Background The ability to predict outcome after stroke is clinically important for planning treatment and for stratification in restorative clinical trials. In relation to the upper limbs, the main predictor of outcome is initial severity, with patients who present with mild to moderate impairment regaining about 70% of their initial impairment by 3 months post-stroke. However, in those with severe presentations, this proportional recovery applies in only about half, with the other half experiencing poor recovery. The reasons for this failure to recover are not established although the extent of corticospinal tract damage is suggested to be a contributory factor. In this study, we investigated 30 patients with chronic stroke who had presented with severe upper limb impairment and asked whether it was possible to differentiate those with a subsequent good or poor recovery of the upper limb based solely on a T1-weighted structural brain scan.
Methods A support vector machine approach using voxel-wise lesion likelihood values was used to show that it was possible to classify patients as good or poor recoverers with variable accuracy depending on which brain regions were used to perform the classification.
Results While considering damage within a corticospinal tract mask resulted in 73% classification accuracy, using other (non-corticospinal tract) motor areas provided 87% accuracy, and combining both resulted in 90% accuracy.
Conclusion This proof of concept approach highlights the relative importance of different anatomical structures in supporting post-stroke upper limb motor recovery and points towards methodologies that might be used to stratify patients in future restorative clinical trials.
Stroke is one of the the most common causes of physical disability worldwide and about 80% of stroke survivors experience impairment of movement on one side of the body.1 Hand and arm impairment in particular is often persistent, disabling and a major contributor to reduced quality of life.2 The main predictor of long-term outcome of upper limb function is the level of initial impairment.3 This can be quantified as the proportional recovery rule which states that by 3 months, patients with stroke will recover about 70% of the initial upper limb motor impairment that has been observed on day 3 post-stroke.4–6 The prediction works extremely well for those presenting with mild to moderate upper limb impairment, but in only about half of those with initially severe upper limb impairment.4–6 In the other half, patients do worse than predicted, that is, there is a failure of proportional recovery. A key question then is, what is the difference between patients with stroke matched for initial severity who go on and have different recovery trajectories? The answer to this will point to the factors that are important for the dynamic process of recovery independent from the causes of initial impairment.
One possibility is the anatomy of the damage may be different in each group. A number of recent studies have proposed that the corticospinal tract (CST) plays a decisive role in this categorical difference7–11 as cortical reorganisation for improved motor function ultimately requires access for cortical motor areas to muscles. However, CST lesion load correlates with initial motor impairment,12 which is the major predictor of long-term outcome. It is therefore reasonable to ask how much CST lesion load can improve prediction of long-term outcome over and above initial severity. Furthermore, most of the patients involved in these studies had suffered from subcortical stroke and recent work has suggested that taking account of cortical damage after stroke can improve prediction of the motor clinical consequences.13 14
In this study, we investigated 30 patients with chronic stroke with a range of lesion locations (cortical and/or subcortical involvement) known to have presented with severe initial upper limb impairment but who had gone on to have quite different recovery trajectories. We applied a support vector machine approach to data representing lesion likelihood derived from structural T1-weighted MRI to answer the following questions. First, how accurately can patients with stroke with severe initial upper limb impairment be classified as having either good or poor recovery using only data extracted from whole brain structural MRI? Second, which brain regions contribute most to the classification? The results have the potential to transform how prediction of long-term upper limb outcome after stroke is achieved in routine clinical practice in future. The ability to easily and accurately predict outcome with standard clinical neuroimaging would have important implications for planning of treatment but also for stratification in future trials of restorative therapies.15[…]