Graves' Orbitopathy. Группа авторов

57] by using a bowl or arc perimeter, with mean excursions shown in Table 4. An age-related decline in excursions, especially elevation, has been noted by some but not all studies [56, 57, 61]. What does not appear to be useful in assessing or monitoring muscle function is imaging. Neither MRI volumes nor ultrasound correlate with muscle function, clinical course, or subjective diplopia [62–65].

      5.Corneal pathology: While minor corneal pathology requires slit-lamp examination to detect punctate fluorescein staining, sight-threatening pathology is evident with simple torch examination. In this situation, the eyelids do not close gently to cover the cornea, which remains visible. The lower conjunctiva is generally red, and if ulceration has developed, a grey opacity or even an abscess will be seen in the inferior cornea. This constitutes an emergency.

      6.Visual disturbance: Clinical assessments for DON comprise the following:

      (a)best-corrected visual acuity of each eye, which is most accurately measured with a logMAR chart, although Snellen charts are more widely available;

      (b)colour vision testing in the blue/yellow axis is most likely to pick up early defects of DON; however, red-green pseudo-isochromatic charts (e.g., Ishihara) are more readily available and remain very useful in this context (see the section “How Do You Decide whether a Patient Has Dysthyroid Optic Neuropathy?”). Each eye is tested separately using a reading correction as required;

Table 4. Mean ocular excursions at all ages in degrees, after Mourits et al.^a [57] and Haggerty et al.^b [56]

Lateral rectus (0°) (abduction)	46.2a to 52b
Superior rectus (67°)	43b
Elevation centrally (90°)	33.8a
Inferior oblique (141°)	46b
Medial rectus (180°) (abduction)	47.5a to 51b
Superior oblique (216°)	49b
Depression centrally (270°)	58.4a
Inferior rectus (293°)	62b
By convention, ocular excursions away from primary fixation (gaze straight ahead) are designated as 0° for pure abduction, i.e., right gaze for the right eye and left gaze for the left eye. The degrees for each gaze direction follow from this reference point such that they increase in a clockwise direction for the right eye and an anticlockwise direction for the left.

      (c)pupil responses are assessed by the swinging flashlight test for a relative afferent pupil defect; artefacts can easily be produced if a consistent method is not followed, particularly in patients with manifest strabismus; the patient fixates on a distant target, and care is taken to give both eyes equal stimulation with the same alignment to the visual axis while the light is moved between alternate eyes;

(d)funduscopy will detect abnormal swelling or pallor of the optic disc and the presence of choroidal folds (Fig. 11) as well as giving valuable information on confounding pathology such as cataract and glaucoma; choroidal folds are thought to develop when the eyeball is mechanically deformed by the secondary effects of enlarged rectus muscles in a restricted space; the folds are horizontal and generally pass straight through the macula unlike retinal folds;

      (e)perimetry is reserved for eyes with suspicion of DON; automated perimetry is most commonly used.

Fig. 11. Fundus showing choroidal folds.

In addition to the above, the intraocular pressure is measured. High readings are commonly found in patients with orbital congestion [66], and although these may increase on upgaze in relation to a tight inferior rectus muscle, the reliability of this sign is poor [4].

How Reproducible Are These Assessments?

There are no peer-reviewed publications on the reproducibility of eyelid dimensions in GO although unpublished data from the co-author’s institution show that the intraclass correlation for palpebral aperture was good at 0.65. Eyelid dimensions in ptosis subjects have been found to be highly repeatable [63]; however, patients with GO frequently exhibit marked variability in upper eyelid positions and are likely to be more difficult to measure accurately.

Exophthalmos accuracy depends significantly on the model of exophthalmometer used together with technique. The Hertel exophthalmometer with straight footplates and a single mirror appears to be more accurate than other Hertel models [54], and although other types of exophthalmometer may be more reliable [64], they are much less commonly available. Reproducibility to within 2 mm is generally quoted [4, 51, 52], and unpublished data from the co-author’s institution showed the intraclass correlation to be 0.71 for exophthalmometry.

      Both UFOF and BSV measurements show high levels of accuracy with UFOF repeatable to within 8° for single muscle measurement [56], and BSV fields accurate to within 4%.

      The reproducibility of assessments for DON is unknown.

How Is Severity Best Classified?

As discussed in the section “What Value Does the Mnemonic ‘NOSPECS’ Have?”, the NOSPECS classification is unsuitable for the management of GO. Instead, EUGOGO has proposed a classification of severity of relevance to patient management [67]. It is based on the impact of GO on the patient’s quality of life and the risk of visual