We performed comprehensive literature searches in PubMed, Embase, Web of Science, and Cochrane databases, covering publications from their inception to 30 October 2024. The following keyword combinations were utilized: (“percutaneous endoscopic lumbar discectomy” OR “PELD” OR “endoscopic discectomy” OR “platelet-rich plasma” OR “PRP” OR “platelet concentrates”) AND (“lumbar disc herniation” OR “LDH” OR “herniated lumbar disc” OR “lumbar intervertebral disc herniation”). Boolean operators (including OR, AND, NOT) were used to ensure both comprehensiveness and precision in the search terms. The complete search strategy for PubMed is provided in Supplementary Table S1. Two authors independently screened the retrieved literature and evaluated it against the inclusion criteria based on titles and/or abstracts. Discrepancies were resolved through discussion with a third senior author. Full-text articles meeting the inclusion criteria were comprehensively reviewed, and their references were manually examined to ensure the inclusion of all relevant studies. This meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (17, 18). The study was registered with the International Prospective Register of Systematic Reviews (PROSPERO) under the ID CRD42024621150 prior to initiating the database search and study selection process.

Inclusion criteria: 1)

The study subjects were adult patients diagnosed with lumbar disc herniation (LDH) who met the diagnostic criteria based on clinical symptoms and imaging examinations (e.g., MRI or CT);

Two independent authors (Hua Song and Ying Zhang) screened titles and abstracts. The extracted information primarily included the first author, year of publication, study design, age, gender, sample size, follow-up duration, PRP injection volume, pain and functional scores at different follow-up time points, and primary outcomes (number of recurrence cases and complications). The summarized data were reviewed and verified by a third author to ensure accuracy.

This study included one RCT and three retrospective cohort studies, with the quality of the RCT assessed using The Cochrane Collaboration's tool (19) and the retrospective cohort studies evaluated using the Newcastle-Ottawa Scale (NOS). Specifically, The Cochrane Collaboration's tool evaluates seven aspects: random sequence generation, allocation blinding, blinding of participants, blinding of outcome measures, incomplete outcome data, selective reporting, and other biases. The judgment of risk of bias is expressed using three categories: “low risk”, “high risk”, and “unclear risk”. A NOS score of ≥7 was considered low risk of bias, a score of 4–6 indicated moderate risk of bias, and a score <4 was classified as high risk of bias. Two independent and experienced authors rated the studies, and final scores were determined through discussion with a senior third author.

This study used the I² statistic and χ² test to evaluate heterogeneity. According to the Cochrane Handbook, heterogeneity was classified as follows: 0%−40% as low heterogeneity, 30%−60% as moderate heterogeneity, 50%−90% as substantial heterogeneity, and 75%−100% as considerable heterogeneity. When I² ≤ 50% and p > 0.10, the combined data were considered to have no significant heterogeneity, and a fixed-effect model was applied. Otherwise, a random-effects model was used for pooled effect analysis (20). Differences in dichotomous variables, such as recurrence rate and the Macnab criteria, were analyzed by calculating the odds ratio (OR), while continuous variables were evaluated by calculating the mean difference (MD), both with 95% confidence intervals (CI) (21). Subgroup analyses were conducted based on the time points for each outcome measure. We planned to assess publication bias using funnel plots and Egger's test if more than 10 studies were included for any outcome. Sensitivity analyses were planned to assess the impact of excluding studies with high risk of bias. All data analyses were performed using RevMan version 5.4 software (The Cochrane Collaboration, Copenhagen, Denmark). A p-value < 0.05 was considered statistically significant.

A comprehensive search of four major electronic databases (PubMed, Embase, Web of Science, and Cochrane databases) initially identified 2,536 articles. After independent screening by one author, 2,143 duplicate articles were manually removed. Subsequently, the titles and abstracts of the remaining articles were reviewed, resulting in the exclusion of an additional 376 articles. Finally, 17 articles were subjected to full-text review, and 4 articles meeting the eligibility criteria were selected for further data analysis. The PRISMA flow diagram for this study is presented in Figure 1.

Table 1 provides detailed information on the study characteristics of the included articles. In terms of study design, all studies were retrospective cohort studies (22–24) except for one (25), which was an RCT. Overall, the studies were published between 2022 and 2024, with patient data originating exclusively from China. A total of 421 patients were included in the data analysis, of which 212 received PELD combined with PRP treatment, while the remaining 209 underwent PELD treatment alone. Specifically, the patients' age range was 31.7–58.35 years, BMI ranged from 18.2 to 28.27, and disease duration ranged from 1.19 to 31.94 months. Surgical levels were concentrated in L3/4, L4/5, and L5/S1. For PRP, the total blood volume used ranged from 18 to 50 ml, with an injection volume of 4–5 ml. Regarding outcome measures, three of the studies reported complications. Two studies comprehensively reported VAS scores for back and leg pain at four follow-up time points, all four studies reported ODI and VAS scores for low back pain at three months, three studies reported the Macnab criteria, and three studies reported the Pfirrmann grade.

Two independent authors meticulously rated the four included studies, with any discrepancies resolved through discussion with a third researcher. The RCT explicitly reported using a random number table for randomization. Apart from unclear risk for allocation concealment and blinding of outcome assessment, all other domains were rated as low risk of bias (Supplementary Figure S1). Two cohort studies received an NOS score of 9, and one study received a score of 8. All studies were evaluated as having a low risk of bias (Supplementary Table S2).

To investigate whether PELD combined with PRP treatment offers a significant advantage over traditional PELD surgery in reducing the overall recurrence rate in patients with lumbar disc herniation, data on recurrence cases from three relevant studies were collected. The heterogeneity test results (I² < 50%, p = 0.87) indicated no significant heterogeneity among the studies. The pooled data from these three studies demonstrated that the recurrence rate was significantly lower in the PRP + PELD group compared to the PELD-only group (OR: 0.21, 95% CI: 0.07 to 0.64, p = 0.006) (Figure 2).

The VAS score for low back pain is commonly used in clinical practice to assess the rehabilitation status of patients after lumbar surgery. All four included studies utilized this measure to evaluate pain outcomes for back and leg pain at different follow-up time points (3 days, 3 months, 6 months, and 12 months) (Figures 3A–D). For the assessment of leg pain, the pooled results demonstrated that the PRP + PELD group was associated with significantly lower pain levels at 3 months (short-term period) (MD: −0.28, 95% CI: −0.45 to −0.10, p = 0.002) and 6 months (mid-term period) (MD: −0.41, 95% CI: −0.62 to −0.20, p = 0.0001). However, pooled analyses for 3 days (shorter-term period) and 12 months (long-term period) showed no significant differences in leg pain scores between the groups.

For low back pain, the pooled results indicated that the PRP + PELD group was associated with significantly lower pain levels at the short-term period (3 months) (MD: −0.32, 95% CI: −0.50 to −0.14, p = 0.0004) and longer-term period (12 months) (MD: −0.28, 95% CI: −0.46 to −0.10, p = 0.002) follow-up time points. However, no significant differences in the VAS score for low back pain were observed between the two groups at shorter-term (3 days) and mid-term (6 months) evaluations (Figures 4A–D). Given the variations in PRP preparation and intervention across studies, a subgroup analysis of the 3-month VAS score for low back pain was conducted based on preparation method, total blood volume, and injection volume. The results indicated that, compared with the PELD-only group, the PRP preparation kits subgroup, total blood volume >30 ml subgroup, and an injection volume of 4.0 ml subgroup may exert a beneficial effect in reducing VAS scores for low back pain. However, due to the relatively small sample sizes within these subgroups, these findings should be interpreted with caution (Supplementary Figure S2).

Although the follow-up time points varied, all four studies reported relevant data on the ODI score. In the shorter-term period (3 days), the pooled results indicated that the PRP + PELD group was associated with a significantly lower functional disability index (MD: −0.53, 95% CI: −0.89 to −0.16, p = 0.005). However, at 3 months, 6 months, and 12 months, no significant differences in ODI scores were observed between the two groups (Figures 5A–D). The subgroup analysis suggested that the total blood volume >30 ml subgroup and the 4.0 ml injection volume subgroup may have a positive effect on improving the 3-month ODI score. However, due to limitations in the number of studies and sample sizes, these findings should be interpreted with caution (Supplementary Figure S3).

A total of three studies reported the Macnab criteria for postoperative patients, and we conducted pooled analyses of the data classified as “excellent”, “good” and “fair”. The results of the pooled analysis showed that the PRP + PELD group had a significantly higher rate of “excellent” outcomes (MD: 1.68, 95% CI: 1.07 to 2.65, p = 0.03), while the PELD group had a higher rate of “good” outcomes (MD: 0.59, 95% CI: 0.37 to 0.94, p = 0.03) (Figures 6A–C).

Three studies reported the Pfirrmann grade outcomes for a total of 266 postoperative patients. The pooled analysis results showed that the PRP + PELD group had significantly fewer patients with Pfirrmann grade IV degeneration postoperatively (OR: 0.48, 95% CI: 0.29 to 0.79, p = 0.004) (Figures 7A–C), suggesting a potential improvement in disc degeneration.

In the evaluation of the JOR score, no significant differences were observed between the two groups (Figure 8). Interestingly, in the analysis of disc height at the final follow-up, the pooled results showed that the PRP + PELD group was associated with significantly greater disc height (MD: 0.88, 95% CI: 0.57 to 1.20, p < 0.00001) (Figure 9).