Differentiating bilateral superior oblique paresis from sensory extorsion

Purpose To determine whether patients historically diagnosed with bilateral superior oblique paresis (BSOP) categorized into (1) immediate-onset and (2) gradual-onset torsional diplopia groups are also distinguishable on the basis of patterns of subjective misalignment in various directions of gaze, consistent with the gradual-onset group being caused by sensory extorsion rather than by BSOP. Methods The medical records of all patients diagnosed with BSOP, V-pattern esotropia, or V-pattern exotropia between 1978 and 2009 were retrospectively reviewed. Those patients with torsional diplopia were classified into immediate- or gradual-onset diplopia groups. The torsional misalignments measured by Lancaster red-green plots were compared, and the surgical outcomes were evaluated. Results Of 38 patients identified, 27 had immediate-onset and 11 gradual-onset diplopia. There was a statistically significant difference in the increase in extorsion from up- to downgaze between the immediate- versus gradual-onset group (17.8 versus -1.5, P < 0.001). Patients in the immediate-onset group fared significantly better with bilateral Harada-Ito procedures than with bilateral inferior oblique-weakening procedures (P = 0.02), whereas patients in the gradual-onset group fared equally well with either procedure (P = 0.72). Conclusions Extorsion in upgaze is largely absent in patients with immediate-onset BSOP but is present in both up- and downgaze in patients with gradual-onset sensory extorsion. Lancaster red-green testing aids in distinguishing these two groups. The bilateral Harada-Ito procedure is a better procedure for true BSOP, whereas a bilateral inferior oblique-weakening procedure may be preferred for patients with sensory extorsion.