Association Between Brain-Derived Neurotrophic Factor Genotype and Upper Extremity Motor Outcome After Stroke
Background and Purpose—The identification of intrinsic factors for predicting upper extremity motor outcome could aid the design of individualized treatment plans in stroke rehabilitation. The aim of this study was to identify prognostic factors, including intrinsic genetic factors, for upper extremity motor outcome in patients with subacute stroke.
Methods—A total of 97 patients with subacute stroke were enrolled. Upper limb motor impairment was scored according to the upper limb of Fugl-Meyer assessment score at 3 months after stroke. The prediction of upper extremity motor outcome at 3 months was modeled using various factors that could potentially influence this impairment, including patient characteristics, baseline upper extremity motor impairment, functional and structural integrity of the corticospinal tract, and brain-derived neurotrophic factor genotype. Multivariate ordinal logistic regression models were used to identify the significance of each factor.
Results—The independent predictors of motor outcome at 3 months were baseline upper extremity motor impairment, age, stroke type, and corticospinal tract functional integrity in all stroke patients. However, in the group with severe motor impairment at baseline (upper limb score of Fugl-Meyer assessment <25), the number of Met alleles in the brain-derived neurotrophic factor genotype was also an independent predictor of upper extremity motor outcome 3 months after stroke.
Conclusions—Brain-derived neurotrophic factor genotype may be a potentially useful predictor of upper extremity motor outcome in patients with subacute stroke with severe baseline motor involvement.
Stroke is a leading cause of acquired disability in adults in Korea, as well as worldwide.1,2 Motor impairment of upper extremities is a significant disability that affects daily life and increases the burden on stroke patients and their families.3 Rehabilitation therapy is currently the best practice in stroke management to maximize functional ability after acute stroke.4 The first step of stroke rehabilitation is to define realistic and attainable goals for improvement. Therefore, accurate prediction of upper extremity motor outcome is important for realistic goal setting in stroke rehabilitation.
Ipsilesional corticospinal tract (CST) integrity has been shown to be an important predictor of upper extremity motor outcome in patients with subacute stroke.5–7 CST integrity can be assessed by clinical findings,8 including electrophysiologic methods, such as transcranial magnetic stimulation–induced motor evoked potential (MEP),9 neuroimaging with diffusion tensor imaging,10 and diffusion tensor tractography (DTT).11 In accordance with the remarkable developments that have been made in functional neuroimaging, the functional and structural integrity of the CST can be well determined by assessing whether transcranial magnetic stimulation–induced MEP and DTT are present or absent, respectively.12 Recently, CST integrity assessments were recommended to predict recovery of upper extremity motor function in patients with subacute stroke.5,10,12,13 In addition, the contribution of neuroplasticity should also be considered, even though the degree of ipsilateral CST integrity reflects the capacity for recovery of upper extremity motor function.5 Neuroplasticity is well known as a mechanism for rehabilitation interventions to enhance motor recovery in stroke patients.14 Brain-derived neurotrophic factor (BDNF) plays a crucial role in the formation of new connections between neurons and in altering the strength of existing connections between neurons.15 A single-nucleotide polymorphism has been identified in the human BDNF gene at codon 66 (Val66Met). Specifically, the replacement of Val with Met has been reported to disrupt cellular processes, such as trafficking and activity-dependent secretion of BDNF.16 BDNF genotype has been associated with functional outcomes in patients with aneurysmal subarachnoid hemorrhage17 and ischemic stroke.18 However, few studies have investigated the influence of BDNF genotype on motor outcome in stroke patients.
In this study, we aimed to investigate the potential of BDNF genotype to predict upper extremity motor outcome in patients with subacute stroke. Our motivation was to identify intrinsic factors for predicting motor outcome because such factors could aid the design of individualized treatment plans in stroke rehabilitation.
One hundred six patients with subacute stroke (mean age, 58.3±12.7 years; 46 women) were enrolled. All patients provided written informed consent, and all procedures were performed in accordance with the Declaration of Helsinki. The study protocol was also approved by the Samsung Medical Center Institutional Review Board. Patient inclusion criteria were hemiparesis caused by first-ever stroke (ischemic or hemorrhagic), >18 years of age, and <4 weeks since stroke onset. Patient exclusion criteria were accompanying neurologic or psychiatric disorder, or any terminal illness.
This was a prospective, observational study to identify factors that influence upper extremity motor outcome after stroke. After baseline assessment (T0), upper extremity motor impairment was assessed at 3 months (day 90; T1) after the standardized rehabilitation program. The standardized inpatient rehabilitation program consisted of 2 hours of physical therapy and 1 hour of occupational therapy daily on workdays. After participating in the standardized inpatient rehabilitation program for 2 weeks, each patient participated in the standardized inpatient rehabilitation program, the standardized outpatient rehabilitation program (1 hour of physical therapy and 30 minutes of occupational therapy 3 days per week), or self-exercise at home until T1.
Motor outcome was assessed using the Fugl-Meyer assessment (FMA) scoring system.19 The upper limb FMA (FMA-UL) score ranges from 0 to 66, with higher scores indicating fewer motor deficits. FMA-UL score at T1 was grouped into the following 4 categories: normal (FMA-UL, 66), slight motor impairment (FMA-UL, 41–65), moderate motor impairment (FMA-UL, 25–40), and severe motor impairment (FMA-UL, 0–24), as previously described.20 In addition, we assessed the lower limb FMA score (range, 0–34) and the total FMA score (range, 0–100) at baseline.
All assessments were performed by investigators who were blinded to patient status of potentially influential factors except sex and side of lesion.
Potentially Influential Factors
For the baseline descriptive characteristics, age at time of stroke and duration from onset to baseline were entered as scalar predictors. Sex was coded as a binary predictor (male, 1; female, 0). Type of stroke, side of lesion, and location of stroke lesion were coded as binary predictors (ischemic, 1; hemorrhagic, 0; right hemisphere, 1; left hemisphere, 0; and supratentorial, 1; infratentorial, 0). Baseline upper extremity motor impairment (FMA-UL score) was entered as a scalar predictor. Because the quality of stroke rehabilitation differed in each patient after the standardized inpatient rehabilitation program for 2 weeks, the intensity of stroke rehabilitation to T1 was entered as a scalar predictor (standardized inpatient rehabilitation program, 2; standardized outpatient rehabilitation program, 1; and self-exercise at home, 0) to analyze predictors at T1.
In addition, 2 assessments of CST integrity obtained at T0 were also considered as potentially influential factors. First, DTT analysis was used to determine the structural integrity of the affected CST, as previously described.21 Briefly, diffusion tensor images were collected using a 3 Tesla MR scanner. The single-shot echoplanar imaging sequence (number of slices, 60; slice thickness, 2.25 mm; matrix size, 112×112 mm; and in-plane resolution, 1.96×1.96 mm) was used, and 46 directional diffusion-weighted images were obtained. Three-dimensional fiber reconstruction was performed using PRIDE software (Philips Medical Systems, Best, the Netherlands).22 The termination criteria used for fiber tracking were fractional anisotropy <0.2 and an angle change >70°.21 To reconstruct the CST, the 2 regions-of-interest method was used. The regions of interest included the motor cortex and the lower anterior pons. Fibers connected to the cerebellum were excluded. Patients were classified into 2 groups according to the degree of CST integrity in the affected hemisphere: visible CST, which included all patients whose CST was preserved; and nonvisible CST, which included all patients whose CST was not preserved according to DTT analysis. DTT status was entered as a binary predictor (visible CST, 1; nonvisible CST, 0).
Second, transcranial magnetic stimulation of the affected motor cortex was used to measure the MEP in the resting paretic first dorsal interosseous (FDI) muscle, as described previously.23 MEPs were assessed using a single magnetic stimulation at 120% of the rMT intensity over the ipsilesional M1 using a 70-mm figure-of-eight coil. A Synergy electromyography/evoked potentials system (Medelec, Bristol, UK) was used to record and monitor the activity of the contralateral FDI muscle. Single-pulse transcranial magnetic stimulation was applied over the ipsilesional M1 with a Magstim BiStim2 stimulator (Magstim; Spring Gardens, Wales, UK). The coil was held tangentially to the scalp, with the handle pointing backward and laterally at 45° from the midsagittal line. Patients were classified into 2 groups according to the presence of MEP on the affected FDI muscle: MEP response, which included all patients who exhibited a MEP in the affected FDI; and no MEP response, which included patients who did not exhibit any MEP with peak-to-peak amplitude of ≥50 µV in the affected FDI according to 3 successive discharges with maximal stimulator output. MEP status was entered as a binary predictor (MEP response, 1; no MEP response, 0).
Finally, the presence of the BDNF Val66Met polymorphism was also assessed as a potentially influential factor. For BDNF genotyping, whole blood was collected into EDTA tubes. Genomic DNA was isolated from peripheral blood leukocytes according to a standard proteinase-K RNase digestion procedure followed by phenol–chloroform extraction. The BDNF Val66Met polymorphism was genotyped via polymerase chain reaction–restriction fragment length polymorphism.24 Patients were classified according to BDNF genotype into either the Val/Val, Val/Met, or Met group, and BDNF genotype was entered as an ordinal predictor according to the number of Met alleles (Val/Val, 0; Val/Met,1; and Met/Met,2).
SPSS version 23.0 (SPSS, Chicago, IL) was used for all statistical analyses. The potentially influential factors for upper extremity motor outcome were tested with multivariate ordinal logistic regression. The dependent variable was upper extremity motor outcome at T1. Upper extremity motor impairment was categorized and coded as an ordinal parameter (severe, 1; moderate, 2; slight, 3; and normal, 4).20 In addition, the potentially influential factors were separately tested via multivariate ordinal logistic regression by severe motor impairment by baseline subgroup (with or without). P<0.05 was considered statistically significant.
Of the first 103 patients, 3 patients failed to complete study follow-up because of personal reasons, and an additional 3 patients were excluded from analysis because of motion artifacts in the diffusion tensor images. Therefore, 97 patients with a complete set of data at 3-month follow-up (T1) were ultimately analyzed. Detailed patient characteristics are presented in Table 1.
Total Stroke Patients
Figure 1A shows the changes in upper extremity motor impairment in total stroke patients. The results from multivariate ordinal logistic regression analysis are summarized in Table 2. Age and FMA-UL at baseline and MEP response were significant independent predictors for upper extremity motor outcome (P<0.001; Nagelkerke R2=0.701).
Stroke Patients Without Severe Motor Impairment at Baseline
Among the 31 patients with slight or moderate motor impairment, the changes in upper extremity motor impairment are described in Figure 1B. Only FMA-UL at baseline was a significant independent predictor for upper extremity motor outcome (P=0.036; Nagelkerke R2=0.467; Table 2).
Stroke Patients With Severe Motor Impairment at Baseline
Among the 66 patients with severe motor impairment at baseline, Figure 1C shows the changes in upper extremity motor impairment. The number of BDNF Met alleles, age, and FMA-UL at baseline were significant independent predictors for upper extremity motor outcome (P<0.001; Nagelkerke R2=0.585; Table 2).
In additional analysis of stroke patients with severe motor impairment at baseline, FMA-UL at baseline showed no significant differences by BDNF genotype (10.9±7.9, 9.4±7, and 8.5±5.3 in the Val/Val, Val/Met, and Met/Met groups, respectively). There was no significant correlation between FMA-UL at baseline and the number of BDNF Met alleles. Six patients (42.9%) in the Val/Val group showed significant improvement at T1. Thirteen patients (31%) in the Val/Met group demonstrated significant improvement at T1. In addition, only 1 patient (10%) in the Met/Met group demonstrated significant improvement at T1 (Figure 2). There was a significant correlation between FMA-UL at T1 and the number of BDNF Met alleles (P=0.045; Spearman coefficient, −0.248).
We found that the predictors for upper extremity motor outcome at 3 months were baseline upper extremity motor impairment, age, and CST functional integrity at the subacute stroke phase. These results confirmed the importance of CST preservation for the upper extremity motor outcome in stroke patients.5,10,12,13 In addition, BDNF genotype was a significant independent predictor for upper extremity motor outcome at 3 months in patients with subacute stroke with severe motor impairment at baseline. These findings demonstrate the potential of BDNF genotype to influence motor outcome in stroke patients with severe motor impairment at baseline.
The preservation of ipsilateral CST integrity is the most important factor that influences upper extremity motor recovery after stroke.5 The integrity of the ipsilateral CST in the acute stage strongly predicts which patients will experience spontaneous neurobiologic recovery from impairment and which patients will not, regardless of initial motor impairment.13 Two factors significantly associated with upper extremity motor outcome were baseline motor impairment and ipsilateral CST functional integrity in this study. These findings might be related to the mechanism of motor recovery with the preservation of ipsilateral CST integrity. In addition, age was a significant independent predictor for upper extremity motor outcome. The number of BDNF Met alleles was a significant independent predictor for upper extremity motor outcome in stroke patients with severe motor involvement at baseline. Synapse-based learning rules could potentially help to create compensatory circuits after stroke; these learning rules can be divided into 2 broad conceptual classes of mechanisms, homeostatic plasticity, and learning-dependent plasticity.25 Age has been significantly correlated with neural plasticity potential,26 and BDNF has been reported to play a critical role in homeostatic plasticity in animal models.27 The related factors in the present study could reflect the mechanisms of motor recovery during the subacute stroke phase with severe motor involvement.
BDNF is synthesized as a precursor, proBDNF, which is proteolytically cleaved to generate mature BDNF. Mature BDNF plays a major role in neuronal survival, synaptic plasticity, learning, and memory.28 In an animal model of stroke, BDNF has been shown to have beneficial effects on motor recovery and rehabilitation.29,30 The BDNF gene has 1 common single-nucleotide polymorphism that significantly affects the intracellular packaging of proBDNF and its activity-dependent release, without affecting structural BDNF.31 Because the majority of BDNF is released via the regulated secretory pathway from neurons, impaired regulated secretion because of this BDNF polymorphism results in a significant decrease in available BDNF. We found that the number of BDNF Met alleles was a significant independent factor related to upper extremity motor outcome at 3 months in stroke patients with severe motor involvement. These findings suggest that BDNF may play a role in upper extremity motor outcome in stroke patients with severe motor involvement. In addition, the amount of mature BDNF released has the potential to influence motor outcome in stroke patients. However, we obtained no direct evidence to support a relationship between the amount of BDNF release and neural plasticity in stroke patients. Therefore, further studies will be needed to test this hypothesis.
One previous study13 reported no significant relationship between BDNF genotype and upper extremity motor outcome in stroke patients. However, most stroke patients in the previous study13 showed mild-to-moderate motor involvement in the early stroke phase. Specifically, we found that BDNF genotype was a significant independent predictor of upper extremity motor outcome only for stroke patients with severe motor involvement.
DTT analysis was not an independent predictor of upper extremity motor outcome in stroke patients. This was in contrast to a previous report stating that DTT analysis had been shown to be a significant predictor of upper extremity motor outcome in stroke patients.21 We assessed CST integrity using 3 different assessments that exhibited collinearity; thus, DTT analysis was not added to the prediction model as a significant factor. In addition, the extent of stroke rehabilitation after the 2-week standardized inpatient rehabilitation therapy period was not an independent predictor of upper extremity motor outcome in stroke patients. It is conceivable that standardized inpatient rehabilitation is far too short in terms of intervention time for patients with severe motor impairment because most participants (80.8%) who showed severe motor involvement at T1 received intensive inpatient rehabilitation until T1. Therefore, it was difficult to extrapolate the effect of prolonged rehabilitation intervention on motor outcome in the upper limbs. Further studies will be needed to determine whether longer intervention would become an independent predictor.
This study has some limitations. We did not enroll many stroke patients with mild upper motor involvement. In addition, we could not control many other potentially influential factors for motor outcome after stroke, such as medication, noninvasive brain stimulation, and the amount of stroke rehabilitation in each patient. Therefore, a multicenter clinical trial with a larger number of patients will be needed in future to determine the extent to which our conclusions are broadly applicable. In spite of these limitations, we demonstrated that BDNF genotype was associated with upper extremity motor outcome in stroke patients with severe motor involvement. We suggest that BDNF genotype could be considered a predictor of upper extremity motor outcome in patients with subacute stroke with severe motor involvement.
Dr Chang contributed to the design and conceptualization of the study, analysis and interpretation of the data, and drafting the article. Dr Park acquired and analyzed the data. Dr Lee analyzed the data. A. Lee acquired the data, and Dr Kim contributed to the design and conceptualization of the study, analysis and interpretation of the data, and revising the article for important intellectual content.
Sources of Funding
This study was supported by the National Research Foundation of Korea (NRF) grant funded by the government of Korea (Ministry of Science, Information and Communications Technology, and Future Planning [MSIP]; NRF-2017R1A2A1A05000730, NRF-2016R1A2B4012054).
- Received August 26, 2016.
- Revision received March 24, 2017.
- Accepted April 10, 2017.
- © 2017 American Heart Association, Inc.
- Buch ER,
- Rizk S,
- Nicolo P,
- Cohen LG,
- Schnider A,
- Guggisberg AG
- Prescott RJ,
- Garraway WM,
- Akhtar AJ
- Milot MH,
- Spencer SJ,
- Chan V,
- Allington JP,
- Klein J,
- Chou C,
- et al
- Bigourdan A,
- Munsch F,
- Coupe P,
- Guttmann CR,
- Sagnier S,
- Renou P,
- et al
- Siironen J,
- Juvela S,
- Kanarek K,
- Vilkki J,
- Hernesniemi J,
- Lappalainen J
- Lu B
- Schabitz WR,
- Berger C,
- Kollmar R,
- Seitz M,
- Tanay E,
- Kiessling M,
- et al
- Ploughman M,
- Windle V,
- MacLellan CL,
- White N,
- Doré JJ,
- Corbett D
- Yoshida T,
- Ishikawa M,
- Niitsu T,
- Nakazato M,
- Watanabe H,
- Shiraishi T,
- et al