Clinical Characteristics and Gait Analysis

OBSERVATION Gait Disturbances in Patients With Pontine Medial Tegmental Lesions Clinical Characteristics and Gait Analysis

Hiroshi Mitoma, MD; Ryoichi Hayashi, MD; Nobuo Yanagisawa, MD; Hiroshi Tsukagoshi, MD

Objective: To determine the clinical characteristics of sions elicited truncal ataxia without limb ataxia. In com- gait disorders in patients with pontine medial tegmental parison, pontine base lesions elicited limb ataxia with- lesions. out truncal ataxia and caused hemiparesis. (2) Instability was more severe and persistent in patients with the former Design: We compared features of gait disorders be- lesions than in those with the latter lesions. Slowness of tween patients with infarcts in the medial tegmentum and walking speed and prolongation of the double-support those with stroke in other areas of the pons (pathologi- period were clearly observed in the former group. (3) Elec- cal control subjects) by measuring electromyographic re- tromyographic changes characteristic of cerebellar ataxia sults of lower limb muscles and several biomechanical were clearly evident in patients with rostral medial teg- parameters. mental lesions. The electromyographic amplitudes of the gastrocnemius and tibialis anterior muscles were al- Patients: Two patients with infarcts in the rostral me- most constant throughout the gait cycle, resulting in the dial tegmentum and 4 control subjects. Two of the con- disappearance of the inherent periodic pattern of each trol patients had lesions in the pontine base, while the muscle. lesions in the other 2 were in the pontine tegmentum and base (combined lesions). Conclusion: Medial tegmental lesions in the rostral pons cause prolonged and severe unstable walking that re- Results: Patients with rostral medial tegmental lesions sembles spinocerebellar ataxic pattern, and impairment and controls with pontine base lesions showed unstable of the spinocerebellar loop might be the pathomecha- walking characterized by irregular angular displace- nism underlying such a gait disturbance. ments and foot pressures. However, they differed by the following 3 features. (1) Rostral medial tegmental le- Arch Neurol. 2000;57:1048-1057

ONTINE HEMORRHAGE or in- dial tegmentum and in the 4 pathological farct causes gait disorders control subjects. Gait analysis was also per- characterized by unstable formed and included recording of electro- broad-based walking.1 Pos- myographic (EMG) activities of lower limb sible structures responsible muscles, angular displacements of the hip Pfor the pontine stroke-induced gait dis- and leg joints, and floor reaction forces dur- turbance include the pyramidal tract, the ing free walking. cerebro-ponto-cerebellar pathway, the medial lemniscus, and the vestibular system.2(pp309-355),3 Involvement of a combina- RESULTS tion of these structures has been considered to determine the clinical picture of GENERAL ASPECTS OF GAIT such a disturbance. From the Kakeyu Rehabilitation To elucidate the pathomechanisms of Healthy Subjects Research Institute, gait disturbances in patients with pontine Maruko-machi, Nagano medial tegmental lesions, we describe 2 pa- An example of a free-walking pattern of (Drs Mitoma and Tsukagoshi), tients with infarcts in this region and 4 pa- a healthy subject is shown in Figure 3. and the Department of Medicine tients with stroke lesions in other regions Foot contacts and angular displacements (Neurology), Shinshu of the pons (pathological control subjects). occurred periodically, and displacement University School of Medicine, Matsumoto (Drs Hayashi and Two of the pathological controls had le- of center of body pressure (COP) was Yanagisawa), Japan. sions in the pontine base, and the other 2 straight from one foot to another (Figure Dr Mitoma is now with the had lesions extending in the pontine teg- 3, A-C). Foot pressures in the vertical Mitoma Neurological Clinic, mentum and base. Clinical features were direction showed a 2-peak pattern, corre- Toshima-ku, Tokyo, Japan. compared in patients with lesions in the me- sponding to step in and kick off (Figure 3,

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1048

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 SUBJECTS AND METHODS Hyperreflexia and the Babinski sign were present on the contralateral side (patient 4) or on both sides (patient 3). Moderate dysmetria, decomposition, and adiadochokine- PATIENTS AND CONTROLS sis were present in both limbs in both patients, whereas truncal ataxia was not evident. In patient 3, sensations of The clinical features of all patients are summarized in touch, pain, and temperature were decreased on the con- Table 1. Patients 1 and 2 suddenly experienced drunk- tralateral side, with the exception of the face. However, the enlike walking. On examination a few days later, both pa- vibratory sensation and proprioception were intact, and the tients had great difficulty in sitting and standing due to un- Romberg sign was not observed. Patient 4 showed no ab- steadiness. Patient 1 did not exhibit hemiparesis, limb ataxia, normalities of sensations without impairment of touch sen- or sensory impairment. Patient 2 showed a slight weak- sation in the left side of the face. Both patients had no nys- ness of the right upper and lower limbs but no limb ataxia tagmus but showed a slightly slurred speech. or sensory impairment. Eye movements were normal in both Patients 5 and 6 showed severe and long-term un- patients. During the following month, the condition gradu- stable gait similar to those with medial tegmental lesions ally improved. However, walking instability persisted, ne- and used canes during walking. Both patients showed slight cessitating the use of canes while walking. hemiparesis and hyperreflexia in contralateral limbs. Muscle On presentation, both patients showed swaying in all tonus was normal in both patients. Moderate dysmetria, de- directions during upright standing and walking. Tandem composition, and adiadochokinesis were noted on the con- gait was impossible. The muscle strength and tonus were tralateral side (patient 5) or on both sides (patient 6). Both normal. Deep tendon reflexes were increased on both sides. patients exhibited severe truncal ataxia. Patient 5 showed The Babinski sign was not observed in patient 1 but was impairments of touch, vibratory, and positional sensa- observed in patient 2. Both patients showed no dysmetria, tions on the contralateral side. The Romberg sign was ob- decomposition, or adiadochokinesis in the limbs. Touch, served in patient 5. Patient 6 showed no sensory distur- pain, and temperature sensations were intact in both pa- bances. Patient 5 had ipsilateral facial palsy, horizontal tients. Vibratory and positional sensations were also pre- nystagmus to the ipsilateral side, and slurred speech. served, and the Romberg sign was not observed. Distur- The interval between the stroke and the present ex- bances of cranial nerves were not present in both patients. amination was similar in patients 1 and 2 (average, 32.5 Patients 3 and 4 showed mild gait disturbances com- months) and control patients (mean±SD, 37.8±23.1 pared with patients with medial tegmental lesions. Al- months). In gait analysis, to minimize speed-dependent though they were unable to walk unsupported due to weak- changes, we also studied healthy elderly subjects (6 men ness and instability of the contralateral limbs in the short- and 6 women; mean±SD age, 75.4±5.8 years) whose term, both patients were able to walk unaided during the present examination. They showed slight hemiparesis. Continued on next page

Table 1. Clinical Characteristics of 6 Patients With Stroke Affecting the Pontine Area* (Cont.)

Time Gait Hemiparesis Age, Lesion Since the Disorder (Brunnstorm DTR Babinski Limb Truncal Touch Patients y/Sex Diagnosis Side Attack, mo Severity Recovery Stage) (rt/lt)† Sign (rt/lt) Ataxia Ataxia Sense Medial tegmental lesions Patient 1 69/M Infarction rt and lt 47 Severe – ↑/↑ −/− – + N Patient 2 60/M Infarction lt 18 Severe – ↑/↑ +/+ – + N Paramedian or lateral pons lesions Patient 3 48/M Hemorrhage lt 24 Mild + rt (5) ↑/↑ +/+ + rtϾlt – ↓ rt Patient 4 73/M Hemorrhage lt 24 Mild + rt (5) ↑/N +/– + rtϾlt – N Combined lesions Patient 5 59/F Infarction lt 31 Severe + rt (5) ↑/N −/− + rt + ↓ rt Patient 6 75/M Infarction lt 72 Severe + rt (5) ↑/↑ +/− + rtϾlt + N

*The tonus was normal in all patients. DTR indicates deep tendon reflexes; rt, right side; lt, left side; −, not observed; +, observed; ↑, augmented; N, normal; and ↓, decreased. †At the biceps brachii and triceps brachii muscles, the knee and the ankle.

B). Electromyographic recordings showed periodic pha- The stride was shortened, and the double-support pe- sic muscle activities (Figure 3, D). riod was prolonged. The floor reaction forces in the vertical axis formed multipeak patterns on both sides. The Patients With Medial Tegmental Lesions range of angular displacement was small in the left and right hips, knees, and ankles, and the pattern was dif- Figure 4, A, demonstrates the gait pattern of patient 1. ferent in each cycle. Phasic EMG activity was noted in Displacement of COP was irregular, and a small loop was both proximal muscles throughout the gait cycle; how- observed when COP shifted from one foot to another. ever, the periodic EMG pattern associated with the gait

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1049

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 preferable walking speed was slow (mean±SD, 63.1±16.3 the sagittal plane were measured using electrogoniom- cm/s). Informed consent was obtained from all subjects. eters.5 Angular displacement values were expressed relative Magnetic resonance imaging was performed using a 0.5-T to the leg position in a relaxed standing posture. superconducting system (repetition time, 3000 milliseconds; Electromyograms of the adductor magnus, gluteus me- echo time, 100 milliseconds). Slice planes were perpendicu- dius, biceps femoris, vastus lateralis, gastrocnemius or so- lar to the long axis of the brainstem, with a section thickness leus (GC), and tibialis anterior (TA) muscles were re- of 5 mm and an interval length of 6 mm. Topographic local- corded bilaterally using surface silver–silver chloride surface ization of the vascular lesion was determined using transverse electrodes placed 3 cm apart.6 Electromyographic signals anatomical templates.4 Infarcts in patients 1 and 2 were lo- were passed through a band-pass filter of 20 to 500 Hz. A cated in the medial tegmentum of the rostral pons (Figure 1, multichannel telemeter box (model 511X; NEC-Sanei, To- A, and Figure 2, A and B). Herein, we defined these lesions kyo, Japan) was attached to the lower back and used to trans- as medial tegmental lesions according to the definition of Sil- mit data of angular changes and EMG signals. verstein.2(pp13-53) Patients 3 and 4 had lesions located in the pon- Data were stored on a microcomputer (model PC- tine base (Figure 1, B, and Figure 2, C). These lesions were 9801; NEC-Sanei) after analog-to-digital conversion. The defined as paramedian or lateral pontine lesions.2(pp13-53) In pa- sampling rate for floor reaction forces and angular dis- tient 5, the area of infarct extended from the paramedian area placements was 100 Hz, while that of EMGs was 1000 Hz. to the medial tegmentum, while in patient 6, the ischemic area A gait cycle was divided into 4 phases: the first double- was located in the lateral area and the medial tegmentum (Fig- support period (phase 1), the single-support period (phase ure 1, C, and Figure 2, D). Furthermore, a magnetic resonance 2), the second double-support period (phase 3), and the imaging scan showed associated infarcts in 3 patients: a fo- swing period (phase 4). Electromyographic signals were rec- cal lesion in the thalamus (in patients 2 and 3) and diffuse ce- tified and integrated at a time constant of 50 milliseconds. rebral white matter abnormalities (in patient 6). The integrated EMG was summed through each phase and then divided by the duration of that phase (the time- GAIT ANALYSIS averaged EMG was calculated in microvolts per second). Data were expressed as the mean±SD. In comparing Gait analysis was performed in all patients but patient 6. Sub- the data of patients with those of healthy controls, the value jects were asked to walk at their own ordinary speed with- recorded in the patient was considered pathological when out support along a 6-m walkway, in which 2 force plates it exceeded the mean±2 SD of controls, according to the (2 m long and 80 cm wide) were serially arranged (model method by Knutsson and Richards.6 Using an unpaired t 1812A; Anima, Tokyo, Japan). The pressure exerted on the test, we examined statistical differences of spatiotemporal force plate by each foot was measured in 3 dimensions. gait parameters and angular displacements between the right Changes in angles around the hip, knee, and ankle joints in and left legs in each subject.

A B C

Rostral 1 Pain and Vibration 4 Temperature and Position Romberg Cranial Nerve 1 Senses Senses Sign Disturbance

NN– – 2 5 NN– – 4 2

↓rt N – – NN– ItV

N↓rt + lt VII and nystagmus to lt 3 4 6 NN– –

4 cycle was not clear in both distal muscles. Patient 2 4 showed a walking pattern similar to that of patient 1. 6

Caudal Pathological Control Subjects 3 RL RL RL Figure 4, B, shows the gait pattern of patient 3. Displace- Figure 1. Lesion sites in each type, superimposed in 4 schematic planes of ment of COP was irregular, the stride was short, and the the pons: 1, section through the decussation of the trochlear nerve; 2, single-support period of the right leg was shorter than that section through the entrance of the trigeminal nerve; 3, section through the of the left leg. The amplitude of the floor reaction forces principal sensory nucleus and motor nucleus of the trigeminal nerve; and 4, section through the vestibular and the facial nerve nuclei. A, Medial in the vertical axis decreased on the right side, and the floor tegmental lesions. B, Paramedian or lateral pons lesions. C, Combined reaction forces in the vertical axis showed a multipeak pat- lesions. Inset numbers represent patients 1 through 6; L, left side; and R, tern on both sides. The range of angular displacements right side.

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1050

C D

Figure 2. T2-weighted (repetition time, 3000 milliseconds; echo time, 100 milliseconds) magnetic resonance images of the lesions of typical cases in each type, classified according to the lesion site. A and B, Medial tegmental lesions (patients 1 and 2, respectively). C, Paramedian or lateral pons lesions (patient 4). D, Combined lesions (patient 5).

decreased in the right knee and ankle joints, and their pat- while the activity of right leg muscles was low com- terns were irregular. The EMG amplitudes of the right leg pared with that of the left leg. muscles were low, compared with those of the left leg. Pa- tient 4 exhibited similar EMG and kinematic patterns. KINEMATIC FACTORS The walking pattern of patient 5 showed features of walking of patients with medial tegmental lesions Patients with medial tegmental lesions and those with com- and of those with paramedian or lateral lesions (data bined lesions showed a widened stance, reduced walking not shown). Displacement of COP was irregular, with speed, and a prolonged double-support period beyond the a small loop. Stride was shortened, and the single- mean±2 SD values of controls (Table 2). In contrast, pa- support period was shorter in the right leg than in the tients with lesions in the paramedian or lateral pons showed left leg. The floor reaction forces in the vertical axis small changes in spatiotemporal parameters, being within showed multipeaks on both sides, while the floor reac- the mean±2 SD, except for a prolonged double-support tion forces in the vertical axis of the right leg period in patient 3 and a widened stance in patient 4. In decreased in amplitude. Patterns of angular excursions patients with paramedian or lateral pons lesions and com- were different on each gait cycle on both sides, and bined lesions, the single-support periods of legs on the con- angular displacements of the knee and ankle joints tralateral side of lesions were significantly shorter than those decreased in amplitude, especially on the right side. of ipsilateral legs (PϽ.01). The periodic EMG pattern associated with the gait Angular displacements decreased in both sides in pa- cycle was not clear in the distal muscles of both sides, tients 1 and 2; the decreases in knee and ankle joints were

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1051

0.2 L

Displacement, m –0.2 R

01 Displacement, m

B D Muscle

100 L Adductor Magnus R Fz L Force, % Force, 50 Gluteus Medius R

0 L Vastus 20 Lateralis R

Fy 0 Force, % Force, Voltage, mV Voltage, –20 L Biceps Femoris R C Hip 20 Flexion Extension 0 L Gastrocnemius R

L Knee Tibialis Flexion Anterior Extension R

Displacement, Degrees 0 L Ankle 10 R Dorsal 123 4 Plantar –10

0.5 mV

L R 1 s 123 4

Figure 3. Sample record of gait in a healthy elderly subject. A, Displacement of the center of pressure. B, Floor reaction forces in the vertical axis (Fz) and anterior-posterior axis (Fy), normalized by body weight. C, Angular displacements of the hip, knee, and ankle. In B and C, the solid line shows the trace of floor reaction forces and angular displacements of the right foot; dotted lines trace those of the left foot. D, Electromyogram (EMG) of the hip and lower leg muscles. L indicates left leg; R, right leg; 1, phase 1 (first double-support period); 2, phase 2 (single-support period); 3, phase 3 (second double-support period); and 4, phase 4 (swing period). The walking speed was 52.6 cm/s.

beyond the mean−2 SD of the control (Table 3). No sig- joints were small, beyond the mean−2 SD of the control nificant differences between the right and left legs were on both sides, and such a decrease was significantly smaller observed (PϾ.05), except for the knee joint of patient 2. on the contralateral side than on the ipsilateral side (PϽ.01). In patients with lesions in the paramedian or lateral pons, angular displacements on the side contralateral to the le- MUSCLE ACTIVITIES sion were significantly smaller than those on the ipsilateral side (PϽ.01). Furthermore, the decrease in the con- Figure 5, A, shows time-averaged EMGs during free tralateral knee joint of patient 3 and those of the walking in patients 1 and 2. The periodic activity pat- contralateral knee and ankle joints in patient 4 exceeded tern associated with the gait cycle was observed in each the mean−2 SD of the controls. In patient 5 with com- proximal muscle but was not seen in each distal muscle. bined lesions, angular displacements of the knee and ankle The amplitude of the time-averaged EMG of the GC

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1052

0.2 L

Displacement, m R RL –0.2

012 Displacement, m

b d Muscle

100 L Adductor Magnus R Fz L Force, % Force, 50 Gluteus Medius R

L 0 Vastus Lateralis 20 R

Fy 0

Voltage, mV Voltage, L Force, % Force, Biceps –20 Femoris R

c L Hip 20 Flexion Gastrocnemius Extension 0 R

60 L Tibialis Anterior Knee R Flexion Extension L L

Displacement, Degrees 0 R R R 10

Ankle –10 Dorsal 0.5 mV Plantar 1 s

L L R R R Continued on next page

Figure 4. Sample record of gait in patient 1 with a medial tegmental lesion (A) and in patient 3 with a lesion in the paramedian or lateral pons (B). a indicates displacement of the center of pressure; b, floor reaction forces in the vertical axis (Fz) and anterior-posterior axis (Fy); c, angular displacements of the hip, knee, and ankle; and d, electromyograms of the hip and lower leg muscles. In b and c, the solid line shows the trace of floor reaction forces and angular displacements of the right foot; dotted lines trace those of the left foot. L indicates left leg; R, right leg.

muscle was higher than the mean+2 SD of the control phases 1 and 4; the gluteus medius muscle during phases during phases 1 and 4. No difference was observed be- 2 and 4; the vastus lateralis muscle during phases 2, 3, and tween right and left legs. 4; the biceps femoris muscle during phase 3; and the GC In patients 3 and 4 (Figure 5, B), the amplitudes of muscle during phases 1 and 2 were higher than the cor- the time-averaged EMGs of the GC muscle during phase responding mean+2 SD of the control. 2 and of the TA muscle during phases 1, 2, and 3 on the In patient 5, the amplitude of the time-averaged EMG contralateral side of the lesion were lower than the mean−2 of each muscle was low on the contralateral side of the le- SD of the control. In contrast, the amplitude of the time- sion (data not shown). The amplitude of the gluteus medius, averaged EMG was augmented on the ipsilateral side. The vastus lateralis, and TA muscles during phase 1 was less than EMG amplitudes of the adductor magnus muscle during the mean−2 SD of the control. On the ipsilateral side, the

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1053

0.2 L

Displacement, m R RL –0.2

012 Displacement, m

b d Muscle

100 L Adductor Magnus R Fz

Force, % Force, 50 L Gluteus Medius R

0 L Vastus 20 Lateralis R

Fy 0 Force, % Force,

Voltage, mV Voltage, L –20 Biceps Femoris R c Hip 20 Flexion L Extension 0 Gastrocnemius R

L Tibialis Knee Anterior Flexion R Extension

Displacement, Degrees 0 L 10 Ankle R Dorsal Plantar –10 0.5 mV

1 s L R

time-averaged EMGs of the proximal muscles showed a nor- walking with irregular angular displacements and foot mal pattern throughout the gait cycle. In ipsilateral distal pressures. Palliyath et al7 defined features of cerebellar muscles,therewasnoclearperiodicpatternduetoaugmented ataxic gait as irregularity of stepping and lack of coordi- activity of the GC muscle during phases 1, 3, and 4, and of nation of limb movements. Thus, both types of gait dis- the TA muscle during phases 2, 3, and 4. turbances exhibited a cerebellar ataxic pattern. How- ever, our studies showed the following 3 differences in COMMENT ataxic symptoms or gait disturbances induced by lesions of the rostral medial tegmentum and pontine base. CHARACTERISTICS OF PONTINE First, patients with lesions in the medial tegmen- STROKE-INDUCED GAIT DISORDERS tum of the rostral pons showed unstable walking accompanied by truncal ataxia, whereas those with lesions in Patients with rostral medial tegmental lesions and pon- the pontine base showed unstable gait with limb ataxia. tine base lesions showed a common feature of unstable Second, instability was more severe and persistent in the

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1054

Double-Support Single-Support Subjects Stance Width, cm Stride Length, cm Speed, cm/s Period, s Period, s Healthy controls (n = 12) 18.8 ± 1.7 39.5 ± 9.7 63.1 ± 16.3 0.19 ± 0.04 0.41 ± 0.06 Medial tegmental lesions Patient 1 22.3 ± 1.9† rt, 16.7 ± 1.4† 20.7 ± 19.3† 0.49 ± 0.14† rt, 0.29 ± 0.07 lt, 17.3 ± 1.5† lt, 0.30 ± 0.09 Patient 2 27.2 ± 2.5† rt, 25.3 ± 4.5 30.0 ± 3.9† 0.37 ± 0.11† rt, 0.44 ± 0.12 lt, 27.5 ± 4.4 lt, 0.46 ± 0.14 Paramedian or lateral pons lesions Patient 3 16.0 ± 2.2 rt, 39.0 ± 1.3 62.0 ± 4.3 0.38 ± 0.12† rt, 0.33 ± 0.12 lt, 41.7 ± 3.6 lt, 0.41 ± 0.13 Patient 4 23.0 ± 1.4† rt, 27.5 ± 0.5 41.0 ± 6.0 0.19 ± 0.05 rt, 0.41 ± 0.11 lt, 25.9 ± 0.7 lt, 0.48 ± 0.13 Combined lesions Patient 5 26.7 ± 2.4† rt, 13.4 ± 1.2† 11.2 ± 15.5† 0.53 ± 0.22† rt, 0.26 ± 0.11† lt, 12.1 ± 1.5† lt, 0.45 ± 0.20

*For stance width, speed, and double-support period, the results of both legs were calculated simultaneously in each subject. For stride length and single-support period, the results of each leg were separately calculated in each subject. Data represent the mean ± SD values calculated from 40 to 55 gait cycles. Control data were obtained from 12 healthy subjects. rt indicates right leg; lt, left leg. †The value is beyond the mean±2SDofhealthy control subjects.

Table 3. Angular Displacements of the Hip, Knee, and Ankle Joints*

Displacement, Degrees

Subjects Hip Knee Ankle Healthy controls (n = 24 legs) 30.6 ± 5.4 58.6 ± 7.8 32.8 ± 7.0 Medial tegmental lesions Patient 1 rt, 23.3 ± 3.1 rt, 36.8 ± 2.3† rt, 18.6 ± 3.1† lt, 23.6 ± 0.8 lt, 34.3 ± 4.0† lt, 17.5 ± 2.0† Patient 2 rt, 20.1 ± 1.5 rt, 14.8 ± 2.0† rt, 16.8 ± 2.2† lt, 21.6 ± 1.8 lt, 29.1 ± 1.8† lt, 16.8 ± 2.2† Paramedian or lateral pons lesions Patient 3 rt, 30.1 ± 2.3 rt, 25.3 ± 6.7† rt, 24.3 ± 6.4 lt, 44.0 ± 2.6 lt, 55.0 ± 4.2 lt, 35.3 ± 4.3 Patient 4 rt, 22.8 ± 1.9 rt, 34.3 ± 5.7† rt, 14.8 ± 5.3† lt, 40.8 ± 2.0 lt, 51.2 ± 3.0 lt, 27.0 ± 2.8 Combined lesions Patient 5 rt, 22.0 ± 2.0 rt, 23.8 ± 4.4† rt, 13.6 ± 5.0† lt, 21.8 ± 1.7 lt, 31.3 ± 2.3† lt, 15.5 ± 2.1†

*For each subject, data from each leg were calculated separately. Data represent the mean ± SD values calculated from 15 to 20 gait cycles. Control data were obtained from 12 healthy subjects. rt indicates right leg; lt, left leg. †The value is beyond the mean±2SDofhealthy control subjects.

former group than in the latter group of patients. Slow- R.H., N.Y., and H.T., unpublished data, 1993). The pres- ness of walking speed and prolongation of the double- ent results showed that in the walking by patients with support period, which are compensatory reactions for in- medial tegmental lesions, GC and TA muscle activities stability in walking8 and reflect the severity of the gait were almost similar in each phase, resulting in the dis- disorder,6 were clearly observed in the former group. appearance of the periodic activity pattern of each muscle Third, EMG features characteristic of cerebellar ataxic associated with the gait cycle. Considered together, the gait were clearly observed in the walking of patients with gait abnormalities of patients with medial tegmental le- rostral medial tegmental lesions. In our previous study,9 sions in the rostral pons closely resembled the unstable features of cerebellar ataxic gait included augmented ac- walking caused by impairment of the cerebellar vermis, tivity of the GC and TA muscles during those phases as- whereas those of patients with pontine base lesions re- sociated with lack of muscle recruitment during normal sembled the unstable walking caused by deficits of the walking. Based on these changes, the periodic activity pat- cerebellar hemisphere. tern associated with the gait cycle was not clear in the In addition to differences in ataxia, another distin- GC and TA muscles. These features were observed in pa- guishable feature was that truncal ataxia was the only tients with lesions in the cerebellar vermis and in those symptom in patients with medial tegmental lesions in the with lesions in the hemisphere, although they were more pons. Lesions in the rostral medial tegmentum elicited marked in the former patients than in the latter (H.M., EMG and kinematic abnormalities on both sides. How-

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1055

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 A Right Leg of Patient 1 Right Leg of Patient 2 B Right Leg of Patient 3 Right Leg of Patient 4 Left Leg of Patient 1 Left Leg of Patient 2 Left Leg of Patient 3 Left Leg of Patient 4

Adductor Magnus Muscle Gluteus Medius Muscle Adductor Magnus Muscle Gluteus Medius Muscle 50 50 50 50

40 40 40 40

V/ms 30 30 V/ms 30 30 µ µ

20 20 20 20

Voltage, Voltage, 10 10 Voltage, 10 10

0 0 0 0

Vastus Lateralis Muscle Biceps Femoris Muscle Vastus Lateralis Muscle Biceps Femoris Muscle 100 100 100 100

80 80 80 75

V/ms 60 V/ms 60 60 µ 50 µ 40 40 40

Voltage, Voltage, 25 Voltage, 20 20 20

0 0 0 0

Gastrocnemius Muscle Tibialis Anterior Muscle Gastrocnemius Muscle Tibialis Anterior Muscle 100 100 100 100

80 80 80 80

V/ms 60 60 V/ms 60 60 µ µ

40 40 40 40

Voltage, Voltage, 20 20 Voltage, 20 20

0 0 0 0 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 Phase Phase Phase Phase

Stance Swing Stance Swing Stance Swing Stance Swing

Figure 5. A, Comparison of time-averaged electromyograms (EMGs) recorded during walking of patients 1 and 2 with medial tegmental lesions with those of healthy subjects. Open bars indicate the time-averaged EMGs of healthy subjects (n=24). B, Comparison of time-averaged EMGs recorded during walkingof patients 3 and 4 with lesions in the paramedian or lateral pons with those of healthy subjects. Open bars indicate the time-averaged EMGs of healthy subjects (n=24). Data are given as the mean±2 SD. Results of each leg were calculated separately in each subject. Phase 1 indicates the first double-support period; 2, the single-support period; 3, the second double-support period; and 4, the swing period.

ever, no hemiparesis, sensory impairment, or vestibular and persistent truncal ataxia without limb ataxia, and symptoms were noted in these patients. In contrast, in walking closely resembled ataxic gait caused by impair- walking by patients with pontine base lesions, limb ataxia ment of the vermis. Stroke-related lesions closely over- on the contralateral side was accompanied with hemi- lapped the area where axons of the fastigial and inter- paresis. positus nuclei neurons are distributed.14-16 Thus, it appears that lesions in the medial tegmentum of the rostral pons POSSIBLE STRUCTURES RESPONSIBLE FOR damage the output fibers from the cerebellar nuclei on GAIT DISORDERS INDUCED BY ROSTRAL the vestibulospinal and reticulospinal tracts, so as to en- MEDIAL TEGMENTAL LESIONS dow the spinocerebellar ataxic nature in this dorsal lesion– induced gait disorder. Differences in impairment of cerebellar afferent or effer- The cerebellar hemisphere receives input signals from ent fibers would explain the difference in ataxia be- the cerebral cortex via the pontine nucleus, and sends tween rostral medial tegmental lesion– and pontine base output signals to the cerebral cortex through the dentato- lesion–induced gait disorders. thalamo-cortical pathways.10 This loop is recruited for The cerebellar vermis receives proprioceptive in- adaptive control of voluntary movements.10 Our pa- formation from the periphery through the spinocerebel- tients with lesions in the pontine base showed ataxic hemi- lar tracts, and in turn sends output signals carried by ves- paresis, and the features of the ataxia resembled cerebel- tibulospinal and reticulospinal neurons in the tegmentum lar hemisphere ataxia. Thus, the cerebrospinal and field via cerebellar nuclei.10 This cerebellospinal loop is cerebro-ponto-cerebellar pathways would have been dam- known to play a crucial role in the regulation of step- aged at the pontine base in these patients,3 which was ping and stabilization.11,12 Hayashi et al13 reported that a confirmed by magnetic resonance imaging observa- 3-Hz postural oscillation, a symptom of cerebellar ver- tions. mis, occurred in patients with strokes in the dorsal pons Such segregated impairments of cerebellar afferent and indicated that such impairment could cause symp- or efferent fibers in the pons appear to adequately ex- toms related to the spinocerebellar loop. The present pat- plain the features of pontine stroke–induced gait distur- tern in rostral medial tegmental lesions included severe bances.

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1056

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 4. Nieuwenhuys R, Voogd J, van Huijzen C. The Human Central Nervous System: CONCLUSIONS A Synopsis and Atlas. Rev 3rd ed. Berlin, Germany: Springer-Verlag; 1988. 5. Finley FR, Cody KA, Finizie RV. Locomotion patterns in elderly women. Arch Phys We reported herein that lesions in the medial tegmen- Med Rehabil. 1969;50:140-146. tum of the rostral pons led to unstable gait, which was 6. Knutsson E, Richards C. Different types of disturbed motor control in gait of hemi- not associated with hemiparesis, limb ataxia, sensory im- paretic patients. Brain. 1979;102:405-430. 7. Palliyath S, Hallet M, Thomas SL, Lebiedowska MK. Gait in patients with cer- pairment, or vestibular-related symptoms. The severe in- ebellar ataxia. Mov Disord. 1988;6:958-964. stability persisted for more than 2 years after the stroke. 8. Conrad B, Benecke R, Carnehl J, Ho¨hne J, Meinck HM. Pathophysiological as- The EMG and kinematic properties of this type of gait pects of human locomotion. In: Desmedt JE, ed. Motor Control Mechanisms in abnormality were of spinocerebellar nature. Impair- Health and Disease. New York, NY: Raven Press; 1983:717-726. 9. Mitoma H, Hayashi R, Yanagisawa N, Tsukagoshi H. Characteristics of parkin- ment of the spinocerebellar loop might be one of the un- sonian and ataxic gaits: a study using surface electromyograms, angular dis- derlying pathomechanisms of marked instability. placements, and floor reaction forces. J Neurol Sci. 2000;174:22-39. 10. Brodal A. Neurological Anatomy in Relation to Clinical Medicine. Rev 3rd ed. New Accepted for publication November 30, 1999. York, NY: Oxford University Press Inc; 1981. 11. Mori S. Integration of posture and locomotion in acute decerebrate cats and in Reprints: Hiroshi Mitoma, MD, Mitoma Neurological awake freely moving cats. Prog Neurobiol. 1987;28:161-195. Clinic, 1-2-10 Minami-Ikebukuro, Toshima-ku, Tokyo 171- 12. Dietz V. Human neuronal control of automatic functional movements: interac- 0022, Japan. tion between central programs and afferent input. Physiol Rev. 1992;72:33-69. 13. Hayashi R, Tako K, Tokuda T, Yanagisawa N. Three-hertz postural oscillation in patients with brain stem or cerebellar lesion. Electromyogr Clin Neurophysiol. REFERENCES 1997;37:431-434. 14. Walberg F, Pompeiano O, Westrum LE, Hauglie-Hanssen E. Fastigioreticular fibers in the cat: an experimental study with silver methods. J Comp Neurol. 1962; 1. Nutt JG. Gait disorders. In: Jankovic J, Tolosa E, eds. Parkinson’s Disease and 119:187-199. Movement Disorders. Baltimore, Md: Urban & Schwarzenberg; 1988:377-383. 15. Batton RR III, Jayaraman A, Ruggiero D, Carpenter MB. Fastigial efferent pro- 2. Vinken PJ, Bruyn GB, eds. Handbook of Clinical Neurology. Vol 12. Amsterdam, jections in the monkey: an autoradiographic study. J Comp Neurol. 1977;174: the Netherlands: North-Holland Publishing Co; 1969. 281-306. 3. Fisher CM, Cole M. Homolateral ataxia and crural paresis: a vascular syndrome. 16. Shinoda Y, Futami T, Mitoma H, Yokota J. Morphology of single neurons in the J Neurol Neurosurg Psychiatry. 1965;28:48-55. cerebello-rubrospinal system. Behav Brain Res. 1988;28:59-64.

Call for Papers

Theme Issues

The ARCHIVES is planning a theme approach for several issues in 2000 and 2001. The topics for these special issues will include cerebrovascular diseases and stroke, demyelinating diseases and multiple sclerosis, the epilepsies, neuromuscular diseases, aging of the nervous system and the dementias, molecular and ge- netic neurology, and ethical issues in neurology. Papers are requested for con- sideration that represent important information related to these areas. Both clinical and basic science papers are requested. We look forward to receiving your contributions to these important issues.

Roger N. Rosenberg, MD Editor

(REPRINTED) ARCH NEUROL / VOL 57, JULY 2000 WWW.ARCHNEUROL.COM 1057