Comparative efficacy of ultrasound-guided intra-articular injection of platelet-rich plasma (PRP), hydrolyzed type I collagen, and hyaluronic acid in patients with knee osteoarthritis: a randomized controlled trial

Abstract

Background

Intra-articular therapies are widely used for knee osteoarthritis (KOA), but direct comparisons between emerging biomaterials such as hydrolyzed collagen (HC) and established options like platelet-rich plasma (PRP) and hyaluronic acid (HA) are limited.

Objective

To compare the efficacy of intra-articular PRP, HC, and HA on pain, stiffness, and functional outcomes in KOA.

Methods

In this randomized, assessor- and statistician-blinded trial, 135 patients with Kellgren–Lawrence grade 2–3 KOA were assigned (1:1:1) to receive three ultrasound-guided Intra-articular injections of PRP, HC, or HA at two-week intervals. The primary outcome was change in WOMAC total score at 6 months. Secondary outcomes included WOMAC subscales, VAS pain, and the Lequesne Index.

Results

One hundred twenty-six patients completed follow-up. All groups showed significant improvements (p < 0.001). HA demonstrated greater improvements in some outcome measures at 1 month. PRP and HC demonstrated more sustained improvements than HA during 1 to 6 months follow-ups, although statistically significant between-group differences were limited to selected outcomes. No serious adverse events occurred.

Conclusion

Intra-articular PRP, HC, and HA improved outcomes in mild to moderate knee OA; PRP and HC showed a trend toward more sustained benefits during 1 to 6 months. Longer-term trials are necessary to confirm these findings.

PLAIN LANGUAGE SUMMARY

Why was this study done?

Knee osteoarthritis is one of the most common causes of pain and disability in older adults. Several injectable treatments are available, but it is still unclear which provides the greatest clinical benefit. This study compared three commonly used ultrasound-guided intra-articular treatments: platelet-rich plasma (PRP), hydrolyzed type I collagen and hyaluronic acid.

What was done?

A total of 135 adults with mild-to-moderate knee osteoarthritis were randomly assigned to receive one of the three treatments. Participants received three ultrasound-guided injections into their knee joints at 2-week intervals and were followed for 6 months. Pain, stiffness and function were evaluated using validated clinical outcome measures.

What were the results?

All three treatments improved patients’ symptoms over time. Improvements during the first month were generally comparable among the treatment groups. At 6 months, PRP demonstrated greater improvement than hyaluronic acid in several outcomes. Hydrolyzed collagen also showed advantages over hyaluronic acid for selected outcomes. No serious treatment-related adverse events were observed, and all three treatments were generally well tolerated.

What do these results mean?

These findings suggest that ultrasound-guided PRP may provide greater mid-term clinical benefit than hyaluronic acid for patients with mild-to-moderate knee osteoarthritis. Hydrolyzed collagen may also represent a promising treatment option for selected clinical outcomes. Larger studies with longer follow-up are needed to confirm these findings and to determine the long-term comparative effectiveness of these biologic therapies.

ARTICLE HIGHLIGHTS

• This randomized assessor- and statistician-blinded controlled trial compared ultrasound-guided intra-articular platelet-rich plasma (PRP), hydrolyzed type I collagen, and hyaluronic acid in patients with mild-to-moderate knee osteoarthritis.

• All three treatments improved pain and functional outcomes over the 6-month follow-up, with generally comparable improvements during the first month.

• PRP demonstrated greater improvement than hyaluronic acid in several clinically important outcomes at 6 months, while hydrolyzed collagen showed advantages over hyaluronic acid for selected outcomes.

• All treatments were generally well tolerated, and no serious treatment-related adverse events were observed.

• These findings provide comparative evidence for three commonly used intra-articular treatments and support further evaluation of PRP and hydrolyzed collagen in larger studies with longer follow-up.

1. Introduction

Knee osteoarthritis (KOA) constitutes a major global health concern, driving chronic pain and musculoskeletal disability worldwide [1]. Its prevalence has been reported to reach approximately 16% and is increasing independently of population aging or obesity [2]. This contributes to impaired quality of life due to pain, stiffness, and reduced mobility, incurring substantial socioeconomic costs [3]. Although guidelines prioritize conservative strategies like therapeutic exercise [4], biofeedback programs [5], low-intensity pulsed ultrasound (LIPUS) [6], low-level laser therapy and high-intensity laser therapy [7,8], adherence challenges often limit their effectiveness [9]. Common pharmacological approaches also face limitations: paracetamol provides minimal benefit, oral NSAIDs carry systemic risks, and opioids demonstrate a poor risk-benefit profile, although topical NSAIDs offer a safer alternative for knee-specific symptoms [10].

Consequently, for patients refractory to initial conservative management, intra-articular (IA) injections represent an important intermediate therapy before considering total knee arthroplasty (TKA). The primary goals of IA therapy are pain relief and functional improvement, potentially delaying surgery, while emerging biologics are also investigated for their ability to modulate the joint environment [11]. Several IA therapies are currently used in clinical practice for KOA, including corticosteroids, hyaluronic acid, platelet-based preparations such as PRP and plasma rich in growth factors (PRGF). Other approaches including ozone therapy, autologous conditioned serum (ACS), and more recently, collagen-based injectates and mesenchymal stem/stromal cells (MSCs), stromal vascular fraction (SVF) and cell-free biological injectates including MSC-derived extracellular vesicles (EVs). Although these therapies are under active clinical investigation, current evidence remains limited, and these therapies have not yet been routinely incorporated into major clinical guidelines [12–20].

Intra-articular corticosteroid injections remain among the most commonly employed IA options because they offer rapid short-term analgesia and are recommended in selected patients by major clinical guidelines [4,11,21]. Nevertheless, their therapeutic benefit is usually transient, and concerns about potential cartilage-related structural effects with repeated use have been reported with long-term repeated injections that require more cautious, episodic administration in current clinical practice [22]. Intra-articular hyaluronic acid (HA) primarily acts as a viscoelastic supplement by improving synovial fluid lubrication and shock absorption [23,24]; Experimental and clinical evidence also suggests anti-inflammatory, chondroprotective and mechanobiological effects [25]. Clinical outcomes vary across HA formulations according to molecular weight, concentration, source, cross-linking characteristics and injection regimen. This variability contributes to ongoing debate and inconsistent recommendations across clinical guidelines [4,21,23,26]. Randomized evidence suggests that high-molecular-weight HA products may yield more favorable outcomes compared to low-molecular-weight variants due to differences in biochemical properties and residence time characteristics [26].

Platelet-rich plasma (PRP), an autologous blood-derived biological therapy enriched with platelets and bioactive molecules, has attracted considerable attention as a regenerative treatment for KOA. Beyond delivering a concentrated source of growth factors, PRP has been proposed to modulate the intra-articular inflammatory milieu, influence chondrocyte metabolism, promote extracellular matrix homeostasis and support tissue repair processes [27,28].

Experimental and translational studies suggest that PRP may reduce pro-inflammatory signaling while enhancing anabolic responses within the osteoarthritic joint [27,28]. Clinical studies have also reported PRP-related changes in MRI-based structural outcomes, although these findings should be interpreted cautiously [29]. Nevertheless, substantial heterogeneity in PRP preparation protocols (including differences in platelet concentration, leukocyte content, activation methods and injection regimens) continues to limit direct comparisons across studies and precludes definitive conclusions regarding the optimal formulation [30,31]. Despite this variability, systematic reviews and meta-analyses have reported that PRP may provide greater improvements in pain and function than HA, particularly in mid- and longer-term follow-ups [30–33]. Randomized clinical trials evaluating PRGF, a standardized platelet-derived formulation, have also reported clinically meaningful improvements in pain and functional outcomes compared to HA [16,34].

More recently, intra-articular administration of low-molecular-weight hydrolyzed collagen (HC) has emerged as a potential therapeutic option for KOA, with preliminary clinical studies suggesting improvements in pain and functional outcomes [35,36]. Proposed mechanisms include providing extracellular matrix precursors, modulating chondrocyte metabolism and promoting hyaline cartilage synthesis [36,37]. Experimental studies have further demonstrated that IA collagen peptides may exert dual-compartment effects by reducing synovial inflammation while enhancing extracellular matrix synthesis within articular cartilage [38]. However, clinical evidence regarding intra-articular HC remains limited and further well-designed randomized controlled trials are warranted to establish its efficacy and long-term safety.

Despite encouraging preliminary findings, the clinical evidence supporting intra-articular HC remains limited and heterogeneous. To our knowledge, no published randomized controlled trials (RCTs) have directly compared the efficacy of intra-articular HC, HA, and PRP within the same cohort of patients with KOA. This knowledge gap limits evidence-based selection among these injectable therapies. Therefore, the present randomized controlled trial was designed to compare these three IA treatment modalities and to evaluate their comparative effects on pain, stiffness, function and disease-specific clinical outcomes in patients with knee osteoarthritis.

2. Methods

2.1. Study design

This was a prospective, randomized, assessor- and statistician-blinded, parallel-group controlled trial. The study protocol was approved by the Ethics Committee of Shahid Beheshti University of Medical Sciences (IR.SBMU.MSP.REC.1403.102) and conducted in accordance with the Declaration of Helsinki. The trial was prospectively registered in the Iranian Registry of Clinical Trials (IRCT20130523013442N34) on 8 February 2024, prior to enrollment of the first participant. Recruitment was conducted at Modava Pain Clinic and the department of Physical Medicine and Rehabilitation, Shahid Modarres Hospital, Tehran, Iran, between April 2024 and August 2025. Written informed consent was obtained from all participants before enrollment.

This randomized controlled trial was designed, conducted and reported in accordance with the CONSORT 2010 Statement.

2.2. Participants

This trial enrolled adults aged 40–70 years with symptomatic knee osteoarthritis (KOA) diagnosed according to the American College of Rheumatology (ACR) clinical classification criteria. Additional inclusion requirements were the presence of symptoms for at least three months, radiographic evidence of Kellgren–Lawrence grade 2 or 3 on standing knee radiographs, a body mass index below 30 kg/m², and the ability to attend scheduled follow-up visits.

Exclusion criteria included inflammatory arthropathies such as rheumatoid arthritis or gout, collagen vascular or autoimmune diseases, uncontrolled diabetes mellitus, active infection or malignancy, and significant prior knee trauma or active skin disorders at the injection site. Additional exclusion criteria comprised anticoagulant or antiplatelet therapy within ten days, NSAIDs use within two days before injection, systemic corticosteroid use within the preceding two weeks, any intra-articular knee injection during the previous six months, platelet count < 150,000/µL, hemoglobin < 11 g/dL, severe varus or valgus deformity (> 20 degrees), hypersensitivity to avian products, bovine collagen, or hyaluronate, pregnancy or breastfeeding, history of vasovagal reactions to injections, clinical or imaging evidence of significant lumbar pathology contributing to leg pain, neuropathy, or evidence of hemarthrosis. If both knees were affected, only the more symptomatic knee was included for randomization and treatment.

At baseline, participants’ eligibility was assessed through medical history (prior intra-articular injections, corticosteroid and anticoagulant/anti-platelet/NSAID use), physical examination, laboratory tests (CBC, ESR, CRP), and knee radiography (standing AP and lateral views). All participants underwent baseline laboratory testing (CBC, ESR, CRP) before randomization to confirm eligibility. Participants randomized to the PRP group subsequently underwent an additional venous blood collection immediately before each treatment session for PRP preparation.

2.3. Randomization and allocation

Eligible participants were randomized in a 1:1:1 ratio to receive PRP, HC or HA. A block randomization sequence with a fixed block size of six was generated using Random Allocation Software (Version 2.0). Eligible participants were enrolled by a study coordinator who was not involved in outcome assessment. To ensure allocation concealment, the assignments (Group A, B, C) were placed in sequentially numbered, opaque, sealed envelopes by the coordinator. Treatment allocation was revealed by opening the next sequential envelope immediately before intervention administration.

Blinding procedures were implemented wherever feasible. The outcome assessor, who performed all follow-up assessments, remained blinded to treatment allocation throughout the study. Outcome data were coded prior to analyses and treatment identities were revealed only after completion of the statistical analyses. All statistical analyses were performed by an independent biostatistician who was also blinded to treatment group allocation. To minimize performance bias, identical injection techniques were used. Additionally, syringes of identical size, covered with opaque sheaths were used for all injections. Despite these precautions, complete injector and participant blinding could not be guaranteed, as experienced clinicians may recognize subtle differences in viscosity between the study injectates (PRP, HC, and HA), and PRP preparation requires a larger blood draw and additional handling steps. Accordingly, the study is reported as an assessor- and statistician-blinded randomized controlled trial.

2.4. Interventions

All intra-articular injections were administered under ultrasound guidance by a single experienced physical medicine and rehabilitation (PM&R) specialist. A sterile medial mid-patellar approach with the knee in full extension was used in all cases, and a 22-gauge needle was employed. Each participant received three intra-articular injections at fixed two-week intervals, with identical patient positioning and injection technique at each visit to enhance reproducibility. A standardized regimen of three injections at fixed two-week intervals was applied across all treatment groups to ensure methodological consistency and facilitate direct comparison of treatment effects. Although no consensus exists regarding the optimal PRP injection schedule, previous randomized trials have used repeated PRP injections administered at varying intervals with three injections at two-week intervals representing one of the frequently adopted regimens. Therefore, an identical injection schedule was intentionally applied across all study arms to minimize procedural heterogeneity, isolate treatment effects rather than differences in injection frequency, and strengthen the internal validity of treatment comparisons [39–42].

2.4.1. Hydrolyzed collagen group

Low-molecular-weight hydrolyzed type I collagen (ChondroGrid^®, Bioteck S.p.A., Italy; approximately 3 kDa) was reconstituted immediately before administration with sterile Water for Injection (WFI) according to manufacturer’s instructions, yielding a final concentration of 2 mg/mL. The final solution (2 mLs) was injected at each treatment session.

2.4.2. PRP group

For participants allocated to the PRP group, 35 mL of peripheral venous blood was collected from the antecubital vein using a 21-gauge needle into sterile tubes containing 5 mL of acid-citrate-dextrose solution A (ACD-A) as anticoagulant. This additional blood collection was performed exclusively for PRP preparation immediately before each treatment session. Samples were processed immediately at room temperature.

PRP was prepared using a registered commercial kit (Royagen kit, Arya Mabna Tashkhis Co., registration number 312569) through a two-step centrifugation protocol:

• First spin at 1600 rpm (=400g) for 12 minutes to separate plasma and buffy coat.

• Second spin at 3500 rpm (=1900g) for 7 minutes to concentrate platelets.

The resulting supernatant platelet-poor plasma was discarded, yielding approximately 5–6 mL of PRP (3 mL per tube) with a mean platelet concentration of 5.24 ± 1.3 times baseline platelet count, with a platelet recovery rate of 70-90%. Platelet (PLT) and leukocyte counts were quantified for quality control. The quality control of PRP characteristics showed samples had an average count of 2.1 ± 0.31 × 10³/μL WBCs. Table 1 shows PRP characteristics in more detail.

Table 1.

PRP characteristics.

PRP characteristics, mean ± SD
Group (#)		PRP (first injection)	PRP (second injection)	PRP (third injection)
PLT Count(×103/µL)		1392.41 ± 312.62	1412.33 ± 266.21	1405.68 ± 271.34
PRP/Blood PLT Ratio		5.34 ± 1.06	5.16 ± 0.65	5.22 ± 1.15
WBC Count(×103/µL)		2.11 ± 0.28	2.25 ± 0.29	2.12 ± 0.31
Leukocyte Differential Count	Lymphocyte (×103/µL) (%)	1712.11 ± 296.20	1820.27 ± 298.02	1719.28 ± 280.65
		81.13 ± 1.52	80.90 ± 1.71	81.09 ± 1.59
	Neutrophil (×103/µL) (%)	242.31 ± 41.21	251.18 ± 40.29	238.52 ± 43.27
		11.48 ± 0.72	11.16 ± 0.65	11.25 ± 0.68
	Monocyte (×103/ µL) (%)	130.11 ± 34.11	140.21 ± 36.28	134.20 ± 32.23
		6.16 ± 1.70	6.23 ± 1.85	6.33 ± 1.44

The final PRP product was gently mixed with a standard shaker and injected within two hours of blood collection. No exogenous platelet activator was used; platelet activation occurred physiologically following IA injection through exposure to endogenous collagen within the joint [27]. Under sterile ultrasound guidance, PRP was injected via the medial mid-patellar approach using a 22-gauge needle. To ensure standardized reporting, the Minimum Information for Studies Evaluating Biologics in Orthopedics (MIBO) checklist was followed.

2.4.3. Hyaluronic acid group

Participants randomized to the HA group received 2 mL of sodium hyaluronate (Hyalgan^®, Fidia Farmaceutici S.p.A., Italy; molecular weight 500–730 kDa, pH 6.8–7.5) from a prefilled syringe at each session.

2.4.4. Post-injection care and co-interventions

Post-injection care was standardized across treatment groups. After each injection, patients were instructed to perform several cycles of knee flexion and extension to facilitate intra-articular uniform distribution of the injectate. They remained under observation for 10–15 minutes before discharge, and were asked to report any adverse events.

Participants were advised to restrict weight-bearing activities for 24–48 hours, apply intermittent cold packs three times daily for 10 minutes, and use Acetaminophen up to 2 g/day as rescue analgesia. For persistent pain, acetaminophen combined with codeine was permitted as needed; Rescue analgesia consumption was not systematically recorded and therefore not included in the statistical analysis. NSAIDs, corticosteroids, and other analgesics or medications affecting platelet count or function were prohibited during the study period.

All participants received an identical standardized home-based exercise program prescribed by a PM&R resident before initiation of the injection protocol. The exercise program consisted of multi-angle isometric strengthening exercises for quadriceps and hip adductor/abductor muscles together with hamstring stretching (10 repetitions with a 10-second hold, performed three times daily). After one month, participants progressed to closed-chain isotonic strengthening exercises. Participants were advised to maintain light to moderate daily activities and gradually increase activity levels as tolerated, while avoiding strenuous exercise.

Written instructions were provided, and patients were given a 24-hour contact number and WhatsApp messaging service for communication with the research team. Adherence to the exercise regimen was monitored through follow-up calls by PM&R residents. No supervised physiotherapy sessions were provided during the study period. These standardized recommendations were applied uniformly across all three treatment groups to minimize co-intervention bias.

2.5. Outcome measures

The primary outcome was the change from baseline in the WOMAC total score (Likert version 3.1, range 0–96) [43] at the 6-month follow-up. Secondary outcomes included changes in WOMAC pain, stiffness and physical function subscale scores, knee pain intensity assessed using a 10-cm visual analog scale (VAS) [44], and the Lequesne Algofunctional Index [45]. Clinical evaluations were performed by the blinded outcome assessor at baseline (prior to the first injection), and at 1 month and 6 months after the final injection. Patient satisfaction was assessed at the 6-month follow-up using a 5-point Likert scale (0 = very dissatisfied, 4 = very satisfied). Adverse events were prospectively recorded at each study visit and classified according to severity (mild, moderate, severe) and relatedness to treatment (unrelated, possibly related, probably related, definitely related).

2.6. Sample size calculation

The sample size was calculated using G*power software (Heinrich Heine University, Dusseldorf, Germany) based on detecting a clinically meaningful difference in the WOMAC total score among the three treatment groups (PRP, HC, and HA). A minimal clinically important difference of approximately 10 points on the WOMAC scale was considered clinically relevant. Given the use of the WOMAC Likert version (range 0–96), this corresponded to a standardized effect size of approximately 0.5, consistent with prior randomized trials of intra-articular injections in knee osteoarthritis [46]. Assuming a standard deviation of 18 points, a two-sided α of 0.05, and 80% power for one-way ANOVA with three parallel groups, a minimum of 38 participants per group was required. To maintain adequate power and account for potential attrition, 45 participants were enrolled in each group at randomization.

2.7. Statistical analysis

Baseline demographic and clinical characteristics were summarized using descriptive statistics. Continuous variables are presented as mean ± standard deviation (SD), whereas categorical variables are expressed as frequencies and percentages. The normality of quantitative variables was evaluated using the one-sample Kolmogorov–Smirnov test.

The primary analysis was conducted on the per-protocol population, including participants who completed the 6-month follow-up. An intention-to-treat analysis was not performed because complete follow-up outcome data were unavailable for participants lost to follow-up, precluding valid imputation without introducing additional assumptions. Within-group changes were analyzed using paired t-tests or Wilcoxon signed-rank tests when normality was not met. Between-group comparisons were performed using one-way analysis of variance (ANOVA) or the Kruskal–Wallis test, as appropriate. No imputation of missing outcome data was performed.

Mean changes from baseline to 1 month, baseline to 6 months and from 1 month to 6 months were calculated for each outcome. Pairwise between-group comparisons were expressed as mean differences (MDs) with 95% confidence intervals (CIs). To control the family-wise error rate resulting from multiple pairwise comparisons, p-values were adjusted using the Holm procedure within each outcome and time-period family.

As a sensitivity analysis, multivariable analyses were performed using change scores to account for potential baseline imbalances. Each model included treatment group together with clinically relevant baseline covariates, including age, sex, body mass index (BMI), symptom duration, Kellgren-Lawrence grade and the corresponding baseline value of the outcome.

All statistical tests were two-sided, and a p-value <0.05 was considered statistically significant. All analyses were performed using R software (R Foundation for Statistical Computing, Vienna, Austria).

2.8. Safety monitoring

Safety outcomes included the incidence and nature of adverse events (AEs) and serious adverse events (SAEs). An AE was defined as any unfavorable or unintended sign, symptom, or medical occurrence temporally associated with the injection, irrespective of causality. SAEs were identified and documented in accordance with the International Council for Harmonization Good Clinical Practice guideline (ICH E6[R2]) [47].

Participants were instructed to report any new symptoms throughout the study. Events were documented prospectively at each visit and through scheduled interim telephone follow-up. AEs were categorized by severity (mild, moderate, severe) and by relatedness to treatment (unrelated, possibly related, probably related, definitely related).

No independent data safety monitoring board was convened, and no predefined stopping rules were applied.

3. Results

3.1. Participant flow

A total of 175 individuals were screened for eligibility. Of these, 135 met the eligibility criteria and were randomized in a 1:1:1 ratio to receive platelet-rich plasma (PRP, n = 45), hydrolyzed collagen (n = 45), and hyaluronic acid (HA, n = 45). During follow-up, nine participants were not included in the final analysis due to protocol deviations or loss to follow-up, leaving 126 participants who completed the 6-month follow-up and were included in the per-protocol analysis.

In the HA group, five participants were not included in the final analysis: one received high-power laser therapy before the 1-month follow-up. Before the 6-month follow-up, two participants received intra-articular corticosteroid injections, one was lost to follow-up, and one received high-power laser therapy.

In the PRP group, three participants were excluded because two initiated knee physiotherapy and one underwent total knee arthroplasty before the 6-month assessment.

In the HC group, no participants withdrew at 1 month, and one participant received an intra-articular corticosteroid injection prior to the 6-month visit (total excluded: 1).

The participant flow is summarized in Figure 1.

Figure 1.

CONSORT flow diagram of participant enrollment, randomization, follow-up and analysis.

3.2. Baseline characteristics

Baseline demographic and clinical characteristics are summarized in Table 2. The treatment groups were generally similar with respect to most demographic and clinical variables. However, significant between-group differences were observed for symptom duration, baseline VAS score and baseline Lequesne pain score. These baseline imbalances were addressed in prespecified multivariable sensitivity analyses.

Table 2.

Baseline demographic and clinical characteristics of participants.

Characteristic	PRP (n = 42)	Collagen (n = 44)	HA (n = 40)	Total (n = 126)	P value
Age, years (mean ± SD)	55.81 ± 6.69	58.16 ± 7.83	57.15 ± 6.24	57.06 ± 6.99	0.298
Sex, male/female	14/28	15/29	13/27	42/84	0.988
BMI, kg/m² (mean ± SD)	27.41 ± 2.93	27.29 ± 3.04	27.35 ± 2.18	27.35 ± 2.74	0.978
K–L grade 2 / 3	19/23	24/20	23/17	66/60	0.506
Symptom duration, years (mean ± SD)	7.69 ± 2.67	9.32 ± 4.49	8.84 ± 2.30	8.62 ± 6.20	0.003
Side of injection, left/right	14/28	18/26	22/18	54/72	0.133
Prior physiotherapy, n (%)	16 (38.0)	15 (34.0)	13 (32.5)	44 (34.9)	0.969
Prior IA injection, n (%)	12 (28.6)	14 (31.8)	12 (30.0)	38 (30.16)	0.073
VAS, Mean ± SD	7.95 ± 1.17	7.45 ± 1.44	8.25 ± 1.10	7.87 ± 1.28	0.015
WOMAC, Mean ± SD
Pain	9.64 ± 1.61	9.41 ± 1.88	9.35 ± 1.64	9.47 ± 1.71	0.714
Stiffness	3.00 ± 1.43	2.75 ± 1.35	2.65 ± 1.05	2.80 ± 1.29	0.449
Function	30.14 ± 8.27	28.80 ± 6.48	31.03 ± 8.94	29.95 ± 7.92	0.431
Total	42.79 ± 9.93	40.32 ± 9.32	42.50 ± 11.25	41.83 ± 10.15	0.470
Lequesne, Mean ± SD
Pain	4.90 ± 1.23	4.80 ± 1.27	5.60 ± 0.96	5.09 ± 1.21	0.004
Walk	1.74 ± 0.86	2.18 ± 1.40	1.65 ± 0.95	1.87 ± 1.12	0.062
ADL	5.64 ± 0.70	5.31 ± 0.97	5.63 ± 0.79	5.52 ± 0.84	0.113
Total	12.29 ± 2.16	12.39 ± 2.02	12.63 ± 2.24	12.43 ± 2.13	0.763

Values are presented as mean ± standard deviation (SD) or number (percentage). P-values were calculated using one-way ANOVA for continuous variables and chi-square test for categorical variables.

Abbreviations: BMI = body mass index; IA = intra-articular; K–L = Kellgren–Lawrence osteoarthritis grade; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; VAS = Visual Analog Scale; ADL = activities of daily living.

3.3. Clinical outcomes

3.3.1. Changes in outcome measures within treatment groups at 1-month and 6-month follow-up

All treatment groups showed statistically significant within-group improvements from baseline to both 1-month and 6-month evaluations across all outcome measures (Table 3 and Figures 2–4).

Table 3.

Within-group changes in outcome measures at 1-month and 6-month follow-ups.

	Outcome Mean ± SD	Before intervention	After 1 month		After 6 months
			Value	p-valuea	Value	p-valuea
WOMAC Pain	PRP	9.64 ± 1.61	5.12 ± 2.04	<0.001	5.17 ± 2.17	<0.001
	Collagen	9.41 ± 1.88	4.75 ± 2.08	<0.001	4.82 ± 2.14	<0.001
	HA	9.35 ± 1.64	4.80 ± 1.59	<0.001	5.28 ± 1.62	<0.001
WOMAC Stiffness	PRP	3.00 ± 1.43	1.64 ± 1.27	<0.001	1.62 ± 1.46	<0.001
	Collagen	2.75 ± 1.35	1.61 ± 1.22	<0.001	1.52 ± 1.39	<0.001
	HA	2.65 ± 1.05	0.98 ± 0.83	<0.001	1.03 ± 0.86	<0.001
WOMAC Function	PRP	30.14 ± 8.27	19.83 ± 7.72	<0.001	18.38 ± 8.89	<0.001
	Collagen	28.80 ± 6.48	19.20 ± 6.56	<0.001	17.32 ± 6.83	<0.001
	HA	31.03 ± 8.94	19.70 ± 6.54	<0.001	20.18 ± 6.79	<0.001
WOMAC Total	PRP	42.79 ± 9.93	26.60 ± 10.30	<0.001	25.17 ± 12.09	<0.001
	Collagen	40.32 ± 9.32	25.57 ± 8.76	<0.001	23.66 ± 9.59	<0.001
	HA	42.50 ± 11.25	25.65 ± 8.08	<0.001	27.18 ± 9.20	<0.001
VAS	PRP	7.95 ± 1.17	2.88 ± 1.35	<0.001	3.76 ± 1.45	<0.001
	Collagen	7.45 ± 1.44	2.98 ± 1.50	<0.001	3.43 ± 1.37	<0.001
	HA	8.25 ± 1.10	2.68 ± 1.19	<0.001	3.78 ± 1.40	<0.001
LEQ Pain	PRP	4.90 ± 1.23	3.86 ± 1.56	<0.001	3.55 ± 1.67	<0.001
	Collagen	4.80 ± 1.27	3.41 ± 1.23	<0.001	3.14 ± 1.44	<0.001
	HA	5.60 ± 0.96	3.68 ± 1.19	<0.001	4.00 ± 1.26	<0.001
LEQ Walk	PRP	1.74 ± 0.86	1.26 ± 0.80	<0.001	1.29 ± 0.89	<0.001
	Collagen	2.18 ± 1.40	1.59 ± 0.97	<0.001	1.66 ± 0.91	<0.001
	HA	1.65 ± 0.95	1.13 ± 0.65	<0.001	1.23 ± 0.66	0.001
LEQ ADL	PRP	5.64 ± 0.70	4.49 ± 1.14	<0.001	4.23 ± 1.30	0.002
	Collagen	5.31 ± 0.97	4.14 ± 1.36	<0.001	3.81 ± 1.24	<0.001
	HA	5.63 ± 0.79	4.25 ± 1.01	<0.001	4.61 ± 0.98	<0.001
LEQ Total	PRP	12.29 ± 2.16	9.61 ± 3.00	<0.001	9.06 ± 3.25	<0.001
	Collagen	12.39 ± 2.02	9.18 ± 2.41	<0.001	8.67 ± 2.88	<0.001
	HA	12.63 ± 2.24	9.03 ± 2.06	<0.001	9.94 ± 2.44	<0.001

Values are presented as mean ± standard deviation (SD).

Abbreviation: WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

^ap-values within groups are from paired t-tests.

Figure 2.

Mean VAS scores within and between treatment groups at baseline, 1 month and 6 months.

Figure 3.

Mean WOMAC Pain, Stiffness, Function and Total scores within treatment groups and between-group improvements at baseline, 1 month and 6 months follow-up.

Figure 4.

Mean LEQ Pain, Walk, ADL and Total scores within treatment groups and between-group improvements at baseline, 1 month and 6 months.

3.3.2. Between-group comparison of outcome measures over different time periods

Between-group comparisons of changes in clinical outcomes are provided in Table 4.

Table 4.

Between-group comparisons of mean changes in outcome measures over different time periods.

		Baseline compared to 1st month		Baseline compared to 6th month		1st month compared to 6th month
Outcome	Treatment groups	MD (95% CI)	Multiple comparison*	MD (95% CI)	Multiple comparison*	MD (95% CI)	Multiple comparison*
WOMAC Pain	PRP	−4.52 (−5.2, −3.9)	PRP vs. Collagen:0.725	−4.48 (−5.2, −3.7)	PRP vs. Collagen: 0.717	0.05 (−0.35, 0.45)	PRP vs. Collagen: 0.637
	Collagen	−4.66 (−5.3, −4.0)	PRP vs. HA: 0.826	−4.59 (−5.4, −3.8)	PRP vs. HA: 0.341	0.07 (−0.56, 0.70)	PRP vs. HA: 0.148
	HA	−4.55 (−5.0, −4.1)	Collagen vs. HA: 0.809	−4.08 (−4.6, −3.6)	Collagen vs. HA :0.159	0.48 (0.19, 0.76)	Collagen vs. HA: 0.061
WOMAC Stiffness	PRP	−1.36 (−1.8, −0.89)	PRP vs. Collagen: 0.521	−1.38 (−1.9, −0.87)	PRP vs. Collagen: 0.562	−0.02 (−0.34, 0.30)	PRP vs. Collagen:>0.999
	Collagen	−1.14 (−1.6, −0.71)	PRP vs. HA: 0.09	−1.23 (−1.6, −0.82)	PRP vs. HA: 0.358	−0.09 (−0.44, 0.26)	PRP vs. HA: 0.434
	HA	−1.68 (−1.9, −1.4)	Collagen vs. HA: 0.011	−1.63 (−2.0, −1.3)	Collagen vs. HA: 0.067	0.05 (−0.17, 0.27)	Collagen vs. HA: 0.383
WOMAC Function	PRP	−10.31 (−12, −8.6)	PRP vs. Collagen: 0.661	−11.76 (−14, −9.5)	PRP vs. Collagen: 0.863	−1.45 (−2.6, −0.27)	PRP vs. Collagen: 0.270
	Collagen	−9.59 (−11, −8.1)	PRP vs. HA: 0.208	−11.48 (−13, −9.6)	PRP vs. HA: 0.682	−1.89 (−3.4, −0.35)	PRP vs. HA: 0.002
	HA	−11.33 (−13, −9.9)	Collagen vs. HA: 0.045	−10.85 (−12, −9.3)	Collagen vs. HA :0.500	0.48 (−0.36, 1.3)	Collagen vs. HA: <0.001
WOMAC Total	PRP	−16.19 (−18, −14)	PRP vs. Collagen: 0.416	−17.62 (−21, −15)	PRP vs. Collagen: 0.952	−1.43 (−3.0, 0.15)	PRP vs. Collagen: 0.252
	Collagen	−14.75 (−17, −12)	PRP vs. HA: 0.537	−16.66 (−20, −13)	PRP vs. HA: 0.472	−1.91 (−4.2, 0.34)	PRP vs. HA: 0.002
	HA	−16.85 (−19, −14)	Collagen vs. HA: 0.137	−15.33 (−17, −14)	Collagen vs. HA: 0.160	1.53 (−0.34, 3.4)	Collagen vs. HA: <0.001
VAS	PRP	−5.07 (−5.5, −4.6)	PRP vs. Collagen: 0.123	−4.19 (−4.7, −3.7)	PRP vs. Collagen: 0.678	0.88 (0.56, 1.2)	PRP vs. Collagen: 0.073
	Collagen	−4.48 (−5.0, −4.0)	PRP vs. HA: 0.145	−4.02 (−4.5, −3.6)	PRP vs. HA: 0.467	0.45 (0.09, 0.82)	PRP vs. HA: 0.261
	HA	−5.58 (−5.9, −5.2)	Collagen vs. HA :0.003	−4.48 (−5.0, −4.0)	Collagen vs. HA: 0.240	1.10 (0.79, 1.4)	Collagen vs. HA: 0.004
LEQ Pain	PRP	−1.05 (−1.4, −0.65)	PRP vs. Collagen: 0.222	−1.36 (−1.9, −0.84)	PRP vs. Collagen: 0.663	−0.31 (−0.64, 0.02)	PRP vs. Collagen: 0.982
	Collagen	−1.39 (−1.8, −1.0)	PRP vs. HA: <0.001	−1.66 (−2.1, −1.2)	PRP vs. HA: 0.752	−0.27 (−0.58, 0.04)	PRP vs. HA: 0.007
	HA	−1.93 (−2.2, −1.7)	Collagen vs. HA: 0.017	−1.60 (−1.9, −1.3)	Collagen vs. HA: 0.922	0.33 (0.12, 0.53)	Collagen vs. HA: 0.004
LEQ Walk	PRP	−0.48 (−0.66, −0.29)	PRP vs. Collagen: 0.327	−0.45 (−0.66, −0.24)	PRP vs. Collagen: 0.943	0.02 (−0.16, 0.21)	PRP vs. Collagen: 0.946
	Collagen	−0.59 (−0.81, −0.37)	PRP vs. HA: 0.865	−0.52 (−0.80, −0.24)	PRP vs. HA: 0.437	0.07 (−0.13, 0.27)	PRP vs. HA: 0.555
	HA	−0.53 (−0.74, −0.31)	Collagen vs. HA: 0.496	−0.43 (−0.66, −0.19)	Collagen vs. HA:0.573	0.10 (−0.06, 0.26)	Collagen vs. HA: 0.563
LEQ ADL	PRP	−1.15 (−1.5, −0.83)	PRP vs. Collagen: 0.413	−1.42 (−1.8, −1.0)	PRP vs. Collagen: 0.352	−0.26 (−0.55, 0.02)	PRP vs. Collagen: 0.335
	Collagen	−1.17 (−1.5, −0.89)	PRP vs. HA: 0.045	−1.50 (−1.9, −1.1)	PRP vs. HA: 0.186	−0.33 (−0.73, 0.07)	PRP vs. HA: 0.001
	HA	−1.38 (−1.6, −1.2)	Collagen vs. HA: 0.204	−1.01 (−1.2, −0.78)	Collagen vs. HA: 0.007	0.40 (0.18, 0.55)	Collagen vs. HA: <0.001
LEQ Total	PRP	−2.68 (−3.3, −2.0)	PRP vs. Collagen: 0.111	−3.23 (−4.0, −2.4)	PRP vs. Collagen: 0.319	−0.55 (−1.1, 0.01)	PRP vs. Collagen: 0.276
	Collagen	−3.20 (−3.9, −2.5)	PRP vs. HA: 0.013	−3.72 (−4.5, −2.9)	PRP vs. HA: 0.175	−0.51 (−1.2, 0.14)	PRP vs. HA: <0.001
	HA	−3.60 (−4.1, −3.1)	Collagen vs. HA: 0.469	−2.69 (−3.2, −2.2)	Collagen vs. HA: 0.014	0.91 (0.40, 1.4)	Collagen vs. HA: <0.001

Values are presented as mean difference (MD) with 95% confidence interval (CI). *Between-group comparisons were performed using independent t-tests on the mean changes from baseline. Bold p-values indicate statistically significant differences (p < 0.05).

Abbreviations: WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; VAS = Visual Analog Scale; LEQ = Lequesne Index; ADL = activities of daily living.

From baseline to 1 month, all treatment groups demonstrated marked improvements. Compared with HC, HA showed significantly greater improvements in WOMAC stiffness, WOMAC function, VAS, Lequesne pain and Lequesne total scores. Compared with PRP, HA also demonstrated significantly greater improvements in Lequesne pain, Lequesne ADL scores.

From baseline to 6 months, HC produced significantly greater improvements than HA in Lequesne ADL and total scores, while differences in other measures remained numerical and did not reach statistical significance.

During the interval 1 to 6 months, PRP and HC demonstrated greater improvements than HA across several clinical outcomes including WOMAC function and total scores, VAS pain, Lequesne pain, ADL and total scores.

Patient satisfaction at the 6-month follow-up also differed across treatment groups. The HC group reported the highest satisfaction (2.90 ± 1.25), followed by the PRP group (2.84 ± 1.31), whereas the HA group reported the lowest satisfaction (2.24 ± 1.28).

3.3.3. Multivariable sensitivity analyses

To assess the robustness of the primary findings and account for baseline imbalances, prespecified multivariable sensitivity analyses were performed for changes from baseline to 1 month, from baseline to 6 months and from 1 month to 6 months. The models were adjusted for the corresponding baseline outcome, age, sex, BMI, Kellgren-Lawrence grade and symptom duration. Overall, the adjusted analyses were consistent with the primary analyses. During the interval from 1 to 6 months, the superiority of PRP and/or HC remained directionally consistent after adjustment, whereas no significant differences were observed between the PRP and HC groups. Detailed adjusted estimates are presented in Supplementary Table S1.

3.4. Adverse events

No serious adverse events (SAEs) or treatment-related complications were observed throughout the study period. All adverse events reported were localized and self-limited. Injection-site pain during or immediately after the procedure, as well as mild local post-injection reactions (including transient swelling, warmth and erythema) were the only adverse events reported. These reactions occurred more frequently in the PRP group compared with the collagen and HA groups but were not statistically significant.

All events were classified as mild or moderate in severity and resolved spontaneously within 48 hours without requiring medical intervention. No systemic, infectious, or delayed adverse events were observed at any point during the 6-month follow-up period (Table 5). Importantly, no participant discontinued treatment due to an adverse event, and no adverse event contributed to loss to follow-up.

Table 5.

Summary of treatment-related adverse events.

Adverse event	PRP (n = 42)	Collagen (n = 44)	HA (n = 40)
Injection-site pain, n (%)	14 (33.3%)	8 (18.2 %)	6 (15.0%)
Local swelling or erythema, n (%)	2 (4.8%)	6 (13.6%)	1 (2.5%)
Systemic reactions, n (%)	0 (0%)	0 (0%)	0 (0%)
Serious adverse events (SAEs), n (%)	0 (0%)	0 (0%)	0 (0%)
Discontinuation due to AE, n (%)	0 (0%)	0 (0%)	0 (0%)

Values are presented as number (percentage). All adverse events were classified as mild or moderate and were assessed for relatedness to treatment.

Abbreviation: SAE = serious adverse event.

4. Discussion

This randomized, assessor- and statistician-blinded controlled trial compared the clinical efficacy of ultrasound-guided intra-articular platelet-rich plasma (PRP), hydrolyzed type I collagen (HC), and hyaluronic acid (HA) injections in patients with mild-to-moderate knee osteoarthritis (Kellgren–Lawrence grades 2–3). All three interventions were associated with significant improvements in pain, stiffness, and function and overall clinical status at both 1-month and 6-month assessments.

The principal finding of this study is that the temporal pattern of treatment response differed among the three interventions. HA produced greater short-term improvements at the 1-month follow-up, whereas PRP and HC demonstrated a more sustained pattern of clinical benefits during the interval from 1 to 6 months. By the 6-month assessment, HC showed greater improvement than HA in Lequesne ADL and total scores, whereas PRP and HC demonstrated comparable clinical effectiveness. Importantly, these findings remained materially unchanged after prespecified multivariable sensitivity analyses adjusting for baseline imbalances, supporting the robustness of the primary findings.

All three interventions were well tolerated. No serious adverse events or treatment-related complications occurred during the study, and all reported adverse events were mild or moderate, self-limited, and resolved without medical intervention. No participant discontinued treatment due to an adverse event.

Our findings are broadly consistent with the existing literature comparing intra-articular biologic therapies for knee osteoarthritis. Several recent high-quality meta-analyses of randomized controlled trials have demonstrated that intra-articular PRP provides greater mid-term improvements in pain and function than HA in KOA. While early (1–3 months) differences between treatments are generally small or not statistically significant, this temporal pattern (early similarity followed by divergence at mid-term) closely parallels the trajectory observed in the present study, in which all groups improved at 1 month, but PRP demonstrated greater or more sustained improvements than HA in several clinical outcomes at the 6-month follow-up, although statistically significant between-group differences were limited to selected outcome measures [31–34].

International clinical practice guidelines also reflect this interpretation. The 2021 American Academy of Orthopedic Surgeons (AAOS) guideline on non-arthroplasty management of KOA notes that the evidence for HA is modest and associated with only small effect sizes, whereas PRP is increasingly viewed as a promising biologic therapy capable of providing more sustained symptomatic improvements, despite ongoing heterogeneity in preparation methods and optimal protocols [48].

In contrast, the evidence base for intra-articular HC remains comparatively limited. Initial clinical investigations, including retrospective analyses and small pilot trials, have suggested that collagen injections are associated with meaningful improvements in pain and function with a favorable safety profile [35,36].

Recently, two retrospective comparative studies have been published. Heredia-Sulbaran and Gomez reported greater 12-month WOMAC improvement with hydrolyzed collagen than HA and outcomes comparable to PRP in symptomatic KOA; however, the analysis was retrospective and non-randomized, with unequal injection regimens across groups (three collagen or PRP injections versus a single HA injection) and non-standardized treatment intervals, limiting internal validation and comparability. Similarly, In et al. found that intra-articular atelocollagen was non-inferior to HA across two injection cycles, but treatment allocation was determined by patient-physician decision within a retrospective cohort of limited size, without randomization. Therefore, direct comparison of treatment efficacy across injectates in these studies should be interpreted cautiously [49,50].

Thus, head-to-head randomized trials directly comparing hydrolyzed collagen with established intra-articular injectates remain scarce. The present study contributes novel comparative evidence by showing that hydrolyzed collagen achieved mid-term improvements comparable to PRP and greater improvements than HA in selected clinical outcomes.

These findings strengthen emerging evidence supporting hydrolyzed collagen as a viable biologic injectable option for management of knee osteoarthritis while underscoring the need for larger randomized trials to confirm these findings.

The temporal response pattern observed across the three treatment groups may be interpreted in the light of the known biological properties and mechanistic profiles of each injectate. HA primarily functions as a viscoelastic supplement, improving joint lubrication, shock absorption, endogenous HA production, and mechanoreceptor modulation. These effects may yield relatively rapid but often transient symptomatic relief [23,51]. This mechanistic profile is consistent with the greater short-term improvements observed in the HA group at 1 month assessment, although these effects were not sustained over time.

PRP, in contrast, provides a concentrated mixture of autologous growth factors and cytokines, including platelet-derived growth factor, transforming growth factor-β, and vascular endothelial growth factor. These factors contribute to anti-inflammatory signaling, modulation of synovial homeostasis, and promotion of anabolic and reparative processes within cartilage and periarticular tissues [27,52]. These biological processes evolve more gradually than the mechanical effects of HA and may underlie the more sustained symptomatic improvements observed at 6 months in the PRP group. The delayed yet more durable pattern of clinical response observed in the PRP group is consistent with this slower biologic activation profile.

Hydrolyzed type I collagen consists of low–molecular-weight bioactive peptides capable of diffusing across the synovial membrane and interacting with chondrocytes. Preclinical and early clinical studies suggest that collagen peptides can stimulate extracellular matrix synthesis, downregulate pro-inflammatory mediators, and enhance chondrocyte metabolism, thereby supporting joint homeostasis [35,36]. These mechanisms provide a biologically plausible explanation for the mid-term benefits observed in the collagen group, which were comparable to PRP and greater than those observed with HA in selected clinical outcomes.

Collectively, these mechanistic differences suggest that HA may be best suited for achieving faster short-term symptom relief, whereas PRP and hydrolyzed collagen may be more effective when a sustained therapeutic response is desired. This differential temporal profile may support an individualized, patient-centered approach to selecting intra-articular injectates based on clinical priorities and expected treatment timelines. Notably, these temporal differences remained consistent after multivariable adjustment for baseline imbalances, supporting the robustness of these mechanistic interpretations.

The results of this trial provide several clinically meaningful insights for the management of symptomatic knee osteoarthritis. All three injectates (PRP, HA, and hydrolyzed type I collagen) produced clinically significant improvements in pain, stiffness, and function over the 6-month follow-up, and no serious adverse events were observed, supporting their acceptable short- and mid-term safety and suitability for outpatient use. The mild and self-limited nature of local post-injection reactions further supports the favorable tolerability of these treatments.

The temporal pattern of treatment response suggests that the choice of injectate may be tailored to individual clinical goals. HA may be preferable when rapid short-term symptom relief is the primary objective, as reflected by its greater early improvements.

In contrast, PRP and hydrolyzed collagen may offer a more sustained pattern of clinical benefits than HA over the 6-month follow-up. Importantly, PRP and hydrolyzed collagen demonstrated comparable clinical effectiveness, providing clinicians with flexibility when selecting between these biologic treatment options based on patient characteristics, availability and cost considerations.

These findings further support the role of intra-articular injectables as part of a multimodal rehabilitation strategy, particularly for patients with symptomatic knee osteoarthritis, who have not responded adequately to first-line treatment options. Importantly, no participant discontinued treatment due to an adverse effect, which further supports the acceptable safety profile of all three interventions over the 6-month follow-up period. Integrating these intra-articular therapies within comprehensive, individualized management plans may help optimize pain control, functional outcomes, and patient satisfaction.

This trial had several methodological strengths that enhance confidence in its findings, including the direct head-to-head comparison of PRP, HC and HA under an identical injection schedule. It was a randomized, assessor- and statistician-blinded, parallel-group study, and despite some attrition after randomization, all 126 analyzed participants completed the planned follow-up assessments, and no treatment discontinuations due to adverse effects occurred. Injections were performed under ultrasound guidance by an experienced physician, and co-interventions were standardized across groups. Outcomes were assessed at prespecified time points using validated measures, including the WOMAC, VAS, and Lequesne Index, which strengthens internal validity and supports rigorous comparison among the three injectates. Furthermore, prespecified multivariable sensitivity analyses with Holm-adjusted p values were performed to account for baseline imbalances and control for multiplicity. The consistency between the adjusted and primary analyses further strengthened the robustness of the study findings.

However, some limitations should be acknowledged. Complete blinding was not feasible because of the inherent differences in injectate preparation and administration, particularly for platelet-rich plasma. To minimize potential bias, outcome assessment and statistical analyses were performed by blinded assessors and a blinded statistician. The study was restricted to individuals with Kellgren–Lawrence grade 2–3 osteoarthritis and BMI < 30, which may reduce generalizability of the findings to broader KOA populations. Rescue acetaminophen use was not systematically recorded and therefore could not be incorporated into statistical analyses. Differential analgesic use may have influenced pain-related outcomes, including VAS and WOMAC pain scores, representing a potential source of unmeasured co-intervention bias. Nevertheless, rescue medication recommendations were standardized across all groups, NSAIDs and corticosteroids were prohibited throughout the follow-ups, and the consistency of treatment effects across multiple outcome domains and adjusted sensitivity analyses supports the main findings. Future studies should prospectively record rescue analgesic consumption and account for it in statistical analyses. Also, the absence of imaging or biomarker measures limits mechanistic interpretation. Although the attrition rate was low, the absence of a formal intention-to-treat analysis represents a methodological limitation and should be considered when interpreting the findings. Considering various sources of collagen preparations (bovine, avian, and porcine) that are available, our study was confined to type I hydrolyzed collagen. Religious or cultural considerations across different nationalities should be accounted for when using different collagen sources in clinical practice. Finally, the 6-month follow-up limits conclusions regarding the longer-term durability and safety of treatments, particularly given the evolving nature of biologic injectates. Accordingly, the present findings should be interpreted as reflecting mid-term rather than long-term outcomes.

Future studies should address the current trial’s limitations through larger, multicenter randomized controlled trials with standardized treatment protocols as well as follow-up of 12 to 24 months to assess durability. Trials designed to directly compare PRP and HC with harmonized preparation methods would help clarify relative efficacy and inform optimal patient selection.

Dose–response studies exploring PRP composition (including platelet concentration and leukocyte content) and collagen characteristics (such as molecular weight, source, and peptide profile) could optimize protocols.

Beyond optimization of current injectable therapies, future research should increasingly integrate advances in osteoarthritis biology into targeted clinical interventions. Emerging insights into cellular senescence, extracellular vesicles, immune modulation, regenerative biomaterials and disease-modifying molecular pathways may facilitate the development of mechanism-based therapies capable of modifying disease progression rather than providing only symptomatic relief. Translational studies linking these preclinical advances with well-designed clinical trials will be essential to identify patient subgroups most likely to benefit from targeted biologic therapies [53,54].

Furthermore, future studies incorporating imaging modalities, biochemical biomarkers, and patient-centered outcomes, such as return to daily activities and quality of life, would provide a more comprehensive assessment of the therapeutic impact of these interventions. Cost-effectiveness analyses have previously been conducted for PRP and HA [55], and performing similar analyses for hydrolyzed collagen may further help contextualize the role of these injectates within the contemporary multimodal management of knee osteoarthritis.

5. Conclusion

This randomized assessor- and statistician-blinded clinical trial provides one of the first head-to-head comparisons of hydrolyzed type I collagen with established intra-articular injectables for mild-to-moderate knee osteoarthritis.

HC achieved clinical outcomes comparable to platelet-rich plasma while demonstrating superior mid-term improvements to HA in several outcome measures, with an acceptable short- and mid-term safety profile. Although HA may offer greater early symptomatic improvement, PRP and HC appear to provide more sustained clinical benefits over the 6-month follow-up.

Collectively, these findings support HC as a promising biologic injectable and expand the evidence base for individualized treatment selection in patients with mild-to-moderate knee osteoarthritis.

6. Summary

This randomized clinical trial demonstrated that ultrasound-guided intra-articular injections of platelet-rich plasma (PRP), hydrolyzed type I collagen (HC) and hyaluronic acid (HA) all produced significant improvements in pain, stiffness, function and overall clinical status in patients with mild-to-moderate knee osteoarthritis. The temporal pattern of treatment response differed among the interventions, with HA providing greater short-term improvements, whereas PRP and HC demonstrated more sustained mid-term clinical benefits during the interval from 1 month to 6 months. Importantly, these findings remained consistent after prespecified multivariable sensitivity analyses, supporting the robustness of the primary findings. All three interventions were well tolerated, with no serious adverse events and no treatment discontinuations related to adverse events.

Funding Statement

This paper was not funded.

Ethics approval

All procedures were conducted in compliance with the ethical principles outlined in the Declaration of Helsinki. Approval was granted by the Ethics Committee of Shahid Beheshti University of Medical Sciences (IR.SBMU.MSP.REC.1403.102). The trial was registered with the Iranian Registry of Clinical Trials (IRCT20130523013442N34), on 08 February 2024, prior to enrollment. The anonymity of participants and data confidentiality were strictly maintained throughout the study and all the individuals provided written informed consent with the right to withdraw from the study at any stage. All authors have read and approved the final manuscript and consent to its publication.

Use of artificial intelligence tools

The authors used ChatGPT (OpenAI, GPT-5.3 and GPT-5.5) solely for language editing, refinement of the manuscript wording, and preparation the response letter during the revision process. All scientific decisions, study design, data collection, statistical analyses, interpretation of the results and the final approval of the manuscript were performed exclusively by the authors.

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in this manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending or royalties.

Data availability statement

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.