Comparing Descemet Stripping Automated Endothelial Keratoplasty and Penetrating Keratoplasty. Am J Ophthalmol. Oct 2017; 182: 62-71. PMID 28739420
  11. Heinzelmann S, Bohringer D, Eberwein P, et al. Outcomes of Descemet membrane endothelial keratoplasty, Descemet stripping automated endothelial keratoplasty and penetrating keratoplasty from a single centre study. Graefes Arch Clin Exp Ophthalmol. Mar 2016; 254(3): 515-22. PMID 26743748
  12. Wacker K, Baratz KH, Maguire LJ, et al. Descemet Stripping Endothelial Keratoplasty for Fuchs' Endothelial Corneal Dystrophy: Five-Year Results of a Prospective Study. Ophthalmology. Jan 2016; 123(1): 154-60. PMID 26481820
  13. Li JY, Terry MA, Goshe J, et al. Three-year visual acuity outcomes after Descemet's stripping automated endothelial keratoplasty. Ophthalmology. Jun 2012; 119(6): 1126-9. PMID 22364863
  14. Dapena I, Ham L, Melles GR. Endothelial keratoplasty: DSEK/DSAEK or DMEK--the thinner the better?. Curr Opin Ophthalmol. Jul 2009; 20(4): 299-307. PMID 19417653
  15. Rose L, Kelliher C, Jun AS. Endothelial keratoplasty: historical perspectives, current techniques, future directions. Can J Ophthalmol. Aug 2009; 44(4): 401-5. PMID 19606160
  16. Deng SX, Lee WB, Hammersmith KM, et al. Descemet Membrane Endothelial Keratoplasty: Safety and Outcomes: A Report by the American Academy of Ophthalmology. Ophthalmology. Feb 2018; 125(2): 295-310. PMID 28923499
  17. Singh A, Zarei-Ghanavati M, Avadhanam V, et al. Systematic Review and Meta-Analysis of Clinical Outcomes of Descemet Membrane Endothelial Keratoplasty Versus Descemet Stripping Endothelial Keratoplasty/Descemet Stripping Automated Endothelial Keratoplasty. Cornea. Nov 2017; 36(11): 1437-1443. PMID 28834814
  18. Pavlovic I, Shajari M, Herrmann E, et al. Meta-Analysis of Postoperative Outcome Parameters Comparing Descemet Membrane Endothelial Keratoplasty Versus Descemet Stripping Automated Endothelial Keratoplasty. Cornea. Dec 2017; 36(12): 1445-1451. PMID 28957976
  19. Li S, Liu L, Wang W, et al. Efficacy and safety of Descemet's membrane endothelial keratoplasty versus Descemet's stripping endothelial keratoplasty: A systematic review and meta-analysis. PLoS One. 2017; 12(12): e0182275. PMID 29252983
  20. Oellerich S, Baydoun L, Peraza-Nieves J, et al. Multicenter Study of 6-Month Clinical Outcomes After Descemet Membrane Endothelial Keratoplasty. Cornea. Dec 2017; 36(12): 1467-1476. PMID 28957979
  21. Tourtas T, Laaser K, Bachmann BO, et al. Descemet membrane endothelial keratoplasty versus descemet stripping automated endothelial keratoplasty. Am J Ophthalmol. Jun 2012; 153(6): 1082-90.e2. PMID 22397955
  22. van Dijk K, Ham L, Tse WH, et al. Near complete visual recovery and refractive stability in modern corneal transplantation: Descemet membrane endothelial keratoplasty (DMEK). Cont Lens Anterior Eye. Feb 2013; 36(1): 13-21. PMID 23108011
  23. Ham L, Dapena I, van Luijk C, et al. Descemet membrane endothelial keratoplasty (DMEK) for Fuchs endothelial dystrophy: review of the first 50 consecutive cases. Eye (Lond). Oct 2009; 23(10): 1990-8. PMID 19182768
  24. Dapena I, Ham L, Droutsas K, et al. Learning Curve in Descemet's Membrane Endothelial Keratoplasty: First Series of 135 Consecutive Cases. Ophthalmology. Nov 2011; 118(11): 2147-54. PMID 21777980
  25. Price MO, Giebel AW, Fairchild KM, et al. Descemet's membrane endothelial keratoplasty: prospective multicenter study of visual and refractive outcomes and endothelial survival. Ophthalmology. Dec 2009; 116(12): 2361-8. PMID 19875170
  26. Guerra FP, Anshu A, Price MO, et al. Descemet's membrane endothelial keratoplasty: prospective study of 1-year visual outcomes, graft survival, and endothelial cell loss. Ophthalmology. Dec 2011; 118(12): 2368-73. PMID 21872938
  27. Anshu A, Price MO, Price FW. Risk of corneal transplant rejection significantly reduced with Descemet's membrane endothelial keratoplasty. Ophthalmology. Mar 2012; 119(3): 536-40. PMID 22218143
  28. McCauley MB, Price MO, Fairchild KM, et al. Prospective study of visual outcomes and endothelial survival with Descemet membrane automated endothelial keratoplasty. Cornea. Mar 2011; 30(3): 315-9. PMID 21099412
  29. Ivarsen A, Hjortdal J. Clinical outcome of Descemet's stripping endothelial keratoplasty with femtosecond laser-prepared grafts. Acta Ophthalmol. Aug 2018; 96(5): e655-e656. PMID 29372934
  30. Cheng YY, Schouten JS, Tahzib NG, et al. Efficacy and safety of femtosecond laser-assisted corneal endothelial keratoplasty: a randomized multicenter clinical trial. Transplantation. Dec 15 2009; 88(11): 1294-302. PMID 19996929
  31. Sorkin N, Mednick Z, Einan-Lifshitz A, et al. Three-Year Outcome Comparison Between Femtosecond Laser-Assisted and Manual Descemet Membrane Endothelial Keratoplasty. Cornea. Jul 2019; 38(7): 812-816. PMID 30973405
  32. Singhal D, Maharana PK. RE: Three-Year Outcome Comparison Between Femtosecond Laser-Assisted and Manual Descemet Membrane Endothelial Keratoplasty. Cornea. Nov 2019; 38(11): e51. PMID 31414998
  33. Hosny MH, Marrie A, Karim Sidky M, et al. Results of Femtosecond Laser-Assisted Descemet Stripping Automated Endothelial Keratoplasty. J Ophthalmol. 2017; 2017: 8984367. PMID 28695004
  34. Vetter JM, Butsch C, Faust M, et al. Irregularity of the posterior corneal surface after curved interface femtosecond laser-assisted versus microkeratome-assisted descemet stripping automated endothelial keratoplasty. Cornea. Feb 2013; 32(2): 118-24. PMID 23132446
  35. Trinh L, Saubamea B, Auclin F, et al. A new technique of endothelial graft: the femtosecond and excimer lasers-assisted endothelial keratoplasty (FELEK). Acta Ophthalmol. Sep 2013; 91(6): e497-9. PMID 23607667
  36. National Institute for Health and Care Excellence (NICE). Corneal endothelial transplantation [IPG304]. 2009; https://www.nice.org.uk/guidance/IPG304. Accessed February 9, 2021.

Coding Section  

Codes

Number

Description

CPT 65756 Keratoplasty (corneal transplant); endothelial
  65757 Backbench preparation of corneal endothelial allograft prior to transplantation (List separately in addition to code for primary procedure)
ICD-10-CM H18.10-H18.13 Bullous keratopathy, code range
  H18.50-H18.59 Hereditary corneal dystrophies, code range
  T85.390A-T85.398S Other mechanical complication of other ocular prosthetic devices, implants and grafts, code range
  T86.840-T86.849 Complications of corneal transplant, code range
ICD-10-PCS   ICD-10-PCS codes are only used for inpatient services.
  08R83KZ, 08R93KZ Replacement, cornea, percutaneous, nonautologous tissue substitute code list (right and left cornea codes)
  08U83KZ, 08U93KZ Supplement, cornea, percutaneous, nonautologous tissue substitute code list (right and left cornea codes)
Type of Service Ophthalmology  
Place of Service Inpatient/Outpatient  

Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.

This medical policy was developed through consideration of peer-reviewed medical literature generally recognized by the relevant medical community, U.S. FDA approval status, nationally accepted standards of medical practice and accepted standards of medical practice in this community, Blue Cross Blue Shield Association technology assessment program (TEC) and other non-affiliated technology evaluation centers, reference to federal regulations, other plan medical policies, and accredited national guidelines.

"Current Procedural Terminology© American Medical Association. All Rights Reserved" 

History From 2013 Forward     

10/01/2021  Annual review, no change to policy intent. Updating rationale and references. 
10/01/2020  Annual review, no change to policy intent. Updating background, description, guidelines, coding, rationale and references. 
10/01/2019  Annual review, language regarding femtosecond laser-assisted surgery updated to state "investigational and/ or unproven and therefore is considered not medically necessary". Also updating rationale and references. 
10/09/2018  Annual review, no change to policy intent. Updating description, rationale and references. 
10/12/2017  Annual review, no change to policy intent. 
10/04/2016  Annual review, no changes to policy intent. Updating background, description, rationale and references. 
10/13/2015  Annual review, no change to policy intent. Updated background, description, rationale, references and coding. Added guidelines
10/08/2014  Annual review. Added to policy: Endothelial keratoplasty is NOT MEDICALLY NECESSARY when endothelial dysfunction is not the primary cause of decreased corneal clarity. Updated rationale, references and related policies. Added coding section. 
10/16/2013  Revisions and changes made in title, policy, and rationale. References updated.
Complementary Content
${loading}