SEARCH

SEARCH BY CITATION

Keywords:

  • fMRI;
  • MRI;
  • prognosis;
  • rehabilitation;
  • stroke

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical assessment
  5. Structural imaging
  6. Functional imaging
  7. Combined measures
  8. Conclusion
  9. References

Neuroimaging plays an important role in acute stroke diagnosis and management, but it is not routinely used in rehabilitation settings. Incorporating imaging information in rehabilitation planning may eventually translate to better outcomes after stroke. Here we review the prediction of outcomes after stroke using magnetic resonance imaging. There are clear and specific relationships between the anatomy of the stroke lesion and impairments at the time of scanning, and at later time points in recovery. However, most studies demonstrate these relationships in groups of patients at the chronic stage. In order to be useful for rehabilitation, neuroimaging needs to provide prognostic information for individual patients at a much earlier stage. Recent studies have used diffusion tensor imaging and functional neuroimaging to address this, with promising results. Combining neuroimaging with clinical and neurophysiological assessments may also be useful. Future work in this area may support the tailoring of rehabilitation for individual patients based on their capacity for neural reorganization and recovery.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical assessment
  5. Structural imaging
  6. Functional imaging
  7. Combined measures
  8. Conclusion
  9. References

Computed tomography (CT) and magnetic resonance imaging (MRI) techniques play an important role in acute stroke diagnosis and management [1-3]. MRI is particularly useful for identifying patients who may benefit from thrombolysis, based on the presence of potentially salvageable ischemic tissue [4, 5]. The usefulness of neuroimaging in stroke rehabilitation is less clear. At present, rehabilitation clinicians do not routinely review patients’ imaging when planning treatment. However, this may change as the relationships between MRI measures and the likely response to rehabilitation become better understood. Imaging may help clinicians to identify each patient's potential for recovery, set realistic rehabilitation goals, and select therapy techniques on the basis of residual connections between key elements of the central nervous system. This article reviews recent work using structural and functional MRI (fMRI) techniques to predict and chart the course of recovery after stroke.

Clinical assessment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical assessment
  5. Structural imaging
  6. Functional imaging
  7. Combined measures
  8. Conclusion
  9. References

Age and early neurological status (often measured with the National Institutes of Health Stroke Scale, NIHSS) are reasonable predictors of functional independence six-months after stroke [6]. Some studies have also reported that early measures of impairment, such as the Fugl-Meyer [7] or the presence of both finger extension and shoulder abduction [8], can predict subsequent outcomes. These types of measures appear to have good prognostic accuracy for groups of patients, using fairly coarse, dichotomized outcome measures such as the Barthel Index or modified Rankin Scale. However, it is not clear whether they will be useful in predicting the recovery of meaningful functional activities of daily living for individual patients, such as the ability to walk independently, use a keyboard, or have a conversation in a crowded room. Patients often ask these questions, but our ability to answer them accurately is relatively poor.

In general, patients with less initial impairment have better functional outcomes, but the prognosis for patients with more severe motor impairment is far more difficult to predict [7]. The story is similar for those with moderate to severe aphasia [9], suggesting that important factors for predicting functional outcomes are still missing from current models. A recent meta-analysis found that initial upper limb impairment and function, and the integrity of ascending and descending white matter pathways measured with neurophysiological and neuroimaging techniques, were the strongest predictors of subsequent recovery of upper limb function [10]. This supports growing interest in the idea that neuroimaging data can add to predictive models, so they provide useful information for individual patients.

Structural imaging

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical assessment
  5. Structural imaging
  6. Functional imaging
  7. Combined measures
  8. Conclusion
  9. References

High resolution images produced by MRI allow stroke lesions to be identified with relative ease. Relating the size of the lesion to the impairments produced by stroke was the focus of many early studies [see [11] for review]. However, small lesions of the subcortical white matter and brain stem can produce a disproportionate degree of impairment, so lesion location also needs to be considered. Impairment and function at the time of scanning are related to the extent of damage to key locations and connections in the brain (Table 1). In the motor system, the degree of overlap between the stroke lesion and the corticospinal tract (CST) appears to be a particularly important factor limiting upper limb recovery [12-16], though the impact on gait is less clear [17-19]. However, early studies found that the predictive power of clinical scores was not improved by adding MRI information [20-22], and consequently interest in this approach waned.

Table 1. Magnetic resonance imaging predictors of recovery after stroke
T1-weighted anatomical imageUseful for identifying lesion location. Extent of lesion overlap with posterior limb of the internal capsule related to motor impairment at the time of scanning.
Diffusion tensor imagingUseful for assessing structural integrity of white matter pathways. Greater damage is related to greater impairment at the time of scanning and at later stages of recovery.
Functional magnetic resonance imagingPattern of cortical activity during voluntary movement of the paretic limb is related to impairment at the time of scanning, and may predict the reduction of impairment at later stages of recovery.

There has been something of a revival with the advent of diffusion-weighted imaging, which is sensitive to the diffusion of water molecules within tissue. In particular, diffusion tensor imaging (DTI) tractography has been used to assess the structural integrity of white matter pathways in the brain. Several studies have demonstrated that greater damage to the CST is associated with more impairment in chronic stroke patients [see [23] for review]. In addition, integrity of CST fibers originating from nonprimary motor areas (i.e. premotor cortices) can also be related to motor impairment [24].

These measures may also be used to predict outcome. DTI parameters of CST integrity acquired within three-weeks of subcortical stroke correlate with both initial impairment and upper limb outcome at six-months [25]. The predictive power of DTI within 12 h of symptom onset has been explored by Puig and colleagues in a study of 60 patients [26]. Damage to the CST at the posterior limb of the internal capsule (PLIC) correlated well with motor impairment using the upper and lower limb components of the NIHSS at 30 and 90 days. Furthermore, the sensitivity and specificity of this measure were superior to those for damage to the corona radiata or cortex where the white matter fibers of the CST are more widely spread out. Importantly, these measures were also superior to lesion volume and baseline clinical scores. Prediction of functional outcomes does not need to be made quite so early after stroke for the purposes of planning rehabilitation, but these data once again highlight the impact of damage to the PLIC as a major determinant of motor deficit after stroke. Assessment of nonmotor pathways is also possible; for example, arcuate fasciculus damage measured with DTI, but not seen on conventional MRI, is associated with severity of aphasia [27]. While there are clear relationships between the extent of stroke damage to key brain structures and the resulting impairments, most studies in this area have examined for correlations between imaging parameters and relatively simple outcome scores. In general, these are not useful for goal setting in rehabilitation, nor necessarily meaningful for patients or their carers in terms of understanding their own recovery. Furthermore, while these studies confirm our general clinical intuition, they do not allow prediction of outcomes at the level of the individual patient. They do confirm, however, that there is clinically relevant information to be gained by careful quantification of the anatomy of the damage.

The PLORAS (Predicting Language Outcome and Recovery After Stroke) system is based on this principle: that the ability to comprehend and produce speech after stroke depends on whether the areas of the brain that support language have been damaged [28]. This system requires a large number of patients with the following: (1) a detailed assessment of various language capabilities, (2) brain scans which are converted into three dimensional descriptions of the lesion, and (3) time since stroke. A new subject's lesion image is compared with those from all the other patients already in the database to find one with a similar lesion. The language scores for all the similar patients are plotted over time, enabling the time course of recovery for the new patient to be estimated. This can be repeated for a number of different types of language skills (expression and comprehension, for example) but will also depend on patient demographics, comorbidities, and degree and type of therapeutic intervention. Using this type of system, it should be possible to tell a patient what the chances are of recovering a specific functional ability by six-months or even a year after stroke. This will be useful for both patients and carers as well as treating physicians by facilitating realistic goal setting. Improvements in imaging analysis technology might allow standard clinical scans (both MRI and CT) to be converted into accurate lesion maps, paving the way for automated prediction software.

One of the limitations of structural neuroimaging is that it provides no information about whether or how surviving tissues are working. The functional connectivity between cortical and subcortical components of neural networks may provide important information about the capacity for reorganization and recovery. Therefore a growing number of studies have explored whether functional imaging data might be useful in predicting recovery and planning rehabilitation.

Functional imaging

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical assessment
  5. Structural imaging
  6. Functional imaging
  7. Combined measures
  8. Conclusion
  9. References

Functional MRI measures the blood oxygen level dependent (BOLD) signal from the brain during task performance and at rest, and this signal is thought to reflect neural activity. BOLD activation patterns after stroke are closely correlated with aspects of impairment, in both the motor and language domains [29, 30], but these correlation studies do not allow true prediction, which requires a different methodological approach. Saur et al. [31] performed fMRI in 21 patients with at least moderate language comprehension impairment. Scanning was performed in the second week after stroke and language function was assessed at six-months. A multivariate machine learning technique was used to calculate the characteristics of scans in those with good or bad outcome. Using fMRI data alone, 76% of patients were correctly assigned to the good or poor recovery group. Adding age and baseline language score increased this to 86%. This is promising, though the binarized outcome measure was a composite of eight different test scores, so the recovery of specific elements of language function was not predictable.

A similar approach has been used in the motor domain. Functional MRI data acquired in the first few days after stroke was used to predict a subsequent change in motor impairment [32, 33]. A particular pattern of brain activation was highly predictive of the change in Fugl-Meyer score over the next three-months, a finding that was independent of initial stroke severity and lesion volume. Although the multivariate analysis used did not allow anatomical inference to be made, it is clear that there is something about the way the function of the brain responds to injury, over and above the anatomy of the damage, that holds clues about future clinical progression. The pattern was distributed and certainly not confined to the motor system, even though clinical improvement was measured in the motor domain. The idea that motor improvement may not be solely related to the structural integrity of the CST, but also with other characteristics of the motor cortex, is supported by the finding that upper limb function at three-months correlates only weakly with CST integrity, but strongly with intracortical excitability within primary motor cortex (M1), measured with transcranial magnetic stimulation (TMS) [34].

There is still a long way to go before these studies influence how best to treat the impairment suffered by patients after stroke. The next question is whether imaging and/or neurophysiological data can contribute to predictive models, not of outcome, but of the potential for therapy-driven improvements in function. For example, a recent study demonstrated that the beneficial effects of facilitatory repetitive TMS over ipsilesional M1 on motor function of the affected hand were seen in patients with subcortical stroke, but not in those with extension of the infarct into ipsilesional sensorimotor cortex [35]. Furthermore, task-related activity in ipsilesional M1 measured with fMRI at baseline correlated with improvement of motor performance induced by repetitive TMS. While this seems an obvious result, this kind of stratification based on residual functional and structural anatomy is rarely considered, although it clearly has the potential to improve trial design [36].

Combined measures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical assessment
  5. Structural imaging
  6. Functional imaging
  7. Combined measures
  8. Conclusion
  9. References

The idea of combining neuroimaging data with clinical and/or neurophysiological measures to predict the response to therapy has also received some attention. Cramer et al. [37] assessed 13 baseline clinical and radiological measures in 24 chronic stroke patients, and determined whether each was able to predict subsequent gains made during six-weeks of rehabilitation therapy. Only two baseline measures were significant and independent predictors of clinical improvement. The first was a lower level of impairment, and the second was less ipsilesional motor cortex activation as measured with fMRI. In a subsequent analysis of the same data set, integrity of corticospinal fibers from M1 was also found to correlate with gains made during treatment [38]. These studies suggest that there is something of biological interest in the imaging data that is independent of baseline clinical impairment, which predicts clinical improvement. Lower baseline motor cortex activation was also associated with larger increases in motor cortex activation after treatment, and so it was suggested that low baseline cortical activity represents underuse of surviving cortical resources. When used carefully, it appears that measures of brain function as well as structure may be useful for predicting which patients are more likely to benefit from a restorative intervention at the chronic stage, although further studies involving larger cohorts of patients will be required before the correct biomarkers of treatment response are identified and can be used in clinical practice.

Stinear and colleagues [39] also set out to determine whether the state of the motor system at the chronic stage of stroke could predict an individual's capacity for further functional improvement during a four-week motor practice program. A variety of tools were used, including TMS, DTI, and fMRI. The presence or absence of motor evoked potentials (MEPs) to TMS in the affected upper limb, and fractional anisotropy values calculated from the posterior limbs of the internal capsules, were both used to assess the integrity of the descending white matter pathways. Not surprisingly, in patients with MEPs, meaningful gains with motor practice were still possible three-years after stroke. The prognosis for patients without MEPs has always been more difficult to predict in the clinical setting, but the absence of MEPs is often taken as a sign of poor prognosis [40]. Here, the functional potential in patients without MEPs was predicted by PLIC disruption as assessed with fractional anisotropy from DTI. Specifically, these authors found that past a certain threshold of PLIC disruption, little therapy-induced functional improvement was possible. Conversely, functional improvement was possible in some patients without MEPs whose DTI data indicated that PLIC disruption had not passed this ‘point of no return’. Interestingly, the patients also performed a simple motor task during fMRI, but the results as assessed by the degree of lateralization of motor cortex activity to one hemisphere or the other did not contribute to the predictive model. Nevertheless, this kind of study illustrates how multimodal imaging and neurophysiological data could be used to assess the state of the motor system and predict the potential for therapy-driven functional improvements.

One of the limitations of studies that have combined different types of neuroimaging modalities, and neuroimaging with neurophysiology measures, is that most have been conducted with patients at the chronic stage of stroke recovery. For imaging to play a role in rehabilitation, it needs to provide useful prognostic information at the acute and sub-acute stages, when therapy is being planned. One recent study combined DTI and TMS measures of 58 patients made within four-weeks of stroke and found that TMS had higher positive predictive value than DTI, while DTI had higher negative predictive value, for upper limb function six-months later [41]. Clearly, it will not be efficient to gather neurophysiological and neuroimaging data on all patients. An algorithm has been proposed for sequentially combining clinical, neurophysiological, and neuroimaging measures to predict the potential for recovery of upper limb function [42]. Called the PREP (Predicting REcovery Potential) algorithm, it has recently been tested in a sample of 40 sub-acute stroke patients, and seems to provide a potentially useful starting point for tailoring rehabilitation to individual patients, based on their capacity for recovery [43]. It may also support accurate stratification of patients in clinical trials of rehabilitation techniques. Further work is needed to develop similar approaches for other motor functions, such as gait, and for other functional domains such as communication.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical assessment
  5. Structural imaging
  6. Functional imaging
  7. Combined measures
  8. Conclusion
  9. References

Neuroimaging plays a vital role in the acute diagnosis and medical management of stroke. There are clear relationships between stroke lesion characteristics and impairments at the time of scanning, and with broadly defined clinical outcomes at the chronic stage. Neuroimaging may also come to play an important role in rehabilitation after stroke. Findings to date suggest that the anatomy of the damage may set a limit on the extent of recovery, but that other parameters, perhaps preserved cortico-cortical and cortico-spinal connectivity, might be important when considering whether a patient has the capacity or potential to improve. Whether a patient realizes their full potential for improvement will also depend, at least partly, on therapy dose. Future studies will need to consider the type and dose of therapy when exploring the relationships between acute neuroimaging measures and subsequent recovery of function. Neuroimaging holds great promise for improving rehabilitation outcomes, by providing useful prognostic information that allows therapy to be tailored to individuals according to their capacity for neural reorganization and recovery.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical assessment
  5. Structural imaging
  6. Functional imaging
  7. Combined measures
  8. Conclusion
  9. References
  • 1
    Burgess RE, Kidwell CS. Use of MRI in the assessment of patients with stroke. Curr Neurol Neurosci Rep 2011; 11:2834.
  • 2
    Kidwell CS, Wintermark M. Imaging of intracranial haemorrhage. Lancet Neurol 2008; 7:256267.
  • 3
    Farr TD, Wegener S. Use of magnetic resonance imaging to predict outcome after stroke: a review of experimental and clinical evidence. J Cereb Blood Flow Metab 2010; 30:703717.
  • 4
    Olivot JM, Albers GW. Diffusion-perfusion MRI for triaging transient ischemic attack and acute cerebrovascular syndromes. Curr Opin Neurol 2011; 24:4449.
  • 5
    Merino JG, Warach S. Imaging of acute stroke. Nat Rev Neurol 2010; 6:560571.
  • 6
    Veerbeek JM, Kwakkel G, van Wegen EE, Ket JC, Heymans MW. Early prediction of outcome of activities of daily living after stroke: a systematic review. Stroke 2011; 42:14821488.
  • 7
    Prabhakaran S, Zarahn E, Riley C et al. Inter-individual variability in the capacity for motor recovery after ischemic stroke. Neurorehabil Neural Repair 2008; 22:6471.
  • 8
    Nijland RH, van Wegen EE, Harmeling-van der Wel BC, Kwakkel G. Presence of finger extension and shoulder abduction within 72 hours after stroke predicts functional recovery: early prediction of functional outcome after stroke: the EPOS cohort study. Stroke 2010; 41:745750.
  • 9
    Lazar RM, Speizer AE, Festa JR, Krakauer JW, Marshall RS. Variability in language recovery after first-time stroke. J Neurol Neurosurg Psychiatry 2008; 79:530534.
  • 10
    Coupar F, Pollock A, Rowe P, Weird C, Langhorne P. Predictors of upper limb recovery after stroke: a systematic review and meta-analysis. Clin Rehabil 2012; 26:291313.
  • 11
    Schiemanck SK, Kwakkel G, Post MW, Prevo AJ. Predictive value of ischemic lesion volume assessed with magnetic resonance imaging for neurological deficits and functional outcome poststroke: a critical review of the literature. Neurorehabil Neural Repair 2006; 20:492502.
  • 12
    Lindenberg R, Renga V, Zhu LL, Betzler F, Alsop D, Schlaug G. Structural integrity of corticospinal motor fibers predicts motor impairment in chronic stroke. Neurology 2010; 74:280287.
  • 13
    Schaechter JD, Fricker ZP, Perdue KL et al. Microstructural status of ipsilesional and contralesional corticospinal tract correlates with motor skill in chronic stroke patients. Hum Brain Mapp 2009; 30:34613474.
  • 14
    Shelton FN, Reding MJ. Effect of lesion location on upper limb motor recovery after stroke. Stroke 2001; 32:107112.
  • 15
    Wenzelburger R, Kopper F, Frenzel A et al. Hand coordination following capsular stroke. Brain 2005; 128(Pt 1):6474.
  • 16
    Schiemanck SK, Kwakkel G, Post MW, Kappelle LJ, Prevo AJ. Impact of internal capsule lesions on outcome of motor hand function at one year post-stroke. J Rehabil Med 2008; 40:96101.
  • 17
    Dawes H, Enzinger C, Johansen-Berg H et al. Walking performance and its recovery in chronic stroke in relation to extent of lesion overlap with the descending motor tract. Exp Brain Res 2008; 186:325333.
  • 18
    Tang PF, Ko YH, Luo ZA, Yeh FC, Chen SH, Tseng WY. Tract-specific and region of interest analysis of corticospinal tract integrity in subcortical ischemic stroke: reliability and correlation with motor function of affected lower extremity. AJNR Am J Neuroradiol 2010; 31:10231030.
  • 19
    Lee JS, Han MK, Kim SH, Kwon OK, Kim JH. Fiber tracking by diffusion tensor imaging in corticospinal tract stroke: topographical correlation with clinical symptoms. Neuroimage 2005; 26:771776.
  • 20
    Schiemanck SK, Kwakkel G, Post MW, Kappelle LJ, Prevo AJ. Predicting long-term independency in activities of daily living after middle cerebral artery stroke: does information from MRI have added predictive value compared with clinical information? Stroke 2006; 37:10501054.
  • 21
    Feys H, Hetebrij J, Wilms G, Dom R, De Weerdt W. Predicting arm recovery following stroke: value of site of lesion. Acta Neurol Scand 2000; 102:371377.
  • 22
    Hand PJ, Wardlaw JM, Rivers CS et al. MR diffusion-weighted imaging and outcome prediction after ischemic stroke. Neurology 2006; 66:11591163.
  • 23
    Jang SH. Prediction of motor outcome for hemiparetic stroke patients using diffusion tensor imaging: a review. NeuroRehabilitation 2010; 27:367372.
  • 24
    Schulz R, Park CH, Boudrias MH, Gerloff C, Hummel FC, Ward NS. Assessing the integrity of corticospinal pathways from primary and secondary cortical motor areas after stroke. Stroke 2012; 43:22482251.
  • 25
    Radlinska B, Ghinani S, Leppert IR, Minuk J, Pike GB, Thiel A. Diffusion tensor imaging, permanent pyramidal tract damage, and outcome in subcortical stroke. Neurology 2010; 75:10481054.
  • 26
    Puig J, Pedraza S, Blasco G et al. Acute damage to the posterior limb of the internal capsule on diffusion tensor tractography as an early imaging predictor of motor outcome after stroke. AJNR Am J Neuroradiol 2011; 32:857863.
  • 27
    Kim SH, Lee DG, You H et al. The clinical application of the arcuate fasciculus for stroke patients with aphasia: a diffusion tensor tractography study. NeuroRehabilitation 2011; 29:305310.
  • 28
    Price CJ, Seghier ML, Leff AP. Predicting language outcome and recovery after stroke: the PLORAS system. Nature reviews. Neurology 2010; 6:202210.
  • 29
    Ward NS, Brown MM, Thompson AJ, Frackowiak RS. Neural correlates of outcome after stroke: a cross-sectional fMRI study. Brain 2003; 126(Pt 6):14301448.
  • 30
    Saur D, Hartwigsen G. Neurobiology of language recovery after stroke: lessons from neuroimaging studies. Arch Phys Med Rehabil 2012; 93(1 Suppl.):S1525.
  • 31
    Saur D, Ronneberger O, Kummerer D, Mader I, Weiller C, Kloppel S. Early functional magnetic resonance imaging activations predict language outcome after stroke. Brain 2010; 133(Pt 4):12521264.
  • 32
    Marshall RS, Zarahn E, Alon L, Minzer B, Lazar RM, Krakauer JW. Early imaging correlates of subsequent motor recovery after stroke. Ann Neurol 2009; 65:596602.
  • 33
    Zarahn E, Alon L, Ryan SL et al. Prediction of motor recovery using initial impairment and fMRI 48 h poststroke. Cereb Cortex 2011; 21:27122721.
  • 34
    Swayne OB, Rothwell JC, Ward NS, Greenwood RJ. Stages of motor output reorganization after hemispheric stroke suggested by longitudinal studies of cortical physiology. Cereb Cortex 2008; 18:19091922.
  • 35
    Ameli M, Grefkes C, Kemper F et al. Differential effects of high-frequency repetitive transcranial magnetic stimulation over ipsilesional primary motor cortex in cortical and subcortical middle cerebral artery stroke. Ann Neurol 2009; 66:298309.
  • 36
    Ward NS. Getting lost in translation. Curr Opin Neurol 2008; 21:625627.
  • 37
    Cramer SC, Parrish TB, Levy RM et al. Predicting functional gains in a stroke trial. Stroke 2007; 38:21082114.
  • 38
    Riley JD, Le V, Der-Yeghiaian L et al. Anatomy of stroke injury predicts gains from therapy. Stroke 2011; 42:421426.
  • 39
    Stinear CM, Barber PA, Smale PR, Coxon JP, Fleming MK, Byblow WD. Functional potential in chronic stroke patients depends on corticospinal tract integrity. Brain 2007; 130(Pt 1):080.
  • 40
    Heald A, Bates D, Cartlidge NE, French JM, Miller S. Longitudinal study of central motor conduction time following stroke. 2. Central motor conduction measured within 72 h after stroke as a predictor of functional outcome at 12 months. Brain 1993; 116(Pt 6):13711385.
  • 41
    Kwon YH, Son SM, Lee J, Bai DS, Jang SH. Combined study of transcranial magnetic stimulation and diffusion tensor tractography for prediction of motor outcome in patients with corona radiate infarct. J Rehabil Med 2011; 43:430434.
  • 42
    Stinear CM. Prediction of recovery of motor function after stroke. Lancet Neurol 2010; 9:12281232.
  • 43
    Stinear CM, Barber PA, Petoe M, Anwar S, Byblow WD. The PREP algorithm predicts potential for upper limb recovery after stroke. Brain 2012; 135:25272535.