Reply The Sensory-Limbic Model Suggests Testable Hypotheses About the Learned Component of the Affective Dimension of Pain FA. Lenz, t R. H. Gracely, t A. T. Zirh, t A. J. Romanoski, * P. Staats, and P. M. Doughertyt C asey is correct in stating that the thalamic data cannot support or challenge the hypotheses that "limbic structures are conditioned by pre- vious experience of pain with a strong affec- tive dimension" or that "thalamic corticolimbic" activation of these conditioned limbic circuits will reproduce this pain. The hypotheses are supported by connections between the area of the thalamus in which stimulation reproduced pain with a strong affective dimension and the parasylvian area of the cortex and amygdala in which stimulation can evoke sensory experiences along with emotions accompanying the original experience [5,24]. These cortically evoked sensory experiences did not include pain, although the exact structures implicated in nociceptive processing may not have been stimulated. This cortical connectivity is congruent with Mishkin's model of somatosensory memory mediated through connections from the parasylvian cortex to the amyg- dala and parahippocampal gyrus (reference 20 and Fig- ure 1 of the Focus article). Gloor has reported that stim- ulation of the amygdala evoked reproduction of sensory experiences, including their emotional tone [5]. The tha- lamic data are consistent with evidence that the pos- teroinferior region projects to the parasylvian cortex (see the discussion of anatomic connections in the From the Departments of Neurosurgeryt and Psychiatry: Johns Hop- kins Hospital, Baltimore, MD, and Neurobiology and Anesthesiology Branch,'] National Institute of Dental Research, National Institutes of Health, Bethesda, MD. Reprint requests: Frederick Lenz, Department of Neurosurgery, Meyer Building 7-113, Johns Hopkins Hospital, 600 North Wolfe Street, Balti- more, MD 21287-7713. Pain Forum 6(1):41-43,1997 Focus article), that both the posteroinferior region and parts of this cortical area are involved in pain-signalling pathways (see the section on physiologic evidence of nociceptive inputs to the region of Vc in the Focus article), and that these cortical areas may be involved in sensory memory through connections with the limbic system (see the section on evidence for involvement of the parasylvian cortex in nociceptive pathways and memory in the Focus article). However, we have no direct evidence of activation of the limbic system by the thalamic stimulation. Casey is correct in pointing out that plasticity of multi- ple central nervous system structures has been identi- fied during classical conditioning. However, stimulation- evoked memory has not been evoked from multiple structures. Stimulation in awake humans has been described in dorsal columns [6,23] the spinothalamic tract at several levels .I6,1 0,26,27], midbrain [9,22,28], cerebellar cortex [1] and nuclei [21], intralaminar/media [10,26-28] and lateral nuclei of the thalamus [4,7, 15,28], striatum and pallidum [3,16], and much of the human cortical surface [25]. Stimulation-evoked sen- sory responses with an abnormal "sense of reality, wealth of detail and sense of immediacy that sets them apart from the normal process of recollection" [24] have only been evoked from the posteroinferior thalamus (see the Focus article), parasylvian cortex [24], and amygdala [5]. Specifically, this type of experiential response was not evoked at other sites in which pain was evoked by stimulation [4,6,7,9,10,15,22,25-28]. These findings strongly suggest that these structures are specifically involved in sensory memory, as sug- gested by the Focus article and by Mishkin [20]. 41