This article is more than 5 years old. Information may no longer be current.
Mini-bubble technique may enhance DSEK, DALK interfaces
Micro bubbles help separate interconnected stromal layers, providing a smooth interface similar to that of LASIK flaps.
Click Here to Manage Email Alerts
CHICAGO — Low femtosecond laser pulse energy and wide spot spacing may enable a novel method of donor preparation for ultra-thin Descemet’s stripping endothelial keratoplasty and deep anterior lamellar keratoplasty, a speaker said here.
The mini-bubble technique is an alternative to the big-bubble technique, Roger F. Steinert, MD, OSN Cornea/External Disease Board Member, said during the joint meeting of the American Academy of Ophthalmology and Asia-Pacific Academy of Ophthalmology.
If proven effective in vivo, the method may vastly improve donor and host preparation for DSEK and DALK, Steinert said.
“We can have the potential for a more reliable technique than big bubble. It also has implications for a potential method of donor preparation for a very thin DSEK or almost [Descemet’s membrane endothelial keratoplasty] type of tissue by central eye banks that could supply it to you already pre-cut and ready to go,” Steinert said.
Roger F. Steinert
The big-bubble technique involves a bare Descemet’s membrane, which obviates stromal interface issues. However, success rates in obtaining a big bubble are low, and compression wrinkles can be problematic, particularly in cases of keratoconus, Steinert said.
“The classic challenge of deep lamellar keratoplasty is that, classically, the optical outcomes are not that good,” Steinert said. “The reason is typically ascribed to the interface more than anything else.”
Manual dissection typically yields poor optical results, and microkeratomes are associated with problems controlling dissection depth and consistency, Steinert said.
Widely spaced pulses, low energy
Femtosecond laser cutting is different for LASIK flaps and DSEK or DALK because of structural differences between the anterior and posterior layers of the cornea, Steinert said.
“[Excellent interfaces] that we get with LASIK in the cornea typically don’t replicate when you get to the posterior half of the cornea,” he said. “The reason for this has been [attributed to] the difference in the lamellar structure of the anterior vs. the posterior cornea, where the posterior cornea is fully lamellar and does not have ridging fibers running front to back that help to tighten the tissue. … We’ve tried all sorts of things, including curved interfaces, and have never been able to get rid of the problem with ridges with lamellar dissection as you get deeper and deeper with the femtosecond laser in the cornea.”
LASIK flap cutting involves closely spaced femtosecond laser pulses at high energy. Steinert and colleagues discovered that proper dissection of the posterior stroma involves widely spaced pulses at low energy.
“That, then, generates a smoother lamellar dissection,” Steinert said.
Steinert and colleagues mounted human corneas on an artificial anterior chamber. They used an IntraLase iFS 150-kHz femtosecond laser (Abbott Medical Optics) to perform cuts with varied spot spacing and pulse energy of 0.6 mJ to 3.5 mJ. Laser passes were performed in alternating raster and spiral patterns.
Trypan blue staining, lissamine red staining and scanning electron microscopy were used to assess endothelial cell loss.
“So far, we have not done any live implants because we still feel like we are defining the technique,” Steinert said. “If this were commercialized, this would be automated to come from different directions. But right now, this is all a manual technique.”
Findings
Testing showed that the micro-bubbles helped peel apart interconnected stromal layers. Descemet’s stromal interface was smooth, similar to a LASIK flap, Steinert said.
“What we postulate is happening is that they find a lamellar layer and slowly peel and slowly separate that,” he said.
Tissue was manually peeled off, showing a smooth stromal bed. The interface was about 50 µm in front of Descemet’s membrane, possibly closer, Steinert said.
Wide spot spacing and low energy are critical to creating a smooth interface, he said.
In another case, central corneal thickness was 35 µm.
“That’s pretty much just Descemet’s membrane,” Steinert said.
Staining and microscopy showed no evidence of endothelial cell loss, he said. – by Matt Hasson