Skip to content
Cornea & External Eye

Descemet's Stripping Endothelial Keratoplasty (DSEK/DSAEK)

Descemet stripping endothelial keratoplasty (DSEK) is a type of corneal endothelial transplant performed for corneal endothelial dysfunction. It was reported by Price & Price in 2005 and established as a procedure involving stripping the recipient’s Descemet membrane (Descemetorhexis) and transplanting a donor graft consisting of posterior stroma, Descemet membrane, and endothelium 1). In 2006, Gorovoy introduced automated excision of donor tissue using a microkeratome, which became known as DSAEK (Descemet stripping automated endothelial keratoplasty) 1).

Features of DSEK/DSAEK

Transplant tissue: Posterior stroma (50–150 μm) + Descemet membrane + endothelium 1)

Donor preparation: DSEK = manual excision, DSAEK = microkeratome

Incision width: 4–5 mm (smaller when using BUSIN glide or Endo-Inserter)

Adhesion method: Intracameral air tamponade

Comparison with DMEK

DMEK: Transplant of Descemet membrane + endothelium only (approximately 15 μm)

DSAEK: Includes posterior stroma, making manipulation easier and unfolding more reliable

Visual recovery: DMEK is slightly superior, but the difference is narrowing 2)

Indications: DSAEK can also be used for eyes with iris abnormalities and complex eyes.

Subsequently, graft thickness has been reduced, leading to the development of ultra-thin DSAEK (UT-DSAEK: <100 μm) and nanothin DSAEK (<50 μm) 1). In the United States, as of 2022, approximately 62% of all corneal transplants are endothelial keratoplasty (EK), with DSAEK/DSEK and DMEK being the main types of EK.

General advantages of endothelial keratoplasty include: no sutures on the corneal surface, thus no suture-related complications; minimal induced astigmatism because the anterior corneal curvature is preserved; and resistance to trauma due to the small incision.

The indication for DSAEK is bullous keratopathy due to corneal endothelial dysfunction.

  • Fuchs endothelial corneal dystrophy (FECD): the most common indication
  • Pseudophakic bullous keratopathy (PBK): endothelial damage after cataract surgery
  • Bullous keratopathy after penetrating keratoplasty
  • Bullous keratopathy after argon laser iridotomy
  • Graft failure after previous EK

In principle, pseudophakic eyes are preferred. If there is scarring in the corneal stroma, visual improvement with endothelial keratoplasty is limited, so penetrating keratoplasty should be considered.

Iridocorneal endothelial (ICE) syndrome: In cases with extensive iris synechiae causing severe anterior chamber distortion, DSAEK after anterior chamber reconstruction with near-total iridectomy has been reported 4). In three eyes, all grafts remained clear at a mean follow-up of 53 months 4).

Congenital aniridia: DSEK can be performed even in the absence of a stable iris-lens diaphragm. Ultrathin DSEK for congenital aniridia has been reported to improve not only corneal edema but also the overall appearance of the cornea 5). In aniridia, DSEK is preferred over DMEK because of easier air bubble management 5).

DMEK is becoming the first choice for standard endothelial failure with good anterior chamber visibility. On the other hand, DSAEK is selected in the following complex eyes:

  • Iris abnormalities (aniridia, extensive peripheral anterior synechiae, large iris defects)
  • Aphakic eyes
  • Eyes after vitrectomy
  • Eyes with a history of glaucoma surgery
  • Cases with extremely shallow anterior chamber

Additionally, there have been reports of cases where forme fruste keratoconus underlying FECD became manifest after DMEK, causing monocular diplopia 6). When the contralateral eye was switched to DSAEK, the posterior corneal irregularity was corrected and visual disturbance did not occur 6). Because the stromal component of DSAEK stabilizes the posterior corneal curvature, DSAEK is preferable in cases with suspected keratoconus 6).

Q Which should I choose, DSAEK or DMEK?
A

DMEK is slightly superior in visual recovery, but DSAEK is safer in cases with iris abnormalities, aphakia, or after vitrectomy. In cases with potential keratoconus, the stromal component of DSAEK stabilizes the posterior cornea, making it advantageous. The choice depends on the surgeon’s experience and the condition of the eye.

The donor corneoscleral button is fixed on an artificial anterior chamber device, and a free cap of 300–350 μm thickness is created using a microkeratome. The remaining approximately 100 μm of posterior stroma, Descemet’s membrane, and endothelium are used as the donor graft. A trephine of 8 mm diameter is used to punch from the endothelial side. Pre-cut tissue from an eye bank is also often used.

Performed under retrobulbar anesthesia or Tenon’s capsule anesthesia. In cases with high vitreous pressure, preoperative intraocular pressure is lowered using a Honan balloon.

In cataract surgery for FECD patients, it is important to minimize the burden on the corneal endothelium.

  • Continuous curvilinear capsulorhexis (CCC) should be made smaller
  • Avoid hydrophilic acrylic intraocular lenses (IOLs) (risk of calcification)
  • Monofocal IOL is recommended: Multifocal IOLs are relatively contraindicated in FECD due to decreased contrast sensitivity
  • Consider the hyperopic shift of DSAEK (−1.00 to −1.25 D myopic target) in IOL power calculation
  • Soft-shell technique: Coat the corneal endothelium with a dispersive OVD and maintain the anterior chamber with a cohesive OVD. During phacoemulsification, the dispersive OVD remains on the endothelial surface, providing protection
  • Use low ultrasound energy and avoid nuclear manipulation near the cornea. Phaco-chop technique results in less endothelial cell loss than divide-and-conquer

Surgical Procedure

Descemetorhexis: Use a reverse Sinskey hook to create a circular descemetorhexis of the recipient’s Descemet membrane (under BSS irrigation)

Graft insertion: Insert into the anterior chamber through a 4–5 mm incision using a BUSIN glide or NS Endo-Inserter®

Centering: Center the graft by external corneal tapping and anterior chamber fluid flow

Wound closure: Suture with 2–3 stitches of 10-0 nylon

Air injection and interface fluid drainage

Air injection: Inject air directly under the graft using a 32G needle from the corneal limbus to sufficiently elevate intraocular pressure.

Interface fluid drainage: Thoroughly drain fluid from the interface through a stab incision.

Air tamponade: Completely replace the anterior chamber with air for about 10 minutes to adhere the graft.

Postoperative positioning: Maintain a supine position to stabilize the graft position with the air bubble.

Use atropine eye drops while air remains to prevent air-induced pupil block. Start steroid eye drops (prednisolone) every 3 hours immediately after surgery, then taper after 1 week. Monitor graft adhesion and corneal thickness improvement over time using anterior segment OCT.

4. Complications and postoperative management

Section titled “4. Complications and postoperative management”

Graft detachment: The most common complication after DSAEK. If detachment occurs, reinject air under the microscope (rebubbling). DSAEK tends to have a lower rebubbling rate than DMEK2).

Air-induced pupil block: Air bubbles in the anterior chamber obstruct the pupil, causing acute intraocular pressure elevation. Preventive measures include atropine eye drops, reducing air at the end of surgery, and performing an iridotomy at the 6 o’clock position. If it occurs, drain a small amount of air through a side port.

Primary graft failure: The graft never clears despite uneventful surgery. Occurs in about 0.1% of all corneal transplants; consider regrafting.

Postoperative intraocular pressure elevation: Occurs in 20–30% of cases after corneal endothelial transplantation. Differentiating between steroid-induced glaucoma and secondary glaucoma determines the treatment strategy.

Rejection: Occurs in about 5–10% after DSAEK, lower than full-thickness corneal transplantation (about 15%). Often resolves with steroid eye drops alone, but tapering must be done very slowly.

Endothelial cell loss: After DSEK, a median cell loss of 32% is expected in the perioperative period, followed by a linear decline of approximately 110 cells/mm² per year.

Donor-derived herpes simplex virus infection and systemic dissemination: The first life-threatening case has been reported of necrotizing hepatitis and hemophagocytic lymphohistiocytosis caused by transmission of herpes simplex virus-1 from donor cornea after DMEK7). A 73-year-old man developed symptoms 10 days after DMEK, and the herpes simplex virus-1 DNA level in the anterior chamber fluid was 220 million copies/mL7). Early intravenous acyclovir was effective7). Current donor screening does not mandate herpes simplex virus testing, but herpes simplex virus dissemination should be considered in unexplained systemic illness after transplantation7).

Q Does the graft function normally after rebubbling?
A

Yes, the graft usually functions well after rebubbling. Since the graft was floating in the aqueous humor, endothelial function is typically maintained after reattachment. However, if multiple rebubblings are needed, additional endothelial cell loss occurs, increasing the risk of graft failure.

In a multicenter RCT by Dunker et al. (54 eyes), the mean best-corrected visual acuity (BCVA) (logMAR) at 12 months postoperatively was 0.08 in the DMEK group vs 0.15 in the UT-DSAEK group, with no significant difference, but the rate of achieving 20/25 or better was 66% in the DMEK group vs 33% in the UT-DSAEK group, a significant difference (P=0.02)3). Endothelial cell density (12 months) was not significantly different between groups3).

ItemDMEKUT-DSAEK
Rate of achieving 20/2566%3)33%3)
12-month best corrected visual acuity0.08 logMAR3)0.15 logMAR3)
Hyperopic shift+0.22D3)+0.58D3)

In a meta-analysis by Sela et al. (376 eyes), the 12-month best corrected visual acuity was significantly better in DMEK (mean difference −0.06 logMAR; 95% CI −0.10 to −0.02)2). However, for DSAEK grafts thinner than 70 μm, there was no significant difference in best corrected visual acuity2). The rebubbling rate of DMEK is significantly higher than that of DSAEK (OR 2.76; 95% CI 1.46-5.22)2).

UT-DSAEK and the effect of graft thickness

Section titled “UT-DSAEK and the effect of graft thickness”

UT-DSAEK (thinner than 100 μm) provides better visual recovery than standard DSAEK, and even thinner nanothin DSAEK (thinner than 50 μm) may offer visual acuity comparable to DMEK. Thinner grafts improve the optical interface but increase surgical difficulty.

Endothelial keratoplasty (DSAEK) induces less astigmatism, provides better wound stability, and allows faster visual rehabilitation compared to penetrating keratoplasty1). However, some studies report no difference in graft rejection rates or improvement in best corrected visual acuity. The 5-year graft survival rate is considered equivalent between the two procedures1).

Q How much visual improvement can be expected with DSAEK?
A

With UT-DSAEK, approximately 33% of eyes achieve a best corrected visual acuity of 20/25 or better at 12 months postoperatively. With standard-thickness DSAEK, this proportion is slightly lower, but most patients achieve a functional visual acuity of 20/40 or better. If there is no corneal stromal scarring or other complications, good visual recovery can be expected.

Recovery of corneal endothelial pump function

Section titled “Recovery of corneal endothelial pump function”

Corneal endothelial cells use the Na⁺/K⁺-ATPase pump to expel water from the corneal stroma into the anterior chamber, maintaining corneal hydration at approximately 78%. In FECD and PBK, this pump function fails, leading to excessive water accumulation in the stroma, resulting in edema and opacity.

In DSAEK, Descemetorhexis removes the diseased Descemet’s membrane and dysfunctional endothelium, and a healthy donor endothelium, Descemet’s membrane, and posterior stroma are transplanted onto the host’s posterior corneal surface 1). Once the graft adheres via air tamponade, the donor endothelium exerts pump function, improving corneal edema.

The posterior stroma of the DSAEK graft acts as an added concave lens on the posterior corneal surface, causing a hyperopic refractive shift. Thinner grafts result in smaller hyperopic shifts, with DMEK producing the minimal shift 1).

7. Recent Research and Future Perspectives

Section titled “7. Recent Research and Future Perspectives”

Graft thinning is rapidly advancing. Meta-analyses suggest that for DSAEK grafts thinner than 70 μm, the difference in best-corrected visual acuity compared to DMEK may disappear 2). Nanothin DSAEK (less than 50 μm) may achieve outcomes equivalent to DMEK, but it remains an additive transplant, not a true replacement surgery.

Reports of near-total iridectomy combined with DSAEK for severe anterior chamber deformation in ICE syndrome 4) and ultrathin DSEK for congenital aniridia 5) indicate that DSAEK indications are expanding to complex eyes.

Cases have been reported where keratoconus latent in FECD becomes manifest after DMEK, and it has been confirmed that the stromal component of DSAEK stabilizes the posterior corneal curvature 6). Screening for such coexisting cases and optimizing surgical technique selection are future challenges 6).

Transmission of herpes simplex virus from donor corneas is rare, but a life-threatening complication of systemic dissemination has been reported for the first time 7). Current screening does not include herpes simplex virus testing, which is a topic for future consideration 7).

Q Can UT-DSAEK achieve visual acuity comparable to DMEK?
A

Meta-analysis data show that for ultrathin DSAEK grafts less than 70 μm, the difference in best corrected visual acuity compared to DMEK is no longer statistically significant. However, thinner grafts make intraoperative manipulation more difficult, so it is important to choose the surgical technique based on the facility’s experience and skill level.

  1. Lee WB, Jacobs DS, Musch DC, Kaufman SC, Reinhart WJ, Shtein RM. Descemet’s stripping endothelial keratoplasty: safety and outcomes: a report by the American Academy of Ophthalmology. Ophthalmology. 2009;116(9):1818-1830. doi:10.1016/j.ophtha.2009.06.021.
  1. Sela TC, Iflah M, Muhsen K, Zahavi A. Descemet membrane endothelial keratoplasty compared with ultrathin Descemet stripping automated endothelial keratoplasty: a meta-analysis. BMJ Open Ophthalmol. 2023;8:e001397.
  1. 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. 2020;127(9):1152-1159.
  1. Chaurasia S, Senthil S, Choudhari N. Outcomes of Descemet stripping endothelial keratoplasty combined with near total iridectomy in iridocorneal endothelial syndrome. BMJ Case Rep. 2021;14:e240988.
  1. Tsatsos M, Athanasiadis I, Ziakas N. Descemet stripping endothelial keratoplasty for congenital aniridia: an interesting and challenging story. Turk J Ophthalmol. 2022;52:208-211.
  1. Yung M, Li JY. Unmasking of subclinical keratoconus with Descemet membrane endothelial keratoplasty in Fuchs endothelial dystrophy. Taiwan J Ophthalmol. 2023;13:110-113.
  1. Azeem A, Baartman B, Conrady CD, Meier JL, El-Herte R. Herpes simplex virus dissemination with necrotizing hepatitis following Descemet membrane endothelial keratoplasty. BMC Infect Dis. 2023;23:465.

Copy the article text and paste it into your preferred AI assistant.