We evaluated patients with hypertropia compatible with a diagnosis of superior oblique (SO) palsy to ascertain whether the 3-step test (3ST) can distinguish SO atrophy characteristic of trochlear nerve pathology from masquerading conditions.

Prospective cross-sectional study.

In an academic practice, we performed quasi-coronal plane, surface coil magnetic resonance imaging in 83 patients clinically diagnosed with SO palsy. We evaluated alignment, SO cross-sectional area, SO contractility, and rectus muscle pulley positions.

A total of 57 patients with mean age 39 years (SD = 21 years) had unilateral SO palsy manifested by SO atrophy (22 congenital and 35 acquired). There was normal SO size in 26 patients with an average age of 39 years (SD =16 years) considered masquerades (8 congenital and 18 acquired). Maximum palsied SO cross-section averaged 9.5 ± 3.8 mm², less than 18.4 ± 3.9 mm² contralaterally (P < 10⁻²⁴). In masquerades, maximum hypertropic SO cross-section was 20.7 ± 3.1 mm², which was not different from the hypotropic SO or the contralesional muscle in SO palsy. Head tilt testing in masquerades was indistinguishable from SO palsy. In SO palsy, central hypertropia averaged 13.2 ± 9.4Δ, increasing to 21.1 ± 14.0Δ in ipsilateral tilt, and decreasing to 4.3 ± 5.3Δ in contralateral tilt. In masquerades, central hypertropia averaged 13.1 ± 8.7Δ, and was 17.7 ± 11.1Δ in ipsilateral and decreasing to 4.9 ± 5.1Δ in contralateral tilt. Upright hypertropia was larger at 17.7 ± 9.9Δ in congenital than 12.0 ± 8.4Δ in acquired SO palsy (P = 0025) but was indistinguishable from congenital masquerades. Contractile change in SO cross-section was bilaterally similar in masquerades. Relevant coordinates of rectus pulleys were similar bilaterally in masquerades.

The 3ST pattern characteristic of unilateral SO palsy may be mimicked in all respects by masquerades.