Platelet-Rich Plasma, Hyaluronic Acid, or Their Combination for Knee Osteoarthritis in Athletes

Background: Intra-articular platelet-rich plasma (PRP) has shown clinical benefit in mild to moderate knee osteoarthritis (OA), but results are heterogeneous. Differences in preparation, platelet concentration, injected volume, leukocyte content and absolute platelet dose make comparisons difficult, and most previous randomized trials poorly report the characteristics of the biologic product. Recent evidence suggests a relationship between higher platelet exposure and better clinical response. The combination of PRP and hyaluronic acid (HA) has shown favorable results compared with HA alone, but it remains uncertain whether the combination adds benefit to PRP alone when both are applied with standardized technique, full product characterization, blinding and long follow-up. Population: In this study, the term athletes is used operationally to describe physically active adults who regularly practice structured exercise (at least 150 minutes per week of moderate intensity or 75 minutes per week of vigorous intensity for at least 6 months, according to the World Health Organization 2020 guidelines), confirmed by the short International Physical Activity Questionnaire (IPAQ) or a Tegner activity level of 3 or higher, and not exclusively to competitive athletes. Design: Single-center, prospective, randomized, controlled, double-blind, superiority trial with three parallel groups (1:1:1) conducted at the Institute of Orthopaedics and Traumatology, Hospital das Clinicas, University of Sao Paulo School of Medicine (IOT-HCFMUSP), Brazil, and reported according to CONSORT 2025. Interventions: All participants undergo blood collection to preserve blinding. Interventions are performed in a single session, intra-articularly, under ultrasound guidance, with opaque syringes. The PRP group receives approximately 5 mL of leukocyte-poor PRP; the PRP+HA group receives approximately 5 mL of leukocyte-poor PRP followed, in the same session, by 4 mL of cross-linked HA (20 mg/mL); and the HA group receives 4 mL of cross-linked HA (20 mg/mL). PRP is prepared at the IOT Regenerative Medicine Laboratory, used fresh and without exogenous activation, with a final concentration of at least 1,000,000 platelets/uL and a minimum total dose of 5 billion platelets. For every participant, whole blood and PRP characteristics are recorded (anticoagulant, spin protocol, platelet recovery, platelet concentration, leukocytes with differential, red blood cells, concentration factor, volume, total platelet dose, activation, delivery and post-injection care), following MIBO and the AAOS/NIH U-13 consensus. Nonsteroidal anti-inflammatory drugs must be stopped at least 5 days (recommended 7 days) before blood collection. Acetaminophen up to 3 g/day is allowed as rescue medication, and all rescue medication use (yes/no, drug, dose and frequency) is recorded at every follow-up visit. Randomization and blinding: The allocation sequence is generated by an independent statistician using permuted blocks of variable size, stratified by sex and Kellgren-Lawrence grade, and allocation is performed through a central electronic system after eligibility confirmation, informed consent and baseline assessment. Participants, outcome assessors and the statistician are blinded. The injecting physician cannot be blinded and does not take part in outcome assessment or analysis. Outcomes: The primary outcome is the change in WOMAC total score from baseline to 12 months, comparing PRP+HA versus PRP alone. Secondary outcomes include WOMAC total and subscales, KOOS (five subscales), Lysholm, Tegner, IKDC subjective, pain by visual analog scale (VAS), Patient Global Impression of Improvement, satisfaction, OMERACT-OARSI responder rate, rescue medication use, cointerventions, new injections, surgery and adverse events, assessed at 3, 6, 12 and 24 months, plus the association between clinical response and PRP characteristics. Questionnaires are administered electronically in REDCap, in validated Brazilian Portuguese versions, with assistance from a trained blinded assessor when needed. Sample size and analysis: 300 participants (100 per group), assuming a clinically relevant difference of 10 points in the normalized WOMAC, standard deviation of 20 points, 80% power, adjusted significance level and up to 15% loss to follow-up. The primary analysis follows the intention-to-treat principle using analysis of covariance adjusted for baseline WOMAC and stratification factors; repeated measures are analyzed with linear mixed models, with Holm correction for multiple comparisons and multiple imputation for missing data.