Proportional Motor Recovery After Stroke
Implications for Trial Design
Background and Purpose—Recovery of upper-limb motor impairment after first-ever ischemic stroke is proportional to the degree of initial impairment in patients with a functional corticospinal tract (CST). This study aimed to investigate whether proportional recovery occurs in a more clinically relevant sample including patients with intracerebral hemorrhage and previous stroke.
Methods—Patients with upper-limb weakness were assessed 3 days and 3 months poststroke with the Fugl–Meyer scale. Transcranial magnetic stimulation was used to test CST function, and patients were dichotomized according to the presence of motor evoked potentials in the paretic wrist extensors. Linear regression modeling of Δ Fugl–Meyer score between 3 days and 3 months was performed, with predictors including initial impairment (66 − baseline Fugl–Meyer score), age, sex, stroke type, previous stroke, comorbidities, and upper-limb therapy dose.
Results—One hundred ninety-two patients were recruited, and 157 completed 3-month follow-up. Patients with a functional CST made a proportional recovery of 63% (95% confidence interval, 55%–70%) of initial motor impairment. The recovery of patients without a functional CST was not proportional to initial impairment and was reduced by greater CST damage.
Conclusions—Recovery of motor impairment in patients with intact CST is proportional to initial impairment and unaffected by previous stroke, type of stroke, or upper-limb therapy dose. Novel interventions that interact with the neurobiological mechanisms of recovery are needed. The generalizability of proportional recovery is such that patients with intracerebral hemorrhage and previous stroke may usefully be included in interventional rehabilitation trials.
Clinical Trial Registration—URL: http://www.anzctr.org.au. Unique identifier: ANZCTR12611000755932.
Motor impairment is common after stroke, and motor recovery is crucial for regaining independence.1 For the upper-limb, most patients recover ≈70% of the difference between their baseline Fugl–Meyer (FM) score and the maximum FM score.2–7 For the remainder of patients, recovery is not proportional to initial impairment and is often poor. Previous studies have shown that in patients with first-ever ischemic stroke, the integrity of the corticospinal tract (CST) determines whether motor recovery will be proportional to initial impairment.5–7 Patients in whom transcranial magnetic stimulation can elicit motor evoked potentials (MEPs) in the paretic upper-limb within 5 days of stroke experience proportional recovery (≈70%) regardless of their initial impairment, whereas patients without MEPs experience limited recovery and poor motor outcomes.5 Proportional recovery has also been observed in nonmotor domains such as aphasia8 and thus reflects a generalizable form of recovery.
The aim of this study was to investigate whether proportional recovery occurs in a more clinically relevant sample of patients with stroke, including those with intracerebral hemorrhage and previous stroke. We hypothesized that proportional recovery of upper-limb impairment reflects a fundamental neurobiological process and would therefore be unaffected by these factors. Such a finding could have important implications for the design of stroke rehabilitation trials.
Materials and Methods
First-ever or recurrent ischemic stroke and intracerebral hemorrhage patients with new upper-limb weakness and aged ≥18 years were recruited. Patients with cerebellar stroke, contraindications to transcranial magnetic stimulation and magnetic resonance imaging, or cognitive or communication impairment precluding informed consent, were excluded. Written informed consent was obtained from all patients. The parent observational study tested the predicting recovery potential algorithm (PREP)9,10 for predicting recovery of upper-limb function, measured with the Action Research Arm Test. This report focuses on the recovery of upper-limb motor impairment, measured with the FM scale.
The FM upper extremity scale was administered 3 days and 3 months poststroke by assessors who were not involved in treating patients. Upper-limb therapy dose was recorded in minutes by treating therapists during inpatient rehabilitation.
Transcranial magnetic stimulation was used to test the functional integrity of the ipsilesional corticomotor pathway within 7 days poststroke, as previously described.5,10 Motor evoked potentials (MEPs) were recorded from the paretic wrist extensors with the limb at rest, and the patient was MEP positive (MEP+) if MEPs of any amplitude were observed at a consistent latency on at least 50% of trials. For patients without MEPs (MEP−), T1-weighted magnetic resonance imaging was used to evaluate the lesion load on the CST 12 days poststroke. A stroke lesion mask was hand-drawn on each patient’s T1-weighted image. A template sensory and motor CST was constructed using probabilistic fiber tracking in the contralesional hemisphere of 85 patients. Preprocessing of diffusion-weighted images was performed with motion and eddy current correction, skull stripping, estimation and fitting of diffusion parameters, and modeling of crossing fibers.11 Seed masks were placed at the pyramid and medial lemniscus near the inferior border of the pons and the primary motor and somatosensory cortices. Tractography was conducted with a curvature threshold of 0.2 and step-length of 0.5. Tracts were then nonlinearly transformed to standard (Montreal Neurological Institute [MNI]) space and mirrored along the midsagittal axis as required before being combined into a template. The template tract was thresholded at 75% probability to ensure that only fibers at the core of the CST were used for calculation of lesion load. The template tract was then nonlinearly transformed into each patients’ T1 space, and the percentage of tract voxels that overlapped the stroke lesion was calculated.6,12,13
Initial impairment (FMii) was calculated as maximum FM score (66) minus baseline FM score. Recovery from motor impairment (ΔFM) was calculated as the difference between baseline and 3-month FM scores. Participants were grouped for analysis into MEP+ or MEP−,5 and bivariate regressions used to examine relationships between ΔFM and FMii, age, sex, previous stroke, stroke type (ischemic, hemorrhagic), thrombolysis, comorbidities (Charlson Comorbidity Index), hand affected (dominant, nondominant), and upper-limb therapy dose (hours). Lesion load (%) was included for MEP− patients. Categorical variables with <5 patients per category were excluded from analysis. All variables with P<0.1 were entered into multivariable linear regression with ΔFM as the dependent variable, with the intercept set to the origin. Recovery from motor impairment was also evaluated in the subset of 45 participants who had experienced a hemorrhagic stroke and/or a previous stroke (either ischemic or hemorraghic), using the same approach and including a constant in the multivariable linear regression model.
One hundred ninety-two patients (86 [45%] women; median [range], age 74 [18–98] years) were recruited, of whom 157 completed follow-up assessment 3 months poststroke (Table 1).
For patients with functional CSTs (MEP+), initial impairment (FMii), age, hand affected, and therapy dose were entered into multivariable linear regression. For these MEP+ patients, only initial impairment predicted ΔFM (Table 2). The proportion of initial impairment that recovered was 63% (95% confidence interval 55%–70%; Figure [A]).
For patients without functional CSTs (MEP−), FMii, age, sex, hand affected, lesion load, previous stroke, and therapy dose were entered into multivariable linear regression. In these MEP− patients, recovery (ΔFM) was not proportional to initial impairment (FMii) but was greater in patients with lower lesion load (Figure [B]).
For the subset of 45 patients with hemorrhagic and/or previous stroke, FMii, MEP status (MEP+, MEP−), hand affected (dominant, nondominant), and therapy dose were entered into multivariable linear regression. Recovery (ΔFM) was proportional to initial impairment (FMii) in patients who were MEP+ (Table 2; Figure [C]).
This study confirms and extends previous work by showing that recovery from motor impairment is unaffected by type of stroke (ischemic, intracerebral hemorrhage), previous stroke, age, sex, comorbidities, or dominance of the affected hand, and is proportional to initial impairment in patients with a functional CST.
For MEP+ patients, we confirmed that recovery is proportional to initial impairment.5 Previous work has found that magnetic resonance imaging measures of the CST are not related to proportional recovery in MEP+ patients.5 However, this could not be confirmed by the present study as magnetic resonance imaging data were not obtained from MEP+ patients. For MEP− patients, we confirmed that recovery is not proportional to initial impairment and is affected by the extent of stroke damage to white matter pathways.5,6 We also report for the first time that recovery is proportional to initial impairment provided the CST is functional (MEP+) in patients with hemorrhagic and/or previous stroke.
For MEP+ patients, the proportion of initial impairment that recovered was 63% (95% confidence interval 55%–70%). This is lower than previously reported,2,4,5,7 perhaps because the final assessment occurred at 3 rather than 6 months.5 The precise proportion of recovery is less important than the existence of the proportional relationship, which may reflect a ubiquitous recovery mechanism. Proportional recovery was confirmed in a more clinically relevant sample, including patients with hemorrhagic and/or previous stroke. We found that these patients also experienced proportional recovery from impairment, provided they were MEP+.
The present and recent findings indicate that recovery from motor impairment reflects a neurobiological process involving the CST.5–7 Provided the CST was functional (MEP+), even patients with severe initial impairment or previous stroke experienced proportional recovery. The type of stroke (ischemic or intracerebral hemorrhage) had no influence on recovery, proportional or otherwise, illustrating independence between the vascular events that cause stroke and the mechanisms of neural recovery. Proportional recovery in MEP+ patients was unaffected by therapy dose or dominance of the paretic hand, both of which influence the amount of poststroke hand use. This may mean that greater use of the paretic hand does not necessarily increase the proportion of recovery.
There are 3 main implications for trial design. First, CST integrity is a relevant biomarker that predicts proportional versus poor recovery from upper-limb impairment. Treatment and control groups need to be matched on a CST biomarker, such as MEP status, not just initial impairment. Biomarkers for recovery potential in other domains should be sought. Second, interventions that increase spontaneous recovery beyond 70% are urgently needed for patients with a functional CST14; interventions that increase recovery potential should be sought for those without. No single interventional approach is likely to achieve both. Finally, the present findings indicate that patients with previous stroke and intracerebral hemorrhage can be included in rehabilitation trials targeting recovery from impairment, increasing potential recruitment.
Sources of Funding
This study was supported by the Health Research Council New Zealand (11/270).
- Received October 3, 2016.
- Revision received December 5, 2016.
- Accepted December 16, 2016.
- © 2017 American Heart Association, Inc.
- Prabhakaran S,
- Zarahn E,
- Riley C,
- Speizer A,
- Chong JY,
- Lazar RM,
- et al
- Zarahn E,
- Alon L,
- Ryan SL,
- Lazar RM,
- Vry MS,
- Weiller C,
- et al
- Winters C,
- van Wegen EE,
- Daffertshofer A,
- Kwakkel G
- Buch ER,
- Rizk S,
- Nicolo P,
- Cohen LG,
- Schnider A,
- Guggisberg AG
- Lazar RM,
- Minzer B,
- Antoniello D,
- Festa JR,
- Krakauer JW,
- Marshall RS
- Stinear CM,
- Barber PA,
- Petoe M,
- Anwar S,
- Byblow WD
- Zhu LL,
- Lindenberg R,
- Alexander MP,
- Schlaug G