Journal of Rehabilitation Medicine 51-5 | Page 11

Recovery-promoting drugs after stroke: a systematic review Adjuvant physical therapy was inconsistently repor- ted and insufficient to allow for replication. There was extreme variation between therapy amount, type (i.e. Bobath vs Arm Ability training; physiotherapy vs oc- cupational therapy, etc.) and duration. In 15 studies, no adjuvant therapy was reported (15, 17, 18, 21, 34, 35, 48, 49, 51, 52, 54, 55, 61, 62, 64), while in another 13 studies dose of adjuvant therapy was not reported (25, 30–32, 40, 42, 44, 45, 50, 53, 56, 57, 63). In 3 studies, a total of ≤ 60 min of adjuvant therapy was provided over the duration of the trial (16, 22, 26). It was deemed impossible to perform data syntheses (meta-analyses) to compare RPDs regardless of drug class, due to the large variability in design, duration and outcome measures. DISCUSSION Eighteen of 28 drug interventions identified in this review demonstrated recovery-promoting poten- tial without associated increased rates of mortality or SAEs. Yet, there were high attrition rates and bias, and variable outcomes used, which prevented meta- analysis. These issues are not isolated to RPD; the Stroke Recovery and Rehabilitation Roundtable group highlighted this as common in stroke rehabilitation trials (10, 65, 66). Nevertheless, several classes of RPDs should be discussed in more detail. Three SSRIs were found to have some evidence of efficacy and safety: citalopram, escitalopram and fluoxetine (17, 18, 24, 30, 45, 60). Typically used as antidepressants, SSRIs inhibit serotonin reuptake into presynaptic neurones thereby enhancing nerve transmission. Motor excitability over the unaffected hemisphere is thought to be decreased, whilst neuro- protective capacity and hippocampal neurogenesis is promoted (67). Of all SSRIs reviewed, fluoxetine was most extensively studied (4/7 SSRI trials with largest cohorts n = 8–118) (18, 28, 30, 45). It is therefore unsurprising that fluoxetine is involved in 3 current international trials (FOCUS, AFFINITY, EFFECTS, combined n =5,045 at May 2018), the results of which will provide reliable estimates of effect (68, 69). Levodopa (as single-drug intervention) was the subject of 5 studies in this review, 4 of which were favourable (15, 16, 19, 20, 43). Replenishing deple- ted striatal dopamine, levodopa stimulates dopamine pathways to increase motor activity (67). Trials admi- nistered immediate-release levodopa preparations just prior to motor retraining, in order to favourably exploit levodopa’s short duration of action, theoretically pri- ming the brain and maximizing remodulation of neural pathways with minimal side-effects or potential dose tolerance (67). Timing of dose administration relative to physical rehabilitation is an important consideration. 327 Current trials provide insufficient evidence to guide these decisions. Nevertheless, further exploration of levodopa as an RPD appears worthwhile. Safety measurement was inconsistent. When as- sessed, mortality and AE were predominantly not different to placebo. Assessment of safety may have been overlooked, in part, due to dosages tested being consistent with dosages used for other indications, with previously established safety profiles. Implementation of standardized guidelines for measurement of safety e.g. International Council for Harmonisation Harmo- nised Tripartite Guideline S7a – Safety Pharmacology Studies For Human Pharmaceuticals, would improve trial rigour and increase potential for meta-analyses in future (70). This review demonstrates the challenge of com- prehensively and easily identifying all RPD studies, even with a robust systematic approach. While 1,548 articles were identified for screening from the com- prehensive database search, yielding 29 studies for inclusion in this review, a further 3,231 citations were identified from the references and forward citations of these included studies. This probably highlights the inconsistent categorization of RPD studies within research databases, which relies on several variables, including limitations of current non-specific MeSH and key terms to adequately tag publications, and the personal preferences and perspectives of the submitting authors when ascribing MeSH and key terms to their submissions (71). If RPD research is to continue to gain momentum as an important field of study, developing a dedicated MeSH term, such as “recovery-promoting drug”, is worth consideration. Personalization of RPD intervention for stroke sur- vivors based on individual recovery needs, medical profile, personal preferences and character traits is an exciting prospect. With several RPDs demonstrating potential efficacy, how and for whom they are prescri- bed requires careful consideration. Coupling a more detailed understanding of RPD pharmacology and biological processes responsible for motor recovery may aid the development of a more ordered classifica- tion system for RPDs based on their biological targets. Differing mechanisms of action and varied indica- tions for use of the drugs in this review offer future possibilities of combining RPDs to exploit synergistic effects. Pilot testing of combination therapy would be necessary to establish safety. Based on this review, combined daily dosing of an SSRI, i.e. fluoxetine, and levodopa, administered 60–90 min prior to a clinician- led rehabilitation regimen of evidence-based adjuvant physical therapy, has potential to maximize therapeu- tic value by capitalizing on different mechanisms of action. The results of the fluoxetine mega-trials are awaited with interest. J Rehabil Med 51, 2019