Re-bubbling risk: Intraop issues more significant than donor characteristics

Take-home message: In an analysis of 340 cases of Descemet membrane endothelial keratoplasty with a standardized technique by both experienced and novice surgeons, the re-bubbling rate was 10.6%.

By Cheryl Guttman Krader; Reviewed by Mark A. Terry, MD

Portland, OR-Donor tissue characteristics-including diabetic status-do not appear to affect Descemet membrane endothelial keratoplasty (DMEK) graft adherence when using eye bank-provided donor material, according to results of a study conducted by Mark A. Terry, MD.

“Our previous studies have shown that donor characteristics in tissue approved by the eye bank do not influence a variety of early and longer-term outcome measures in Descemet stripping automated endothelial keratoplasty (DSAEK),” said Dr. Terry, director, Corneal Services, Devers Eye Institute, Portland, OR. “However, the influence of donor characteristics in the outcomes of DMEK has not been analyzed in detail in a large series.

“Our analyses found no statistically significant differences between grafts that needed re-bubbling and those that did not in terms of traditional graft characteristics, including donor age, preoperative endothelial cell density (ECD), death-to-preservation time, death-to-use time, and pre-stripping-to use time,” he said.

In addition, the adherence rate was found similar when comparing grafts from diabetic and non-diabetic donors, he noted.

“Therefore, special requests by surgeons with high cell counts, younger or ‘fresher’ donors simply are not warranted,” Dr. Terry said. “We believe that corneal surgeons should feel comfortable accepting any tissue for DMEK that the eye bank has passed according to today’s standards.”

Donor characteristics

To investigate if donor characteristics influenced the adherence of DMEK grafts, Dr. Terry and colleagues undertook a retrospective review that included 340 consecutive eyes with a diagnosis of Fuchs’ corneal dystrophy undergoing DMEK performed with a standardized surgical technique.

“The cases also included those done by novice surgeons (cornea fellows) who were still in their DMEK learning curve but who used our standardized technique,” Dr. Terry said.

Eyes with glaucoma tubes, an anterior chamber IOL, aphakia, a history of trabeculectomy, vitrectomy, or failed penetrating keratoplasty were excluded from analysis.

When to perform re-bubbling?

The decision to perform re-bubbling was based on optical coherence tomography (OCT)-defined criteria or subjective findings. Specifically, re-bubbling was done if on OCT, more than 33% of the graft was separated at 1 week postoperatively; if there was progressive graft separation over 2 consecutive visits; or if central graft separation was present over the visual axis and persistent at the next visit.

In addition, re-bubbling was performed if there was any evidence of graft separation associated with reduced vision and the patient was asking for intervention rather than being willing to wait for spontaneous clearing.

Of the 340 eyes, 36 (10.6%) underwent re-bubbling.

“This rate is lower than that reported in other large DMEK series with multiple surgeons,” Dr. Terry said. “It is higher than the 6% rate we initially reported, however, and that is due to the relaxation of our re-bubbling criteria, which now includes a patient’s request for faster visual rehabilitation.”

The analyses comparing the grafts from cases with and without re-bubbling showed the mean donor age was about 66 years in both groups. Mean preoperative ECD in the no re-bubbling and re-bubbled grafts was 2,690 and 2,625 cells/mm², respectively.

Other comparisons between the non-re-bubbling and re-bubbled grafts showed that mean death-to-preservation time, death-to-use time, and pre-stripping-to-use time were all slightly longer in the group that was not re-bubbled, but none of the differences was statistically significant.

Dr. Terry and colleagues also undertook an analysis of whether diabetic status of the donor influenced graft adherence. That issue was investigated based on a paper by Greiner et al. that reported an increased risk of tissue tearing while preparing DMEK grafts from diabetic donors [Cornea. 2014;33:1129-1133].

“When the Greiner paper was published, we had 17 re-bubble cases in our DMEK series, and almost half were from diabetic donors,” Dr. Terry said. “Since diabetic donors seemed to be over-represented in the re-bubble group relative to their prevalence in the general population, we decided to reduce at that time our use of diabetic donors.”

In the final analysis of the total cohort of 340 DMEKs, however, diabetic donors accounted for only 22% of the re-bubbled grafts. The rate of re-bubbling among the 272 grafts from non-diabetic donors was 9.5% and the rate of re-bubbling of the 68 diabetic donors was 12%. The difference between groups was not statistically significant.

“We believe that the level of trauma at the time of surgery and interface obstacles are more important factors for adherence than diabetic status, donor endothelial cell count, age, and preservation time,” Dr. Terry said.

Standardized DMEK technique

Dr. Terry and colleagues described their DMEK surgical technique in a recently published article [Cornea. 2015;34:845-852]. They use donor tissue pre-stripped by an eye bank technician that is also marked by the eye bank with an “S” stamp on the Descemet’s side of the tissue to guide intraoperative orientation and eliminate the possibility of upside-down grafts.

Other novel elements include use of a Straiko glass injector (Gunther-Weiss) attached to a syringe, a modified Yoeruek tap technique with a shallow anterior chamber, and injection of 20% SF6 gas to achieve 70% to 90% fill at the end of the case, which they have found encourages prolonged graft support.

Mark A. Terry, MD

E: mterry@deverseye.org

This article is adapted from Dr. Terry’s presentation at the World Cornea Congress VII preceding the 2015 meeting of the American Society of Cataract and Refractive Surgery. Dr. Terry and his colleagues have no financial interests in any aspect of DMEK surgery or this study.