Effect of Prostaglandin Analogues on Corneal Biomechanical Parameters Measured With a Dynamic Scheimpflug Analyzer

Précis: Treatment with topical prostaglandin analogues (PGAs) induces increased corneal compliance in glaucoma eyes measured with a dynamic Scheimpflug analyzer. Purpose: The purpose of this study was to evaluate the effect of topical PGAs on the corneal biomechanical properties. Methods: We retrospectively studied the biomechanical parameters of 31 eyes of 19 consecutive patients with glaucoma measured using a dynamic Scheimpflug analyzer (Corvis ST) before and after initiation of treatment with topical PGAs. No patients had a history of glaucoma treatment before the study and no other antiglaucoma medication was used during the study period. Nine biomechanical parameters were evaluated before and 61.6±28.5 days (range: 21 to 105 d) after initiation of the treatment. The changes in the corneal biomechanical parameters before and after treatment were analyzed using multivariable models adjusting for intraocular pressure and central corneal thickness. The Benjamini–Hochberg method was used to correct for multiple comparison. Results: In multivariable models, PGA treatment resulted in shorter inward applanation time (P=0.016, coefficient=−0.151) and larger deflection amplitude (P=0.023, coefficient=0.055), peak distance (P=0.042, coefficient=0.131), and deformation amplitude ratio at 1 mm (P=0.018, coefficient=0.028). These associations consistently indicated increased corneal compliance (deformability) after PGA treatment. Conclusion: Topical PGAs resulted in greater corneal compliance, suggesting that the changes in the corneal biomechanical properties may lead to overestimation of the intraocular pressure–lowering effects.

O cular biomechanics recently has received a great deal of attention due to its relevance for several aspects of glaucoma practice such as intraocular pressure (IOP) measurements and risk assessments. Accurate measurement and IOP control are essential for glaucoma treatment, as IOP is the only modifiable risk factor for glaucoma. In clinical practice, the IOP is calculated from the corneal deformation when an external force is applied. Thus, the measured IOP is affected by the corneal properties. 1,2 It is well established that the central corneal thickness (CCT) affects the IOP measurements in that a thinner cornea results in lower measured IOP. 1 It is also known that a thin cornea independently increases the risk of developing glaucoma. 3 The corneal biomechanics exert an even greater impact on the measured IOP than the corneal thickness. 4 Therefore, evaluating the corneal biomechanics is essential in glaucoma practice both for accurate IOP measurements and risk assessment.
Despite its importance, assessments of corneal biomechanics in clinical practice have been challenging. With the recent clinical introduction of the ocular response analyzer (ORA; Reichert Ophthalmic Instruments, Buffalo, NY) and the dynamic Scheimpflug analyzer (Corvis ST; OCULUS, Wetzlar, Germany), in vivo assessments of corneal biomechanical parameters in a clinical setting are possible. 5 Using the dynamic Scheimpflug analyzer, we confirmed that the corneal biomechanical properties of glaucoma eyes are distinct from those of healthy control eyes in that glaucoma eyes are more compliant than healthy eyes. 6,7 Other factors such as age, myopia, and corneal thickness significantly affect the corneal biomechanical parameters. 8 The effects of these confounding factors should be controlled to accurately quantify the impact of any disease on the corneal biomechanics. Prostaglandin analogues (PGAs), the first-line glaucoma drugs with strong IOPlowering effects, 9 are among the potential confounding factors that affect the corneal biomechanics. PGAs cause changes in corneal tissue properties such as corneal thinning presumably via activation of matrix metalloproteinases. [10][11][12] The resultant corneal thinning may lead to larger corneal (elastic) compliance (smaller resistance to corneal deformation caused by external forces). Larger compliance may result in underestimation of the IOP.
Few studies have investigated differences in the corneal biomechanical parameters between untreated eyes and those treated with PGAs. [13][14][15] In the current study, we investigated the changes in the biomechanical parameters measured using the dynamic Scheimpflug analyzer after initiation of PGA treatment in previously untreated glaucoma eyes.

Participants and Design
This retrospective study evaluated the changes in the ocular biomechanical parameters induced by antiglaucoma medication. The medical records of consecutive patients with glaucoma who were examined using the Corvis ST from November 2016 to March 2019 were studied retrospectively. Participants were examined with the Corvis ST before and 61.6 ± 28.5 days (range: 21 to 105 d) after the initiation of the PGA treatment (66.8 ± 26.9 d after the first examination). The exclusion criteria included closed angles, secondary causes of glaucoma, history of antiglaucoma medication, history of any intraocular surgery, other intraocular diseases except for cataract, best-corrected visual acuity worse than 0.5, and low quality score of the Corvis ST measurement. The last is an indicator of the examination reliability determined by the device software based on edge detection, alignment, and pressure property.
Glaucoma was defined by the optic disc appearance (presence of neuroretinal rim thinning, excavation, notching, or characteristic retinal nerve fiber layer defect) based on fundus photography and corresponding visual field abnormalities detected during standard automated perimetry based on the Anderson and Patella criteria 16 of 1 or more of the following: a cluster of 3 or more nonedge points with a probability of < 5% including 1 point or more with a probability of < 1% on the pattern deviation map in at least 1 hemifield; a pattern standard deviation with a probability of < 5%; or glaucoma hemifield test results outside the normal limits.
This study was conducted as a part of the evaluation of a noninvasive, noncontact anterior-segment imaging study. The Institutional Review Board (IRB) of Osaka University Hospital approved the study protocol, which adhered to the tenets of the Declaration of Helsinki. The IRB waived the need for written informed consent because of the noninvasive and retrospective nature of the study. The study protocol was published on the department's Web site, and each participant provided oral consent before the initiation of the examination (www.med.osaka-u.ac.jp/pub/ophthal/ www/attend/research/index.html).

Examinations
The baseline demographic data such as age, sex, and ocular data such as refractive error, axial length (AL), and lens status were collected from the medical charts. All ocular examinations were performed within 3 months of the Corvis ST measurement. The AL was measured by laser interferometry (IOLMaster; Carl Zeiss Meditec, Jena, Germany).
All participants underwent corneal deformation response measurements using the Corvis ST. The high-speed Scheimpflug camera obtains 140 images in the horizontal section of the cornea and anterior chamber up to 8.5 mm in diameter with a resolution of 640×480 pixels and a speed of 4330 frames/second. This imaging system allows visualization of the corneal reaction to an air impulse.
The dynamic Scheimpflug biomechanical parameters were calculated in 3 defined states during deformation: inward applanation or applanation 1 (A1), outward applanation or applanation 2 (A2), and highest concavity (HC). The applanation phase was defined as the transition from a convex to a concave shape (A1) or from a concave to a convex shape (A2). HC was defined as the time at which total of both cornea and whole eye motion deformed maximally. The new analysis software (version 1.3r1538) of the dynamic Scheimpflug analyzer provides 38 parameters, including IOP, CCT, and 36 parameters that show the deformation responses.
Analyses of too many parameters may cause confusion and an issue with multiple comparisons. Therefore, we selected 9 relevant parameters, including 7 relevant parameters in our previous study 6,7,17 and 2 additional newly developed parameters based on reproducibility and correlation with clinical factors in previous studies. 18 The 7 relevant parameters in our previous study included the A1 time (A1T), the time of the first applanation; A1 velocity (A1V), the speed of the corneal apex at the first applanation; A2 time, the time of the second applanation; A2 velocity, the speed of the corneal apex at the second applanation; deflection amplitude (DeflA), the maximum corneal deflection calculated by subtracting whole eye motion from corneal deformation, in other words, cornea-only maximum motion; and the peak distance (PD), the distance between the 2 bending peaks created in the cornea at the HC. The whole eye movement is a slow linear motion in the posterior direction of the whole eye during air-puff tonometry. We selected DeflA as a parameter that shows the vertical movement of the corneal apex, instead of the deformation amplitude (DA), which has been investigated in many previous studies. The 2 newly developed parameters include the integrated inverse radius (Iradius) and deformation amplitude ratio at 1 mm (DAR1). 17 The Iradius is plotted over the duration of the air pulse and the integrated sum is calculated between the first and the second applanation events. DAR1 describes the ratio between the DA at the apex and the average DA measured 1 mm from the center. The 9 relevant parameters evaluated in this study are shown in Table 1.

Statistical Analysis
Descriptive statistics such as the mean, SD, and range were computed for the baseline clinical factors and Corvis ST parameters. The changes in the biomechanical parameters induced by antiglaucoma medications were evaluated using multivariable linear regression analyses. For each biomechanical parameter, multivariable models were fit with medication (before or after the initiation of medication), IOP, CCT, and AL as covariates. P-values < 0.05 were considered significant. All statistical analyses were performed using the statistical programming language R (The R Foundation for Statistical Computing, Vienna, Austria).

Descriptive Statistics
Thirty-one eyes of 19 patients with glaucoma were included. There was no significant difference between CCT before and after treatment. The participant characteristics are summarized in Table 2. All patients started treatment with PGA drugs; tafluprost (17 eyes, 12 patients), latanoprost (10 eyes, 5 patients) and travoprost (4 eyes, 2 patients).

Biomechanical Parameters
The values of the 9 clinically relevant biomechanical parameters at the pretreatment and posttreatment examinations are shown in Table 3. The results of the multivariable regression analyses evaluating the effect of antiglaucoma medication on the biomechanical parameters are shown in Table 4. Seven parameters differed significantly between the pretreatment and posttreatment examinations. Of these, the DAR1, HC DeflA, and PD were associated positively with treatment, whereas the A1T was associated negatively with treatment.
Other factors also were associated with the biomechanical parameters. The IOP was associated positively with the A1T and negatively with the A1 velocity, A2 time, HC DeflA, PD. The CCT was associated negatively with the DAR1 and Iradius. The AL was not associated significantly with any parameters.

DISCUSSION
The current results confirmed that topical PGAs significantly altered the biomechanical parameters. The changes in the biomechanical parameters suggested that the PGAs caused the cornea to be more compliant (less resistant), which potentially leads to underestimation of IOP. These drug-induced changes in ocular biomechanics have important clinical implications for the accurate IOP measurements and risk assessments in patients with medically controlled glaucoma.
Treatment with PGAs resulted in significant decreases of the A1T and increases of the HC DeflA, PD, and DAR1. These changes consistently indicated increased corneal compliance (decreased stiffness). Because the air pressure increases with time in the inward indentation phase, decreased A1T means lower pressure is necessary to applanate the cornea, thus suggesting larger compliance. Higher HC DeflA values, defined as the vertical movement of the cornea at the HC, also suggested less corneal resistance (larger compliance). PD is the horizontal length of the deformed cornea at the HC, which also is an indicator of corneal compliance. A higher DAR1 value means less resistance to deformation and thus greater compliance.  Coefficients and P-values were obtained from multivariable regression analyses. A1T indicates applanation 1 time; A1V, applanation 1 velocity; A2T, applanation 2 time; A2V, applanation 2 velocity; DAR1, deformation amplitude ratio at 1 mm; DeflA, deflection amplitude; HC, highest concavity; Iradius, integrated inverse radius; PD, peak distance; WEM, whole eye movement.
Increased corneal compliance (decreased corneal stiffness) caused by topical PGAs potentially leads to underestimation of the IOP. Applanation tonometry measures IOP by the force required to flatten a constant corneal area. This is based on the Imbert-Fick law that states that the pressure in a sphere is equal to the applanation force needed to flatten part of the sphere divided by the area flattened. 19 The assumptions to this theory are that the sphere is dry, infinitely thin, perfectly elastic, and perfectly flexible, none of which is true for the cornea. 19,20 As such, it follows from the laws of elasticity that the flattened area is decided not only by the force causing the flattening but also by the spherical stiffness. With increased compliance, the force required to deform the cornea becomes less even if the true IOP remains the same, thus leading to lower IOP readings. Underestimation of the IOP may be associated with insufficient IOP reduction, resulting in a worse visual prognosis. When assessing the treatment effect of PGAs, changes in biomechanical properties need to be considered.
In addition to the effect on the measured IOP, druginduced changes in the biomechanical properties may directly affect the susceptibility to glaucoma insult. Several previous reports have demonstrated correlations between the corneal biomechanical parameters and glaucoma development and progression. 6,7,21,22 The relationship between the corneal biomechanics and the risk of glaucoma may be related to an association in tissue properties between the cornea and sclera at the level of the optic nerve head. Some investigators have reported an association between the corneal biomechanics and the configurations of the lamina cribrosa. [23][24][25] Therefore, changes in the corneal biomechanical properties induced by topical PGAs could alter the susceptibility to glaucoma. We reported previously that glaucoma eyes are more compliant than healthy control eyes. 6,7 The current results suggested that PGAs further increase the compliance of glaucoma eye's cornea. 5 A previous study reported that there may be a relationship between the biomechanics of the cornea and the optic nerve head. 26 Therefore it is possible that PGAs also affect the susceptibility of the optic nerve head. Further research will be needed to clarify the influence of PGAs on the optic nerve head.
The current study underscored the dynamic nature of the corneal biomechanical properties. Significant changes in multiple corneal biomechanical parameters after topical PGA treatment imply that the corneal biomechanical properties may change in response to physiological, pathologic, and interventional events. In addition, we observed a large variation in the responses of the corneal biomechanical parameters induced by PGAs. Regular assessments throughout follow-up may be necessary to capture the dynamic changes in the corneal biomechanics.
Several previous studies have reported small but significant corneal thinning after PGA treatment. 11,12 In the current study, no significant change was seen in the CCT after initiation of topical PGAs, possibly due to the differences in the participant population and study design. A few previous studies that investigated the difference in corneal biomechanics between eyes with and without topical PGA treatment showed that glaucoma eyes treated with topical PGAs have higher corneal compliance than glaucoma eyes not treated with PGAs. 13,15 Only 1 study that longitudinally evaluated the influence of PGA treatment on corneal biomechanics found reduction in corneal stiffness after PGA treatment. 27 The results of the current longitudinal study basically agreed with those of previous comparative studies.
Several study limitations are worth noting. This was a retrospective observational study. The follow-up duration was relatively short (∼2 mo on average). All participants were Japanese, and most participants had normal tension glaucoma (NTG). It has been reported that eyes with NTG were more compliant and have different biomechanical properties compared with eyes with other types of glaucoma, 28 and the effect of PGA on the biomechanics of NTG may be different from that of other types of glaucoma. The limited number of participants in this study did not allow investigation of whether the drug-induced corneal biomechanical changes differed among different types of PGA formulations.
In conclusion, this study confirmed that topical PGAs significantly altered the ocular biomechanical properties. Increased compliance may potentially lead to underestimation of the IOP and altered susceptibility to glaucoma in eyes treated with topical PGAs. This study also suggested the dynamic nature of the corneal biomechanics and the necessity for regular examination of the biomechanics for accurate IOP measurements and improved risk assessments in glaucoma practice.