[ARTICLE] Time Course of Wrist Hyper-Resistance in Relation to Upper Limb Motor Recovery Early Post Stroke – Full Text

Abstract

Background. Patients with an upper limb motor impairment are likely to develop wrist hyper-resistance during the first months post stroke. The time course of wrist hyper-resistance in terms of neural and biomechanical components, and their interaction with motor recovery, is poorly understood. Objective. To investigate the time course of neural and biomechanical components of wrist hyper-resistance in relation to upper limb motor recovery in the first 6 months post stroke. 

Methods. Neural (NC), biomechanical elastic (EC), and viscous (VC) components of wrist hyper-resistance (NeuroFlexor device), and upper limb motor recovery (Fugl-Meyer upper extremity scale [FM-UE]), were assessed in 17 patients within 3 weeks and at 5, 12, and 26 weeks post stroke. Patients were stratified according to the presence of voluntary finger extension (VFE) at baseline. Time course of wrist hyper-resistance components and assumed interaction effects were analyzed using linear mixed models. 

Results. On average, patients without VFE at baseline (n = 8) showed a significant increase in NC, EC, and VC, and an increase in FM-UE from 13 to 26 points within the first 6 months post stroke. A significant increase in NC within 5 weeks preceded a significant increase in EC between weeks 12 and 26. Patients with VFE at baseline (n = 9) showed, on average, no significant increase in components from baseline to 6 months whereas FM-UE scores improved from 38 to 60 points. 

Conclusion. Our findings suggest that the development of neural and biomechanical wrist hyper-resistance components in patients with severe baseline motor deficits is determined by lack of spontaneous neurobiological recovery early post stroke.

Introduction

Recovery of post-stroke upper limb motor impairment is heterogeneous. Recent studies suggest that most patients follow a predictable pattern of spontaneous neurobiological recovery within the first 3 months after stroke, while 20% to 30% of the patients fail to show any motor recovery.^1–3 Previous observational studies have shown that early control of voluntary finger extension (VFE) is an important determinant of upper limb motor recovery at 6 months post stroke.^4,5 In addition, several studies suggested that patients with poor motor recovery are likely to show increased resistance to passive muscle stretch,^6–8 that is, hyper-resistance. This hyper-resistance is hypothesized to be caused by a poorly understood and complex interaction between pathological neuromuscular activation due to damage to descending pathways as well as non-neural biomechanical changes in the muscles and soft tissues spanning the joint post stroke.^9–11 The neural components of hyper-resistance may be divided into velocity-dependent stretch hyperreflexia (altered set point and/or gain of the stretch reflex, ie, spasticity following the definition of Lance)^9,12 and non-velocity-dependent involuntary activation (ie, increased background levels of contraction).^11,13 Biomechanical components of joint hyper-resistance include altered tissue properties, for example, elasticity, viscosity and muscle shortening.^11,14

In particular, there is a lack of knowledge about the time course of wrist hyper-resistance in terms of its neural and biomechanical components, and its interaction with motor recovery early post stroke,¹⁵ yet this is important for understanding observed improvements in motor control of the upper paretic limb in terms of behavioral restitution and compensation strategies.^16,17 Development of the velocity and non-velocity-dependent neural components, among which spasticity, as a reflection of reorganization of spared descending pathways, might reflect neural repair processes during upper limb recovery, further influencing behavioral restitution.^16,18 Moreover, information about the time course of different components of wrist hyper-resistance may help to optimize individualized treatment decisions, for example, when and to whom to apply botulinum toxin treatment^19–21 during the early post-stroke phase. Considering the target mechanism of botulinum toxin, blocking neural signal transmission to the muscle, it is expected that patients with an increased neural component of wrist hyper-resistance will benefit most from this treatment. Recently, a new measurement technique, called NeuroFlexor (Aggero MedTech, AB), has been developed for the quantification of neural and biomechanical elastic and viscous components of wrist hyper-resistance, which has proved to be valid^22,23 and reliable^23,24 in patients with chronic stroke.

The first aim of the present study was to investigate the time course of wrist hyper-resistance in the first 6 months post stroke, separated into its neural and biomechanical elastic and viscous components. This was done in patients with and without VFE within 3 weeks post stroke, in relation to the critical time-window of spontaneous neurobiological recovery as reflected by improvements observed using the Fugl-Meyer upper extremity scale (FM-UE). Findings were compared with healthy reference data. The second aim was to investigate the association between neural and biomechanical elastic and viscous components of wrist hyper-resistance in the first 6 months post stroke.

We hypothesized that in patients without VFE within 3 weeks post stroke, in the absence of spontaneous neurobiological recovery, both the neural and biomechanical components of wrist hyper-resistance would gradually increase over time.¹⁵ In addition, we hypothesized that the neural component would increase within the time-window of spontaneous neurobiological recovery, while an increase of the biomechanical components would not be restricted to this specific time-window. In a similar vein, we hypothesized that an increase in the neural component would be accompanied by an increase in biomechanical components, in reaction to a pathological neuromuscular activation. In patients with VFE within 3 weeks post stroke, that is, those showing spontaneous neurobiological recovery, components of wrist hyper-resistance were hypothesized to normalize to values seen in age- and gender-matched healthy subjects.

Methods

Participants

All patients with stroke who were admitted to Revant Rehabilitation Center Breda, The Netherlands, for inpatient rehabilitation were screened for eligibility between July 2015 and July 2016. The inclusion criteria for this study were (1) a first-ever ischemic stroke within the past 3 weeks, with an initial upper limb deficit as defined by the National Institutes of Health Stroke Scale item 5 a/b score >0 (ie, not able to hold the affected arm at a 90° angle for at least 10 seconds), (2) ≥18 years of age, (3) able to sit in a chair for at least 1 hour, and (4) sufficient cognitive ability to follow test instructions as indicated by a score higher than 17 on the Mini Mental State Examination.²⁵ Exclusion criteria were (1) a history of other neurological impairments and (2) limitations of arm-hand function of the affected side prior to the stroke. A group of healthy, right-handed, age- and gender-matched adults without wrist function restrictions served as a reference group. Ethical approval was obtained from the Medical Ethics Reviewing Committee of the VU University medical center, Amsterdam, The Netherlands (protocol number 2014.140). In accordance with the Declaration of Helsinki (2013), all participants gave written informed consent.

Study Design and Procedures

In this prospective cohort study, repeated measurements were performed at fixed times post stroke, that is within 3 weeks, and at 5, 12, and 26 weeks. The first measurement was performed as soon as possible after stroke onset, with more intensive repeated measurements within the window of nonlinear spontaneous neurobiological recovery within the first 12 weeks post stroke³ and a follow-up measurement at the start of the chronic phase after stroke.²⁶ Demographics and stroke characteristics were collected at baseline. All measurements were performed by a trained assessor. In the healthy controls, neural and biomechanical components of wrist hyper-resistance were determined for the dominant arm. All patients received usual care. The use of botulinum toxin injections was recorded throughout the study period.

At baseline, patients were stratified into 2 groups, based on the presence or absence of VFE within 3 weeks post stroke^4,5: (1) a group of patients showing any VFE, according to the FM-UE item of finger extension >0, within 3 weeks and (2) a group of patients showing no VFE (FM-UE item finger extension = 0) within 3 weeks.

Outcome Measures

Neural and biomechanical elastic and viscous components of resistance to passive wrist extension were assessed with a validated and commercially available measurement technique, the NeuroFlexor, feasible for use in clinical practice (Figure 1).²² This motor-driven device imposes isokinetic wrist displacements with extended fingers from 20° palmar flexion to 30° dorsal flexion at 2 controlled velocities (5 and 236 deg/s), for which a minimal passive wrist extension of 40° is needed. A force sensor, placed underneath the moveable hand platform, measures the resistance trace during the passive wrist movement. The participant was seated comfortably parallel to the device with the shoulder in 45° of abduction, 0° of flexion, the elbow in 90° of flexion, with the forearm fastened to the device in pronation, and the hand with extended fingers fastened to the hand platform. Participants were instructed to relax their arm and to look ahead of them during the measurements. The experimental session consisted of 5 slow movements (5 deg/s) followed by 10 fast movements (236 deg/s). The first movement at both velocities was excluded from analysis to avoid bias from startle reflexes and mechanical hysteresis. The resting torque of the hand before onset of stretch was subtracted from the resistance traces prior to further calculations. Using the biomechanical model described by Lindberg et al,²² the different components of wrist hyper-resistance, that is, the velocity-dependent part of the neural component (NC), the biomechanical elastic component (EC), and viscous component (VC), were derived from the resistance traces (using the NeuroFlexor Scientific v0.06 software program, Supplemental File 1 and Supplemental Figure 1). The NC was determined as a derivative of the velocity-dependent resistance to passive wrist extension, which is to reflect the neural, velocity-dependent part of wrist hyper-resistance, that is, assumed proxy of spasticity as defined by Lance,¹² not including the non-velocity-dependent part of neural activity, that is, involuntary background activation. The length-dependent EC was determined as the resistance at the end of the slow movement. It was assumed that the velocity-dependent VC was highest during the initial acceleration and continued at a lower level, that is, 20%, during further extension movement. The developers of the NeuroFlexor have previously underpinned the validity of the NC based on 3 arguments: (1) the NC as measured by the device was reduced after an ischemic nerve block, (2) the NC correlated with the integrated electromyography (EMG) across subjects and in the same subject during the ischemic nerve block, and (3) the NC was found to be velocity dependent.²² In a recent study,²³ the NeuroFlexor method was suggested to be construct-valid against clinical assessments using the modified Ashworth and Tardieu scales. In addition, good to excellent reliability was shown for the quantification of the different components.^23,24 As a result of the positioning of the fingers, the measured resistance was a combination of resistance caused by wrist and finger flexor muscle groups. Measurements were performed twice at the same occasion, and mean values were used for further analysis.

Figure 1. NeuroFlexor method. (A) Measurement set-up. (B) An example of the force trace (red line) obtained during a fast movement (236°/s). (C) An example of the force trace obtained during a slow movement (5 deg/s). The blue line represents the angle of the wrist joint. The recorded force traces, measured in newton (N), are analyzed by a biomechanical model, which results in the quantification of the velocity-dependent part of the neural component (NC), elastic component (EC), and viscous component (VC) of wrist hyper-resistance. The total measured resisting force (F_m) during passive wrist extension is a summation of passive elastic force (F_p), viscous force (F_v), reflexive force (F_r), and inertial forces of the limb and the moving parts of the device (F_in), described as: F_m(θ) = F_p(θ) + F_v(θ) + F_r(θ) + F_in(θ), where θ denotes a specific angle. In the model, 4 force points in the resistance trace of the slow and fast movements are used to estimate the different components of the total measured passive force. P0 is the resting torque of the hand before onset of stretch. Two force points are defined within the fast passive wrist extension movement (236 deg/s): P1, the initial peak in resistance, and P2, the late peak in resistance. One force point (P3) is defined at the end position of the slow wrist extension movement (5 deg/s). Resting torque (P0) is subtracted from P1, P2, and P3 prior to further calculations. Detailed information about the biomechanical model can be found in Supplemental File 1.

[…]

Continue