Intraosseous Infiltration of Platelet-Rich Plasma for Knee Osteoarthritis: A Systematic Review of Literature and Limited Meta-analysis

Abstract

Background

The role of subchondral bone as a key mediator in pathogenesis of osteoarthritis (OA) is emerging and various injectable modalities are being tried to address the subchondral pathology. Intraosseous (IO) infiltrations of PRP is biological way of addressing the pathology. This review aimed to explore the role of IO PRP in OA knees.

Methods

A systematic literature search was conducted on the electronic databases of PubMed, Embase, Scopus, Web of Science and Ovid for relevant articles on IO PRP with a predefined search strategy. Data from eligible studies were extracted and was analyzed.

Results

There were only five studies on the study topic which included 112 patients (112 knees) who received IO PRP injection. IO PRP was associated with significant improvements in VAS, KOOS and WOMAC scores. The most commonly reported complication was pain at injection site. Meta-analysis revealed that combined intraarticular (IA) and IO PRP were associated with significantly better WOMAC scores, compared to only IA PRP. However, VAS and KOOS scores were equivocal.

Conclusion

IO PRP infiltration for OA knees is safe and effective and results in improved patient-reported outcome scores based on the current limited literature. However, its added advantage over only IA PRP is not well established and clear due to limited available published data.

Introduction

With the continued increasing life expectancy globally, the prevalence of knee osteoarthritis (OA) has increased which has caused significant social, economic and health-related burden worldwide [1, 2]; it accounts for almost four fifths of the burden of OA worldwide [3, 4]. There are several non-pharmacological, pharmacological, and surgical modalities used for treatment in different stages of OA knee; however, surgery in the form of partial or total knee replacement is most commonly used for advanced stages of OA knee [5, 6].

Through the dynamically evolving research in the treatment of OA knee, a newer therapy coming into vogue, which targets the disease process is platelet-rich plasma (PRP) [7, 8]. PRP is an orthobiologic prepared from the patient’s own blood, and contains 4–5 times higher concentration of platelets than the base line levels in blood. PRP is rich in bioactive molecules and growth factors which potentially modify the disease process of OA knee [8, 9].

PRP is conventionally given as an IA injection, acts on the cartilage and synovial membrane altering the biological environment, and thereby slowing the progression of the degenerative changes [10]. Multiple clinical trials have established the safety and positive role of PRP injections.[7] Strategies to improve the efficacy of PRP are being tried and focuses on the type of PRP, number of injections and modification by combining with other agents.[11] One of the strategies to improve the efficacy is by IO infiltration of PRP to target the subchondral bone pathology and was described and popularized by Sanchez et al. [9] It is given in supine position under local infiltration of subcutaneous plane and periosteum, with 2 ml of 2% lignocaine hydrochloride; puncture is performed with 13-G trocar (a bone marrow biopsy needle) on the tibial plateau and femoral condyle at an angle of 45⁰ to the long axis of the limb, under fluoroscopic guidance, the tip being placed within 1 cm of the joint line at the desired position in the subchondral bone. 2–5 ml PRP injection is administered slowly into the condyles; dressing is done and compression bandage is applied. Patient is advised partial weight bearing on the limb for 24 h with ice application. [9, 12, 13]

They have postulated that a combination of IA PRP and IO PRP helps in targeting all pathologic sites, wherein the IA PRP would target the synovium and superficial zone of cartilage; whereas IO PRP would target the deeper zone of cartilage and subchondral bone. Subchondral injections are administered to medial tibial and femoral condyle. They have extended the use to advanced stages of OA too wherein traditionally IA PRP have poorer results. The addition of IO infiltration of PRP to target this specific anatomical region could potentially lead to additional therapeutic benefits. The purpose of this systematic review was to explore the role of IO PRP injection in improving clinical outcomes in OA knees.

Methodology

Study Design

This systematic review was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. [14] A protocol of the systematic review was formulated and registered on PROSPERO vide registration number CRD42021262681.

Search Strategy

The primary electronic search was conducted on Medline (PubMed), Embase, Scopus, Web of Science, and Ovid databases for the published literature from the year of inception to 22nd February 2022 by two authors (RKR and SP) independently, using a well-defined search strategy priorly formulated (Table 1), without any initial restriction on the language and country of publication. A secondary search of bibliography of included studies from the primary search, as well as bibliography of any review articles was done for any additional studies. Finally, a total of 2277 results were obtained.

Table 1.

Search strategy used for literature search

Database	Period-inception to 22nd of February 2022 with Keywords	Results
Medline (PubMed)	(((“intraosseal” [All Fields] OR “intraosseous” [All Fields] OR “intraosseously” [All Fields]) AND (“inject”[All Fields] OR “injectability” [All Fields] OR “injectant” [All Fields] OR “injectants” [All Fields] OR “injectate” [All Fields] OR “injectates” [All Fields] OR “injected” [All Fields] OR “injectible” [All Fields] OR “injectibles” [All Fields] OR “injecting” [All Fields] OR “injections” [MeSH Terms] OR “injections” [All Fields] OR “injectable” [All Fields] OR “injectables” [All Fields] OR “injection” [All Fields] OR “injects” [All Fields]) AND (“platelet rich plasma” [MeSH Terms] OR (“platelet rich” [All Fields] AND “plasma” [All Fields]) OR “platelet rich plasma” [All Fields] OR (“platelet” [All Fields] AND “rich” [All Fields] AND “plasma” [All Fields]) OR “platelet rich plasma” [All Fields])) OR (“pharmacol res perspect” [Journal] OR “prp” [All Fields])) AND (“osteoarthritis, knee” [MeSH Terms] OR (“osteoarthritis” [All Fields] AND “knee” [All Fields]) OR “knee osteoarthritis” [All Fields] OR (“knee” [All Fields] AND “osteoarthritis” [All Fields]))	478
Embase	(intraosseous AND injection AND platelet AND rich AND plasma OR prp) AND knee AND osteoarthritis	730
Scopus	(ALL (intraosseous AND injection) AND ALL (platelet AND rich AND plasma) OR ALL (prp) AND ALL (knee AND osteoarthritis))	217
Web of Science	((Intraosseous injection) AND (Platelet Rich Plasma) OR (PRP) AND (Knee osteoarthritis))	558
Ovid	(((Intraosseous injection and Platelet Rich Plasma) or PRP) and Knee osteoarthritis).af	294
Total		2277

Inclusion and Exclusion Criteria

The current review included any prospective or retrospective studies in English language, which have evaluated the outcomes of IO injection of platelet-rich plasma in knee osteoarthritis. Studies that evaluated outcomes of any other IO injections, case reports, conference abstracts, e posters, book chapters, reviews, tips and tricks, animal studies, and articles not in English language were excluded.

Study Selection

All the studies were screened based on their titles and abstracts, independently by three authors (PK, VB, RKR) and ones related to the study question were identified. Subsequently their full texts were accessed, and relevant studies based on the inclusion and exclusion criteria were included in the current review. Any discrepancies between the authors were resolved by mutual agreement.

Data Extraction

The data extraction was performed by three independent authors (RKR, PK,VK) from each included article and were tabulated mentioning the names of the first authors, year of publication, level of evidence, total number of patients included, mean follow-up period, degree of knee alignment, body mass index (BMI), relevant demographic parameters, Kellgren–Lawrence or Ahlback–Lawrence grades, primary and secondary outcome measures of interest, and information regarding risk of bias (Tables 2, 3). All the selected articles were finally reviewed and discussed by all the authors of the current review to reduce all possible operator-dependent bias. Finally, five studies relevant to the study question were included in the current review (Fig. 1).

Table 2.

Characteristics of included studies

Sl. no	References	Study Design	Level of Evidence	Treatment Groups	Number of Patient/Cases	Male/Female	Mean age in Years with SD	BMI	PRP Preparation Technique	Number and Site of IO PRP Injections	Mean Follow up
1	Sanchez et al. [12]	Prospective	IV	Intraarticular + Intraosseous PRP	13	4/9	62.23 ± 9.6	20–33	Leukocyte-poor PRP	3 injections, weekly, 1st IA and 2 IO into medial femoral condyle and tibial plateau	6 months
2	Su et al. [18]	RCT	II	Intraarticular + Intraosseous PRP	27	10/17	51 ± 90	28.2 ± 1.3	Leukocyte-rich PRP	2 IA + IO (into medial tibial plateau and femoral condyle) injections at 14 days interval	18 months
				Intraarticular PRP	25	11/14	54 ± 70	28.17 ± 1.43		2 injections at an interval of 14 days	18 months
				Intraarticular HA	30	12/18	53 ± 60	28.69 ± 1.13		5 injections at an interval of 7 days	18 months
3	Sanchez et al. [19]	Prospective	III	Intraarticular PRR	30	17/13	67.9 ± 7.3	30.9 ± 5.6	Leukocyte-poor PRP	3 injections on a weekly basis	12 months
				Intraarticular + Intraosseous PRP	30	12/18	63.4 ± 9.0	30.7 ± 5.6		Total-3, 1st IO + IA (IO into the medial tibial plateau and medial femoral condyle) followed by 2 more IA injections in the following 2 weeks	12 months
4	Lychagin et al. [20]	Prospective	IV	Intraosseous PRP	17	7/10	41.7 ± 14.3	18–35	Not mentioned	Single injection, site not mentioned	12 months
				Control (healthy volunteers)	17	10/7	40.3 ± 12.1	NR		–	12 months
5	Barman et al. [13]	RCT	I	Intraarticular + Intraosseous PRP	25	8/17	57.12 ± 4.27	25.44 ± 2.39	Leukocyte-rich PRP	One injection medial femoral condyle, 2-cm proximal to the joint line; & at medial tibial condyle, 2-cm distal to the joint line	6 months
				Intraarticular PRP	25	9/16	57.0 ± 4.96	25.63 ± 2.46		One injection	6 months

PRP platelet-rich plasma, IA Intraarticular, IO Intraosseous, HA Hyaluronic Acid, BMI Body mass index, RCT Randomized control trial

Table 3.

Data reporting study outcomes

Sl no	References	Treatment Groups	Kellgren–Lawrence	Ahlback Scale	VAS Score		WOMAC (pain)		KOOS		Complications	Conversion to Arthroplasty
					Base line	End Point	Base Line	End Point	Base Line	End Point
1	Sanchez et al. [12]	Intraarticular + Intraosseous PRP	NR	III-9, IV-4	6.77 ± 1.75	2.88 ± 2.48	NR	NR	61.55 ± 14.11	74.60 ± 19.19	Fever/Flu-1 Pain-1	2/13 (15.4%)
2	Su et al. [18]	Intraarticular + Intraosseous PRP	II-16 III-11	NR	7.09 ± 0.31	3.78 ± 0.27	50.15 ± 1.10	36.41 ± 1.74	NR	NR	Pain-1 Swelling-1	NR
		Intraarticular PRP	II-13 II-12	NR	7.02 ± 0.27	6.66 ± 0.31	50.17 ± 1.60	48.07 ± 1.9	NR	NR	Pain-1 Swelling-2 Headache-4%	NR
		Intraarticular HA	II-14 III-16	NR	7.04 ± 0.33	6.54 ± 0.34	49.88 ± 1.54	46.88 ± 3.8	NR	NR	Pain-1 Swelling-1	NR
3	Sanchez et al. [19]	Intraarticular PRP	NR	III-29, IV-1	4.7 ± 1.60	5.0 ± 2.1	33.5 ± 11.8	33.9 ± 14.9	53.2 ± 14.8	53.2 ± 21.7	None	NR
		Intraarticular + Intraosseous PRP	NR	III-27, IV-3	4.1 ± 1.6	3.0 ± 1.6	26.5 ± 11.9	19.9 ± 11.4	56.7 ± 15.2	67.7 ± 17.2	None	NR
4	Lychagin et al. [20]	Intraosseous PRP	II-5 III-10 IV-2	NR	5.14 ± 0.69	1.11 ± 0.93	65.1 ± 6.42	55.38 ± 12.85	52.78 ± 13.38	77.03 ± 7.48	NR	NR
		Control (healthy volunteers)	I-12 II-5	NR	NR	NR	NR	NR	NR	NR	NR	NR
5	Barman et al. [13]	Intraarticular + Intraosseous PRP	III-25	NR	6.46 ± 0.64	3.66 ± 1.02	NR	NR	49.88 ± 7.60  − 20.2	70.08 ± 9.79	Pain-23 Stiffness-22 Swelling-19 Headache-22	NR
		Intraarticular PRP	III-25	NR	6.38 ± 0.82	3.84 ± 0.75	NR	NR	49.92 ± 8.26 − 28.24	67.32 ± 10.64	Pain-9/25 Stiffness-5/25 Swelling-2/25 Headache-5/25	NR

HA Hyaluronic Acid, PRP Platelet-rich plasma, VAS Visual Analog Scale, WOMAC Western Ontario and McMaster Universities, KOOS Knee injury and Osteoarthritis Outcome Score, NR Not Reported

Fig. 1.

PRISMA flow diagram for the study selection

Outcome Measure

The primary outcomes of interest were patient-reported outcomes -Visual Analog Scale (VAS), Western Ontario and McMaster Universities (WOMAC), Knee injury and Osteoarthritis Outcome Score (KOOS), and Conversion to knee arthroplasty. The secondary outcomes of interest were any complication during and/or related to treatment.

Statistical Analysis

The statistical analysis was done using Review Manager Software version 5.4. [15] Wherever appropriate meta-analysis was performed if two or more studies reported a specific outcome of interest. To describe the measure of treatment effects, mean difference or standard mean difference was used for continuous variables and odds ratio or relative risk was used for dichotomous variables. All the results were expressed along with 95% confidence intervals. Forest plots was used to provide a visual summary for each outcome of interest. Reasons for clinical heterogeneity, if any, were also explored. Statistical heterogeneity was evaluated by the I² test. Fixed-effects model was used if the heterogeneity was considered low, otherwise random-effects model was used.

Risk of Bias Assessment

Risk of bias of the included studies was assessed independently by three observers (PK, RKR, SP) using The Cochrane Collaboration’s risk of bias tool’ for randomized controlled trial (RCTs) [16] and MINORS tool consisting of 8 items for non-randomized comparative studies. [17]. Any discrepancies were resolved by mutual agreement.

Results

Literature Search and Screening

A total of 2277 results were obtained from all the five databases searched. After excluding duplicates and studies not relevant to the current review 107 studies were identified for further evaluation. After screening of titles and abstracts, 10 studies were isolated and their full texts were assessed for eligibility. Finally, 5 studies were included. (Fig. 1).

Characteristics of the Studies

All the five included studies were prospective in design. [12, 13, 18–20] One initial study from same authors [20] with same set of patients was excluded and their subsequent study with 12-month follow-up was included. The level of evidence was I [13], II [18] and III [19] in one study each and level IV in two studies [12, 20]. Four studies used IA + IO PRP [12, 13, 18, 19] while one study used only IO PRP [20]. Four studies were comparative studies [12, 13, 18, 19] while in one study control used was healthy human volunteers. [20] Su et al. compared IA + IO PRP with isolated IA PRP injection and isolated IA hyaluronic acid (HA) injections [18]. Sanchez et al. compared combined IA and IO PRP injections with only IA PRP injection [19]. Lychagin et al. compared IO PRP injection with a control group that did not receive any intervention [20]. Barman et al. in a recent RCT compared combined IA and IO PRP injections with only IA PRP injection. [13] The last one was a cohort study evaluating the outcomes of combined IO and IA PRP [12].

A total of 112 patients with (112 knees) received IO PRP infiltration of which 41 were male and 71 were females. The overall mean age ranged from 41.7 to 63.4 years across five studies. The mean body mass index of the included patients ranged from 18 to 30.7. The mean follow-up period ranged from 6 to 18 months. Kellgren–Lawrence grading was reported by three studies [13, 18, 20] while Ahlback–Lawrence Scale grading of knees was used by other two [12, 19] (Kellgren–Lawrence grade II- 21, III-46, and IV-2; Ahlback–Lawrence Scale grading III-36 and IV-7.)

Assessment of Risk of Bias

There were 2 RCTs [13, 18] and three studies were non-randomized prospective in design. [12, 19, 20] The risk of bias calculated for RCTs was found low, and the MINORS tool Score was > 20 for two non-randomized prospective comparative studies. [19, 20] The overall risk of bias assessment of all included studies is shown in Figs. 2 and 3.

Fig. 2.

a Risk of bias graph for RCTs, b Non-comparative study, and c non-randomized comparative studies

Fig. 3.

a Risk of bias summary for RCTs, b Non-comparative study, and c non-randomized comparative studies

Primary Outcomes

Patient-Reported Outcomes

1. Visual Analog Scale (VAS) for pain

In their cohort study, Sanchez et al. [12] reported statistically significant improvements in VAS score with the use of combined IO and IA PRP, from a mean of 6.77 at baseline to 2.88 after 6 months. In another comparative study by Sanchez et al. [19] found that the use of isolated IA PRP injection did not lead to any significant improvement in VAS at 2 and 6 months; whereas the group that received combined IA and IO PRP demonstrated significant improvement from baseline. Su et al. [18] reported that combined IA and IO PRP led to significantly improved VAS pain scores when compared to isolated IA PRP and HA injections at 18 months follow-up.

Lychagin et al. [20] reported that IO infiltration of PRP led to statistically significant reduction of pain based on the VAS score compared to a control group that did not receive any intervention at 12 months follow-up. Berman et al. [13] reported a greater reduction in pain based of VAS score in combined IA and IO PRP injection group at the end of 6 months during follow-up compared to alone IA PRP injection. However, they did not find a statistically significant difference between two groups.

• 2.Western Ontario and McMaster Universities (WOMAC)

Su et al. [18] reported significant improvements in WOMAC pain subscale in the group that received both IA and IO PRP at every stage of follow-up up to 18 months, compared to isolated IA PRP or HA. Sanchez et al. [19] observed that combined IA and IO PRP led to significant improvement in WOMAC scores at 2, 6 and 12 months, whereas no such improvement was noted in the group that received only IA PRP [11]. Lychagin et al. [20] noted that WOMAC scores in patients that received IO PRP improved significantly from a mean 65.1 at baseline to 56.46 one month after intervention and the improvement was maintained with scores of 55.33, 54.23 and 55.38 at 3-, 6-, and 12-month follow-up, respectively.

• 3.Knee injury and Osteoarthritis Outcome Score (KOOS)

Sanchez et al. [12] noted significant improvement in KOOS scores in their cohort study for patients treated with combined IO and IA PRP; from a mean 61.55 baseline to 74.60 at 24 weeks. In another comparative study, Sanchez et al. [19] reported significant improvement in KOOS in the group that received IA as well as IO PRP whereas there was no significant improvement in the group that received only IA PRP. IO PRP was found to significantly improve KOOS by Lychagin et al. [20] from a baseline 58.78 to 72 within one month and thereafter 72.13, 75.12 and 77.03 at three, six and 12 months, respectively. Berman et al. [13] in their RCT, compared a single injection of combined IO and IA PRP with a single IA PRP injection and observed a significant improvement in KOOS scores in both the groups. However, the improvement in KOOS score was not statistically significant between the two groups.

Conversion to Knee Arthroplasty

Sanchez et al. [14] in their cohort study observed that two out of the 13 patients (15.4%) required a total knee arthroplasty, as they did not respond to combined IO and IA PRP at 24-week follow-up.

Secondary Outcomes

Complications

In Sanchez et al.’s cohort of 13 patients that received combined IA and IO PRP, one had an episode of fever/flu-like symptoms, and one patient had exacerbation of knee pain [12]. Su et al. in their comparative study reported complications in five patients out of 27 (18.5%) who received combined IO and IA PRP (1 ankle pain, 1 nausea, 1 knee pain and swelling, 1 numbness, 1 inability to squat) [18]. In the group that received only IA PRP, 8/25 patients (32%) developed complications (3 knee pain/swelling, 2 low back pain, 2 respiratory infections, 1 headache). The group that received IA HA had 3 (10%) complications (2 knee pain/swelling, 1 upper respiratory infection). Berman et al. [13] in their comparative study found more adverse events in combined IO and IA PRP injection group compared to alone IA PRP injection group. In combined IO and IA PRP, 23 patients reported knee pain, 19 swelling, knee joint stiffness and headache in 22 each, while IA injection group had knee pain in 9, swelling in 2, joint stiffness and headache each in 5 patients.

Results of Meta-analysis

We compared the outcomes of IA versus combined IA and IO PRP with available data. Meta-analysis of comparative studies comparing IA PRP with combined IA and IO PRP showed that the combination is associated with significantly better outcomes in terms of WOMAC scores (Mean difference 11.55 [95% CI 10.30, 12.80; p < 0.05] compared to isolated IA PRP injection (Fig. 4) [18, 19]. However, VAS pain scores (Mean difference 1.56 [95% CI 0.44, 3.55; p = 0.13] (Fig. 5) [13, 18, 19], and KOOS scores (Mean difference 5.23 [95% CI − 2.11, 12.57; p = 0.16] (Fig. 6) [13, 19] showed no statistical difference between two treatment groups.

Fig. 4.

Forest plot comparing WOMAC scores

Fig. 5.

Forest plot comparing VAS pain scores

Fig. 6.

Forest plot comparing KOOS scores

Discussion

This systematic review has revealed that there is a paucity of high-quality RCTs evaluating the outcomes of IO PRP infiltration in the current literature. It has been found that IO PRP injections are relatively safe, and lead to improved functional outcomes and patient satisfaction as per currently available research.

Recent research has highlighted the role of subchondral bone in the pathogenesis of OA by its communication with underlying cartilage [17]. It has been hypothesized that sclerotic subchondral bone could reduce the amount of load that is transferred to the metaphysis leading to osteoporotic changes [12]. Subchondral bone may also be subjected to microdamage that stimulates remodeling and ingrowth of fibrovascular tissues. These fibrovascular channels could invade and damage the articular cartilage leading to accelerated degeneration [18]. Combined IA and IO PRP has the potential to act on the cartilage, synovium, subchondral bone and its cartilage communication, as well as on mesenchymal stem cells modulating tissue regeneration [19]. In our systematic review, from all the five included studies, it was concluded that the use or addition of IO PRP had led to significant benefits in outcomes with statistically significant improvement in outcome scores compared to baseline. The procedure was safe as there was no serious treatment-related complications reported in any of the studies.

In a limited meta-analysis including two studies [18, 19], the combination (IO + IA PRP) had significantly better WOMAC scores compared with IA PRP group. However, there was no significant difference in VAS score or KOOS scores.

Though the combination (IO + IA PRP) is effective, with the available literature it cannot be conclusively stated to be better than IA PRP alone as IA PRP group also had clinically significant outcome scores compared to baseline. The small sample size and limited studies are not powered enough to pick up minor superiority of one treatment over the other as both are effective treatment and to detect minor changes would need higher sample sizes. It should also be noted that IO PRP is a more invasive and time consuming procedure requiring C-arm guidance compared to IA PRP; and we believe that the available evidence is not strong enough to establish its utility over IA PRP which is a much simpler and quicker procedure.

This review has some limitations. The available studies included relatively small number of patients with short follow-up and there is paucity of high-quality RCTs evaluating the outcomes of IO PRP with longer follow-up. Most of the included studies lacked serial MRI evaluations to assess for the healing response after IO injections, which could have been a limiting factor because of the associated costs.

Conclusion

Intraosseous PRP injections are safe, and its combination with IA PRP injections provides significantly improved outcomes compared to baseline. However, its added advantage over only IA injection is not well established and future high-quality RCTs are required to corroborate these findings and further substantiate the evidence.

Funding

None.

Declarations

Conflict of Interest

The authors declare no competing interests.

Ethical Approval

Ethical approval does not require as this a review article.

Informed Consent

Not applicable as this a review article.