This article is best viewed in a PDF Format.
|
|
With accurate screening being the forefront of optometrical discussion, many recent studies have examined corneal thickness and its effect on the measurement of IOP. Its importance has also been highlighted by the growing numbers of patients now undergoing laser surgery in which biomechanical properties have been found to change.
In recent years, extensive research has been conducted on the sensitivity and specificity of using the current 'gold standard' Goldmann applanation tonometer (GAT), as a reliable screening tool for ocular hypertension. Its accuracy has been questioned, with many concluding that biomechanical features such as thickness, rigidity and hydration are indeed variable and can have a significant effect on IOP readings.
Goldmann designed his applanation tonometer to measure IOP for an average CCT of 520µm, an assumption which does not hold true given thickness can vary significantly in both normal and glaucomatous populations. A thinner cornea leads to a measured IOP that is less than the actual value and vice versa. So it appears there is no universally agreed correction factor that can be applied to give a true reading. Some researchers advise a nominal reduction in IOP for every 100µm reduction below the average central thickness, but these suggested values fluctuate widely between 2.0 and 7.5mmHg.1 Furthermore, the alternative theory of applying a linear corrective algorithm is also unfounded as indicated by findings of the ocular hypertension treatment study (OHTS),2 in which CCT was shown to have a weak association with IOP. In fact, this 2002 US study suggested that IOP alteration may be in the wrong direction. This is evident in patients with corneal dystrophies, such as Fuchs' endothelial dystrophy where CCT is generally high while corresponding GAT measurements tend to be lower than average.3 It is apparent that CCT alone cannot account for these variations in IOP.
Since the publication of the OHTS trial, several centres worldwide have researched the more complex material properties of the cornea, leading to the development of a new parameter called corneal hysteresis (CH). Dynamic tonometers are being designed to measure this viscoelastic behaviour, using it to correct the measured IOP reading accordingly. These include the Ocular Response Analyser (ORA), Pascal Dynamic Contour Tonometer (DCT) and, in a certain capacity, the I-Care tonometer.
This article considers the ORA (Reichert), comparing it with the traditional static air-puff tonometer NT-3000 (Nidek) in a small randomised sample study.
Theory
The ORA and NT-3000 both use a non-contact applanation system in which air pressure is used to apply a force to the cornea, but they differ mainly in the basic principles of measurement. The NT-3000 is essentially a 'static' tonometer which determines IOP from a single applanation of the cornea.
Comparatively, the ORA has been termed a 'dynamic' tonometer, involving a bi-directional system of two separate applanation events.
In both instruments, a collimated rapid jet of air of increasing intensity is initially directed toward the central cornea. The applanation event is monitored by an advanced infrared electro-optical system consisting of an emitting and photoreceptor diode arranged at oblique angles to the corneal surface. When the cornea is relaxed in a convex configuration, the diode receives no signal. At the point of full applanation, maximum infra-red light reflects off the corneal surface and into the receptor diode (Figure 1). With the NT-3000, this triggers the air pump to switch off. The pressure required for this full applanation is monitored by a transducer and recorded as the final IOP. Nidek recommends taking three readings to minimise any variability with cardiac pulse.
Unlike the NT-3000, once the precisely metered air pulse has achieved the first (inward) applanation using the ORA, the air puff does not stop but increases at a higher strength, deforming the cornea beyond full applanation to concavity (Figure 2). The air pump only stops once the full concave formation has been achieved, as monitored by the electro-optical system. As the pressure applied subsequently reduces, the cornea is able to resist the opposing force and once again retain its normal curvature. In the process, it passes through applanation (outward) for the second time. This produces two independent pressure values corresponding to the points of inward and outward applanations. The viscoelastic properties of the cornea which enable it to absorb and dissipate the applied kinetic energy, create a time delay between inward and outward applanations. This resistance of the air puff produces a disparity in pressure readings where the inward applanation occurs at a higher pressure to the outward event.
The readings are instantly displayed in a graphical plot which a trained operator can interpret to correctly identify a reliable measurement (Figure 3a). The graph displays three curves:
? A bell-shaped curve (green) representing the pressure applied to the cornea during the bi-directional applanation event
? The second curve (red) indicates the strength of signal detected by the electro-optical system. It characteristically has two spikes corresponding to the inward and outward applanations
? The final curve (blue) is essentially a filtered version of the 'signal strength' curve which is used to identify the optimum points of applanation.
The programme is able to display a maximum of four readings at any one time. Reichert recommends taking more than one measurement to ensure the reliability of the results as indicated by consistently 'strong' signal graphical displays.
However, consecutive low amplitude or jagged signals may be indicative of corneal pathology (Figure 4).
The ORA software has been designed to manipulate the raw data to produce four figures - two IOP measurements and two indices of corneal biomechanics (Figure 3b):
? IOPg: Traditional Goldmann-correlated IOP measured for historical reference. It is determined from the average of the two applanation IOP readings
? IOPcc: True IOP reading corrected for corneal biomechanics
? CH: Corneal hysteresis is the difference between IOP readings at the two applanation events. It is a unique measure and only has a weak link with central corneal thickness values
? CRF: Corneal resistance factor is a general indicator of the overall resistance of the cornea influenced by CCT, corneal topography and biomechanics. It has been shown to have a relatively strong correlation with central corneal thickness, and a weak correlation with the cornea-compensated IOP.
The ORA can also be set to measure CCT using an in-built 20MHz ultrasonic pachymeter.4
Method
A study of 30 eyes was undertaken to compare the IOP readings obtained using the ORA and the NT-3000. Patients aged 21-65 years of Caucasian, Afro-Caribbean, Asian and Indian origin were sampled from a normal population without any ocular pathology. Both ORA and NT-3000 measurements were taken on the same day within a two-hour period to eliminate variability in IOP. An average of two and three readings were taken for the ORA and NT-3000 respectively. Erroneous measurements indicated by poor signal curves in the ORA display and 'APL/BLK' on the NT-3000 CCTV screen were not included in this trial.
Analysis
The results of the study revealed several key features summarised below and shown in Table 1.
Mean IOPcc Vs NT-3000:
The overall mean IOPcc was found to be 1.7mmHg higher than the average NT-3000 readings. This is a 9.5 per cent difference, although it is important to note that some eyes showed a discrepancy of up to 30 per cent. Traditional models of IOP variation in the normal population display a Gaussian distribution curve with a positive skew towards higher pressures. Results of the ORA reinforce this finding as shown by the relative skew of +0.51. Interestingly, the NT-3000 readings produced a negative skew value of -0.04, indicating a shift to the lower IOP range. These trends are also reflected in the frequency chart (Figure 5) where the IOPcc curve peaks at a higher pressure relative to the NT-3000. In addition, the range of IOP readings were found to be significantly higher in ORA readings at 13.4mmHg compared to a smaller inter-patient variability of 7.7mmHg as measured by the NT-3000.
Mean IOPg vs NT-3000
As with the IOPcc comparison, IOPg readings were higher than NT-3000 measurements by an average of 6.3 per cent. A wide variation was again evident with individual percentage differences of up to 25 per cent produced by an apparent disparity in IOP of 5mmHg. This is slightly unusual as NCT instruments have been shown to correlate well with GAT in the normal IOP range. The relationship only tends to break down at higher IOP measurements above 30mmHg. Therefore, this limited study either questions the precision of the NT-3000 tonometer, or implies that IOPg readings do not correlate exactly with GAT. Further studies using a larger patient sample are required to investigate these preliminary hypotheses.
Mean IOPcc vs IOPg
IOPcc values were on average over 5 per cent higher than the corresponding IOPg readings at 15.7mmHg. If all other variables are assumed to be controlled, this apparent increase in IOP can only be attributed to the influence of corneal biomechanics. In a population study of 48 eyes surveyed in 2006, the ORA was found to consistently overestimate the IOP comparative to GAT measurements.5
CH and CRF
The ORA measurements of bio-mechanics have been tabulated in Table 2. The findings of this trial are not dissimilar to that conducted in 2006 on 207 eyes, in which a mean hysteresis and CRF of 10.7mmHg +/- 2.0 SD and 10.3mmHg +/- 2.0SD were derived respectively.6
Another noteworthy observation, was the variability in IOP measurements present between individual readings using each instrument for a given eye. The three values measured per eye using the NT-3000 differed by a maximum of 4mmHg with an overall average disparity of 1.62mmHg. Comparatively, the ORA showed greater intra-patient variability, with differences in individual eyes of up to 5.8mmHg and an average disparity of just under 2mmHg.
Discussion
Corneal hysteresis can essentially be considered as the 'damping' effect within the corneal tissues. A higher hysteresis is created by a greater difference in IOP between applanation events, indicating a 'stiffer' cornea. This study supports current literature, in which correction for corneal influence by and large generates higher true IOP measurements with a wider range. This has a significant bearing on glaucoma screening in general optometrical practice, with particular relevance to the diagnosis of normal tension glaucoma (NTG).
Corneal hysteresis and glaucoma screening
Glaucoma referrals are based on three main criteria - visual fields, optic disc appearance and IOP readings. The latter is the only modifiable risk factor in the decision making process, making accurate screening vitally important. Most high street practices at present use NCT instruments to screen for ocular hypertension. While this method has been employed relatively successfully since the early 1970s, we may be missing individuals with potential risk of NTG.
Patients with NTG have been found to have lower-than-average hysteresis with a wider range comparative to the normal population. In such circumstances, traditional static tonometers would measure falsely low IOP readings which may fall within what is considered to be the 'normal' range of pressures. This is supported by a 2006 clinical study of 24 NTG eyes, in which corneal corrected IOP readings were an average of 2.25mmHg higher than the respective IOPg measurements taken by ORA.7
Furthermore, the OHT study involved a multivariate analysis of 1,094 eyes over a five-year period, identifying thin central cornea as an independent risk factor in the development of glaucoma.8 Essentially, a patient with high IOP and a thicker cornea appears to be less at risk of developing glaucoma than a subject with high IOP and a thinner cornea.
More recent hypotheses advocate the use of CH as a direct screening tool in the diagnosis of glaucoma. A 2005 study indicated a 28 per cent lower average hysteresis of 8.74mmHg in glaucomatous eyes compared to the normal sample.9 This is positive evidence that corneal biomechanics may be indicative of the integrity of the lamina cribrosa, providing a measure of the susceptibility of nerve fibres to glaucomatous damage at this anatomical level.10
Corneal hysteresis and laser surgery
Extensive research has been conducted to evaluate changes in corneal biomechanics with laser surgery. The authors conclude a universal post-surgical reduction in CH,11 leading researchers to believe that alongside reduction in CCT, creation of a flap in Lasik also weakens the corneal structure affecting both hydration and viscoelasticity.
Traditional static tonometers are expected to produce lower post-surgical IOP readings as they do not account for this 'softening' of the cornea. This is supported by post-Lasik studies which have found an average reduction in IOP of 2-6mmHg using GAT compared to only 1mmHg in IOPcc.12
The ORA also has potential to be used as a screening tool to identify refractive surgery candidates, improving the prediction and control of successful outcomes. CH has proved an important criterion in assessing the risk of post-surgical complications such as corneal ectasia more effectively than traditional methods of corneal topography and CCT currently under use.
Corneal hysteresis and ocular disease
CH may be used as an indicator in both the diagnosis and management of ocular disease. Research into the variation of CH in normal and compromised corneae have shown a significant disparity, where eyes with corneal pathology tend to have a lower than average hysteresis. In a study of 182 eyes, patients with Fuchs' dystrophy and keratoconus had a lower mean hysteresis of 8.41mmHg and 8.74mmHg respectively13 comparative to the normal mean of 10.7mmHg. Also, general trends show that eyes with low CH may be at higher risk of developing corneal pathologies in later life, such as Fuchs' endothelial dystrophy or pellucid marginal degeneration.
Conclusion
The non-contact ORA tonometer is easy to use with little risk of corneal trauma and transfer of pathological microbes. It can also be conducted by non-optical staff who have undergone training to interpret the graphical displays. One major limitation is the error which may arise from variation in IOP with cardiac pulse and other short-term fluctuations, as measurements are only taken over a small time period of 20 milliseconds.
This may account for the large intra-individual variability present between readings for a given eye. It can be speculated that three or more readings may be required to nullify the influence of these factors. At present, Reichert recommends 'multiple readings' but does not specify an absolute figure this can only be determined through further trials.
The Pascal DCT has overcome this problem using a 'contact' method. The slit-lamp mounted device with curved tonometer head complements the central corneal shape, minimising corneal distortion and the corresponding corneal mechanical effects. Furthermore, it takes an average of 100 measurements per second to account fully for variation with the cardiac pulse.
The ORA has great potential for the future in the screening of laser surgery candidates, the diagnosis and management of glaucoma, as well as in the identification of corneal pathology. Its importance will become more pronounced over the next 10-20 years as laser surgery becomes increasingly popular and baby-boomers reach the age of higher risk for developing glaucoma. This allows a more detailed analysis of IOP and permits earlier referral and treatment, which may delay the onset of glaucoma in ocular hypertensives as indicated by findings of the OHT Study.14 This makes it all the more likely that dynamic tonometry will eventually replace its static counterparts, one day making instruments such as the NT-3000 purely a historical reference.
References
1 Kotecha A. Central corneal thickness and IOP, OT, 22, 2005.
2 Palmerg P. Answers from the Ocular Hypertension Treatment Study (editorial) Arch Ophthalmol, 2002120: 829-830.
3 Anderson J P et al. IOP and Corneal Biomechanical Metrics in Eyes with Keratoconus and Fuchs' Dystrophy Compared to Pachymetry matched Controls. ARVO abstract No 2267.
4 ORA Manual (Reichert 2006).
5 Martinez-de-la-Casa J M et al. Ocular Response Analyzer versus Goldmannn Applanation Tonometry for Intraocular Pressure Measurements. Investigative Ophthalmology and Visual Science, 200647:4410-4414.
6 Shah S et al. The use of the Reichert ocular response analyser to establish the relationship between ocular hysteresis, CRF and CCT in normal eyes. Contact Lens and Anterior Eye, 200629(5):257-62.
7 Luce D, Taylor D. Reichert ORA Measures Corneal Biomechanical Properties and IOP.
http://www.ocularresponseanalyzer.com/pressure.html (2006).
8 Gordon MO et al (OHTS Group). The Ocular Hypertension Treatment Study: Baseline Factors that Predict the Onset of Primary Open-Angle Glaucoma. Arch Ophthalmol, 2002120:714-720.
9 Shimmyo M. http://www.ophmanagement.com/article.aspx?article=86422.
10 Tanimoto S A et al. Change in Axial Length Following Surgical Decompression. A Surrogate for Material Properties of the Eye Wall. ARVO abstract No 4423, 2006.
11. Pan X et al. Comparing Lasik and Lasek: Indications of Biomechanical Response. ARVO abstract No 2728, 2005.
13 Shah S. Assessment of Biomechanical Properties of the Cornea using the Reichert ORA in Normal and Keratoconic Eyes. World Cornea Congress V, 2005.
14 Higginbotham E J at el (OHTS Group): The Hypertension Treatment Study: Topical medication delays or prevents primary open-angle glaucoma in African American individuals. Ophthalmology, 2004122:813-820.
Priya Dabasia is a visiting clinician at the City University Optometry Clinic