Long-term efficacy of platelet-rich plasma for knee osteoarthritis: A systematic review and meta-analysis of randomized controlled trials

Study design and reporting

This systematic review and meta-analysis were conducted to assess the long-term efficacy and safety of PRP injections for KOA. The study adhered to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 reporting standards and followed a predefined research protocol. All stages – from literature search to statistical synthesis – were performed independently by two reviewers, with discrepancies resolved by consensus or consultation with a senior investigator.

Search strategy

A comprehensive search of PubMed/MEDLINE, Embase, Web of Science, Scopus, and the Cochrane Central Register of Controlled Trials was undertaken from database inception to March 2025. The following search terms and their Boolean combinations were used: (“knee osteoarthritis” OR “gonarthrosis”) AND (“platelet-rich plasma” OR “PRP”) AND (“randomized controlled trial” OR “RCT”). Reference lists of relevant reviews and meta-analyses were screened manually, and additional records were identified through ClinicalTrials. gov and the World Health Organization International Clinical Trials Registry Platform. No restrictions were applied during the electronic search; however, only full-text English-language studies were included in the final analysis. This criterion was adopted to ensure accurate data extraction but may introduce language-related selection bias, which is acknowledged as a methodological limitation.

Eligibility criteria

Studies were selected using the Population, intervention, comparator, outcomes (PICO) framework:

• Population: Adults (≥18 years) with clinically and/or radiographically confirmed primary KOA (Kellgren– Lawrence grade I-III).

• Intervention: Intra-articular PRP injections of any preparation (leukocyte-rich [LR] or leukocyte-poor [LP], activated or non-activated).

• Comparator: HA corticosteroids, ozone, saline, or alternative PRP regimens.

• Outcomes: Primary outcomes included validated pain and function measures – Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), International Knee Documentation Committee score (IKDC), Knee Injury and Osteoarthritis Outcome Score (KOOS), and visual analog scale (VAS) – with a minimum follow-up of 6 months. Secondary outcomes were structural changes (e.g., cartilage volume or joint space width on MRI), reintervention rates, and treatment-related adverse events.

• Study design: RCTs (parallel-group, single- or multicenter).

Exclusion criteria encompassed non-randomized or observational studies, secondary osteoarthritis, Kellgren– Lawrence grade IV disease, adjunctive surgical procedures, incomplete outcome data, or follow-up shorter than 6 months.

Study selection process

Two reviewers independently screened titles and abstracts to identify eligible studies. Full texts of potentially relevant articles were then reviewed against the inclusion criteria. Any disagreements were resolved through discussion or arbitration by a third reviewer. The selection process followed the PRISMA 2020 framework and is summarized in a flow diagram that documents the number of records identified, screened, excluded, and finally included.

Data extraction and management

Data extraction was performed independently by two reviewers using a standardized electronic form. Extracted variables included:

• Study characteristics (first author, year, country, and study design)

• Participant demographics and baseline OA grade

• PRP preparation details (platelet enrichment level, leukocyte content, activation method, and volume per injection)

• Comparator type and dosing schedule

• Duration of follow-up and assessment intervals

• Reported outcomes for pain, function, and structural change

• Adverse events and funding source

To enhance transparency, a comprehensive summary table details PRP formulation parameters across studies, including platelet concentration, leukocyte profile, activation technique, and injection frequency. This enables interpretation of heterogeneity in preparation protocols and dosing schedules.

Quality and risk-of-bias assessment

The methodological quality of each included trial was evaluated using the Cochrane Risk of Bias 2.0 (RoB 2) tool. The assessment covered five key domains:

1. Randomization process

2. Deviations from intended interventions

3. Missing outcome data

4. Measurement of outcomes

5. Selective reporting of results

Each domain was rated as low risk, some concerns, or high risk. Quality appraisal was conducted independently by two reviewers, and discrepancies were resolved by consensus. Results were synthesized in a summary table and illustrated through a traffic-light diagram for clarity.

Statistical analysis

Meta-analyses were performed using a random-effects model (DerSimonian–Laird method) to accommodate variability among studies. Continuous data were expressed as mean differences (MD) or standardized MDs (SMD) with corresponding 95% confidence intervals (CI). When only change-from-baseline values were available, data were standardized to maintain consistency across trials.

Heterogeneity was quantified using the I² statistic (low = 25%, moderate = 50%, and high = 75%) and Cochran’s Q test.

Pre-specified subgroup analyses were conducted to explore sources of heterogeneity, including:

• PRP formulation (LR vs. LP)

• Dosing frequency (single vs. multiple injections)

• Comparator type (HA, corticosteroid, saline, or ozone)

• Duration of follow-up (≤12 months vs. >12 months)

Sensitivity analyses were performed by excluding high-risk-of-bias studies and those with imputed data. Publication bias was assessed using funnel plots and Egger’s regression test when ≥10 studies were available.^[15-17]

All statistical analyses were executed using review manager and comprehensive meta-analysis software packages.

Study selection

The initial database search identified 2,500 records, and an additional 40 studies were retrieved from clinical trial registries and manual searches of reference lists. After the removal of duplicates, 2,200 unique records were screened based on titles and abstracts. Of these, 2,120 articles were excluded for not meeting the inclusion criteria. A total of 80 full-text articles were reviewed for eligibility. Following a detailed assessment, 70 studies were excluded for the following reasons:

• Non-randomized or observational design (n = 25)

• Follow-up duration shorter than 6 months (n = 18)

• Studies involving secondary or non-KOA populations (n = 12)

• Incomplete or insufficiently reported outcome data (n = 9)

• Presence of adjunctive surgical procedures (n = 6)

Ultimately, 10 RCTs met all inclusion criteria and were incorporated into both the qualitative and quantitative syntheses of this meta-analysis.

These studies compared intra-articular PRP injections with HA, corticosteroids, ozone therapy, saline placebo, or alternative PRP dosing regimens in patients with mild-to-moderate KOA. A detailed overview of the selection process is presented in Figure 1 (PRISMA 2020 Flow Diagram), while the inclusion and exclusion criteria are summarized in Table 1.

Close

Figure 1: Preferred reporting items for Systematic Reviews and Meta-Analyses 2020 flow diagram. RCT: Randomized control trial, OA: Osteoarthritis.

Export to PPT

Study characteristics

Across the 10 included RCTs, participants with mild-to-moderate KOA were enrolled in single- and multicenter settings across Asia and Europe. Sample sizes ranged from small single-center studies (~70–90 participants) to a large multicenter trial (n = 610). Follow-up periods spanned 6–60 months, with two trials reporting 5-year outcomes. Most protocols used multi-dose PRP (typically 3 weekly injections); PRP type varied between LP-PRP and LR-PRP. Primary outcomes commonly included WOMAC, IKDC, and VAS, with select studies reporting structural MRI changes and synovial cytokines. Full trial details are provided in Table 2.

Across the ten RCTs, participants with mild-to-moderate KOA were enrolled in both single- and multicenter settings across Asia and Europe [Table 2]. Sample sizes ranged from ~70 to 90 in single-center studies to a large multicenter trial (n= 610), and follow-up spanned 6–60 months, with two trials reporting 5-year outcomes. Comparators included HA, saline placebo, ozone, and within-PRP dose-regimen arms, allowing both head-to-head and optimization comparisons. Most protocols used multi-dose schedules – typically 3 weekly injections – with injected volumes commonly 2–5 mL/dose. Both LP and LR preparations were represented; activation methods were variably reported (e.g., a freeze-thawed LR-PRP protocol in one multicenter RCT). Primary endpoints most often included WOMAC, IKDC, and VAS; several studies additionally reported structural MRI measures and synovial cytokines. Taken together, the trials show a consistent clinical signal favoring PRP over comparators at 6–12 months, with multi-dose regimens generally outperforming single-dose schedules. Durability is supported by long-term data against saline out to 60 months, whereas differences versus HA attenuate by 5 years in at least one multicenter comparison. Where measured, PRP was associated with lower synovial pro-inflammatory cytokines and less MRI-detected cartilage loss, aligning biologic plausibility with clinical outcomes. Variability in PRP type (LP vs. LR) and dosing frequency likely contributes to between-study heterogeneity, motivating the pre-specified subgroup analyses reported later.

RoB assessment

All ten included RCTs were appraised using the Cochrane RoB 2 tool across five domains: randomization process, deviations from intended interventions, missing outcome data, outcome measurement, and selection of the reported result. Most trials demonstrated low risk in all domains, particularly the large multicenter RCTs. A minority of single-center studies had some concerns – most often due to limited reporting of allocation procedures, incomplete details on blinding, or sparse information on protocol registration and handling of attrition. No study met criteria for high RoB, and none were excluded on this basis. Figure 2 summarizes domain-level judgments per study (Low = “Low risk”; Some = “Some concerns”).

Close

Figure 2: Risk of bias 2 – traffic-light. RCT: Randomized controlled trials.

Export to PPT

Quantitative synthesis: Meta-analysis – detailed table of all included RCTs

A comprehensive quantitative synthesis of all ten eligible RCTs provides a granular overview of the primary efficacy outcomes for PRP versus comparators (including HA, saline, ozone, plasma rich in growth factors [PRGF], and corticosteroids), with details on time points, MD or SMD in pain and function scores, structural or biomarker outcomes, and safety. Table 3 below, the table compiles extracted outcome data and effect sizes from each individual study, facilitating side-by-side comparison.

Across the 10 randomized trials, intra-articular PRP consistently outperformed comparators for pain and function, and the advantage was clearest when PRP was given as a multi-dose series. By 12 months, trials reporting percentage change showed ~40–45% reductions from baseline on WOMAC and VAS with PRP, compared with ~21–33% for HA and ≤20% for saline or ozone. These benefits were not only statistically significant but also clinically meaningful: PRP was the only regimen that reliably maintained minimal clinically important difference (MCID) thresholds beyond the early months. In contrast, responses to HA and saline typically waned after 3–6 months.

A dose–response pattern was evident. Protocols using three injections (weekly or every 2 weeks) produced the most durable improvements; at 1 year, triple-dose groups reported average VAS scores around 1.5, versus ~3.4–3.7 in single- or double-dose arms. Signals of structure modification buttressed the symptomatic gains: In the largest multicenter RCT with 5-year follow-up, PRP slowed cartilage loss on MRI (mean change about −1171 mm³ with PRP versus −2311 mm³ with saline), suggesting a possible disease-modifying effect. Biological plausibility was supported by synovial biomarker data – PRP was associated with greater reductions in pro-inflammatory cytokines (e.g., Interleukin (IL)-1β, IL-6, tumor necrosis factor-alpha, and IL-17A) over 1 year than HA, paralleling the clinical improvements.

Safety findings were reassuring. Adverse events were mild and transient (typically post-injection soreness or swelling), and no serious PRP-related events were reported across the KOA trials. Taken together, the evidence indicates that PRP – especially in three-injection schedules – delivers larger and more durable improvements in validated pain and function scales at 12 months and likely through 24 months, with supportive MRI and cytokine data pointing toward potential disease modification. Seetharamaiah et al., in a randomized trial from India involving 90 elbows with tennis elbow, reported that a single PRP injection produced better pain outcomes at 6 months than triamcinolone or saline, while steroid treatment was associated with adverse effects including hypopigmentation and skin atrophy, suggesting greater durability and a better safety profile for PRP in this condition.^[20]

Primary outcome: WOMAC total score – meta-analytic quantitative synthesis

A detailed quantitative synthesis of pooled RCTs demonstrates that intra-articular PRP injections provide superior improvements in WOMAC total score (a validated instrument for pain, stiffness, and function in KOA) at both 6 and 12 months compared to comparators such as HA, ozone, and saline (placebo).

Safety and tolerability

Adverse events for PRP are predominantly mild and transient (e.g., short-term post-injection soreness); no serious, long-lasting, or treatment-limiting effects have been reported across the included studies.

Figure 3 representative forest plot displaying the mean difference (MD) and 95% confidence intervals (CI) for WOMAC total score at 12 months, favoring PRP.

Close

Figure 3: Forest plot–Western Ontario and McMaster Universities Osteoarthritis Index total score at 6 and 12 months. PRP: Platelet rich plasma, HA: Hyaluronic acid, WOMAC: Western Ontario and McMaster Universities Osteoarthritis index.

Export to PPT

PRP injections result in statistically significant, robust, and clinically meaningful reductions in WOMAC total score at both 6 and 12 months compared to HA, ozone, and placebo for patients with mild to moderate KOA. Multi-dose protocols (typically 3 weekly injections) provide the most pronounced and durable benefit. These findings confirm that PRP is an effective, safe, and durable therapy for pain and functional improvement in knee OA, with advantages extending to 1–2 years and beyond.

WOMAC total score at 6 and 12 months

Across high-quality RCTs, PRP consistently outperformed HA, ozone, and saline in improving WOMAC total scores. At 6 months, Raeissadat et al.^[16] reported a mean change in WOMAC total of −19.0 (95% CI: −20.9, −17.1) for PRP, corresponding to a 39% improvement from baseline. This was significantly greater than HA (−14.7; ~29.8% improvement) and ozone (−11.9; ~22.2% improvement). Only PRP and its derivative, PRGF, sustained benefits beyond 6 months. At 12 months, the PRP group in Raeissadat et al.^[16] maintained a −15.5 (95% CI: −17.4, −13.6) change (~36.5% improvement), whereas HA (−8.4) and ozone (−6.4) declined to non-significant benefits. The between-group difference (PRP vs. HA) was −7.1 (95% CI: −9.6, −4.7), both statistically and clinically significant [Table 4].

Pooled estimates showed MD ranging from −1.9 to −4.5 in favor of PRP at 6–12 months, consistently above the MCID threshold of 30% improvement [Figure 4a].

Close

Figure 4: Forest plots summarizing pooled mean differences (95% confidence intervals) for (a) Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total score at 6 and 12 months, (b) International Knee Documentation Committee (IKDC) score scores at 12–60 months, and (c) WOMAC-pain, WOMAC-function, and visual analog scale (VAS) pain scores at 12– 60 months, favoring platelet-rich plasma (PRP) over hyaluronic acid (HA), ozone, and saline. Vertical dashed line = no effect. The vertical axis lists the included studies (no numeric units on the y-axis). The effect measure is Mean Difference (MD) with 95% confidence interval (CI) on the horizontal axis. Outcome scales: WOMAC total (0–96 points; Likert), IKDC (0–100 points), VAS pain (0–10 points/cm), and WOMAC subscales reported in points (pain 0–20, stiffness 0–8, function 0–68; total 0–96).

Export to PPT

IKDC functional outcomes (12–60 months)

Three major RCTs^[6,2,18] demonstrated the superiority of PRP in IKDC scores compared with HA, saline, or different PRP protocols. At 12 months, Lin et al.^[6] reported an IKDC improvement of +14.0 points (95% CI: 6.0, 22.0) with PRP, exceeding the MCID and significantly greater than HA and saline. Subramanyam et al.^[18] found the triple PRP regimen achieved a mean IKDC score of 71.7 versus 61.8 (double PRP) and 60.6 (single PRP). Long-term follow-up from Chu et al.^[2] showed PRP’s functional superiority persisted at 24, 36, and 60 months, with a pooled MD of 12.6 points (95% CI: 8.7, 16.5) across trials [Figure 4b].

WOMAC subscales and VAS pain (12–60 months)

PRP produced greater and more durable improvements in both WOMAC-pain and WOMAC-function subscales compared to HA, ozone, and saline. At 12 months, Raeissadat et al.^[16] observed WOMAC-pain reductions of −4.4 (95% CI: −4.9, −4.0) with PRP versus −3.1 (HA) and −1.7 (ozone). WOMAC-function improved by −10.0 (95% CI: −11.0, −8.8) in PRP patients, compared to −5.8 (HA) and −4.4 (ozone). PRP also provided superior pain relief measured by the visual analog scale (VAS): −3.3 versus HA (−2.6) and ozone (−1.3) at 12 months. In Yurtbay et al.,^[19] triple PRP dosing achieved a −1.6 (95% CI: −2.7, −0.5) difference versus saline at 24 months, and Chu et al.^[2] reported −2.7 (95% CI: −3.6, −1.8) at 60 months. Pooled effects across studies showed MDs of −3.2 (−4.5, −1.9) for WOMAC subscales and −2.3 (−3.3, −1.5) for VAS, both exceeding clinical relevance thresholds. [Figure 4c].

Structural outcomes

Among trials reporting imaging, only the large multicenter RCT by Chu et al.^[2] (60-month follow-up) quantified cartilage volume change. PRP halved tibiofemoral cartilage loss versus saline (≈6.7% vs. 13.0% over 5 years), consistent with a potential disease-modifying effect.

Over 5 years, PRP was associated with ~50% less cartilage loss than saline, aligning structural change with the superior pain/function outcomes observed at 12–24 months [Table 5].

Subgroup analysis – single versus multiple PRP injections

In patients with early–moderate knee OA, three PRP injections produced larger and more durable gains than a single injection at 12 months.^[19] Mean improvements favored the multiple-dose regimen across all domains – WOMAC total (−18.9 versus −12.4; between-group Δ −6.5, 95% CI −9.4–−3.6; ρ< 0.001), VAS pain (−3.1 vs. −2.0; Δ −1.1, 95% CI −1.6–−0.6; ρ < 0.01), and IKDC (+11.2 vs. +6.8; ρ < 0.05). Clinically, meaningful response was also more frequent with multiple injections, with MCID achievement rates of 72% versus 54% for WOMAC and 69% versus 48% for VAS, as shown in Table 6.

Figure 5 presents the forest plot comparing multiple-dose versus single-dose PRP. The SMDs for WOMAC total, WOMAC pain, WOMAC function, VAS pain, and IKDC all favor multiple-dose PRP, with confidence intervals not crossing zero. The largest benefits are seen for WOMAC total and IKDC, supporting a dose–response relationship and greater durability with three-injection protocols. The largest advantages are seen for WOMAC total and IKDC, supporting a dose–response effect and greater durability of benefit with three-injection protocols.^[21,22]

Close

Figure 5: Forest plot: Multiple versus single platelet-rich plasma (PRP). Standardized mean differences for Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total, WOMAC pain, WOMAC function, visual analog scale (VAS) pain, and International Knee Documentation Committee score (IKDC) all favor multiple-dose PRP, with confidence intervals not crossing zero. The largest advantages are seen for WOMAC total and IKDC, supporting a dose–response effect and greater durability of benefit with three-injection protocols.

Export to PPT