Growth Factors in Bone Healing

The family. Bone morphogenetic proteins are members of the TGF-β superfamily and the most potent osteoinductive growth factors known. Over 20 family members have been identified; BMP-2, BMP-4, BMP-6 and BMP-7 have strong osteoinductive activity, and BMP-2 and BMP-7 are the most clinically relevant. Structurally they are dimeric proteins, the two chains linked by disulphide bonds, and they signal through the Smad1/5/8 pathway to activate Runx2.

Where they come from. In healing bone, BMPs are produced by osteoblasts and osteoprogenitor cells, stored in bone matrix and released during resorption, and carried in small amounts by platelets. Expression peaks at week 2-3 after fracture (days 14-21). They are essential for both fracture healing and skeletal development.

View larger

The Smad pathway. BMP signal transduction follows the canonical Smad pathway; the name comes from the Drosophila gene Mothers against decapentaplegic. The receptor complex is built of serine/threonine kinases, and the Smad that gets phosphorylated is what makes the signal a bone signal.

Inhibitors. The pathway is braked at every level:

• Noggin binds BMPs extracellularly and prevents receptor binding
• Chordin is a similar extracellular BMP antagonist, important in development
• Smad6 and Smad7, the inhibitory I-Smads, block R-Smad phosphorylation
• Smurf1/2 are E3 ubiquitin ligases that degrade Smads and receptors

Not every BMP makes bone. The family names GDF-5 (BMP-14 / CDMP-1) and BMP-5 as cartilage- and joint-related members and BMP-3 as an inhibitor. These non-osteogenic BMPs are worth distinguishing, because examiners use them to test whether you understand the family beyond rhBMP-2.

GDF-5. Growth/differentiation factor-5 (also BMP-14 or CDMP-1, cartilage-derived morphogenetic protein-1) is a TGF-β-superfamily member that signals through BMP receptors, notably BMPR-IB / ALK-6, but is principally chondrogenic and tendinogenic rather than osteogenic. It is essential for synovial joint formation, marking the developing joint interzone, and for cartilage, tendon and ligament development and repair, and it is investigated as a biologic for tendon, ligament and intervertebral disc repair, where a chondrogenic and tendinogenic rather than bone-forming profile is precisely the goal.

GDF5 genetics. Loss-of-function mutations cause brachydactyly and chondrodysplasias (Hunter-Thompson and Grebe types), confirming its role in limb skeletal patterning. A common GDF5 regulatory variant is a recognised susceptibility locus for osteoarthritis and developmental dysplasia of the hip, a high-yield genetics point.

Other non-osteogenic members. BMP-3 (osteogenin) is a negative regulator that inhibits bone formation, and BMP-15 / GDF-9 act in ovarian biology. "BMP" is a functionally diverse family, not a synonym for bone-forming.

The most abundant growth factor in bone matrix. TGF-β is a key regulator of bone remodelling and fracture healing, and its effects are complex and context-dependent, changing with the stage of repair. Mammals have three isoforms:

• TGF-β1 - most abundant in bone and the main regulator
• TGF-β2 - development and wound healing, with functions similar to TGF-β1
• TGF-β3 - anti-scarring; scarless wound healing with a less fibrotic response

Sources. Platelets release TGF-β during clotting in the fracture haematoma, which initiates healing. Osteoclastic resorption releases it from bone matrix, which is the coupling mechanism. Osteoblasts and inflammatory cells actively produce it.

Secretion and activation. TGF-β is secreted as an inactive latent complex, bound to latency-associated peptide (LAP), and is activated by proteases (plasmin, matrix metalloproteinases), acidic pH or mechanical stress. Active TGF-β then binds the type II TGF-β receptor (TβR-II). Activation is tightly regulated, which is how its local effects are controlled.

View larger

Biphasic on osteoblasts. TGF-β first expands the progenitor pool, then must be downregulated to allow terminal differentiation. Persistent high TGF-β blocks bone formation despite abundant progenitors.

Other cells. TGF-β is chemotactic for mesenchymal stem cells, recruiting them to the healing site along a gradient. It acts on osteoclasts indirectly, by regulating osteoblast production of RANKL and OPG. It promotes chondrocyte proliferation and cartilage matrix synthesis in the soft callus, and it stimulates fibroblast collagen synthesis, which can lead to fibrosis if excessive.

The Smad2/3 pathway. TGF-β signals through a different pair of R-Smads from the BMPs:

1. TGF-β binds the type II receptor (TGFβR-II, a constitutive serine/threonine kinase)
2. Type II recruits and phosphorylates the type I receptor (TGFβR-I, also called ALK-5)
3. Activated type I receptor phosphorylates Smad2 and Smad3 (the R-Smads specific to TGF-β)
4. Smad2/3 bind Smad4 (the Co-Smad, common to both TGF-β and BMP pathways)
5. The complex translocates to the nucleus and regulates gene transcription

The difference from BMP signalling. TGF-β uses Smad2/3 via ALK-5; BMPs use Smad1/5/8 via ALK-2/3/6. Both pathways share Smad4 as the common mediator, which allows differential gene activation despite similar receptor mechanisms.

Inhibitors. The TGF-β pathway has its own brakes:

• Smad7 is the inhibitory I-Smad specific for the TGF-β pathway
• Decorin is an extracellular antagonist that binds and sequesters TGF-β
• Follistatin binds activins (TGF-β family) and prevents receptor binding

No vessels, no bone. VEGF is absolutely essential for bone healing because angiogenesis is coupled to osteogenesis: without vessels there is no bone formation. Its roles in fracture healing:

• Vascular invasion of the cartilage callus during endochondral ossification
• Coupling angiogenesis to osteogenesis, spatially and temporally
• Osteoblast survival (anti-apoptotic, prevents cell death)
• Osteoclast recruitment, via indirect RANKL regulation
• Delivery of osteoprogenitors via Type H vessels

Sources. Hypertrophic chondrocytes are the highest producers and signal vascular invasion of the soft callus. Osteoblasts maintain the blood supply to forming bone; macrophages and inflammatory cells contribute early and throughout healing; platelets release the first wave in the haematoma.

The family. VEGF is a homodimeric glycoprotein with a heparin-binding domain for matrix retention, and the family has several members:

• VEGF-A - the most important for bone, with multiple splice variants: VEGF121 is freely diffusible with no heparin binding; VEGF165 is the most abundant, partially matrix-bound, and the key isoform for bone angiogenesis; VEGF189 and VEGF206 are fully matrix-bound. The balance of isoforms affects vessel stability
• VEGF-B - limited role in bone; cardiac angiogenesis
• VEGF-C and VEGF-D - primarily lymphangiogenesis
• PlGF (placental growth factor) - synergises with VEGF-A; placental and pathological angiogenesis

Receptors. Three, with very different jobs:

• VEGFR-1 (Flt-1) - a decoy receptor with low signalling activity that sequesters VEGF
• VEGFR-2 (KDR/Flk-1) - the primary signalling receptor on endothelial cells, driving angiogenesis
• VEGFR-3 - primarily lymphatic endothelium

The VEGFR-2 cascade. VEGF-A binds VEGFR-2 on endothelial cells; the receptor dimerises and autophosphorylates (tyrosine kinase activation) and switches on several downstream pathways: PI3K/Akt for survival and migration, MAPK for proliferation, and PLCγ for calcium signalling and permeability. The result is endothelial proliferation, migration and tube formation, which is angiogenesis.

What drives VEGF expression. Hypoxia is the major stimulus, acting through HIF-1α (hypoxia-inducible factor). BMPs and TGF-β stimulate VEGF production in osteoblasts, inflammatory cytokines (IL-1, TNF-α) increase it in macrophages, and mechanical stress upregulates it in osteocytes.

What they are. Type H vessels are a specialised subset of capillaries that couple angiogenesis to osteogenesis. They express high levels of CD31 (PECAM-1) and Endomucin (CD31-high, Emcn-high), sit at the metaphysis and in fracture callus, and create a strongly pro-osteogenic environment: they deliver osteoprogenitors from the circulation and marrow and support perivascular mesenchymal cell differentiation to osteoblasts.

Type L vessels, by contrast, are CD31-low and Emcn-low, lie in the diaphysis (cortical bone), are less osteogenic and primarily deliver nutrients, and do not actively support bone formation.

How the coupling works. The sequence runs:

1. VEGF and other factors induce Type H vessel sprouting
2. Type H vessels grow into the callus along specific trajectories
3. Perivascular mesenchymal cells associate with the vessels
4. Local BMP and Wnt signals drive osteoblast differentiation
5. Bone formation follows the vessel network - osteoblasts line up along vessels and deposit osteoid

View larger

First on the scene. PDGF is the earliest growth factor at the fracture site, released immediately from platelet alpha granules when the fracture haematoma forms. It is released within minutes to hours, peaks at 24-48 hours in the haematoma, is sustained by macrophages and fibroblasts during inflammation (days 3-7), and declines as inflammation resolves.

Isoforms. PDGF is a dimer:

• PDGF-AA - two A chains, binds PDGFR-α
• PDGF-BB - two B chains, the most potent, binds both receptors
• PDGF-AB - heterodimer
• PDGF-CC and PDGF-DD - newer isoforms, less studied

What it does. PDGF is chemotactic, recruiting neutrophils and macrophages to the fracture and attracting mesenchymal stem cells from periosteum, marrow and circulation. It is mitogenic for fibroblasts, smooth muscle cells and osteoblasts, stimulates collagen and proteoglycan synthesis, and promotes angiogenesis indirectly by inducing VEGF expression in stromal cells.

Signalling. PDGF binds the α or β PDGF receptor, both receptor tyrosine kinases, and activates the Ras-MAPK pathway (proliferation), the PI3K-Akt pathway (survival and migration) and the PLCγ pathway (calcium signalling and the cytoskeletal reorganisation needed for migration).

The family. Over 20 fibroblast growth factors are involved in skeletal development, fracture healing and homeostasis. The key FGFs in bone:

• FGF-2 (basic FGF, bFGF) - the most studied in bone healing
• FGF-18 - cartilage development and homeostasis
• FGF-23 - phosphate homeostasis; produced by osteocytes, it regulates renal phosphate handling

FGF-2. A mitogen for osteoprogenitors, chondrocytes and endothelial cells, FGF-2 stimulates alkaline phosphatase expression in osteoblasts, enhances angiogenesis in synergy with VEGF, modulates BMP signalling (enhancing or inhibiting it depending on concentration and context) and maintains the mesenchymal cell pool by preventing premature terminal differentiation. It is produced by osteoblasts and chondrocytes during the repair phases (weeks 2-8), persists at low levels during remodelling and homeostasis, and is upregulated by mechanical loading.

FGF-18. Critical for growth plate chondrocyte proliferation and differentiation and chondroprotective for articular cartilage. Sprifermin, recombinant FGF-18, is in clinical trials for knee osteoarthritis.

FGFR mutations. Activating FGFR3 mutations cause achondroplasia, the most common skeletal dysplasia; FGFR1/2 mutations cause the craniosynostosis syndromes, with premature suture fusion.

IGF-1. The primary systemic anabolic factor for bone and cartilage, IGF-1 stimulates osteoblast proliferation and differentiation, increases type I collagen synthesis, enhances mineralisation of osteoid and is anti-apoptotic, prolonging osteoblast lifespan. It mediates many of the skeletal effects of growth hormone.

Sources. Endocrine IGF-1 is produced by the liver in response to GH and circulates bound to IGFBPs. Autocrine and paracrine IGF-1 is produced locally by osteoblasts, stored in bone matrix and released during osteoclastic resorption.

Signalling and binding proteins. The IGF-1 receptor (IGF-1R) is a receptor tyrosine kinase homologous to the insulin receptor, and it activates the PI3K-Akt pathway (survival, protein synthesis, glucose uptake) and the MAPK pathway (proliferation). Six binding proteins, IGFBP-1 to IGFBP-6, modulate bioavailability by sequestering IGF in the circulation and bone matrix; proteolysis of the IGFBPs releases active IGF at target sites, and IGFBP-3 is the most abundant in the circulation.

Other factors work synergistically with the primary growth factors across the phases of healing:

View larger

BMPs are the only growth factors with robust FDA approval and widespread clinical use in orthopaedics. The products, their carriers and their labels:

How rhBMP-2 works in the wound. BMP bound to its collagen carrier is placed at the fracture or fusion site and released locally over weeks. It induces mesenchymal stem cells to differentiate into osteoblasts and stimulates bone formation without autograft, overcoming the poor osteogenic potential of the graft site: true osteoinduction, bone in a non-skeletal site if that is where the sponge ends up.

Advantages over bone graft. The case for a recombinant protein rests on four points:

• No donor site morbidity (iliac crest pain in 10-20%, infection, haematoma)
• Unlimited supply, not limited by the patient's bone stock
• Consistent osteoinductive potential, whereas autograft quality varies with age and comorbidity
• Avoids a second surgical site, which shortens operative time and is what patients prefer

The evidence. The BESTT trial (2002) was a Level I randomised trial in open tibial fractures in which rhBMP-2 was superior to standard care, with reduced infection, faster healing and fewer secondary procedures. Multiple spine RCTs showed non-inferiority to autograft for ALIF with less donor-site pain, and industry-funded trials showed higher fusion rates with BMP. Meta-analyses find fusion rates 5-10% higher but complications higher too, and the effects are dose-dependent: higher doses increase both efficacy and complications.

rhBMP-7 (OP-1). Held a Humanitarian Device Exemption for recalcitrant long-bone nonunions and required IRB approval for use in the United States. It has lower osteoinductive potency than BMP-2 and potentially fewer inflammatory complications, and it has been withdrawn from many markets.

Platelet-rich plasma. The rationale is to concentrate platelets, and with them growth factors, from autologous blood and deliver supraphysiological doses of PDGF, TGF-β, VEGF, FGF-2 and IGF-1 to the healing site. Whole blood is centrifuged to separate the platelet-rich layer, which is activated with thrombin, calcium chloride or freeze-thaw to release the granule contents, at a concentration factor of 2-8 times platelet baseline (a typical target is 1 million/μL). Growth factor content varies widely with the preparation method:

• PDGF-BB: 2-10 times baseline concentration
• TGF-β1: 2-5 times baseline
• VEGF: 2-10 times baseline
• IGF-1 and FGF-2: present in variable amounts

Why the evidence is weak. Preparation methods are not standardised across studies or clinics, and the studies are heterogeneous in preparation method, dosing, patient selection and outcome measures; platelet-poor plasma may work similarly, since physiological doses may be sufficient. Activated platelets release pro-inflammatory mediators as well as growth factors, the placebo effect of an injection and belief in the treatment is substantial, and the cost is not covered by insurance. Systematic reviews and meta-analyses conclude there is insufficient evidence to recommend routine use; PRP may have a limited role as an adjunct in refractory tendinopathies and is not recommended for bone healing on current evidence.

Bone marrow concentrate. The rationale is to concentrate the nucleated cells of a bone marrow aspirate to deliver mesenchymal stem cells and endogenous growth factors. Marrow is aspirated, typically from the posterior iliac crest, centrifuged to concentrate the nucleated cells (removing red cells and plasma), and the concentrated fraction injected at the treatment site. MSCs make up only 0.001-0.01% of the nucleated cells; haematopoietic progenitors are the majority, cytokines and chemokines come with them, and the growth factor content is lower than PRP because marrow aspirate has fewer platelets.

BMC evidence and limits. For fracture nonunion, case series show benefit but there are no high-quality RCTs against autograft; for osteoarthritis the evidence is minimal, short-term pain reduction at best with no structural benefit; for cartilage defects the benefit is limited and inferior to established procedures (ACI, OATS). MSC potency declines dramatically with age, so efficacy is limited in the elderly; aspiration technique and cell yield vary between operators; there is no FDA-approved indication, so use is off-label on surgeon judgement, and neither preparation nor dosing is standardised. PRP and BMC are autologous and relatively low-risk adjuncts, but they lack robust evidence of efficacy for most applications.

Global regulatory status of growth-factor products

Approved indications are deliberately narrow and consistent across major regulators; almost all heavy clinical use of rhBMP-2 worldwide is off-label.

Major guideline / advisory positions, side by side

Registry and surveillance evidence

Unlike arthroplasty, growth-factor use is not systematically captured by joint registries (NJR England/Wales, AJRR USA, AOANJRR Australia, SHAR Sweden, Norwegian, NZJR). Safety knowledge therefore comes from regulatory pharmacovigilance and large administrative datasets rather than implant registries.

Global practice variation