Fostering Poststroke Recovery

Despite countless experimental studies demonstrating neuroprotective effects in the treatment of acute ischemic stroke, they all could not live up their promises when it came to translation into patients. Recent research therefore increasingly shifted to therapies enhancing poststroke recovery with the important advantage of a much wider time window up to several months. To further increase poststroke recovery, the combination of different therapy principles is a logical step. However, several burning questions concerning the modus of combination are unresolved. This review will briefly mention the principle poststroke recovery approaches and discuss the mostly used combination designs. In the second part, the central issues of combination paradigms will be discussed, and some most promising paradigms will be presented.

Biological Background

Ischemic stroke immediately triggers a cascade of molecular and cellular events in the perilesional tissue which both further escalate the primary damage but contrariwise also are part of an endogenous response aided to reduce brain injury. Even more, these primarily adverse effects are the starting point to remove irreversibly damaged tissue and to repair and to reorganize surviving structures with remapping of related brain regions later on. The alterations in the peri-ischemic brain become the cornerstone for a particularly sensitive period of enhanced poststroke plasticity with a short window of a regenerative friendly milieu (Zeiler and Krakauer).¹ This is reflected by the fact that the poststroke recovery is widely restricted to the first 3 to 6 months in human patients and about 1 month in rodent models of ischemic stroke. One of the key events thought to prepare the ground for regenerative processes plays the inflammatory response which starts very early after ischemia.² Triggered by inflammatory cells and the activation of extracellular matrix-degrading enzymes, the tissue surrounding the infarct becomes permissive and allows sprouting of axonal processes with the formation of new connections and dendritic remodeling for a small time window. However, these plastic processes are not restricted to the immediate peri-infarct region but occur also in contralateral and remote areas, including the spinal cord.^3,4 Stroke also induces neurogenesis and angiogenesis which are reliant on a supportive milieu. Although in the very acute phase after focal cerebral ischemia the shift in the balance between excitation and inhibition toward excitation increases excitotoxic effects, a long-lasting inhibition in the subsequent chronic phase impedes regenerative processes.⁵

Regenerative Therapy Strategies

A great variety of therapeutic principles to enhance poststroke recovery are available which will briefly be discussed.

The rationale of cell-based strategies is to replace damaged cells and to provide trophic support to progenitor and mature surviving cells. In rodent models of ischemic stroke, various cell types, such as embryonic stem cells, induced pluripotent stem cells, neural stem cells, glial precursor cells, bone morrow–derived stem cells, including hematopoietic stem cells, mesenchymal stem cells, and mononuclear cells, have been used.⁶ To which extend these cells differentiate into mature functional cells and directly contribute to poststroke functional recovery is controversially discussed. However, there is broad consensus that indirect effects like production of growth factors, immunomodulation, and modulation of cortical excitability play a major role. Although these cell-based strategies are very promising in experimental stroke models and ideal candidates for combination therapies, the translation into the clinic as a standard treatment in stroke therapy is far from being achievable. Current issues are the ideal route and the time window of cell administration and safety concerns.

The pharmacological approach comprises a broad spectrum of substances and biological active molecules targeting various components of the repair process. The use of neuromodulators and neuroenhancers in this context is long known. Amphetamine and levodopa are prominent examples, but multiple other pharmaceuticals have shown beneficial effects in particular when combined with neurorehabilitative therapies.⁷ Recently, the serotonin reuptake inhibitor fluoxetine has shown beneficial effects in motor recovery in stroke patients.⁸

Within the group of bioactive molecules, beneficial effects have been demonstrated for a multitude of growth factors, for example, epidermal growth factor (EGF), brain-derived neurotrophic factor (BDNF), erythropoietin (EPO), granulocyte-colony stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), glial cell line-derived neurotrophic factor (GDNF), ciliary neurotrophic factor (CNTF), and interferon-γ just to cover the most frequently tested ones.⁹ The effect profile of the various factors is multimodal and typically more or less overlapping with an impact on neurogenesis, angiogenesis, axonal sprouting, dendritic plasticity, synaptogenesis, the inflammatory response, glial scar formation, or the balance between excitation and inhibition. Because many of these factors have clinically already approved for treatment of other diseases, they are particularly attractive candidates for a fast translation into the clinic.

Other biologicals include monoclonal antibodies targeting specific recovery hostile molecules. The most prominent example is an antibody against Nogo-A which is able to blockade its growth-inhibiting activity.⁴ The extracellular matrix may also be targeted pharmacologically by manipulating the chondroitin sulfate proteoglycans and their formation into perineuronal nets to produce a regeneration friendly milieu free from growth-inhibiting molecules.¹⁰

During the past years, the important role of noncoding RNAs, including miRNAs for brain remodeling and in this context for stroke recovery, has become evident.¹¹ These molecules are able to decrease the expression of specific genes through mRNA destabilization and translational repression with a particular impact on neurogenesis, angiogenesis, oligodendrogenesis, and axonal remodeling. The understanding how miRNAs interplay with histone deacetylases and exosomes which are also involved in remodeling processes will provide a powerful tool to modulate cellular pathways facilitating recovery after stroke. Currently, one major issue is the development of cell type–specific modulation of these molecules. Although of high interest, this therapeutic track is still in an early phase and far from translation into the clinic.

Another fascinating and rapidly emerging therapeutic option which is perfectly predestinated for combination therapy is the use of biomaterials.^12,13 Biomaterial hydrogels can be directly injected into the damaged brain providing a supportive matrix that promotes cellular infiltration and axonal growth. Furthermore, these hydrogels can serve as delivery vehicles for cells or bioactive molecules to enhance and complement endogenous repair processes. The development of transplantable living scaffolds represents another facet in the evolving field of biomaterials.¹⁴

Various neurorehabilitative training paradigms have shown beneficial effects in animal models of chronic stroke, although negative results have been reported too.¹⁵ Roughly, these training strategies can be subdivided into task-specific paradigms and more unspecific ones such as physical exercise in general. Further, one can discriminate between voluntary training and forced training such as constraint-induced movement therapy (CIMT) or treadmill training which seem to provide more benefit.¹⁵ The optimal time point for initiation of the training is under discussion because some studies show that a too early beginning will worsen neurological outcome. Combination with all the other mentioned therapeutic strategies seems theoretically plausible (Figure 1). Importantly, these rehabilitative approaches currently represent the cornerstone in the treatment of stroke-induced disability in human patients without evidence that one specific therapy is superior to another.¹⁶

• Download figure
• Download PowerPoint

Enriched environment is an alternative therapeutic approach which is thought to improve regenerative processes per se in a multifaceted way, including induction of neurogenesis and increase of the production of trophic factors.¹⁷ Importantly, enriched environment can easily be combined with any other therapeutic strategy. One has also to concern that in the clinical setting deprivation of physical, cognitive, and social activity is a real problem. Astonishingly, few attempts have been made to translate this cost-effective therapy into the clinic.

Last but not least, there is a plethora of literature on noninvasive brain stimulation techniques as a further option for stroke recovery by modulating cortical excitability.¹⁸ Similar to rehabilitative approaches, they are already in clinical use and may be theoretically combined with other therapeutic strategies but are not in the focus of the present review.

Experimental and Clinical Approaches to Combination Treatments

Each of these therapy principles mentioned above harbors specific technical problems, for example, systemic versus local application of cells, biologicals or pharmaca. One more general problem is that all these regenerative mechanisms are well orchestrated and highly interdependent. The major challenge is to combine therapeutic approaches in a way that potentiates beneficial effects and avoids neutralizing or even worsening ones. Therefore, a deep understanding of the individual repair mechanisms and their interplay is a prerequisite for a rationale planning of a combination therapy. In principle, combination treatments are possible among all of the above listed modalities, for example, synergistically acting drug combinations, including bioactive molecules and antibodies, stem cell grafting plus drug treatment, and combinations of biomaterials, miRNAs and drug treatment (Figure 1). The most common approach, however, represents the combination of a drug treatment (eg, neurotransmitters or neurotransmitter modulating drugs) to a specific or forced rehabilitative training (Table). In the landmark prospective, randomized and controlled trial in 53 primary stroke patients, 100 mg levodopa was concurrently combined to intensive physiotherapy and proofed to enhance motor function compared with physiotherapy alone.¹⁹ A later and smaller study in 18 chronic stroke patients tested dopamine treatment (3 doses of 100 mg) concurrently combined to procedural motor learning of the paretic hand by using a modified version of a serial reaction time task. Patients treated with this combination showed improved procedural motor function compared with placebo.²⁰ Dopamine treatment in both studies was well tolerated with no major side effects compared with placebo.

Another neuromodulator, d-amphetamine, was combined to different modes of physical rehabilitation represented by a control environment, enriched environment, or enriched environment with additional sessions of focused activity after permanent ischemia.²¹ Animals treated with amphetamine (treatment 10 minutes before motor training on days 2, 5, and 8 poststroke) and intensive rehabilitation performed significantly better than any other treatment group achieving complete motor recovery by 8 weeks poststroke. These effects were largely confirmed in a primate study using a cortical infarction model in the squirrel monkey demonstrating that a single dose of d-amphetamine on the first day of training initiated 10 days after stroke onset facilitated the rate of recovery and improved performance (68% improvement from first day of training) compared with monkeys treated with saline (27% improval).²²

In a relatively large randomized, controlled stroke recovery trial, the serotonin reuptake inhibitor fluoxetine was added to standard physiotherapy in 118 subacute stroke patients. Those subjects treated with the combination showed a significantly better outcome measured with the Fugl-Meyer motor scale (almost 45% improval).⁸ Overall, tolerability of study medication was good. Of the main adverse events, just nausea, diarrhea, and abdominal pain occurred more frequent in the fluoxetine group, whereas hepatic enzyme disorders occurred more frequent in the placebo group. A recent experimental study in mice suggested worse outcome when the beginning of poststroke motor training (skilled prehension task) was delayed to day 7 compared with day 1.²³ Importantly, fluoxetine treatment immediately after stroke (day 1) was able to overcome this training gap resulting in complete motor recovery comparable to the mice trained early after stroke.

Growth factors, such as BDNF, GDNF, insulin-like growth factor (IGF), basic fibroblast growth factor (bFGF), EPO, or G-CSF, represent the most intensively studied group of drugs able to promote neuronal growth and recovery after ischemic stroke. An intraventricular 2-week infusion of epidermal growth factor and erythropoietin after cortical stroke in rats resulted when followed by an 8-week enriched rehabilitation program (including voluntary access to reach a training chamber) in significant acceleration in motor recovery (at 4 weeks compared with 10 weeks in those without enriched rehabilitation).²⁴ Interestingly, when growth factor treatment with BDNF is combined to early CIMT, the benefit of the sole BDNF treatment on sensorimotor function disappears.²⁵ Importantly, blocking BDNF function by antisense BDNF oligonucleotide in animals receiving either a graduated rehabilitation program, including running exercise and skilled reaching training, or no rehabilitation negated the beneficial effects of rehabilitation on recovery of motor function, such as skilled reaching.³⁴ Although the effect on recovery in various experimental models and paradigms was conclusively shown, clinical translation was not successful to date. Treatment of acute stroke patients with growth factors resulted either in increased intracranial bleedings and an increased mortality (EPO stroke trial)³⁵ or was simply ineffective (AXIS-II G-CSF trial [AX200 for the Treatment of Ischemic Stroke]).³⁶ Of all growth factors, only G-CSF was tested in a randomized, double-blinded stroke recovery trial. In this study, 41 patients were treated with a 10-day intensive training paradigm in hand motor function and finger tapping speed and motor and verbal learning combined to daily doses of G-CSF (day 6–10) after a 5-day period of pretreatment.²⁷ The study proofed feasibility of this approach but failed to show significant benefits of the combination therapy. This approach was recently replicated in a blinded 2×2 factorial design study, where G-CSF was applied (5 doses subcutaneously) together with moderate physical training (18 home-based therapy sessions) and compared with no treatment or no training.²⁸ Again feasibility was confirmed, but the study also failed to show efficacy of this approach—which can at least partly be explained by the small group sizes and in particular length of induction of intervention from stroke event (about 1 year after stroke onset). In both studies, more severe adverse events occurred in G-CSF–treated patients, and some of these (infections and vascular events) were possibly related to G-CSF treatment. This finding was expected and correlated to data from hematology and oncology studies, where G-CSF treatment was well tolerated, and side effects self-limiting and clinically acceptable.

Recently, cerebrolysin, a neuropeptide preparation with growth factor–like effects, was combined in a multicenter, randomized, double-blinded design over a treatment period of 21 days to intensive rehabilitation therapy (2 hours of physical and 1 hour of occupational therapy daily on workdays) in 70 patients with subacute stroke.²⁹ Combination treatment in the overall group failed to show additional efficacy, but patients with severe motor impairment (Fugl-Meyer score <50) improved significantly after combination therapy. Overall, cerebrolysin was well tolerated. Severe adverse events were similar between treatment groups and unrelated to study medication.

Pharmacologically blocking the growth inhibitory effects of various myelin-associated proteins is another approach to enhance poststroke regenerative capacity. In rats, the continuous intraventricular infusion of myelin-associated neurite outgrowth disinhibitor NEP 1 to 40 (Nogo extracellular peptide residues 1-40) showed, when concurrently combined to motor training for 4 weeks, better sensorimotor outcome over the whole observation period compared with no training or no NEP application.³⁰ Myelin-associated inhibitors limit axonal regeneration via activation of the Rho–Rho-associated protein kinase pathway that can be inhibited by, for example, fasudil. Combination of fasudil to CIMT beginning 1 day after focal cerebral ischemia in rats resulted in better motor outcome (skilled reaching and foot fault test) compared with all other groups (fasudil and CIMT) over the 4-week period after stroke.³¹ A novel approach is the inhibition of Nogo-A function using blocking antibodies (anti-Nogo-A) which in rats showed early after stroke improval of motor function because of sprouting in peri- and contralesional areas.^37–39 Interestingly, anti-Nogo-A treatment resulted in worse outcome (even compared with no treatment at all) when simultaneously applied with intensive forced-use training of the forelimb during the first 2 weeks after the stroke. When intensive training (skilled forelimb reaching) was delayed to 2 weeks poststroke in animals treated with this growth-enhancing antibody, motor function could be restored.⁴ These results complement earlier animal studies were early intensive motor training using CIMT after stroke proofed to be detrimental,^40,41 and combination studies with concurrent G-CSF treatment failed to achieve synergistic effects.²⁶

Other examples include anti-inflammatory treatment with either indometacin or minocycline (intraperitoneal application twice daily during the first 2 weeks after cortical stroke in rats) combined with rehabilitative training.³² The combination approach strongly improved sensorimotor performance over 4 weeks poststroke compared with training alone. Activation of the cAMP/CREB pathway is thought to be crucial for experience-dependent neural plasticity, potentially involved in poststroke recovery processes and can be enhanced by phosphodiesterase inhibitors. Treatment with rolipram, the classical phosphodiesterase inhibitor 4 or the novel HT-0712 inhibitor concomitantly given during rehabilitation treatment in rats, significantly enhanced motor recovery poststroke and induced expansion of distal movement representations that extended beyond residual motor cortex.³³

Deductions for the Optimal Timing of Combination Treatments

The major issues that arise are (1) what kinds of therapeutic principles can be combined, (2) when is the ideal start point for the respective therapy, (3) should the different therapies be applied in a concurrent or sequential manner, and (4) what is the optimal dosage of the respective therapy. Bringing together our current knowledge of the spontaneous poststroke recovery processes with the available experimental and clinical data in the literature concerning combination therapies, one can circumspectly answer some of the most urgent questions concerning combination therapy.

What Kinds of Therapeutic Principles Can Be Combined?

Trying to classify the available poststroke recovery approaches, one can roughly distinguish priming therapies from consolidating ones (Figure 1). All the different therapy paradigms which are primarily aimed to enhance spontaneously occurring recovery such as neurogenesis, angiogenesis, axonal sprouting, dendritic remodeling or synaptogenesis, and so on would be considered as priming ones. Every therapy which primarily is intended to exercise newly formed connections thereby stabilizing and strengthen them whereas nonused become pruned and eventually disappear would be classified as consolidating. This subgroup primarily includes all forms of training paradigms but also any form of noninvasive brain stimulation. Importantly, the consolidating therapies can additionally be combined with a pharmacotherapy, for example, with neuromodulators or neuroenhancers that additionally strengthen learning effects. And in fact, the most frequently tested combination therapies are designed according to this rationale showing beneficial effects both in experimental animals and humans (Table).

However, there are yet many theoretically promising combinations (Figure 1) which have not been tested so long in humans and only partly in experimental stroke models. This is primarily because of the fact that in the group of the priming strategies some are just upcoming ones such as biomaterials or targeting miRNAs. Other priming approaches, such as cell-based strategies, are not feasible yet because of various technical problems or safety concerns, whereas the therapeutic approaches in the consolidating group are more or less well established.

In this context, enriched environment takes a special position because it clearly has beneficial effects in the early priming stage as shown by its positive effects on, for example, neurogenesis, although on the other hand it has also been proven effective when applied in the consolidating phase by stabilizing newly formed connections and pathways.¹⁷

When Is the Ideal Starting Point for the Respective Therapy?

Taking into account the biological basis of stroke recovery, all the priming paradigms may be started as early as possible after stroke. The situation is more complicated for the consolidating training therapies. There is strong evidence from experimental stroke models that very early intensive training impairs functional outcome when initiated within minutes to hours after onset of cerebral ischemia.^25,41–43 This early impulse is even so counterproductive that long-term motor learning and final outcome are impaired 6 weeks after cortical stroke.^25,41 On the other hand, skilled motor training in mice has been shown to be less effective when started 7 days compared with 1 day after photothrombotic stroke.²³ Therefore, the optimal time point for starting and the duration of neurorehabilitative therapies continue to be a matter of debate (Figure 2). There are also data supporting a direct combination of priming and consolidation therapy (Table; Figure 2). The situation in patients is even more complex because of the different cause, size, and location of the infarcts, as well as comorbidity and multimedication. To find the ideal time point to start the respective therapy (ie, not too early and too late), remains a major challenge yet unresolved.

• Download figure
• Download PowerPoint

Should the Different Therapies Be Applied in a Concurrent or Sequential Manner?

Conceptually, a priming therapy has to be followed by a consolidating approach (Figures 1 and 2). Direct evidence comes from an elegant study by Wahl et al.⁴ Using the photothrombotic stroke model in rats, they could show that only the sequential therapy with first neutralization of the growth inhibitory molecule Nogo-A for 2 weeks and then a subsequent skilled training paradigm resulted in nearly complete recovery, whereas the concurrent therapy failed. Successful recovery could be attributed to an enhanced sprouting of midline crossing fibers from the intact motor cortex and spinal cortical tract. In the concurrent paradigm, hyperinnervation and aberrant sprouting could be identified as causal for failure of recovery. Interestingly, the concurrent treatment group not only failed to benefit from the combination but in fact performed worse than all other groups.⁴ Additional indirect evidence comes from a study of our group. Using also the photothrombotic stroke model in rats, we tested treatment with the growth factor G-CSF and a forced training paradigm in concurrent and sequential applications.²⁶ Although all combinations with G-CSF resulted in improved sensorimotor outcome, CIMT alone did not reach significance level. Importantly, starting with CIMT followed by G-CSF did not further enhance recovery.²⁶ Nevertheless, there are also examples of beneficial combination therapies different from this theoretical concept.

What Is the Optimal Dosage of the Respective Therapy?

Currently, knowledge of the optimal dosage, which is defined by frequency, duration, and intensity of the respective therapy, is more or less based on try and error approaches. From the literature available, it becomes also clear that there exists a dosage maximum which by going beyond will result in less benefit. This holds true not only for pharmacological treatment but has also been convincingly demonstrated for rehabilitative therapy. In the VECTORS study (Very Early Constraint-Induced Movement During Stroke Rehabilitation), stroke patients with high-intensity CIMT therapy did worse compared with the standard CIMT group and even compared with a standard treatment control group.⁴⁴ Theoretically, infinite combinations are possible by varying the single parameters of the dosage of each single therapy. Furthermore, this variation is multiplied when combining priming and consolidating therapy. In the future, stroke researchers should more involve biostatisticians and mathematicians to develop combination treatments with maximum benefit on a rational basis. However, also for single therapies, such approaches are exceptionally rare.⁴⁵

General Practical Issues

It is clear that testing the concept of priming and consolidating therapies in preclinical and clinical practice poses major problems. There is no simple solution for this challenge. Large sample sizes and sophisticated experimental designs will be necessary to discriminate distinct therapeutic effects of the various components. We should not repeat the mistake as with neuroprotective strategies where primarily young, healthy, male animals were investigated just to avoid problems of large numbers, costs, and complex experimental designs. Ignoring the reality of human stroke in experimental models has certainly contributed to the persistent translational failure.⁴⁶ In light of the high absolute number of stroke victims with an increasing percentage of stroke survivors, even the smallest progress in regenerative stroke medicine may justify these efforts.

Examples of Top Combination Therapies

Considering the discussed evidence under translational aspects, intensive physiotherapy or training should be started not earlier as 7 or 14 days poststroke. The early poststroke interval should instead be targeted with priming therapies to enhance the subsequent training. Good candidates for such priming include neuromodulators such as dopamine or fluoxetine, which already showed good tolerability in humans and some promising signs of efficacy.^8,19,20,23 Newer and mechanistically exciting drugs include growth inhibitory molecules such Anti-Nogo-A, although for those just experimental evidence is available.⁴ Because clinical evidence does not exist, a study comparing a concurrent versus a sequential treatment design would be desirable.

Conclusions

Poststroke recovery can be enhanced by combination therapies. The initial approach to this therapy includes early poststroke priming with the goal to enhance naturally occurring recovery processes to make the brain susceptible for subsequent rehabilitative training. Different modes of rehabilitation may be additionally supported with neuromodulators or neuroenhancers to strengthen learning effects. Optimal efficacy may be achieved in future clinical or experimental poststroke recovery studies by focusing on such a sequential design irrespective of type and mode of treatments combined. Scientifically unexplored is at present superiority of different combinations over other ones and duration and dose of treatment.

Footnotes