Mania following deep brain stimulation for Parkinson’s disease

Cases.

All patients had presurgical motor fluctuations with disabling “on” dyskinesias and severe parkinsonism in “off” periods (table 1). None had a history of psychiatric disorders, impaired cognition, or mood fluctuations with dopaminergic drugs. The quadripolar electrodes (DBS 3389, Medtronic, Minneapolis, MN) were implanted stereotactically in a single operative session using microelectrode confirmation. The STN target was defined by brain CT, and the postoperative location of the electrodes was confirmed by MRI. All preoperative medications continued to be given at the same doses, and DBS with the lower contact of the electrodes (see table 1) was initiated after 24 hours using external screener stimulators (Medtronic, 3625) and, at day 7, 2 subcutaneous programmable pulse generators (ITREL 2, 7424, Medtronic). There was no rationale for the unusual practice of initially stimulating through the lowest contact in each patient other than the routine taken from our previous experience in DBS of the pars interna of the globus pallidus. Active electrode contacts and stimulation parameters were progressively adjusted in the follow-up and resulted in significant improvement in parkinsonism (see table 1). The initial monopolar stimulation through the lower contacts of the quadripolar electrodes was changed in each patient at the end of the 1st postoperative week because they all showed a better motor response with higher contacts. In all three cases, this change was followed by a gradual resolution of mania over the following week. The diagnoses of mania were based on Diagnostic and Statistical Manual of Mental Disorders, 4th edition2 (DSM-IV), and severity was assessed with the Young Mania Rating Scale (YMRS).3 Blinded raters of the postsurgical MRI and the microrecording charts of the 15 STN-DBS patients correctly identified the three manic patients by both a lower positioning of the tip of the electrodes at the midbrain level surpassing the substantia nigra reticulata and a longer presumed STN activity in the microrecording (extending in all three manic patients from 0 to 70 mm), suggestive of excessive deepening of the tip of the electrode. A more precise anatomic localization of the tip of the electrode was made difficult by artifacts.

Discussion.

All three patients fulfilled DSM-IV diagnostic criteria for manic disorder due to a general medical condition2 because all had the classic features of mania (i.e., elation, overactivity, grandiose delusions) in the absence of impaired consciousness or hallucinations. Although the behavioral change could represent a psychological reaction to improvement in mobility4 after surgery or successful outcome of a life-threatening intervention, the current findings suggest that the misplacement of electrodes at the midbrain level—probably surpassing the substantia nigra reticulata—and the initiation of DBS with the lowest contacts were responsible for the manic behavior. The artifact induced by the electrodes makes the exact position of the tip of the electrode uncertain (figure). However, the electrode trajectory spared the limbic-related regions (orbitofrontal and basotemporal cortex, thalamus),5-7⇓⇓ previously linked with the emergence of manic behavior after brain injury. In our three cases, DBS might have affected projections destined from the midbrain to the orbitofrontal or anterior cingulate striato-pallido-thalamo-cortical circuits or fibers from the ventral tegmental area to the limbic striatum, causing symptoms typical of mania.5-8⇓⇓⇓ Intracranial stimulation of the ventral midbrain in experimental animals elicits self-stimulation, flight, ambivalence, and other complex behaviors that are phylogenetically involved in the development of motor and cognitive functions and likely also in the behavioral manifestations of major psychiatric illnesses.9 The tip of the electrode in our patients—caudal to the STN—might have affected one or more of the described neuroanatomic projections to the periaqueductal gray and its connections, mediating their behavioral alterations.9

• Download figure
• Open in new tab
• Download powerpoint

Figure. Postoperative coronal T1-weighted MRI of Patient 1 illustrating the position of the deep brain electrodes in the midbrain (similar in Patients 2 and 3).

The fact that in all three cases manic episodes were observed only after the stimulation was turned “on,” the recurrence of the manic syndrome after repeating the stimulation with the lower right contact, coupled with the disappearance of the manic state after modifying the stimulation to higher contacts in Patient 1 lend support to a stimulation effect. Nevertheless, the concurrent edema and partial lesion effect due to the implantation procedure may constitute a favorable scenario for mania to develop in the first postoperative days.

Contrarily, as it occurred with transient depression associated with DBS in the left substantia nigra reticulata in an earlier study,1 mood changes in all our three patients did not rapidly return to euthymia when the stimulation was turned off. Thus, we reasoned that systematic trials of stimulation trying to reproduce mania using the lower contacts might put our patients at risk for lengthy mood changes.

The question also arises why DBS induced transient acute depression in a previous case1 but induced mania in our three patients. Although this is not easy to solve because our knowledge about the action of DBS on neurotransmission is still limited,10 manic behavior in our patients would be the expected reaction if DBS enhances dopaminergic transmission facilitating strong positive emotional states (i.e., elation). Depression and mania are the opposite poles of a continuum of mood changes that may alternate in the same patient as a result of oscillatory changes in brain dopamine release,8 presumably associated in these PD cases to local changes induced by DBS in midbrain structures. Moreover, relatively subtle differences in electrode placement may cause the effects of DBS to largely surpass the nigra reticulata spreading to the retrorubral region and other ventral midbrain regions,9 such as the ventral tegmental area, as well as to other nondopaminergic nuclei, thus explaining contrasting mood responses (mania or depression).