Side-to-side characterisation of cellular content, soluble factors and in vitro potential on chondrocytes for bone marrow aspirate concentrate and adipose-derived stromal vascular fraction

Abstract

Purpose: Orthobiologics gained popularity for the treatment of musculoskeletal pathologies, including osteoarthritis (OA). Bone marrow aspirate concentrate (BMAC) and adipose-derived stromal vascular fraction (SVF) were reported to reduce OA symptoms, contributing to the restoration of joint homoeostasis. Variability in production protocols and lack of extensive characterisation hinder a clear indication for the choice of a product over the other. The purpose of this study was to characterise side-by-side BMAC and SVF obtained with the same family of devices, by assessing cell immunophenotype, release of soluble factors and their ability to reduce inflammation in a pathologic in vitro chondrocyte model.

Methods: BMAC (iliac crest) and SVF (abdomen liposuction) were obtained from 28 (55 years old ± 8) and 39 patients (56 years old ± 9), with Hy-Tissue BMAC and Hy-Tissue SVF. BMAC/SVF were characterised for cellular content. The number of mesenchymal stromal cells (MSCs) was investigated by flow cytometry (CD45^-CD31^-CD34⁺CD90⁺CD105^-/LCD146^-, adipose-MSCs; CD45^-CD271⁺, bone marrow-MSCs). Two-hundred factors (cytokines, chemokines, receptors, growth factors and inflammatory molecules) were tested by enzyme-linked immunosorbent assay (ELISA). Anti-inflammatory potential was assayed in vitro on interleukin-1 beta (IL1β)-treated chondrocytes by quantitative reverse transcription polymerase chain reaction (qRT-PCR) arrays of 84 genes involved in inflammatory processes.

Results: BMAC had higher concentration for white cells (213x), erythrocytes (49x) and platelets (25x), while the number of MSCs resulted comparable between the two products (1000 cells/mL). One-hundred and twenty-one soluble factors were identified in all analysed samples, with 88 more abundant in BMAC and one in SVF. Gene ontology revealed that the higher concentrated molecules were mainly growth factors and/or involved in differentiation processes. Both orthobiologics reduced inflammation in the in vitro chondrocyte model, with BMAC showing higher efficacy.

Conclusions: Using specific commercial systems, both orthobiologics showed anti-inflammatory effects in vitro. BMAC had higher blood cell and growth factor concentrations than SVF, with greater efficacy. However, variability in commercial systems limits generalisation, requiring further study to draw conclusions when different devices are employed.

Level of evidence: NA.

Keywords: cartilage; mesenchymal stromal cells; orthobiologics; osteoarthritis; soluble factors.

Figure 1

Orthobiologics characterisation. (a) Stromal vascular fraction (SVF) and bone marrow aspirate concentrate (BMAC) donor age, body mass index (BMI) and Kellgren–Lawrence (KL). Violin plots with quartiles and median are shown. N = 39 for SVF and 28 for BMAC. ns stands for not significant. (b) Haematological data for SVF and BMAC. Violin plots with quartiles and median are shown. N = 39 for SVF and 27 for BMAC. **** for p value < 0.0001. (c) Mesenchymal stromal cells (MSCs) content per millilitre in SVF and BMAC. Violin plots with quartiles and median are shown. N = 27 for SVF and 23 for BMAC. ns stands for not significant.

Figure 2

Functional association network for identified bone marrow aspirate concentrate (BMAC) and stromal vascular fraction (SVF) soluble factors. Blue connections are for proteins with known interactions based on curated databases. Empty nodes, proteins of unknown three‐dimensional (3D) structure; filled nodes, known or predicted 3D structure.

Figure 3

Soluble factors profile comparison between bone marrow aspirate concentrate (BMAC) and atromal vascular fraction (SVF) samples. (a) Principal component analysis of the ln(x)‐transformed pg/million values of detected factors. X and Y‐axis show principal component 1 and principal component 2, which explain 64.2% and 9% of the total variance. (b) Heat map of hierarchical clustering analysis of the ln(x)‐transformed pg/million values of detected factors with sample clustering tree at the top and factors clustering tree on the left side.

Figure 4

Functional association network for proteins that resulted more concentrated in bone marrow aspirate concentrate (BMAC). Blue connections are for proteins with known interactions based on curated databases. Empty nodes, proteins of unknown three‐dimensional (3D) structure; filled nodes, known or predicted 3D structure.

Figure 5

Inflammation‐related gene expression comparison profile for Interleukin‐1 beta (IL1β)‐chondrocytes treated with bone marrow aspirate concentrate (BMAC) or stromal vascular fraction (SVF). (a) Principal component analysis of the RPLP0‐normalised C _t values of detected genes. X and Y axes show principal component 1 and principal component 2, which explain 34.5% and 12.7% of the total variance. (b) Heat map of hierarchical clustering analysis of the RPLP0‐normalised C _t values of detected genes with sample clustering tree at the top and gene clustering tree on the left side. For both A and B panels, CTRL and IL1β‐treated chondrocytes mean values are reported for ease of Figure readability.

Figure 6

Gene expression modulation in inflamed chondrocytes by bone marrow aspirate concentrate (BMAC) or stromal vascular fraction (SVF) treatment. Interleukin‐1 β (IL1β)‐inflamed (IL1B) chondrocytes treated with BMAC (IL1B + BMAC) or SVF (IL1B + SVF) were analysed for the expression of genes involved in inflammation and their amount compared to untreated (CTRL) cells set as 1. Only significantly (fold ≥ 2 or ≤ 0.5 and p value ≤ 0.05) modulated genes are shown. * for p value ≤ 0.05, **≤ 0.01, ***≤ 0.001 and ****≤ 0.0001. N = 7 for CTRL and IL1B, N = 10 for IL1B + BMAC and IL1B + SVF. Box and whiskers are shown with min to max values.

Figure 7

Correlation analysis of bone marrow aspirate concentrate (BMAC) and stromal vascular fraction (SVF) parameters with modulation in gene expression in inflamed chondrocytes. Correlations with ‘r’ values being ≤–0.8 or ≥0.8 with p value ≤ 0.05 are presented. N = 8 for BMAC and N = 10 for SVF.