Platelet-Rich Plasma in Regenerative Medicine: Biology, Preparation, Classification, and Clinical Applications
Márcio Hiroaki Kume¹*, Bianca Furlan², Camila Gobatto Boaventura², Mônica Andréa Probst², Edson Peracchi², and Carmen Austrália Paredes Marcondes Ribas3
¹Sugisawa Hospital, Department of Regenerative Medicine, Curitiba, Brazil
²CeUnina, Department of Biologic Science, Curitiba, Brazil
3Mackenzie University, Curitiba, Brazil
*Corresponding author: Márcio Hiroaki Kume, Sugisawa Hospital, Department of Regenerative Medicine, Curitiba, Brazil
Citation: Kume MH, Furlan B, Boaventura CG, Probst MA, Peracchi E, Ribas CAPN. Impact of Adipose-Derived Mesenchymal Stem Cell Isolation Parameters on Cell Yield, Growth Factor Secretion, and Clinical Outcomes in Knee Osteoarthritis: A Randomized Controlled Trial. J Orthop Study Sports Med. 4(1):1-42.
Received: August 11, 2026 | Published: August 28, 2026
Copyright© 2026 Genesis Pub by Kume MH et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0). This license permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are properly credited.
DOI: https://doi.org/10.52793/JOSSM.2026.4(1)-36
Abstract
Background: Platelet-rich plasma (PRP) is the most widely used autologous orthobiologic in the world, yet it remains one of the least standardised. Twenty-five years after Marx defined the platelet-concentration threshold that still anchors the field, PRP is prepared by dozens of incompatible protocols, described by at least six competing classification systems, marketed for more than thirty indications, and reimbursed for exactly one. The result is a literature in which every meta-analysis pools products that are not the same product.\
Objectives: To assemble a single, source-grounded synthesis of PRP that links platelet biology to the physics of centrifugation, the arithmetic of dose, the logic of classification, and the measured clinical effect size in every major indication; and to state explicitly, for each indication, whether the evidence supports use, does not support use, or cannot yet answer the question.
Methods: Narrative review with structured evidence appraisal. Six independent source-grounded evidence briefs were compiled covering platelet biology and growth-factor cargo; preparation, devices and classification;
musculoskeletal clinical outcomes; dermatological, wound, dental and non-orthopaedic outcomes; combination with mesenchymal stromal cells, exosomes and next-generation autologous products; and safety, regulation, reimbursement and anti-doping status. Every numerical value reported in this review was extracted from a retrieved primary source; 226 sources are cited. Effect estimates are compared against published minimal clinically important differences rather than against statistical significance alone.
Results: Platelet α-granules carry more than 300 proteins and PRP releasate exceeds 1,300 identified proteins, but only a minority have been assayed in any clinical trial, and no trial has reported the delivered dose of any single growth factor. Centrifugation parameters that differ by a factor of ten in relative centrifugal force are all described in the literature as producing “PRP”. In knee osteoarthritis, the largest network meta-analysis (53) randomised trials, 5,031 patients) found no optimal platelet dose, no effect of leukocyte content, and no benefit of exogenous activation at six months, while the largest adequately blinded trial (RESTORE, 288 patients) found no structural or symptomatic superiority over saline on its primary endpoint. Effects in lateral elbow tendinopathy, diabetic foot ulceration and venous leg ulceration are more consistent; effects in facial rejuvenation, recurrent implantation failure, erectile dysfunction and periodontal intrabony defects are not supported. All 79 systematic reviews of PRP for knee osteoarthritis contained spin. Pooled complication rates are higher with PRP than with comparators (18.66% versus 9.14%, number needed to harm 11), although severe events are rare; documented harms include septic arthritis, irreversible visual loss after periocular injection, and a five-case HIV cluster linked to unlicensed cosmetic PRP microneedling.
Conclusions: PRP is neither the panacea of its marketing nor the placebo of its harshest critics. It is a family of biologically plausible, procedurally heterogeneous products whose measured benefit is small, indication-specific, and frequently below the threshold a patient would notice. The path forward is not another underpowered trial of an undescribed product; it is mandatory dose-and-composition reporting, blinded comparators, and honest alignment of clinical claims with regulatory reality.
Keywords
Platelet-rich plasma; Regenerative medicine; Growth factors; Mesenchymal stem cells; Exosomes; Tissue repair; orthopedics.
Abbreviations
A2M, alpha-2-macroglobulin; ACD-A, acid citrate dextrose solution A; ACL, anterior cruciate ligament; ACS, autologous conditioned serum; AD-MSC, adipose-derived mesenchymal stromal cell; AGA, androgenetic alopecia; AOFAS, American Orthopaedic Foot and Ankle Society; APS, autologous protein solution; A-PRF, advanced platelet-rich fibrin; BMAC, bone marrow aspirate concentrate; CAGR, compound annual growth rate; CAPE-V, Consensus Auditory-Perceptual Evaluation of Voice; CBER, Center for Biologics Evaluation and Research; CFM, Conselho Federal de Medicina; CGF, concentrated growth factor; CI, confidence interval; COX-2, cyclo-oxygenase-2; CPT, Current Procedural Terminology; DEPA, dose, efficiency, purity, activation; DFU, diabetic foot ulcer; EAU, European Association of Urology; EDQM, European Directorate for the Quality of Medicines and HealthCare; EDTA, ethylenediaminetetraacetic acid; EV, extracellular vesicle; FBS, fetal bovine serum; FDA, US Food and Drug Administration; GAIS, Global Aesthetic Improvement Scale; GMP, good manufacturing practice; GRIIP, Groupe de Recherche sur les Injections Intra-articulaires de PRP; HA, hyaluronic acid; HCT/P, human cells, tissues, and cellular and tissue-based product; hPL, human platelet lysate; ICER, incremental cost-effectiveness ratio; ICRS, International Cartilage Regeneration and Joint Preservation Society; IGF, insulin-like growth factor; IIEF, International Index of Erectile Function; IKDC, International Knee Documentation Committee; IL, interleukin; IL-1Ra, interleukin-1 receptor antagonist; i-PRF, injectable platelet-rich fibrin; ISTH, International Society on Thrombosis and Haemostasis; KL, Kellgren–Lawrence; KOOS, Knee injury and Osteoarthritis Outcome Score; L-PRF, leukocyte- and platelet-rich fibrin; LP-PRP, leukocyte-poor platelet-rich plasma; LR-PRP, leukocyte-rich platelet-rich plasma; MARSPILL, method, activation, red blood cells, spin, platelet number, image guidance, leukocyte, light activation; MASI, Melasma Area and Severity Index; MCID, minimal clinically important difference; MD, mean difference; MF-AT, microfragmented adipose tissue; MIBO, Minimum Information for Studies Evaluating Biologics in Orthopaedics; MMP, matrix metalloproteinase; MRONJ, medication-related osteonecrosis of the jaw; MSC, mesenchymal stromal cell; NMA, network meta-analysis; NNH, number needed to harm; NSAID, non-steroidal anti-inflammatory drug; OA, osteoarthritis; OARSI, Osteoarthritis Research Society International; ONFH, osteonecrosis of the femoral head; OR, odds ratio; OSDI, Ocular Surface Disease Index; PAW, platelets, activation, white cells; PDGF, platelet-derived growth factor; POSAS, Patient and Observer Scar Assessment Scale; PRF, platelet-rich fibrin; PRGF, plasma rich in growth factors; PRP, platelet-rich plasma; PUSH, Pressure Ulcer Scale for Healing; QALY, quality-adjusted life year; RCF, relative centrifugal force; RCT, randomised controlled trial; RR, risk ratio; SMD, standardised mean difference; SoHO, substances of human origin; STEMI, ST-elevation myocardial infarction; TGF-β, transforming growth factor beta; TGA, Therapeutic Goods Administration; T-PRF, titanium-prepared platelet-rich fibrin; VAS, visual analogue scale; VEGF, vascular endothelial growth factor; VHI-10, Voice Handicap Index-10; VISA-A, Victorian Institute of Sport Assessment–Achilles; VLU, venous leg ulcer; WADA, World Anti-Doping Agency; WBC, white blood cell; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index; YAP, Yes-associated protein.
Introduction
Twenty-five years of an unstandardised standard
Platelet-rich plasma occupies a peculiar position in modern medicine. It is simultaneously the most frequently administered autologous biological product in outpatient musculoskeletal practice and one of the least characterised. In Japan, where every regenerative-medicine procedure must be registered under the Act on the Safety of Regenerative Medicine, platelet-rich plasma accounted for 3,041 of 4,621 registered provision plans, or 65.8% of all regenerative medicine delivered in the country, and 74.4% of privately delivered Class III activity [1]. The global market was valued at USD 650.13 million in 2025 with a projected compound annual growth rate of 13.2% to USD 1,751.45 million by 2033 [2]. Yet no platelet-rich plasma device fetched for this review is cleared by the United States Food and Drug Administration for a single intra-articular, tendinous, dermatological, trichological or sexual-medicine indication, and the mandatory labelling for the relevant device class states that the safety and effectiveness of the device for in vivo indications for use has not been established [3].
A short history with a long shadow
The modern era begins with Marx and colleagues, who defined platelet-rich plasma as an autologous concentrate containing approximately 1,000,000 platelets per microlitre in 5 mL of plasma, and who documented that roughly 70% of stored growth factor is released within ten minutes of activation and essentially 100% within one hour [4]. That single arithmetic definition, produced for maxillofacial bone grafting, became the de facto standard for every subsequent clinical application, including applications in which platelet number has never been shown to correlate with outcome. Three consequences followed and have never been undone. First, a number derived from one tissue was exported to all tissues. Second, because the number is a concentration rather than a total dose, protocols delivering wildly different absolute platelet loads could all claim compliance. Third, because concentration is trivially achievable by reducing plasma volume, the definition rewarded arithmetic rather than biology.
Two decades of proliferation followed. Preparation moved from open manual double-spin protocols in blood-bank tubes to closed proprietary systems, gravity-filtration devices, buffy-coat separators and automated apheresis platforms. Nomenclature proliferated in parallel: pure platelet-rich plasma, leukocyte-rich platelet-rich plasma, platelet-rich fibrin, advanced platelet-rich fibrin, injectable platelet-rich fibrin, titanium-prepared platelet-rich fibrin, concentrated growth factors, plasma rich in growth factors, autologous protein solution, autologous conditioned serum and hyperacute serum. Each is a different product with different cellular content, different release kinetics and different regulatory status, and yet the clinical literature routinely pools them.
What this review does differently
This review is built on six independently compiled, source-grounded evidence briefs, and every numerical value reported below was extracted from a retrieved primary document rather than from a secondary summary. Three editorial rules were applied throughout. First, effect estimates are judged against published minimal clinically important differences, not against p values; a statistically significant improvement smaller than the smallest change a patient can perceive is reported as a negative result. Second, where a proprietary product was used, the product, the centrifugation parameters and the delivered platelet dose are reported when the source states them, and their absence is reported when it does not. Third, where the evidence base is contaminated by retraction, spin or undisclosed funding, this is stated in the same paragraph as the effect estimate rather than deferred to a limitations section.
|
Instrument |
Population / setting |
Minimal clinically important difference |
Source anchor |
|
WOMAC total (0–100 normalised) |
Knee osteoarthritis, intra-articular injection |
≈12 points |
Exercise-controlled randomised trial pre-specification [5] |
|
WOMAC pain (0–10 numeric) |
Knee osteoarthritis |
≈2 points |
Exercise-controlled randomised trial pre-specification [5] |
|
KOOS Activities of Daily Living |
Knee osteoarthritis |
9.2 points |
Network meta-analysis of 53 trials [6] |
|
KOOS Quality of Life |
Knee osteoarthritis |
10.3 points |
Network meta-analysis of 53 trials [6] |
|
IKDC subjective score |
Knee osteoarthritis and cartilage lesions |
≈11 points at 6–12 months |
Microfragmented adipose versus PRP trial [7] |
|
IIEF erectile-function domain |
Erectile dysfunction |
≥4 points |
Meta-analysis of 10 studies [8] |
|
Hair density |
Androgenetic alopecia |
No validated threshold; ≈20 hairs/cm² used pragmatically |
Bayesian network meta-analysis [9] |
|
POSAS observer scale |
Post-traumatic and post-burn scars |
No validated threshold reported |
Randomised laser trial [10] |
|
Complete wound closure |
Diabetic foot and venous leg ulcers |
Binary endpoint; no MCID required |
Cochrane review [11] |
Table 1: Interpretive thresholds used throughout this review. Where an indication has no validated minimal clinically important difference, that absence is itself a reportable limitation of the evidence base.
Figure 1: The platelet-rich plasma evidence landscape. Each cell positions an indication by the maturity of its randomised evidence (horizontal) against the consistency and magnitude of measured benefit relative to the minimal clinically important difference (vertical). Indications cluster in the lower-left quadrant — abundant enthusiasm, immature evidence — far more often than in the upper-right.
Molecular Mechanisms: What Is Actually In The Syringe
The therapeutic rationale for platelet-rich plasma rests on a single premise: that concentrating the platelet, and with it the platelet secretome, delivers a supraphysiological pulse of mitogenic, angiogenic, chemotactic and matrix-modulating signals to a tissue whose intrinsic repair capacity is insufficient. The premise is biologically sound. The problem is that the secretome is far larger, far more variable, and far more context-dependent than the five or six growth factors that appear in most clinical papers.
The platelet as a secretory organelle
A resting platelet contains 50 to 80 α-granules of 200 to 500 nm diameter, occupying approximately 10% of platelet volume and carrying more than 300 distinct proteins [12]. Proteomic interrogation of the activated product raises this figure substantially: platelet releasate contains more than 1,300 identified proteins and platelet lysate more than 3,800, reflecting the additional cytoplasmic and membrane content liberated by freeze-thaw disruption [13]. Dense granules contribute adenosine diphosphate, adenosine triphosphate, serotonin, calcium and polyphosphates; lysosomes contribute acid hydrolases and cathepsins. This molecular inventory means that the clinical variable is not “growth factor concentration” but the entire composition of a complex secretome whose proportions shift with anticoagulant, centrifugal force, leukocyte carry-over, erythrocyte contamination, activation method and time to injection.
Figure 2: Platelet ultrastructure and the granule-derived secretome. Alpha granules dominate protein cargo; dense granules supply small-molecule mediators; lysosomal and membrane-derived components add proteolytic and vesicular signalling. Protein counts shown are those reported for releasate and lysate proteomes rather than for intact platelets.
Growth-factor concentrations: what has actually been measured
Direct measurement of growth-factor content in platelet concentrates is surprisingly rare and surprisingly discordant. Kobayashi and colleagues assayed cumulative ten-day release from platelet-rich plasma, leukocyte-platelet-rich fibrin and advanced platelet-rich fibrin prepared from the same donors, and reported PDGF-AA of 6,176, 9,262 and 11,048 pg/mL respectively; TGF-β1 of 1,105, 1,110 and 1,589 pg/mL; VEGF of 847, 732 and 847 pg/mL; EGF of 363, 512 and 659 pg/mL; and IGF of 54, 166 and 129 pg/mL [18]. Amable and colleagues normalised to platelet number and found TGF-β1 of 318.6 ± 118.4 pg per 10⁶ platelets, while IGF-1 at 105.30 ± 48.44 ng/mL did not differ from plasma (P = 0.9734), confirming that insulin-like growth factor 1 in platelet-rich plasma is overwhelmingly plasma-derived rather than platelet-derived [19]. This single observation invalidates a common marketing claim, namely that platelet concentration raises IGF-1 delivery.
Figure 3: Measured growth-factor content across platelet products. Values are those reported by the source studies; note the different normalisations used (concentration in the final product versus content per 10⁶ platelets) and the near-identity of IGF-1 between platelet-rich plasma and plasma.
|
Growth factor |
Principal source |
Reported quantity |
Established biological role |
|
PDGF-AA / -AB / -BB |
Platelet α-granule |
6,176 pg/mL (PRP), 9,262 (L-PRF), 11,048 (A-PRF) over 10 days [18] |
Mitogen and chemoattractant for mesenchymal cells; pericyte recruitment |
|
TGF-β1 |
Platelet α-granule |
1,105 / 1,110 / 1,589 pg/mL [18]; 318.6 ± 118.4 pg per 10⁶ platelets [19] |
Matrix synthesis; chondrogenesis; also profibrotic and osteophyte-associated |
|
VEGF |
Platelet and leukocyte |
847 / 732 / 847 pg/mL [18] |
Angiogenesis and vascular permeability |
|
EGF |
Platelet α-granule |
363 / 512 / 659 pg/mL [18] |
Epithelial and fibroblast proliferation |
|
IGF-1 |
Plasma, not platelet |
105.30 ± 48.44 ng/mL, no difference from plasma (P = 0.9734) [19] |
Anabolic; myogenesis |
|
bFGF / FGF-2 |
Platelet and matrix-bound |
Not quantified in the retrieved comparative assays |
Angiogenesis; fibroblast mitogenesis |
|
HGF |
Platelet α-granule |
Not quantified in the retrieved comparative assays |
Anti-fibrotic; anti-inflammatory via NF-κB suppression |
|
IL-1β (catabolic) |
Leukocyte, chiefly neutrophil |
3.67 (leukocyte-rich) versus 0.31 (leukocyte-poor) [21] |
Pro-inflammatory; matrix degradation |
|
MMP-9 (catabolic) |
Neutrophil |
222 versus 40 [21] |
Matrix degradation |
Table 3: Growth factors and counter-regulatory mediators in platelet concentrates. The final column is clinically consequential: platelet-rich plasma itself is not prohibited in sport, but several of the growth factors it contains are prohibited when administered in isolated or recombinant form.
Cellular content: the leukocyte and erythrocyte question
The most consequential compositional variable is not platelet number but leukocyte content. Sundman and colleagues compared leukocyte-rich and leukocyte-poor preparations from the same donors and found neutrophil concentrations of 8,455 versus 109 per millilitre, interleukin-1β of 3.67 versus 0.31, matrix metalloproteinase-9 of 222 versus 40, and TGF-β1 of 89 versus 20 ng/mL [21]. Leukocyte-rich preparations therefore deliver both more anabolic and more catabolic signal simultaneously, which is why the leukocyte debate cannot be resolved by asserting that white cells are “inflammatory” and therefore undesirable. Erythrocyte contamination is less ambiguous: haemolysis releases free haemoglobin and iron, and red-cell contamination has been associated with synoviocyte death and with the pain reported after injection [22].
The clinical translation of this in-vitro dichotomy is weaker than expected. In the largest network meta-analysis available, leukocyte content was not an effect modifier for any clinical outcome in knee osteoarthritis [6]. However, in the largest dedicated safety meta-analysis, the excess of adverse events over hyaluronic acid was significant only for leukocyte-rich preparations [23]. The defensible synthesis is therefore that leukocyte depletion has not been shown to improve efficacy but has been shown to reduce short-term reactogenicity.
Figure 4: Cellular composition across preparation categories and its measured consequences. Leukocyte-rich preparations deliver more of both anabolic and catabolic mediators; the clinical signal from this difference appears in reactogenicity rather than in efficacy.
Fibrin architecture, activation and the release curve
Activation determines not only how much cargo is released but over what interval. Exogenous activation with calcium chloride or thrombin produces a rapid burst; in-situ activation by contact with collagen and tissue thromboplastin produces a slower and more sustained release; and polymerised fibrin scaffolds such as platelet-rich fibrin release over days rather than minutes [24]. The fibrin matrix is not inert packaging: fibre thickness, branch-point density and clot stiffness govern cell migration into the construct [25]. This is the mechanistic basis for the platelet-rich fibrin family described in Section 5, and for the clinically important observation that the network meta-analysis found exogenous activation to have no effect at six months but a significant effect on twelve-month KOOS Activities of Daily Living, Sport and Quality of Life subscales [6].
Classification Systems: Six Answers To One Question
Classification exists to make trials comparable. In platelet-rich plasma it has achieved the opposite, because at least six systems are in simultaneous use [26], none has been adopted by a majority of authors, and several classify on axes that the clinical literature does not report.
Dohan Ehrenfest 2009: the architectural dichotomy
The first widely adopted framework divided platelet concentrates on two axes, leukocyte content and fibrin architecture, generating four families: pure platelet-rich plasma, leukocyte- and platelet-rich plasma, pure platelet-rich fibrin, and leukocyte- and platelet-rich fibrin [27]. Its enduring value is conceptual: it recognised that a liquid injectable and a polymerised solid membrane are different pharmaceutical dosage forms and cannot be pooled. Its limitation is that it is categorical rather than quantitative and therefore permits a ten-fold difference in delivered platelet dose within a single category.
PAW 2012: platelets, activation, white cells
DeLong and colleagues introduced the first quantitative system, grading absolute platelet concentration, whether exogenous activation was used, and the presence or absence of white cells [28]. PAW is the ancestor of every subsequent dose-aware classification and the first to make explicit that activation is a protocol decision rather than an inherent property.
MARSPILL, depa, mautner and the isth proposal
MARSPILL extended the axes to method, activation, red blood cells, spin, platelet number, image guidance, leukocyte content and light activation [29], and formalised the observation that platelet concentrations above roughly 1.5 million per microlitre may inhibit rather than stimulate stem-cell proliferation [30]. The DEPA classification shifted the emphasis decisively from concentration to absolute dose, purity and activation; when applied retrospectively to twenty commercial devices it revealed absolute platelet doses ranging from 0.21 to 5.43 billion and platelet recovery efficiencies from 13.1% to 79.3%, with no device exceeding 90% recovery [31]. Mautner and colleagues proposed a pragmatic clinical coding system for use in trial reporting [32], and the Subcommittee on Platelet Physiology of the International Society on Thrombosis and Haemostasis issued formal guidance advocating standardised terminology and platelet-dose reporting [33]. Each proposal is defensible; their coexistence is not.
|
System |
Year |
Classifying axes |
Principal strength |
|
Dohan Ehrenfest |
2009 |
Leukocyte content; fibrin architecture |
Separates liquid injectables from solid fibrin membranes |
|
PAW (DeLong) |
2012 |
Platelet concentration; activation; white cells |
First quantitative platelet grading |
|
Mautner et al. |
2015 |
Platelet count, leukocytes, red cells, activation |
Designed for trial reporting |
|
DEPA (Magalon) |
2016 |
Dose of injected platelets; efficiency of production; purity; activation |
Shifts focus from concentration to absolute dose |
|
MARSPILL (Lana) |
2017 |
Method, activation, red cells, spin, platelet number, image guidance, leukocytes, light activation |
Most complete; captures image guidance and photoactivation |
|
ISTH SSC guidance |
2018 |
Terminology and minimum characterisation |
Authority of a haemostasis standards body |
|
Mishra threshold |
as applied 2026 |
Platelet enrichment ≥ 5× baseline; leukocyte presence |
Applicable to 55 of 57 arms (96.5%) in a network meta-analysis |
Table 4: The six coexisting classification systems and a threshold-based scheme applied in the most recent network meta-analysis. No system classifies on an axis that has been shown to predict clinical outcome.
Figure 5: The classification landscape. Each system is plotted by the number of axes it grades against the proportion of the published literature that reports enough information for the system to be applied. Completeness and applicability are inversely related.
Why classification has failed: the reporting audit
The reason no classification has succeeded is that the underlying trials do not report the variables. Chahla and colleagues audited the clinical orthopaedic literature and found that only 10% of studies provided a reproducible preparation protocol, 16% reported the composition of the final product and 20% reported whether a second centrifugation was used [34]. Murray and colleagues responded with the Minimum Information for Studies Evaluating Biologics in Orthopaedics checklist, comprising twenty-three reporting items [35]. Adherence remains below 55% [36], and subsequent audits confirm persistent non-reporting of platelet dose, leukocyte content and red-cell contamination [37][38][39]. The single most striking demonstration of this failure comes from aesthetic dermatology, where an overview of thirteen systematic reviews covering twenty-eight primary studies found that the final platelet number was never reported in any primary study [40].
|
Audit |
Scope |
Key reporting deficiency |
Implication |
|
Chahla et al. 2017 |
Clinical orthopaedic PRP literature |
Reproducible protocol in 10%; composition in 16%; second-spin status in 20% [34] |
Most published PRP cannot be replicated |
|
MIBO checklist |
23 items proposed for PRP and MSC studies |
Consensus standard, not journal-mandated [35] |
Compliance is voluntary and therefore low |
|
MacElroy et al. |
Adherence audit of the PRP literature |
MIBO adherence below 55% [36] |
Nearly half of required items still missing after a decade |
|
Subsequent audits |
Randomised trials and composition reporting |
Persistent non-reporting of dose, leukocytes and red cells [37][38][39] |
No improvement trend demonstrable |
|
Cruciani et al. 2024 |
13 systematic reviews, 28 primary studies, facial rejuvenation |
Final platelet number never reported in any primary study [40] |
An entire indication rests on undescribed products |
|
Richardson et al. 2025 |
All 79 systematic reviews of PRP for knee osteoarthritis |
Spin in 100%; mean spin score 5.5 ± 2.3; benefit claimed despite high risk of bias in 84.8% [41] |
Secondary literature systematically overstates benefit |
|
Chou et al. 2023 |
87 randomised trials |
Undisclosed funding status predicted a positive result (odds ratio 3.61, 95% CI 1.1–11.9, p = 0.035) [42] |
Non-disclosure, not industry funding itself, tracks positivity |
Table 5: Reporting-quality and evidence-integrity audits of the platelet-rich plasma literature. These findings constrain how confidently any pooled effect estimate in this review can be interpreted.
Figure 6: Reporting completeness across the platelet-rich plasma literature. Bars show the proportion of studies reporting each item; the dashed reference line marks full compliance with the Minimum Information for Studies Evaluating Biologics in Orthopaedics checklist.
Preparation: the physics that determines the product
Every compositional property of platelet-rich plasma is set by decisions made before the needle enters the patient. Those decisions are governed by sedimentation physics, and the governing quantity is relative centrifugal force rather than rotational speed. The conversion is g = 1.118 × 10⁻⁵ × R × rpm², where R is the rotor radius in centimetres [43]. Because R differs between devices, two laboratories following the same revolutions-per-minute instruction can generate substantially different forces. This alone explains a portion of the irreproducibility documented in Section 3.
Single-spin versus double-spin
A single centrifugation separates whole blood into a plasma layer, a buffy coat and a red-cell fraction; platelet-rich plasma is drawn from the plasma layer immediately above the buffy coat. A second centrifugation of that plasma pellets the platelets, allowing supernatant plasma to be discarded and a higher concentration to be resuspended in a smaller volume. Saqlain and colleagues quantified the difference directly: single spin produced 594.6 × 10³ platelets per microlitre against 923.06 × 10³ for double spin (p < 0.01), with white-cell counts of 6.056 versus 1.06 × 10³ per microlitre [44]. Double spin therefore raises concentration and lowers leukocyte content, at the cost of an additional handling step, greater platelet activation and lower recovery.
The clinical consequence of this laboratory difference is, so far, undetectable. In androgenetic alopecia, a meta-analysis of three randomised trials and ninety participants found no difference in delivered platelets (mean difference 66.14 × 10⁹/L, p = 0.77) and no difference in hair density (4.10%, 95% CI −4.74 to 12.93, p = 0.36, I² = 0%) [45], a finding corroborated by a separate randomised comparison [46].
Force and time: where the optima actually lie
Piao and colleagues modelled and then validated cell-recovery rates across centrifugation conditions, identifying 700 to 800 g for four to five minutes as the practical optimum for recovery, 240 g for fifteen minutes as the condition of maximal platelet integrity, and a critical force of 350 g at twelve minutes [47]. Perez and colleagues approached the same question through activation markers and found that soluble P-selectin rises at 800 and 1,200 g, while 70 to 100 g gave the best platelet recovery with the least activation [48]. These two conclusions are not contradictory: they optimise different endpoints. High force maximises yield at the cost of premature activation and granule loss; low force preserves integrity at the cost of yield. No clinical trial has ever randomised patients between these two philosophies.
Figure 7: Centrifugation parameter space. Published protocols are plotted by relative centrifugal force against duration, with the recovery optimum, the integrity optimum and the activation threshold marked. Protocols separated by an order of magnitude in force are all described in the literature as producing platelet-rich plasma.
|
Protocol / parameter |
Reported condition |
Measured consequence |
Source |
|
Recovery optimum |
700–800 g for 4–5 min |
Highest validated platelet recovery |
[47] |
|
Integrity optimum |
240 g for 15 min |
Maximal platelet integrity |
[47] |
|
Critical force |
350 g at 12 min |
Threshold beyond which recovery modelling changes regime |
[47] |
|
Least-activation condition |
70–100 g |
Best recovery with least activation |
[48] |
|
Activation threshold |
800 g and 1,200 g |
Soluble P-selectin rises |
[48] |
|
Single spin |
Variable |
594.6 × 10³ platelets/µL; WBC 6.056 × 10³/µL |
[44] |
|
Double spin |
Variable |
923.06 × 10³ platelets/µL (p < 0.01); WBC 1.06 × 10³/µL |
[44] |
|
PRGF (Anitua) |
580 g for 8 min, single spin |
Leukocytes 0.15 ± 0.08 × 10³/µL |
[49] |
|
RESTORE trial protocol |
1,500 g for 5 min, single spin, 5 mL |
1.6–5× enrichment, ≈80% platelet recovery |
[50] |
|
Exercise-controlled trial protocol |
830 g for 5 min, 26 mL processed to 6 mL |
3.4× enrichment, 76% recovery |
[5] |
|
Commercial protocol range |
350–2,008 g, 5–21 min |
Kit cost US$50–500 |
[51] |
|
Freeze-dried protocol |
800 g × 5 min then 580 g × 20 min |
4.1-fold enrichment (130.1 ± 41.1 versus 31.5 ± 7.3 × 10⁴/µL) |
[52] |
|
L-PRF |
≈400 g (2,700 rpm) for 12 min |
Solid fibrin membrane; cannot be stored |
[53] |
|
A-PRF |
230 g (1,500 rpm) for 14 min |
Looser fibrin, higher leukocyte retention |
[53] |
|
A-PRF+ |
200 g (1,300 rpm) for 8 min |
Further reduced force |
[53] |
|
i-PRF |
60 g (700 rpm) for 3 min |
Injectable liquid fibrinogen-rich product; transfer within 2 min 30 s |
[53] |
|
T-PRF |
Titanium tubes, 2,800 rpm for 12 min |
Avoids silica activator |
[53] |
|
CGF |
2,400–3,300 rpm variable-speed programme |
Concentrated growth factor fraction |
[53] |
Table 6: Centrifugation parameters reported across the platelet-rich plasma and platelet-rich fibrin literature, with the consequence each condition was shown to produce. Relative centrifugal force spans more than a thirty-fold range.
Platelet dose: the variable that may matter most
If any compositional variable predicts outcome, the strongest candidate is absolute platelet dose rather than concentration. A systematic analysis of dose in randomised trials of intra-articular platelet-rich plasma found that 90% of arms delivering more than 5.5 billion platelets reported improvement, whereas arms delivering 2.3 billion or fewer failed (p < 0.01) [54]. Against this, the largest network meta-analysis of 53 trials and 5,031 patients found no optimal platelet dose and could not identify a dose threshold that predicted response [6]. The competing published thresholds are themselves irreconcilable: one analysis proposes a cut-point below versus at or above 800,000 platelets per microlitre, another proposes bands of fewer than 5 billion, 5 to 10 billion and more than 10 billion total platelets, and a third proposes a three-fold enrichment criterion [6,55].
Complicating matters further, the dose–response relationship is probably not monotonic. Platelet concentrations above approximately 1.5 million per microlitre may inhibit stem-cell proliferation [30], and in culture the optimum for human bone-marrow stromal cells lies near 5% platelet-rich plasma with colony-forming efficiency maximal at 1% to 2%, while 20% to 30% represses proliferation and 40% to 50% causes cell death [56]. Straum has argued that much of the apparent bell-shaped curve is an artefact of heterogeneous methodology rather than a genuine biological ceiling [57], and the honest position is that the shape of the human dose–response curve is unknown.
Figure 8: The platelet dose–response problem. Panel arrangement contrasts the trial-level dose signal, the competing published thresholds, and the in-vitro evidence for an inhibitory ceiling. The three bodies of evidence do not converge on a single recommendation.
|
Preparation category |
Platelets |
Leukocytes |
Erythrocytes |
Catabolic mediators |
Clinical signal |
|
Leukocyte-poor (pure) PRP |
Moderate to high |
8,455 → 109 neutrophils/mL versus leukocyte-rich [21] |
Low |
IL-1β 0.31; MMP-9 40 [21] |
Lower reactogenicity; releasate inhibits inflammatory processes in osteoarthritic chondrocytes [58]; no proven efficacy advantage [6][23] |
|
Leukocyte-rich PRP |
Moderate to high |
High neutrophil load [21] |
Variable |
IL-1β 3.67; MMP-9 222; TGF-β1 89 ng/mL [21] |
Only category with a significant adverse-event excess over hyaluronic acid [23]; inferior to pure PRP for cartilage regeneration in vitro [59] |
|
Red-cell contaminated PRP |
Variable |
Variable |
High |
Free haemoglobin and iron [22] |
Synoviocyte toxicity in vitro [60]; associated post-injection pain [22] |
|
PRGF (plasma rich in growth factors) |
≈2× baseline |
0.15 ± 0.08 × 10³/µL [49] |
Minimal |
Minimal |
Used across ophthalmic and musculoskeletal indications [61] |
|
Autologous protein solution |
Concentrated |
Concentrated |
Low |
IL-1 receptor antagonist ≈30,000 pg/mL versus <250 in plasma [62] |
Primary endpoint not met in a randomised knee trial [63] |
|
Platelet lysate |
Disrupted, no intact platelets |
Variable |
Low |
>3,800 proteins [13] |
Culture supplement, not an injectable [64] |
Table 7: Compositional and clinical profile of the principal preparation categories. The consistent finding is that leukocyte depletion reduces reactogenicity without a demonstrated efficacy penalty or benefit.
Anticoagulant, activation and co-injected anaesthetic
Anticoagulant choice measurably alters platelet function. In a systematic review of technical procedures, acid citrate dextrose solution A preserved platelet function at 310% relative to reference, against 110% for ethylenediaminetetraacetic acid and 100% for sodium citrate [65]. Activation with bovine thrombin, still used to produce platelet gel, carries a documented immunological hazard: the product label carries a boxed warning that it can cause fatal severe bleeding or thrombosis through antibodies against bovine thrombin and factor V that may cross-react with human factor V, with post-treatment seroconversion of 18.4% in one study and 12.7% in another [66][67], and antibody positivity persisting to three years in 15.6% of recipients with more than one hundred reported cases of factor V inhibition [68]. Autologous thrombin or calcium chloride avoids this entirely and should be the default where activation is required.
The co-injected local anaesthetic is a frequently overlooked determinant of product viability. Bupivacaine 0.75% increased reactive oxygen species, dysregulated calcium handling, induced apoptosis, and reduced platelet adhesion and viability, whereas lidocaine 1% and ropivacaine 0.5% behaved essentially like saline and may be used at up to a 1:1 ratio with platelet preparations [69]. Growth-factor release was not assessed in that study, so the recommendation rests on viability rather than on function.
Pre-procedural medication washout is the weakest link in routine practice. Naproxen suppressed PDGF and interleukin-6 in leukocyte-rich platelet-rich plasma, but all factors normalised after a one-week washout; aspirin reduced growth-factor release; and cyclo-oxygenase-2 selective inhibitors did not inhibit platelet activation or growth-factor release at all [70][71]. No washout interval for non-steroidal anti-inflammatory drugs generally, for aspirin, for paracetamol or for corticosteroids has been established by outcome data, and the widely quoted forty-eight-hour figure carries no stated evidence level in its source document [72].
|
Additive or co-intervention |
Evidence |
Practical position |
|
Acid citrate dextrose solution A |
Platelet function preserved at 310% of reference [65] |
Preferred anticoagulant |
|
Sodium citrate |
Function 100% of reference [65] |
Acceptable; inferior to ACD-A |
|
EDTA |
Function 110% of reference but causes platelet swelling [65] |
Not recommended for therapeutic preparations |
|
Heparin |
0.6 IU/mL required to prevent gelling in platelet-lysate manufacture [73] |
Manufacturing use only |
|
Bovine thrombin |
Boxed warning for fatal bleeding or thrombosis; seroconversion 18.4% and 12.7% [66]; antibody persistence 15.6% at three years [68] |
Avoid; use autologous thrombin or calcium chloride |
|
Calcium chloride |
Standard exogenous activator; produces immediate burst release [24] |
Acceptable when activation is desired |
|
Exogenous activation, clinical effect |
No effect at 6 months on WOMAC (p = 0.94), VAS (p = 0.79) or IKDC (p = 0.17); significant at 12 months for KOOS ADL (p ≤ 0.01), Sport (p = 0.042) and QoL (p = 0.030) [6] |
Reasonable but not mandatory |
|
Bupivacaine 0.75% |
Increased reactive oxygen species, apoptosis; reduced adhesion and viability [69] |
Do not mix with platelet preparations |
|
Lidocaine 1% / ropivacaine 0.5% |
Behave like saline apart from mild reactive-oxygen-species rise; usable up to 1:1 [69] |
Acceptable |
|
Naproxen |
PDGF and interleukin-6 suppressed; normalised after one week [70] |
One-week washout has an experimental anchor |
|
Aspirin |
Reduced growth-factor release [70] |
Washout advisable; interval unestablished |
|
COX-2 selective inhibitors |
No inhibition of platelet activation or growth-factor release [70] |
Washout not required |
|
Corticosteroid co-injection |
No primary experimental study on platelet viability retrieved; relative contraindication by consensus only [72] |
Avoid mixing; interval by consensus |
|
Any additive (Brazil) |
Prohibited by Resolução CFM nº 2.464/2026 art. 5 outside approved research [74]; additives also void the ANVISA conventional-therapy classification [75] |
Regulatory prohibition, not merely a preference |
Table 8. Anticoagulants, activators, anaesthetics and pre-procedural medication, with the evidence supporting each practical position.
Devices, cost and image guidance
Fitzpatrick and colleagues processed blood from the same donors through four common commercial kits and obtained platelet concentrations of 412 ± 140 for one system, 964 ± 551 for a second, 1,224 ± 560 for a third and 1,266 ± 831 × 10⁹/L for a fourth [76]. A three-fold difference in delivered product from identical input blood is the single clearest argument for device-level reporting. Published protocols span 350 to 2,008 g and five to twenty-one minutes, with kit costs of US$50 to US$500 [51].
Injection accuracy is a separate and more tractable variable. Ultrasound guidance achieved correct intra-articular placement in 356 of 373 attempts (95.4%) against 268 of 327 (82.0%) for landmark-guided injection [77]. Injection number also matters: a network meta-analysis found three injections optimal, with fifty-two-week WOMAC scores of 61.03, 31.80 and 28.17 for one, two and three injections respectively [78].
Figure 9: Same-donor output across commercial preparation systems, with the reported absolute-dose and recovery-efficiency range from the DEPA device audit. Identical input blood yields a three-fold difference in delivered platelet concentration depending on device.
|
System / kit |
Reported output or specification |
Regulatory note |
|
Autologous Conditioned Plasma |
412 ± 140 × 10⁹ platelets/L, same-donor comparison [76] |
Device clearance covers preparation only [3] |
|
GPS III |
964 ± 551 × 10⁹/L [76]; cleared with specifications of at least 250,000 platelets/µL and pH above 6.2 [79] |
Predecessor K030555 limited to plasma and concentrated platelets for diagnostic tests [80] |
|
SmartPrep2 / SmartPReP2 |
1,224 ± 560 × 10⁹/L [76] |
510(k) K103340 [81] |
|
Magellan |
1,266 ± 831 × 10⁹/L [76] |
Preparation clearance [3] |
|
Angel |
60 mL processed; 7.23-fold platelet enrichment in a randomised knee trial [82] |
Preparation clearance [3] |
|
Regen Lab (RESTORE) |
Single spin 1,500 g × 5 min, 5 mL, 1.6–5× enrichment, ≈80% recovery [50] |
Preparation clearance [3] |
|
T-Lab (exercise-controlled trial) |
26 mL processed to 6 mL, 830 g × 5 min, 3.4-fold, 76% recovery [5] |
Preparation clearance [3] |
|
EmCyte PureBMC |
Marrow and platelet concentrate system [83] |
510(k) K183205 [83] |
|
Autologous protein solution device |
60 mL of blood processed [63] |
Preparation clearance [3] |
|
2024 CBER clearances |
Six devices cleared, all as “Platelet And Plasma Separator For Bone Graft Handling” [84] |
No 2024 clearance covers an intra-articular or aesthetic indication [84] |
|
Device audit across 20 systems |
Absolute dose 0.21–5.43 billion platelets; recovery 13.1–79.3%; no device above 90% [31] |
DEPA framework proposed in response [31] |
|
Protocol and cost range |
350–2,008 g, 5–21 min, kit cost US$50–500 [51] |
— |
Table 9: Commercial preparation systems with reported output and regulatory position. No system in this table is cleared for the clinical indications for which it is most commonly used.
Growth-Factor Release: Kinetics, Scaffolds And The Fibrin Family
The therapeutic window of a platelet concentrate is defined not only by what it contains but by when the contents appear. Marx's foundational description of platelet-rich plasma established both the working definition of one million platelets per microlitre in five millilitres of plasma and the observation that roughly 70% of stored growth factors are released within ten minutes and close to the entire reservoir within the first hour [4]. This burst kinetic is a pharmacological liability as much as an asset: tissue repair unfolds over weeks, while an unmodified liquid injectable delivers its payload in minutes.
Burst versus sustained release
Comparative release studies show that fibrin architecture converts the burst into a gradient. Kobayashi and colleagues measured release from platelet-rich plasma, platelet-rich fibrin and advanced platelet-rich fibrin and found that the liquid preparation released most of its growth-factor content early while the fibrin-based preparations released progressively over ten days, with advanced platelet-rich fibrin giving the highest total ten-day release of platelet-derived growth factor, vascular endothelial growth factor and insulin-like growth factor [18]. Fibrin therefore functions as a provisional scaffold with three simultaneous roles: mechanical matrix, reservoir that retards diffusion, and substrate for integrin-mediated cell recruitment [24]. Photoactivation has been proposed as an alternative route to sustained release [85], and freeze-drying permits storage of a 4.1-fold enriched product without loss of activity [52], but neither modification has been tested against standard preparation in a randomised clinical trial.
What is measurable and what is inferable
Absolute growth-factor concentrations vary with preparation, donor and assay [19,86], and the mechanistic pathways downstream of receptor engagement have been mapped in detail [25]. Yet the inferential chain from a measured growth-factor concentration to a clinical outcome has never been closed in humans. No randomised trial has measured growth-factor content in the injected product and correlated it with the primary endpoint. The closest available evidence is a case series of thirty knees in which baseline platelet-rich plasma immune-cell and protein composition was correlated with change in Knee injury and Osteoarthritis Outcome Score, where nineteen of thirty knees improved and the calprotectin subunit correlated at r = 0.65 with outcome [14]. A single hypothesis-generating series is a thin foundation for a field that routinely markets growth-factor content as its mechanism.
Two further biological properties deserve mention because they are frequently overstated. Platelet concentrates possess antimicrobial activity, but a systematic review concluded that this activity is bacteriostatic rather than bactericidal and inactive against established biofilm [15]. And platelet-rich plasma exerts genuine antinociceptive activity through a peripheral endocannabinoid-related mechanism, since analgesia was abolished by cannabinoid receptor CB1 and CB2 antagonism [87] — which means that an early analgesic response cannot be read as evidence of structural repair.
Figure 10: Growth-factor release kinetics. The liquid preparation releases the majority of its payload within the first hour, whereas fibrin-based preparations sustain release across ten days. The shaded band marks the interval over which tissue repair actually proceeds.
Orthopaedics And Sports Medicine: The Largest And Least Conclusive Evidence Base
Knee osteoarthritis is the most studied indication and the one on which the credibility of the field rests. It is also the indication where the gap between meta-analytic enthusiasm and rigorous trial results is widest. Interpretation requires anchoring to minimal clinically important differences: visual analogue scale 1.37, WOMAC total 6.4 and WOMAC pain 1.5 for knee osteoarthritis [91], VISA-A 12 points for Achilles tendinopathy [92], and visual analogue scale 0.9 with Foot Function Index 7 for plantar fasciitis [93].
The placebo-controlled trials
The RESTORE trial randomised 288 patients with knee osteoarthritis to leukocyte-poor platelet-rich plasma prepared by single spin at 1,500 g for five minutes or to saline placebo, and was negative on both co-primary endpoints of pain and medial tibial cartilage volume [50]. Global improvement at two months favoured platelet-rich plasma, at 48.2% versus 36.2%, giving a risk ratio of 1.37 (95% CI 1.05 to 1.80, p = 0.02), but this was a secondary outcome and did not persist [50]. Secondary reporting identified cartilage thinning in three or more subregions in 17.1% of platelet-rich plasma recipients against 6.8% of placebo recipients (p = 0.02) [94], a signal that has never been replicated but has also never been refuted.
Adding platelet-rich plasma to supervised exercise produced increments of −0.5 for pain (p = 0.69) and −3 for WOMAC (p = 0.97), against minimal clinically important differences of 2 and 12 respectively [5], with a retrospective cohort reaching a comparable conclusion [95]. In ankle osteoarthritis, the PRIMA trial found an American Orthopaedic Foot and Ankle Society score difference of −1 (95% CI −6 to 3) [96]. In chronic Achilles tendinopathy, the landmark randomised trial found no benefit against the twelve-point VISA-A threshold [92], and in acute muscle injury the hazard ratio for return to play was 0.96 with a between-group difference of zero days (95% CI −11 to 11) [97].
Figure 11: Forest plot of pooled effect estimates for intra-articular platelet-rich plasma in knee osteoarthritis alongside the placebo-controlled trials. The vertical reference line marks the minimal clinically important difference rather than the null, which is the comparison that matters clinically.
Why the meta-analyses disagree with the trials
Meta-analyses report large effects. Standardised mean differences of −1.38 against saline and −2.29 against alternative comparators have been published [98], and across eighteen randomised trials and 1,995 patients WOMAC improvements of −8.15 to −15.90 were reported [55]. Yet across thirty-four randomised trials the superiority of platelet-rich plasma over comparators was documented in only a minority of individual studies [99]. The reconciling explanation is methodological: all seventy-nine systematic reviews of platelet-rich plasma for knee osteoarthritis were found to contain spin, with a mean spin score of 5.5 ± 2.3, and 84.8% claimed benefit despite high risk of bias in the underlying trials [41]. Undisclosed funding status predicted a positive result with an odds ratio of 3.61 (95% CI 1.1 to 11.9, p = 0.035) [42]. Five retractions have affected the platelet-rich plasma literature, including a knee meta-analysis retracted in April 2026 and a subacromial injection trial withdrawn in December 2025 [100][101][102][103][104].
The most methodologically careful synthesis, a frequentist network meta-analysis of fifty-six randomised trials and 5,251 participants, with fifty-three trials and 5,031 patients in the connected network, reached three conclusions that constrain practice: exogenous activation was null at six months but favoured at twelve months on several Knee injury and Osteoarthritis Outcome Score subscales; no optimal platelet dose could be identified; and leukocyte content was not an effect modifier [6].
|
Trial or synthesis |
Design and n |
Preparation |
Primary result |
Interpretation against MCID |
|
RESTORE [50] |
Randomised, saline-controlled, n = 288 |
LP-PRP, single spin 1,500 g × 5 min, 5 mL, 1.6–5× |
Negative on pain and medial tibial cartilage volume; 2-month global improvement 48.2% vs 36.2%, RR 1.37 (1.05–1.80), p = .02 |
Below MCID on primary endpoints; cartilage thinning in 17.1% vs 6.8%, p = .02 [94] |
|
Exercise-controlled trial [5] |
Three-arm randomised, n = 84 (NCT04697667) |
830 g × 5 min, 26 mL → 6 mL, 3.4×, 76% recovery |
Increment of adding PRP to exercise: pain –0.5 (p = 0.69); WOMAC –3 (p = 0.97) |
Far below MCIDs of 2 and 12 |
|
Retrospective cohort [95] |
PRP + exercise vs PRP vs exercise |
Not specified |
KOOS and OMERACT-OARSI responder analysis |
Consistent with the randomised finding |
|
Network meta-analysis [6] |
56 RCTs / 5,251 participants; 53 / 5,031 in network |
Mixed; Mishra threshold applicable to 55/57 arms (96.5%) |
Activation null at 6 mo; 12-mo KOOS ADL p ≤ 0.01, Sport p = 0.042, QoL p = 0.030; no optimal dose; leukocytes not an effect modifier |
MCID 9.2 for ADL and 10.3 for QoL not consistently exceeded |
|
Dose analysis [54] |
Systematic analysis of dose across randomised arms |
Stratified by absolute platelet number |
90% of arms above 5.5 billion platelets improved; arms at or below 2.3 billion failed (p < 0.01) |
Strongest available dose signal; contradicted by the network analysis |
|
Bensa et al. [55] |
18 RCTs / 1,995 patients |
Mixed |
WOMAC improvements –8.15, –15.90, –15.32, –14 |
Exceeds MCID of 6.4 but derived from high-risk-of-bias trials [41] |
|
Nie et al. [98] |
Meta-analysis of randomised trials |
Mixed |
SMD –1.38 vs saline; –2.29 vs other comparators |
Implausibly large relative to placebo-controlled data |
|
Filardo et al. [99] |
34 RCTs |
Mixed |
Superiority over comparators shown in a minority of individual trials |
Pooling amplifies a signal absent in most trials |
|
Dulic et al. [82] |
Randomised, n = 175 |
Angel, 60 mL, 7.23-fold enrichment |
WOMAC BMAC 44.34 → 24.34 versus PRP 48.12 → 31.06; BMAC vs PRP p = 0.306 |
No difference between orthobiologics |
|
Anz et al. [105] |
Randomised, n = 90, 2-year follow-up |
PRP versus bone-marrow concentrate |
IKDC P = .909 |
Equivalence, not superiority |
|
MF-AT comparison [7] |
Randomised, n = 118 |
PRP versus microfragmented adipose tissue |
6-month MCID attainment 75.0% vs 34.6%, P = .005 |
PRP inferior to adipose tissue in this trial |
|
Injection number [78] |
Network meta-analysis |
One, two or three injections |
52-week WOMAC 61.03 / 31.80 / 28.17 |
Three injections optimal |
|
Intra-articular plus intraosseous [106] |
5 prospective studies / 112 knees |
Combined IA + IO delivery |
WOMAC mean difference 11.55 (10.30–12.80) |
Exceeds MCID; small non-randomised base |
|
Autologous protein solution [63] |
Double-blind randomised, n = 40 |
APS, 60 mL processed |
12-month WOMAC mean difference –10.4 (–24.4 to 3.6), p = 0.141 |
Primary endpoint not met |
|
Autologous conditioned serum [107] |
Randomised, n = 376 |
Orthokine ACS versus hyaluronic acid and saline |
Superior to hyaluronic acid through 104 weeks |
Positive but predates modern standards |
|
Alpha-2-macroglobulin [108] |
Double-blind randomised, n = 75 enrolled |
A2M vs PRP vs methylprednisolone |
No between-group difference |
No advantage over PRP |
|
Composition biomarkers [14] |
Case series, 30 knees |
Composition profiled per patient |
19/30 improved; calprotectin subunit r = 0.65; effector-memory CD4 r = 0.52 |
Hypothesis-generating only |
Table 11: Randomised trials and syntheses of platelet-rich plasma for knee osteoarthritis, interpreted against minimal clinically important differences rather than against the null hypothesis.
Figure 12: Head-to-head comparisons of platelet-rich plasma against other orthobiologics and against hyaluronic acid in knee osteoarthritis. Where differences reach significance they favour the comparator as often as they favour platelet-rich plasma.
Tendon, muscle, nerve and spine
Outside the knee the pattern repeats: positive pooled estimates in indications with small heterogeneous trials, and null results wherever placebo control and adequate power coincide. In lateral elbow tendinopathy a multicentre controlled trial of 230 patients reported success in 71.5% versus 56.1% [109], but the Cochrane review found a standardised mean difference of 0.26 with evidence graded insufficient [110] and a pooled analysis restricted to placebo-controlled trials found no benefit [111]. In rotator cuff repair augmentation the re-tear risk ratio of 0.70 (95% CI 0.56 to 0.88) [112] moved to 0.91 (95% CI 0.69 to 1.19) after trim-and-fill adjustment for publication bias [113], which is the clearest single illustration of how small-study effects generate apparent efficacy in this field.
Figure 13: Pooled effect estimates across non-knee musculoskeletal indications, with the rotator-cuff estimate shown before and after trim-and-fill adjustment. Heterogeneity, indicated beside each estimate, is extreme in several indications.
|
Indication |
Best available evidence |
Effect estimate |
Assessment |
|
Lateral elbow tendinopathy |
Multicentre controlled trial, n = 230 [109]; Cochrane review [110]; placebo-controlled pooling [111] |
Success 71.5% vs 56.1%; SMD 0.26, evidence insufficient; no benefit over placebo |
Positive open trials, null against placebo |
|
Rotator cuff repair augmentation |
Meta-analysis [112] with trim-and-fill reanalysis [113] |
Re-tear RR 0.70 (0.56–0.88) → 0.91 (0.69–1.19) |
Apparent benefit explained by small-study effects |
|
Chronic Achilles tendinopathy |
Randomised trial [92] |
No benefit against the 12-point VISA-A MCID |
Negative |
|
Acute muscle injury |
Randomised trial [97] |
Return-to-play HR 0.96; difference 0 days (–11 to 11) |
Negative |
|
Plantar fasciitis |
Randomised evidence and MCID definition [93] |
MCIDs of VAS 0.9 and Foot Function Index 7 required |
MCID-anchored interpretation essential |
|
Ankle osteoarthritis |
PRIMA randomised trial [96] |
AOFAS difference –1 (–6 to 3) |
Negative |
|
Osteonecrosis of the femoral head |
Two contradicting meta-analyses [114][115] |
Arthroplasty conversion RR 0.29 (0.16–0.53) versus no significant reduction |
Irreconcilable; direction unresolved |
|
Carpal tunnel syndrome |
Meta-analysis, 191 patients [116] |
Boston questionnaire SMD –2.06 |
Effect size implausible for the sample size |
|
Lumbar radicular pain (epidural) |
Meta-analysis [117] |
Pooled effect –0.09, I² = 96.7% |
Null with uninterpretable heterogeneity |
|
Discogenic low back pain (intradiscal) |
Randomised trial [118] |
21/44 vs 16/45 responders, p = 0.244; one spondylodiscitis |
Negative with a serious infection signal |
|
Glenohumeral osteoarthritis |
Randomised comparison with hyaluronic acid [119] |
No between-group difference |
No advantage |
|
Trapeziometacarpal arthritis |
Randomised trial, n = 33 [121] |
VAS 20 vs 65, p = 0.015 |
Positive but very small |
|
Injection accuracy (all joints) |
Systematic review, 13 studies [77] |
95.4% (356/373) with ultrasound vs 82.0% (268/327) landmark |
Ultrasound guidance is the strongest modifiable variable |
Table 12: Non-knee musculoskeletal indications with the best available evidence and its assessment. Ultrasound guidance is the only variable in this table with a large, consistent and mechanistically plausible effect.
What guidelines conclude
Guideline bodies have converged on caution rather than on prohibition or endorsement. The 2019 American College of Rheumatology and Arthritis Foundation guideline strongly recommends against platelet-rich plasma [122], and the Osteoarthritis Research Society International guidance likewise does not support it [123]. The American Academy of Orthopaedic Surgeons rates the strength of recommendation as Limited [124]. The National Institute for Health and Care Excellence permits use only with special arrangements for clinical governance, consent and audit or research [125]. In contrast, the ESSKA Orthobiologic Initiative regards platelet-rich plasma as a valid option in Kellgren–Lawrence grades 1 to 3 with a course of two to four injections [126], the joint ESSKA–ICRS statement extends this to grades 0 to III in patients up to eighty years of age [127], and French-speaking experts have issued a supportive consensus [128]. The 2026 American Academy of Physical Medicine and Rehabilitation guidance statement issued five evidence-based recommendations accompanied by eleven consensus-based best practices [129] — a ratio that accurately reflects how much of current practice rests on expert opinion. A Delphi exercise found 77.5% strong agreement restricted to Kellgren–Lawrence grade II, against a claimed efficacy of 78% and a mean charge of US$714 per knee [130][131].
Esthetic dermatology and hair restoration
Dermatological use is the fastest-growing application and the least regulated. It is also where the reporting failure documented in Section 3 reaches its extreme: an overview of thirteen systematic reviews covering twenty-eight primary studies of facial rejuvenation found that no primary study reported the final platelet number, and rated twelve of thirteen reviews as low or critically low confidence on AMSTAR-2 [40].
Androgenetic alopecia
Hair restoration carries the most internally consistent evidence in the whole platelet-rich plasma literature. The original placebo-controlled trial reported a total hair-density change of +45.9 hairs per square centimetre in treated scalp against −3.8 in control [132]. Meta-analyses converge: a pooled hair-density mean difference of 27.55 hairs per square centimetre across fourteen studies and 431 patients [133], and 25.09 hairs per square centimetre (95% CI 9.03 to 41) across ten randomised trials and 318 participants [134]. Adding platelet-rich plasma to topical minoxidil produced an increment of 21.81 hairs per square centimetre (95% CI 10.64 to 33.00) at three months [135].
The comparative question is more instructive than the placebo question. A Bayesian network meta-analysis of twenty-seven trials and 1,110 patients found that minoxidil 5% combined with microneedling exceeded platelet-rich plasma monotherapy by 16 hairs per square centimetre (95% CI 2.57 to 28.66) [9]. Platelet-rich plasma therefore works in androgenetic alopecia, but it is not the best-performing option, and its cost per hair gained is far higher than that of a topical agent. In alopecia areata, a half-head placebo- and active-controlled trial reported benefit [136], and single-spin and double-spin preparations performed identically [45][46].
Facial rejuvenation, scars and pigmentation
Here the evidence turns negative as soon as the design tightens. A systematic review of thirty-six studies and 3,172 patients reported broadly favourable results for rejuvenation and scar management [137], but a randomised double-blind split-face trial of platelet-rich plasma versus saline with full-face microneedling in eighteen women found no improvement and a slight worsening of laxity (p ≤ 0.004) and rhytids [138]. Adding platelet-rich plasma to microneedle fractional radiofrequency for melasma produced no significant between-side difference [139], and adding intradermal platelet-rich plasma to fractional carbon dioxide laser for post-burn and post-traumatic scars gave an observer Patient and Observer Scar Assessment Scale p value of 0.793 [10]. Periorbital dark circles have been reviewed across fourteen studies without a controlled synthesis [140]. The pattern is unambiguous: in split-face designs with an adequate energy-device comparator, platelet-rich plasma adds nothing measurable.
Chronic wounds: the strongest clinical case
If platelet-rich plasma has a genuinely established indication, it is the chronic wound. The Cochrane review of autologous platelet-rich plasma for chronic wounds reported a risk ratio for complete healing of 1.19 [11], and subsequent and larger syntheses have strengthened rather than eroded this. In diabetic foot ulcers, twenty-two randomised trials and 1,559 patients yielded a complete-healing risk ratio of 1.42 (95% CI 1.30 to 1.56) with high certainty [141]. Across twenty-nine randomised trials and 2,198 wounds of mixed aetiology the odds ratio for complete healing was 5.32 (95% CI 3.37 to 8.40) [142]. In venous leg ulcers, sixteen randomised trials and 699 patients gave a closure odds ratio of 5.06 (95% CI 2.35 to 10.89) and a recurrence odds ratio of 0.16 (95% CI 0.05 to 0.50) [143]. For pressure injuries the healed-area standardised mean difference was 1.38 square centimetres (p = 0.02) with a Pressure Ulcer Scale for Healing difference of 1.69 (p = 0.01) [144].
A best-evidence summary drawing on three guidelines, five consensus statements and nine systematic reviews graded venous ulcers at level 1a with grade A and diabetic foot ulcers at a comparable level [145]. Allogeneic preparations extend this further, with a pooled odds ratio for complete healing of diabetic foot ulcers of 6.19 (95% CI 2.32 to 16.56), although no quantified immunogenicity data accompany that estimate [146]. The regulatory position matches the evidence: the only United States national coverage determination that permits platelet-rich plasma reimbursement does so for diabetic wounds and for twenty weeks only [147].
Dentistry, maxillofacial surgery and other specialties
Dental and maxillofacial use is where platelet concentrates originated, and it is where the fibrin-based members of the family dominate. The evidence is nonetheless indication-specific rather than uniformly positive. In periodontal intrabony defects, a network meta-analysis of thirty-two randomised trials found that platelet-rich plasma did not improve probing depth or clinical attachment level (p > 0.05) and ranked among the least effective adjuncts [148]. In maxillary sinus augmentation, thirteen quantitative studies covering 369 patients and 621 augmentations produced a bone-formation mean difference of −0.63 mm (p = 0.81) [149]. A four-arm randomised comparison in third-molar extraction sockets is available but small [150].
The exception is medication-related osteonecrosis of the jaw, where complete healing was achieved in 480 of 557 lesions (86.2%) treated with any platelet concentrate across fifty-eight articles [151]. This is an uncontrolled pooled proportion, but it addresses a condition with poor alternatives and a biologically coherent rationale.
Ophthalmology, urology, gynaecology, otolaryngology and neurology
Beyond these three domains, platelet-rich plasma has entered almost every specialty, generally through uncontrolled series. In Sjögren-related dry eye a randomised comparison against autologous serum drops left the Ocular Surface Disease Index and Schirmer I unchanged [152]; neurotrophic keratitis evidence is a fifteen-patient uncontrolled series [61]; and subretinal platelet-rich plasma for large macular hole achieved 93.3% closure versus 90.0% for an inverted internal limiting membrane flap (p = 0.158) [153].
In erectile dysfunction a double-blind placebo-controlled randomised trial reported benefit [154] and a meta-analysis of ten studies and 559 patients found a six-month International Index of Erectile Function difference of 3.22 (95% CI 2.13 to 4.31) [8], yet both the American Urological Association and the European Association of Urology state that platelet-rich plasma should be considered experimental and not offered outside an approved research protocol [155][156]. Direct-to-consumer pricing averages US$1,507 ± 388 with only 9% of offering urologists [157]. Peyronie disease evidence is a thirty-six-man retrospective cohort [158], and in female stress urinary incontinence a sham-controlled trial reported subjective cure of 32% versus 4% (p < 0.01) with objective cure of 0% [159] — one of the cleanest demonstrations in medicine that patient-reported and objective endpoints can diverge completely.
In reproductive medicine, a critical appraisal of thirteen randomised trials and twelve meta-analyses of intrauterine platelet-rich plasma for recurrent implantation failure found all thirteen trials at high risk of bias [160]; thin-endometrium evidence rests on two small randomised trials [161]; and intraovarian use across fourteen studies could not be meta-analysed [162]. In otolaryngology, tympanic membrane perforation closure carries an odds ratio of 3.69 (95% CI 2.02 to 6.74) overall and 2.70 (95% CI 1.27 to 5.76) restricted to randomised trials [163], post-viral olfactory dysfunction evidence is discordant across ten studies [164], and vocal fold scar evidence is a fifteen-adult uncontrolled trial [165]. A randomised trial in painful diabetic polyneuropathy compared platelet-rich plasma with pregabalin over one year [166].
The commercial context explains this breadth. The global platelet-rich plasma market was valued at US$650.13 million in 2025 and is projected to reach US$1,751.45 million by 2033, a compound annual growth rate of 13.2% [2]. Growth of that magnitude in a field with the reporting deficits documented in Section 3 is itself a patient-safety concern, and the United States Federal Trade Commission has acted against unsubstantiated regenerative-medicine advertising [167], as has the United Kingdom Advertising Standards Authority [168].
Figure 14: Non-orthopaedic indications ranked by the strength of the best available evidence. Chronic wounds and androgenetic alopecia separate clearly from the remainder, where uncontrolled series predominate.
|
Indication |
Best evidence |
Effect estimate |
Verdict |
|
Diabetic foot ulcer |
22 RCTs / 1,559 patients [141] |
Complete healing RR 1.42 (1.30–1.56), high certainty |
Established |
|
Venous leg ulcer |
16 RCTs / 699 patients [143][145] |
Closure OR 5.06 (2.35–10.89); recurrence OR 0.16 (0.05–0.50); level 1a grade A |
Established |
|
Chronic wounds, mixed |
29 RCTs / 2,198 wounds [142]; Cochrane [11] |
Complete healing OR 5.32 (3.37–8.40); RR 1.19 |
Established |
|
Pressure injury |
Meta-analysis [144] |
Healed area SMD 1.38 cm² (p = 0.02); PUSH 1.69 (p = 0.01) |
Probable |
|
Allogeneic PRP, diabetic foot ulcer |
Review with pooled estimate [146] |
Complete healing OR 6.19 (2.32–16.56); immunogenicity unquantified |
Promising, unregulated |
|
Androgenetic alopecia |
14 studies / 431 patients [133]; 10 RCTs / 318 [134]; placebo trial [132] |
MD 27.55 and 25.09 hairs/cm²; +45.9 versus –3.8 hairs/cm² |
Effective but not first-line [9] |
|
PRP added to minoxidil |
5 RCTs [135] |
Hair density 21.81 (10.64–33.00) at 3 months |
Additive benefit |
|
Minoxidil + microneedling versus PRP |
27 trials / 1,110 patients, Bayesian NMA [9] |
Comparator exceeds PRP by 16 hairs/cm² (2.57–28.66) |
PRP is not the best option |
|
Alopecia areata |
Half-head placebo- and active-controlled trial [136] |
Benefit reported |
Possible |
|
Facial rejuvenation |
13 reviews / 28 studies [40]; split-face RCT [138] |
Platelet number never reported; no improvement, laxity worsened (p ≤ 0.004) |
Not supported |
|
Melasma |
Split-face RCT with fractional radiofrequency [139] |
No significant between-side difference |
Not supported |
|
Post-burn and post-traumatic scars |
Randomised observer-blinded trial, 60 completers [10] |
Observer POSAS p = 0.793 |
Not supported |
|
Skin rejuvenation and scars, pooled |
36 studies / 3,172 patients [137] |
Broadly favourable, uncontrolled designs predominant |
Uncertain |
Table 13: Non-orthopaedic indications with the best available evidence, effect estimate and verdict. Only chronic wounds reach a high-certainty positive conclusion.
Combination with mesenchymal stromal cells and extracellular vesicles
The most consequential contribution of platelet-rich plasma to regenerative medicine may not be as an injectable at all. As a xeno-free culture supplement it has already displaced fetal bovine serum in cell manufacturing, and as a source of extracellular vesicles it defines a plausible next generation of acellular products.
Human platelet lysate as a culture supplement
A systematic review and meta-analysis comparing human platelet lysate with fetal bovine serum for mesenchymal stromal cell expansion confirmed superior proliferation with preserved immunophenotype and differentiation capacity [169]. The magnitude is large: doubling time at day seven was 40.61 ± 1.11 hours with a four-cycle platelet lysate against 119.70 ± 32.36 hours with fetal bovine serum (p < 0.001), with heparin at 0.6 international units per millilitre required to prevent gelling [73]. A good-manufacturing-practice cell-factory protocol achieved more than twenty population doublings and approximately 250 × 10⁶ mesenchymal stromal cells per batch, and specified a platelet threshold below 1.5 × 10⁶ per microlitre [64]. Because platelet lysate is manufactured from disrupted platelets and contains more than 3,800 identified proteins [13], it is not interchangeable with an injectable platelet-rich plasma releasate, which contains more than 1,300 [13].
Dose-dependent effects on stem cells
The interaction between platelet concentrate and mesenchymal stromal cells is biphasic, and this is the single most important practical fact for clinicians who combine the two products in one syringe. In human bone-marrow mesenchymal stem cell culture, 5% platelet-rich plasma gave the highest cell number, 20% to 30% repressed proliferation, and 40% to 50% caused immediate cell death [56]. Platelet concentrations above roughly 1.5 million per microlitre may inhibit stem-cell proliferation [30][57]. Chondrogenic differentiation of both adipose- and bone-marrow-derived cells was inhibited by platelet-rich plasma even where proliferation was enhanced [170], which means that a preparation optimised for cell expansion may be actively counterproductive for cartilage formation. Given that mesenchymal stromal cells constitute approximately one in ten thousand marrow cells [56], the concentration of platelet product delivered alongside a bone-marrow concentrate is rarely calculated and almost never reported.
Platelet-derived extracellular vesicles
Platelet-rich plasma exosomes measuring 40 to 100 nanometres and expressing CD9, CD63, CD81 and CD41 promoted re-epithelialisation of chronic cutaneous wounds through activation of the Yes-associated protein pathway in a diabetic rat model [171], and exosomes isolated from both platelet-rich plasma and mesenchymal stem cells promoted functional recovery after muscle injury [16]. Multiomic characterisation has begun to distinguish vesicles derived from platelet lysate, fresh platelets and aged platelets, with yields on the order of 9 × 10⁸ ± 2 × 10⁸ particles per microgram of protein [17].
Clinical translation is at the earliest possible stage. The first-in-human phase I randomised, double-blind, placebo-controlled intra-subject trial of allogeneic platelet-derived extracellular vesicles enrolled eleven participants and did not meet its efficacy endpoint, although no serious adverse events occurred [90]. The regulatory position is unambiguous: there are no approved exosome products and the United States Food and Drug Administration has issued a public safety notification documenting multiple serious adverse events [89], reinforced by a warning letter to a company marketing umbilical-cord-derived products under platelet-rich plasma branding [172]. Aesthetic exosome evidence consists of a twelve-week split-face randomised trial of twenty-eight participants [173] and a nine-woman observational series [174].
|
Product or approach |
Evidence base |
Key quantitative finding |
Status |
|
Human platelet lysate for MSC expansion |
Systematic review and meta-analysis, 22 studies [169] |
Superior proliferation; immunophenotype and differentiation preserved |
Established manufacturing practice |
|
Blood-bank standardised platelet lysate |
Process-development study [73] |
Doubling time 40.61 ± 1.11 h versus 119.70 ± 32.36 h for FBS (p < 0.001); heparin 0.6 IU/mL |
Established |
|
GMP cell-factory protocol |
Cell-factory experience [64] |
>20 population doublings; ≈250 × 10⁶ MSC per batch; platelet threshold <1.5 × 10⁶/µL |
Established |
|
PRP dose–response in MSC culture |
Controlled in-vitro study [56] |
5% optimal; 20–30% repressive; 40–50% lethal |
Mechanistically important, clinically ignored |
|
PRP and chondrogenesis |
Controlled in-vitro study [170] |
Chondrogenic differentiation inhibited despite enhanced proliferation |
Cautionary |
|
PRP-derived exosomes, wound model |
Diabetic rat model [171] |
40–100 nm; CD9/CD63/CD81/CD41 positive; YAP activation |
Preclinical |
|
PRP and MSC exosomes, muscle injury |
Animal model [16] |
Functional recovery promoted |
Preclinical |
|
Platelet-derived extracellular vesicles |
Multiomic characterisation [17] |
9 × 10⁸ ± 2 × 10⁸ particles per µg protein |
Characterisation stage |
|
Allogeneic platelet EV (Plexaris / Plexoval II) |
Phase I RCT, n = 11, ACTRN12620000944932 [90] |
No serious adverse events; efficacy endpoint not met |
Phase I |
|
Exosome products generally |
FDA public safety notification [89]; warning letter [172] |
No approved product; multiple serious adverse events documented |
Unapproved and enforced against |
|
Adipose stem-cell exosome solution, aesthetic |
12-week split-face RCT, n = 28 [173]; series, n = 9 [174] |
GAIS significantly higher on the treated side |
Early clinical |
|
Autologous conditioned serum (Orthokine) |
RCT, n = 376 [107] |
Superior to hyaluronic acid through 104 weeks |
Marketed, dated evidence |
|
Autologous protein solution |
Double-blind RCT, n = 40 [63] |
12-month WOMAC MD –10.4 (–24.4 to 3.6), p = 0.141; IL-1Ra ≈30,000 pg/mL [62] |
Primary endpoint not met |
|
Alpha-2-macroglobulin |
Double-blind RCT, n = 75 [108] |
No difference versus PRP or methylprednisolone |
No advantage |
|
GOLDIC |
Multicentre open-label, 65 patients / 106 knees [175] |
Uncontrolled |
Investigational |
|
Hyperacute serum |
Open-label pilot, n = 24 [176] |
Three weekly 3 mL injections, no control group |
Investigational |
|
Photoactivated PRP |
In-vitro release study [85] |
Sustained growth-factor release |
Investigational |
|
Freeze-dried PRP |
Preparation and activity study [52] |
4.1-fold enrichment (130.1 ± 41.1 versus 31.5 ± 7.3 × 10⁴/µL) |
Investigational, enables storage |
|
Allogeneic PRP |
Review with pooled estimate [146] |
Diabetic foot ulcer healing OR 6.19 (2.32–16.56); immunogenicity unquantified |
Promising, unregulated |
|
Phenotype-targeted formulation |
Conceptual framework [131] with composition biomarkers [14] |
Calprotectin subunit r = 0.65 with KOOS change |
Hypothesis |
Table 14: Adjacent and next-generation platelet-derived products, with the evidence base and development status of each. Culture-supplement use is the only application in this table that is genuinely established.
Safety, contraindications and regulation
Platelet-rich plasma is autologous, which is often presented as equivalent to being safe. It is not. Adverse-event reporting in this literature is poor, the complication rate exceeds that of comparators in the largest analyses, and the regulatory frameworks under which it is delivered differ so profoundly between jurisdictions that the same procedure is a registered drug in one country and an unregulated clinical act in another.
Measured adverse-event rates
A systematic review and meta-analysis of twenty-four randomised trials found that platelet-rich plasma injections for knee osteoarthritis carry a higher complication rate than comparators, at 18.66% versus 9.14%, giving a number needed to harm of eleven [177][178]. A separate analysis of thirty-two randomised trials and 1,268 knees found a pooled adverse-event rate of 18.7% with no severe events, and identified an excess only in the leukocyte-rich subgroup [23]. In the foot and ankle, complications occurred in 41.1% versus 33.7% (p < 0.01) [179]. Reporting quality undermines all of these figures: only eleven of twenty-four studies reported adverse reactions at all, and two of those eleven reported them incompletely [180].
Serious events, though rare, are documented. The PATH-2 trial of platelet-rich plasma for acute Achilles tendon rupture recorded one ST-elevation myocardial infarction two and a half hours after injection [181], and a sham-controlled trial in chronic midportion Achilles tendinopathy provides comparable placebo-controlled safety data [182]. A systematic collection of twenty published adverse-event reports identified seven infection-related events and, most seriously, six patients with visual loss, with a calcium-citrate hapten mechanism proposed [183]. Complete patellar tendon rupture has been reported after a series of four injections in a recreational athlete [184], and one case of spondylodiscitis followed intradiscal injection [118]. The most severe documented harm did not arise from platelet-rich plasma itself: five cases of human immunodeficiency virus transmission were traced to unsafe injection practice at a spa offering platelet-rich plasma microneedling [185]. Red blood cell contamination is associated with post-injection pain [22] and synoviocyte toxicity in vitro [60], and bovine thrombin activation carries a boxed warning [66][67][68].
Figure 15: Adverse-event rates for platelet-rich plasma against comparators across the largest available meta-analyses, with the proportion of trials that reported adverse events at all shown alongside. The reporting gap is larger than the event rate difference.
|
Safety domain |
Evidence |
Quantitative finding |
|
Knee complication rate |
24 RCTs, meta-analysis [177] |
18.66% versus 9.14%; number needed to harm 11 |
|
Knee adverse events |
32 RCTs / 1,268 knees [23] |
Pooled rate 18.7%; no severe events; excess confined to leukocyte-rich preparations |
|
Foot and ankle complications |
Systematic review and meta-analysis [179] |
41.1% versus 33.7% (P < .01) |
|
Adverse-event reporting quality |
24 studies [180] |
Only 11/24 reported adverse reactions; 2 of those 11 incompletely |
|
Cardiovascular event |
PATH-2 randomised trial [181] |
One ST-elevation myocardial infarction 2.5 h post-injection |
|
Placebo-controlled safety, tendinopathy |
Sham-controlled RCT [182] |
Comparative safety data in chronic midportion Achilles tendinopathy |
|
Published adverse-event series |
20 reports collected [183] |
Seven infection-related events; six patients with visual loss; calcium-citrate hapten mechanism proposed |
|
Tendon rupture |
Case report [184] |
Complete patellar tendon rupture after four injections |
|
Spinal infection |
Randomised trial [118] |
One case of spondylodiscitis after intradiscal injection |
|
Bloodborne transmission |
Public health investigation [185] |
Five HIV transmissions linked to unsafe injection practice at a spa |
|
Cartilage signal |
RESTORE secondary reporting [94] |
Cartilage thinning in ≥3 subregions in 17.1% versus 6.8% (p = 0.02) |
|
Red-cell contamination |
Composition studies [22][60] |
Post-injection pain; synoviocyte toxicity in vitro |
|
Bovine thrombin activation |
Product label [66][67]; review [68] |
Boxed warning; seroconversion 18.4% and 12.7%; antibody persistence 15.6% at 3 years; >100 factor V inhibition cases |
|
Local anaesthetic co-injection |
In-vitro study [69] |
Bupivacaine 0.75% harmful; lidocaine 1% and ropivacaine 0.5% acceptable at 1:1 |
Table 15: Documented safety findings. The autologous nature of the product does not confer immunity from procedural, formulation or practice-related harm.
Contraindications and procedural rules
Contraindication lists circulate widely and are almost entirely consensus-derived. The International Cellular Medicine Society guidelines present absolute and relative contraindication lists without any stated evidence level [72]. The GRIIP consensus of thirty-one French-speaking experts produced twenty-three recommendations, most graded D, and explicitly concluded that thrombocytopenia above 50,000 per cubic millimetre is not a contraindication [186][128]. The only experimentally anchored washout interval in routine practice is the one-week interval for naproxen [70].
|
Rule or contraindication |
Basis |
Evidence level |
|
Active malignancy |
Consensus contraindication list [72] |
No stated evidence level |
|
Active infection at the injection site |
Consensus [72]; documented infection events [183] |
Consensus, supported by case reports |
|
Haemodynamic instability or sepsis |
Consensus [72] |
No stated evidence level |
|
Thrombocytopenia |
GRIIP consensus [186] |
Above 50,000/mm³ explicitly not a contraindication |
|
Anticoagulant therapy |
Consensus [72][186] |
Relative; no outcome data |
|
Pregnancy and lactation |
Consensus [72] |
No stated evidence level |
|
Naproxen washout |
Controlled study [70][71] |
One week; PDGF and IL-6 normalise — the only anchored interval |
|
Aspirin washout |
Controlled study [70] |
Growth-factor release reduced; interval unestablished |
|
COX-2 selective inhibitors |
Controlled study [70] |
No inhibition; washout not required |
|
Corticosteroid interval |
Consensus only [72] |
No primary experimental study retrieved |
|
Bupivacaine co-injection |
In-vitro study [69] |
Avoid; direct evidence of platelet harm |
|
Bovine thrombin activation |
Boxed warning [66][67] |
Avoid; regulatory warning |
|
Additives of any kind (Brazil) |
Resolução CFM nº 2.464/2026 art. 5 [74]; ANVISA NT 29/2024 [75] |
Legally binding prohibition outside approved research |
|
Ultrasound guidance |
13-study systematic review [77] |
Accuracy 95.4% versus 82.0%; the strongest modifiable variable |
|
Injection number |
Network meta-analysis [78] |
Three injections optimal; 52-week WOMAC 61.03 / 31.80 / 28.17 |
|
Kellgren–Lawrence restriction |
ESSKA [126]; ESSKA–ICRS [127]; Delphi [130] |
Grades 1–3 or 0–III; 77.5% strong agreement limited to grade II |
Table 16: Contraindications and procedural rules with the basis and evidence level of each. Only three entries in this table rest on experimental data.
Regulation: the same procedure, six legal identities
In the United States, platelet-rich plasma is not regulated as a human cell, tissue or cellular and tissue-based product [187][188]; instead the centrifuge is cleared as a class 1 device under product code QBV, subject to a mandatory labelling statement that the safety and effectiveness of the product for any specific therapeutic use has not been established [3]. The period of enforcement discretion for regenerative-medicine products ended on 31 May 2021 [189], and enforcement has followed: a warning letter of 22 April 2026 found a preparation kit cleared only for bone-graft handling to be adulterated and misbranded when promoted for musculoskeletal use [190]. Every device cleared in 2024 was cleared as a platelet and plasma separator for bone-graft handling [84], and the historical predicates were narrower still [80][81][83].
In Europe, the Competent Authorities for Blood and Blood Components recorded that platelet-rich plasma and platelet-rich fibrin fall within different legal frameworks across Member States [191]. Regulation (EU) 2024/1938 on substances of human origin brings autologous substances into scope through Article 2(5), with application from 7 August 2027 [192], the European Directorate for the Quality of Medicines and HealthCare has published a twenty-second edition of the Blood Guide adding relevant chapters [193][194], and the United Kingdom explanatory memorandum lists platelet-rich plasma among the preparations brought into scope [195].
Brazil now has the most specific rules in the world. Resolução CFM nº 2.464 of 2 July 2026, published on 15 July 2026, authorises platelet-rich plasma as an adjuvant procedure for four musculoskeletal conditions only, with additives prohibited outside approved research [74][196]. ANVISA classifies platelet-rich plasma as a conventional-therapy product not subject to sanitary registration provided it acts through its own mechanism without additives [75]. The Federal Nursing Council filed a civil public action on 29 July 2026 challenging articles 8 to 12 of the resolution [197]. Australia regulates platelet-rich plasma under the autologous human cells and tissues framework [198][199], Health Canada states that platelet-rich plasma meets the definition of a drug and that no applications have been received [200], and Japan provides it as Class II and Class III regenerative medicine, accounting for 65.8% of provision plans [1].
Figure 16: Regulatory identity of platelet-rich plasma by jurisdiction, arranged from device-only oversight through blood-product frameworks to full drug classification. The same autologous procedure occupies six distinct legal categories.
|
Rule or contraindication |
Basis |
Evidence level |
|
Active malignancy |
Consensus contraindication list [72] |
No stated evidence level |
|
Active infection at the injection site |
Consensus [72]; documented infection events [183] |
Consensus, supported by case reports |
|
Haemodynamic instability or sepsis |
Consensus [72] |
No stated evidence level |
|
Thrombocytopenia |
GRIIP consensus [186] |
Above 50,000/mm³ explicitly not a contraindication |
|
Anticoagulant therapy |
Consensus [72][186] |
Relative; no outcome data |
|
Pregnancy and lactation |
Consensus [72] |
No stated evidence level |
|
Naproxen washout |
Controlled study [70][71] |
One week; PDGF and IL-6 normalise — the only anchored interval |
|
Aspirin washout |
Controlled study [70] |
Growth-factor release reduced; interval unestablished |
|
COX-2 selective inhibitors |
Controlled study [70] |
No inhibition; washout not required |
|
Corticosteroid interval |
Consensus only [72] |
No primary experimental study retrieved |
|
Bupivacaine co-injection |
In-vitro study [69] |
Avoid; direct evidence of platelet harm |
|
Bovine thrombin activation |
Boxed warning [66][67] |
Avoid; regulatory warning |
|
Additives of any kind (Brazil) |
Resolução CFM nº 2.464/2026 art. 5 [74]; ANVISA NT 29/2024 [75] |
Legally binding prohibition outside approved research |
|
Ultrasound guidance |
13-study systematic review [77] |
Accuracy 95.4% versus 82.0%; the strongest modifiable variable |
|
Injection number |
Network meta-analysis [78] |
Three injections optimal; 52-week WOMAC 61.03 / 31.80 / 28.17 |
|
Kellgren–Lawrence restriction |
ESSKA [126]; ESSKA–ICRS [127]; Delphi [130] |
Grades 1–3 or 0–III; 77.5% strong agreement limited to grade II |
Table 17: Regulatory classification of platelet-rich plasma by jurisdiction. No two major jurisdictions regulate it identically.
Cost, reimbursement and the commercial gradient
Reimbursement follows the evidence closely, which is unusual and instructive. The only United States national coverage determination permits autologous platelet-rich plasma for chronic non-healing diabetic wounds and for twenty weeks [147], while a local coverage determination establishes non-coverage for all musculoskeletal use [204]. Commercial payers classify it as experimental or unproven for all indications [205][206], and Current Procedural Terminology code 0232T carries zero relative value units and is not covered by most payers [207][208]. In Brazil it is not in the mandatory coverage list [209].
Costs consequently fall on patients, and the gradient between price and evidence is steep. United States direct-to-consumer intra-articular platelet-rich plasma has a median price of US$800 with a range of US$350 to US$2,815 [210], the RESTORE trial costed its injections at US$2,032 each [50], a Delphi survey found a mean charge of US$714 ± 144 against a claimed efficacy of 76% ± 11% [130], and urological direct-to-consumer pricing averages US$1,507 ± 388 [157]. Where a health system pays, the arithmetic inverts entirely: an Italian analysis reported a cost of €82.62 per cycle with an incremental cost-effectiveness ratio of €1,524 per quality-adjusted life year [211], and an Iranian analysis reported 7,583 international dollars per quality-adjusted life year [212].
|
Setting |
Figure |
Source |
|
US Medicare national coverage |
Diabetic chronic non-healing wounds only, 20 weeks |
[147] |
|
US Medicare local coverage |
Non-coverage for all musculoskeletal injections |
[204] |
|
Commercial payer, Aetna |
Experimental or unproven for all indications, ≈30 named conditions |
[205] |
|
Commercial payer, UnitedHealthcare |
Unproven and not medically necessary, effective 1 January 2026 |
[206] |
|
Procedure coding |
CPT 0232T carries zero relative value units; not covered by most payers |
[207][208] |
|
Brazil, supplementary health |
Not included in the mandatory coverage list |
[209] |
|
US direct-to-consumer, knee |
Median US$800 (range US$350–2,815) |
[210] |
|
RESTORE trial costing |
US$2,032 per injection |
[50] |
|
Delphi survey of US practice |
US$714 ± 144 per knee; claimed efficacy 76% ± 11% |
[130] |
|
US direct-to-consumer, urology |
US$1,507 ± 388; offered by 9% of urologists |
[157] |
|
Italy, health-system perspective |
€82.62 per cycle; ICER €1,524 per QALY |
[211] |
|
Iran, health-system perspective |
ICER 7,583 international dollars per QALY |
[212] |
|
Preparation kit cost |
US$50–500 depending on system |
[51] |
|
Global market |
US$650.13 million in 2025 to a projected US$1,751.45 million by 2033 (CAGR 13.2%) |
[2] |
|
Advertising enforcement |
Federal Trade Commission action [167]; Advertising Standards Authority ruling upheld |
[168] |
Table 18: Reimbursement position and cost of platelet-rich plasma across settings. Cost-effectiveness is favourable only where the price is set by a health system rather than by a consumer market.
Discussion: what would have to change
Twenty-five years after Marx defined the product [4], platelet-rich plasma occupies an unusual position in medicine: a therapy with coherent biology, an enormous trial literature, a growing market, and an evidence base that cannot support most of the claims made for it. This review has assembled the reasons, and they are structural rather than biological.
The four structural failures
The first failure is definitional. There is no product called platelet-rich plasma; there is a family of preparations spanning a thirty-fold range of centrifugal force [51][47][48], a three-fold range of delivered platelet concentration from identical donor blood [76], and a twenty-six-fold range of absolute platelet dose across commercial devices [31]. Pooling such preparations in a meta-analysis is not a statistical operation but a category error.
The second failure is reporting. Only 10% of clinical studies provide a reproducible protocol [34], adherence to the Minimum Information for Studies Evaluating Biologics in Orthopaedics checklist remains below 55% [35][36][37][38][39], and in an entire aesthetic indication the final platelet number was never reported in any primary study [40]. Composition cannot be an effect modifier in an analysis that does not measure composition.
The third failure is interpretive. All seventy-nine systematic reviews of platelet-rich plasma for knee osteoarthritis contain spin, with 84.8% claiming benefit despite high risk of bias [41], undisclosed funding predicts positivity with an odds ratio of 3.61 [42], five retractions have affected the literature [100][101][102][103][104], trim-and-fill adjustment eliminates the rotator-cuff effect [112][113], and adverse events go unreported in more than half of studies [180]. Meta-analytic conclusions in this field are systematically more favourable than the trials they summarise.
The fourth failure is regulatory and commercial. The same procedure is a class 1 device output in the United States [3], a substance of human origin in the European Union from 2027 [192], a drug in Canada with no authorised product [200], and a four-indication adjuvant act in Brazil [74]. Patients pay a median of US$800 and up to US$2,815 out of pocket [210] for a therapy that commercial payers classify as unproven for all indications [205][206], in a market growing at 13.2% annually [2], with advertising enforcement actions in two jurisdictions [167][168].
Figure 17: The evidence-integrity picture. Spin prevalence, risk-of-bias distribution, funding-disclosure association and adverse-event reporting completeness are shown together, because each independently inflates the apparent efficacy of platelet-rich plasma.
What the current trial pipeline will and will not resolve
Several trials now in progress are designed to answer the compositional questions directly. Two Rizzoli trials randomise leukocyte-rich against leukocyte-poor preparations in epicondylitis and in hip osteoarthritis [213][214], which is the correct design for the leukocyte question that the network meta-analysis could not settle [6]. A Toronto phase 2/3 trial randomises both bone-marrow aspirate and leukocyte-poor platelet-rich plasma against saline [215], providing the placebo control that most orthobiologic comparisons lack. Trials of platelet-rich plasma with hyaluronic acid [216][217] and of a lyophilised preparation [218] address formulation. A San Francisco trial comparing young with old donors [219] addresses the donor-variability question that sex-based compositional differences have already raised [220].
What the pipeline will not resolve is the definitional problem. A 10,000-participant unmasked registry running to 2035 [221] will generate a very large quantity of uncontrolled data across undefined preparations. Unless preparation reporting becomes mandatory at the journal and registry level, the next decade of trials will reproduce the heterogeneity of the las.
|
Registry identifier |
Sponsor |
Design and enrolment |
Question addressed |
|
NCT05517434 [215] |
University Health Network, Toronto |
Phase 2/3, n = 148; BMA versus saline and LP-PRP versus saline |
Placebo-controlled efficacy of two orthobiologics in parallel |
|
NCT06040203 [213] |
Istituto Ortopedico Rizzoli |
n = 240; LR-PRP versus LP-PRP versus placebo in epicondylitis |
Does leukocyte content matter? |
|
NCT05497349 [214] |
Istituto Ortopedico Rizzoli |
n = 230; LR-PRP versus LP-PRP in hip osteoarthritis |
Leukocyte content in a second joint |
|
NCT06685120 [216] |
Istituto Ortopedico Rizzoli |
n = 288, double-blind; PRP plus hyaluronic acid in knee osteoarthritis |
Does combination add benefit? |
|
NCT05727371 [217] |
Multicentre |
n = 280, triple-masked; PRP with cross-linked hyaluronic acid |
Formulation with a carrier matrix |
|
NCT06932614 [218] |
Multicentre |
n = 90, quadruple-masked; lyophilised growth factor versus PRP, WOMAC primary |
Does a storable product match fresh PRP? |
|
NCT06003101 [222] |
Massachusetts General Hospital |
Phase 3, n = 160; PRP, platelet-poor plasma and BMAC in hip arthroscopy |
Is the platelet fraction the active fraction? |
|
NCT05603468 [223] |
Second Affiliated Hospital of Zhejiang University |
Phase 4, n = 387; PRP versus corticosteroid in rotator cuff tendinopathy |
Comparison against standard care at scale |
|
NCT05412381 [224] |
Hospital for Special Surgery |
Phase 3, n = 56; PRP in ACL reconstruction, platelet count as primary outcome |
Links delivered dose to a trial endpoint |
|
NCT05378815 [225] |
Assistance Publique – Hôpitaux de Paris |
n = 210, triple-blind; change in numerical rating scale pain |
Rigorously blinded pain outcome |
|
NCT06451120 [219] |
University of California San Francisco |
Phase 2, n = 60; PRP in young versus old subjects |
Donor-age effect on the product |
|
NCT06680856 [226] |
Ospedale Policlinico San Martino |
Phase 1/2, n = 72, completed; allogeneic PRP for diabetic foot ulcers |
Feasibility of an allogeneic product [146] |
|
NCT07231471 [221] |
University of Utah |
Planned n = 10,000, no masking, primary completion 2035 |
Large-scale observational data; will not resolve heterogeneity |
Table 19: Registered trials in progress and the specific question each address. Three trials directly test the leukocyte and dose questions that current syntheses cannot answer
A research agenda
Five changes would alter the trajectory of this field more than any additional trial of unspecified platelet-rich plasma against saline. First, journals should refuse publication of clinical studies that do not report absolute platelet dose, leukocyte and erythrocyte content, centrifugal force in units of g, spin duration, anticoagulant, activation status and injected volume — the essential subset of the existing checklist [35][33]. Second, trials should power for the minimal clinically important difference rather than for statistical significance [91][92][93]. Third, dose should be randomised rather than described, since the only strong dose signal available derives from cross-arm comparison [54] and is contradicted by network analysis [6]. Fourth, composition biomarkers should be collected prospectively in every trial, following the single existing series that correlated composition with outcome [14][131]. Fifth, ultrasound guidance should be standard, since it is the one variable with a large, reproducible and mechanistically transparent effect [77].
Conclusion
Platelet-rich plasma is neither the regenerative breakthrough its marketing describes nor the placebo its critics assert. It is a heterogeneous family of autologous preparations with a demonstrable and clinically meaningful effect in chronic wounds, a real but second-line effect in androgenetic alopecia, a plausible effect in medication-related osteonecrosis of the jaw and tympanic membrane perforation, an unresolved effect in knee osteoarthritis where the highest-quality placebo-controlled trial was negative, and no demonstrable effect in facial rejuvenation, periodontal intrabony defects, sinus augmentation, ankle osteoarthritis, Achilles tendinopathy, acute muscle injury or recurrent implantation failure.
The four-tier framework in Table 20 separates these levels of confidence. Applying it would mean offering platelet-rich plasma routinely for venous and diabetic ulcers, offering it with explicit discussion of alternatives for androgenetic alopecia and for Kellgren–Lawrence grade II knee osteoarthritis, confining the remaining indications to research, and declining to charge patients for indications where controlled evidence is absent. The single most valuable technical change available today is not a new formulation but complete reporting of the formulation already in use.
Figure 18: The four-tier framework. Indications are placed by the strength of controlled evidence, with the recommended clinical posture for each tier shown alongside.
|
Tier |
Definition |
Indications |
Clinical posture |
|
Tier 1 — Established |
Consistent randomised evidence with high certainty and effects exceeding the MCID |
Diabetic foot ulcer [141]; venous leg ulcer [143][145]; chronic wounds of mixed aetiology [142][11] |
Offer as part of standard wound care; reimbursed for diabetic wounds for 20 weeks [147] |
|
Tier 2 — Probable, not first-line |
Consistent effect demonstrated, but a comparator performs as well or better |
Androgenetic alopecia [133][134][9]; pressure injury [144]; medication-related osteonecrosis of the jaw [151]; tympanic membrane perforation [163]; KL grade II knee osteoarthritis [126][130] |
Offer only after discussing better-evidenced alternatives and the out-of-pocket cost [210] |
|
Tier 3 — Research only |
Trials are small, discordant, or confined to unblinded designs |
Knee osteoarthritis beyond KL II [50][6]; rotator cuff augmentation [113]; osteonecrosis of the femoral head [114][115]; carpal tunnel syndrome [116]; olfactory dysfunction [164]; erectile dysfunction [155][156] |
Enrol in a trial; do not charge for treatment outside research [125] |
|
Tier 4 — Not supported |
Adequately designed controlled trials show no benefit |
Facial rejuvenation [138][40]; melasma [139]; scars with laser comparator [10]; periodontal intrabony defect [148]; sinus augmentation [149]; ankle osteoarthritis [96]; Achilles tendinopathy [92][182]; acute muscle injury [97]; recurrent implantation failure [160]; stress urinary incontinence [159] |
Do not offer |
Table 20: Four-tier framework for the clinical use of platelet-rich plasma. Tier assignment reflects the strength of controlled evidence interpreted against minimal clinically important differences, not the volume of published literature.
Declarations
Funding
No funding.
Conflicts of interest
No competing interests exist.
Ethics statement
This work is a narrative review of published literature and does not involve new human participants, animals, or identifiable patient data. Ethical approval was therefore not required.
Author contributions
All authors critically revised the manuscript and approved the final version.
Data availability
All data discussed in this review are contained in the cited publications, regulatory documents and trial registry records, each of which is referenced with its source location.
Acknowledgements
None
References
- Arita A, Tobita M. Regenerative medicine in Japan under the Act on the Safety of Regenerative Medicine. Regen Ther. 2023;23:37–43.
- Grand View Research. Platelet-rich plasma market size, share and trends analysis report. Global market USD 650.13 million in 2025 rising to a projected USD 1,751.45 million by 2033, a compound annual growth rate of 13.2%.
- US Food and Drug Administration device classification, product code QBV: centrifuge for cell or plasma concentrate, 21 CFR 862.2050, class 1, subject to the mandatory labelling statement that the safety and effectiveness of the device for in vivo indications for use has not been established.
- Marx RE. Platelet-rich plasma (PRP): what is PRP and what is not PRP? Implant Dent. 2001;10(4):225–8.
- Three-arm randomised controlled trial of exercise, platelet-rich plasma and their combination in knee osteoarthritis (NCT04697667), n = 84. Incremental benefit of adding PRP to supervised exercise: pain mean difference –0.5 (95% CI –1.2 to 0.4, p = 0.69) and WOMAC –3 (95% CI –12 to 5, p = 0.97).
- Anitua E, et al. Composition of platelet-rich plasma and clinical outcomes in knee osteoarthritis: a systematic review and frequentist network meta-analysis of 56 randomised controlled trials and 5,251 participants (PROSPERO CRD42025646778).
- Meta-analysis of platelet-rich plasma for erectile dysfunction: 10 studies including 3 randomised trials and 559 patients. Transl Androl Urol. 2024;13(6):970–82.
- Gupta AK, et al. Bayesian network meta-analysis of 27 trials and 1,110 patients with androgenetic alopecia: minoxidil 5% with microneedling exceeded PRP monotherapy by 16 hairs/cm² (95% CI 2.57–28.66); PRP surface under the cumulative ranking curve 34.9%.
- Godara S, et al. Fractional carbon dioxide laser with or without intradermal platelet-rich plasma for post-burn and post-traumatic scars: a randomised observer-blinded trial of 60 completers. Observer POSAS between-group p = 0.793; pain and erythema in 11 of 30 versus 30 of 30.
- Martinez-Zapata MJ, Martí-Carvajal AJ, Solà I, Expósito JA, Bolíbar I, Rodríguez L, et al. Autologous platelet-rich plasma for treating chronic wounds. Cochrane Database Syst Rev. 2016;5:CD006899.
- Golebiewska EM, Poole AW. Platelet secretion: from haemostasis to wound healing and beyond. Blood Rev. 2015;29(3):153–62. Platelets contain 50–80 alpha-granules of 200–500 nm occupying approximately 10% of platelet volume and releasing more than 300 soluble proteins.
- Comparative proteomic characterisation of platelet releasate and platelet lysate: more than 1,300 proteins identified in releasate and more than 3,800 in lysate.
- Case series of 30 knees correlating baseline platelet-rich plasma immune-cell and protein composition with change in KOOS, including effector memory CD4 r = 0.52 and calprotectin subunit r = 0.65.
- Systematic review of the antimicrobial activity of platelet concentrates: activity is bacteriostatic rather than bactericidal and is inactive against established biofilm.
- Iyer SR, Scheiber AL, Yarowsky P, Henn RF 3rd, Otsuru S, Lovering RM. Exosomes isolated from platelet-rich plasma and mesenchymal stem cells promote recovery of function after muscle injury. Am J Sports Med. 2020;48(9):2277–86.
- Amengual-Tugores AM, et al. Multiomic characterisation of platelet-lysate-, fresh-platelet- and aged-platelet-derived extracellular vesicles.
- Kobayashi E, Flückiger L, Fujioka-Kobayashi M, Sawada K, Sculean A, Schaller B, Miron RJ. Comparative release of growth factors from PRP, PRF and advanced PRF. Clin Oral Investig. 2016;20(9):2353–60.
- Amable PR, Carias RBV, Teixeira MVT, da Cruz Pacheco I, Corrêa do Amaral RJF, Granjeiro JM, Borojevic R. Platelet-rich plasma preparation for regenerative medicine: optimization and quantification of cytokines and growth factors. Stem Cell Res Ther. 2013;4(3):67.
- World Anti-Doping Agency. Prohibited List 2026. The term platelet-rich plasma does not appear in the List; section S2.3 prohibits named growth factors including platelet-derived growth factor, insulin-like growth factor 1, mechano growth factors, fibroblast growth factors, hepatocyte growth factor and vascular endothelial growth factor, and section M1 prohibits blood and blood-component manipulation.
- Sundman EA, Cole BJ, Fortier LA. Growth factor and catabolic cytokine concentrations are influenced by the cellular composition of platelet-rich plasma. Am J Sports Med. 2011;39(10):2135–40.
- Review of the effect of erythrocyte contamination in platelet-rich plasma, including eryptosis-mediated cytotoxicity and the absence of any direct clinical comparison of red-cell-rich versus red-cell-poor preparations.
- Nakagawa Y, et al. Adverse events of intra-articular platelet-rich plasma injection for knee osteoarthritis: systematic review and meta-analysis of 32 randomised trials and 1,268 knees. PM R. 2026. Pooled adverse-event rate 18.7%, mild pain or swelling 10.6%, no severe adverse events, and excess over hyaluronic acid confined to leukocyte-rich preparations.
- Wu Y, et al. Molecular mechanisms of platelet-rich plasma in tissue regeneration. Int J Mol Sci. 2025;26(21):10804.
- Everts P, Onishi K, Jayaram P, Lana JF, Mautner K. Platelet-rich plasma: new performance understandings and therapeutic considerations in 2020. Int J Mol Sci. 2020;21(20):7794.
- Dohan Ehrenfest DM, Rasmusson L, Albrektsson T. Classification of platelet concentrates: from pure platelet-rich plasma (P-PRP) to leucocyte- and platelet-rich fibrin (L-PRF). Trends Biotechnol. 2009;27(3):158–67.
- DeLong JM, Russell RP, Mazzocca AD. Platelet-rich plasma: the PAW classification system. Arthroscopy. 2012;28(7):998–1009.
- Lana JFSD, Purita J, Paulus C, Huber SC, Rodrigues BL, Rodrigues AA, et al. MARSPILL classification of platelet-rich plasma.
- Lana JF, Purita J, Paulus C, Huber SC, Rodrigues BL, Rodrigues AA, et al. Contributions for classification of platelet rich plasma – proposal of a new classification: MARSPILL. Regen Med. Includes the observation that platelet concentrations above 1.5 million/µL may inhibit stem-cell proliferation in vitro, that only 11.5% of 105 studies reported all processing variables, and the comparative composition of bone-marrow aspirate concentrate versus PRP.
- Magalon J, Chateau AL, Bertrand B, Louis ML, Silvestre A, Giraudo L, et al. DEPA classification: a proposal for standardising PRP use and a retrospective application of available devices. BMJ Open Sport Exerc Med. 2016;2(1):e000060. Across 20 devices the absolute platelet dose ranged from 0.21 to 5.43 billion and platelet recovery from 13.1% to 79.3%, with no device exceeding 90%.
- Mautner K, Malanga GA, Smith J, Shiple B, Ibrahim V, Sampson S, Bowen JE. A call for a standard classification system for future biologic research: the rationale for new PRP nomenclature. PM R. 2015;7(4 Suppl):S53–9.
- Harrison P; Subcommittee on Platelet Physiology of the International Society on Thrombosis and Haemostasis. The use of platelets in regenerative medicine and proposal for a new classification system: guidance from the SSC of the ISTH. J Thromb Haemost. 2018;16(9):1895–900.
- Chahla J, Cinque ME, Piuzzi NS, Mannava S, Geeslin AG, Murray IR, et al. A call for standardization in platelet-rich plasma preparation protocols and composition reporting: a systematic review of the clinical orthopaedic literature. J Bone Joint Surg Am. 2017;99(20):1769–79.
- Murray IR, Geeslin AG, Goudie EB, Petrigliano FA, LaPrade RF. Minimum information for studies evaluating biologics in orthopaedics (MIBO): platelet-rich plasma and mesenchymal stem cells. J Bone Joint Surg Am. 2017;99(10):809–19.
- MacElroy A, et al. Adherence to the MIBO reporting guideline in the platelet-rich plasma literature remains below 55%.
- Reporting-quality audit of PRP randomised trials against the MIBO checklist.
- Second reporting-quality audit of the PRP literature against the MIBO checklist.
- Third reporting-quality audit of PRP composition reporting in orthopaedic trials.
- Cruciani M, et al. Platelet-rich plasma for facial rejuvenation: an overview of systematic reviews. Blood Transfus. 2024;22(5):429–40. Thirteen systematic reviews, 28 primary studies, AMSTAR-2 confidence low or critically low in 12 of 13; the only poolable outcome was patient satisfaction (mean difference 0.63, 95% CI 0.25–1.00); the final platelet number was never reported in any primary study.
- Richardson MK, et al. Spin is present in all systematic reviews and meta-analyses of platelet-rich plasma for knee osteoarthritis. Eur J Orthop Surg Traumatol. 2025;35(1):348.
- Chou K, Gazendam A, Vemulakonda J, Bhandari M. The influence of industry funding on randomised controlled trials of platelet-rich plasma. Knee Surg Sports Traumatol Arthrosc. 2023;31(2):626–31.
- Dhurat R, Sukesh MS. Principles and methods of preparation of platelet-rich plasma: a review and author's perspective. J Cutan Aesthet Surg. 2014;7(4):189–97.
- Saqlain N, Mazher N, Fateen T, Siddique A. Comparison of single and double centrifugation methods for preparation of platelet-rich plasma. Pak J Med Sci. 2023. Single spin 594.6 × 10³/µL versus double spin 923.06 × 10³/µL platelets (p < 0.01), with white-cell counts of 6.056 versus 1.06 × 10³/µL.
- Meta-analysis of single-spin versus double-spin platelet-rich plasma for androgenetic alopecia: three randomised trials, 90 participants; hair-density mean difference 4.10% (95% CI –4.74 to 12.93, p = 0.36). Front Med. 2025;12:1631087.
- Randomised comparison of single-spin versus double-spin PRP for androgenetic alopecia: no difference in clinical outcome between protocols.
- Piao L, Park H, Jo CH. Theoretical prediction and validation of cell recovery rates in preparing platelet-rich plasma through a centrifugation. PLoS One. 2017;12(11):e0187509.
- Perez AGM, Lana JFSD, Rodrigues AA, Luzo ACM, Belangero WD, Santana MHA. Relevant aspects of centrifugation step in the preparation of platelet-rich plasma. ISRN Hematol. 2014;2014:176060
- Anitua E, Prado R, Orive G. Plasma rich in growth factors: a single-spin protocol at 580 g for 8 minutes yielding leukocyte counts of 0.15 ± 0.08 × 10³/µL.
- Bennell KL, Paterson KL, Metcalf BR, Duong V, Eyles J, Kasza J, et al. Effect of intra-articular platelet-rich plasma vs placebo injection on pain and medial tibial cartilage volume in patients with knee osteoarthritis: the RESTORE randomized clinical trial. JAMA. 2021;326(20):2021–30.
- Oudelaar BW, Peerbooms JC, Huis in 't Veld R, Vochteloo AJH. Concentrations of blood components in commercial platelet-rich plasma separation systems: a review of the literature. Am J Sports Med. 2019;47(2):479–87. Protocols span 350 to 2,008 g and 5 to 21 minutes, with kit costs of US$50 to US$500.
- Shiga Y, Kubota G, Orita S, Inage K, Kamoda H, Yamashita M, et al. Freeze-dried platelet-rich plasma maintains activation and growth-factor content: a laboratory comparison of storage conditions over eight weeks.
- Farshidfar N, Miron RJ, et al. Systematic review of injectable platelet-rich fibrin versus platelet-rich plasma: 23 studies, 72% favouring i-PRF and 24% showing no difference.
- Systematic analysis of platelet dose in randomised trials of intra-articular PRP: 90% of arms delivering more than 5.5 billion platelets reported improvement whereas arms delivering 2.3 billion or fewer failed (p < 0.01).
- Bensa A, Sangiorgio A, Deabate L, Illuminati A, Pompa B, Filardo G. Corrigendum and analysis of platelet-rich plasma injections for knee osteoarthritis: 18 randomised controlled trials and 1,995 patients. WOMAC improvements of –8.15, –15.90, –15.32 and –14.69 at 1, 3, 6 and 12 months.
- Nguyen PD, Pham PV. Human platelet lysate and platelet-rich plasma dose–response in human bone-marrow mesenchymal stem cell culture: 5% gave the highest cell number, and 40–50% caused immediate cell death.
- Straum OK. The optimal platelet concentration in platelet-rich plasma for proliferation of human cells in vitro – diversity, biases, and possible basic experimental principles for further research in the field: a review. PeerJ. 2020;8:e10303. Argues that the reported bell-shaped dose–response is largely a methodological artefact.
- van Buul GM, Koevoet WLM, Kops N, Bos PK, Verhaar JAN, Weinans H, et al. Platelet-rich plasma releasate inhibits inflammatory processes in osteoarthritic chondrocytes. Am J Sports Med. 2011;39(11):2362–70.
- Xu Z, Yin W, Zhang Y, Qi X, Zhang Y, Cao J, Zhang JY. Comparative evaluation of leukocyte- and platelet-rich plasma and pure platelet-rich plasma for cartilage regeneration. Sci Rep. 2017;7:43301.
- Braun HJ, Kim HJ, Chu CR, Dragoo JL. The effect of platelet-rich plasma formulations and blood products on human synoviocytes: implications for intra-articular injury and therapy. Am J Sports Med. 2014;42(5):1204–10.
- Retrospective series of plasma rich in growth factors (PRGF-Endoret) eye drops in 15 patients and 26 eyes with stage 1 neurotrophic keratitis, without a control group.
- The concentration of plasma provides additional bioactive proteins: interleukin-1 receptor antagonist reaches approximately 30,000 pg/mL in autologous protein solution versus less than 250 pg/mL in plasma.
- Ross M, et al. Autologous protein solution versus saline for knee osteoarthritis: a double-blind randomised controlled trial. Bone Joint J. 2024. Twelve-month WOMAC mean difference –10.4 (95% CI –24.4 to 3.6, p = 0.141), with worst pain favouring saline (mean difference 21.5, 95% CI 6.2 to 36.8, p = 0.008).
- Becherucci V, Piccini L, Casamassima S, Bisin S, Gori V, Gentile F, et al. Human platelet lysate in mesenchymal stromal cell expansion according to a GMP grade protocol: a cell factory experience. Stem Cell Res Ther. 2018;9(1):124.
- Pachito DV, et al. Technical procedures for preparation and administration of platelet-rich plasma: a systematic review. Front Cell Dev Biol. 2020;8:598816.
- THROMBIN-JMI (thrombin, topical, bovine, USP) full prescribing information. Boxed warning: can cause fatal severe bleeding or thrombosis through antibodies against bovine thrombin and factor V that may cross-react with human factor V. Post-treatment seroconversion 18.4% in one study and 12.7% in another.
- US Food and Drug Administration. Bovine thrombin, topical – prescribing information with boxed warning regarding antibody formation and coagulopathy.
- Lomax C, Traub O. Topical thrombins: benefits and risks. Pharmacotherapy. 2009;29(7 Pt 2):8S–12S.
- Dregalla RC, Uribe Y, Bodor M. Effect of local anesthetics on platelet physiology and function. J Orthop Res. 2021;39(12):2744–54.
- Gupta A, Jeyaraman M, Maffulli N. Common medications which should be stopped prior to platelet-rich plasma injection. Biomedicines. 2022;10(9):2134
- Effect of non-steroidal anti-inflammatory drugs and aspirin on platelet function and on the growth-factor content of platelet-rich plasma.
- International Cellular Medicine Society. Guidelines for the use of platelet rich plasma, section 10. Absolute and relative contraindication lists are provided without any stated evidence level.
- Bianchetti A, Chinello C, Guindani M, Braga S, Neva A, Verardi R, et al. A blood bank standardized production of human platelet lysate for mesenchymal stromal cell expansion. Doubling time at day 7: 40.61 ± 1.11 h for a four-cycle lysate versus 119.70 ± 32.36 h for fetal bovine serum (p < 0.001).
- Conselho Federal de Medicina (Brazil). Resolução CFM nº 2.464, de 2 de julho de 2026, published in the Diário Oficial da União on 15 July 2026. Authorises platelet-rich plasma as an adjuvant procedure only for knee osteoarthritis, lumbar discopathy, lateral elbow epicondylitis and meniscal repair; prohibits additives and promises of cure; revokes Resolução CFM nº 2.128/2015.
- Agência Nacional de Vigilância Sanitária (Brazil). Nota Técnica nº 29/2024-GSTCO. Platelet-rich plasma is a conventional-therapy product not subject to sanitary registration provided it acts through the original functions of platelets and undergoes only minimal manipulation; production is permitted within the same surgical act, in a closed system, or exceptionally in a validated open system.
- Fitzpatrick J, Bulsara MK, McCrory PR, Richardson MD, Zheng MH. Analysis of platelet-rich plasma extraction: variations in platelet and blood components between 4 common commercial kits. Orthop J Sports Med. 2017;5(1):2325967116675272.
- Systematic review of injection accuracy across 13 studies: 95.4% (356/373) with ultrasound guidance versus 82.0% (268/327) with a landmark technique.
- Zhuang W, et al. Optimal number of intra-articular platelet-rich plasma injections for knee osteoarthritis: a network meta-analysis. Three injections were optimal, with 52-week WOMAC scores of 61.03, 31.80 and 28.17 for one, two and three injections respectively.
- US Food and Drug Administration. BK230844 summary – Gravitational Platelet Separation (GPS III) system, cleared with specifications of at least 250,000 platelets/µL and pH above 6.2.
- US Food and Drug Administration 510(k) summary K030555, Biomet Gravitational Platelet Separation system, cleared for producing plasma and concentrated platelets that can be used for diagnostic tests.
- US Food and Drug Administration 510(k) summary K103340, SmartPReP2 platelet concentrate system.
- Dulić O, Rasović P, Lalić I, Keković V, Gavrilović D, Abazović D, et al. Bone marrow aspirate concentrate versus platelet-rich plasma or hyaluronic acid for the treatment of knee osteoarthritis. Medicina (Kaunas). 2021;57(11):1193.
- US Food and Drug Administration 510(k) summary K183205, EmCyte PureBMC.
- US Food and Drug Administration, Center for Biologics Evaluation and Research. 2024 biological device application approvals. All platelet-rich plasma devices cleared in 2024 were cleared as platelet and plasma separators for bone graft handling.
- Irmak G, Demirtaş TT, Gümüşderelioğlu M. Sustained release of growth factors from photoactivated platelet-rich plasma. Eur J Pharm Biopharm. 2020;148:67–76.
- Boswell SG, Cole BJ, Sundman EA, Karas V, Fortier LA. Platelet-rich plasma: a milieu of bioactive factors. Arthroscopy. 2012;28(3):429–39.
- Descalzi F, Ulivi V, Cancedda R, Piscitelli F, Luongo L, Guida F, et al. Platelet-rich plasma exerts antinociceptive activity by a peripheral endocannabinoid-related mechanism. Tissue Eng Part A. 2013;19(19-20):2120–9.
- Pavlovic V, Ciric M, Jovanovic V, Trandafilovic M, Stojanovic P. Platelet-rich fibrin: basics of biological actions and protocol modifications. Open Med (Wars). 2021;16(1):446–54.
- US Food and Drug Administration. Public safety notification on exosome products, 6 December 2019. There are no FDA-approved exosome products and multiple serious adverse events were documented in Nebraska.
- First-in-human phase I randomised, double-blind, placebo-controlled intra-subject trial of allogeneic platelet-derived extracellular vesicles (Plexaris, Plexoval II, ACTRN12620000944932), n = 11: no serious adverse events and mean healing of 22.8 ± 8.7 days in both arms.
- Determination of the minimal clinically important difference for pain and function outcomes in knee osteoarthritis: visual analogue scale 1.37, WOMAC total 6.4 and WOMAC pain 1.5.
- de Vos RJ, Weir A, van Schie HTM, Bierma-Zeinstra SMA, Verhaar JAN, Weinans H, Tol JL. Platelet-rich plasma injection for chronic Achilles tendinopathy: a randomized controlled trial. JAMA. 2010;303(2):144–9. Minimal clinically important difference for VISA-A taken as 12 points; between-group difference –0.9 (95% CI –12.4 to 10.6).
- Secondary reporting of the RESTORE trial, including the finding of cartilage thinning in three or more subregions in 17.1% of PRP recipients versus 6.8% of placebo recipients (p = 0.02).
- Kawahara S, et al. Retrospective cohort of platelet-rich plasma with exercise, platelet-rich plasma alone and exercise alone in knee osteoarthritis, assessed with KOOS and OMERACT-OARSI responder criteria.
- Johnson-Lynn S, et al. PRIMA randomised clinical trial of platelet-rich plasma for ankle osteoarthritis: AOFAS difference –1 (95% CI –6 to 3). JAMA. 2021;326(16):1595–605.
- Reurink G, Goudswaard GJ, Moen MH, Weir A, Verhaar JAN, Bierma-Zeinstra SMA, et al. Platelet-rich plasma injections in acute muscle injury. Hazard ratio for return to play 0.96 and a between-group difference of 0 days (95% CI –11 to 11). https://repub.eur.nl/pub/79696/160212_Reurink-Gustaaf.pdf
- Nie LY, Zhao K, Ruan J, Xue J. Effectiveness of platelet-rich plasma in the treatment of knee osteoarthritis: a meta-analysis of randomized controlled clinical trials. Orthop J Sports Med. 2021;9(3):2325967120973284.
- Filardo G, Previtali D, Napoli F, Candrian C, Zaffagnini S, Grassi A. PRP injections for the treatment of knee osteoarthritis: a meta-analysis of randomized controlled trials. Cartilage. 2021;13(1 Suppl):364S–75S.
- Retraction note: meta-analysis of platelet-rich plasma combined with hyaluronic acid for knee osteoarthritis. J Orthop Surg Res. Retracted 1 April 2026.
- Retracted randomised controlled trial of subacromial platelet-rich plasma injection, withdrawn December 2025.
- Retraction notice for a platelet-rich plasma randomised trial.
- Retraction notice for a platelet-rich plasma systematic review.
- Further retraction notice affecting the platelet-rich plasma literature.
- Anz AW, Plummer HA, Cohen A, Everts PA, Andrews JR, Hackel JG. Bone marrow aspirate concentrate is equivalent to platelet-rich plasma for the treatment of knee osteoarthritis at 2 years: a prospective randomized trial. Am J Sports Med. 2022;50(3):618–29.
- Patel S, et al. Combined intra-articular and intraosseous platelet-rich plasma versus intra-articular platelet-rich plasma for knee osteoarthritis: a systematic review of five prospective studies and 112 knees. WOMAC mean difference 11.55 (95% CI 10.30–12.80).
- Baltzer AWA, Moser C, Jansen SA, Krauspe R. Autologous conditioned serum (Orthokine) is an effective treatment for knee osteoarthritis. Osteoarthritis Cartilage. 2009;17(2):152–60. Randomised comparison of 376 patients against hyaluronic acid and saline.
- Thompson E, et al. Alpha-2-macroglobulin versus platelet-rich plasma versus methylprednisolone for knee osteoarthritis: a double-blind randomised controlled trial of 75 enrolled patients.
- Mishra AK, Skrepnik NV, Edwards SG, Jones GL, Sampson S, Vermillion DA, et al. Efficacy of platelet-rich plasma for chronic tennis elbow: a double-blind, prospective, multicenter, controlled trial of 230 patients. Am J Sports Med. 2014;42(2):463–71. Success 71.5% versus 56.1% (p = 0.019).
- Karjalainen TV, Silagy M, O'Bryan E, Johnston RV, Cyril S, Buchbinder R. Autologous blood and platelet-rich plasma injection therapy for lateral elbow pain. Cochrane Database Syst Rev. 2021;9:CD010071. Standardised mean difference 0.26 with evidence graded insufficient.
- Placebo-controlled randomised trials of platelet-rich plasma for lateral elbow tendinopathy: pooled analysis showing no benefit over placebo.
- Meta-analysis of platelet-rich plasma augmentation in arthroscopic rotator cuff repair: re-tear risk ratio 0.70 (95% CI 0.56–0.88).
- Reanalysis of rotator cuff augmentation meta-analyses with trim-and-fill adjustment: re-tear risk ratio 0.91 (95% CI 0.69–1.19).
- Meta-analysis of platelet-rich plasma in osteonecrosis of the femoral head: risk ratio for conversion to total hip arthroplasty 0.29 (95% CI 0.16–0.53).
- Contradicting meta-analysis of adjuvant biologics in osteonecrosis of the femoral head reporting no significant reduction in arthroplasty conversion.
- Meta-analysis of platelet-rich plasma for carpal tunnel syndrome: Boston Carpal Tunnel Questionnaire standardised mean difference –2.06 derived from 191 patients.
- Meta-analysis of epidural platelet-rich plasma for lumbar radicular pain: pooled effect –0.09 with I² = 96.7%.
- Randomised comparison of intradiscal platelet-rich plasma with control for discogenic low back pain: 21 of 44 versus 16 of 45 responders (p = 0.244), with one reported case of spondylodiscitis.
- Kirschner JS, et al. Randomised comparison of platelet-rich plasma and hyaluronic acid for glenohumeral osteoarthritis: no significant between-group difference.
- El-Kadiry AE, Lumbao C, Rafei M, Shammaa R. Autologous BMAC therapy improves spinal degenerative joint disease and knee osteoarthritis: a retrospective comparison with platelet-rich plasma.
- Malahias MA, Roumeliotis L, Nikolaou VS, Chronopoulos E, Sourlas I, Papagelopoulos PJ. Platelet-rich plasma versus corticosteroid intra-articular injections for the treatment of trapeziometacarpal arthritis. VAS 20 versus 65 (p = 0.015) in 33 patients.
- Kolasinski SL, Neogi T, Hochberg MC, Oatis C, Guyatt G, Block J, et al. 2019 American College of Rheumatology/Arthritis Foundation guideline for the management of osteoarthritis of the hand, hip, and knee. Platelet-rich plasma is strongly recommended against.
- Bannuru RR, Osani MC, Vaysbrot EE, Arden NK, Bennell K, Bierma-Zeinstra SMA, et al. OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthritis Cartilage. 2019;27(11):1578–89.
- American Academy of Orthopaedic Surgeons. Management of osteoarthritis of the knee (non-arthroplasty), third edition evidence-based clinical practice guideline. Strength of recommendation for platelet-rich plasma: Limited.
- National Institute for Health and Care Excellence. Platelet-rich plasma injections for knee osteoarthritis. Interventional procedures guidance HTG497: use only with special arrangements for clinical governance, consent and audit or research.
- ESSKA Orthobiologic Initiative (ORBIT) consensus on the use of platelet-rich plasma in knee osteoarthritis: a valid treatment option in Kellgren–Lawrence grades 1 to 3 with a recommended course of two to four injections.
- ESSKA–ICRS consensus recommendation on platelet-rich plasma injections for the management of knee osteoarthritis in patients up to 80 years of age with Kellgren–Lawrence grades 0 to III.
- Eymard F, Ornetti P, Maillet J, Nogués É, Sordet C, Cadet C, et al. Intra-articular injections of platelet-rich plasma in symptomatic knee osteoarthritis: a consensus statement from French-speaking experts. Knee Surg Sports Traumatol Arthrosc. 2021;29(10):3195–210.
- American Academy of Physical Medicine and Rehabilitation. Platelet-rich plasma for knee osteoarthritis: an AAPM&R evidence-based guidance statement. PM R. 2026;18(Suppl 2):S20–35. Five evidence-based recommendations and eleven consensus-based best practices.
- Piuzzi NS, Ng M, Kantor A, Ng K, Kha S, Mont MA, Muschler GF. What is the price and claimed efficacy of platelet-rich plasma injections for the treatment of knee osteoarthritis in the United States? J Knee Surg. 2019;32(9):879–85. Among 179 centres contacted by secret shopper, the mean price of a single unilateral injection was US$714 ± 144 and the mean claimed clinical efficacy was 76% ± 11%.
- Andia I, Atilano L, Maffulli N. Moving toward targeting the right phenotype with the right platelet-rich plasma formulation for knee osteoarthritis. Ther Adv Musculoskelet Dis. 2021;13:1759720X211004336.
- Gentile P, Garcovich S, Bielli A, Scioli MG, Orlandi A, Cervelli V. The effect of platelet-rich plasma in hair regrowth: a randomized placebo-controlled trial. Stem Cells Transl Med. 2015;4(11):1317–23. Twenty men analysed; total hair density +45.9 hairs/cm² with PRP versus –3.8 with placebo (p < 0.0001).
- Kieling BdA, et al. Efficacy of platelet-rich plasma for androgenetic alopecia: a systematic review and meta-analysis of 14 studies and 431 patients. An Bras Dermatol. 2024. Pooled hair-density mean difference 27.55 hairs/cm² (95% CI 14.04–41.06), I² = 95.99%, GRADE low.
- Li K, et al. Efficacy of platelet-rich plasma for androgenetic alopecia: a meta-analysis of 10 randomised controlled trials and 318 participants. Aesthetic Plast Surg. 2023. Pooled hair density 25.09 hairs/cm² (95% CI 9.03–41.15, p = 0.002), with effect-size inflation in small trials (40.64 versus 9.57 hairs/cm², p for interaction 0.0004).
- Yao Y, et al. Platelet-rich plasma added to topical minoxidil for androgenetic alopecia: systematic review and meta-analysis of five randomised trials. PLOS ONE. 2024;19(8):e0308986. Hair density 21.81 (95% CI 10.64–33.00) at three months, GRADE low to very low.
- Trink A, Sorbellini E, Bezzola P, Rodella L, Rezzani R, Ramot Y, Rinaldi F. A randomized, double-blind, placebo- and active-controlled, half-head study to evaluate the effects of platelet-rich plasma on alopecia areata. Br J Dermatol. 2013;169(3):690–4. Forty-five patients.
- Xiao H, et al. Efficacy and safety of platelet-rich plasma in skin rejuvenation and scar management: a systematic review of 36 studies and 3,172 patients. Clin Cosmet Investig Dermatol. 2021.
- Pincelli MS, et al. Randomised double-blind split-face trial of platelet-rich plasma versus saline with full-face microneedling in 18 women: no improvement and slight worsening of laxity (p ≤ 0.004) and rhytids, with no difference between PRP and saline (p = 1.00).
- Randomised evaluator-blinded split-face trial of microneedle fractional radiofrequency with or without platelet-rich plasma for melasma, 30 enrolled and 29 completed: hemi-modified MASI fell on both sides with no significant between-side difference and no incremental change in ultrasound dermal thickness.
- Sollitto RA, et al. Platelet-rich plasma and platelet-rich fibrin for periorbital dark circles: systematic review of 14 studies. 2025.
- Deng W, et al. Platelet-rich plasma for diabetic foot ulcers: systematic review and meta-analysis of 22 randomised trials and 1,559 patients. J Orthop Surg Res. 2023. Complete healing risk ratio 1.42 (95% CI 1.30–1.56), certainty high for healing rate.
- Meznerics FA, et al. Platelet-rich plasma in chronic wound management: systematic review and meta-analysis of 29 randomised trials and 2,198 wounds. J Clin Med. 2022. Complete healing odds ratio 5.32 (95% CI 3.37–8.40).
- Systematic review and meta-analysis of platelet-rich plasma for venous leg ulcers: 16 randomised trials and 699 patients; complete closure odds ratio 5.06 (95% CI 2.35–10.89), recurrence odds ratio 0.16 (0.05–0.50). Int Wound J. 2024.
- Hu Z, et al. Platelet-rich plasma for pressure injuries: meta-analysis reporting healed ulcer area standardised mean difference 1.38 cm² (p = 0.02) and PUSH score 1.69 (p = 0.01). Int J Low Extrem Wounds. 2024.
- Best-evidence summary of platelet-rich plasma for chronic wounds drawn from three guidelines, five consensus statements and nine systematic reviews. Regen Ther. 2025;31:101053. Venous ulcers level 1a, grade A; diabetic foot ulcers level 1a, grade B; pressure and arterial ulcers insufficient.
- Review of allogeneic platelet-rich plasma for diabetic foot ulcers, including a pooled odds ratio for complete wound healing of 6.19 (95% CI 2.32–16.56) and the absence of any quantified immunogenicity data.
- Centers for Medicare and Medicaid Services. MLN Matters MM12403 revised: national coverage determination 270.3, blood-derived products for chronic non-healing wounds. Autologous platelet-rich plasma is nationally covered for chronic non-healing diabetic wounds for 20 weeks from 13 April 2021 when the device clearance includes management of exuding cutaneous wounds.
- Network meta-analysis of 32 randomised trials of regenerative adjuncts in periodontal intrabony defects. Odontology. 2024. Platelet-rich plasma did not improve probing depth or clinical attachment level (p > 0.05) and was among three interventions associated with worse radiographic bone fill.
- Systematic review and meta-analysis of platelet-rich plasma with bone graft in maxillary sinus augmentation: 13 quantitative studies, 369 patients and 621 augmentations. Bone formation mean difference –0.63 mm (95% CI –5.91 to 4.65, p = 0.81); implant survival risk ratio 1.95 (0.67–5.69, p = 0.22). Int J Oral Maxillofac Surg.
- Bennardo F, et al. Autologous platelet concentrates in the treatment of medication-related osteonecrosis of the jaw: a comprehensive review of 58 articles. Complete healing 480 of 557 lesions (86.2%) with any concentrate and 113 of 135 (83.7%) with platelet-rich plasma; no study has compared PRP, PRGF and L-PRF directly.
- Randomised comparison of 20% platelet-rich plasma drops with 20% autologous serum drops in 36 women with Sjögren-related dry eye: Ocular Surface Disease Index and Schirmer I unchanged, tear break-up time improved in both arms with no between-group difference. Sci Rep. 2023;13:19256.
- Pilot randomised trial of subretinal platelet-rich plasma versus inverted internal limiting membrane flap for large macular hole, 60 eyes: closure 93.3% versus 90.0% (p = 0.158). Indian J Ophthalmol. 2020.
- Poulios E, Mykoniatis I, Pyrgidis N, Zilotis F, Kapoteli P, Kotsiris D, et al. Platelet-rich plasma (PRP) improves erectile function: a double-blind, randomized, placebo-controlled clinical trial. J Sex Med. 2021;18(5):926–35. Minimal clinically important difference achieved by 20 of 29 versus 7 of 26 (risk difference 42%, 95% CI 18–66, p < 0.001).
- American Urological Association. Erectile dysfunction guideline, statement 25: platelet-rich plasma therapy should be considered experimental and should not be offered outside an institutional review board approved research protocol.
- European Association of Urology. Sexual and reproductive health guideline: do not use platelet-rich plasma to treat erectile dysfunction outside the confines of a clinical trial.
- Shahinyan GK, Weinberger JM, Mitchell WG, Shahinyan RH, Yang SC, Mills JN, Eleswarapu SV. Direct-to-consumer internet prescription platforms and platelet-rich plasma for erectile dysfunction. JAMA Netw Open. 2022;5(5):e2214187. Among 109 United States clinics, the mean price was US$1,507 ± 388, only 10 clinicians (9%) were urologists and 24 (22%) were not physicians.
- Retrospective uncontrolled single-centre cohort of intralesional platelet-rich plasma for chronic Peyronie disease, 36 men: curvature change –6.3° (95% CI –7.66 to –5.31, p < 0.001) with 9 of 36 (25%) achieving a 10° or greater reduction. Medicina (Kaunas). 2026;62(1):221.
- Scoping review of platelet-rich plasma in female stress urinary incontinence: 13 manuscripts and 320 patients, with a sham-controlled randomised trial reporting subjective cure of 32% versus 4% (p < 0.01) but objective cure of 0%.
- Katsika ET, et al. Intrauterine platelet-rich plasma for recurrent implantation failure: a critical appraisal of 13 randomised trials and 12 meta-analyses. Hum Reprod. 2025;40(5):771–84. All 13 trials at high risk of bias; restricted to the single low-risk trial, live birth odds ratio 1.10 (95% CI 0.38–3.14).
- Randomised evidence on intrauterine platelet-rich plasma in thin endometrium, including Eftekhar 2018 and Nazari 2019. Arch Gynecol Obstet. 2025;312(3):745–53.
- Systematic review of intraovarian platelet-rich plasma for poor ovarian response and primary ovarian insufficiency: 14 studies, no meta-analysis possible; the best-designed randomised trial reported clinical pregnancy in 8 of 30 with PRP versus 18 of 30 with control (p = 0.018). Int J Reprod Biomed. 2025;23(6):459–74.
- Meta-analysis of platelet-rich plasma in tympanic membrane perforation repair: 8 studies including 5 randomised trials and 455 participants; closure odds ratio 3.69 (95% CI 2.02–6.74) overall and 2.70 (1.27–5.76) in randomised trials alone. PLOS ONE. 2021;16(1):e0245968.
- Systematic review of platelet-rich plasma for post-viral olfactory dysfunction: 10 studies and 531 patients including 3 randomised trials, with discordant results and no meta-analysis. J Clin Med. 2024;13(3):782.
- Prospective uncontrolled two-institution trial of platelet-rich plasma for vocal fold scar and sulcus, 15 adults: Voice Handicap Index-10 –8.7 points and CAPE-V –18.8 points, without a comparator. Laryngoscope. 2024;134(12):5021–7.
- Randomised controlled trial of platelet-rich plasma injections versus pregabalin in 60 patients with painful diabetic polyneuropathy, one-year follow-up. Anaesth Pain Intensive Care.
- United States Federal Trade Commission. Stem Cell Institute of America co-founders and companies banned from marketing stem cell treatments and ordered to pay more than US$5.1 million, January 2025.
- Advertising Standards Authority (United Kingdom). Ruling on The Regenerative Clinic Ltd, reference A20-1056974. Claims that treatment was ground-breaking and extremely successful were unsubstantiated and misleading; complaint upheld under CAP Code rules 3.1, 3.7 and 12.1.
- Palombella S, Perucca Orfei C, Castellini G, Gianola S, Lopa S, Mastrogiacomo M, et al. Systematic review and meta-analysis of human platelet lysate versus fetal bovine serum for mesenchymal stromal cell expansion: 35 articles screened, 22 in meta-analysis. Stem Cell Res Ther. 2022;13(1):142.
- Liou JJ, Rothrauff BB, Alexander PG, Tuan RS. Effect of platelet-rich plasma on chondrogenic differentiation of adipose- and bone-marrow-derived mesenchymal stem cells. Tissue Eng Part A. 2018;24(19-20):1432–43.
- Guo SC, Tao SC, Yin WJ, Qi X, Yuan T, Zhang CQ. Exosomes derived from platelet-rich plasma promote the re-epithelialization of chronic cutaneous wounds via activation of YAP in a diabetic rat model. Theranostics. 2017;7(1):81–96.
- US Food and Drug Administration, Center for Biologics Evaluation and Research. Warning letter to Platinum Biologics LLC, CBER 25-705090, 15 August 2025. Umbilical-cord-derived products marketed as Nano PRP Jelly and Nano Flex declared unapproved new drugs and unlicensed biological products.
- Twelve-week prospective randomised split-face trial of a human adipose stem-cell exosome-containing solution with microneedling versus saline with microneedling, n = 28: Global Aesthetic Improvement Scale significantly higher on the treated side (p = 0.005). https://pubmed.ncbi.nlm.nih.gov/37377400/
- Split-face observational series of exosome-augmented microneedling and laser resurfacing in nine women aged 30 to 50.
- Tulpule S, et al. Gold-induced cytokine (GOLDIC) therapy in advanced knee osteoarthritis: a multicentre open-label study of 65 patients and 106 knees. https://pubmed.ncbi.nlm.nih.gov/37908900/
- Open-label pilot study of hyperacute serum (OrthoSera hypACT Inject) in knee osteoarthritis, n = 24, three weekly 3 mL injections, six-month follow-up without a control group.
- Fucaloro SP, et al. Platelet-rich plasma injections for knee osteoarthritis are associated with a higher complication rate than comparators: a systematic review and meta-analysis of 24 randomised controlled trials and 2,751 patients. Arthroscopy. 2025;41(11):4789–803.
- Meta-analysis of adverse events after intra-articular knee injection of platelet-rich plasma: 18.66% versus 9.14% with control, giving a number needed to harm of 11. https://pubmed.ncbi.nlm.nih.gov/40409439/
- Fucaloro SP, et al. Complications of platelet-rich plasma injections in the foot and ankle: systematic review and meta-analysis. Arthroscopy. 2025;41(10):4357–66.
- Xiong Y, et al. Efficacy and safety of platelet-rich plasma injections for osteoarthritis: a systematic review. Front Med (Lausanne). 2023;10:1204144.
- Keene DJ, Alsousou J, Harrison P, Hulley P, Wagland S, Parsons SR, et al. Platelet rich plasma injection for acute Achilles tendon rupture: PATH-2 randomised, placebo controlled, superiority trial. BMJ. 2019;367:l6132.
- Kearney RS, Ji C, Warwick J, Parsons N, Brown J, Harrison P, et al. Effect of platelet-rich plasma injection vs sham injection on tendon dysfunction in patients with chronic midportion Achilles tendinopathy: a randomized clinical trial. JAMA. 2021;326(2):137–44.
- Arita A, Tobita M. Adverse events related to platelet-rich plasma therapy and future issues to be resolved. Regen Ther. 2024;26:496–501.
- Redler LH, et al. Complete patellar tendon rupture in a recreational athlete after a series of four platelet-rich plasma injections. Clin J Sport Med. 2020;30(1):e20–2
- Centers for Disease Control and Prevention. Investigation of presumptive HIV transmission associated with receipt of platelet-rich plasma microneedling facials at a spa – New Mexico, 2018–2023. MMWR Morb Mortal Wkly Rep. 2024;73(16):372–6.
- Eymard F, Ornetti P, Maillet J, Noel E, Adam P, Legre-Boyer V, et al. Intra-articular injections of platelet-rich plasma in symptomatic knee osteoarthritis: a consensus statement from French-speaking experts (GRIIP). Knee Surg Sports Traumatol Arthrosc. 2025;33(6):2293–306.
- US Food and Drug Administration. Regulatory considerations for human cells, tissues, and cellular and tissue-based products: minimal manipulation and homologous use. Guidance for industry and FDA staff, July 2020. Section V.A states that platelet-rich plasma is not a human cell, tissue or cellular and tissue-based product under 21 CFR Part 1271 because it is a blood product.
- Electronic Code of Federal Regulations, 21 CFR 1271.10: an HCT/P is regulated solely under section 361 of the Public Health Service Act only if it is minimally manipulated, intended for homologous use, not combined with another article, and either without systemic effect and independent of living-cell metabolic activity or intended for autologous, first- or second-degree relative, or reproductive use.
- US Food and Drug Administration. Advancing the development of safe and effective regenerative medicine products. FDA Voices. The period of enforcement discretion for regenerative-medicine products ended 31 May 2021 and will not be extended further.
- US Food and Drug Administration, Center for Biologics Evaluation and Research. Warning letter to Estar Technologies Ltd, CBER 26-716831, 22 April 2026. Promotion of a PRP kit cleared only for bone-graft handling towards musculoskeletal, osteoarthritis, wound, ophthalmic and dental indications rendered the devices adulterated under section 501(f)(1)(B) and misbranded under section 502(o) of the Federal Food, Drug, and Cosmetic Act.
- European Commission. Minutes of the meeting of the Competent Authorities for Blood and Blood Components, 18–19 June 2019. Platelet-rich plasma and platelet-rich fibrin fall within different legal frameworks across Member States: tissues and cells legislation in three, blood legislation in five, medicinal-products legislation in two, other frameworks in three, and no regulation in six.
- Regulation (EU) 2024/1938 of the European Parliament and of the Council of 13 June 2024 on standards of quality and safety for substances of human origin intended for human application. Article 2(5) brings autologous substances that are processed or stored before application into scope; Article 85 repeals Directives 2002/98/EC and 2004/23/EC and Article 87 sets application from 7 August 2027.
- European Directorate for the Quality of Medicines and HealthCare. Twenty-second edition of the Guide to the preparation, use and quality assurance of blood components, published 21 May 2025, adding chapters on blood components for topical use or injection and on blood supply contingency planning.
- Pflieger D, et al. The EDQM Blood Guide and its new role as expert body under the SoHO Regulation. Blood Transfus. 2025;24(1):86–92.
- United Kingdom Government explanatory memorandum on Regulation (EU) 2024/1938. Paragraph 9 lists platelet-rich plasma among blood preparations not used for transfusion that are brought into scope, and paragraph 87 states that the Medicines and Healthcare products Regulatory Agency is not aware of any manufacture of platelet-rich plasma under a manufacturing licence.
- Conselho Federal de Medicina regulates the use of platelet-rich plasma for four musculoskeletal conditions. Folha de S. Paulo, 15 July 2026.
- Conselho Federal de Enfermagem (Brazil). Press release on the civil public action filed on 29 July 2026 challenging articles 8 to 12 of Resolução CFM nº 2.464/2026.
- Therapeutic Goods Administration (Australia). Australian regulatory guidelines for biologicals: autologous human cells and tissues products regulation, version 2.0, July 2019. Platelet-rich plasma, platelet-rich fibrin and conditioned serum are autologous human cell and tissue products across three tiers, and biologicals and excluded autologous products must not be advertised to consumers.
- Therapeutic Goods Administration (Australia). Regulating platelet-rich plasma, platelet-rich fibrin and conditioned serum.
- Health Canada. Information update RA-70559: Health Canada clarifies its position on platelet-rich plasma treatments, 26 July 2019. Platelet-rich plasma meets the definition of a drug under the Food and Drugs Act; preparation within a single practitioner-delivered procedure falls under provincially regulated practice of medicine; no application to market platelet-rich plasma or conduct clinical trials has been received and no scientific evidence has been reviewed.
- World Anti-Doping Agency. WADA Executive Committee approves the 2010 Prohibited List, under which platelet-derived preparations were prohibited when administered by the intramuscular route, with other routes requiring a declaration of use.
- World Anti-Doping Agency. The 2011 Prohibited List is now published: platelet-derived preparations were removed after consideration of the lack of current evidence concerning the use of these methods for purposes of performance enhancement.
- Athletics Integrity Unit. Understand the Prohibited List: platelet-rich plasma procedures remain not prohibited.
- Noridian Healthcare Solutions. Local coverage determination L39058: a non-coverage policy for all platelet-rich plasma injections or applications as a means of managing musculoskeletal injuries or joint conditions, revision effective 11 September 2025.
- Aetna. Clinical policy bulletin 0784: platelet-rich plasma. Platelet-poor and platelet-rich plasma injection are considered experimental, investigational or unproven for all indications, including approximately 30 named conditions, and codes 0232T, 0481T, G0460, G0465, P9020 and S9055 are listed as not covered.
- UnitedHealthcare. Commercial medical policy 2026T0498CC, prolotherapy and platelet-rich plasma therapies, effective 1 January 2026: due to insufficient evidence of efficacy, platelet-rich plasma is unproven and not medically necessary for any condition or indication.
- Blue Shield of California medical policy on orthopaedic applications of platelet-rich plasma. Current Procedural Terminology code 0232T covers injections of platelet-rich plasma at any site including image guidance, harvesting and preparation, is effective from 1 July 2010 and carries zero relative value units.
- Reimbursement analysis of PRP under Current Procedural Terminology code 0232T, which is not covered by most United States payers.
- Agência Nacional de Saúde Suplementar (Brazil). Parecer Técnico nº 04/GCITS/GGRAS/DIPRO/2024, 30 August 2024: the application of platelet-rich plasma is not included in the mandatory coverage list because coverage is not compulsory for experimental clinical or surgical treatment.
- Survey of direct-to-consumer pricing of intra-articular PRP in the United States: median US$800 with a range of US$350 to US$2,815.
- Cost-effectiveness of intra-articular PRP in Italy: €82.62 per treatment cycle with an incremental cost-effectiveness ratio of €1,524 per quality-adjusted life-year.
- Cost-effectiveness analysis of PRP for knee osteoarthritis in Iran: incremental cost-effectiveness ratio of 7,583 international dollars per quality-adjusted life-year.
- ClinicalTrials.gov NCT06040203. Istituto Ortopedico Rizzoli, leukocyte-rich versus leukocyte-poor platelet-rich plasma versus placebo in epicondylitis, n = 240.
- ClinicalTrials.gov NCT05497349. Istituto Ortopedico Rizzoli, leukocyte-rich versus leukocyte-poor platelet-rich plasma in hip osteoarthritis, n = 230.
- ClinicalTrials.gov NCT05517434. University Health Network Toronto, bone marrow aspirate versus saline and leukocyte-poor platelet-rich plasma versus saline, phase 2/3, n = 148.
- ClinicalTrials.gov NCT06685120. Istituto Ortopedico Rizzoli, platelet-rich plasma plus hyaluronic acid in knee osteoarthritis, n = 288, double-blind.
- ClinicalTrials.gov NCT05727371. Platelet-rich plasma combined with cross-linked hyaluronic acid (RegenMatrix), n = 280, triple-masked
- ClinicalTrials.gov NCT06932614. Lyophilized self growth factor versus platelet-rich plasma in knee osteoarthritis, n = 90, quadruple-masked, primary outcome WOMAC.
- ClinicalTrials.gov NCT06451120. University of California San Francisco, platelet-rich plasma in young versus old subjects, phase 2, n = 60.
- Xiong G, Lingampalli N, Koltsov JCB, Leung LL, Robinson WH, Maloney WJ, Chu CR. Men and women differ in the biochemical composition of platelet-rich plasma. Am J Sports Med. 2018;46(2):409–19.
- ClinicalTrials.gov NCT07231471. University of Utah, platelet-rich plasma for musculoskeletal conditions, planned enrolment 10,000, no masking, primary completion 2035.
- ClinicalTrials.gov NCT06003101. Massachusetts General Hospital, platelet-rich plasma, platelet-poor plasma and bone marrow aspirate concentrate in hip arthroscopy, phase 3, n = 160.
- ClinicalTrials.gov NCT05603468. Second Affiliated Hospital of Zhejiang University, platelet-rich plasma versus corticosteroid for rotator cuff tendinopathy, phase 4, n = 387.
- ClinicalTrials.gov NCT05412381. Hospital for Special Surgery, platelet-rich plasma in anterior cruciate ligament reconstruction, phase 3, n = 56, primary outcome platelet count.
- ClinicalTrials.gov NCT05378815. Assistance Publique – Hôpitaux de Paris, randomised triple-blind trial, n = 210, primary outcome change in numerical rating scale pain.
- ClinicalTrials.gov NCT06680856. Ospedale Policlinico San Martino, allogeneic platelet-rich plasma for diabetic foot ulcers, phase 1/2, n = 72, completed.
Genesis Scientific Publication is licensed under CC BY-NC-ND 4.0
