Longitudinal Multimodal Brain Mapping and Post-Stroke Neuroplasticity: A Systematic Review and Meta-Analysis of Combined Neuroimaging and Molecular Biomarker Evidence

Why clinicians should know about it

• Picked for Complementary and Alternative Medicine (paper of the day, 21 September 2026): Systematic review/meta-analysis of acupuncture for stroke recovery
• Picked for Rehabilitation (paper of the day, 20 September 2026): Systematic review/meta-analysis of post‑stroke neuroplasticity and acupuncture
• Picked for Epidemiology (top studies of the week, 20 September 2026): Systematic review/meta‑analysis of stroke neuroplasticity, not epidemiology
• Picked for Biochemistry (medical) (top studies of the week, 20 September 2026): Brain mapping and acupuncture in stroke

Abstract

BACKGROUND: Stroke is the leading cause of long-term neurological disability worldwide. Multimodal neuroimaging and blood-based molecular biomarkers enable longitudinal characterization of post-stroke neuroplasticity, and acupuncture and Tuina are proposed mechanistic modulators of recovery. Evidence across these domains has not been quantitatively synthesized within a PRISMA 2020-compliant framework. OBJECTIVES: To quantitatively synthesize (i) the prognostic value of multimodal neuroimaging and blood-based molecular biomarkers and (ii) the efficacy of acupuncture and electroacupuncture (EA) on motor and neurological recovery after stroke. METHODS: Following PRISMA 2020, PubMed, EMBASE, Web of Science, CENTRAL, and CNKI were searched from 1 January 2015 to 15 April 2026. Eligible studies were RCTs, prospective cohort studies, and prognostic biomarker studies in adult stroke populations. Two reviewers independently screened, extracted, and appraised risk of bias (RoB 2, NOS, QUIPS). Random-effects (DerSimonian-Laird) meta-analyses were used, with heterogeneity quantified by I². Certainty of evidence was rated using GRADE. Two primary outcomes were pre-specified per synthesis arm; all other analyses were pre-specified as secondary or exploratory. The protocol was submitted to PROSPERO (record under review; registration identifier to be inserted upon assignment). RESULTS: Of 3,412 identified records, 48 studies were included and 21 contributed to meta-analyses. Blood NfL predicted 3-month unfavourable mRS (pooled OR 1.54, 95% CI 1.29-1.84; I² = 38%; 6 studies). DTI-CST fractional anisotropy asymmetry at 2-4 weeks predicted 3-6-month Fugl-Meyer outcome (pooled SMD -0.78, 95% CI -1.02 to -0.54; 7 studies). Acupuncture/electroacupuncture plus rehabilitation improved FMA (pooled MD 8.14, 95% CI 6.22-10.05; 12 RCTs) and reduced NIHSS (pooled MD -2.41, 95% CI -3.18 to -1.63; 9 RCTs). Electroacupuncture and subacute-phase initiation produced larger effects. Tuina was represented by a single trial and was summarized narratively only (not meta-analysed). Sensitivity analyses were concordant; funnel asymmetry was detected for the FMA outcome (Egger p = 0.04; trim-and-fill adjusted MD 7.23). CONCLUSIONS: This synthesis suggests that moderate-certainty evidence is compatible with NfL and DTI-CST fractional anisotropy acting as prognostic markers of post-stroke outcome, and that low-to-moderate-certainty evidence is compatible with a possible adjunctive benefit of acupuncture/electroacupuncture when added to conventional rehabilitation. Important limitations include substantial clinical and methodological heterogeneity, detectable small-study effects for the pooled motor-recovery outcome, a predominance of single-country (Chinese) unblinded trials in the acupuncture evidence base, and limited ability to blind acupuncture providers. Findings should therefore be interpreted cautiously and regarded as hypothesis-refining rather than confirmatory. Adequately powered, pre-registered, STRICTA-compliant multicentre trials with blinded outcome assessment are the principal priority before strong clinical recommendations are issued.

Abstract as published, via PubMed.