|Sep 15, 2006|
While the Goldmann applanation tonometer is a very accurate instrument for measuring IOP compared with previous instruments, ophthalmologists now know that central corneal thickness (CCT) is much more variable than was believed when the instrument was developed.
Leon W. Herndon, MD, advocates adjusting IOP for CCT alone because studies have shown that CCT is an important risk factor in glaucoma, while James D. Brandt, MD, counters that Goldmann tonometry is an inherently imprecise measurement, and that attempting to add precision to an imprecise measurement without knowing the impact of other factors is a step backward, not forward.
It is generally accepted that the Goldmann applanation tonometer gives a correct value for the pressure when CCT is approximately 520 μm. When the thickness is below normal, the applanation reading is too low, and when it is above normal, the reading is too high, with a 70-μm change in CCT corresponding to a difference of about 5 mm Hg of IOP, Dr. Herndon said.
While this relationship has been known for a number of years, it garnered widespread attention with publication of the results of the Ocular Hypertension Treatment Study (OHTS) in 2002. Although pachymetry was not part of the initial evaluation, ultimately CCT measurements were obtained for 82% of the patients in that study.
"Those with ocular hypertension had significantly thicker corneas than normal, and fully 24% of subjects had corneal thicknesses greater than 600 μm," said Dr. Herndon, associate professor of ophthalmology, Duke University Eye Center, Durham, NC.
Results of OHTS showed that the risk of developing primary open-angle glaucoma (POAG) was greatest among participants with the thinnest central corneas (Gordon et al. The Ocular Hypertension Treatment Study: baseline factors that predict the onset of primary open angle glaucoma. Arch Ophthalmol 2002;120:714-720.) In addition, patients with a CCT of 555 μm or lower had a threefold greater risk of developing POAG than participants with a CCT higher than 588 μm.
Since publication of the OHTS data, many other researchers have also studied the relationship between CCT and glaucoma. In one recent study published last fall,1 the authors looked at predictive models for 5-year risk of conversion to glaucoma, derived from OHTS results. Twenty-five percent of the 126 subjects in the longitudinal study developed glaucoma over follow-up.
"A scoring system was used to estimate the 5-year risk by adding up six predictive factors," Dr. Herndon said. "CCT carries the greatest weight on this conversion table. There's almost perfect agreement between the predictive risk as estimated from the point system and those estimated directly from the Cox regression equations."
One of the seminal papers on this topic, published in 1999,2 looked at determination of true IOP by using mathematical models to assess the influence of CCT and corneal radius of curvature.
"They showed that adjusting one's corneal radius of curvature would have very little influence on the applanating IOP," Dr. Herndon said. "However, varying one's corneal thickness would have a significant impact on the true IOP.
"There have been numerous published studies that have shown the importance of CCT in glaucoma," he concluded. "Risk models show that CCT is the most important factor."
Dr. Brandt began his portion of the debate by explaining that his opinion had changed with recent research findings and that he no longer supports adjusting pressure measurements for CCT in individual patients.
"I believe that the data now suggest that attempting to adjust or correct IOP based on corneal thickness alone is overly simplistic, may be harming patient care as much as it may help, and really doesn't represent a significant advance in glaucoma management," said Dr. Brandt, professor of ophthalmology and director, glaucoma service, University of California, Davis, Sacramento, CA.
When OHTS reported that CCT was a powerful independent predictive factor for the development of glaucoma, many clinicians began to use pachymetry routinely in their practices.
Their enthusiasm was based in part on the belief that measuring CCT would provide more accurate IOP measurements through an easily applied adjustment nomogram.
"What is not appreciated by most is that the OHTS investigators have attempted to correct IOP using the various published algorithms and that CCT does not drop out of the multivariate model of risk," Dr. Brandt said.
"There are several plausible explanations for this," he continued. "One, perhaps we don't have the right algorithm. Second, perhaps we're missing part of the picture to adjust IOP properly. And finally, perhaps corneal thickness is linked biologically to glaucoma risk in ways that we don't yet appreciate."
Dr. Brandt also stated that engineering models suggest that the impact of corneal thickness is smaller than has been thought, that it is fallacy to try to correct an already imprecise measurement, and that even if IOP could be corrected, it wouldn't be much of an advance.
He recommended an article by Liu and Roberts3 outlining a sophisticated bioengineering model that looks at corneal thickness, radius of curvature, and Young's Modulus, a measure of stiffness.
"Corneal curvature in the range seen clinically explains less than 2 mm Hg of variance, and corneal thickness over the range seen clinically explains a bit less than 3 mm Hg of variance," Dr. Brandt added. "Young's Modulus, however, explains more than 17 mm Hg of variance. To add additional complexity, the impact of Young's Modulus varies depending on the corneal thickness and true IOP. To summarize, curvature has a negligible effect, corneal thickness has a modest effect, and material properties have a huge effect. Furthermore, the impact of material properties varies enormously with corneal thickness and true pressure."
To see how this affects a real-life clinical experience, consider two individuals with the same pachymetry measurements: one with a stiff cornea and the other with a soft cornea.
"Each has a large error, but in opposite directions," Dr. Brandt said. "It is therefore possible to adjust IOP in the wrong direction in individuals with the same pachymetry measurement."
However, he added that the most compelling reason for not adjusting IOP based on CCT is that it is a fallacy to try to correct an inherently imprecise measurement.
"Attempting to adjust Goldmann measurements assumes that the underlying measurement is precise. It simply isn't," Dr. Brandt said. "Let's face it, Goldmann tonometry is an old mechanical technology, and in the real world, up to 50% of these machines are out of calibration.
"Adjusting tonometry with corneal thickness represents an attempt to add precision to a very noisy individual measurement," he continued. "In contrast, using large data sets can average out the noise of imprecision and explore the impact of corneal thickness, as we did in the OHTS."
Even if it were possible to correct pressure to add precision, this would not necessarily be a big advance, Dr. Brandt said.
"Glaucoma is unique among chronic diseases in that its primary risk factor, IOP, is measured randomly only a few times a year in most patients," he explained.
He contrasted this situation to the management of diabetes, noting that as their understanding of diabetes has progressed, physicians and patients have developed an excellent picture of glycemic control.
"I would argue that if back in the 1930s a technology to add precision to random, nontargeted glucose measurements had been introduced, it would not have represented much of an advance in patient care . . . and yet that is where we are today in the field of glaucoma," he added.
"My recommendation is that clinicians categorize corneas as thin, average, and thick, and integrate that determination into their global assessment of risk," Dr. Brandt said. "Attempting to be more precise than this is simply not supported by the data."
In his rebuttal, Dr. Herndon also referred to the paper on biomechanical properties of IOP measurement.
"The simulation results indicate that differences in corneal biomechanics across individuals may have a greater impact on IOP measurement errors than corneal thickness or curvature," Dr. Herndon said. He cited a passage from the paper:
"The magnitude of this influence has yet to be substantiated by experimental data, which is contingent on the establishment of in vivo methods for independent measurement of corneal biomechanical properties.
"In summary, measurement of CCT has been incorporated into clinical practice for years now. We should continue to measure this important variable and adjust treatment decisions accordingly, but as part of global risk assessment," Dr. Herndon said. "The bottom line is we treat the patient and not the cornea. We eagerly await further study on corneal biomechanics as related to measuring IOP."
Dr. Brandt reminded the audience that investigators studying CCT and other glaucoma risk factors are just beginning to scratch the surface of an incredibly complex problem.
"Data from the OHTS are telling us to look at the material properties of the eye, and I believe that they may underlie the basic pathophysiology of glaucoma."
1. Medeiros FA, Weinreb RN, Sample PA, Gomi CF, Bowd C, Crowston JG, Zangwill LM. Validation of a predictive model to estimate the risk of conversion from ocular hypertension to glaucoma. Arch Ophthalmol 2005;123:1351-1360.
2. Orssengo GJ, Pye DC. Determination of the true intraocular pressure and modulus of elasticity of the human cornea in vivo. Bull Mathematical Biol 1999;61:551-572.
3. Liu J, Roberts CJ. Influence of corneal biomechanical properties on intraocular pressure measurement: quantitative analysis. J Cataract Refract Surg 2005;31:146-155.