October 2020


Research Highlight
Manual vs. femtosecond DMEK

by Maxine Lipner Contributing Writer

Intraoperative image following stripping of Descemet’s membrane demonstrating the symmetric descemetorhexis circle created by the femtosecond laser with no remnant Descemet’s islands or tags
Source: Nir Sorkin, MD

Using femtosecond DMEK in patients with Fuchs dystrophy may be an opportunity to avoid certain complications, according to Nir Sorkin, MD. New study results indicate that graft detachment and rebubble rates were significantly reduced using a femtosecond vs. manual approach.1
“Graft detachment is the most common complication after DMEK,” Dr. Sorkin said, adding that in the literature it averages about 13% but has been reported to go as high as 80% in some settings. Graft detachment with the femtosecond approach in the study was just over 6%, as was the need for rebubbling, compared with a rate of more than 30% using the manual approach, Dr. Sorkin said.
Investigators in the retrospective study wanted to explore the applicability of femtosecond DMEK to see if it could improve outcomes. Fuchs dystrophy patients who were undergoing cataract extraction in conjunction with either femtosecond or manual DMEK were included. “In order to prevent bias of comorbidities, we excluded people who had a complicated anterior segment or patients with previous keratoplasties, scarring, or any other visually significant comorbidity,” Dr. Sorkin said.

Weighing complication rates

The study indicated that there was no statistically significant difference in best spectacle- corrected visual outcomes between the study groups at up to 3 years postoperatively. Detachment, rebubbling, and endothelial cell loss rates were a different story. With the femtosecond approach, the rate of significant detachment was 6.25% vs. 35.6% with the manual technique. The rebubbling rate was 6.25% and 33.3% for femtosecond and manual DMEK, respectively. Endothelial cell loss rates at the end of the first year were 26.8% for femtosecond DMEK vs. 36.5% for manual DMEK; by 3 years, cell loss was 37% for the femtosecond approach and 47.5% for the manual technique.
Dr. Sorkin said lower complication rates are important. In the case of a detachment where air needs to be injected into the anterior chamber of the eye to make the graft attach, this necessitates having the patient lay down for 48 hours after the procedure.

Considering cell loss

The reason the cell loss tends to be less with the femtosecond technique is because the size of Descemet’s membrane removed from the patient is the same as the graft. With manual DMEK a larger amount of tissue has to be stripped from the patient’s eye to ensure that the donor tissue does not overlap the host’s, which could lead to possible detachment if not adhered correctly. It also removes more of the patient’s own endothelial cells in the process. “The more Descemet’s you remove from the periphery, the more it works against you because your purpose is not only to transplant the new cells, it’s also to preserve the patient’s cells,” Dr. Sorkin said. “They still have good cells in the periphery.”
With the manual approach, even if the amount of tissue removed from the patient is oversized by just 0.25 mm in conjunction with the graft that is typically 8.25 mm, this increases the affected stripped area by 6%. “That means that the cells on the graft have to redistribute themselves over a larger area,” Dr. Sorkin said. “They have to spread over an area that’s 6% larger. That effectively reduces the cell density by 6%.”
In manual DMEK, it’s also hard to create a perfect circle. Similar to tearing a piece of paper, it’s not uncommon for this to inadvertently take off in another direction, Dr. Sorkin noted, adding that this results in removing a greater amount of tissue from the patient than was intended with more cell loss.
The clinical advantages of the femtosecond approach translate into fewer graft detachments and a reduced rebubble rate, as well as potentially less endothelial cell loss. “We get these advantages without compromising the visual outcome,” Dr. Sorkin said. One downside to the femtosecond approach is the expense because you’ll need the laser and personnel to operate.
Currently, manual DMEK is more commonly performed. Though a good approach, Dr. Sorkin said with femtosecond technology showing fewer complications, it may come down to the patient’s preference and whether they are willing to risk, for example, a higher chance of a rebubble.
Overall, Dr. Sorkin views manual DMEK as an excellent surgery with great outcomes, both visually and mechanically, and low detachment rates. “The femtosecond laser gives us another tool, which can further reduce the complications, as well as improve long-term endothelial cell survivability,” he said.
Dr. Sorkin mentioned that their study group found reduced detachment and rebubbling rates in a similar study that included patients with a failed penetrating keratoplasty graft who underwent DMEK,2 and are currently conducting a prospective study to evaluate the femtosecond DMEK approach.

About the doctor

Nir Sorkin, MD
Cornea Fellow
Department of Ophthalmology
University of Toronto
Toronto, Canada


1. Sorkin N, et al. Three-year outcome comparison between femtosecond laser-assisted and manual Descemet membrane endothelial keratoplasty. Cornea. 2019;38:812–816.
2. Sorkin N, et al. Comparison of manual and femtosecond laser-assisted Descemet membrane endothelial keratoplasty for failed penetrating keratoplasty. Am J Ophthalmol. 2020;214:1–8.

Relevant disclosures

Sorkin: None


Sorkin: nir.sorkin@gmail.com

Manual vs. femtosecond DMEK Manual vs. femtosecond DMEK
Ophthalmology News - EyeWorld Magazine
283 110
220 215
True, 10