April 2011
Volume 52, Issue 5
Free
Glaucoma  |   April 2011
Retinal Nerve Fiber Layer Normative Classification by Optical Coherence Tomography for Prediction of Future Visual Field Loss
Author Affiliations & Notes
  • Kyung Rim Sung
    From the Department of Ophthalmology and
  • Sophia Kim
    From the Department of Ophthalmology and
  • Youngrok Lee
    From the Department of Ophthalmology and
  • Sung-Cheol Yun
    the Department of Clinical Epidemiology and Biostatistics, College of Medicine, University of Ulsan, Asan Medical Center, Seoul, Korea.
  • Jung Hwa Na
    From the Department of Ophthalmology and
  • Corresponding author: Kyung Rim Sung, Department of Ophthalmology, University of Ulsan, College of Medicine, Asan Medical Center, 388-1 Pungnap-2-dong, Songpa-gu, Seoul, Korea 138-736; sungeye@gmail.com
Investigative Ophthalmology & Visual Science April 2011, Vol.52, 2634-2639. doi:10.1167/iovs.10-6246
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Kyung Rim Sung, Sophia Kim, Youngrok Lee, Sung-Cheol Yun, Jung Hwa Na; Retinal Nerve Fiber Layer Normative Classification by Optical Coherence Tomography for Prediction of Future Visual Field Loss. Invest. Ophthalmol. Vis. Sci. 2011;52(5):2634-2639. doi: 10.1167/iovs.10-6246.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose.: To evaluate the utility of baseline Stratus optical coherence tomography (OCT; Carl Zeiss Meditec, Dublin, CA) retinal nerve fiber layer (RNFL) normative classification in the prediction of future visual field (VF) loss.

Methods.: Eighty-eight eyes with suspected glaucoma with abnormal RNFL classification by Stratus OCT were followed up for more than 4 years. VF conversion in three consecutive tests was assessed after baseline Stratus OCT and VF examination. Baseline intraocular pressure, VF global indices, OCT RNFL thickness, and number of abnormal OCT sectors were compared between VF converters (CG) and nonconverters (NCG). Positive and negative predictive values (PPV, NPV) of OCT sectors with abnormal classifications were calculated with respect to VF conversion. Hazard ratios (HRs) of various risk factors, including abnormal OCT classification, with respect to future VF conversion, were determined by use of the Cox proportional hazard model.

Results.: Twenty-one (23.9%) eyes showed VF conversion during follow-up. Baseline OCT RNFL thickness was significantly lower and the number of abnormal OCT RNFL sectors significantly greater in CG than in NCG patients (P = 0.022 for both). The PPV and NPV of normative OCT RNFL classification was highest in the inferior quadrant (50%, 87.1%, respectively). Baseline VF mean deviation (MD) and the number of abnormal OCT RNFL sectors were both associated with future VF conversion (HR, 0.788 and 1.290, respectively).

Conclusions.: In patients with suspected glaucoma, an abnormal RNFL classification in the inferior area of the optic disc or an elevated number of abnormal RNFL sectors, as determined by Stratus OCT, were both associated with future VF conversion.

Stratus optical coherence tomography (OCT; Carl Zeiss Meditec, Inc., Dublin, CA) has been used for glaucoma diagnosis for the past 8 years. Many cross-sectional studies have verified the glaucoma diagnostic capability afforded by retinal nerve fiber layer (RNFL) thickness measurement as performed by Stratus OCT. 1 11 However, only a few studies have addressed its ability to detect longitudinal glaucoma progression. 12,13  
The most helpful feature of Stratus OCT, shared with other imaging devices used in clinical practice, is an objective, quantitative output permitting subsequent normative classification. Such quantitative data are clearly superior to qualitative assessment using conventional optic disc photography in patients with optic discs ambiguous in appearance or in the earlier stages of optic disc change. Although optic disc examination remains the gold standard for the assessment of structural glaucomatous damage, there have been large variability and disagreement among clinicians in the assessment of such optic disc change. 14 17 Categorical classification of RNFL thickness measurements using terms such as abnormal, borderline, or within normal limits, based on comparison with a normative database, is another advantage aiding clinicians in assessing the structural status of glaucoma objectively and conveniently. Therefore, eyes with suspected glaucoma with optic discs of suspicious appearance have usually been analyzed with imaging devices. If patients do not fall within normal limits in OCT RNFL thickness analysis, but nonetheless have apparently normal visual fields (VFs), a clinician can expect that some functional glaucomatous change may be evident in future VF examinations. 
It has been reported that glaucomatous optic disc or RNFL damage precedes VF defect. 18 20 Therefore, detection of progressive structural change may predict future VF loss. Medeiros et al. 21 observed more than 600 eyes with suspected glaucoma eyes for an average of 8 years and reported that the presence of progressive optic change was highly predictive of future functional loss. Chauhan et al. 22 reported that longitudinally measured optic disc change was predictive of subsequent VF progression and may be an efficacious end point for functional outcomes in clinical studies and trials in glaucoma. Research outcomes regarding the predictability of structural change assessed by imaging devices for future VF development is relatively scarce. 4,23 25 For various reasons, long-term follow-up studies tend to be more difficult to perform than cross-sectional studies. Constant updating of versions of imaging devices may make it difficult to examine patients with the same equipment longitudinally. 
Clinicians employing OCT to evaluate patients have often asked which patients would be expected to show future VF changes and when apparent OCT RNFL abnormalities will develop into detectable glaucomatous VF deficits. In other words, clinicians are interested in the predictive power of a Stratus OCT RNFL abnormal classification with respect to development of future VF loss. An allied question is that, if it is accepted that some proportion of suspect eyes with Stratus OCT RNFL abnormalities will develop VF loss, is it possible to quantify the risk? Is there any difference in OCT baseline characteristics between patients in whom an RNFL abnormality will progress to VF loss and those in whom such progression will not occur? Finally, do some particular eye sectors showing Stratus OCT RNFL abnormalities progress more frequently to VF loss? In this study, we sought to answer these questions. 
Methods
Subjects
We retrospectively reviewed the medical records of patients with suspected glaucoma who were seen in the glaucoma service clinic of the Asan Medical Center, Seoul, Korea. On initial evaluation, each subject underwent a complete ophthalmic examination, including the taking of medical, ocular, and family history; visual acuity (VA) testing; visual field perimetry (Humphrey field analyzer [HFA] Swedish Interactive Threshold Algorithm (SITA) 24-2 full threshold; Carl Zeiss Meditec); multiple intraocular pressure (IOP) measurements using Goldmann applanation tonometry (GAT); central corneal thickness (CCT) assessment (DGH-550 instrument; DGH Technology, Inc., Exton, PA); stereoscopic optic nerve photography; and Stratus OCT (Carl Zeiss Meditec, Inc.). 
All participants had to meet the following criteria: best corrected VA (BCVA) of 20/30 or better, with a spherical equivalent within ±5 D and cylinder correction within +3 D; presence of a normal anterior chamber and open angle on slit lamp and gonioscopic examination; and a reliable visual field test result with a false-positive error <15%, a false-negative error <15%, and a fixation loss <20%. Subjects with any other ophthalmic disease that could result in an abnormal HFA test result, or with a history of diabetes mellitus or intraocular surgery, were excluded. One eye was randomly selected if both eyes were found to be eligible for the study. Those with suspected glaucoma with at least 4 years of follow-up after baseline Stratus OCT measurement were consecutively selected. Suspect eyes were defined by an IOP greater than 21 mm Hg on at least three GAT measurements (ocular hypertension [OHT]), the presence of a glaucomatous optic disc showing increased cupping, a difference in vertical cup-disc ratio >0.2 between eyes considering optic disc size, the presence of diffuse or focal neural rim thinning, or evident hemorrhage. All suspect eyes had normal VF results on at least two repeated tests. Among such subjects, suspect eyes with abnormal classification defined in any of the 12 clock-hour sectors or the four quadrants, as evaluated by Stratus OCT RNFL assessment, were consecutively entered into the final analysis. The baseline VF and OCT examinations were performed within a period of 2 weeks. To minimize the learning effect, the second VF test result was used in analysis. Follow-up VF tests were usually performed at 6-month intervals. 
All participants had newly diagnosed suspected glaucoma; therefore, none of the participants was using antiglaucoma medication at baseline examination. 
All procedures conformed to the Declaration of Helsinki and the study was approved by the Institutional Review Board of the Asan Medical Center at the University of Ulsan, Seoul, Korea. 
VF Conversion
Glaucomatous VF changes were defined as the appearance of a cluster of three points with a probability of less than 5% on a pattern standard deviation (PSD) map, including at least one point with a probability of less than 1%, or a cluster of two points with a probability of less than 1% in the corresponding area of OCT RNFL abnormality; a glaucoma hemifield test (GHT) result outside normal limits; and a PSD significantly elevated beyond the 5% level. Subjects with normal VFs did not show any of these features. If glaucomatous VF changes were apparent on three consecutive examinations, the eye was considered to be a VF converter. Therefore, after baseline examination, at least three reliable follow-up VF results were required. Since we excluded the first VF for minimizing learning effect and considered the second VF as the baseline, all participants needed to have at least five VF tests. Eyes were categorized into two groups: converters and nonconverters. In the converter group (CG), follow-up time was defined as the time from baseline OCT imaging to the date of the first abnormal VF examination. In the nonconverter group (NCG), follow-up time was defined as the time from baseline OCT imaging to the last normal VF examination date. We also assessed the prevalence of VF mean deviation (MD) deterioration of more than 2 dB compared with baseline in three consecutive examinations. 
Optical Coherence Tomography
The basic principles and technical characteristics of the Stratus OCT have been described elsewhere. 26 The fast RNFL scan protocol was used. We excluded all poor-quality scans, defined as those of signal strength less than 7, overt misalignment of the surface detection algorithm on at least 10% of consecutive A scans or 15% of cumulative A scans, or overt displacement of the measurement circle, as assessed subjectively. Stratus OCT provides RNFL thickness maps for four quadrants (superior, inferior, nasal, and temporal) and 12 clock hours and includes classifications derived by use of an internal normative database. Four normative classifications are provided in the standard printout. The 95th to 100th percentiles are hypernormal (shown in white on thickness maps), the 5th to 95th percentiles are normal (green); the 1st to 5th percentiles are borderline (yellow); and the percentile <1 is abnormal (red). In the current work, white and green signified normal, and yellow and red signified abnormal. Eyes with an abnormal classification in ≥1 quadrant or ≥1 clock-hour were included in the final analysis. All OCT data were aligned according to the orientation of the right eye. Thus, clock hour 9 of the circumpapillary scan represented the temporal optic disc side of both eyes. 
Statistical Analysis
Categorical variables are presented as numbers and percentages, and the χ2 test was used to compare clinical characteristics between the CG and NCG. The Shapiro-Wilk test was used to determine the distribution of numerical data. Continuous variables are expressed as means with SD (normally distributed data), or as medians with interquartile ranges (non-normally distributed data), as appropriate, and Student's unpaired t-test (normally distributed data) or the Mann-Whitney U test (non-normally distributed data) was used for between-group comparisons. Positive and negative predictive values (PPV, NPV) of OCT RNFL classifications in the four quadrants and 12 clock-hour sectors, with respect to future VF conversion, were assessed. Hazard ratios (HRs) for associations between potential predictive factors and VF conversion were obtained using Cox proportional hazards models. Univariate analyses were performed separately for each variable. Variables with probability values ≤0.10 in univariate analyses were included in multivariate Cox proportional hazards models. A backward-elimination process was used to develop the final multivariable model, and adjusted HRs with 95% confidence intervals (CIs) were calculated. Schoenfeld residuals and the log [−log(survival rate)] test were used to verify that the proportional hazards assumptions were not violated. Model fit was assessed using residual analysis. P < 0.05 was considered statistically significant. All statistical analyses were performed with commercial software (SAS, ver. 9.1; SAS Institute Inc., Cary, NC; MedCalc, ver. 9.6; Mariakerke, Belgium;, or SPSS, ver. 15.0; SPSS Inc., Chicago, IL). 
Results
Among 156 eyes with suspected glaucoma that had baseline OCT and at least 4 years' follow-up, 13 had OCT images of poor quality, 34 had normal OCT RNFL classification, and 21 had an insufficient number of reliable follow-up VF results. 
Hence, the final analysis included 88 suspected glaucomatous eyes of 88 subjects who had abnormal OCT RNFL thickness classifications but normal VFs at baseline, at least three follow-up VF tests after baseline measurement, and more than 4 years' follow-up. Forty-one subjects were men, 47 were women, and all were Asian (Korean). Thirteen subjects had OHT, and 75 had glaucomatous optic disc changes. The mean follow-up time (time to VF conversion) was 28.4 ± 16.4 months in the CG and 52.9 ± 4.6 months in the NCG. Twenty-one eyes (23.9%) showed VF conversion during the follow-up period and were thus classified in the CG. In other words, 67 eyes did not show the predefined VF conversion criteria during follow-up and were thus placed in the NCG. Six (6.8%) eyes showed VF conversion at the 1-year follow-up after baseline OCT and VF measurement, 14 (15.9%) eyes at 2 years, 16 (18.2%) at 3 years, and 19 (21.6%) at 4 years. By the worsening VF MD criterion, 14 subjects showed MD deterioration more than 2 dB in three consecutive tests. Baseline OCT RNFL thickness was significantly lower in those eyes that showed worsening VF MD than those eyes that did not show (83.4 ± 8.6 μm vs. 92.2 ± 10.1 μm; P = 0.003). Among 14 subjects, 11 showed predefined VF conversion criteria during the follow-up period. Thus, among the 88 participants, 11 eyes showed both VF MD deterioration and predefined VF conversion. Among the 13 OHT eyes, only 1 eye showed VF conversion, among 75 eyes with glaucomatous optic disc change; 20 eyes showed VF conversion during the follow-up period. The mean baseline IOP, baseline CCT, mean number of VF tests, and mean follow-up IOP did not significantly differ between subjects in the CG and NCG (P = 0.65, 0.31, 0.66, and 0.91, respectively). The mean baseline age was marginally different between the CG and NCG (P = 0.06). However, the baseline VF MD was significantly lower and the PSD significantly higher in the CG than in the NCG subjects (P = 0.018 and 0.001, respectively). The baseline OCT RNFL thickness was significantly lower and the number of abnormal OCT RNFL sectors significantly greater in the CG than in the NCG subjects (P = 0.022 for both). Fifty-six of 88 eyes received IOP-lowering medication during the follow-up period. The percentage of eyes treated with such medication was significantly greater in the CG than in the NCG (P < 0.001). The eyes treated with IOP-lowering medication had a significantly lower baseline average RNFL thickness (87.7 vs. 96.3 μm, P < 0.001) and a lower baseline VF MD (−2.14 dB vs. −1.42 dB, P = 0.041), than did the untreated eyes. The characteristics of the CG and NCG subjects are compared in Table 1
Table 1.
 
The Baseline Characteristics and Ocular Parameters the CG and NCG
Table 1.
 
The Baseline Characteristics and Ocular Parameters the CG and NCG
CG (n = 21) NCG (n = 67) P
Baseline age, y, mean ± SD 56.2 ± 15.6 48.7 ± 15.6 0.06*
Sex, n, male/female 10/11 37/30 0.359†
Spherical equivalent, D, mean ± SD −0.79 ± 2.01 0.67 ± 2.21 0.272*
Mean follow-up time, mo, mean ± SD 28.4 ± 16.4 52.9 ± 4.6 <0.001
Baseline VF MD, dB, mean ± SD −2.7 ± 1.7 −1.6 ± 1.8 0.018*
Baseline VF PSD, median dB (interquartile range) 2.51 (1.94–3.03) 1.80 (1.63–2.23) 0.001‡
Baseline OCT RNFL thickness, μm, mean ± SD 86.3 ± 8.8 92.2 ± 10.5 0.022*
Baseline number of abnormal sectors by OCT, median n (interquartile range) 3.0 (1.5–6.0) 2.0 (1.0–3.0) 0.022‡
Baseline CCT, μm, mean ± SD 538.4 ± 38.1 549.4 ± 33.8 0.31*
Follow-up VF tests, n, mean ± SD 5.9 ± 1.5 5.7 ± 1.4 0.66*
Mean baseline IOP, median mm Hg (interquartile range) 16.0 (13.5–18.0) 15.0 (13.0–18.0) 0.65‡
Mean follow-up IOP, median mm Hg (interquartile range) 15.0 (13.0–16.5) 14.0 (12.0–18.0) 0.91‡
IOP-lowering treatment, n, yes/no 20/1 36/31 <0.001†
The PPV of the normative OCT RNFL classification with respect to VF conversion was highest in the inferior quadrant (50%), followed by that in the 6 and 7 o'clock segments (45.0% and 47.4%, respectively). NPV was also highest in the inferior quadrant and next highest at 6 and 7 o'clock (87.1%, 82.4%, and 82.6%, respectively). The PPVs and NPVs of the normative OCT RNFL classification in the four quadrants and 12 clock-hour sectors are listed in Table 2
Table 2.
 
Predictive Values of Baseline OCT RNFL Normative Classifications with Respect to Future VF Conversion
Table 2.
 
Predictive Values of Baseline OCT RNFL Normative Classifications with Respect to Future VF Conversion
Sector PPV (95% CI) NPV (95% CI)
Temporal quadrant 16.7 (0.88–63.5) 75.6 (64.7–84.1)
Superior quadrant 38.1 (19.0–61.3) 80.6 (68.8–88.9)
Nasal quadrant 30.8 (10.4–61.1) 77.3 (65.9–85.9)
Inferior quadrant 50.0 (30.4–69.6) 87.1 (75.6–93.9)
9 o'clock 0 (0–53.7) 25.3 (16.7–36.2)
10 o'clock 30.8 (10.4–61.1) 77.4 (66.7–85.5)
11 o'clock 41.7 (16.5–71.4) 78.9 (67.8–87.1)
12 o'clock 35.7 (14.0–64.3) 78.4 (70.0–86.8)
1 o'clock 27.8 (10.7–53.6) 72.2 (46.4–89.3)
2 o'clock 13.0 (3.4–34.7) 72.3 (59.6–82.3)
3 o'clock 44.4 (15.3–77.3) 78.5 (67.5–86.6)
4 o'clock 38.5 (15.1–67.7) 78.7 (67.4–87.0)
5 o'clock 28.6 (12.2–52.3) 77.6 (65.5–86.5)
6 o'clock 45.0 (23.8–68.0) 82.4 (70.8–90.2)
7 o'clock 47.4 (25.2–70.5) 82.6 (71.2–90.3)
8 o'clock 16.7 (0.88–63.5) 75.6 (64.7–84.1)
Table 3 shows the univariate and multivariate HRs for each putative risk factor, including abnormal OCT RNFL sectors with respect to VF conversion. Baseline VF MD, PSD, average OCT RNFL thickness, and number of abnormal OCT RNFL sectors, were all found to be significant predictive factors for VF conversion by univariate analysis. Of the OCT sectors, abnormal classification in the inferior quadrant and in the 3-, 6-, and 7-o'clock segments, were also significant predictive factors of VF conversion. By multivariate analysis, both baseline VF MD and number of abnormal OCT sectors were associated with future VF conversion (HR [95% CI]; 0.788 [0.628–0.987], 1.290 [1.099–1.587], respectively). 
Table 3.
 
Univariate and Multivariate Cox's Proportional Hazard Models for Prediction of VF Conversion
Table 3.
 
Univariate and Multivariate Cox's Proportional Hazard Models for Prediction of VF Conversion
Univariate Analysis Multivariate Analysis
HR (95% CI) P HR (95% CI) P
Baseline Characteristics
Age (per year) 1.028 (0.997–1.061) 0.078 1.036 (1.003–1.070) 0.32
Sex 1.369 (0.580–3.232) 0.474
Spherical equivalent (per D) 0.947 (0.922–0.998) 0.352
VF MD (per 1 dB) 0.768 (0.613–0.961) 0.021 0.788 (0.628–0.987) 0.038
VF PSD (per 1 dB) 1.541 (1.106–2.147) 0.011 1.07 (0.627–0.714) 0.887
Average RNFL thickness by OCT (per μm) 0.956 (0.916–0.998) 0.042 1.047 (0.983–1.114) 0.151
Number of abnormal sectors by OCT (per sector) 1.313 (1.110–1.553) 0.002 1.290 (1.099–1.587) 0.016
Mean IOP (per mm Hg) 1.007 (0.896–1.130) 0.912
CCT (per μm) 0.990 (0.976–1.004) 0.990
Baseline Abnormal OCT Sectors
Temporal 0.627 (0.084–4.67) 0.649
Superior 2.02 (0.83–4.88) 0.12
Nasal 1.84 (0.61–5.56) 0.279
Inferior 4.05 (1.65–9.90) 0.002 1.99 (0.73–5.38) 0.177
9 o'clock 0.31 (0.087–21.5) 0.428
10 o'clock 2.19 (0.50–9.52) 0.297
11 o'clock 2.21 (0.81–6.06) 0.122
12 o'clock 1.73 (0.63–4.72) 0.288
1 o'clock 1.37 (0.50–3.76) 0.539
2 o'clock 0.49 (0.14–1.68) 0.259
3 o'clock 3.08 (1.02–9.26) 0.046 0.66 (0.08–5.35) 0.695
4 o'clock 2.36 (0.86–6.49) 0.097 1.54 (0.38–6.25) 0.545
5 o'clock 1.32 (0.51–3.39) 0.571
6 o'clock 2.74 (1.14–6.62) 0.025 1.36 (0.50–3.73) 0.549
7 o'clock 3.70 (1.56–8.85) 0.003 2.49 (0.96–6.41) 0.059
8 o'clock 0.69 (0.092–5.18) 0.719
Follow-up Parameters
Mean follow-up IOP (per mm Hg) 0.991 (0.873–1.126) 0.891
Discussion
Our longitudinal observational study showed that 23.9% of eyes with suspected glaucoma with abnormal baseline OCT RNFL classification showed VF conversion during follow-up. In other words, more than three quarters of eyes did not show VF conversion during more than 4 years of follow-up. The percentage of VF conversion was similar to that seen by Lalezary et al., 4 using OCT 2 instrumentation (20%, average of 4.2 years of follow-up). As OCT 2 did not permit normative classification of RNFL thickness, the cited authors compared raw RNFL thicknesses, both overall and in the four quadrants, and reported that VF converters had significantly lower baseline RNFL thickness in the inferior quadrant. Mohammadi et al. 23 found a lower percentage (10%) of VF conversion using a nerve fiber analyzer (GDx; Carl Zeiss Meditec, Inc.), which may be attributable to the shorter follow-up period (2.7–3.8 years) used. 
Of note, although the baseline VF of all participants was normal by the predefined VF abnormality criteria (PSD <5%, a GHT result outside normal limits, and clustering in the pattern deviation map), both VF MD and PSD differed significantly between the CG and the NCG. This is in line with previous data obtained with other devices. 4,23 Both cited reports found that baseline VF PSD was significantly higher in VF converters. Therefore, we speculate that CG eyes already have subtle VF changes although the commonly used VF abnormality criteria were not met. Baseline average RNFL thickness and the number of sectors classified as abnormal differed significantly between the CG and the NCG. The observation that a significantly greater percentage of eyes in the CG were treated with IOP-lowering medication can be explained by the fact that eyes of thinner RNFL thickness, or with a greater number of abnormal RNFL sectors at baseline, showed an increased tendency toward a requirement for treatment, and that eyes showing VF conversion during follow-up were subsequently treated with IOP-lowering medication. This finding is similar to that of Lalezary et al., 4 in that a higher percentage of eyes were treated with glaucoma medication in the CG than in the NCG. However, mean baseline and follow-up IOP and CCT did not differ significantly between the CG and NCG in our current analysis. 
Of the four quadrants (superior, interior, nasal, and temporal), the inferior quadrant RNFL classification showed the highest PPV (50%) and lowest NPV (87.1%) with respect to VF conversion. Thus, 50% of eyes with abnormal inferior quadrant classifications at baseline showed VF conversion during follow-up. Of the 12 clock-hour sectors, the 7 o'clock data showed the highest PPV (47.4%) and the lowest NPV (82.6%). These observations suggest that abnormal RNFL classification in the inferior region of the optic disc should be carefully assessed to predict future VF conversion. 
In a univariate Cox proportional hazard model, all VF MD, VF PSD, average RNFL thickness by OCT, and number of abnormal OCT sectors, were associated with VF conversion. Of the OCT sectors, abnormal classification of the inferior quadrant, and the 3-, 6-, and 7-o'clock segments, were predictive of VF conversion. In multivariate analysis adjusted for age; VF PSD; average RNFL thickness; and abnormal classification of the inferior quadrant and 3-, 4-, 6-, and 7-o'clock sectors, VF MD and the number of abnormal OCT sectors were strongly associated with VF conversion. Thus, we believe that reduced VF sensitivity and a greater extent of abnormal RNFL readings at baseline are strongly predictive of VF conversion. 
As our study was not prospectively designed, IOP-lowering medication was not randomly distributed among the participants. Hence, we were unable to identify any effect of IOP reduction on VF conversion. 
Although spectral domain (SD)-OCT instruments with higher resolutions and a faster scan speed than Stratus OCT have recently become commercially available, many clinics worldwide continue to use the Stratus OCT. Since TD- and SD-OCT data have been reported to be incomparable in many studies, 27 31 we believe longitudinal studies with the same device are advantageous in detecting glaucomatous progression. Therefore, for longitudinally tracing the patients' status with the same equipment, even some SD-OCT users may prefer to scan their patients with Stratus OCT if the patients have been scanned by that system previously. Moreover, considering the recent launch time of the commercially available SD-OCT and the slow progression characteristic of glaucoma, we must wait for longitudinal SD-OCT data with long enough follow-up to become available. In other words, longitudinal data were available only from Stratus OCT at the time of this writing. Therefore, we believe longitudinal assessment of OCT RNFL measurement using Stratus OCT for evaluation of structural change predicting VF change is necessary at this time. In a busy clinic, categorical classification and color coding in a standard printout corrected by an age-matched normative database are more helpful than thickness data. Therefore, it would be useful to study future development of those eyes classified as abnormal by OCT RNFL measurements but presenting with normal VFs. Few longitudinal studies using Stratus OCT instrumentation have appeared in the literature. 12,13,32 To the best of our knowledge, the present work is the first to employ longitudinal observation of eyes with suspected glaucoma classified as abnormal by Stratus OCT with respect to future VF conversion. 
In summary, approximately 24% of eyes with abnormal OCT RNFL classifications developed VF abnormalities during 4 years of follow-up. More than three quarters of eyes classified as abnormal by OCT RNFL thickness measurement did not show any apparent VF abnormality during an average of 52.9 months of follow-up. Eyes developing VF abnormalities during follow-up had a higher tendency to show subtle VF changes, a thinner RNFL, and a greater number of RNFL sectors that were classified as abnormal at baseline. Such classification in the inferior region of the optic disc was strongly associated with VF conversion. Clinicians employing Stratus OCT should pay particular attention to eyes yielding abnormal RNFL data from the inferior region of the optic disc or showing a greater number of abnormal sectors and mild changes in VF global indices, with respect to future VF conversion. 
Footnotes
 Disclosure: K.R. Sung, None; S. Kim, None; Y. Lee, None; S.-C. Yun, None; J.H. Na, None
References
Burgansky-Eliash Z Wollstein G Chu T . Optical coherence tomography machine learning classifiers for glaucoma detection: a preliminary study. Invest Ophthalmol Vis Sci. 2005;46:4147–4152. [CrossRef] [PubMed]
Huang ML Chen HY . Development and comparison of automated classifiers for glaucoma diagnosis using Stratus optical coherence tomography. Invest Ophthalmol Vis Sci. 2005;46:4121–4129. [CrossRef] [PubMed]
Kanamori A Nagai-Kusuhara A Escano MF . Comparison of confocal scanning laser ophthalmoscopy, scanning laser polarimetry and optical coherence tomography to discriminate ocular hypertension and glaucoma at an early stage. Graefes Arch Clin Exp Ophthalmol. 2006;244:58–68. [CrossRef] [PubMed]
Lalezary M Medeiros FA Weinreb RN . Baseline optical coherence tomography predicts the development of glaucomatous change in glaucoma suspects. Am J Ophthalmol. 2006;142:576–582. [CrossRef] [PubMed]
Leung CK Chan WM Yung WH . Comparison of macular and peripapillary measurements for the detection of glaucoma: an optical coherence tomography study. Ophthalmology. 2005;112:391–400. [CrossRef] [PubMed]
Manassakorn A Nouri-Mahdavi K Caprioli J . Comparison of retinal nerve fiber layer thickness and optic disk algorithms with optical coherence tomography to detect glaucoma. Am J Ophthalmol. 2006;141:105–115. [CrossRef] [PubMed]
Medeiros FA Zangwill LM Bowd C . Evaluation of retinal nerve fiber layer, optic nerve head, and macular thickness measurements for glaucoma detection using optical coherence tomography. Am J Ophthalmol. 2005;139:44–55. [CrossRef] [PubMed]
Naithani P Sihota R Sony P . Evaluation of optical coherence tomography and heidelberg retinal tomography parameters in detecting early and moderate glaucoma. Invest Ophthalmol Vis Sci. 2007;48:3138–3145. [CrossRef] [PubMed]
Nouri-Mahdavi K Nikkhou K Hoffman DC . Detection of early glaucoma with optical coherence tomography (StratusOCT). J Glaucoma. 2008;17:183–188. [CrossRef] [PubMed]
Parikh RS Parikh S Sekhar GC . Diagnostic capability of optical coherence tomography (Stratus OCT 3) in early glaucoma. Ophthalmology. 2007;114:2238–2343. [CrossRef] [PubMed]
Sihota R Sony P Gupta V . Comparing glaucomatous optic neuropathy in primary open angle and chronic primary angle closure glaucoma eyes by optical coherence tomography. Ophthalmic Physiol Opt. 2005;25:408–415. [CrossRef] [PubMed]
Medeiros FA Zangwill LM Alencar LM . Detection of glaucoma progression with stratus OCT retinal nerve fiber layer, optic nerve head, and macular thickness measurements. Invest Ophthalmol Vis Sci. 2009;50:5741–5748. [CrossRef] [PubMed]
Leung CK Cheung CY Weinreb RN . Evaluation of retinal nerve fiber layer progression in glaucoma: a study on optical coherence tomography guided progression analysis. Invest Ophthalmol Vis Sci. 2010;51:217–222. [CrossRef] [PubMed]
Varma R Steinmann WC Scott IU . Expert agreement in evaluating the optic disc for glaucoma. Ophthalmology. 1992;99:215–221. [CrossRef] [PubMed]
Abrams LS Scott IU Spaeth GL . Agreement among optometrists, ophthalmologists, and residents in evaluating the optic disc for glaucoma. Ophthalmology. 1994;101:1662–1667. [CrossRef] [PubMed]
Reus NJ de Graaf M Lemij HG . Accuracy of GDx VCC, HRT I, and clinical assessment of stereoscopic optic nerve head photographs for diagnosing glaucoma. Br J Ophthalmol. 2007;91:313–318. [CrossRef] [PubMed]
Reus NJ Lemij HG Garway-Heath DF . Clinical assessment of stereoscopic optic disc photographs for glaucoma: the European Optic Disc Assessment Trial. Ophthalmology. 2010;117:717–723. [CrossRef] [PubMed]
Quigley HA Katz J Derick RJ . An evaluation of optic disc and nerve fiber layer examinations in monitoring progression of early glaucoma damage. Ophthalmology. 1992;99:19–28. [CrossRef] [PubMed]
Sommer A Katz J Quigley HA . Clinically detectable nerve fiber atrophy precedes the onset of glaucomatous field loss. Arch Ophthalmol. 1991;109:77–83. [CrossRef] [PubMed]
Zeyen TG Caprioli J . Progression of disc and field damage in early glaucoma. Arch Ophthalmol. 1993;111:62–65. [CrossRef] [PubMed]
Medeiros FA Alencar LM Zangwill LM . Prediction of functional loss in glaucoma from progressive optic disc damage. Arch Ophthalmol. 2009;127:1250–1256. [CrossRef] [PubMed]
Chauhan BC Nicolela MT Artes PH . Incidence and rates of visual field progression after longitudinally measured optic disc change in glaucoma. Ophthalmology. 2009;116:2110–2118. [CrossRef] [PubMed]
Mohammadi K Bowd C Weinreb RN . Retinal nerve fiber layer thickness measurements with scanning laser polarimetry predict glaucomatous visual field loss. Am J Ophthalmol. 2004;138:592–601. [CrossRef] [PubMed]
Weinreb RN Zangwill LM Jain S . Predicting the onset of glaucoma: the confocal scanning laser ophthalmoscopy ancillary study to the Ocular Hypertension Treatment Study. Ophthalmology. 2010;117:1674–1683. [CrossRef] [PubMed]
Strouthidis NG Gardiner SK Owen VM . Predicting progression to glaucoma in ocular hypertensive patients. J Glaucoma. 2010;19:304–309. [PubMed]
Paunescu LA Schuman JS Price LL . Reproducibility of nerve fiber thickness, macular thickness, and optic nerve head measurements using Stratus OCT. Invest Ophthalmol Vis Sci. 2004;45:1716–1724. [CrossRef] [PubMed]
Vizzeri G Weinreb RN Gonzalez-Garcia AO . Agreement between spectral-domain and time-domain OCT for measuring RNFL thickness. Br J Ophthalmol. 2009;93:775–781. [CrossRef] [PubMed]
González-García AO Vizzeri G Bowd C . Reproducibility of RTVue retinal nerve fiber layer thickness and optic disc measurements and agreement with Stratus optical coherence tomography measurements. Am J Ophthalmol. 2009;147:1067–1074. [CrossRef] [PubMed]
Sung KR Kim DY Park SB . Comparison of retinal nerve fiber layer thickness measured by Cirrus HD and Stratus optical coherence tomography. Ophthalmology. 2009;116:1264–1270. [CrossRef] [PubMed]
Cho JW Sung KR Hong JT Um TW Kang SY Kook MS . Detection of glaucoma by spectral domain-scanning laser ophthalmoscopy/optical coherence tomography (SD-SLO/OCT) and time domain optical coherence tomography. J Glaucoma. 2011;20:15–20. [CrossRef] [PubMed]
Knight OJ Chang RT Feuer WJ . Comparison of retinal nerve fiber layer measurements using time domain and spectral domain optical coherent tomography. Ophthalmology. 2009;116:1271–1277. [CrossRef] [PubMed]
Lee EJ Kim TW Park KH . Ability of Stratus OCT to detect progressive retinal nerve fiber layer atrophy in glaucoma. Invest Ophthalmol Vis Sci. 2009;50:662–668. [CrossRef] [PubMed]
Table 1.
 
The Baseline Characteristics and Ocular Parameters the CG and NCG
Table 1.
 
The Baseline Characteristics and Ocular Parameters the CG and NCG
CG (n = 21) NCG (n = 67) P
Baseline age, y, mean ± SD 56.2 ± 15.6 48.7 ± 15.6 0.06*
Sex, n, male/female 10/11 37/30 0.359†
Spherical equivalent, D, mean ± SD −0.79 ± 2.01 0.67 ± 2.21 0.272*
Mean follow-up time, mo, mean ± SD 28.4 ± 16.4 52.9 ± 4.6 <0.001
Baseline VF MD, dB, mean ± SD −2.7 ± 1.7 −1.6 ± 1.8 0.018*
Baseline VF PSD, median dB (interquartile range) 2.51 (1.94–3.03) 1.80 (1.63–2.23) 0.001‡
Baseline OCT RNFL thickness, μm, mean ± SD 86.3 ± 8.8 92.2 ± 10.5 0.022*
Baseline number of abnormal sectors by OCT, median n (interquartile range) 3.0 (1.5–6.0) 2.0 (1.0–3.0) 0.022‡
Baseline CCT, μm, mean ± SD 538.4 ± 38.1 549.4 ± 33.8 0.31*
Follow-up VF tests, n, mean ± SD 5.9 ± 1.5 5.7 ± 1.4 0.66*
Mean baseline IOP, median mm Hg (interquartile range) 16.0 (13.5–18.0) 15.0 (13.0–18.0) 0.65‡
Mean follow-up IOP, median mm Hg (interquartile range) 15.0 (13.0–16.5) 14.0 (12.0–18.0) 0.91‡
IOP-lowering treatment, n, yes/no 20/1 36/31 <0.001†
Table 2.
 
Predictive Values of Baseline OCT RNFL Normative Classifications with Respect to Future VF Conversion
Table 2.
 
Predictive Values of Baseline OCT RNFL Normative Classifications with Respect to Future VF Conversion
Sector PPV (95% CI) NPV (95% CI)
Temporal quadrant 16.7 (0.88–63.5) 75.6 (64.7–84.1)
Superior quadrant 38.1 (19.0–61.3) 80.6 (68.8–88.9)
Nasal quadrant 30.8 (10.4–61.1) 77.3 (65.9–85.9)
Inferior quadrant 50.0 (30.4–69.6) 87.1 (75.6–93.9)
9 o'clock 0 (0–53.7) 25.3 (16.7–36.2)
10 o'clock 30.8 (10.4–61.1) 77.4 (66.7–85.5)
11 o'clock 41.7 (16.5–71.4) 78.9 (67.8–87.1)
12 o'clock 35.7 (14.0–64.3) 78.4 (70.0–86.8)
1 o'clock 27.8 (10.7–53.6) 72.2 (46.4–89.3)
2 o'clock 13.0 (3.4–34.7) 72.3 (59.6–82.3)
3 o'clock 44.4 (15.3–77.3) 78.5 (67.5–86.6)
4 o'clock 38.5 (15.1–67.7) 78.7 (67.4–87.0)
5 o'clock 28.6 (12.2–52.3) 77.6 (65.5–86.5)
6 o'clock 45.0 (23.8–68.0) 82.4 (70.8–90.2)
7 o'clock 47.4 (25.2–70.5) 82.6 (71.2–90.3)
8 o'clock 16.7 (0.88–63.5) 75.6 (64.7–84.1)
Table 3.
 
Univariate and Multivariate Cox's Proportional Hazard Models for Prediction of VF Conversion
Table 3.
 
Univariate and Multivariate Cox's Proportional Hazard Models for Prediction of VF Conversion
Univariate Analysis Multivariate Analysis
HR (95% CI) P HR (95% CI) P
Baseline Characteristics
Age (per year) 1.028 (0.997–1.061) 0.078 1.036 (1.003–1.070) 0.32
Sex 1.369 (0.580–3.232) 0.474
Spherical equivalent (per D) 0.947 (0.922–0.998) 0.352
VF MD (per 1 dB) 0.768 (0.613–0.961) 0.021 0.788 (0.628–0.987) 0.038
VF PSD (per 1 dB) 1.541 (1.106–2.147) 0.011 1.07 (0.627–0.714) 0.887
Average RNFL thickness by OCT (per μm) 0.956 (0.916–0.998) 0.042 1.047 (0.983–1.114) 0.151
Number of abnormal sectors by OCT (per sector) 1.313 (1.110–1.553) 0.002 1.290 (1.099–1.587) 0.016
Mean IOP (per mm Hg) 1.007 (0.896–1.130) 0.912
CCT (per μm) 0.990 (0.976–1.004) 0.990
Baseline Abnormal OCT Sectors
Temporal 0.627 (0.084–4.67) 0.649
Superior 2.02 (0.83–4.88) 0.12
Nasal 1.84 (0.61–5.56) 0.279
Inferior 4.05 (1.65–9.90) 0.002 1.99 (0.73–5.38) 0.177
9 o'clock 0.31 (0.087–21.5) 0.428
10 o'clock 2.19 (0.50–9.52) 0.297
11 o'clock 2.21 (0.81–6.06) 0.122
12 o'clock 1.73 (0.63–4.72) 0.288
1 o'clock 1.37 (0.50–3.76) 0.539
2 o'clock 0.49 (0.14–1.68) 0.259
3 o'clock 3.08 (1.02–9.26) 0.046 0.66 (0.08–5.35) 0.695
4 o'clock 2.36 (0.86–6.49) 0.097 1.54 (0.38–6.25) 0.545
5 o'clock 1.32 (0.51–3.39) 0.571
6 o'clock 2.74 (1.14–6.62) 0.025 1.36 (0.50–3.73) 0.549
7 o'clock 3.70 (1.56–8.85) 0.003 2.49 (0.96–6.41) 0.059
8 o'clock 0.69 (0.092–5.18) 0.719
Follow-up Parameters
Mean follow-up IOP (per mm Hg) 0.991 (0.873–1.126) 0.891
×
×

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.

×