The fine structure of human extraocular muscle spindles and their potential proprioceptive capacity.

Twenty seven muscle spindles from six extraocular muscles removed following ocular enucleation from patients aged 58, 76 and 74 years were examined throughout all or most of their length by means of light and electron microscopy using serial transverse sections. Five others were prepared in longitudinal section. Twelve spindles of the superior rectus muscle from three sheep orbits were studied in a similar manner to provide a comparison. The human spindles contained a total of 90 (42%) nuclear chain and 5 (2%) nuclear bag fibres with the usual complement of sensory endings, and 120 (56%) fibres were anomalous with continuous, unattenuated myofibrils throughout their length, a constant width and peripherally placed nuclei. Eight anomalous fibres received sensory terminals similar in form to those of chain and bag fibres. Most (26) spindles contained at least one chain and one anomalous fibre. The periaxial space was limited or absent and the inner capsule was often segmented and in contact with the outer capsule. Abrupt termination of some chain fibres including several with one pole missing, together with evidence of fibre fragmentation and other structural anomalies, were indicative of degeneration. Eight further encapsulated fibre groups were identified as false spindles containing only anomalous fibres; associated nerves failed to terminate in the encapsulations. Sheep spindle content was of regular form, all spindles containing several chain and at least one bag fibre enclosed by an inner capsule and surrounded by a substantial periaxial space equatorially. The human extraocular muscle spindles have lost, either by aging or phylogenetically, the privilege of contractile chambers isolated by a fluid periaxial space from extrafusal fibre activity and sensory terminals are subject to the direct mechanical influences of anomalous intrafusal fibres. These, and the other departures from normal structure described, must jeopardize monitoring of muscle activity in the manner normally attributed to spindles and their capacity to provide useful proprioceptive information is questionable.