Thalamic Deep Brain Stimulation: Title and subTitle BreakComparison Between Unilateral and Bilateral Placement FREE

ESSENTIAL TREMOR (ET) and Parkinson disease (PD) can produce high-amplitude, disabling tremors that can clinically and pathophysiologically overlap.^{1 - 7} In most cases, the tremor is bilateral, although one side may predominate.

Through mechanisms that are not yet well understood, high-frequency thalamic deep brain stimulation (DBS) improves tremor contralateral to the site of implantation.^{8 - 12} Results from these studies have demonstrated remarkably uniform tremor improvement of about 75% in those who undergo unilateral thalamic DBS. Although less frequently reported, bilateral thalamic DBS appears to safely improve bilateral tremors.^{13 - 16} There are, however, little data directly comparing the relative effects of unilateral with bilateral thalamic DBS. We report our results in 21 patients who underwent staged bilateral DBS to compare the efficacy and adverse events (AEs) 3 months after the first DBS placement with those 3 months after the second DBS placement.

Thirteen patients with ET and 8 with tremor-dominant PD recruited from the Baylor College of Medicine Parkinson's Disease Center and Movement Disorders Clinic, Houston, Tex, completed the entire protocol. Two other patients with ET underwent their second implantation without an adequate first side evaluation and were not included. The diagnosis of ET was based on that proposed by the Tremor Investigational Group.¹⁷ The diagnosis of PD required the presence of 2 of 3 cardinal manifestations (tremor, rigidity, and bradykinesia). All patients with PD reported at least transient improvement with levodopa, and all were taking some PD medications at the time of surgery. Exclusion criteria included significant dementia (Mini-Mental State Examination score <24), a history of brain surgery, significant comorbid disease that resulted in unacceptable surgical risk, and age older than 80 years. Tremor medications for patients with ET were discontinued 1 month before the study. Patients with PD continued to take medications, but no medication changes were allowed from 1 month before enrollment or during the study.

All preoperative baseline evaluations were performed within 3 days of surgery, and all second side (bilateral) evaluations were performed 3 months after the second implantation. The first side "3-month" evaluations were performed at 3 months after the first surgery (14 patients), at 2 to 2½ months after the first surgery (6 patients), and at 6 months after the first surgery (1 patient). The 6 patients who underwent the first side evaluation before 3 months did so at their request to expedite second side placement. The single patient with a 6-month evaluation missed the scheduled 3-month visit. The actual duration between first and second side implantations ranged from 72 to 1059 days (median, 139 days).

Adjustments of the stimulator device settings for optimal tremor control, with minimal AEs, were performed at both postoperative visits, and at other times as necessary. Adjustable device settings included voltage (1-8 V), pulse frequency (90-400 Hz), pulse width (100-450 µs), and the electrode montage (4 electrodes, 1.5 mm apart from each other).

Patients were assessed according to the North America Multi-Center Deep Brain Stimulation Trial protocol.⁸ All assessments were performed by a single investigator (W.O.) who was familiar with the study protocol. The ET evaluation included subjective questions based on the Unified Tremor Rating Assessment,¹⁷ clinical assessments of arm, leg, voice, head, face, and tongue tremors (0-4 scale based on amplitude of tremor). For example, for arm tremor, 0 indicates no tremor; 1, slight tremor, <1 cm; 2, moderate tremor, 1 to 2 cm; 3, marked tremor, 2 to 6 cm; and 4, severe tremor, >6 cm. Subjects also drew a series of spirograms, straight lines, and cursive writing samples and performed a water pouring test (0-4 scale) on both sides. Patients with PD were evaluated with the Unified Parkinson's Disease Rating Scale (UPDRS), which included a 0 to 4 rating of rest, posture, and kinetic tremors in the arms and legs. Overall subjective assessments (much worse to much better) and the Mini-Mental State Examination were scored in patients with ET and PD. Adverse events included all symptoms that the patient believed occurred immediately after surgery or only occurred while the device was activated.

At the 3-month second side evaluation, all patients also underwent a "blinded" motor evaluation identical to that used in the North America Trial except that both stimulators were randomized (random number generator) to either "on" or "off" at the device settings from the previous visit, without the rater being aware of the activation status. The blinded examination included the UPDRS part 3, with posture and kinetic tremor ratings for both arms and legs in patients with PD and tremor ratings of arms, legs, voice, head, face, and tongue in patients with ET. After stimulator adjustments, patients underwent the complete open-label assessments. We herein describe only the 3-month second side blinded data, as the first side blinded data were previously reported.¹² The surgical methods for device placement are also identical to those previously reported.¹²

Statistical evaluation consisted of Wilcoxon signed rank tests to compare preoperative with postoperative blinded and unblinded results, and postoperative first side with postoperative second side results. Data are given as mean ± SD unless otherwise indicated.

Twenty-three patients completed the 3-month bilateral DBS evaluations; however, 2 did not have adequate first side 3-month evaluations for comparison. The mean age of the 13 patients with ET who completed all data points was 71.5 ± 4.9 years, and the mean age of the 8 patients with PD was 71.4 ± 6.1 years. All patients with ET chose to initially implant the side that would improve their dominant hand. Of the 8 patients with PD, 6 elected to initially improve their dominant side.

After the initial implantation, visual assessments of tremor contralateral to the implantation site improved significantly (Table 1). Contralateral pouring and writing scores also improved. There were no significant (P = .22) changes in any midline tremor assessments or measurements ipsilateral to the implantation site. All subjective measures of functional improvement also significantly improved in patients with ET after the first implantation. Patients rated themselves as "markedly improved" (n = 10), "moderately improved" (n = 2), and "mildly improved" (n = 1). No patient believed that there was any worsening.

After implanting the second side, tremor measures contralateral to that side significantly improved, mostly without significantly changing those referable to the initial implantation. Writing scores contralateral to the initial side (dominant hand in all cases), however, did worsen after the second side implantation, but were still better than at baseline (Table 1). In 9 patients with ET, head tremor significantly improved only after the second implantation. Functional scores and disability scores tended to further improve after the second placement, but were not statistically different from those after the first implantation. When compared with their first implantation, patients rated themselves as markedly improved (n = 7), moderately improved (n = 4), mildly improved (n = 1), and "no change" (n = 1) after the second implantation. Mini-Mental State Examination scores did not change during the study: baseline, 27.8 ± 1.1; 3-month first side implantation, 28.5 ± 1.0; and 3-month second side implantation, 28.7 ± 1.4.

The blinded motor examination results closely resembled those of the unblinded data (Table 2). At baseline, the 2 groups were similar except that the off stimulation randomized group showed less leg tremor. In patients with ET randomized to on stimulation, the arm tremor (n = 12 arms) and leg tremor (n = 12 legs) ratings improved, whereas patients randomized to off stimulation (n = 10 for arms and legs) showed no improvement.

After the first implantation, total visual assessments of contralateral tremor significantly improved (Table 3). Subjective disability scores and total UPDRS part 2 subjective scores also improved. Patients rated themselves as markedly improved (n = 5), moderately improved (n = 2), and no change (n = 1). None of the patients believed that their tremor worsened.

After the second side implantation, total visual assessment of tremor contralateral to that side significantly improved (P = .01), without significantly affecting the first side tremor scores (Table 3). The 6 patients with PD who had facial tremor also improved only after the second implantation. Functional scores, however, were no longer significantly improved compared with the preoperative baseline status, although there was no significant worsening of functional scores between the 3-month first and the 3-month second side assessments. Nevertheless, after the second procedure, patients rated themselves as markedly improved (n = 3), moderately improved (n = 2), mildly improved (n = 1), no change (n = 1), and "mildly worse" (n = 1). Mini-Mental State Examination scores did not change during the study: baseline, 27.0 ± 2.0; 3-month first side implantation, 27.3 ± 1.9; and 3-month second side implantation, 27.5 ± 2.5.

The blinded PD scores closely resembled unblinded scores. Unfortunately, only 2 patients were randomized to "on" status, so no meaningful analysis could be performed (Table 4). In patients with PD randomized to on stimulation (n = 4 arms), the blinded arm tremor scores improved from 9.3 ± 0.9 to 1.0 ± 0.5, whereas patient arms randomized to off stimulation (n = 12) showed no change from their preoperative scores (Table 4). Concurrent adjustment of bilateral devices was generally more arduous than that of a single device. In almost all cases, the first side implantation required readjustment. There was, however, no clear pattern, as settings were increased in some and decreased in others, resulting in no overall difference (Table 5). There were also no differences between settings for PD and ET.

Adverse events were usually mild to moderate, but tended to occur more frequently in patients with ET (12 of 13 or 92%) than in patients with PD (4 of 8 or 50%), and more frequently in those with bilateral compared with (16 of 21 or 76%) unilateral placement (11 of 21 or 52%) (Table 5). The most problematic stimulation-related AEs were gait disorders and dysarthria. There were no serious perioperative AEs.

Our results show that unilateral and bilateral ventral intermediate nucleus DBS procedures are safe and effective treatments for ET and the tremor of PD. The addition of a second side DBS significantly improved tremor measures contralateral to that placement without meaningfully affecting tremor contralateral to the first side DBS by most measures in patients with ET and PD. Midline tremors, such as head and voice, also improved only after the second implantation. Subjectively, patients with ET robustly improved after the second implantation, whereas PD global impressions improved less and activity of daily living scores actually tended to worsen.

The most common AEs included dysarthria, gait or disequilibrium problems, and paresthesias, and were more problematic after bilateral implantation in both groups. The gait or disequilibrium problems were phenotypically varied but most resembled unsteadiness. Only one patient (a patient with PD) actually reported the onset of falls after implantation. Adverse events were more frequent in patients with ET; however, when they occurred, dysarthria and gait problems tended to be more significant in patients with PD. Overall, patients with ET had fairly consistent but mild problems that usually responded well to adjustments, whereas the AEs in patients with PD were inconsistent but more frequently compromised the benefits of DBS. Compared with other reports of thalamic stimulation, our patients tended to experience more gait difficulty but less paresthesia and dystonia. This could reflect differences in patient selection, differences in placement of the electrodes, or programming tendencies.

The lack of functional improvement seen after the second placement in patients with PD, as determined by UPDRS part 2, despite clear improvement in tremor, probably results from the fact that the dominant hand was already improved from the first placement in most cases. Furthermore, part 2 of the UPDRS is not weighted to assess disability directly resulting from tremor, and many of the functional assessments only require one hand. The improved global assessment after second implantation despite lack of part 2 UPDRS improvement also highlights the fact that tremor is the major concern for most patients with PD.¹⁸ This contrasts to the view of most physicians who believe that bradykinesia causes the most disability.

Several other factors could also explain the relative lack of subjective improvement in PD following the second implantation. First, although the results showed robust second side improvement, we were not always able to maximize bilateral tremor improvement without incurring intolerable AEs, usually balance problems. Second, gait and bulbar AEs, seen mostly with bilateral implantations, increased overall disability in patients with PD. Finally, disease progression in tremor and other aspects of PD between surgical procedures could have negated some observable functional improvement with second side placement. Although we did not believe that there was meaningful disease progression between surgical procedures in most cases, 2 patients with PD waited more than 12 months before their second implantation, which may have lessened their apparent improvement.

Our results are generally similar to those of others who have assessed bilateral thalamic DBS. Pahwa et al,¹⁵ in the only other report comparing bilateral with unilateral placement in the same patients, reported that 9 patients with ET had significant improvement in tremor examination and activity of daily living scores after bilateral DBS. Results tended to be better than those after the first procedure and were maintained for 1 year. Taha et al¹⁶ reported midline tremor improvement after bilateral placement but did not compare this with unilateral placement. The increase in AEs after the second implantation, especially balance and bulbar problems, also appears to be a consistent finding, although less than that seen with bilateral thalamotomies.^{19 - 21}

We included only those patients with PD in whom tremor was the main cause of disability. These patients tended to have relatively little bradykinesia and rigidity. We would not, therefore, advocate thalamic DBS in patients with PD in whom tremor was a minor contributor to overall disability. In contrast to ventral intermediate nucleus DBS, subthalamic DBS improves all the cardinal manifestations of PD.^{22 - 23} The extent of tremor suppression has been less thoroughly evaluated; however, if it improves tremor as much as ventral intermediate nucleus DBS, then subthalamic DBS should be considered even for tremor-dominant PD. Disadvantages of subthalamic DBS include its greater technical requirements, its more complex device programming, and the lack of Food and Drug Administration approval for the procedure in the United States.

Weaknesses of our study include the relatively small sample size, the short period of evaluation, and the intrinsic inability to control for disease progression. All patients were given the choice of whether to undertake the second procedure; therefore, our group of patients who chose the second procedure are biased toward more severe cases. Nevertheless, our results provide further support for the use of unilateral and bilateral ventral intermediate nucleus DBS in patients with ET and in selected patients with tremor-dominant PD who are refractory to medications.

Accepted for publication June 23, 2000.

This study was supported in part by grants from Medtronic Inc, Minneapolis, Minn; by the Methodist Hospital, Houston, Tex; and by the National Parkinson Foundation, Miami, Fla.

Corresponding author and reprints: William Ondo, MD, Department of Neurology, Baylor College of Medicine, 6550 Fannin Dr, Suite 1801, Houston, TX 77030 (e-mail: wondo@bcm.tmc.edu).