advertisement
Disc-field relationships in EMGT and OHTS
Anders Heijl with input from Bo Bengtsson, Cristina Leske and Boel Bengtsson
We have been asked to comment on the apparent differences in the relative value of visual field testing versus disc photography in EMGT and OHTS. In EMGT, many more patients were found to progress based on the results of field testing than when disc photographs were used. In fact, 86% of the progressions were based on field outcomes alone, 13% on both visual field and optic disc outcomes, and only 1% on optic disc outcomes alone (Heijl et al., 2002). In OHTS, on the other hand, 55% of endpoints were defined by optic disc analysis, 10% by concurrent disc and field, and 35% by field changes alone (Kass et al., 2002). What are the explanations for these large differences?Firstly, we should note that the key issue in OHTS is to detect incident glaucoma damage in patients with ocular hypertension without signs of such damage, while, in EMGT, it is a question of detecting the progression of existing glaucoma damage in patients with early glaucoma. This is a fundamental distinction, since each trial is designed to identify very different stages of the disease process.
Secondly, EMGT and OHTS also differ markedly in their methods and criteria for defining field and disc outcomes. With regard to fields, both EMGT and OHTS use Humphrey Full Threshold tests, but EMGT uses the 76-point, 30-2 pattern, while OHTS employs the smaller 54-point, 24-2 pattern. The collection of additional perimetric data in EMGT cannot, in our opinion, explain the differences, although a few additional progressing eyes may have been revealed by testing those extra mid-peripheral test point locations. However, visual field interpretation is different in both EMGT and OHTS. EMGT uses Pattern Deviation Glaucoma Change Probability Maps (Bengtsson et al., 1997), which are based on a mathematical model differentiating between random and abnormal variability in glaucomatous fields. Included in the model are the initial deviation of the threshold value of each individual test point location, the eccentricity of the location, and the general status of the field, as expressed by the MD value. The criterion of (definite) EMGT visual field progression a minimum of three significantly deteriorated locations in three consecutive tests compared to a baseline of two fields (preceded by a minimum of two learning fields) (Leske et al., 1999), was defined after an extensive retrospective analysis of many thousands of glaucoma fields. Compared to this machine-based , objective criterion, OHTS uses grading performed by masked certified readers at a Visual Field Reading Center. Initially, results from two consecutive tests were used to determine field progression, and later from three tests.
With regard to discs, OHTS and EMGT obtained photographs every six months. OHTS used a Reading Center for analysis of the stereo disc photographs from that study. In EMGT, we used monoscopic disc photography, and our Disc Photography Reading Center utilized flicker chronoscopy in order to detect small changes in optic disc topography. Flicker chronoscopy can be a very sensitive type of analysis, especially if photographs are obtained in a controlled way, minimizing parallax errors. We have previously shown that this type of analysis reveals incident glaucoma damage in patients with high-risk ocular hypertension at approximately the same time as standard automated threshold perimetry (Heijl and Bengtsson, 1989).
The EMGT protocol was designed for high specificity; over-sensitive and non-specific analyses could prove disastrous in a controlled treatment trial such as EMGT, where examinations are repeated frequently and definite progression permits a change of treatment. Thus, optic disc photograph graders were masked as to sequence of photographs, and changes found by the flicker test had to be based on the consensus of two graders, as well as being confirmed by independent side-by-side grading by a third observer. Optic disc progression was defined by the presence of such confirmed changes in at least two sets of photographs. Our quality control scheme confirmed the high specificity of the EMGT optic disc grading protocol (Heijl et al., 2002).
In summary, as noted in this comment, any comparisons of disc and field findings in EMGT and OHTS must consider the large differences in their respective outcome measures. The measures used in each trial were designed to achieve different objectives, as well as being based on different methods and using different definitions and criteria.
We have begun a masked, quantitative, planimetric analysis of all EMGT
disc photographs, but will need another year of research work before being
able to provide a more detailed explanation of the apparent differences
between EMGT and OHTS results. However, we would expect that regression
analyses of disc parameters obtained from the many photographs of EMGT eyes
may increase the number of disc progressions, compared to those identified
from flicker analyses followed by masked side-by-side comparisons. Nevertheless,
it still seems likely that field analysis will prove superior to analysis
of optic disc photographs in these eyes with manifest glaucoma, which was
present from the start of the follow-up. The relative value of disc and
field analysis will always depend on the methods used, but disease stage
is probably also very important. As such, the advantages of disc analysis
that are suggested by OHTS results may be eliminated in patients with manifest
glaucoma who already have quantifiable field defects.
Note from the Editor: It would be very interesting to apply the type
of disc evaluation used in EMGT to OHTS discs. Unfortunately, the reverse
would not be possible because EMGT does not use stereophotographs.
References
- Bengtsson Boel, Lindgren A, Heijl A, Lindgren G, Åsman P, Patella M: Perimetric probability maps to separate change caused by glaucoma from that caused by cataract. Acta Ophthalmol Scand 75:184-188, 1997.
- Heijl A & Bengtsson Boel: Diagnosis of early glaucoma using flicker comparisons of serial disc photographs. Invest Ophthalmol Vis Sci 30:2376, 1989.
- Heijl A, Leske C, Bengtsson B, Hyman L, Bengtsson Boel, Hussein M and the Early Manifest Glaucoma Trial Group: Reduction of intraocular pressure and glaucoma progression. Results from the Early Manifest Glaucoma Trial. Arch Ophthalmol 120:1268-1279, 2002.
- Kass MA, Heuer DK, Higginbotham EJ et al: The Ocular Hypertension Treatment Study. A randomized trial determines that topical ocular hypotensive medication delays or prevents the onset of primary open-angle glaucoma. Arch Ophthalmol 120:701-711, 20022.
- Leske MC, Heijl A, Hyman L, Bengtsson B and the Early Manifest Glaucoma Trial Group: Early Manifest Glaucoma Trial, design and baseline data. Ophthalmology 106:2144-2153, 1999.
