Two intra-articular platelet-rich plasma injections were associated with greater reductions in pain and improvements in function over 6 months compared with a single corticosteroid injection or oral nonsteroidal anti-inflammatory drug therapy among patients with advanced knee osteoarthritis awaiting knee replacement, according to an article-in-press randomized trial in Journal of Orthopaedic Surgery and Research.
The trial included 90 patients aged 40 to 80 years with symptomatic Kellgren-Lawrence grade 3 or 4 knee osteoarthritis who were already on arthroplasty waiting lists. Patients were randomly assigned in equal groups to receive 2 autologous platelet-rich plasma (PRP) injections administered 1 week apart, a single intra-articular betamethasone injection, or oral aceclofenac 100 mg twice daily.
The study was prospective, randomized, and active-controlled. Participants were not blinded to treatment assignment, and the trial did not include a placebo or sham-injection group. The researchers assessed visual analogue scale pain scores, Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores, rescue opioid use, and serum biomarkers associated with cartilage turnover, inflammation, angiogenesis, pain signaling, and tissue remodeling at baseline, 3 months, and 6 months.
Eighty-two patients completed the 6-month follow-up. Two patients in the corticosteroid group and 4 patients in the nonsteroidal anti-inflammatory drug group discontinued after 3 months because of insufficient pain control and underwent earlier total knee replacement. Two additional patients in the nonsteroidal anti-inflammatory drug group were excluded because of worsening ischemic heart disease. No patients in the PRP group withdrew. The researchers reported no statistically significant baseline differences among groups in age, sex, body mass index, pain or WOMAC scores, or serum biomarker concentrations.
Pain scores improved only in the PRP group. Mean visual analogue scale pain scores declined from about 6 at baseline to about 3 at 3 months and about 4 at 6 months among patients who received PRP. Scores remained close to baseline in the corticosteroid and nonsteroidal anti-inflammatory drug groups. The researchers reported statistically significant differences favoring PRP over both comparators at 3 and 6 months.
Functional outcomes followed a similar pattern. Mean WOMAC total scores improved from 49 at baseline to 28 at 3 months and 33 at 6 months among patients who received PRP. The researchers reported statistically significant improvements from baseline in WOMAC pain, stiffness, function, and total scores only in the PRP group, with statistically significant differences favoring PRP over both comparator groups at 3 and 6 months.
Rescue opioid use, standardized as morphine milligram equivalents, was lower in the PRP group at 3 months compared with both comparator groups. At 6 months, opioid use remained lower with PRP compared with corticosteroid injection, but the researchers did not report a statistically significant difference between PRP and nonsteroidal anti-inflammatory drug therapy.
The researchers also reported differences in several serum biomarkers. At 3 months, PRP was associated with lower levels of cartilage oligomeric matrix protein, matrix metalloproteinase 3, interleukin 6, interleukin 18, tumor necrosis factor alpha, and calcitonin gene-related peptide compared with corticosteroid injection. Compared with nonsteroidal anti-inflammatory drug therapy, PRP was associated with lower levels of interleukin 18, tumor necrosis factor alpha, and calcitonin gene-related peptide.
At the same time point, PRP was associated with higher levels of soluble receptor for advanced glycation end products and transforming growth factor beta 1 compared with both comparator groups. Soluble triggering receptor expressed on myeloid cells 2 was higher with PRP compared with corticosteroid injection, but not compared with nonsteroidal anti-inflammatory drug therapy.
Some biomarker differences persisted at 6 months. The researchers reported lower levels of cartilage oligomeric matrix protein, interleukin 6, and interleukin 18 with PRP compared with corticosteroid injection, and lower levels of interleukin 6 and tumor necrosis factor alpha with PRP compared with nonsteroidal anti-inflammatory drug therapy. No statistically significant between-group differences were observed for monocyte chemoattractant protein 1, vascular endothelial growth factor, fractalkine, brain-derived neurotrophic factor, or beta nerve growth factor.
The biomarker findings should be interpreted cautiously. The researchers described the biomarker analyses as exploratory and did not apply correction for multiple comparisons. The PRP product also was not quantitatively characterized before injection, although the researchers noted that blood cell content had been evaluated in a previous study using the same preparation protocol.
Adverse events in the injection groups were mild and transient. Three patients in the PRP group and 2 patients in the corticosteroid group experienced transient increases in knee pain and swelling that resolved within 1 day. No other adverse events were reported in either injection group.
The trial’s design limits the conclusions that can be drawn. Because there was no placebo or sham-injection arm, the study compared PRP with 2 active treatment strategies under open-label conditions but did not determine whether PRP would outperform an inert control. The treatment groups also differed in route of administration, dosing frequency, and overall treatment intensity: 2 injections for PRP, 1 injection for corticosteroids, and oral therapy for the nonsteroidal anti-inflammatory drug group.
The researchers used a per-protocol analysis as the primary analysis, stating that dropouts reflected treatment failure. Because the 6-month analysis excluded patients who discontinued, including those who underwent earlier knee replacement because of insufficient pain control, the effect of missing data on between-group comparisons cannot be fully assessed from the reported results.
Generalizability may also be limited. The trial was conducted at a single center, included a relatively small sample, and followed patients for 6 months. Patients were excluded if they had a body mass index greater than 40, recent intra-articular corticosteroid or hyaluronic acid injections, infectious diseases, severe cardiovascular disease, diabetes mellitus, rheumatoid arthritis, hematologic disorders, malignancy, immunosuppression, anticoagulant or antiaggregant therapy, or known allergies to study medications.
The researchers noted that clinical guidelines remain cautious on PRP for knee osteoarthritis because of heterogeneity in PRP formulations, dosing protocols, leukocyte content, outcome measures, and long-term evidence. They wrote that larger, longer-duration randomized trials are needed to confirm the findings and determine whether benefits persist beyond 6 months.
The authors concluded that PRP may have a role as a bridge therapy for patients awaiting arthroplasty, but said larger and longer-duration randomized trials are needed to confirm whether the benefits persist beyond 6 months.
Disclosures: The study was funded by the Scientific Grant Agency of the Ministry of Education, Science, Research and Sports of the Slovak Republic through VEGA no. 1/0686/24. The researchers reported no competing interests.