Influence of corneal parameters in keratoconus on IOP readings obtained with different tonometers

Authors


Abstract

Background

Accurate intraocular pressure (IOP) measurement is important but of unsure reliability in patients with keratoconus. Different types and models of tonometers are available. This study investigated the influence of corneal parameters on IOP readings obtained by a Goldmann applanation tonometer, a non-contact tonometer and a dynamic contour tonometer.

Methods

IOP readings with the Goldmann applanation, non-contact and dynamic contour tonometers were obtained from 52 patients with keratoconus and from 50 normal subjects and their corneal parameters were measured using a Pentacam.

Results

The mean IOP measurements in keratoconus obtained with the Goldmann applanation, non-contact and dynamic contour tonometers were statistically significantly different from the mean IOP measurements in the normal subjects (p < 0.0001; p < 0.005; p < 0.0001, respectively). In the keratoconus group, the thinnest corneal thickness (TCT), steepest keratometry, the corneal curvature (CC), central corneal thickness (CCT) and the posterior corneal curvature (PCC) had a significant effect on the Goldmann applanation and non-contact tonometers but not on the dynamic contour tonometer. In the control group, thinnest and central corneal thicknesses had a significant effect on findings with the Goldmann and non-contact tonometers but not on the dynamic contour tonometer. The corneal volume (CV) had no significant effect on the three tonometers in both groups.

Conclusions

The corneal parameters affecting the IOP readings of the Goldmann applanation tonometers, non-contact tonometers and the dynamic contour tonometers are not the same. While the Goldmann applanation and non-contact tonometers were significantly affected by the corneal parameters that were measured, the dynamic contour tonometer was not affected by any of these corneal parameters.

Keratoconus is a non-inflammatory disease of the cornea characterised by steepening, distortion and corneal ectasia. It is usually seen bilaterally and has a progressive nature. A decrease in corneal thickness and corneal volume (CV) and an increase in corneal curvature (CC) have been demonstrated.[1] These corneal alterations not only cause a reduction in visual acuity (VA) but also make accurate measurement of intraocular pressure (IOP) problematic. In addition to being a risk factor for glaucoma, IOP is an important parameter in the diagnosis and management of glaucoma.

The Goldmann applanation tonometer has been the world standard for IOP measurement since the first intervention with the device by Goldmann. Unfortunately, under- and over-estimations occur with this device when the corneal thickness is outside normal limits.[2]

Non-contact tonometers are basically applanation tonometers but they measure the IOP without a topical anaesthetic. The applanation force is achieved by a column of air that flattens the cornea in a very short time. The force for flattening divided by the area of applanation is the IOP. Like Goldmann tonometry, many studies have shown that corneal abnormalities influence the IOP measured by non-contact tonometers.[3-5]

The Pascal dynamic contour tonometer is a recently introduced tonometer, which does not rely on the principle of flattening or indenting the cornea. Thus, IOP measured with the dynamic contour tonometer is expected to be unaffected by corneal properties. The Pascal tonometer measures IOP with a hypothetical corneal contour that is achieved when the pressures on the anterior and posterior of the cornea are equal. The applied force gently fits the corneal surface to that hypothetical contour that counterbalances the force distribution generated by the IOP. The tonometer's curved tip takes the measurements by duplicating the corneal curvature with a sensor located in its centre. The dynamic contour tonometer can also measure ocular pulse amplitude by differentiating between diastolic and systolic IOP.[6]

The corneal parameters are the main source of error for all three devices;[2, 7, 8] however, there are limited data on whether IOP measurements obtained with the Goldmann applanation, non-contact and dynamic contour tonometers are influenced by the same corneal parameters in normal and keratoconic eyes. The purpose of this study was to determine the agreement between IOP readings obtained by these three tonometers and to determine the influence of corneal parameters such as central corneal thickness (CCT), thinnest corneal thickness (TCT), steepest keratometry (K), corneal curvature, corneal volume and posterior corneal curvature (PCC) on the IOP readings in keratoconic and normal eyes.

Methods

Data were obtained on 52 eyes of 52 patients with keratoconus and 50 eyes of 50 healthy subjects. In each of the subjects, one eye was randomly selected for the study. Informed consent was obtained from each individual before the study, which was conducted in accordance with the ethical standards of the Declaration of Helsinki and approved by our institutional ethics board.

Each subject underwent a full ophthalmic examination, including VA, slitlamp biomicroscopy and stereoscopic fundus evaluation on the slit-lamp using a 90 D lens. The diagnosis of keratoconus was made as described previously.[9] The diagnosis was also confirmed topographically using a Pentacam.

Patients with glaucoma, those suspected as having glaucoma and those with any disease that can affect IOP were excluded.

The order of the examination was keratometry and refractive error measurements using an ophthalmic unit (Topcon Delta Ophthalmic Unit IS-775, Tokyo, Japan) and IOP measurements with the three tonometers. The interval between the IOP measurements was approximately 15 minutes. Three consecutive IOP readings were obtained with each of the three tonometers and the means of the readings were recorded. First, the IOP of the subjects was measured using the non-contact tonometer (Topcon CT-80A Computerized Tonometer, Tokyo, Japan) and a Pascal dynamic contour tonometer (SMT Swiss Microtechnology AG, Port, Switzerland) and the mean of the three measurements was taken. Three readings were taken using the Pascal tonometer. Finally, the IOP was measured with the Goldmann applanation tonometer (Haag-Streit, Bern, Switzerland).

The Pascal tonometer uses a digital score in the quality assessment. IOP readings with a score of 1, 2, and 3 were recorded.

Following the IOP measurements, the corneal parameters were measured using a Pentacam rotating Scheimpflug camera (Oculus, Wetzlar, Germany).

All the statistical analyses were performed using SPSS for windows, version 15 (SPSS, Inc, Chicago, IL, USA). The Shapiro–Wilk test was used to test the distribution of the numerical data. Non-normally distributed data of the groups were compared by the Kruskal–Wallis H test. Multiple comparisons were carried out using the Conover test. The Chi-square test was used to compare categorical data. The relationships among the CCT, TCT, K, CC, CV, PCC and IOP measurements were assessed by Spearman's correlation analysis for each group. Bland–Altman plots were used to evaluate the agreement between the three tonometers. In all the statistical analyses, p < 0.05 was considered statistically significant.

Results

The mean age and standard deviation of the patients in the keratoconus group (21 women and 31 men) was 25.5 ± 1.36 years and in the normal group (25 women and 25 men) was 26.7 ± 1.13 years. The mean values of the corneal parameters are given in Table 1.

Table 1. Corneal parameters of the groups
 KeratoconusNormalp
  1. TCT: thinnest corneal thickness, K: steepest corneal keratometry, CC: corneal curvature, CCT: central corneal thickness, CV: corneal volume, PCC: posterior corneal curvature, μm: micron, D: dioptre.
TCT, μm455 (139–556)530 (238–592)0.000
K, D49.5 (42.1–67)43.8 (40.9–46.5)0.000
CC, D48.3 (40.3–64)43.1 (40.8–46)0.000
CCT, μm475.5 (384–589)549 (477–605)0.000
CV, mm357.2 (48–70)60.5 (47.4–71.5)0.001
PCC, D-7.0 (-10.3–5.3)-6.2 (-6.7–4.5)0.000

The mean IOP readings of the Goldmann, non-contact and dynamic contour tonometers are shown in Figure 1. The mean IOP measurements in the keratoconus group using the three tonometers were statistically significantly different from the mean IOP measurements in the normal subjects (p < 0.0001, p < 0.005, p < 0.0001, respectively). The median (minimum to maximum) IOP measurements using the Goldmann, non-contact and dynamic contour tonometers in the keratoconus group were 10 (4–16), 11 (5–22) and 15.4 (10–21.4) mmHg, respectively. The median (min-max) IOP measurements using the Goldmann, non-contact and dynamic contour tonometers in the normal subjects were 14 (9–17), 15 (8–21) and 16.8 (6.2–26.3) mmHg, respectively.

Figure 1.

Mean intraocular pressure readings of the groups obtained using the Goldmann applanation (GAT), non-contact (NCT) and dynamic contour (DCT) tonometers

There were correlations among the IOP measurements of the three tonometers (Table 2) both in the keratoconus group and in the normal group.

Table 2. The correlation coefficients for the Goldmann applanation tonometer (GAT), non-contact tonometer (NCT) and dynamic contour tonometer (DCT)
 KeratoconusNormal
GATNCTDCTGATNCTDCT
  1. * Correlation is significant at the 0.01 level
GAT10.694*0.766*10.507*0.456*
NCT0.694*10.621*0.507*10.376*
DCT0.766*0.621*10.456*0.376*1

Figures 2, 3 and 4 show Bland–Altman plots of the agreement among the Goldmann, non-contact and dynamic contour tonometers. Table 3 shows the mean and standard deviation differences and 95 per cent limits of agreement between the three tonometers.

Figure 2.

Bland–Altman plot showing the agreement between the non-contact (NCT) and dynamic contour (DCT) tonometers

Figure 3.

Bland–Altman plot showing the agreement between the non-contact (NCT) and Goldmann applanation (GAT) tonometers

Figure 4.

Bland–Altman plot showing the agreement between the dynamic contour (DCT) and Goldmann applanation (GAT) tonometers

Table 3. Results of Bland–Altman analyses
 

Mean ± SD

(mmHg)

95% LoA

(mmHg)

  1. NCT: non-contact tonometer, DCT: dynamic contour tonometer, GAT: Goldmann applanation tonometer, LoA: limits of agreement.
NCT-DCT-3.01 ± 3.86-10.59 to 4.55
NCT-GAT1.88 ± 2.90-3.81 to 7.57
DCT-GAT4.89 ± 3.12-1.22 to 11.02

The corneal parameters affecting the IOP measurements were not the same in the keratoconus and the control groups. In the keratoconus group, the steepest keratometry, anterior and posterior corneal curvatures and thinnest and central corneal thicknesses and had a significant effect on both the Goldmann and non-contact tonometry but not on the dynamic contour tonometry. In the control group, the thinnest and central corneal thicknesses had a significant effect on both the Goldmann and non-contact tonometric readings but not on dynamic contour tonometry. Corneal volume had no significant effect on the three tonometers in both groups (Table 4).

Table 4. The correlation coefficients for the Goldmann applanation tonometer (GAT), non-contact tonometer (NCT) and dynamic contour tonometer (DCT) and corneal parameters
 KeratoconusNormal
GATNCTDCTGATNCTDCT
  1. TCT: thinnest corneal thickness, K: steepest corneal keratometry, CC: corneal curvature, CCT: central corneal thickness, CV: corneal volume, PCC: posterior corneal curvature. * Correlations significant at 0.05 levels; ** Correlations significant at 0.01 levels
TCT0.282*0.330*0.1060.274*0.287*0.076
K-0.413*-0.519**-0.2120.103-0.172-0.071
CC-0.304*-0.477**-0.1740.027-0.021-0.042
CCT0.448**0.568**0.2720.285*0.318*-0.175
CV-0.0080.0920.0760.0740.224-0.087
PCC0.290*0.422**0.144-0.560.047-0.060

Finally, the keratoconus patients were divided into three groups based on mean K readings obtained from the Pentacam: mild (up to 47 D), moderate (between 47 and 52 D), and severe (52 D or greater), as described previously.[10] While there were statistically significant differences among the three keratoconus groups in the IOP readings obtained with the Goldmann and non-contact tonometers, there were no statistically significant differences among the IOP readings for the three keratoconus groups obtained with the dynamic contour tonometer.

Discussion

A major factor in the diagnosis of glaucoma and management is an accurate determination of IOP. Due to corneal changes in keratoconus, exact IOP measurement has been a challenge. Currently, two applanation tonometers, the Goldmann and non-contact, are the most widely used and accepted tonometers around the world; however, many previous studies have demonstrated errors with applanation tonometers in altered corneas.[11, 12] These errors led to a search for a new tonometer, which can measure IOP independently of the corneal parameters. The recently developed Pascal tonometer uses contour matching instead of applanation to measure IOP. Several studies with this dynamic contour tonometer in patients with keratoconus have demonstrated that IOP readings are not affected by corneal parameters.[13, 14] Conversely, other studies have suggested that corneal changes do affect IOP readings with the dynamic contour tonometer.[2, 9, 15]

In the current study, we investigated the effect of several corneal parameters on IOP readings with the Goldmann, non-contact and dynamic contour tonometers in patients with keratoconus and also in normal subjects, to identify difference in corneal factors affecting IOP measurements with the two conditions. We demonstrated that the corneal parameters that affect IOP measurement with these three tonometers in keratoconic and normal eyes are not the same. In keratoconic eyes, the thinnest corneal thickness, keratometric reading, corneal curvature, central corneal thickness and posterior corneal curvature significantly affected IOP measurements with the Goldmann and non-contact tonometers. In a similar study with keratoconic eyes, Read and Collins[14] also found that the average anterior axial curvature, the steepest anterior curvature and the average posterior axial curvature significantly influence IOP readings with the non-contact tonometer in patients with keratoconus. These findings are in agreement with ours.

An interesting finding in both studies is that the posterior corneal curvature affects the IOP readings. Anterior and posterior curvatures are higher in patients with keratoconus and a weaker correlation has been demonstrated between the anterior and the posterior curvature in clinical keratoconic eyes than in normal ones.[16] This finding can help explain why the posterior corneal curvature affects IOP readings in keratoconic eyes but not in normal ones. In contrast to Read and Collins' study,[14] we also investigated the effect of corneal parameters on normal eyes with the Goldmann, non-contact and dynamic contour tonometers. We found that the thinnest and central corneal thicknesses significantly influence IOP measurement in normal eyes using the Goldmann and non-contact tonometers. Three corneal factors—the steepest keratometric reading, corneal curvature and posterior corneal curvature—influenced the IOP readings of the Goldmann and non-contact tonometers in keratoconic eyes but not in normal eyes. In light of these results, it seems that a normal range of keratometry and the anterior and posterior curvature values do not influence IOP readings with applanation tonometers.

A number of recent studies have investigated the effect of central corneal thickness on IOP readings in normal and in keratoconic eyes. These studies have reported a significant association between central corneal thickness and IOP readings using the Goldmann and non-contact tonometers;[17-19] however, the effect of the thinnest corneal thickness on IOP readings was not previously studied. The thinnest corneal and central corneal thicknesses were the common corneal parameters in both of the groups.

In our study, the dynamic contour tonometer did not influence any of the corneal parameters that were measured in both the keratoconic and normal eyes. This finding suggests that the dynamic contour tonometric IOP readings are more reliable than those of the Goldmann and non-contact tonometers. In a clinical study of keratoconus, pellucid marginal degeneration and penetrating keratoplasty patients, Ozbek and colleagues[7] compared the IOP readings of the Goldmann, Tono-Pen and dynamic contour tonometers. They demonstrated that unlike the Goldmann and tono-pen, the dynamic contour tonometer was not affected by central corneal thickness. Papastergiou, Kozobolis and Siganos[20] showed that the dynamic contour tonometer was not significantly affected by central corneal thickness in normal and keratoconic eyes. The results of our study are consistent with these studies. In our study, the dynamic contour tonometer seemed to be unaffected by the corneal parameters that were measured but the IOP measurements in the keratoconus group were significantly different from those in normal eyes. The biomechanical parameters of the cornea provide a possible explanation for this difference. In a recent study, Bayer and colleagues[9] showed that the dynamic contour tonometer was significantly affected by corneal hysteresis and the corneal resistance factor in keratoconic eyes. They suggested that this result may be attributable to the design of the dynamic contour tonometer, which is not an applanation tonometer. The conformable design of the device is influenced by the changes in the cornea of keratoconic eyes, such as corneal geometrics and viscoelasticity.

We also divided the patients with keratoconus into three groups according to their corneal curvature. We found that more keratoconic changes lead to an underestimation of IOP with the Goldmann and non-contact tonometers but not with the dynamic contour tonometer. In contrast to some previous studies,[2, 20] we found a significant correlation of IOP measurement with the Goldmann and non-contact tonometers and corneal curvature. In a study with 30 eyes, Mark[21] demonstrated that IOP increases as the corneal curvature increases. Many other clinical studies support this finding.[22, 23] Our findings are also in general agreement with these previous studies; however, instead of a positive correlation between corneal curvature and IOP, we found a negative correlation.

Measurements of IOP using the Goldmann, non-contact and dynamic contour tonometers appear to be unaffected by corneal volume. Studies have demonstrated that corneal volume is significantly reduced due to corneal tissue loss during the progression of keratoconus.[10, 24] This finding is consistent with our results; however, in our study, corneal volume did not influence IOP readings with the three tonometers, perhaps because of more dominant variables, such as corneal hysteresis and the curvature having a more significant effect than the corneal volume.

In conclusion, we demonstrated that corneal parameters affecting IOP readings of the Goldmann applanation tonometers, non-contact tonometers and the dynamic contour tonometers are not the same. Although the Pascal dynamic contour tonometer measured the highest IOP both in keratoconic and in normal eyes, it was not affected by any of the corneal parameters in this study. The results of our study suggest that using a tonometer, which was not affected by corneal parameters is necessary equipment, especially in cornea departments.

Acknowledgement

Authors have no financial or proprietary interest in any instrument or product used in this study.

Ancillary