To our knowledge this is the first study worldwide with a long term follow (> 4 years) of patients who suffer from severe degenerative osteoarthritis of the knee (grade III and grade IV) and underwent an injection of bone marrow aspirate concentrate (BMAC, also described as “MSCs, mesenchymal stem cells” or “stem cell therapy”13,27).
In 2013 one of the first studies worldwide on the application of autologous pluripotent (stem) cells described the injection of autologous expanded bone marrow MSCs in 12 patients with knee arthritis, followed for a year17. However, as a matter of fact, a major limitation in the whole present literature is its paucity of detailed information, what the authors actually did in the respective studies: Of the 1580 studies identified, considerable deficiencies in the reporting of key variables, including the details of stem cell processing, possible culture conditions, and the characteristics of cell populations delivered, were noted. Overall, studies reported only 52% of variables that may critically influence outcome33,34.
In 2018 a review article of 1832 papers dealing with [“bone marrow aspirate” and “cartilage”] or [“mesenchymal stem cells” and “cartilage”] demonstrated promising results in the clinical application for repair of chondral defects and early clinical data suggests BMAC may help stimulate a more robust hyaline cartilage repair tissue response. Therefore, the current study was started18.
In 2021 a review on BMA or BMAC in osteoarthritis confirmed an remarkably increased publication trend over time. Safety was again confirmed by all studies, with a low number of adverse events, and an overall improvement in pain and function was documented in most of the studies, however, no long term results available yet35. This is well in line with our findings, where no mild or severe adverse effects were noted.
The outcome of BMAC in knee OA was already compared to the current gold standard therapies in the last years: Undoubted PRP and BMAC are superior to hylauronic acid (HA), as confirmed in a large review in the year 20238,36. Regarding BMAC alone, a significant improvement in WOMAC after BMAC injection plus a reduction of VAS values as compared to PRP was found at six months15. In another study, the BMAC group significantly improved in VAS, KOOS, and WOMAC scores between baseline and 12 months; in contrast, the PRP group (n = 13 knees) witnessed nonsignificant improvement in all scores37. Dulic showed, that OA patients treated with BMAC alone showed a significantly better VAS postoperatively, as compared to PRP or HA alone9.
However, these positive publications are in contrast to a review of 8 studies with a mean follow-up of 13 months, reporting the outcome of BMAC injections in OA knees: BMAC has not demonstrated clinical superiority in this paper in relation to other biologic therapies commonly used in the treatment of OA, including platelet-rich plasma and microfragmented adipose tissue, or in relation to placebo38. Fortunately in 34 of 36 (94.4%) patient-reported outcomes assessed across these 8 studies in the respective paper, BMAC demonstrated a significant improvement, and BMAC injection is effective in improving pain and patient-reported outcomes in patients with knee OA at short- to midterm follow-up. On a closer look, the follow-up of 13months in the respective studies was probably too short to notice a superior effect, which was demonstrated in the patients of the current study, where improvement was significant from the year 2 on and later, but not significant in the first year.
We found no effect of BMI and patient’s age on outcome in our data of 37 knees. This was confirmed in a study of 111 participants by Rasovic, who reported, that participants’ age and BMI did not influence the clinical outcome, but there was an influence of OA severity, especially among older patients. His study shows that BMAC therapy is effective and he states, that younger patients with milder OA changes could be better candidates for long-lasting and more efficient BMAC therapy39. Kim assessed the outcome of BMAC injection in patients with grade I to grade IV degenerative arthritis of the knee and confirmed this findings, that a BMAC injection significantly improved both knee pain and functions in the patients with degenerative arthritis of knee. Also, the injection would be more effective in early to moderate phases40.
It can be assumed, that the possibly inferior effect in elderly patients with severe arthritis as reported by the authors above is rather a result of a poor donor site (osteoarthritis) than age alone, since we published earlier, that the number of pluripotent cells in the bone marrow remains constant in adults, up to eighty years21. From former studies we already assessed, that we injected 140 ± 98 CD34 cells /µl with a viability of 85 ± 8% and 12 ± 4 Leucocytes / nl with our methodology in this study22. However, the accurate number of mesenchymal stem cells cannot be calculated, since it can only be measured indirectly: In this context it should be emphasized, that the use of the term ‘pure MSCs’ should be avoided since mesenchymal cells are not separated from the other cells of the graft’s microenvironment, and consequently the procedures performed are straightforwardly addressed at concentrating tissues with minimal manipulation, originating bone marrow aspirate concentrates (BMAC)26.
In order to reduce the risk of manipulation and infection, to simplify the technique and to avoid the costly and time-consuming procedure of centrifugation in order to generate BMAC from BMA, multiple attempts have been made, to use BMA alone, which was then harvested with an improved technique: The use of a special reorientation technique and the use of a needle with multiple lateral holes (Magellan® or MarrowCellution®) seem to be two different options to improve the cell yield, as described22. This new technique can lead to a 50% reduction in the visual analog scale score for pain at 64 ± 26 weeks post-procedure, even in patients with severe arthritis of the knee, using this pure bone marrow aspiration technique41. Varady used BMA harvested with a specific cannula instead of BMAC and found significant improvements in early pain and function scores for knee OA42. Knee OA patients treated with this technique exhibited significant reductions in VAS pain scores and significant improvements in WOMAC scores that exceeded the minimal clinically important difference thresholds43. Vigano published a single-step technique with BMA obtained with a centrifuge-free process, employing a dedicated aspiration device and showed the effectiveness of the study device to harvest pure bone marrow with minimal peripheral blood contamination. The relevant content of MSCs resulted in the ability to counteract the catabolic cascade through a paracrine action. The clinical outcomes in patients affected by unicompartmental knee OA were encouraging in terms of pain reduction and functional improvement up to mid-term evaluation44.
Another attempt, to improve clinical results is, to inject the BMAC at the subchondral site instead of intraarticularily. Hernigou was the first, to publish impressive 12 year results in patients with corticoid induced Grad IV arthritis: During the same anesthesia, one knee received TKA; for the other knee, a bone marrow graft was delivered to the subchondral bone of the femur and tibia45. At the most recent follow-up (average of 12 years), six (out of 30) TKA knees needed subsequent surgery versus only one with cell therapy in patients and patients liked the BMAC knee more. A similar improvement of a subchondral injection of BMAC was described by Vad, who published 14 month results with an improvement of the mean WOMAC score from 58.2 points pre-procedure to 35.3 points post-procedure (p < 0.01) and a mean NRS-Pain score drop from 8.6 points to 2.8 (p < 0.01), respectively46.
Kon combined the intraarticular and the subchondral injection of BMAC and showed safety and positive clinical and radiological outcome up to 24 months in the treatment of symptomatic knee OA, with durable clinical results, a low failure rate, and a significant reduction of bone marrow edema32. The role of a bone marrow edema in arthritis is not yet fully understood and it might be a negative predictor of outcome31. We can confirm this hypothesis, since we had unfavourable results leading to total knee arthroplasty in two patients outside of this study, where we injected BMA in severe osteoarthritis knees with a long lasting Vitamin D deficiency or a history of persistent bone marrow edema in the respective knee one year prior to the operation. It is worth to conduct further studies with MR imaging preoperatively to identify the influence of bone marrow edema. Lychagin is currently working on this, he published already a significant reduction of pain based on the VAS score and a significant improvement in the patients’ WOMAC score and in the overall KOOS score of patients who underwent a subchondral (= intraosseous) injection of PRP and suffered from knee arthritis and bone marrow edema31. However, injection into the edema may improve the outcome.
But BMAC can also be compared to other cell-based therapies: Pintore compared BMAC and ADSC (adipose-derived stromal cells) in patients with knee OA: Both treatment groups demonstrated significant improvement pre-procedure to post-procedure in knee KOOS scores, knee OKS scores, and VAS pain scores. Patients with K-L grade II showed better functional and clinical outcomes than patients with K-L grades III and IV47. Mautner compared BMAC and ADSC and his data demonstrate significant improvement in pain and function with both injections in patients with symptomatic knee OA without a significant difference in improvement when comparing the two autologous tissue sources48. A meta-analysis was already performed and the ADSC injection had a significantly greater effect on pain reduction than did the BMAC injection, but the clinical effect of BMAC versus ADSC knee injection in patients with knee OA regarding WOMAC was equivalent49.
Park compared BMAC and Human Umbilical Cord Blood-Derived Mesenchymal Stem Cells (hUCB-MSC): Improved clinical outcomes were found in both BMAC and hUCB-MSC groups; however, no significant difference was observed. On second-look arthroscopy, the hUCB-MSC group showed better International Cartilage Repair Society Cartilage Repair Assessment grade compared with the BMAC group in varus knee OA patients50.
Reading about all these different techniques, it becomes increasingly complicated to pick the right patient for the right therapy. Therefore recently Smith developed a CDR (clinical decision rule (CDR) ) to identify people with knee osteoarthritis who are likely or unlikely to benefit from bone marrow aspirate concentrate (BMAC) injection: Multiple logistic regression analysis was used to determine which combination of risk factors predicted BMAC responsiveness. A simple CDR containing three variables predicted responsiveness to a single intraarticular knee BMAC injection with high accuracy. Those with lower initial pain levels (< 7/10), or high pain levels with previous surgery, could be predicted to benefit from a single BMAC injection51.
The downside of all these cellular therapies is, that they fall under restrictive legislative constraints, making it almost impossible to offer these therapies to patients in Germany, Austria or Switzerland, except you have a respective license as a hospital, as we do. Some of the cited applications can be viewed as Advanced Therapy Medicinal Products (ATMPs): The use of those requires the knowledge of diverse regulations and an extensive (year long) communication with the national/international authorities, up to the hospital exemption52.
We do not see total knee arthroplasty (TKA) as an alternative to BMAC injection, since the patients differ, but the results have to be compared. Therefore we offer BMAC injections to the following three groups of patients:
To old to undergo TKA (e.g. due to cardic, pulmonary issues)
Too young to undergo TKA (biological age versus demographic age)
Ideal age for TKA, but does not like to have a big operation right now due to personal circumstances (takes care of ill family member, long holiday journey planned,..) TKA will be planned some years later.
Since we do not know, how long the positive effects of BMAC will last, we always have to bear in mind, that TKA is a well proven long lasting option with a low rate of failure: The pooled revision rate per 100 observed component years for TKA of all registry based studies worldwide was 0.6 and can be compared to 1.7 regarding unicondylar knee arthroplasty (UKA). This equals a 6% revision rate for TKA and a 17% revisions rate for UKA at 10 year follow up3,5,6,53,54. Since these data include all manufacturers, it can be seen in the Australian registries, that specific modern prostheses (my knee® by MEDACTA or ATTUNE® bei DePuy Synthes,…) perform slightly better. These data of knee replacement surgery have to be compared to a failure rate of 5% in the patients of the current study, which equals a failure rate of 1.35 per 100 observed patient years, which is therefore between TKA and UKA (2 failed / (4 years x 37 patients) x 100).
Another even lesss invasive therapy is Cognitive behaviour therapy (CBT): It showed sustainable effects through health-led cognitive behavioral-based group therapy and has a positive effect on pain, functional disability and psychological outcomes for knee osteoarthritis patients55. A recent meta-analysis suggests the durability of CBT-associated treatment benefits, supporting its role as a potential promising alternative or complementary intervention for patients with knee/hip osteoarthritis, especially against pain and insomnia56. The effect of CBT is also influenced by age, education and sociocultural environment: Patients with high mean pain catastrophizing scores may particularly benefit from CBT57.
A limitation of the current study is, that the CDR51. was not available when we started the study. We furthermore made no extensive pre- and postoperative MRI evaluation, since approximately half of the patients were indicated on plain radiographs and medical history only—according to our inclusion and exclusion criteria. We also made no double-blind assessment against placebo, since this would not have passed the ethics commission and we also did not use a control group (PRP or ADSC or HA), since most patients had already undergone therapy with PRP or HA. We furthermore did not assess all SF 36 questions and made no extensive psycho-social anamnesis (to exclude bias and confounders), but the unchanged SF36 values pre- and postoperatively indicate a stable social environment of our patients.
A strength of the current study, evaluating the positive 4-year effects of BMAC injections in OA- knees, is the long follow up and the sound and solid statistics (see supplementary files). The late onset (year 2–4) of the beneficial effects might be explained rather by the role of the cells (cytokines, exosomes and cartilage generation) than by a placebo effect (Supplementary Information 1), which was reported only in short term studies, see above.
In summary, this is the first study on BMAC injections (“mesenchymal stem cell therapy”) into 37 KL III and IV OA knees with a long term (four year) follow up and a relevant and significant improvement in terms of IKDC (from 56 ± 12 to 73 ± 13, p < 0.001) and WOMAC (40 ± 23 to 18 ± 18, p < 0.001), with a 95% success rate and significant improvement in walking distance.