Bone Marrow Aspiration Concentrate in the Treatment of Osteoarthritis: A Review of its Current Clinical Application

Abstract

Background:

Bone marrow aspiration concentrate (BMAC) has gained acceptance as a safe orthobiologic for treating osteoarthritis (OA), despite lacking robust supporting evidence. Although several publications have documented the use of BMAC in OA, evidence confirming its unequivocal efficacy remains limited.

Methods:

This review aims to summarize the current clinical evidence regarding BMAC as a therapeutic for OA, while also presenting the author’s perspective. Sixteen studies were reviewed, including ten randomized clinical trials (RCTs) and six cohort studies.

Results:

From the review of existing literature, BMAC injections do not appear to significantly improve pain and function compared to conventional therapies such as hyaluronic acid and corticosteroids, although some studies report a longer duration of effectiveness. Furthermore, the evidence for structural improvement, which was the original rationale for cell therapy, is seldom reported.

Conclusion:

In light of these findings, it is suggested that high-quality data from a large patient cohort is needed to determine the role of BMAC injections in OA treatment and address reimbursement issues. From the author’s perspective, the introduction of a national registry system that provides valuable information on the cost-effectiveness of various orthopedic procedures may offer a solution.

Introduction

Osteoarthritis (OA) is the most common form of arthritis, causing pain and loss of function, which leads to reduced quality of life [1, 2]. Non-surgical management for OA traditionally includes 1) lifestyle changes such as weight loss and activity modification, 2) exercise and physical therapy, 3) pain and anti-inflammatory medications, and 4) intra-articular (IA) injections of substances like corticosteroids and hyaluronic acid (HA). In cases of advanced disease, surgical treatments, primarily total joint replacement (TJR), are utilized [3].

While TJR offers the ultimate solution for end-stage OA, it requires an extensive rehabilitation and recovery period, and the cost of the operation imposes a significant financial burden on the healthcare system [4, 5]. Additionally, not all patients are satisfied with TJA outcomes, 20% report dissatisfaction with the results, and there is a potential risk of severe complications such as pulmonary embolism and periprosthetic infection. For these reasons, among others, many patients prefer a non-surgical solution for OA. There is thus an unmet clinical need to improve the pain and quality of life for patients suffering from OA without resorting to surgical options, even at advanced stages. Recent interest in regenerative medicine (RM) for OA stems from these considerations [6].

RM for OA has primarily focused on cell-based injection therapies. IA injections in cell delivery offer benefits such as increased local availability and minimal invasiveness [7]. Particularly, mesenchymal stromal/stem cells (MSCs) have been extensively researched for their potential therapeutic applications [8–10]. The therapeutic mechanisms of MSCs, including paracrine, trophic, anti-inflammatory, and immunomodulatory effects, have been elucidated [11–13]

Bone marrow aspiration concentrate (BMAC), which can be obtained in outpatient settings without a culture process, has been considered as an alternative to MSCs, which require time for culture expansion and face complex regulatory hurdles for patient application. BMAC contains relatively few MSCs (only 0.001–0.01% of the cellular content of BMAC are stromal cells) [8, 9]. The US Food and Drug Administration (FDA) has issued guidance regarding the use of BMAC as minimally manipulated under the Code of Federal Regulations; however, to date, the FDA has not formally approved any therapy or intervention that cures OA or halts its progression [14–16].

BMAC has gained recognition as a safe biologic because it is autologous, and few complications have been associated in harvesting. Bone marrow is normally extracted from the iliac crest, but also available from the proximal tibia or calcaneus. The aspiration usually takes only a few minutes, and concentrating the aspirates takes between 15–30 min. Manufacturers provide a centrifuge and special sterile equipment to concentrate the bone marrow [17].

BMAC has been used in various countries to treat OA patients without robust supporting evidence or evaluation. Recently, BMAC use has been approved in Korea as new advanced therapeutics for KL 1–3 OA knees. While it is not reimbursed by National Health Insurance, some private insurers cover BMAC for OA, leading to its indiscriminate use for OA patients. Despite several studies documenting the use of BMAC in OA [16], there is limited evidence to confirm its unequivocal efficacy. Therefore, the purpose of this review is to summarize the current knowledge on clinical evidence regarding BMAC as a therapeutic for OA and to introduce the author’s perspective.

Methods

Data on BMAC for OA were gathered from PubMed, Medline, Embase, and Cochrane Library databases using the search string: “osteoarthritis” or “bone marrow” or “stem cell” or “BMAC”. A total of 453 papers were identified after removing duplicates. Excluding preclinical studies, reviews, proceedings, case series, reports, and studies that used cultured cells or those other than BMAC without control groups, sixteen studies were found suitable for review; of these, ten were randomized clinical trials (RCTs) and six were cohort studies.

Summary of findings from clinical studies

1) BMAC versus placebo treatments.

Varma et al. first reported a randomized trial of BMAC in 50 patients with mild to moderate knee OA. Twenty-five patients underwent arthroscopic debridement alone, while the other 25 received a BMAC injection along with the arthroscopic debridement. According to the authors, the BMAC injection group experienced significantly improved pain, function, and quality of life at 2,3, and 6 months of follow-up [18].

Two studies by Shapiro et al. studying the same patient cohort compared a single BMAC plus platelet poor plasma (PPP) IA injection to a saline IA injection [16, 19] in patients with bilateral symptomatic knee OA with Kellgren-Lawrence (KL) grades 1–3. The first study was a single-blind, placebo-controlled trial. Twenty-five patients with bilateral knee pain from bilateral OA were randomized to receive BMAC in one knee and a saline placebo in the other. There were no serious adverse events associated with the BMAC procedure. OARSI (Osteoarthritis Research Society International) Intermittent and Constant Osteoarthritis Pain and VAS pain scores in both knees decreased significantly from baseline at 1 week, 3 months, and 6 months (P ≤ 0.019 for all). Although the pain relief was dramatic, it did not differ significantly between the treated knees (P > 0.09). At a 1-year follow-up, pain remained reduced in both groups, and the quality of life improvement in BMAC-injected knees was comparable to that in the placebo group. MRI cartilage sequences did not demonstrate a regenerative benefit from a single BMAC injection at the 6-month follow-up quantitative mapping [19].

2) BMAC versus exercise.

Centeno et al. compared exercise therapy with BMAC and platelet products for treating knee OA. Patients with symptomatic knee OA (N = 48, KL 2–3) were randomized into either an exercise therapy control group or treatment group receiving a single injection of autologous BMAC and platelet products. At the end of 3 months, all patients originally assigned to exercise therapy transitioned to to the BMAC treatment (N = 22). After the BMAC treatment, patients' clinical outcome scores, except for the SF-12 Mental Health, were significantly improved through the 2-year follow-up compared to baseline. No serious adverse events were reported [20]

3) BMAC versus HA.

Goncars et al. reported significantly improved clinical outcome scores with a single BMAC IA knee injection compared to three IA sodium HA injections at 1 year in patients with moderate to severe (KL 2–3) knee OA in a randomized trial. Twenty-eight patients received a single BMAC IA injection while a control group of 28 patients received three sodium HA IA injections administered weekly. After 12 months, the knee injury and osteoarthritis outcome score (KOOS) showed significant improvement (P < 0.05) in both groups across the pain, activity, daily living, quality of life subscales, and the total KOOS score. The Knee Society Score (KSS) also showed significant improvements in the symptom subscale and the function subscale (P < 0.001) in both groups. The BMAC group demonstrated statistically significant (P < 0.05) superiority in the KOOS pain subscale over the HA group at 6 and 12 months after injection. In all other score subscales, outcomes were better in the BMAC group, although the differences were not statistically significant [21].

Boffa et al. similarly compared a single IA injection of BMAC versus HA for knee OA in a double-blind randomized controlled trial (RCT). Fifty-six patients (35 men, 21 women, mean age 57.8 years), with bilateral knee OA were randomized to receive one IA injection of tibial-derived BMAC in one knee and one HA injection in the contralateral knee. The International Knee Documentation Committee (IKDC) subjective score improved from baseline at all follow-ups for BMAC (p < 0.0005), while it improved until 12 months (p < 0.0005) and then declined at 24 months (p = 0.030) for HA. Compared to HA, BMAC demonstrated greater improvement in visual analogue scale (VAS) pain scores at 12 (p = 0.041) and 24 months (p = 0.002). The analysis stratified by OA severity confirmed this superiority only in KL 1–2 knees, with comparable outcomes in moderate to severe OA. Radiographic evaluations showed no progression of knee OA in either treatment group, with no significant differences between them. Over time, while HA results diminished, BMAC maintained more durable outcomes in mild OA knees [22].

4) BMC versus PRP.

Dulcic et al. directly compared clinical outcomes of knee injections of BMAC, PRP, or HA in the OA treatment. A total of 175 patients with knee osteoarthritis KL grade 2–4 were randomized to receive a BMAC, PRP, and HA injection; 111 patients received BMAC injections, 30 received HA injections, and 34 received PRP injections. The mean VAS scores after 3, 7, 14, and 21 days showed significant differences between groups, with a reduction in VAS in all groups but more notably in the BMAC group compared to the others (p < 0.001). Statistically significant differences were observed between baseline scores and those after 12 months (p < 0.001) in the Western Ontario and McMaster Universities Arthritis Index (WOMAC), KOOS, KOOS pain, and IKDC scores, with distinctions noted in the BMAC group compared to other groups. The authors concluded that BMAC could be better in terms of clinical improvements in the treatment of knee OA than PRP and HA up to 12 months [23].

On the other hand, Anz [3, 24] found no significant difference in outcome scores (WOMAC and IKDC) between 90 patients with mild to moderate knee OA (KL1-3) randomized to receive platelet-rich plasma (PRP) versus autologous BMAC single IA injection, although improvements were observed in both groups compared to baseline. All IKDC and WOMAC scores for both groups significantly improved from baseline to 1 month after the injection (P < 0.001). These improvements persisted for 12 and 24 months after the injection, with no differences between PRP and BMAC at any examined time point.

5) BMAC versus SVP or microfragmented adipose tissue.

Mautner evaluated patients with knee OA (KL 1–4) who received either BMAC or microfragmented adipose tissue (MFAT). They retrospectively reviewed prospectively collected data from patients who received BMAC (17 females and 24 males, average age 59) or MFAT (23 females and 12 males, average age 63) injections for symptomatic knee OA. Mean follow-up time was 1.8 years for BMAC and 1.09 years for MFAT. Both groups exhibited similar functional outcomes [KOOS and mean quality of life (EQOL)], and pain (VAS) scores that significantly improved from pre-procedure [25].

Mautner et al. conducted a phase 2/3, four-arm parallel, multicenter, single-blind, randomized, controlled clinical trial involving 480 patients with knee OA (KL 2–4) to evaluate the safety and efficacy of cell injections from autologous BMAC, autologous SVP, and allogeneic human umbilical cord blood-derived MSC, compared to corticosteroid injection (CSI). Participants were distributed into three different arms with a 3:1 ratio. Arm 1 consisted of autologous bone marrow aspirate concentrate (n = 120) and CSI (n = 40); arm 2 included umbilical cord tissue-derived mesenchymal stromal cells (n = 120) and CSI (n = 40); arm 3 comprised stromal vascular fraction (n = 120) and CSI (n = 40). At 1 year post -injection, none of the orthobiologic injections proved superior to another or to the CSI control. Additionally, there was no significant change in the magnetic resonance imaging osteoarthritis score from baseline in any of the four groups. No procedure-related serious adverse events were reported during the study period [26].

Pintore et al. compared the results of IA injections of BMAC in 51 patients with those of SVP in 51 knee OA patients. Knee KOOS scores, Oxford knee scores (OKS), and VAS pain scores exhibited similar changes in the two treatment groups. Significant improvements from pre-procedure to post-procedure were observed in the knee KOOS scores (p < 0.0001), knee OKS scores (p < 0.0001), and VAS pain scores (p < 0.0001). Patients with KL grade 2 demonstrated better functional and clinical outcomes than those with KL grades 3 and 4 (p < 0.0001) [27].

Similarly, Estrada et al. retrospectively studied 89 patients with a median age of 61 years who had painful knee OA (KL grades 1–3). Patients were assigned to one of three treatments based on OA severity, as indicated by symptoms and radiography: PRP (KL 1, 29 patients), BMAC (KL 2, 27 patients), or microfragmented adipose tissue (KL 3, 33 patients). Clinical score values (KSS and IKDC) were comparable among the treatment groups at baseline. Statistically significant improvements were observed in all three groups according to clinical scores at all time points during follow-up (90, 180, and 265 days) compared with baseline. No differences were found among treatment types [28].

6) BMAC versus autologous serum.

Vitali et al. compared the efficacy of a single BMAC injection with a cycle of four autologous conditioned serum (ACS) injections in treating early-stage knee OA. Twenty-four patients were treated with either a single BMAC injection (12 patients) or a cycle of four ACS injections (12 patients). The BMAC group showed significantly greater improvements in the WOMAC after the BMAC injection at one month (p = 0.001) and six months (p < 0.001), as well as a reduction in VAS values at the six-month follow-up (p = 0.024). In contrast, no significant differences in ROM between the two groups were observed [29].

7) BMAC application associated with surgical procedure.

In a retrospective cohort study, Jin et al. compared BMAC with placebo in patients suffering from medial unicompartmental knee OA who underwent high tibial osteotomy (HTO). They retrospectively analyzed clinical outcomes in patients treated with HTO for medial unicompartmental OA with varus deformity, specifically comparing the microfracture alone group (43 cases) to the microfracture with BMAC augmentation group (48 cases). At the final follow-up, no significant intergroup differences were observed in the Knee Society Score for pain and function (p > 0.05) or WOMAC scores (p > 0.05). Additionally, second-look arthroscopy of 64 patients revealed no significant difference in defect coverage (p = 0.187) [30].

In a study by Yang, patients with medial compartment OA (KL 3) who underwent HTO were compared. Two treatment groups were evaluated: one receiving BMAC and the other human umbilical cord blood-derived MSCs (hUCB-MSC). At 33 weeks, both groups demonstrated improvements in functional scores from their preoperative baselines, though no significant differences were noted between them. However, significantly better healing of regenerated cartilage, as assessed by second-look arthroscopy, was observed in the hUCB-MSC group compared to the BMAC group [31].

Discussion

Although no significant harm from BMAC application has been reported, its acceptance as a suitable therapy for knee OA is limited due to insufficient supporting data. While BMAC resource utilization is minimal when used in the operating room, its use in outpatient clinics demands considerable resources for patient monitoring, analgesia, and sterile preparation. Additionally, BMAC is usually not covered 1 by insurance, thus necessitating that costs be absorbed either by the hospital or the patient.

Table 1.

Summary of clinical trials that used BMAC to treat OA

	Study type	Number of patients	KL grade	Intervention	Control	Follow up period	Key findings	References
1	RCT	50	Mild to moderate OA	BMAC + AD	AD	6M	BMAC injections significantly improved pain, function, and quality of life compared to the control group	[18]
2	RCT	25 (both knees)	KL1-3	BMAC	Saline	6M	There was no difference in pain relief between the groups	[19]
3	RCT	25 (both knees)	KL1-3	BMAC	Saline	1Y	Pain and QoL were similar to those in the saline group. T2 quantitative MRI mapping revealed no significant changes resulting from the treatment	[16]
4	RCT	48 (BMAC 26, exercise 22)	KL 2–3	BMAC + platelet product	Exercise	2Y	All patients in exercise group crossed over to BMAC group	[20]
5	RCT	56 (BMAC 28, HA 28)	KL2-3	BMAC	HA × 3	1Y	BMAC group showed significant superiority in the pain subscale	[21]
6	RCT	56 (both knees)	KL1-3	BMAC	HA	2Y	BMAC group exhibited greater improvement in VAS pain scores than the HA group	[22]
7	RCT	175 (BMAC 111, HA 30, PRP 34)	KL1-4	BMAC	PRP HA × 3	1Y	BMAC demonstrated superior pain relief and functional improvement compared to other groups	[23]
8	RCT	BMAC 45 PRP 39	KL 1–3	BMAC	PRP	1Y	All IKDC and WOMAC scores for both groups improved significantly from baseline, with no differences between the groups	[3]
9	RCT	BMAC 25 PRP 17	KL 1–3	BMAC	PRP	2Y	Both PRP and BMAC enhanced pain relief and functional outcomes, with no significant differences between the two groups	[24]
10	Cohort	76 (BMAC 41, MFAT 35)	KL1-4	BMAC	MFAT	1-2Y	Both groups exhibited improved pain relief and functionality, showing no significant differences between them	[25]
11	RCT	480 (BMAC 120, UCMSC 120, SVF 120, CS 40)	KL2-4	BMAC, UBMC, SVF	CS	1Y	None of the three orthobiologic injections proved superior to the others or to the CS control, with no group demonstrating a significant change in MRI OA score from baseline	[26]
12	Prospective Cohort	102 (BMAC 51, SVP 21)	KL2-4	BMAC	SVP	6M	Both groups demonstrated improvements in pain and function with no significant differences between them	[27]
13	Cohort	89 (BMAC 27, PRP 29, MFAT 33)	KL1-3	BMAC	PRP, MFAT	1Y	All three groups showed clinical improvement over baseline with no differences among them	[28]
14	Cohort	24	KL1-2	BMAC	ACS × 4	6M	The BMAC group exhibited significantly better improvement in pain and function compared to the ACS group	[29]
15	Cohort	91	KL3-4	BMAC + HTO + MFX	HTO + MFX	2Y	No significant intergroup differences were observed in pain, function, or defect coverage	[30]
16	Cohort	110 (BMAC 55, hUBMSC 55)	KL3	BMAC + HTO + MFX	hUBMSC + HTO + MFX	2Y	No significant differences were found between groups. The hUCB-MSC group demonstrated significantly better healing of regenerated cartilage	[31]

ACS: autologous conditioned serum; AD: arthroscopic debridement; CS: corticosteroid; HA: hyaluronic acid; HTO: tibial osteotomy; MFAT: microfragmented adipose tissue; MFX: microfracture; QoL: quality of life; SVF: stromal vascular fraction; hUBMSC: umbilical cord blood-derived mesenchymal stromal cell

When interpreting findings from cited articles, it is crucial to consider the heterogeneity of the studies included. The ability to draw definitive conclusions on the efficacy of BMAC is constrained by challenges such as controlling for variables like patient age, comorbidities, methods of BMAC preparation, and the quantity and quality of cell concentrates.

Two high-quality studies concluded that a single IA injection of BMAC plus platelet poor plasma was no more effective than a placebo saline injection in terms of function, MRI-based cartilage appearance, or pain control [16, 19]. Given that the injection itself acts as a strong placebo, and considering that it was typically administered when patients were experiencing peak pain, these studies demonstrated that BMAC does not have therapeutic effects lasting for a long-term period time (1 year). Conversely, another high-quality study reported that BMAC plus platelet product intraarticular injection outperformed exercise therapy functionally, with all patients switching treatments within three months of the study's commencement. The BMAC group exhibited better outcomes at three months post-injection [20]. This improvement could be attributed to a strong placebo effect in the injection group. If the exercise group had not transitioned to the BMAC group, long-term results at one year might have shown similarities, with both groups regressing to the mean.

Several RCTs compared BMAC injections with the injection of other substances, including HA, PRP, corticosteroids, and adipose tissue-derivatives. Two articles indicated that both BMAC and HA improved pain and function at one or two years post-treatment compared to baseline, with BMAC achieving better pain scale results [21, 22], with one study showing that BMAC have more durable effect on pain scale in KL1-2 knee [22]. Given that HA is a relatively inexpensive injectable used in outpatient clinics for symptomatic improvement of OA, it is very important that expensive therapeutic such as BMAC should show definite advantage over HA. According to these studies, BMAC seems to have demonstrated comparable or slightly better results than HA, particularly in alleviating pain if not else including structural improvement [21, 22]. PRP is another biologic that is easier and less invasive to prepare than BMAC, making it less costly. One RCT comparing BMAC, HA, and PRP revealed clinical improvements with all three injectables; however, BMAC showed superior results a year after treatment compared to the other two [23] while other RCT that compared BMAC and PRP showed no difference between the two groups at 2 year follow up [24]. Based on these findings, BMAC appears to have similar effects on the pain and function in OA patients as PRP, but whether BMAC is a superior orthobiologic remains unclear.

Several studies compared BMAC with adipose-derived cell products, including SVF or MFAT. Each reported improvements in pain and function from baseline, without detecting differences in outcomes between BMAC and these products [25, 27–29]. Of notes, a high-quality study that prospectively compared BMAC, PRP, HA, and corticosteroids found no differences in long-term follow-up over one year and showed no significant changes in MRI findings [26]. These results suggest strong placebo effects of injection therapy and raise questions about the long-term efficacy of BMAC.

In conclusion, according to the literature, BMAC injections, which have emerged as a promising orthobiologic for treating OA, do not significantly enhance pain relief or function compared to conventional therapies such as HA and CS, despite some studies reporting longer duration of effectiveness. Also, the evidence of structural improvement which was the original rationale of cell therapy is rarely reported. Given the costs associated with BMAC preparation and patient morbidity during harvesting, BMAC injection should show distinctively better results than conventional therapies, which has not come true so far. In view of these findings, it is suggested that high-quality data from large number of patients is necessary to define the place of BMAC injection for OA treatment and resolve reimbursement issue. From the author’s perspective, the introduction of a national registry system that provides valuable information on the cost-effectiveness of several orthopedic procedures may answer the issues. Until the value of BMAC for OA treatment is determined, the use of BMAC should be restricted to patients who have provided informed consent and are under the care of physicians participating in the registry system.

Ethical statement

There are no animal experiments carried out for this article.