Guided by PRISMA standards, this meta-analysis of nine Randomized Controlled Trials (RCTs) evaluates Buyang Huanwu Decoction as a surgical adjunct. Results confirm its clinical efficacy in enhancing sensory-motor recovery and neurological outcomes for postoperative spinal cord injury patients.
Research Article
Clinical Efficacy and Postoperative Nerve Function Recovery of Buyang Huanwu Decoction for Spinal Cord Injury: A Meta-Analysis
* These authors contributed equally
In This Article
Summary
Abstract
This meta-analysis evaluated whether adding Buyang Huanwu Decoction (BYHWD) to standard surgery improves spinal-cord nerve recovery after traumatic spinal-cord injury (SCI). Following PRISMA and PROSPERO protocol CRD42024612859, we systematically searched PubMed, Cochrane, Embase, Web of Science, China National Knowledge Infrastructure (CNKI), Wanfang and VIP to January 2025 for randomized controlled trials (RCTs) comparing surgery plus usual care with the same regimen plus BYHWD. Data were pooled with RevMan 5.3.
Nine Chinese RCTs (614 patients, 306 BYHWD, 308 control) were included. Relative to controls, BYHWD significantly raised post-operative American Spinal Injury Association(ASIA) sensory score (MD 7.51; 95% CI 5.08–9.93; P <0.01), ASIA motor score (MD 7.65; 5.81–9.49; P <0.01), somatosensory evoked potential amplitude (MD 0.37 µV; 0.18–0.55; P <0.01), motor evoked potential amplitude (MD 0.43 µV; 0.19–0.68; P <0.01) and the proportion of patients achieving ≥1 ASIA grade improvement (RR 1.23; 1.09–1.38; P <0.01). Heterogeneity was low for most outcomes. No serious herb-related adverse events were reported. Adjunctive BYHWD appears to accelerate short-term neurological recovery after surgical decompression for SCI. However, all trials were Chinese, most lacked allocation concealment or blinding, follow-up was brief and outcome sets varied. Large-scale, rigorously blinded, multicentre RCTs with long-term functional endpoints are needed before routine clinical adoption.
Introduction
Spinal Cord Injury (SCI) is a devastating central nervous system trauma typically caused by external mechanical forces, resulting in partial or complete loss of sensory, motor, and autonomic functions1,2. Globally, the incidence and prevalence of SCI are increasing, with an estimated 250,000 to 500,000 new cases reported annually3,4. In China specifically, the burden is substantial, with approximately 759,300 cumulative cases and 66,374 newly diagnosed instances each year5. The average age of onset is 31.7 years, predominantly affecting individuals aged 15–25 years and males (male to female ratio of 4:1). Notably, traffic accidents (36%–48%) are the leading cause, followed by violence (5%–29%), falls (17%–21%), and recreational activities (7%–16%)6. SCI imposes a dual burden: a significant physical and psychological toll on patients and a substantial economic strain on society. The Need for Adjuncts Current standard management for SCI primarily involves surgical decompression to restore spinal stability, pharmacotherapy (e.g., corticosteroids) to mitigate secondary inflammation, and comprehensive rehabilitation to maximize functional independence7. While surgery is critical for relieving compression, it does not directly facilitate neural regeneration. Traditional Chinese Medicine (TCM) offers potential complementary strategies. Buyang Huanwu Decoction (BYHWD), a classic TCM formulation known for its qi-supplementing and blood-activating properties, is theorized to alleviate the "stasis" and "blockage" associated with SCI in TCM pathology. Previous research indicates that BYHWD can lower post-operative plasma C-reactive protein (CRP) levels, improve neurological scores, shorten hospital stays, and reduce complications8.
However, despite its widespread clinical use in China, the current evidence base remains fragmented. While individual randomized controlled trials (RCTs) have reported benefits, there is a lack of high-level evidence quantitatively synthesizing the efficacy of BYHWD specifically as an adjunct to surgical intervention. To address this gap, this study presents the first systematic review and meta-analysis to strictly evaluate the synergistic effect of BYHWD combined with surgical decompression. We hypothesized that adding BYHWD to standard surgical care would significantly improve neurological recovery, specifically ASIA sensory and motor scores and evoked potentials (SEP/MEP)—compared with surgery alone. By strictly adhering to PRISMA guidelines, this study aims to resolve inconsistencies in existing small-scale trials and provide robust, evidence-based recommendations for the integration of TCM into standardized postoperative SCI management.In traditional Chinese medicine (TCM), spinal cord injury is typically categorized as “bi syndrome” or “wei syndrome.” The pathogenesis is primarily associated with trauma-induced damage to the Du meridian, which governs the body’s yang energy. Once damaged, the yang energy cannot be properly distributed, leading to extravasation of blood outside the vessels and the formation of stasis, which obstructs the meridians and deprives the sinews and vessels of nourishment, thereby causing disease. TCM posits that the treatment of spinal cord injury should focus on replenishing and regulating qi, as well as activating blood circulation to resolve stasis, to restore the normal function of the Du meridian and ensure the smooth flow of qi and blood9.
Protocol
This systematic review and meta-analysis were conducted in strict accordance with the PRISMA 2020 guidelines. As this study utilizes secondary data from previously published trials, institutional review board (IRB) approval was not required; however, all included original studies adhered to ethical standards involving human subjects. The protocol was prospectively registered with the PROSPERO database immediately upon study initiation (Registration number: CRD42025637800).
Search Strategy and Selection Criteria Two independent researchers performed a comprehensive systematic search for Randomized Controlled Trials (RCTs) investigating Buyang Huanwu Decoction (BYHWD) for spinal cord injury (SCI). The search spanned from the inception of each database to January 2025. Electronic databases included PubMed, Embase, The Cochrane Library, Web of Science, CNKI, VIP, Wanfang, and CBM. The search strategy utilized a combination of controlled vocabulary (e.g., MeSH Terms in PubMed, Emtree in Embase) and free-text terms (Title/Abstract fields). Specifically, the boolean operators AND and OR were used to combine terms related to Spinal Cord Injury (e.g., Traumatic SCI, Spinal Injury), Buyang Huanwu Decoction (e.g., Buyang Huanwu Tang), and Randomized Controlled Trial. Taking PubMed as an example:
#1: ((Spinal Cord Injuries[Mesh] OR Spinal Cord Injury[Mesh] OR Spinal Cord Injuries, Traumatic[Mesh] OR Spinal Cord Injury, Traumatic[Mesh]) OR (Spinal Cord Injury OR SCI OR Spinal Injury))
#2: ((Buyang Huanwu Tang[Mesh] OR Buyang Huanwu Decoction[Mesh]) OR (Buyang Huanwu Tang OR Buyang Huanwu Decoction))
#3: (randomized controlled trial[pt] OR controlled clinical trial[pt] OR randomized[tiab] OR placebo[tiab] OR drug therapy[sh] OR randomly[tiab] OR trial[tiab] OR groups[tiab]) NOT (animals[mh] NOT humans[mh])
#4: #1 AND #2 AND #3
Filters and Limits: The Humans filter was activated where available. No language restrictions were imposed. Export Formats: All retrieved records were exported in standard citation formats (e.g., .nbib, .ris, or .ciw) and imported into reference management software (Table of Materials). Duplicates were removed using EndNote’s automatic deduplication function followed by manual verification.
Study Selection and Data Extraction Outcomes were screened based on strict inclusion criteria: (1) RCT study design; (2) participants diagnosed with SCI undergoing surgical treatment; (3) an experimental group receiving BYHWD plus routine surgical care versus a control group receiving routine surgical care alone; and (4) reporting of quantitative outcome measures including ASIA scores, evoked potentials (SEP/MEP), or clinical efficacy rates. Exclusion criteria encompassed animal studies, non-RCTs, duplicates, and studies with incomplete data. Two researchers independently screened titles, abstracts, and full texts. Disagreements regarding study eligibility were resolved operationally through immediate discussion; if consensus could not be reached, a third senior reviewer (Z.P.W.) made the final decision based on the full-text review. Data extraction was performed using a standardized spreadsheet form (see Table of Materials), capturing first author, year, demographics, interventions, and outcomes.
Risk of Bias Assessment The methodological quality of included RCTs was assessed independently by two reviewers using the Cochrane Risk of Bias Tool (RoB 2). For each included study, detailed justifications were provided for judgments across five domains: (1) Randomization process—sequence generation was considered low risk if computer-generated random numbers or randomization tables were used, and high risk if alternation, date of birth, or case record number was used; allocation concealment was judged low risk if centralized allocation, sealed opaque envelopes, or pharmacy-controlled randomization was employed, and high risk if open allocation schedule or unconcealed envelopes were used; (2) Deviations from intended interventions—blinding of participants and personnel was assessed as low risk for double-blind studies, some concerns for single-blind or objective outcomes, and high risk for open-label studies with subjective outcomes; (3) Missing outcome data—judged as low risk if <5% missing data with intention-to-treat analysis, moderate risk if 5%–20% missing, and high risk if >20% missing or per-protocol analysis with significant attrition (>15% difference between groups); (4) Measurement of the outcome—assessed as low risk if outcome assessors were blinded or outcomes were objective (e.g., mortality), and high risk if assessors were aware of group allocation and outcomes were subjective; (5) Selection of the reported result—evaluated by comparing protocols (if available) or methods sections against reported results, rated high risk if primary outcome was changed or selectively reported. Discrepancies between reviewers were resolved through discussion or consultation with a third reviewer (Sun).
Evidence quality was graded using the GRADE system across four levels: High quality (further research is very unlikely to change our confidence in the effect estimate; started as high for RCTs and rated down for limitations); Moderate quality (further research is likely to have an important impact on confidence and may change the estimate); Low quality (further research is very likely to have an important impact and is likely to change the estimate); and Very low quality (any estimate of effect is very uncertain). Quality was rated down from high for: risk of bias (serious limitations in randomization or blinding in >25% weight of studies), inconsistency (I2 > 50% or non-overlapping confidence intervals), indirectness (differences in population, intervention, comparator, or outcome measures), imprecision (optimal information size not met [<300 events for dichotomous outcomes or <400 participants for continuous outcomes] or 95% CI includes both meaningful benefit and harm), and publication bias (evident from funnel plot asymmetry, Egger's test p < 0.10, or <10 included studies with apparent small-study effects).
Statistical Analysis and Software Workflow Meta-analysis was performed using meta-analysis software (Table of Materials). Continuous data (e.g., ASIA scores, SEP/MEP amplitude) were entered as mean and standard deviation (SD) to calculate Mean Differences (MD) with 95% Confidence Intervals (95% CI). For dichotomous outcomes (e.g., efficacy rates), data were entered as event counts (number of events/total sample size) to calculate Risk Ratios (RR) with 95% Confidence Intervals (CI) using the Mantel-Haenszel method; for continuous outcomes (e.g., ASIA scores), the Inverse Variance method was employed to calculate Mean Differences (MD) with 95% CI. All analyses were initially performed using a fixed-effect model in the meta-analysis software (Table of Materials).
Heterogeneity assessment and model selection: Statistical heterogeneity across studies was quantified using the I2 statistic and Cochran’s Q test. The pre-specified threshold for significant heterogeneity was set at I2 > 50% or P < 0.10. If heterogeneity exceeded these thresholds, indicating that true effect sizes likely varied between studies rather than arising from chance alone, the analysis model was switched to a random-effects model (Der Simonian-Laird method) to provide a more conservative pooled estimate that accounts for between-study variance.
Sensitivity analysis: To test the robustness of findings, particularly for outcomes exhibiting high heterogeneity, sensitivity analyses were conducted by sequentially omitting one study at a time (leave-one-out method) and recalculating the pooled effect estimate for the remaining studies. If the omission of a single study substantially altered the direction or statistical significance of the overall effect (e.g., causing the 95% CI to cross the null value of 1.0 for RR or 0 for MD), that study was considered influential, and the stability of the results was interpreted with caution.
Publication bias assessment: Funnel plots were generated to visually assess publication bias only when at least 10 studies were included in the meta-analysis, as fewer studies result in insufficient power to distinguish true asymmetry from random scatter. When applicable, funnel plot asymmetry was interpreted considering the involvement of small-study effects; for analyses with fewer than 10 studies, publication bias assessment was not performed due to inadequate statistical power.
Results
The final meta-analysis comprised nine randomized controlled trials with 614 well-characterized patients, providing a robust evidence base to test our hypothesis8,9,10,11,12,13,14,15,16. The pooled sample demonstrated balanced allocation between treatment arms (306 patients receiving BYHWD plus surgery versus 308 receiving surgery alone), ensuring adequate statistical power to detect clinically meaningful differences in neurological recovery (Table 1). This substantial, homogeneous patient cohort—derived from rigorous Chinese-language RCTs—provided the necessary methodological rigor to evaluate whether BYHWD supplementation significantly enhances post-surgical neurological outcomes beyond standard care. Controls: The surgical-only control group (n = 308) served as the critical counterfactual, controlling for surgical technique variability and spontaneous recovery trajectories. The near-equal distribution between experimental and control groups (306 vs. 308) minimized allocation bias and confirmed that observed improvements in motor/sensory scores (detailed in subsequent results) could be attributed to the herbal intervention rather than baseline imbalances or procedural confounders.
Risk of Bias and Quality Assessment
Among the 9 included studies, 47,12,14,15explicitly utilized random number tables or other randomization methods for group allocation. In contrast, the remaining studies did not describe their specific randomization procedures. Regrettably, none of the studies mentioned blinding, allocation concealment, or detailed study protocols. Figure 2 presents the overall risk of bias summary across all domains, illustrating the proportion of studies judged as low, unclear, and high risk for each methodological criterion. Figure 3 provides the detailed risk of bias assessment for each individual study across the five RoB 2 domains, offering specific insights into the methodological quality of each included trial.
Statistical Analysis
ASIA Sensory Score
A total of 8 studies8,9,10,11,12,13,14,15 reporting ASIA sensory scores were included in this meta-analysis. Heterogeneity testing revealed low heterogeneity among the studies, prompting the use of a fixed-effects model for the meta-analysis. The results indicated that the experimental group had a significantly higher ASIA sensory score compared to the control group, with a mean difference (MD) of 7.51 (95% CI: 5.08 to 9.93; P < 0.01), indicating a statistically significant difference. For detailed results, refer to Figure 4.
ASIA Motor Score
A total of 6 studies8,11,12,13,15 reporting ASIA motor scores were included in this meta-analysis. Heterogeneity testing indicated low heterogeneity among these studies, leading to the selection of a fixed-effects model for the meta-analysis. The results demonstrated that the experimental group had a significantly higher ASIA motor score compared to the control group, with a mean difference (MD) of 7.65 (95% CI: 5.81 to 9.49; P < 0.01), indicating a statistically significant difference between the two groups. For detailed results, refer to Figure 5.
Mechanistic Validation from Independent Studies
The significant improvement in ASIA motor scores observed in our meta-analysis (MD 7.65; P < 0.01) correlates with recent independent experimenta&l findings. Specifically, Luchun et al17. demonstrated in a spinal cord injury model that BYHWD exerts a neuroprotective effect by inhibiting ferroptosis via the regulation of the GPX4/ACSL4 axis. This suppression of iron-dependent cell death preserves motor neurons, providing a biological mechanism that supports the pooled clinical efficacy data reported here.
Somatosensory Evoked Potentials (SSEPs)
A total of 2 studies12,15 reporting somatosensory evoked potentials (SSEPs) were included in this meta-analysis. Heterogeneity testing revealed low heterogeneity among the studies, prompting the use of a fixed-effects model. The results indicated that the experimental group had a significantly higher SSEP score compared to the control group, with a mean difference (MD) of 0.37 (95% CI: 0.18 to 0.55; P < 0.01), indicating a statistically significant difference. For detailed results, refer to Figure 6.
Motor Evoked Potentials (MEPs)
A total of 2 studies12,15 reporting motor evoked potentials (MEPs) were included in this meta-analysis. Heterogeneity testing revealed high heterogeneity among the studies (I2 = 75%), prompting the use of a random-effects model. The results indicated that the experimental group had a significantly higher MEP score compared to the control group, with a mean difference (MD) of 0.43 (95% CI: 0.19 to 0.68; P < 0.01), indicating a statistically significant difference. For detailed results, refer to Figure 7.
Number of Cases with Spinal Cord Injury Improvement
A total of 3 studies11,14,15reporting the number of cases with spinal cord injury improvement were included in this meta-analysis. Heterogeneity testing revealed low heterogeneity among the studies, prompting the use of a fixed-effects model. The results indicated that the experimental group had a significantly higher number of cases with spinal cord injury improvement compared to the control group (RR 1.23; 95% CI: 1.09 to 1.38; P < 0.01). For detailed results, refer to Figure 8.
Publication Bias and Quality Assessment of Outcome Measures
Publication bias is typically assessed for outcome measures with >10 included studies. Given that all outcome measures in this study included fewer than 10 studies, funnel plots were not generated. An evidence grading software was used to rate the quality of the outcome measures included in the studies. See Table 2.
Sensitivity Analysis
Given the high heterogeneity observed in the motor evoked potentials (MEPs) outcome (I2 = 75%), we scrutinized the two included studies12,15. Due to the limited number of studies (n = 2), a standard statistical sensitivity analysis was constrained. However, a qualitative comparison revealed that both studies independently favored the experimental group with statistical significance. This suggests that while the magnitude of the effect size varies—likely attributable to differences in baseline injury severity and specific herbal modifications between the trials—the direction of the therapeutic benefit is consistent and robust.
Data Availability
The raw data file containing the meta-analysis are publicly available. The complete dataset has been uploaded to an open-access cloud repository and can be accessed at: https://pan.baidu.com/s/11iW6T5afrYQ4qJndHxJZCQ?pwd=kfqw. Detailed characteristics of all included studies are cited within the References section.
Figure 1. Flowchart of Literature Screening. Systematic search and selection process following PRISMA 2020 guidelines, depicting the flow of records from initial database searches (n = 484), through duplicate removal (n = 235), full-text screening (n = 249), to final inclusion of nine RCTs. Please click here to view a larger version of this figure.
Figure 2. Bias Risk Assessment of Included Studies. Individual risk of bias ratings for each of the nine included studies across five RoB 2 domains (randomization process, deviations from intended interventions, missing outcome data, measurement of outcome, and selection of reported result), categorized as low risk, some concerns, or high risk. Please click here to view a larger version of this figure.
Figure 3. Summary of Bias Risk Assessment of Included Studies. Overall proportion of studies rated as low risk, some concerns, or high risk for each of the five RoB 2 domains, presented as horizontal stacked bar graphs. Please click here to view a larger version of this figure.
Figure 4. Forest Plot of ASIA Sensory Scores. Meta-analysis of eight studies comparing postoperative ASIA sensory scores between BYHWD plus surgery (experimental) and surgery alone (control), showing individual and pooled mean differences with 95% confidence intervals using a fixed-effects model (I2 = 0%). Please click here to view a larger version of this figure.
Figure 5. Forest Plot of ASIA Motor Scores. Meta-analysis of six studies comparing postoperative ASIA motor scores between groups, demonstrating individual effect sizes and pooled estimates using a fixed-effects model (I2 < 50%). Please click here to view a larger version of this figure.
Figure 6. Forest Plot of Somatosensory Evoked Potentials. Meta-analysis of two studies comparing SSEP amplitudes between treatment and control groups using a fixed-effects model, indicating significant improvement with BYHWD adjunctive therapy. Please click here to view a larger version of this figure.
Figure 7. Forest Plot of Motor Evoked Potentials. Meta-analysis of two studies comparing MEP amplitudes using a random-effects model (I2 = 75%), noting significant heterogeneity between studies but consistent directional benefit favoring BYHWD. Please click here to view a larger version of this figure.
Figure 8. Forest Plot of the Number of Cases with Spinal Cord Injury Improvement. Meta-analysis of three studies reporting clinical improvement rates (defined as ≥1 ASIA grade improvement), displaying risk ratios and 95% confidence intervals using a fixed-effects model. Please click here to view a larger version of this figure.
| Investigator | Sample size (n) T/C | Age (years) | Intervention(C) | Intervention(T) | observational indicators |
| Zhang2014 | 31/32 | 33.6 | Posterior decompression + Pedicle screw fixation | Interventions in the Control Group+Buyang Huanwu Decoction | ①② |
| Zhang2019 | 18/18 | 38.6 | Traction + Surgical Decompression | Interventions in the Control Group+Buyang Huanwu Decoction | ②⑥ |
| Chen2015 | 30/30 | 35.36±5.32 | Posterior decompression + Pedicle screw fixation | Interventions in the Control Group+Buyang Huanwu Decoction | ② |
| Guo2015 | 36/36 | 34.3±7.2/33.4±6.9 | Posterior decompression + Pedicle screw fixation | Interventions in the Control Group+Buyang Huanwu Decoction | ①②③ |
| Lu2015 | 42/42 | 36.80±9.70/ 37.60±10.10 | Posterior decompression + Pedicle screw fixation | Interventions in the Control Group+Buyang Huanwu Decoction | ①②⑦⑧ |
| Deng2016 | 52/53 | 48.5±5.7/ 48.9±5.3 | Posterior decompression + Pedicle screw fixation | Interventions in the Control Group+Buyang Huanwu Decoction | ①② |
| Yang2017 | 31/31 | 44.50± 1.76/ 42.00±1.23 | Posterior decompression + Pedicle screw fixation | Interventions in the Control Group+Buyang Huanwu Decoction | ①②③ |
| Qing2021 | 30/30 | 50.26±13.15/50.40±13.89 | Intradural Tumor Resection + Routine Medication | Interventions in the Control Group+Buyang Huanwu Decoction | ④⑤⑨⑩ |
| Zhang2022 | 36/36 | 34.19 ± 5.47/34.74 ± 5.08 | Posterior decompression + Pedicle screw fixation | Interventions in the Control Group+Buyang Huanwu Decoction | ①②③⑦⑧ |
Table 1: Basic Characteristics of the Literature. T: Experimental Group; C: Control Group; ① ASIA Motor Score; ② ASIA Sensory Score; ③ Number of Cases with Spinal Cord Injury Improvement; ④ Patient Benefit; ⑤ Neuron-Specific Enolase (NSE) Score; ⑥ Number of Adverse Events; ⑦ Somatosensory Evoked Potential (SSEP) Amplitude; ⑧ Motor Evoked Potential (MEP) Amplitude; ⑨ JOA Score; ⑩ C-Reactive Protein (CRP)
| Outcome indicator | Number of studies | Study design | Risk of bias | Inconsistency | Imprecision | Indirectness | Other consideration | Certainty |
| ASIA Sensory Score | 8 | randomised trials | seriousa | not serious | seriousb | not serious | None |  ◯◯Low |
| ASIA Motor Score | 6 | randomised trials | seriousa | not serious | seriousb | not serious | None | ◯◯Low |
| SSEPs | 2 | randomised trials | seriousa | seriousc | seriousb | not serious | None | ◯◯◯Very low |
| MEPs | 2 | randomised trials | seriousa | not serious | seriousb | not serious | None | ◯◯low |
| Improvement in Spinal Cord Injury | 3 | randomised trials | seriousa | not serious | not serious | not serious | None | ◯Moderate |
Table 2: Quality Assessment of Outcome Indicators Using the GRADE System
Discussion
Postoperative Assessment of Spinal Cord Injury
Spinal Cord Injury is a severe central nervous system disorder characterized by diverse and complex treatment and assessment methods. In terms of therapeutic approaches, conservative management encompasses pharmacological interventions, such as the use of corticosteroids to alleviate edema and neurotrophic agents to promote nerve repair7,18. Rehabilitation therapies include physical, occupational, and psychological interventions aimed at enhancing muscle strength, restoring activities of daily living, and addressing the psychological state of patients19,20. Surgical treatments primarily involve decompressive surgery to relieve spinal cord compression and fixation-fusion surgery to stabilize the spine. Spinal cord repair surgeries, such as neural stem cell transplantation, are still in the experimental phase21,22.
In the assessment of outcome measures, neurologic function assessment holds a central position23,24. The American Spinal Injury Association (ASIA) Impairment Scale is widely used to assess the severity of injury 25,26, ranging from complete injury (Grade A) to normal (Grade E), providing a key basis for evaluating treatment efficacy. ASIA sensory and motor scores quantify neurologic recovery by grading sensory perception and muscle strength at key points. Quality of life is assessed using the SCI-QOL27,28 and Activities of Daily Living (ADL) assessments, measuring patients' quality of life from multiple dimensions. Complications are evaluated by the incidence of pressure sores, urinary tract infections, and pulmonary infections 29,30, reflecting the quality of patient care and rehabilitation outcomes. Walking ability is assessed using the Walking Functional Measure (WFMS), 10-meter walk test, and 6-minute walk test 31,32, measuring patients' walking capacity and motor function recovery. These comprehensive assessment metrics provide a holistic reference for the treatment and rehabilitation of spinal cord injury, aiding in the development of personalized treatment plans and thereby improving patients' quality of life and prognosis.
Among the studies included in this review, a lack of uniformity in outcome measures was observed. Specifically, research on ASIA sensory and motor scores, number of improved cases, and evoked potentials was relatively abundant, whereas studies involving complications, SCI-QOL, ADL, and walking ability assessments were less frequent. This situation suggests that future research should broaden the scope of outcome measures and strive to develop a more comprehensive and systematic assessment framework.
Mechanistic Insights and Clinical Implications
Buyang Huanwu Decoction (BYHWD) is a traditional Chinese herbal formula that plays a significant role in postoperative neurological recovery following spinal cord injury. Its multi-pathway and multi-target mechanisms provide comprehensive support for rehabilitation. In terms of immune cell polarization, BYHWD has been shown to significantly downregulate the expression of proteins in the TLR4/NF-κB signaling pathway, thereby inhibiting the polarization of macrophages towards the pro-inflammatory M1 phenotype and alleviating inflammation33,34. Quercetin, an active component of BYHWD, promotes the polarization of macrophages towards the anti-inflammatory M2 phenotype, which protects neurons. Additionally, BYHWD may regulate the polarization of microglia, inhibit inflammation, and promote neurovascular regeneration. Microglial polarization towards the M1 phenotype can trigger inflammation, inhibit axonal remyelination of neurons, and exacerbate neuronal damage, whereas the M2 phenotype exerts regulatory effects.
BYHWD also reduces the formation of glial scars35,36, significantly inhibiting the expression of GFAP (glial fibrillary acidic protein) and thereby reducing glial scar formation, which in turn promotes the repair of SCI and the recovery of motor function. Furthermore, by inhibiting the expression of BMP2 and BMP4, BYHWD suppresses the differentiation of oligodendrocyte precursor cells into astrocytes, thereby reducing glial scar formation. Regarding the promotion of substance P expression37, BYHWD intervention in a rat model of SCI significantly increased the expression of substance P in the dorsal horn of the rat spinal cord, protecting the injured spinal cord neurons. Although the specific mechanisms require further investigation, this effect highlights the potential neuroprotective role of BYHWD.
BYHWD also inhibits the expression of platelet-activating factor (PAF)38. SCI leads to increased release of PAF, which triggers inflammation. In rats treated with BYHWD, the expression of this factor in spinal cord tissue was significantly lower than that in the model group. The anti-inflammatory and blood-activating effects of BYHWD are closely related to its active components, which improve platelet aggregation and inhibit the expression of inflammatory factors. For example, carthamin, a component of BYHWD, inhibits PAF-induced platelet aggregation and the release of inflammatory factors, thereby providing a favorable microenvironment for nerve repair35.
BYHWD can reduce the levels of C-reactive protein (CRP)35, an important inflammatory factor in the inflammatory microenvironment of SCI. Elevated CRP levels can exacerbate inflammation. When combined with conventional Western medicine for the treatment of cervical SCI, BYHWD is more effective in reducing CRP levels than Western medicine alone. Meta-analyses have also confirmed the significant advantages of combined treatment with BYHWD, which can significantly alleviate inflammation and provide a favorable microenvironment for SCI repair. Furthermore, BYHWD inhibits the JAK2/STAT3 signaling pathway39, which is activated following SCI and is associated with cell apoptosis and nerve regeneration. BYHWD significantly inhibits the expression of p-JAK2 and p-STAT3 proteins in SCI tissues, thereby reducing inflammation and treating traumatic paraplegia. Additionally, BYHWD inhibits ferroptosis by regulating the GPX4/ACSL4 axis17.Following SCI, the expression of GPX4 is downregulated, while that of ACSL4 is upregulated, leading to significant ferroptosis. BYHWD increases GPX4 expression and reduces ACSL4 expression, thereby inhibiting ferroptosis, reducing the accumulation of lipid peroxides, exerting neuroprotective effects, and promoting the recovery of neurological function.
In conclusion, Buyang Huanwu Decoction exerts its therapeutic effects on spinal cord injury (SCI) through a multi-pathway and multi-target mechanism of action. Specifically, it regulates the polarization of immune cells, reduces the formation of glial scars, promotes the expression of substance P, inhibits the expression of platelet-activating factor, lowers the levels of C-reactive protein, suppresses the JAK2/STAT3 signaling pathway, and regulates the GPX4/ACSL4 axis to inhibit ferroptosis. Collectively, these mechanisms alleviate post-injury inflammation and facilitate the recovery of neurological function. By providing comprehensive support for postoperative neurological rehabilitation, Buyang Huanwu Decoction demonstrates significant potential in the treatment of SCI.
The clinical efficacy observed in this study is strongly supported by modern pharmacological evidence, which offers alternative ways of understanding the "Qi-supplementing and blood-activating" effects of BYHWD. The therapeutic potential of BYHWD appears to be mediated through multi-target mechanisms: Immunomodulation and Anti-inflammation: Secondary injury after SCI is largely driven by an uncontrolled inflammatory cascade. BYHWD has been shown to downregulate the TLR4/NF-κB signaling pathway, inhibiting the polarization of microglia/macrophages towards the pro-inflammatory M1 phenotype while promoting the reparative M2 phenotype32,33. This shift is crucial for preserving spared neural tissue. Additionally, components like quercetin and carthamin inhibit platelet-activating factor (PAF) and reduce C-reactive protein (CRP) levels, creating a microenvironment conducive to repair35,37. Inhibition of Glial Scarring: Glial scar formation acts as a physical and chemical barrier to axonal regeneration. Previous studies indicate that BYHWD significantly inhibits the expression of GFAP and modulates the BMP2/BMP4 signaling pathway, thereby suppressing astrocyte hypertrophy and scar formation34,35. Neuroprotection and Anti-oxidation: Recent research highlights the role of BYHWD in regulating the GPX4/ACSL4 axis to inhibit ferroptosis, a form of iron-dependent cell death prevalent in SCI39. Furthermore, by suppressing the JAK2/STAT3 pathway, BYHWD reduces neuronal apoptosis and promotes the expression of neuroprotective Substance P in the dorsal horn36,38.
It is worth noting that while high-dose methylprednisolone (MP) has historically been a standard pharmacological treatment for acute SCI, its use remains controversial due to the high risk of systemic complications such as gastrointestinal bleeding and infection40. In contrast, our meta-analysis suggests that BYHWD provides significant neurological benefits with a favorable safety profile, as no serious herb-related adverse events were reported in the included studies. Unlike MP, which primarily targets acute inflammation, BYHWD offers a multi-faceted approach suitable for the sub-acute and recovery phases. While direct head-to-head trials comparing BYHWD versus neuroprotective agents are currently lacking, the present data support BYHWD as a safe and effective adjunctive therapy that may complement or reduce reliance on high-risk pharmacological interventions.
Limitations and Future Perspectives
Despite the positive outcomes achieved in this study, several limitations must be acknowledged. Firstly, the majority of the included studies did not mention allocation concealment, the implementation of blinding, or blinding in outcome assessment. Only three studies reported the use of blinding, which compromises the reliability of the effectiveness evaluation. The high heterogeneity observed for MEPs (I2 = 75%, Figure 7) likely reflects differences between the two included studies in injury severity, assessment timing, and herbal formulations. Despite this variability, both trials showed consistent benefit favoring BYHWD, supporting the robustness of the pooled estimate under the random-effects model. These findings should be interpreted cautiously pending future RCTs with standardized electrophysiological protocols. Secondly, all the included studies were published in Chinese. Although China has abundant traditional Chinese medicine (TCM) resources with in-depth research and extensive clinical practice, SCI is a globally relevant condition. Therefore, the findings may be subject to language and publication bias. Additionally, most studies had short trial durations and inconsistent postoperative follow-up times, lacking long-term evaluations of therapeutic outcomes. Finally, there is an absence of a comprehensive spinal cord injury (SCI) assessment system, with inconsistent evaluation criteria across studies. Thirdly, the analysis of evoked potentials (SSEPs and MEPs) was restricted to two studies each. While both showed consistent benefits, the small number of trials limits the precision and generalizability of these estimates. Future RCTs should incorporate standardized electrophysiological assessments to corroborate these preliminary observations.
Alternative validation approaches beyond clinical randomized trials, the neuroprotective effects of BYHWD can be further validated through animal studies. Using transgenic animal models (e.g., genetically modified mice with fluorescent markers) to observe the effects of BYHWD on spinal cord gliosis and inflammation resolution in real time. Patient stratification: Classifying patients based on inflammatory biomarkers (e.g., CRP, IL-6 levels) to identify specific subgroups with optimal responses to BYHWD, which may also help explain heterogeneity in electrophysiological outcomes (MEP) across studies. Imaging studies: Employing MRI diffusion tensor imaging (DTI) to objectively visualize nerve fiber regeneration and overcome inter-center variability in electrophysiological measurements41,42.
To facilitate the global adoption of this protocol, future research should focus on three key areas. First, standardization requires large-scale, multicenter, double-blind randomized controlled trials to confirm these findings with rigorous controls for bias. Second, optimization should determine the optimal therapeutic window, including dosage and duration, and investigate potential drug–drug interactions with standard neuroprotective agents. Third, mechanistic depth requires further translational research using transgenic animal models to pinpoint the precise molecular targets of BYHWD’s active ingredients, facilitating its acceptance in mainstream Western medicine.
This systematic review and meta-analysis provide quantitative evidence that Buyang Huanwu Decoction (BYHWD), when administered as an adjunct to surgical decompression, significantly enhances neurological recovery and functional outcomes in patients with traumatic spinal cord injury—findings that converge with extensive wet lab validation from independent laboratories32,33,34,35,36,37,38,39. Specifically, the observed improvements in ASIA motor scores align with animal studies demonstrating BYHWD-mediated inhibition of ferroptosis39and glial scarring34, while the reduced inflammatory markers in clinical subjects corroborate mechanistic findings of suppressed TLR4/NF-κB signaling32,33and JAK2/STAT3 pathways38. By synthesizing data from nine randomized controlled trials, this study advances the field by offering a scientific basis for integrating Traditional Chinese Medicine into standardized postoperative rehabilitation protocols. However, the strength of this recommendation is tempered by the methodological limitations of the current evidence, specifically the lack of rigorous blinding, allocation concealment, and long-term follow-up in the primary studies. Consequently, while BYHWD demonstrates promising clinical efficacy as a complementary therapy, its routine global adoption warrants further validation through large-scale, multicenter, double-blind randomized controlled trials adhering to international reporting standards.
Disclosures
The authors declare that they have no competing financial interests or personal relationships that could have influenced the work reported in this paper.
Acknowledgements
Scientific Research Project of Sichuan Provincial Administration of Traditional Chinese Medicine (GrantNos.2023MS182,2023MS192,2024MS276); Special Scientific Research Project of Chengdu University of Traditional Chinese Medicine (Grant No. DJYB2021009).
Materials
| Name | Company | Catalog Number | Comments |
|---|---|---|---|
| Chinese National Knowledge Infrastructure (CNKI) | Tongfang Knowledge Network | Database | https://www.cnki.net |
| Cochrane Library | Wiley | Database | https://www.cochranelibrary.com |
| Embase | Elsevier | Database | https://www.embase.com |
| EndNote X9 | Clarivate Analytics | Software | https://endnote.com |
| GRADEpro system tool | Evidence Prime Inc. | Software | https://www.gradepro.org/product |
| PubMed | National Library of Medicine | Database | https://pubmed.ncbi.nlm.nih.gov |
| R Software (Version 4.5.2) | R Foundation for Statistical Computing | Software | https://www.r-project.org |
| Review Manager (RevMan) 5.3 | The Cochrane Collaboration | Software | https://training.cochrane.org |
| Web of Science | Clarivate Analytics | Database | https://www.webofscience.com |