November 2008
Volume 49, Issue 11
Free
Glaucoma  |   November 2008
Variability of the Standard Reference Height and Its Influence on the Stereometric Parameters of the Heidelberg Retina Tomograph 3
Author Affiliations
  • Christophe Breusegem
    From the Department of Ophthalmology, Leuven University Hospitals, Leuven, Belgium; and the
  • Steffen Fieuws
    Biostatistical Centre, Catholic University Leuven, Leuven, Belgium.
  • Ingeborg Stalmans
    From the Department of Ophthalmology, Leuven University Hospitals, Leuven, Belgium; and the
  • Thierry Zeyen
    From the Department of Ophthalmology, Leuven University Hospitals, Leuven, Belgium; and the
Investigative Ophthalmology & Visual Science November 2008, Vol.49, 4881-4885. doi:https://doi.org/10.1167/iovs.08-2331
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Christophe Breusegem, Steffen Fieuws, Ingeborg Stalmans, Thierry Zeyen; Variability of the Standard Reference Height and Its Influence on the Stereometric Parameters of the Heidelberg Retina Tomograph 3. Invest. Ophthalmol. Vis. Sci. 2008;49(11):4881-4885. https://doi.org/10.1167/iovs.08-2331.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. To investigate the variability of the standard reference height (SRH) in consecutive Heidelberg Retina Tomograph 3 (HRT3) examinations and its influence on five main stereometric parameters.

methods. HRT3 results of all patients attending our glaucoma center from August to October 2007 were retrospectively reviewed. Only one eye per patient with at least three consecutive HRTs and a quality control label of acceptable or better was selected. An SRH difference ≥10% compared with baseline was considered as excess SRH variability.

results. A review of 641 patients yielded 150 patients (150 eyes) who met the inclusion criteria, representing a total of 556 examinations. The mean total number of HRTs per patient was 3.7 (range, 3–7), and the mean follow-up time was 3.5 years (range, 1.4–6.7). The mean proportion of consecutive HRT3 examinations with intraindividual excess SRH variability was 46% (95% CI, 40–52), whereas the mean intraindividual excess SRH variability was 24% (95% CI, 20–28). The median absolute SRH difference was 8.6% (interquartile range, 3.9%–16.2%). Furthermore, 55.5%, 75.3%, 74.7%, 0.0%, and 19.4% of the variability in rim area, rim volume, retinal nerve fiber layer thickness, cup shape measure, and height variation contour, respectively, could be attributed to SRH variability.

conclusions. There is considerable SRH variability between HRT3 examinations of the same eye, and this could explain more than half the variability of the parameters RA, RV, and RNFL. These findings indicate that changes of HRT3 parameters should be considered with caution when excess SRH variability is present.

Glaucoma is a chronic optic neuropathy characterized by optic nerve damage, visual field defects, and often also elevated intraocular pressure. 1 2 Changes in the optic nerve head morphology are of major importance in glaucoma progression, and these often precede the onset of visual field defects. 3 4 5 Optic disc biometry by means of scanning laser ophthalmoscopy with the Heidelberg Retina Tomograph (HRT; Heidelberg Engineering, Heidelberg, Germany) has been widely used to assess structural changes of the optic nerve head and is expressed in different descriptive parameters. 6 7 8 Other imaging techniques used for this purpose are optical coherence tomography, scanning laser polarimetry, or stereophotographs. 9 10 11 12  
Optic disc changes with the HRT3 can be assessed by either of two modalities, stereometric parameters or topographic change analysis (TCA). The first method depends on the standard reference height (SRH) and a manually drawn contour line essential for cross-sectional studies, whereas the second method is less dependent on these. 7 8 During each HRT3 scan, an SRH is automatically determined. The reference height is expressed in micrometers and represents the location of the reference plane relative to the mean height of the peripapillary retinal surface. 8 This reference plane is required to separate the cup and the neuroretinal rim of the disc. Structures located more deeply than the reference plane are considered cup, and structures located above the reference plane are considered rim. Ideally, the reference plane should be located at the inner surface of the sclera and beneath the retinal nerve fiber layer. As an approximation of this, the reference plane is automatically defined such that it is located 50 μm deeper than the average surface height of the papillomacular bundle at the contour line (350°–356°) and is parallel to the peripapillary retinal surface. The papillomacular bundle was chosen because it is assumed to be the last to change in the process of glaucomatous structural damage to the optic nerve head. 
Variability of the SRH does occur in consecutive HRT3 measurements. 8 Therefore, we wanted to investigate the frequency and magnitude of SRH variability and its impact on the variability of the stereometric HRT3 parameters. 
Methods
This retrospective noninterventional cohort study required no submission for approval according to the policies of our local ethics committee. However, this study adhered to the tenets of the Declaration of Helsinki. 
The HRT3 technique was used for scanning laser ophthalmoscopy of the optic nerve head to detect morphologic changes over time characteristic of glaucoma. HRT3 outcomes of patients with glaucoma consecutively attending the glaucoma division of the University Hospitals of Leuven from August to October 2007 were reviewed. Patients with a minimum of three consecutive mean HRT3 scans with a quality control label of acceptable or better—that is, with a scan SD less than 50 μm—were included. Only one eye per patient was selected. The right eye was considered first, and, if the inclusion criteria were not met for this eye, the left eye was then considered. If these criteria were not applicable to either eye, the patient was excluded. A difference of 10% or more in the SRH compared with the baseline was considered excess variability. This cutoff was based on previously published work and was informally confirmed by the HRT3 manufacturer. 13 For each patient included, all parameters of baseline and consecutive HRT3 examinations were exported from the HRT software (HRT Eye Explorer, version 3.0). The five stereometric parameters considered to be the main parameters by the HRT3 manufacturer were subsequently analyzed for their global (circular) values, including rim area (RA), rim volume (RV), mean retinal nerve fiber layer thickness (RNFL), cup shape measure (CSM), and height variation contour (HVC). 
The proportion of intraindividual HRT3 examinations with excess SRH variability, the mean intraindividual excess SRH variability, and the median absolute difference in SRH for all follow-up HRT3 examinations compared with the baseline were calculated, together with a 95% confidence interval (CI) or an interquartile range (IQR). To quantify the stability of the SRH values within a subject, the within-patient variability of the SRH values has been calculated. 14 The effect of SRH variability on the five main HRT3 stereometric parameters was assessed. To explore what extent of the changes in RA, RV, RNFL, CSM, and HVC could be attributed to changes in SRH, plots were constructed for all changes (compared with baseline) for SRH and compared with the changes in the parameter of interest. Given that each subject contributed multiple measurements and that this number differed among the subjects, the squared correlation coefficient obtained from these plots would not be appropriate to quantify the percentage of variability in parameter change versus baseline explained by the change in SRH versus baseline. Instead, a bivariate linear mixed model was used that took into account the repeated measures and modeled the squared correlation between changes as a function of the time since baseline. 15 More specifically, a linear mixed model was used to describe SRH evolution over time. Random intercepts and slopes were used to allow each patient to deviate from the average evolution. A similar model was specified for the parameter of interest (e.g., RA). Both models were joined by allowing the random effects in the model (intercept, slope, and error component) to be correlated. As such, the SRH and RA measurements are allowed to be correlated (and the correlation will depend on the covariance parameters in the joint model). Moreover, changes in SRH can be correlated with changes in RA. The latter correlation can be calculated for specific time intervals. All analyses were performed with a statistical package (SAS, version 9.1; SAS Institute, Cary, NC). 
Results
The HRT3 outcomes of 641 glaucoma patients visiting our department from August 2007 to October 2007 were evaluated for eligibility, according to the inclusion criteria. Of these, 150 patients (150 eyes) were included and entered for final analysis. They represent a total of 556 examinations, of which 406 were follow-up examinations. The mean total number of HRTs per patient was 3.7 (range, 3–7), whereas the mean follow-up time of each patient was 3.5 years (range, 1.4–6.7). The mean proportion of follow-up HRT3 examinations with intraindividual excess variability of SRH was 46% (95% CI, 40–52). The mean intraindividual excess SRH variability was 24% (95% CI, 20–28), indicting that when excess SRH variability was present, the absolute (percentage) difference versus baseline equaled on average 24%. The median absolute difference in the SRH compared with the baseline in all HRT3 examinations was 8.6% (IQR, 3.9%–16.2%), and the percentage differences in SRH compared with the baseline for all follow-up examinations was plotted (Fig. 1) . The mean value of the SRH equals 0.29 μm, with a between-patient SD of 0.105 μm and a within-patient SD of 0.033 μm. The latter implies that 95% of the SRH measurements are expected to fall within a distance of 0.066 μm of the true SRH value. We plotted the SRH variability against the variability of the different stereometric parameters (Fig. 2) . The proportion of variability of the stereometric parameters explained by the variability of SRH was calculated (Table 1)
Discussion
Given the widespread use of the HRT to detect optic disc damage progression, measurements of the stereometric parameters should be expected to be reliable. However, almost all stereometric parameter outcomes are based on an automatically determined SRH, which can vary. 8 Several studies have investigated the usefulness of different reference planes for the HRT. 16 17 18 19 20 Little has been reported on the frequency, magnitude, and impact of SRH variability on the main HRT parameters. Based on previously published work and as informally suggested by the HRT3 manufacturer, a cutoff value of 10% variability was chosen. Our study clearly indicates that excess SRH variability does occur in a considerable proportion of follow-up HRT3 scans (46% of all HRT3 scans) and was 24% on average. In a large sample of consecutive HRT3 examinations of good quality, more than half the variability in RNFL, RA, and RV could be attributed to variability in the SRH. Strouthidis et al. 21 reported that intertest differences in reference height and image quality had a strong relationship with intertest RA differences and that together these were responsible for 70% of the intertest variability of RA measurements. In addition, Tan et al. 22 found that the most frequent contributor to RA variability was the position of the reference plane with respect to the optic nerve head. A more stable reference plane was recently suggested (Kappou V, et al. IOVS 2008;49:ARVO E-Abstract 3650). Regarding the reproducibility of HRT parameters, Miglior et al. 23 concluded that image acquisition-induced variability seemed larger than operator-induced variability. Other factors to be taken into account are measurement variability and the noise of HRT measurements. 24  
In fact, a more stable SRH or an appropriate correction for SRH variability is necessary. In this study, the within-patient SD of the SRH (0.033), representing the SRH fluctuations within a patient, amounted to one-third of the between-patient SD of the SRH (0.105). Considering the mean SRH (0.29), the within-patient SD of the SRH indicated that intraindividual SRH fluctuations were nonnegligible. 
The HVC, as an HRT3 parameter, was only slightly influenced by SRH variability, whereas the CSM appeared to be independent of SRH. The latter finding suggests that CSM could be a preferred HRT3 parameter to determine glaucoma progression. Despite its independence of the SRH, one can graphically observe that considerable variability in CSM still exists. HVC represents the height variation of the retinal surface along the contour line—that is, the height difference between the most elevated and the most depressed points of the contour line. 8 In light of this definition, with HVC also independent of SRH, our results still demonstrate a limited influence of SRH variability on HVC changes over time, probably because of measurement variability. Our findings indicate that TCA, provided appropriate image alignment correction and considering guidelines for clinical versus statistical significant progress, may be preferred to assess glaucoma progression. 
In summary, we investigated SRH variability in consecutive HRT3 examinations of good quality and its influence on stereometric parameters. We found that 45.8% of intraindividual follow-up HRT3 examinations showed excess SRH variability. The impact rates of SRH variability on the variability of the stereometric parameters RA, RV, RNFL, CSM, and HVC were 55%, 75%, 74%, 0%, and 19%, respectively. These findings indicate that changes in stereometric HRT3 parameters should be considered with caution when the variability of the SRH exceeds 10% compared with baseline. Therefore, researchers and the HRT manufacturer are challenged to further refine this technique or the software to obtain a more stable reference plane, resulting in more repeatable and reliable stereometric HRT parameters for the detection of glaucoma progression. 
 
Figure 1.
 
Histogram plotting the differences in SRH compared with baseline for all follow-up HRT3 examinations.
Figure 1.
 
Histogram plotting the differences in SRH compared with baseline for all follow-up HRT3 examinations.
Figure 2.
 
Scatterplots of the changes in SRH against the changes in RA (A), RV (B), RNFL (C), CSM (D), and HVC (E) relative to baseline in consecutive HRT3 examinations.
Figure 2.
 
Scatterplots of the changes in SRH against the changes in RA (A), RV (B), RNFL (C), CSM (D), and HVC (E) relative to baseline in consecutive HRT3 examinations.
Table 1.
 
Variability of HRT3 Parameters Explained by Variability in SRH
Table 1.
 
Variability of HRT3 Parameters Explained by Variability in SRH
HRT3 Stereometric Parameter Percentage of Variability Explained by SRH Variability
Rim area 55.5
Rim volume 75.3
Retinal nerve fiber layer thickness 74.7
Cup shape measure 0.0
Height variation contour 19.4
WeinrebRN, KhawPT. Primary open-angle glaucoma. Lancet. 2004;363:1711–1720. [CrossRef] [PubMed]
KhawPT, ShahP, ElkingtonAR. Glaucoma 1: diagnosis. BMJ. 2004;328:97–99. [CrossRef] [PubMed]
ZeyenTG, CaprioliJ. Progression of disc and field damage in early glaucoma. Arch Ophthalmol. 1993;111:62–65. [CrossRef] [PubMed]
ArtesPH, ChauhanBC. Longitudinal changes in the visual field and optic disc in glaucoma. Prog Retin Eye Res. 2005;24:333–354. [CrossRef] [PubMed]
ReusNJ, LemijHG. Relationships between standard automated perimetry, HRT confocal scanning laser ophthalmoscopy, and GDx VCC scanning laser polarimetry. Invest Ophthalmol Vis Sci. 2005;46:4182–4188. [CrossRef] [PubMed]
ErvinJC, LemijHG, MillsRP, QuigleyHA, ThompsonHW, BurgoyneCF. Clinician change detection viewing longitudinal stereophotographs compared to confocal scanning laser tomography in the LSU Experimental Glaucoma (LEG) Study. Ophthalmology. 2002;109:467–481. [CrossRef] [PubMed]
KalaboukhovaL, FridhammarV, LindblomB. Glaucoma follow-up by the Heidelberg Retina Tomograph—new graphical analysis of optic disc topography changes. Graefes Arch Clin Exp Ophthalmol. 2006;244:654–662. [CrossRef] [PubMed]
Heidelberg Retina Tomograph, Glaucoma Module. Operating Instructions Software Version 3.0. 2005;Heidelberg Engineering GmbH Heidelberg.
HoffmannEM, BowdC, MedeirosFA, et al. Agreement among 3 optical imaging methods for the assessment of optic disc topography. Ophthalmology. 2005;112:2149–2156. [CrossRef] [PubMed]
GiangiacomoA, Garway-HeathD, CaprioliJ. Diagnosing glaucoma progression: current practice and promising technologies. Curr Opin Ophthalmol. 2006;17:153–162. [CrossRef] [PubMed]
ChauhanBC, McCormickTA, NicolelaMT, LeBlancRP. Optic disc and visual field changes in a prospective longitudinal study of patients with glaucoma: comparison of scanning laser tomography with conventional perimetry and optic disc photography. Arch Ophthalmol. 2001;119:1492–1499. [CrossRef] [PubMed]
KourkoutasD, BuysYM, FlanaganJG, HatchWV, BalianC, TropeGE. Comparison of glaucoma progression evaluated with Heidelberg Retina Tomograph II versus optic nerve head stereophotographs. Can J Ophthalmol. 2007;42:82–88. [CrossRef] [PubMed]
VerdonckN, ZeyenT, Van MalderenL, SpileersW. Short-term intra-individual variability in Heidelberg Retina Tomograph II. Bull Soc Belge Ophtalmol. 2002;286:51–57. [PubMed]
QuanH, ShihWJ. Assessing reproducibility by the within-subject coefficient of variation with random effects models. Biometrics. 1996;52:1195–1203. [CrossRef] [PubMed]
ThiebautR, Jacqmin-GaddaH, ChêneG, LeportC, CommengesD. Bivariate linear mixed models using SAS PROC MIXED. Comput Methods Programs Biomed. 2002;69:249–256. [CrossRef] [PubMed]
BurkRO, VihanninjokiK, BartkeT, et al. Development of the standard reference plane for the Heidelberg Retina Tomograph. Graefes Arch Clin Exp Ophthalmol. 2000;238:375–384. [CrossRef] [PubMed]
ParkKH, CaprioliJ. Development of a novel reference plane for the Heidelberg Retina Tomograph with optical coherence tomography measurements. J Glaucoma. 2002;11:385–391. [CrossRef] [PubMed]
TanJC, WhiteE, PoinoosawmyD, HitchingsRA. Validity of rim area measurements by different reference planes. J Glaucoma. 2004;13:245–250. [CrossRef] [PubMed]
VihanninjokiK, BurkRO, TeesaluP, TuulonenA, AiraksinenPJ. Optic disc biomorphometry with the Heidelberg Retina Tomograph at different reference levels. Acta Ophthalmol Scand. 2002;80:47–53. [PubMed]
StrouthidisNG, WhiteET, OwenVM, HoTA, Garway-HeathDF. Improving the repeatability of Heidelberg Retina Tomograph and Heidelberg Retina Tomograph II rim area measurements. Br J Ophthalmol. 2005;89:1433–1437. [CrossRef] [PubMed]
StrouthidisNG, WhiteET, OwenVM, HoTA, HammondCJ, Garway-HeathDF. Factors affecting the test-retest variability of Heidelberg Retina Tomograph and Heidelberg Retina Tomograph II measurements. Br J Ophthalmol. 2005;89:1427–1432. [CrossRef] [PubMed]
TanJC, Garway-HeathDF, FitzkeFW, HitchingsRA. Reasons for rim area variability in scanning laser tomography. Invest Ophthalmol Vis Sci. 2003;44:1126–1131. [CrossRef] [PubMed]
MigliorS, AlbéE, GuareschiM, RossettiL, OrzalesiN. Intraobserver and interobserver reproducibility in the evaluation of optic disc stereometric parameters by Heidelberg Retina Tomograph. Ophthalmology. 2002;109:1072–1077. [CrossRef] [PubMed]
OwenVM, StrouthidisNG, Garway-HeathDF, CrabbDP. Measurement variability in Heidelberg Retina Tomograph imaging of neuroretinal rim area. Invest Ophthalmol Vis Sci. 2006;47:5322–5330. [CrossRef] [PubMed]
Figure 1.
 
Histogram plotting the differences in SRH compared with baseline for all follow-up HRT3 examinations.
Figure 1.
 
Histogram plotting the differences in SRH compared with baseline for all follow-up HRT3 examinations.
Figure 2.
 
Scatterplots of the changes in SRH against the changes in RA (A), RV (B), RNFL (C), CSM (D), and HVC (E) relative to baseline in consecutive HRT3 examinations.
Figure 2.
 
Scatterplots of the changes in SRH against the changes in RA (A), RV (B), RNFL (C), CSM (D), and HVC (E) relative to baseline in consecutive HRT3 examinations.
Table 1.
 
Variability of HRT3 Parameters Explained by Variability in SRH
Table 1.
 
Variability of HRT3 Parameters Explained by Variability in SRH
HRT3 Stereometric Parameter Percentage of Variability Explained by SRH Variability
Rim area 55.5
Rim volume 75.3
Retinal nerve fiber layer thickness 74.7
Cup shape measure 0.0
Height variation contour 19.4
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×