With the anterograde tracers Phaseolus vulgaris-leucoagglutinin (PHA-L) and biotinylated dextranamine (BD), direct spinal connections from the upper cervical spinal cord (UC; C1 and C2) and the cervical enlargement (CE; C5-T1) were demonstrated in various striatal and limbic nuclei in both squirrel monkey and rat. Within each species and from each spinal level, the total number of terminals seen in the limbic and striatal areas was approximately 50-80% of the number seen within the thalamus. Labeled terminal structures were seen in the hypothalamic nuclei, ventral striatum, globus pallidus, amygdala, preoptic area, and septal nuclei. In both species, the number of labeled terminals in limbic and striatal regions was larger from UC than from CE, although the distributions to each nucleus varied with the specific lamina injected. In both species and from both UC and CE, approximately one-half of the projections to striatal and limbic areas terminated in the hypothalamus. The only region that demonstrated a topographical organization was the globus pallidus, where terminals from the CE were located dorsomedially to those from the UC. In the rat, UC and CE injections into the lateral dorsal horn and pericentral laminae resulted in the largest number of limbic and striatal terminations. The proportion of ipsilateral terminations was greatest when the medial laminae in the UC or the lateral dorsal horn in the CE received injections. Analysis of the morphology of these spinohypothalamic and spinotelencephalic terminals showed that, in the squirrel monkey, terminals from CE injections were larger than terminals from UC injections; no such size difference was evident in the rat. However, limbic and striatal terminals in the rat were generally larger than those in the squirrel monkey following injections into the UC or CE. The exact function of these direct spinal projections to various striatal and limbic areas in primates and in rodents remains to be determined. These findings, however, support recent imaging studies that suggest that the limbic system plays an important role in the mediation of chest pain, perhaps directly through these spinolimbic and spinostriatal pathways.

COI is able to rapidly identify areas of cortex containing elicited neuronal activity. The technique allows cortical activation maps to be made rapidly with a very high degree of spatial resolution. COI is reliable and consistent over time. COI, if used carefully, holds promise as an intraoperative technique to study both human and experimental animal cortical function.