DBS in the STn of DADepleted PD Patients Corrects Migration

The finding of separable storage and retrieval mnemonic deficits restricted in the temporal domain in PD was followed up by a new set of experiments asking whether distinct neural pathways underlie those deficits. To this end, we applied the same encode-decode design in a series of studies testing PD patients OFF levodopa undergoing deep brain stimulation (DBS) either in the subthalamic nucleus (STn) or the pallidum. The effect of DBS is reversible, and the stimulation parameters (principally site and voltage) can be varied to achieve the optimal therapeutic result (Benabid et al., 1991, 1998). Moreover, the reversibility of DBS means that patients' cognitive performance can be assessed both with (ON) and without (OFF) DBS. This in turn presents the possibility of identifying specific patterns of impairment, including cognitive ones, associated with the stimulated basal ganglia substructures.

DBS in the STn, which alleviates parkinsonism (akinesia, bradykinesia) quite dramatically, also corrects the migration effect (Limousin et al., 1995a, 1995b; Pollak et al., 2002). The similar order of magnitude in correcting migration by STn DBS and L-Dopa replacement therapy contrasted to a clear difference between the dopaminergic and stimulation effect on encode (Molinuevo et al., 2000). The encode distortion (i.e., proportional overestimation of both time targets) that we described previously as slowed encode is not reversed by STn DBS. Moreover, overestimation occurs under STn DBS regardless of whether feedback is available. It is as though during encoding a lengthened estimate is laid down in memory, and it is not corrected by the presence of feedback. These results are illustrated in Figure 20.4.

In the same PD patients, stimulation has a clear effect on motor symptoms that mimics the dopaminergic effects, as can be seen in the right panel of Figure 20.4. This is an unexpected and quite striking finding, implying that STn stimulation (1) reverses slowness of motor, but not cognitive functions; and (2) fails to alleviate a cognitive (mnemonic) deficit, otherwise corrected by L-Dopa. The neural basis for slowed encode remains unclear and is not to be attributed to neural networks that would involve STn. Contrasts and similarities between the dopaminergic and DBS effects on both encode and decode sessions are illustrated in Figure 20.5.

Alleviating a deficit in temporal memory retrieval by altering neural activity in the STn is, on the other hand, a second unique and novel finding suggesting the role of a specific basal ganglia pathway through this particular nucleus, or the nucleus itself, in cognitive function. The STn stimulation is regarded as essentially correcting (attenuating) the abnormally high neural activity of this nucleus associated with the parkinsonian state (Limousin et al., 1998). The effects reported here were obtained with voltages below 5 V in all patients. Although the exact volume through which stimulation exerts its action is not known, it is generally accepted that at less than 5 V the effect is restricted to a few millimeters around the stimulation contact (Limousin et al., 1998). The anatomical localization of each of the stimulating contacts was carefully determined in each patient (for more details, see Bejjani et al., 1997). Only patients with verified implanted electrodes in the STn were included in the study. One might hence consider the observed effects to reflect modified (reduced) neural activity within the boundaries of the STn. This would suggest that normal activity levels in STn play a role in inhibiting a previously learned temporal memory when attempting to estimate a more recently learned memory.

Although overactivity of the STn is thought to be the key abnormality underlying parkinsonian symptoms in human and nonhuman primates, the evidence is almost exclusively restricted in motor deficits (Mitchell et al., 1986, 1995). STn overactivity is believed to result by overactivity of the striatal outputs to the lateral segment of the globus pallidus (GPe), the so-called "indirect striatal output pathway," due to the lack of dopaminergic inhibition (Crossman, 1989; Miller and DeLong, 1987; Robertson et al., 1990, 1991; Young and Penney, 1984). This implies that the retrieval process may involve the indirect output striatal pathway as a whole, rather than the STn itself.

The dopaminergic effect as conceived by the model implies that deprivation of dopaminergic nigro-striatal innervation reduces the positive feedback via the direct path and increases the negative feedback via the indirect path (described below in more detail). We postulated that the slowed encode may be produced by a delay for neurons in the direct pathway to either initiate decaying or, just after initiation, o

o Train on DBS o Test on DBS

0 0

Post a comment