Tendon and Ligament Healing

OPBSMusculoskeletalBiomechanics

Must Know

  • Tendon and ligament heal by scar with material properties inferior to native tissue. Native microstructure is never restored.
  • Healing runs in three overlapping phases. Inflammatory, proliferative and remodelling.
  • Inflammatory cells (neutrophils, monocytes, lymphocytes, macrophages) clear necrotic tissue in the first week.
  • Tendon fibroblasts start making type III collagen early in healing.
  • Proliferative phase starts within days and lasts several weeks, driven by TGF-β, IGF-1, PDGF, VEGF and bFGF.
  • Blood flow, VEGF and microvessel density peak near 1 week and stay raised for about 4 weeks.
  • Remodelling starts near 6 weeks and lasts months, with consolidation then maturation stages.
  • During consolidation, collagen aligns with tensile load and switches from type III to mainly type I.
  • Healed tendon has more type III collagen, thinner fibrils and abnormal cross-links than native tendon.
  • Epitenon cells proliferate first. Endotenon cells follow later and lay down larger, more mature fibres.
  • MMP-9 and MMP-13 degrade collagen. MMP-2, 3 and 14 both degrade and remodel.

Intrinsic and Extrinsic Healing

FeatureIntrinsicExtrinsic
Cell sourceEpitenon and endotenon tenocytesCells invading from sheath or synovium
ResultBetter gliding, fewer complicationsScar and adhesions
Clinical roleGoal of treatmentDominant mode in clinical ruptures
  • Intrasynovial tendons, such as hand flexors, have a sheath that lowers friction and allows greater excursion.

Intra-articular and Extra-articular Tissue

FeatureIntra-articular (ACL, rotator cuff)Extra-articular (MCL, Achilles)
Healing capacityPoor, slower and limitedBetter
Complete ruptureUsually needs surgeryMCL and Achilles can do well nonoperatively
ReasonSynovial fluid, proteases, no stable fibrin clotClot forms as scaffold
  • Fibrinolysates in the joint stop a stable fibrin clot forming.
  • PDGF, TGF-β and FGF levels may be too low after intra-articular ligament rupture.
  • Patellar tendon is extra-articular but is almost always repaired surgically.
  • A torn ACL grows an α-smooth muscle actin synovial layer over its ends, causing stump retraction and no bridging.

Tendon to Bone Healing

  • Direct insertions (rotator cuff, ACL) have four zones. Tendon, unmineralised fibrocartilage, mineralised fibrocartilage, bone.
  • Tendon to bone healing forms fibrovascular scar. The fibrocartilage transition zone does not regenerate.
  • Causes include avascular fibrocartilage, bone changes, complex loading and a tendon to bone gap.
  • Bone mineral density falls at the insertion after tendon rupture and repair, even late.
  • Rotator cuff repairs fail at the attachment, with tendon pulling through sutures.
  • Acute ruptures heal better than chronic tendinopathy, though some preexisting tendinopathy is usually present.
  • Prostaglandin E2 and IL-6 can induce tendinopathy-like histology.
More on footprint and tendon edge biology
  • In rats, an untreated footprint or microfracture beat radiofrequency ablation or excavation to cancellous bone.
  • Vitamin D deficiency lowered load to failure in a rat cuff repair model.
  • Near the torn edge, caspases 3, 8 and 9, BAX, TNF-α, IL-1β and interferon gamma rise.
  • Further from the edge, Bcl-2, IL-10 and type I collagen expression rise.
  • Postnatal adult healing is driven by inflammation. Embryos heal without scar before the immune system develops.
  • Macrophage depletion in a rat ACL model gave a narrower scar interface and greater strength.
  • ED1 proinflammatory macrophages come first, then ED2 proregenerative macrophages acting via TGF-β.

Mechanical Load

  • Some load is needed. Stress shielding harms tendon and bone.
  • Early excessive load impairs healing, possibly by stopping the inflammatory response resolving.
  • Brief initial immobilisation gives a stronger tendon to bone complex with better collagen organisation.
  • Prolonged immobilisation weakens the complex and raises MMP-13 activity.

ACL Reconstruction

  • Soft tissue grafts rely on tendon to bone tunnel healing. BTB adds bone to bone healing.
  • The tunnel interface forms granulation tissue then Sharpey-like type III collagen fibres.
  • Tunnel healing takes 8 to 30 weeks in animal studies.
  • NSAIDs, both selective and nonselective, impair tendon to bone healing in animal models.
  • Longer tunnels, press fit and maximal graft contact improve healing.
  • Interference screws stiffen the construct but may reduce graft to tunnel contact.
  • Graft micromotion is greatest at tunnel apertures, where healing is poorest and osteoclasts gather.
  • BTB plugs undergo ischaemic necrosis and creeping substitution, incorporated by 6 weeks.
  • BTB pull-out strength beats soft tissue at 3 weeks, with no difference at 6 weeks.

Biological Augmentation

StimulatoryInhibitory
BMP, TGF, FGF, PDGFDiabetes, thyroid and renal disease, osteoporosis
BMAC, PRPSmoking, NSAIDs, quinolone antibiotics
Teriparatide (PTH 1 to 34)MMPs, TNF-α, IL-17
Extracellular matrix patch
  • BMAC comes mainly from the iliac crest. Small 10 mL syringes maximise progenitor yield.
  • BMAC stem cell yield is only 0.001 to 0.01 percent, but it carries PDGF, TGF-β, BMP-2 and BMP-7.
  • PRP comes from the upper buffy coat after centrifuging autologous blood.
  • The fibroblast is the main cell making new matrix in tendon and ligament healing.
  • Platelet-rich fibrin matrix at cuff repair was linked with a tendon defect at 12 weeks, odds ratio 5.8.
  • BMP-12 and BMP-13 induce tendon and fibrocartilage. BMP-2, 4 and 7 are osteoinductive.
  • Teriparatide works through intermittent, once daily PTH exposure that preferentially activates osteoblasts.
  • Acellular dermis is the most used cuff patch. Patches heal by macrophage infiltration then fibrovascular scar.
More on augmentation studies
  • PRP gel gave earlier ACL graft homogeneity on MRI, 179 versus 369 days, without reported clinical outcomes.
  • A randomised trial of PRP with BTB found no difference at 2 years.
  • A bovine osteoinductive extract gave more bone and fibrocartilage, but strength per area was unchanged.
  • Xenograft, allograft and synthetic patches report structural healing of 60 to 100 percent in small series.
  • Patches act as augmentation (overlay) or bridging (in series with native tendon).
  • Allografts showed a higher MRI signal to noise quotient at 2 years with similar clinical outcomes.
  • MRI graft maturity does not match clinical scores and should not set return to sport timing.
  • Loading raises tenocyte genes such as tenomodulin, without chondrocyte or osteocyte genes.
  • Botulinum toxin A at birth in supraspinatus reduced fibrocartilage and mineralisation at the insertion.
  • Chondroid cells line the ACL tunnel wall laying down type II collagen by endochondral bone formation.
  • Aggressive rehabilitation with hamstring grafts causes more tunnel widening.
  • PRP comes leucocyte-rich or leucocyte-poor, liquid or semisolid. Results are largely equivocal.
  • Internal bracing with suture tape risks overconstraint and stress shielding of the graft.
  • Healed MCL does well clinically despite persistently abnormal mechanical and histological properties.
  • Symptomatic rotator cuff disease affects up to 30 percent of people.

Numbers to Remember

ItemValue
Inflammatory phaseAbout 1 week
Vascular peakAbout 1 week, raised to 4 weeks
Remodelling startsAbout 6 weeks
ACL epiligamentous regeneration8 to 12 weeks
Tunnel healing (animal)8 to 30 weeks
BTB plug incorporation6 weeks
Primary ACL repair instabilityUp to 94 percent at 5 years
BMAC stem cell yield0.001 to 0.01 percent

Memory Aids

  • Phases I P R. Inflammation 1 week, Proliferation weeks, Remodelling from 6 weeks.
  • Collagen 3 then 1. Scar starts as type III and matures towards type I.
  • Four zones TUMB. Tendon, Unmineralised fibrocartilage, Mineralised fibrocartilage, Bone.

Exam Traps

  • NSAIDs, including COX-2 inhibitors, worsen tendon to bone healing in animal models.
  • Repaired entheses heal by fibrovascular scar, not regenerated fibrocartilage.
  • Extrinsic healing causes adhesions. Intrinsic healing is the desired pathway.
  • Micromotion and healing problems concentrate at the tunnel aperture, not the tunnel exit.
  • The patellar tendon is extra-articular yet still needs surgical repair.
  • BMP-12 and BMP-13 are tenogenic, not osteoinductive.

Related pages. Rotator Cuff Tears, Achilles Tendon Ruptures, Anterior Cruciate Ligament Rupture | ACL, Peroneal Tendon Injuries, Platelet Rich Plasma

Recent Review. Mid-Substance Achilles Tendinopathy

  • Mid-substance tendinopathy lies 2 to 6 cm above the calcaneal insertion, a zone of relative hypovascularity.
  • Achilles fibres spiral up to 90 degrees as they pass distally. Blood supply comes through the paratenon.
  • Histology shows more and disorganised tenocytes, excess glycosaminoglycan, fragmented collagen and neovascularisation, without inflammatory cells.
  • Progression gives mucoid or lipoid degeneration with fibrocartilaginous metaplasia and calcific deposits.
  • Tendon homeostasis depends on a balance between matrix metalloproteinases and tissue inhibitors of metalloproteinases (TIMPs).
  • Neovessels and accompanying nerve ingrowth on the ventral tendon are thought to generate pain.
  • Systemic risks include fluoroquinolones, corticosteroids, diabetes, gout, inflammatory arthropathy and raised BMI.
  • NSAIDs do not help long term because the process is degenerative.
  • Eccentric loading (Alfredson, 3 sets of 15, twice daily for 12 weeks) has the best evidence and beats concentric loading.
  • PRP showed no benefit over saline or sham injection in RCTs. Corticosteroid injection is advised against.
More on diagnosis and second-line options
  • About 75 percent resolve without surgery. About 20 to 29 percent are recalcitrant.
  • VISA-A has 8 items scored 0 to 100. Healthy athletes score above 96, but nonathletes may not exceed about 89.5.
  • Royal London Hospital test is positive when pain eases with the ankle dorsiflexed (sensitivity 51 percent, specificity 93 percent).
  • Arc sign is positive when a nodule moves with ankle motion (sensitivity 25 percent, specificity 100 percent).
  • Ultrasound is the first-line imaging and correlates with histology. MRI is kept for unclear cases.
  • High-volume image-guided injection distends the space between tendon and Kager fat pad to disrupt neovessels.
  • Glyceryl trinitrate acts by releasing nitric oxide, with conflicting RCT results.
  • Tendon transfer, usually FHL, is advised when more than 50 percent of the tendon is debrided.

Venugopal V, Fleming TA, Mee M, Grice JE. An update on current concepts and management of mid-substance Achilles tendinopathy. Bone Joint J. 2026;108-B(3):369-80. PubMed

Recent Review. Biologics in Rotator Cuff Repair

  • Retear rates after rotator cuff repair are reported at 15 to 40 percent.
  • Tendon healing passes through inflammation, proliferation from about day 2, then remodelling from 1 to 2 months.
  • Healing tenocytes first lay down type III collagen, later replaced by type I collagen, which gives tensile strength.
  • Degenerate tendons hold excess type III collagen, implying failure to complete remodelling.
  • Platelets release TGF-β, IGF, PDGF and FGF, recruiting progenitor cells to the repair.
  • The anterolateral watershed zone has less blood supply than the bony insertion. Synovial fluid prevents clot formation in articular-side tears.
  • Orthobiologics are grouped as blood derived (PRP, PRFM), stem cell (bone marrow or adipose MSCs) and tissue grafts.
  • PRP activated with thrombin forms platelet-rich fibrin matrix, giving sustained growth factor release.
  • After multivariate analysis, fatty infiltration, retraction and low bone mineral density predict healing, while age does not independently.
  • Nicotine delays tendon-to-bone healing and lowers type I collagen. Postoperative fluoroquinolones raise failure.
More on augmentation evidence
  • PRP varies with platelet and leucocyte concentration (leucocyte-rich or leucocyte-poor), activation and delivery, which hampers comparison.
  • Intraoperative PRP improved tendon-to-bone healing in some trials without a clinical difference at one year.
  • Bone marrow aspirate concentrate can be harvested from iliac crest, proximal tibia, humeral head or calcaneum.
  • One RCT of bone marrow aspirate improved structural repair without better PROMs at 3.2 years.
  • Dermal allograft augmentation outperformed xenograft and primary repair alone in a systematic review.
  • A bovine collagen implant lowered retear at one year in a level I trial (92 versus 74 percent intact).
  • NSAIDs impair tendon-to-bone healing in animals, but a celecoxib RCT showed no difference in healing.
  • Doxycycline inhibits matrix metalloproteinases and improved tendon-to-bone healing in animal work.
  • Smokers had a failure rate of 29.4 percent against 5.9 percent in matched non-smokers in one series.

Savoie FH 3rd, Delvadia BP, Tate JP, Winter JE, Williams GH, Sherman WF, O’Brien MJ. Biologics in rotator cuff repair. Bone Joint J. 2024;106-B(9):978-85. PubMed

Recent Review. Subacromial Bursa and Cuff Healing

  • About 50 percent of people aged 65 or over have a rotator cuff tear. Repair failure ranges from 20 to 94 percent.
  • Codman (1906) saw the bursa as indispensable for abduction and rotation, limiting disease in adjacent structures.
  • Neer (1972) introduced anterior acromioplasty with bursectomy for chronic impingement on the extrinsic (mechanical) theory.
  • The intrinsic theory attributes impingement symptoms to degenerative change within the cuff tendons.
  • Bursectomy alone matched bursectomy with acromioplasty, and decompression adds no clinically meaningful benefit to repair outcomes.
  • The bursa is highly vascular, while the supraspinatus tendon is relatively avascular.
  • Bursal connective tissue progenitor cells behave like MSCs and form colonies faster than bone marrow progenitors.
  • A reparative healing response is seen on the bursal side of torn cuff tendons.
  • Inflamed bursae hold 50 to 80 percent CD2+ T cells and 10 to 40 percent CD11b+ macrophages, with few plasma cells.
  • Bursitis upregulates TNF-α, IL-1α, IL-1β, IL-6, MMP-1, MMP-9, COX-1, COX-2 and VEGF. Substance P correlates with pain.
More on targeting the bursa
  • Bursae form where tissues glide over bony prominences without a cartilaginous joint, such as subacromial, prepatellar and olecranon sites.
  • With advancing cuff disease the bursa becomes more fibrous, fatty or vascular.
  • Some healing tendon αSMA+ cells may come from the bursa.
  • Complete suppression of inflammation inhibits tendon healing, and high-dose steroid harms tendon.
  • Low-dose dexamethasone protected tendon explants and limited IL-1 driven cartilage breakdown in animal work.
  • The bursa could act as a natural depot for local drug delivery without hydrogels or scaffolds.

Marshall BP, Levine WN, Thomopoulos S. The role of the subacromial bursa in rotator cuff healing. Friend or foe. J Bone Joint Surg Am. 2023;105(5):417-25. PubMed

Source

Lin KM, Rodeo SA. Tendon and Ligament Healing. In. Aaron RK, editor. Orthopaedic Basic Science. Foundations of Clinical Practice. 5th ed. Philadelphia, Wolters Kluwer, 2021.

Practise Type X questions on this topic

Reviewed by Professor Phong Tran, Head of Orthopaedic Surgery, Western Health. Last updated 17 October 2026.