Use of Corneal hysteresis and Corneal Resistance Factor in Target Intraocular Pressure Estimation in Patients with Early Primary Open Angle Glaucoma

Purpose: To introduce a new method for estimation of the target intraocular pressure (TIOP) in naïve eyes with early primary open angle glaucoma (POAG) using Corneal hysteresis (CH) and corneal resistance factor (CRF). Methods: A prospective quasi-experimental study was conducted on naïve 90 eyes of 45 patients who were newly diagnosed with early primary open angle glaucoma (POAG). They were compared to 72 eyes of 36 normal subjects. The TIOP was determined for each eye. The IOP Goldmann (IOPg), IOP corneal-compensated (IOPcc), CH and CRF were estimated by Ocular Response Analyzer (ORA, Reichert) device. Measurements were taken for each patient prior to treatment and after 1, 3, 6, 9 and 12 months of receiving medications; either monotherapy or combination therapy. Results: For all patients, there was a signi�cant negative correlation (p < 0.05) between IOP, either IOPg or IOPcc, and CH while a signi�cant positive relationship (p < 0.05) existed between IOP and CRF. For patients with early POAG, The CH was signi�cantly increased (p ≤ 0.001) while CRF was signi�cantly decreased (p ≤ 0.001) when TIOP was achieved. At IOP levels higher than TIOP, CH value was lower than CRF with a signi�cant negative correlation between them in contrast to controls. This correlation was reversed on reaching TIOP and CH values became higher than CRF similar to controls. Conclusion: CH, CRF and IOP measured by ORA can be used for TIOP estimation. This provides us with a guide for assessing the effectiveness of medications introduced to patients with POAG.


Introduction
The principal goal of glaucoma management is to lower the intraocular pressure (IOP) either by conservative or interventional measures [1].Hence, the term "target IOP" has been established.It can be de ned as the IOP level at which the glaucoma stabilizes with treatment, and, thus, the patient's quality of life is preserved [2].Another de nition is the IOP level at which the sum of the health-related quality of life (HRQOL) from preserved vision and the HRQOL from not having treatment side effects is maximized [3].
In practice, the de nition of target IOP includes the highest IOP level in a given eye at which no clinically apparent optic neuropathy is likely to occur [4].
There are several factors to be considered when setting the target IOP.They include the stage of glaucoma, pre-treatment IOP, age and life expectancy, rate of progression during follow-up, presence of other risk factors such as exfoliations, the adverse effects and risks of treatment, and patient's preference [5].Other factors, such as corneal hydration, connective tissue composition, and bioelasticity may contribute to the response of the corneoscleral shell to the force applied during the IOP measurement [6].
Documenting the corneal biomechanical properties of the cornea is crucial as they are considered a signi cant risk factor for development and progression of glaucoma [7].
The Ocular Response Analyzer (ORA) (Reichert Inc., Depew, New York) is a popular commercially available medical device capable of measuring corneal biomechanics in vivo.It allows measurement of Goldmann-correlated IOP (IOPg), which is correlated to Goldmann applanation tonometry IOP, and corneal-compensated IOP (IOPcc) which is derived from both IOP and corneal biomechanical data.The ORA supplies two additional parameters that re ect the biomechanical properties of the cornea and demonstrate an inter-individual variation of the corneal hysteresis (CH) and corneal resistance factor (CRF) [8].Several studies reported that IOPcc is more accurate indicator of true IOP than IOPg as it is unrelated to central corneal thickness [7,9,10].Lower corneal hysteresis has been reported in previous research to be related to progressive glaucoma [7,[11][12][13].
Furthermore, CH and CRF were found to be lower among glaucoma patients than normal people and much lower in advanced stage of the disease [14,15].But no researches yet studied the relationship between CH and CRF with changes in IOP or the role of corneal biomechanics in target IOP estimation.Till now, there is no hard evidence for the concept or the methods used to determine the target IOP [16].
Therefore, our study aimed to describe a novel method for target intraocular pressure estimation in patients with early primary open angle glaucoma using the ocular response analyzer.

Patients And Methods
The study was conducted as a prospective quasi-experimental study after being approved by our Research Ethics Committee (REC) according to the principles of the Declaration of Helsinki [17].
A total of 125 eyes of 65 consecutive patients with elevated intraocular pressure (IOP) were initially enrolled in the study and assessed at Ophthalmology Outpatient Clinic, Suez Canal University Hospital in the period from January 2018 till December 2019.The study included patients older than 40 years who were newly diagnosed, still untreated, and prone to therapy either with monotherapy (travoprost 0.004% one drop once daily in the evening) or combination therapy (travoprost 0.004% + timolol 0.5% one drop once daily in the evening).All patients having early glaucomatous damage as de ned by Hodapp-Parrish-Anderson criteria for staging glaucoma using Humphrey visual eld testing [18] were included in the study.These criteria are: (1) Glaucoma hemi eld test outside normal limits on at least two consecutive occasions.(2) A cluster of 3 or more non-edge points in a location typical for glaucoma, all of which are depressed on pattern standard deviation (PSD) at P < 5% level and one which is depressed at P < 1% level, on two consecutive occasions.(3) Corrected pattern standard deviation (CPSD) that occurs in less than 5% of normal individuals on two consecutive elds.Glaucoma risk factors in our study cohort involved hypertension, diabetes mellitus and smoking.Patients suffering from any other types of glaucoma, patients on previous regular treatment or previous ocular surgery, patients with corneal pathology or refractive error of more than ± 4 diopters or aphakic patients were excluded from the study.Forty-ve (45) patients with 90 eyes out of 65 patients (125 eyes) were eligible for the study (POAG group) (Fig. 1).They were compared to 72 eyes of 36 normal individuals that came for regular checkup (control group).
Visual eld sensitivity was tested using the 30 − 2 Swedish Interactive Threshold Algorithm (SITA) [19] and standard automated perimetry (SAP) program of the Humphrey Field Analyzer.All participants had to have a reliable xation loss, false-positive, and false-negative error < 10%.
Corneal biomechanical properties and IOP were measured using Ocular Response Analyzer.Device settings and calibration were reviewed.Proper patient alignment was ensured.Readings were examined graphically, and those with poor-quality applanation signals (multiple applanation spikes, asymmetric signals or waveform score (WS) below 7 [20]) were discarded.Four good-quality measurements were recorded by the device for each eye and the results were averaged.Corneal biomechanical properties included corneal hysteresis (CH) and corneal resistance factor (CRF).The CH is derived from the difference between the inward and outward applanation pressures (P1-P2) and thought to represent the viscous-damping capacity of the cornea.The CRF is derived from the relationship between both P1 and P2 and central corneal thickness.It represents a measurement of corneal resistance [21].The intra-ocular pressure (IOP) measured by the ORA included two values, the rst IOP value is an estimate of Goldmann IOP (IOPg) and the second IOP value is an estimate of IOP corrected for the biomechanical properties of the cornea (IOPcc).
Patients were assessed at time of diagnosis (before start of the treatment) as a baseline assessment then regularly assessed after receiving medications at 1, 3, 6, 9 and 12 months follow up period.The target IOP was estimated separately for each eye of every patient according to European Glaucoma Society guidelines considering stage of glaucoma, IOP before treatment, age and life expectancy, rate of progression during follow up, side effects and risks of treatments, status of other eye and patient preference [22].Medications were given to the patients and the responses were followed up till IOPcc reach the predetermined target value.
A baseline peripapillary retinal nerve ber layer (pRNFL) thickness was measured using SD-OCT for documenting glaucomatous retinal nerve ber layer loss.Average nerve ber layer thickness; superior, inferior, temporal, and nasal quadrants thicknesses were assessed in the 12 clock hours vectors.
Analysis of optic nerve head (ONH) was done.Scan covers an area of 6x6 mm centered on the disc with a scan density of 512 (vertical) and 256 (horizontal) scans.Scans with motion artifact, poor quality or incorrect segmentation were rejected and repeat scans were acquired.
Statistical analyses were performed with SPSS software (SPSS for Windows Version 26.0; SPSS Inc., Chicago, IL).Data of all patients was normally distributed.Demographic characteristics and ocular response analyzer parameters were compared between POAG patients and controls using t-test.
Percentages were calculated for the qualitative data while mean and standard deviation were calculated for quantitative data.The Chi-square test was applied for the comparison of the categorical variables in both groups.For POAG patients, parametric analyses between means of IOP, CH and CRF, at baseline and at 1, 3, 6, 9 and 12 months of treatment, were evaluated with t-test.The correlations between ORA parameters were calculated using Pearson correlation test and presented on scatter plots.Changes in ORA parameters through the follow up period were presented as line graphs.P value less than 0.05 was considered statistically signi cant.

Results
Our study included patients with early primary open angle glaucoma (POAG group) who had early glaucomatous damage on visual eld (n = 90 eyes).
The mean target IOP determined according to the European glaucoma society guidelines was (15.50 ± 0.71 mmHg).The IOP measured by GAT (IOP GAT ) showed nonsigni cant difference with IOPg measured by ORA for study and control groups (p = 0.596 and 0.113 respectively).For the POAG patients, The IOPcc reached target levels after one month of treatment.From this point and forward, its measurement and IOPg, CH and CRF measurements were kept stable with nonsigni cant differences (p > 0.05) till the end of the one year follow up.Moreover, the IOPcc was signi cantly higher than the IOPg at baseline (p < 0.001), while after IOP control, the differences decreased to nonsigni cant values (p = 0.834) (Fig. 2).As indicated by the line graph in Fig. 2, at baseline, the CH is lower than the CRF.By reaching the target IOP, the CH was higher than the CRF which was similar to the results of the subjects of the control group, (Table 2).
Comparison of ORA parameters at baseline and after 12 months of treatment revealed signi cant decrease in IOPg, IOPcc and CRF (p ≤ 0.001) that was associated with signi cant increase in CH (p ≤ 0.001).Additionally, the IOPg and the IOPcc showed nonsigni cant differences when compared to the target IOP determined based on the European glaucoma society guidelines (p = 0.708 and 0. 879 respectively).
There was a signi cant negative correlation (p < 0.05) between IOP, either IOPg or IOPcc, and CH among POAG patients and controls.In contrast, a signi cant positive correlation (p < 0.05) existed between IOP and CRF in both groups (Fig. 3a-c).
Our study concluded that there were signi cant negative (p value < 0.05) correlations between CH and CRF at baseline before onset of treatment.This was in contrast to the normal controls whose CH and CRF showed a signi cant positive correlation (p value < 0.05).However, when target IOP was achieved with medications, the CRF showed an upward trend with increasing CH (p value < 0.05) similar to the controls (Fig. 3a-c).
Visual eld indices together with average, superior and inferior RNFL thicknesses revealed signi cant differences when compared to controls.However, following up the glaucoma progression in the study group showed nonsigni cant differences when comparing the 12 months changes in visual eld indices and pRNFL thickness from baseline, (Table 3).

Discussion
This study is a prospective quasi-experimental study that included early glaucomatous patients without treatment as a study group (90 eyes) versus normal subjects as control group (72 eyes).
Our observations con rmed that corneal biomechanical properties can be used for target IOP estimation.
Ocular response analyzer was performed for all patients before onset of treatment and repeated through one year follow up period at 1, 3, 6, 9 and 12 months after receiving medications.Ocular response analyzer is highly valuable device that measures corneal hysteresis (CH), a dependable predictor of glaucoma progression as evidenced by De Moraes et al. [23] and Chandrasekaran et al. [24] who reported that lower corneal hysteresis is associated with rapid glaucoma progression.Corneal hysteresis is an indication of the corneal biomechanical properties differing from corneal thickness or topography, which are geometrical characteristics.
Patients were assessed for achieving target IOP after taking medications after four weeks to ensure maximum drug effect and patients' response as recommended by the European Glaucoma Society guidelines [22].Then treatment was modi ed accordingly.
Eyes with lower CH had faster rates of visual eld loss than those with higher CH [7].Additionally, ocular response analyzer provides corneal compensated intraocular pressure (IOPcc), a better indication of the true pressure, proven to be less in uenced by corneal properties than that measured by Goldmann applanation tonometry (GAT) [9].This met agreement with our ndings as IOPcc was higher than the Goldmann IOP.So, we used the IOPcc as an indication for the real IOP for our glaucoma patients aiming to decreasing it to the target value.
In our study, the IOPg and IOP GAT were nearly similar with nonsigni cant differences.This nding con rmed the results of Ehrlich et al [25] who concluded that IOP GAT and IOPg showed good agreement and the IOPg may be substituted for GAT.So, we depended on IOPg in comparisons as an indicative for the Goldmann tonometry.
Our study reported that CH and CRF, at baseline, were lower among patients with POAG than normal people.This could be explained as the elevated IOP affects the dampening capacity and bioelasticity of the cornea.This nding was matched with the results of Karin R Pillunat et al [14].
Furthermore, the CH is inversely correlated to the IOP elevation while the CRF is directly correlated to the increase in IOP.Our analysis con rmed that CH and CRF are inversely correlated to each other as much as the IOP is higher than the levels of target IOP.Nonetheless, similar to control subjects, CH is directly proportionate to CRF when the target IOP is achieved.This indicates that the IOP controls the relationship between CH and CRF.
In practice, this correlation can con rm target IOP achievement.Moreover, our data suggested that when IOP is higher than its target levels, even if it is within average range, the CH is always lower than CRF in contrast to normal controls.But only at target IOP levels, the CH restore its normal value and be higher than the CRF similar to controls.
In other words, when CH became higher than CRF after treating POAG, the recorded IOP is consequently the target pressure.
The CH and CRF, at the end of follow up period and after IOP control, were signi cantly lower among our POAG patients than normal controls.In accordance to our ndings, Deol et al in 2015 reviewed corneal hysteresis and its relevance to glaucoma.They reported that glaucomatous damage and glaucoma progression is associated with low corneal hysteresis.In addition, corneal hysteresis is dynamically changing and may increase after IOP-lowering interventions are implemented [26].These results built on existing evidence which supports that corneal hysteresis is signi cantly lower in eyes with POAG than normal eyes [21,27].Sullivan-Mee et al [15] demonstrated that corneal hysteresis was signi cantly lower in patients with POAG than with ocular hypertension, glaucoma suspect, or normal subjects.Similar to our results, they concluded that corneal hysteresis could discriminate between the POAG and the normal controls.
The association between CH and IOP in our study have met agreement with Liang et al [28] who reviewed and summarized the characteristics of corneal hysteresis (CH) and its association with glaucoma.They concluded that CH can be used as a predictor of glaucoma risk.The CH also plays an important role in the diagnosis and treatment of glaucoma as it increases when the intraocular pressure decreases and decreases when the intraocular pressure increases [28].
Despite our ndings matched with various studies [29][30][31]; which concluded that the use of prostaglandin (PG) analogues has no effect on the corneal biomechanical properties, Meda et al [32] found a signi cant increase in CH and CRF with chronic PG analogues use.This mismatched results could be attributed to the fact that Meda et al evaluated patients already on treatment for up to seven years, while we evaluated naïve eyes at the start of treatment to control the glaucoma and achieve the target IOP.Accordingly, it seems that the changes in the corneal biomechanics were related to the change of the IOP rather than the effect of the drugs used.Furthermore, they used smaller sample in their study (n = 70) and exposed them to drug wash out in one eye for 6 weeks which is considered an ethical concern.
Stability of IOP and corneal biomechanics of our POAG patients through the one year follow up period was con rmed by nonsigni cant differences in visual led indices and pRNFL thicknesses when comparing baseline assessment to that at the end of the 12 months follow up.
Because our glaucoma patients were detected early and before visual elds defects occurred, the baseline pRNFL thickness depression was small and nonsigni cant.This can be explained by the results of Miki A et al. [33] who reported that eyes which did not developed visual eld defect had an approximately 2.5 times slower rate of RNFL thickness loss compared to eyes which developed them.Moreover, the mean rate of global RNFL loss in eyes which did not develop visual eld damage was − 0.82 µm/year.
We have to report that there is no evidence-based research reporting the relation between CH and CRF and target IOP in glaucoma management.
As our study had some limitations regarding enrollment of only patients with POAG, we recommend further research to evaluate the role of corneal biomechanics in controlling other types of glaucoma.As patients were observed for only one year, longer longitudinal follow-up to precisely assess the stability of corneal biomechanics with IOP control and to study the effect of treatment on CH and CRF are recommended.However, the current study was a prospective study introducing a new tool in determining target IOP in POAG patients.Also, to ensure accuracy, we depended on multiple measurement with four sequential puffs and the measures were averaged.Only one ophthalmologist carried out the measurements and no patients were lost to follow-up.

Figures
Figures

Figure 2 Changes
Figure 2

Table 1
summarized the preoperative clinical and demographic characteristics of our study groups.

Table 1
Demographic and clinical characteristics of the study participants before onset of treatment.
POAG: Primary open angle glaucoma; AL: Axial length; CCT: Central corneal thickness; MD: Mean deviation; BCDVA: best corrected distant visual acuity; pRNFL: Peripapillary retinal nerve ber layer.*ttest; ** Chi square test Age, gender, AL, CCT, MD and BCDVA are expressed as mean ± standard deviation (SD); glaucoma risk factors are expressed as frequencies Ocular response analyzer parameters were compared (IOPg, IOPcc, CH, CRF and WS) between POAG and normal controls at baseline and at the end of 12 months follow up period (Table2).

Table 2
Ocular response analyzer parameters between groups at baseline and after achieving target intra-ocular pressure (TIOP) by treatment.
POAG: Primary open angle glaucoma; IOPg: Goldmann intraocular pressure; IOPcc: cornea corrected intraocular pressure; CH: Corneal hysteresis; CRF: Corneal resistance factor; WS: waveform score.* Baseline ORA parameter of POAG patients compared to that of normal controls.** The ORA parameter of POAG patients at the end of follow up period (at target IOP levels) compared to that of normal controls.P value is based on t test.

Table 3
Characteristics of visual eld indices and peripapillary RNFL thickness for study and control groups.
RNFL: Peripapillary retinal nerve ber layer; POAG: Primary open angle glaucoma; MD: Mean deviation P1, Baseline comparison between POAG patients and control groups P2, POAG patients at baseline versus end of follow up period * t test