Background

Osteoarthritis (OA) is a degenerative disease characterized by the progressive degeneration of articular cartilage, the development of osteophytes, alterations in subchondral bone and synovial hyperplasia, clinically responsible for increasing disability and reduced quality of life [1]. Knee OA is characterized by multiple symptoms such as mechanical pain, stiffness and disabling inflammatory flare-ups that can reduce locomotor autonomy and promote falls [2, 3]. In the absence of a cure, treatment remains symptomatic, combining analgesic and anti-inflammatory drugs with essential non-pharmacological measures (physiotherapy, adapted physical activities, etc.) [4, 5].

Among injectable treatments, intra-articular corticosteroid injections have proved efficacy in cases of inflammatory flare-ups, with limited sustaining effect [6] and their extensive use can have harmful consequences for cartilage [7]. Viscosupplementation with hyaluronic acid may also be recommended to limit the functional impact, but the symptomatic effect remains limited and may last for several months with no long-term chondro-protective effect, according to French recommendations for the management of knee osteoarthritis [8].

More recently, intra-articular injections of autologous Platelet Rich Plasma (PRP) have proved to be an attractive alternative in the treatment of knee osteoarthritis [1, 9]. Indeed, their biological properties appear to be non-negligible, notably via the impact of growth factors on the stimulation and chondrogenic differentiation of mesenchymal stem cells and their effects on certain inflammation pathways [10,11,12]. Several meta-analyses have highlighted the benefits of using PRP as opposed to hyaluronic acid (HA) [13, 14]. Despite their use is not yet recommended by international guidelines, their position need to be specified.

Based on different but complementary modes of action, the PRP-HA combination therefore appears to be a relevant solution with potentially synergistic effects [15,16,17]. In addition to their recognized lubricant action, injections of HA simultaneously with PRP are thought to contribute to overall joint homeostasis, enabling better diffusion of the growth factors supplied by PRP into the joint following platelet degranulation [18]. Taking into account the complex interactions with different joint tissues, it seems that HA could prolong the activity of PRP [18]. In vitro studies also suggest improved stimulation of chondrogenesis, and a reduction in both inflammation and chondrocyte catabolism [19, 20]. In addition, Iio et al. [21] have shown that the release of growth factors is increased in vitro on the fifth day following the addition of HA, which remains to be confirmed clinically. Regarding patient satisfaction, Gao et al. [15] reported a reduction in pain in the short and medium term, and functional improvement with PRP-HA versus hyaluronic acid alone. However, the great heterogeneity of injection protocols and PRP preparations limit the significance of these results.

To date, this combination cannot be recommended in the absence of sufficient evidence of efficacy. In addition, the introduction of a combination mono-injection PRP-HA treatment could have a significant clinical role to play in improving efficacy in the symptomatic treatment of knee osteoarthritis. In this context, the aim of this multicenter, prospective, controlled and randomized study is to assess the efficacy and safety of single injection of a PRP-HA combination versus a crosslinked hyaluronic acid alone in mild-to-moderate symptomatic knee osteoarthritis.

Methods

Study design

A multi-center, randomized, single-blind (investigator-controlled) non-inferiority trial was carried out in three French hospitals (Dijon, Montpellier and Toulouse). The efficacy and safety of the PRP-HA combination in mono-injection was evaluated and compared to mono-injection of a crosslinked HA after 1 month (M1), 3 months (M3) and 6 months (M6) follow-up, with intention-to-treat (ITT) analysis. At 6 months, patients who were still symptomatic (Western Ontario and McMaster Universities Osteoarthritis, WOMAC pain score > 30/100 [22] (Table S1)) were systematically offered an open-label extension phase with another PRP-HA injection (exploratory phase), regardless of randomization group, with a further 6 months’ follow-up (assessment at 9 months (M9) then 12 months (M12)).

Trial registration

The clinical trial “Cellular Matrix Device for the Treatment of Mild to Moderate Knee Osteoarthritis” was registered under the trial registration number NCT03328728 on 24 October 2017. The registration was performed retrospectively, as the first patient was enrolled earlier, on 18 July 2017. The primary endpoint and the statistical analysis plan were defined prior to the start of the study and remained unchanged throughout the trial. No modifications were made after patient inclusion. This study was conducted in compliance with international Good Clinical Practice (Declaration of Helsinki) and French law. This study conforms to all CONSORT guidelines and reports the required information accordingly [23].

Evaluation criteria

The main objective of the study was to compare, in reference to baseline (M0), the 6-month evolution of the WOMAC pain on walking score (WOMAC A1) between the two groups. Secondary objectives were to compare:

  • The evolution of the different WOMAC domains (WOMAC A pain, WOMAC B stiffness, WOMAC C function) during follow-up as well as M9 and M12 for non-responders.

  • Changes in quality of life (SF-36 subdomains and scores) during follow-up and at M9 and M12 for non-responders.

  • Tolerance based on the proportion of adverse events during follow-up and serious adverse events reported in the 2 groups.

  • The percentage of responders at M6 was assessed according to two criteria. Firstly, the OMERACT-OARSI criteria were used [24] (Figure S2). The second criterion was defined as the percentage of patients achieving Minimal Clinically Important Improvement (MCII) and Patient Acceptable Symptom State (PASS) for WOMAC pain at M6. MCII was defined as an improvement of 19.9 mm in pain [25], while PASS was defined as a pain level below 32.3 mm on a scale ranging from 0 to 100 mm [26].

Participants

The main inclusion criteria were: participants aged between 40 and 80, predominantly femoro-tibial knee osteoarthritis (American College of Rheumatology criteria), symptomatic with a WOMAC pain on walking or WOMAC function score between 50 and 90 (using a 0–100 scale), and mild to moderate radiological knee osteoarthritis (Kellgren and Lawrence (K-L) stage II or III). The main non-inclusion criteria were patients with symptomatic bilateral knee osteoarthritis, who have received intra-articular hyaluronic acid or corticosteroid injections in the last 3 months, who received a PRP or PRP-HA injection less than one year ago, with a history of joint-tropic autoimmune diseases or chronic rheumatism, with hereditary or acquired coagulation disorders, with thrombocytopenia (platelets < 150 G/l) or anemia (hemoglobin < 10 g/dl) and with oncological or hematological pathologies or severe immunosupression.

Intervention

The PRP-HA combination was prepared using a dedicated, certified medical device: Cellular Matrix A-CP-HA kit (Regen Lab SA, Switzerland). Each A-CP-HA tube contains 2 ml of a 2% (20 mg/ml) non-crosslinked hyaluronic acid gel (1550 kDa) in phosphate buffer, a thixotropic separator gel composed of a polymer blend enabling isolation of PRP from other blood components located above the hyaluronic acid gel, and 0.6 ml of a sodium citrate solution (4%) acting as an anticoagulant located above the two gels. The vacuum in the tube allows for easy collection of 6 ml of the patient’s blood. Once filled with blood, the tube is centrifuged for 5 min at a force of 1500 x g. Centrifugation separates blood components according to their relative density. The separating gel, thanks to its thixotropic properties and specific density, migrates within the tube and intercalates precisely between the blood components, while the HA gel floats on the surface of the plasma. At the end of centrifugation, the separating gel forms a physical barrier which isolates the plasma and platelets, as well as the HA, in the upper section of the tube. Approximately 3 ml of PRP combined with 2 ml of HA (5 ml in total) are obtained from a 6 ml sample of patient blood. The preparation time for PRP-HA combination was approximately 10–15 min per participant, including 5 min for centrifugation.

According to the PAW classification of DeLong et al. [27] the PRP obtained with this device belongs to the P2-Bβ category, i.e. with a platelet concentration (P) above the baseline value in venous blood, but below 750,000 platelets/µl (P2), a white blood cell content below the baseline value in venous blood (B) and a neutrophil content also below the baseline value in venous blood (β).

Injection of crosslinked HA alone was performed with the Synvisc-One® device (Hylan G-F 20), (Sanofi, France), a 6-ml syringe containing hylan A (average molecular weight 6,000 kDa) and hylan B (hydrated gel, 8.0 mg per ml) in a buffered physiological sodium chloride solution. To maintain the double-blind design, approximately 6 mL of blood was also collected from patients in the HA group, although it was not used. A sham centrifugation step was performed systematically in the presence of the patient immediately prior to the injection. The preparation time for HA combination was approximately 5–10 min per participant.

All injections were carried out under full ultrasound guidance in sterile conditions by expert rheumatologists to ensure the correct positioning of the intra-articular product. The physician performing the injection was not involved in the clinical evaluation of the patient. Similarly, the clinician responsible for the clinical assessments did not participate in blood collection or in the preparation of the injectable product. To prevent patients from observing the procedure, a screen was positioned in front of them during both the preparation and injection phases. The patients are monitored lying on the examination table and asked to refrain from physical activity for one hour after the injection. Additionally, they were asked to have reduced activity during the 48 h following the injection, avoiding all intense activity in the affected lower limb. After this period, normal activity could be resumed.

Sample size

In an intention-to-treat non-inferiority study based on the reduction in pain on walking of the treated knee between M0 and M6 (assessed using the WOMAC score), the parameters for calculating the number of participants were: common standard deviation at 16 mm, non-inferiority limit at 6.8 mm (i.e. less than the minimum perceptible clinical improvement [25, 28]) with an alpha risk of 5% and power of 1 - β = 80%. The calculated sample size was 174 participants. With an estimated 10% drop-out and non-evaluable data, 190 participants, i.e. 95 per group, were planned for inclusion. Patients were randomly allocated (1:1) to the two groups using a computer-generated randomization sequence prepared by an independent statistician. Block randomization with variable block sizes (2, 4, or 6) was used to ensure balanced group sizes throughout enrollment. No stratification was applied. Investigators had no access to the allocation sequence, thereby ensuring allocation concealment.

Statistical analysis

The following characteristics were collected at inclusion, described in terms of descriptive statistics: age, gender, body mass index, radiological stage according to Kellgren Lawrence criteria, presence of knee effusion, use of symptomatic treatments for knee osteoarthritis (level 1, level 2 analgesics or non-steroidal anti-inflammatory drugs). Student’s t-test was used to compare quantitative values, and the Chi2 test for qualitative values. Analyses were performed on an intention-to-treat basis.

For the primary endpoint, non-inferiority of PRP-HA versus HA was tested, followed by exploratory superiority analysis. A repeated-measures ANOVA was also conducted. For secondary endpoints, data were analyzed using a linear model with repeated measures. The percentage of positive responders (according to OMERACT-OARSI criteria at M1, M3 and M6) was analyzed using logistic regression and generalized estimating equation (GEE) modeling. The hypothesis test used was a Wald test based on the regression parameter estimated for treatment group assignment. Differences in percentages of positive responders at 6 months were analyzed using GEE. For tolerance analyses, the percentage of patients presenting at least one adverse event was compared using Chi2 test. The characteristics of patients included in the extension phase were presented on the basis of descriptive statistics and comparisons of the means of WOMAC and SF36 scores. For WOMAC scores during follow-up, an analysis of variance for repeated measures was performed.

Statistical analyses were performed using SPSS (IBM, Armonk, USA) and Matlab (R2024a, The MathWorks, USA) software. A p-value < 0.05 was considered statistically significant. For the proportion of responders at M6, Chi2 test was performed to compare responders in terms of the WOMAC pain criteria OMERACT-OARSI, PASS and MCII.

Results

Population characteristics

Between 2017 and 2021, 156 patients were included: 81 patients in the PRP-HA group and 75 patients in the HA group (Fig. 1). Table 1 shows the population characteristics at inclusion. There were no significant differences in homogeneity between the two groups, except for the higher consumption of analgesics (p = 0.04) in the HA group and the higher proportion of K-L grade III in the PRP-HA group (p = 0.04).

Fig. 1
Fig. 1
Full size image

Participant flow chart (main phase with single-blind randomized controlled trial and open-label extension phase for non-responders at M6 in both arms)

Table 1 Population characteristics at inclusion (only non-responders at M6 continued into the open-label phase)
Full size table

Results of main phase (RCT M0-M6)

Primary endpoint

The evolution of the various WOMAC scores during follow-up is shown in Fig. 2 and Table S3, together with the difference from baseline. Regarding the non-inferiority test, a significant difference between the two groups was demonstrated for the WOMAC pain score on walking at M6, in favor of PRP-HA (−6.34 [−12.51; −0.18], p = 0.04). In the superiority test, a significant difference between the two groups was confirmed for the same score at M6 (−6.34 [−12.45; −0.23], p = 0.04) in favor of the PRP-HA group. There was a significant decrease in both groups during follow-up compared with baseline. A significant difference between the two groups at M1 was demonstrated for WOMAC pain during walking (−6.94 [−13.65; −0.23], p = 0.04) in favor of the PRP-HA group.

Fig. 2
Fig. 2
Full size image

Evolution and difference from M0 of WOMAC scores (WOMAC A1, WOMAC A, B, C, Total) during follow-up according to randomization groups (PRP-HA vs. HA). * = p < 0.05 (comparison with M0 for WOMAC evolution; comparison between treatments for Reduction compared to M0)

Results for algo-functional impairment (others WOMAC score) and quality of life (SF-36)

Whatever the others WOMAC domain, there was a significant decrease in both groups during follow-up compared with baseline. A significant difference between the two groups at M1 was demonstrated for WOMAC pain (−8.40 [−14.83; −1.97], p = 0.01) in favor of the PRP-HA group. No significant difference between the two groups was found for the different domains of the SF-36, except for pain at M1 in favor of the PRP-HA (−10.35 [0.56; 20.14], p = 0.04) (Table S4).

PRH-HA vs. HA - Tolerance

Tolerance analysis showed no significant difference between the two treatments in the percentage of patients with at least one adverse event (PRP-HA 7.4% vs. HA 5.3%, p = 0.59). Adverse events during follow-up were mainly post-injection pain and swelling in both groups, with rapid resolution (Table S5). No adverse event warranted discontinuation of the protocol.

M6 responder patients

The proportions of responders according to the different response criteria for knee osteoarthritis are shown in Table 2, with significant differences in favor of the PRP-HA group for achieving the PASS WOMAC pain criterion (50.6% vs. 33.3%, p = 0.03) (Figure S6), without reaching significance for the OMERACT-OARSI (58% vs. 48%, p = 0.26) and MCII (44.4% vs. 36.0%, p = 0.29) criteria.

Table 2 Percentage of responders at M6 by group and response criterion
Full size table

In univariate analysis, obtaining WOMAC Pain PASS was associated with WOMAC Pain score at inclusion (p < 0.001), WOMAC Total score at inclusion (p < 0.001) and allocated treatment (p = 0.04). In multivariate analysis, the only prognostic factor associated with PASS at the limit of significance was the treatment allocated at inclusion, i.e. the use of PRP-HA (OR = 2.05 [1.07–3.92]; p = 0.06).

Results of open extension phase (M6 to M12)

71 patients took part in the open extension phase (Fig. 1), as they were symptomatic at M6 (WOMAC A pain > 30/100), all receiving a PRP-HA injection. There was no significant difference in homogeneity between the two groups (PRP-HA or HA at M0, symptomatic at M6). The evolution of WOMAC scores and the difference from M0 during follow-up in the extension phase are shown in Fig. 3 and Table S7.

Fig. 3
Fig. 3
Full size image

In non-responders at M6 in the 2 randomization groups, evolution and difference from M0 of WOMAC scores during follow-up: controlled phase (M0-M6) then extension phase (injection of PRP-HA at M6, follow-up M9-M12, grey background). * = p < 0.05 (comparison with M0 for WOMAC evolution; comparison between treatments for Reduction compared to M0)

A significant decrease at M9 versus M0 was observed in the PRP-HA group for WOMAC pain on walking, pain, and total score. A significant decrease at M12 compared with M0 was observed for all scores, irrespective of treatment at inclusion. No significant difference between the two groups was found at follow-up. At 12 months, the proportion of responders was 40.7% (11 of 27 patients) for PRP-HA and 47.4% (18 of 38 patients) for the HA group.

Discussion

This multicenter, randomized, controlled, single-blind trial (investigator-controlled) showed a significant symptomatic effect of both intra-articular mono-injection treatments (PRP-HA or HA alone) at medium-term follow-up, with satisfactory tolerance. The benefit of mono-injection was maintained for up to 6 months for a majority of patients (54%). For pain on walking, non-inferiority of PRP-HA versus HA was demonstrated at M6, with a significant between-group difference in favor of PRP-HA (− 6.34 [− 12.51; −0.18], p = 0.04). At M1, a greater analgesic effect in the PRP-HA group compared with the HA group (− 6.94 [− 13.65; −0.23], p = 0.04) was found. In addition, superiority tests showed a significant gain for the PRP-HA combination at M6 (−6.34 [−12.45; −0.23], p = 0.04). This is corroborated by the fact that over 50% of patients reached an acceptable symptomatic level of pain (PASS WOMAC pain) at 6 months in the PRP-HA group versus 33% in the HA group. With regard to other efficacy criteria, such as quality of life (SF36), no statistically significant difference could be demonstrated between the two groups.

The extension phase presented in this study was carried out to assess the benefit of offering a PRP-HA injection at 6 months for patients still experiencing pain. Based on the different results obtained, the symptomatic benefit of this additional injection performed with the combination PRP-HA seems interesting, as it obtains a significant effect in both groups, which is maintained at one year in the majority of patients. As a result, it would seem that this treatment protocol could be pragmatic in routine clinical practice to improve the management of knee osteoarthritis, i.e. a single injection followed by a second at 6 months, only in patients who remain symptomatic or become symptomatic again. However, this open-label extension phase did not include a control group, and therefore does not allow definitive conclusions regarding the efficacy of PRP-HA at 12 months. The results of the extension phase are therefore framed as exploratory and hypothesis-generating rather than as confirmatory evidence of long-term efficacy. In terms of tolerability, the PRP-HA combination does not appear to increase the occurrence of adverse events [17]. Intra-articular injections for symptomatic relief (corticosteroids and HA) represent conservative therapeutic options for improving the impact of knee osteoarthritis in patients who are often frail with multiple comorbidities. The French Society of Rheumatology [4] and OARSI [29] recommend these conservative therapies, but they do not recommend PRP intra-articular injections. However, the latest meta-analyses reviewing large numbers of patients with moderate knee osteoarthritis (K-L grade II and III) point to the superior efficacy of PRP versus HA, both in terms of pain and function at 12 months post-injection [14, 30]. That said, the different types of PRP, platelet activators and injection protocols make the evaluation of meta-analyses of these randomized controlled trials complex, with sometimes discordant results.

Additionally, the potential synergistic effect of PRP combined with HA represents a key hypothesis of the present study. As it was explained by Iio et al. [21] based on in vitro experiments, the addition of HA can increase the release of growth factors from PRP, particularly from the fifth day onward, although this observation remains to be clinically confirmed. In this context, the clinical outcome observed in the PRP + HA group at one month may be explained by an early biological effect related to enhanced growth factor availability, potentially contributing to inflammation modulation and tissue repair during the subacute phase following injection. The absence of a significant difference between groups at 3 and 6 months suggests that this effect may be transient, with the clinical benefits of HA alone becoming comparable at mid-term follow-up, although comparable outcomes do not necessarily imply shared underlying mechanisms. Importantly, the present study design does not allow discrimination between the individual contribution of PRP, the intrinsic properties of HA, or a true synergistic interaction between the two components. Therefore, while a synergistic mechanism is biologically plausible, it cannot be conclusively established based on the current data. Future studies incorporating appropriate control groups and biomarker analyses are warranted to better elucidate the respective roles of PRP, HA, and their potential synergistic interaction.

There are a growing number of studies comparing the combination of PRP and HA versus PRP alone [31,32,33]. Notably, by comparing the PRP-HA combination versus PRP alone, Aw et al. [17]. showed that PRP-HA combination appears to offer superior symptomatic benefit in terms of pain scores, including VAS, WOMAC and IKDC (International Knee Documentation Committee) [17], as early as 3 months and sometimes up to 1 year. However, these results were not as prominent in a more recent meta-analysis by Zhang et al. [32], but this analysis did find greater reduction in the rate of adverse events with PRP-HA compared with PRP monotherapy.

At the same time, comparisons between the PRP-HA combination versus HA alone, are rare in the current literature [34,35,36,37], which remains the injectable symptomatic treatment most often used in first-line rheumatology practice. Using the identical treatment used in this study, Ciapini et al. [34] performed a randomized trial comprised of 3 injections of PRP-HA, HA alone or PRP alone at 1-month intervals. The results highlighted the superior efficacy of the PRP-HA combination for pain and functional recovery at 6-month follow-up. However, the study appears to lack statistical power. In a single-center, randomized, double-blind, controlled trial, Barac et al. [38] reported better WOMAC scores in the Cellular Matrix PRP-HA group compared with 2 different HA groups at 2, 6 and 12 months. On the other hand, in a recent randomized, double-blind, 12-month trial, Fossati et al. [39] found an improvement in the various WOMAC scores at 12 months, but no significant difference between the different treatment groups (PRP-HA vs. HA), with a protocol proposing 3 intra-articular injections 2 weeks apart. Finally, Yu et al. [35] presented a randomized controlled trial comparing four groups (PRP-HA, PRP, HA and placebo) with 6 injections at one-week intervals in progressively increasing doses if tolerated (e.g. 2, 4, 8, 10, 12 and 14 ml for PRP). This makes for complex interpretation of the results and clearly does not reflect real-world clinical practice. Nevertheless, the results are in favor of the PRP-HA combination in all areas of WOMAC (pain, stiffness, function) compared with other therapies, including HA alone, and placebo.

Moreover, despite the large number of studies comparing PRP + HA with placebo, PRP alone, or HA alone, important limitations remain in the existing literature. Several methodological issues persist, including suboptimal or poorly described randomization procedures and considerable heterogeneity among the results of randomized controlled trials, which would warrant further investigation, particularly through sensitivity analyses [40]. Notably, in a recent review, Bard et al. [41] emphasized the challenges associated with PRP standardization. Specifically, they identified substantial variability among PRP formulations, differing in platelet count and concentration, activation methods, leukocyte content and subtype, injected volume, and plasma quality, among other factors. Similar concerns can be extrapolated to the use of hyaluronic acid, as different HA formulations may also influence clinical outcomes [42]. Collectively, this heterogeneity reduces the robustness of conclusions drawn from comparative studies and meta-analyses, emphasizing the need for standardized protocols and reporting guidelines to improve reproducibility and clinical applicability.

Our study has several limitations. Firstly, the HA used in the PRP-HA combination was different from the crosslinked HA used in the second group and could have partially influenced the results [43]. Notably, intra-articular high molecular-weight HA (≥ 3000 kDa) provides superior efficacy and safety compared with low–molecular-weight HA (< 1500 kDa) [42]. In a recent study, the combination of PRP with two different hyaluronic acids demonstrated comparable efficacy and safety at 6 months [43]. Interactions among PRP and hyaluronic acid could influence treatment outcomes, and the optimal hyaluronic acid for combination with PRP has yet to be determined. Thus, a more appropriate comparison would have used identical HA formulations in both arms. However, the type of viscosupplementation used in this present study remains a common reference HA in clinical trials, and it is recommended as a comparator by the Food and Drug Administration. Secondly, statistically significant baseline imbalances were observed between groups, with a higher proportion of Kellgren–Lawrence grade III patients in the PRP-HA group and a higher proportion of daily analgesic or NSAID use at inclusion in the HA group. Both radiographic disease severity and baseline analgesic consumption are well-established predictors of pain outcomes in knee osteoarthritis and may therefore act as confounding factors, potentially influencing the observed treatment effects. Taken together, these imbalances may have acted in opposite directions, with greater disease severity potentially disadvantaging the PRP-HA group and higher baseline analgesic use potentially favoring the HA group. As a result, the between-group difference may have been attenuated, suggesting that the observed effect in favor of PRP-HA can be considered robust, although its exact magnitude remains uncertain in the absence of adjusted analyses. Thirdly, this study was conducted using a single-blind design, in which patients were aware of their treatment allocation, potentially introducing response bias, particularly for subjective patient-reported outcomes such as WOMAC and SF-36 scores. A double-blind design would have required more complex procedures, including systematic blood sampling in the non-PRP group and simulated centrifugation, in order to maintain patient blinding, which was not feasible in the present study. Fourthly, the number of patients included in our study was lower than the number predicted by the initial calculation. The estimated number was 174 patients, but only 156 could be included in the context of the Covid-19 pandemic. This 10% reduction in sample size may have resulted in insufficient statistical power, increasing the risk of type II error and potentially explaining the absence of statistically significant differences between groups for several secondary outcomes. Consequently, non-significant findings should be interpreted with caution, especially for certain inter-group differences. Finally, the primary endpoint selected for this study was a single WOMAC pain sub-item (pain during walking). While this choice may limit direct comparability with studies using the full WOMAC pain subscale or total WOMAC scores, it was made deliberately based on clinical relevance. Pain during walking is one of the most frequently reported symptoms by patients with knee osteoarthritis and represents a major determinant of functional limitation, mobility impairment, and reduced quality of life. Although the use of a single WOMAC item as a primary endpoint is relatively uncommon, this outcome was considered particularly meaningful from a patient-centered perspective, as it directly reflects a key daily-life concern in this population.

Considering this interesting benefit-risk balance, with a potential synergistic effect of this PRP-HA combination, it remains essential to confirm the results with a larger-scale, multicentric, randomized, double-blind study, in order to minimize these biases. In addition, it would seem appropriate to combine PRP and crosslinked HA. Indeed, the advantage of using crosslinked HA would be a longer intra-articular residence time, and this could therefore enable longer-lasting treatment efficacy. Furthermore, this could make it possible to target a more advanced stage of osteoarthritis (K-L grade IV). Finally, injections should be performed systematically under ultrasound guidance to improve efficacy and tolerance [44]. This method is now recommended by PRP experts [45].

Conclusions

The combination of PRP and hyaluronic acid in knee osteoarthritis appears promising, with encouraging symptomatic results coupled with excellent tolerance, and may offer advantages in specific patient populations. With a view to broaden the scope of treatment options, which are still limited, and currently lacking a treatment capable of slowing structural evolution, a ready-to-use combined treatment such as the one prepared with the Cellular Matrix A-CP-HA kit would appear to be an interesting alternative, and possible mono-injection, for patients suffering from moderate knee osteoarthritis, with numerous comorbidities that limit therapeutic possibilities in terms of medication.