"DSAEK has revolutionized the management of endothelial diseases," began Dr. Jun, associate professor of ophthalmology, Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore.
Researchers have previously concluded that the posterior corneal curvature is a certain radius preoperatively. When a graft is placed, the cornea is actually thickened. The new radius and curvature of the posterior corneal surface then becomes smaller, and the radius of curvature is steeper with more of a refractive effect in the negative direction.
"The implications of this change in the radius of curvature are important because it suggests that you will have a hyperopic shift even with a planar graft," he said. "Even if you didn't have a graft with a thicker edge, simply by putting in additional tissue, you steepen the posterior corneal surface, and that creates negative refraction, which requires hyperopic correction with spectacles."
Effects of adjusting A constants
Dr. Jun and colleagues conducted a study to determine how to minimize the error between expected and achieved refractive outcomes following combined DSAEK, cataract extraction, and IOL implantation.
They included 30 consecutive eyes that underwent uneventful combined DSAEK triple procedure for Fuchs' dystrophy. All surgeries were performed by Dr. Jun, using the SRK/T formula and a proprietary lens (MA50BM, Alcon Laboratories) with an A constant of 119.4 (as provided by the manufacturer), and a standard graft size of 8.5 mm. For calculations, Dr. Jun and colleagues used the Holladay IOL software (Version 1.0, Holladay Consultants).
"This is a nice software that allows any surgeons doing cataract surgery to optimize their own A constants for particular lenses based on postoperative results," Dr. Jun said. "This allowed us to calculate a new A constant, 120.9, for these triple procedures."
For both A constants, they then compared the mean absolute error that resulted in a positive or negative direction from the preoperative predicted refraction to the actual postoperative refraction.
Using the original A constant for the lens, the mean absolute error from the target refraction was 1.09 D (±0.63 D). Using the new A constant of 120.9, the mean absolute error was reduced to 0.61 D (standard deviation: 0.40 D) (p = 0.036).
"With the new A constant, we had a lower absolute error for postoperative refraction, and a tighter standard deviation," he said. "The tighter standard deviation of our data is important because it suggests that an optimized A constant gives greater accuracy than just adding or subtracting a subjective fudge factor to the IOL calculations performed with the standard A constant."
Dr. Jun and colleagues then examined the number of eyes that were outside of the target refraction. Using the original A constant, the percentage of eyes that were outside of the target refraction by less than 0.5 D was only 20%. The percentage of eyes that were greater than 1 D outside the target refraction was 50%.
With the new A constant, more patients (44%) were within 0.5 D of the target refraction and less patients (16%) were more than 1 D outside the target refraction.
"Just as in cataract surgery alone, optimizing an IOL A constant using DSAEK triple procedures improved the predictability and accuracy of postoperative outcomes," Dr. Jun said. "With the advent of endothelial keratoplasty, with a minimal refractive change and quicker recovery, corneal transplant surgery can also be viewed as a refractive procedure.
"The same types of techniques that we use in cataract surgery for optimizing outcomes, therefore, can be applied here as well," he concluded.
Dr. Jun has no disclosures.