Clinicians frequently encounter patients with a painful tendo Achillis (TA). They present with pain, swelling, and reduced function.
The TA is the thickest and strongest tendon in the body. The gastrocnemius–soleus complex is a tri-articular muscle; it crosses the knee, ankle, and subtalar joints. It is innervated by the posterior tibial nerve. The medial and lateral heads of gastrocnemius originate, respectively, from the posterior aspect of the medial and lateral femoral condyles. The soleus muscle has a complex origin from the posterior aspect of the tibia, fibula, and interosseous membrane. The two muscles coalesce to form the TA.
The Tendo Achillis
The TA varies in length, but averages 15 cm in adults (11–26 cm). At the proximal end it can measure up to 8 cm wide, narrowing to its thinnest cross-section in the midportion (1.5–2.5 cm). It then widens to around 4 cm above the insertion, from where it gradually widens and flattens as it reaches the calcaneal insertion. At the posterior calcaneus midpoint it measures 2 to 4.8 cm in width.
The TA is composed of cellular and extracellular matrix. The cellular component is primarily fibroblasts, which make up around 20% of the volume. The extracellular matrix accounts for around 80% and is mainly type I collagen, with lesser quantities of type III and IV. Collagen makes up over 70% of the extracellular matrix, with the remainder being proteoglycan-rich ground substance to stabilize the collagen fibrils, and elastin allowing the network of collagen to stretch and recoil.
The collagen fibrils are wrapped in a thin endotenon, which in turn is enveloped by epitenon. The paratenon, or peritendineum, is a thin layer of areolar connective tissue around the epitenon, which further protects and stabilizes the tendon.
The collagen fibers of the tendon are not vertical but orientated in a spiral manner. As the fibers descend from the musculotendinous junction they spiral through 90° and come to lie in the configuration shown in Figure 15.1. The degree of fiber rotation depends at which level the gastrocnemius and soleus muscles fuse to form the tendon – a more distal fusion resulting in more rotation. A rotational fiber pattern allows tendon elongation and recoil during locomotion, with efficient energy storage and dissipation. It also ensures less fiber buckling when the tendon is lax and less deformation under stress, which increases overall tendon strength.
Figure 15.1 Cross-section of the tendo Achillis 1 cm above the calcaneum. (Fascicles of the adult human tendo Achillis, adapted from Szaro et al.1).
The blood supply to the TA is poor and worsens with age. The tendon is supplied by two arteries – the posterior tibial and peroneal. There are three vascular territories. The midsection is supplied by the peroneal artery, and the proximal and distal sections are supplied by the posterior tibial artery2. The vascularity of the midsection, 2 to 7 cm from the calcaneal insertion, is poorest, with blood entering through the anterior surface. The proximal third of the tendon receives vessels via the muscle bellies. The distal third receives its supply at the level of the insertion and the supply travels proximally. The midsection is the site most commonly affected by tendinopathy.
The paratenon is vascular although vessels may not be uniformly distributed throughout its length.
The tendon and paratenon are innervated by the same nerves as the musculature, with small branches from the cutaneous nerves, particularly the sural nerve.
The posterior calcaneal tuberosity is divided into three facets. The upper and middle facets are separated by a shallow groove, the middle and lower facets by a rough ridge. The upper facet lies anterior to the retrocalcaneal bursa. The TA inserts onto the middle and lower facets, and is confluent with the plantar fascia3.
The superficial calcaneal bursa lies superficial to the TA between the tendon and the skin. Deep to the tendon is the retrocalcaneal bursa. The superior facet of the calcaneal tuberosity forms the anterior wall of the retrocalcaneal bursa and the tendon lies posteriorly. Periosteal fibrocartilage overlies the tuberosity anteriorly. Sesamoid fibrocartilage overlies the tendon posteriorly. On the inner surface of the fibrocartilage is a synovial membrane, it is mainly smooth on the posterior TA surface but richly layered with vascular folds on the anterior aspect.
Kager’s triangle is often described on imaging and is bounded inferiorly by the superomedial calcaneum, anteromedially by the flexor hallucis longus (FHL) tendon, and posteriorly by the TA (Figure 15.2). It contains Kager’s fat pad. Three parts of the fat pad have been identified:
Figure 15.2 The MRI appearance of Kager’s triangle, containing Kager’s fat pad. S: soleus; FHL: flexor hallucis longus; K: Kager’s fat pad. The three arrows delineate the tendo Achillis.
The TA-associated part is encapsulated by the paratenon and held to the TA. The FHL-associated part lies anteriorly and within the FHL tendon sheath. Between these two parts is the retrocalcaneal bursal wedge.
Kager’s fat pad minimizes pressure changes within the retrocalcaneal bursa as the wedge of fat displaces into the retrocalcaneal space on plantar flexion of the foot. It has an excursion of 10 to 12 mm, on loadbearing dorsiflexion/plantar flexion. The fat pad also protects the tortuous vascular supply to the TA in its TA-associated part.
The plantaris tendon is present in approximately 93% of individuals. It arises laterally in the knee from the linea aspera and oblique popliteal ligament. It then runs between the gastrocnemius and soleus muscles. It passes obliquely, coming to lie medial to the TA, inserting on the posterior calcaneus. In 3% of individuals there is a variable insertion, 1 to 16 cm above the calcaneus, into the TA or calf complex itself4. The role of the plantaris in tendinopathy of TA is being increasingly recognized and may explain the reason why symptoms can be more severe on the medial side of the TA.
There are a number of different descriptions of TA disorders. The term “tendinitis” of the TA has been recognized distinctly as having “insertional” and “non-insertional” components. Authors have since suggested abandoning the term “tendinitis” and replacing it with “tendinopathy,” as histologically there is no inflammatory change5; there is in fact an incomplete healing response to repeated micro-injury or overuse. This also helps to distinguish between “paratendinopathy” and “tendinopathy.” Paratendinopathy is inflammation or degeneration of the paratenon, which surrounds the tendo Achillis.
Many incorrectly use the term “insertional tendinopathy” to describe any disorder located around the TA insertion. It is important to appreciate other distinct causes of pain around the TA insertion, which may coexist and make diagnosis difficult. The following conditions are included under insertional disorders:
The incidence of tendinopathy of TA is rising, largely as a result of mass participation in recreational activities with increasingly demanding training regimens. Fifty-five to 65% of tendinopathy of TA is midportion tendinopathy, insertional disorders are about 25%, with 3% ruptures, and 8% myotendinous pain6.
The sports most associated with TA disorders are middle- and long-distance running, with up to 10% of runners affected7.
Normal aging can affect the tendon in a similar fashion as a result of overuse. A cadaveric study demonstrated that 34% of previously healthy individuals have evidence of tendinopathic change8.
There is a correlation between tendinopathy and chronic disease, such as diabetes mellitus, obesity, hypertension, hypercholesterolemia, and mixed hyperlipidemia9.
Tendinopathy of TA is multifactorial but research into the genetic susceptibilities to tendinopathy has revealed links to a variety of factors. Collagen forms 80% of the dry mass of tendons and genetic polymorphisms within the genes for collagen types 1, 2, and 12 have been linked with tendinopathy of TA and rupture. For example, individuals carrying a single nuclear polymorphism, the TT genotype of the GDF5-rs143383 variant, have twice the risk of tendinopathy compared with non-carriers10. Homeostasis and remodeling of the extracellular matrix is controlled in part by matrix metalloprotease (MMP) enzymes, tissue inhibitors of MMPs (TIMPs), and growth factors such as transforming growth factor β (tgf-β). Imbalances in the remodeling of the extracellular matrix can lead to collagen disturbances and tendinopathic change. Genotypes of these enzymes and growth factors have been linked to tendinopathy of AT and rupture. The deleterious effect of quinolone antibiotics (ciprofloxacin) is also mediated by MMPs and TIMPs.
The major cell type in the tendon is the tenocyte. Damaged tenocytes are removed by apoptosis. However, repetitive loading can cause excessive tenocyte apoptosis such that the tenocytes are unable to maintain the extracellular matrix and the tendon does not heal normally. The apoptosis signaling cascade has been implicated in the development of tendinopathic changes11, and in turn this pathway has genetic polymorphisms that are linked to tendinopathy of TA.
Epigenetics, the effect of external or environmental factors on genetic expression, is also an area that is being increasingly recognized as important, as aging and exercise regimes can modulate the expression of genetic material. This may help to explain why some individuals develop musculoskeletal conditions, while others do not.
Insertional Tendinopathy of Tendo Achillis
Predisposing factors to insertional tendinopathy of TA include increasing age, seronegative inflammatory arthropathy, corticosteroids, diabetes mellitus, hypertension, obesity, gout, hyperostotic conditions, hyperlipidemias, and medications such as quinolone antibiotics.
There are extrinsic and intrinsic etiological factors in insertional tendinopathy of AT.
Increased, repetitive loading, which is common in runners7 and athletes in “springing” sports, such as badminton, causes insertional disease. New training regimens can also trigger stress-related structural change. Increased hindfoot eversion, reduced ankle dorsiflexion speed, and knee flexion during running are associated with tendinopathic change12.
Coronal plane malalignment of the knee, ankle, and particularly the subtalar joint predisposes to insertional disorders. Both varus and valgus malalignment may cause the tendon to shorten. Malalignment also causes atypical stress, which can initiate a sequence of repetitive microtears and tendinopathic changes. Thus shoes with uneven wear can contribute to disease as a result of excessive subtalar joint movement. Poor shock absorption and uneven surfaces may also be contributors.
Sagittal malalignment and calcaneal pitch can influence the interface between the tendon and calcaneum producing a bursal inflammatory response, rather than degenerative change.
There are four principal pathologies of the TA insertion: Haglund’s deformity; enthesophytes; the retrocalcaneal bursa; and degeneration of the tendon itself.
1. Prominence of the posterosuperior calcaneum, originally described by Haglund, is associated with local attritional wear of the TA insertion causing pain, swelling, and tenderness. The so-called “Haglund’s deformity” probably also contributes to retrocalcaneal bursitis.
2. The anterior surface of the TA insertion is often more affected than the posterior surface in insertional tendinopathy, despite the posterior surface undergoing a higher strain on dorsiflexion13. Insertional spurs, or enthesophytes, form by endochondral ossification of the enthesis fibrocartilage on the anterior, stress-shielded aspect of the TA14. Thus the precise role loading has to play is complex. Enthesophytes do not appear to need preceding microtears or inflammatory change to initiate them.
3. The retrocalcaneal bursa is lined with synovium and fibrocartilage. These layers are apposed during ankle dorsiflexion and the tendon is compressed against the calcaneum. Synovial fold hypertrophy, calcification of the sesamoid fibrocartilage, and cellular degeneration with bursal debris have been demonstrated in the retrocalcaneal bursa14. This may be the cause of retrocalcaneal bursitis.
4. The poor tendon blood supply seen in normal entheses reduces further with age. Thus midsubstance tendinopathy may exacerbate insertional tendinopathy.
Non-Insertional Tendinopathy of Tendo Achillis
As with insertional tendinopathy of TA, shoes have been linked to non-insertional tendinopathy. The shoes may cause slight varus or valgus of the heel and therefore change the direction of pull of the TA, adversely affecting the tendon.
Positioning of the foot while running can lead to tendinopathy. Hyperpronation of the hindfoot, in particular, is related to the development of tendinopathy of AT. As with insertional tendinopathy, an increase in activity can adversely affect the TA. Similarly, environmental factors such as changes in the surface or surface orientation can influence ankle biomechanics and the resultant strain force on the tendon itself.
The tendon itself is predisposed to tendinopathic change by virtue of its shape. As described above the collagen fibers spiral through 90° so the medial fibers proximally lie posteriorly distally. This produces areas of high stress, which could be a factor in ischemic change in the tendon.
Neovascularization on US scanning of the tendinopathic TA has been the focus of recent study15. The evidence is somewhat conflicting and confusing. It is assumed that normal tendons exhibit no vascular flow on color Doppler US, whereas tendinopathic tendons exhibit increased flow16. Increased vascularity is thought to be a physiological, adaptive response to increased load, although it may be a direct reaction to hypoxia secondary to reduced vascularity and degeneration. Alfredson has correlated pain with histological tendinopathic change, showing that neovascularization at the site of pain correlates with biopsies of tendinopathic material17. On the other hand intratendinous flow does not increase, in fact it may decrease, with repetitive loading18.
The plantaris tendon is closely related to the medial aspect of the tendo Achillis. It is stiffer than the TA and acts as a weak hindfoot invertor. The soleus muscle forms the medial fibers of the TA and is an ankle plantar flexor. The differing action and stiffness of plantaris and the medial TA can produce differing excursions of the two tendons during gait. This is proposed as a cause of TA pain, especially in patients complaining of medial symptoms.
The history of insertional and non-insertional disease differs. Pertinent features for both conditions include whether the onset was associated with an increase in activity, alteration in shoes, and the presence of a generalized inflammatory arthropathy or systemic disease.
Non-Insertional Tendinopathy of Tendo Achillis
The patient with non-insertional tendinopathy often describes the gradual onset of pain, but may also remember an isolated incident. Symptoms may have been present for years. In more severe cases, patients complain of pain even with the activities of daily living. Typically there is stiffness in the morning and after periods of sitting. Night pain can be a feature. There is often variable swelling and tenderness of the tendon.
The history is usually more acute with a rapid onset of symptoms, which include diffuse swelling and pain around the tendon, sometimes associated with erythematous change. The pain is usually in the middle third of the tendon and can be more prominent medially.
Insertional Tendinopathy of Tendo Achillis
The main feature is pain located at the insertion of the TA. Affected populations include the younger athlete in running and jumping sports, who has often started a new regimen, and the sedentary older individual who presents with a more chronic picture, and may or may not describe progressive planovalgus foot deformity. The posterior heel pain is sharp and exacerbated by rubbing shoes, with increased pain on start-up.
Retrocalcaneal pain is worse on walking or running. The pain is “to the side” or “in front” of the tendon insertion. The pain is exacerbated by activity or new shoes. Use of NSAIDs often dramatically improves the symptoms.
For both insertional and non-insertional disorders the general alignment is important, as planovalgus and cavovarus can predispose to heel pain. The patient should be observed standing and walking. The presence of calf wasting should be noted.
In paratendinitis the tendon is exquisitely tender and swollen with crepitus being felt on movement. In both insertional and non-insertional tendinopathy the affected area is tender, firm, and swollen. There is usually a discrete, fusiform swelling of the tendon in non-insertional tendinopathy, although occasionally the tendon may be diffusely involved (Figure 15.3). A posterior prominence at the insertion is seen in insertional tendinopathy (Figure 15.4). With retrocalcaneal bursitis the skin is typically warm and erythematous with the swelling and tenderness localized just anterior and to the sides of the tendon. The range of motion in dorsiflexion is often limited, although plantar flexion strength is usually normal.
Figure 15.3 The clinical appearance of a thickened, diffuse non-insertional tendinopathy of tendo Achillis on the left side.
Figure 15.4 The clinical appearance of an insertional tendinopathy of tendo Achillis with posterior prominence on the right side.
The role of plain radiographs is limited in non-insertional tendinopathy, although calcification of the tendon is occasionally seen. However, in the investigation of “heel pain,” or insertional tendinopathy, plain radiographs, specifically a weightbearing lateral view, are much more useful. The three major features to note are:
2. ossification of the insertion of the tendon (enthesophytes, see Figure 15.6)
3. swelling in Kager’s fat pad, anterior to the usually well-demarcated tendo Achillis.
Figure 15.5 A large posterosuperior prominence of the calcaneum, demarcated by Pavlov’s parallel pitch lines.
Figure 15.6 Enthesophytes at the distal insertion of the tendo Achillis.
Ultrasound is useful both diagnostically and therapeutically in tendo Achillis disease. It is operator dependent, but can also be used to guide injection therapies.
In the prone position, patients are assessed with a linear high-frequency probe (7–18 MHz) examining in the transverse and longitudinal planes. The use of color Doppler shows neovascularization. The normal tendon exhibits an echogenic pattern of parallel, fibrillar lines in the longitudinal plane, and a round or oval shape in the transverse plane (Figures 15.7 and 15.8).
Figure 15.7 A normal tendo Achillis in the longitudinal plane. TA: tendo Achillis; top right arrow indicates the paratenon.
Figure 15.8 A normal tendo Achillis in the transverse plane. Arrows point to the tendon.
Non-insertional tendinopathy of TA may be indistinguishable clinically from paratendinitis, but in paratendinitis the US shows a normal tendon with a circumferential hypoechogenic halo, which may be hypervascular on power Doppler. Contrast-enhanced MRI has probably superseded US for the diagnosis of isolated acute paratendinitis.
In insertional disorders partial tears may be demonstrated, usually on the ventral surface with expansion of the tendon in the AP dimension. There is heterogeneous loss of reflectivity and the normal fibrillar pattern. Enthesophytes and a Haglund’s deformity are also visualized.
The retrocalcaneal and subcutaneous calcaneal bursae are not demonstrable by ultrasonography in healthy people. Abnormal bursae are usually well defined, although US is poorly sensitive but highly specific in making the diagnosis of retrocalcaneal bursitis compared to MRI. In a true bursitis increased blood flow can be seen on color or power Doppler.
Magnetic Resonance Imaging
The normal MRI scan of the TA shows a low signal, homogeneous structure on T1 and T2 weighted images (Figure 15.2). Standard sequences include T1-weighted and fluid-sensitive sequences such as short-tau inversion recovery (STIR) and fat-suppressed T2, or intermediate (proton) density weighted sequences in the axial and sagittal planes.
The sensitivity of MRI scanning in detecting TA abnormalities is 94% with a specificity of 81% and positive predictive value of 90%21. The tendinopathic TA on MRI appears with a variable amount of ill-defined longitudinal high signal, with fusiform expansion of the tendon on the sagittal images. Enhancement of the paratenon is best visualized with gadolinium contrast. Presence of a calcaneal prominence, especially with bone edema, is of significance (Figure 15.9).
Figure 15.9 Sagittal MRI showing edematous changes in the posterosuperior calcaneum, an inflamed retrocalcaneal bursa, and tendinopathic change with inflammatory change in the posterior paratenon.
Magnetic Resonance Imaging or Ultrasound?
Ultrasound shows neovascularization and images the tendon dynamically. It also has a therapeutic role with injection therapies. MRI may be more useful to detect multiple lesions (Figure 15.9). Sagittal MRI is the optimal tool to diagnose a significant retrocalcaneal bursitis, especially when coupled with insertional pathology, although US shows vascularization in cases with insertional tendinopathy. Both modalities clearly demonstrate tendon thickening and partial tears.
The treatment of tendinopathy of TA is multimodal. The requirements of the low-demand patient with multiple comorbidities differ from those of the elite athlete. Treatment also depends on the area of pathology. Pathologies can coexist, for example in non-insertional disorders an acute paratendinitis can coexist with a tendinopathy. There is no one strategy for either insertional or non-insertional disorders and therefore complicated treatment algorithms have been avoided.
Non-Steroidal Anti-Inflammatory Drugs (NSAIDs)
NSAIDs have been shown to have a modest effect on symptoms from three trials reported in the Cochrane Review22, but a randomized study found no benefit over placebo23. Additionally, the scientific basis for their use in chronic tendinopathy is questionable in view of the histological absence of inflammatory cells in the tendinopathic tissue. Thus their short-term benefit is likely to be due to their analgesic effect. There are some studies that highlight the possible detrimental effects of NSAIDs. Celecoxib inhibits tenocyte migration and proliferation24 and NSAIDs increase leukotriene B, which may contribute to the development of tendinopathy of TA25.
Eccentric exercises are supported by multiple studies. Nevertheless, the therapeutic effect of eccentric exercises is poorly understood. Öhberg showed that they lead to normalization of the tendon structure and reduction in the neovascularization seen on US26. Rees demonstrated increased stretching of the tendon with eccentric, when compared to concentric, exercises27. Alfredson, in a randomized study, reported 82% of individuals returning to normal activities at 12 weeks with eccentric, as opposed to 36% with concentric, stretches28. Many now use a 12-week regimen as the gold-standard therapy for treating non-insertional tendinopathy; however, shorter, six-week regimens have been described with reasonable results29. Clinical trials comparing different eccentric protocols are lacking and thus it remains to be seen whether a shorter regimen is as effective as a 12-week regimen.
In trials, extra-corporeal shockwave therapy (ESWT), typically low energy, has given rise to conflicting results, although recent randomized trials have demonstrated improvement when ESWT is combined with eccentric exercises, when compared to eccentric exercises alone30. A non-randomized case-control study of patients failing to improve following at least three non-operative treatments, for a minimum of six months, found improved visual analogue scores at 12 months in the ESWT group31.
The mechanism of action of ESWT is not absolutely clear, but ESWT causes cavitation of the tendon with interstitial and extracellular disruption leading to a healing response32. At the cellular level, ESWT promotes cell growth and collagen synthesis in cultured human tenocytes. It is suggested that this increases the efficacy of tendon repair after injury33. Changes in TGF-β1 and IGF-1 expression and decrease in some interleukins and MMPs have been demonstrated in rat and human cultured tenocytes34.
Extra-corporeal shockwave therapy also causes selective dysfunction of sensory unmyelinated nerve fibers and changes in the dorsal root ganglia. It appears to have a role in treatment if other conservative measures have failed, as it is safe and inexpensive.
Corticosteroid injections are reported to reduce pain and swelling, and improve the US appearance of the TA. This may be through their vasoconstrictive effects35. Corticosteroid injections may have some early benefit but later complications are reported including tendon rupture36 and the risks appear to outweigh the benefits.
A randomized double-blind, placebo-controlled study demonstrated impressive improvement in symptoms from glyceryl trinitrate (GTN) patches applied over the tendon. A follow-up study showed the benefit was maintained in the GTN compared to the control group37. Doubts have been raised as to the blinding of this study, as the GTN group required significantly more paracetamol to treat headaches. A subsequent randomized study demonstrated no additional benefit for GTN patches38. More worryingly, increased nitric oxide levels have also been implicated in the development of degenerative conditions, including tendinopathy39. It has been suggested therefore that GTN may in fact be detrimental to the underlying pathological process.
Dry needling and autologous blood injections are said to provide cellular and humoral mediators, which promote the healing of tendinopathy. Good results are reported in the treatment of medial epicondylitis of the elbow and patellar tendinopathy, although no good quality studies have reported benefit in tendinopathy of TA.
Platelet rich plasma (PRP) is widely used in orthopedic practice; however, recent publications have failed to demonstrate significant improvement using PRP to treat tendinopathy of TA40. A meta-analysis concluded that while there may be benefit in using PRP to increase the healing strength in tendo Achillis repair following acute rupture, there was no evidence to support the use of PRP in the treatment of tendinopathy of TA41.
High-volume injections with limited follow-up have demonstrated reduced pain and improved function following high-volume injection (10 ml bupivacaine and 40 ml normal saline) anterior to the tendo Achillis42. However, they also injected 25 mg hydrocortisone, which has been shown to provide early symptomatic improvement, but is associated with a higher rate of later complications. It should be noted that there were no control groups, and the results were early, at 30 weeks follow-up. It is proposed that the mechanical effect of the volume of fluid causes damage to the neovessels and neural ingrowth.
Intratendinous hyperosmolar dextrose (prolotherapy) has been used since the 1940s and is thought to produce a local inflammatory response and increase in tendon strength. Evidence to support prolotherapy is lacking. A pilot study by Maxwell et al. demonstrated that ultrasonographically guided injection of 25% dextrose led to a reduction in TA pain both at rest and during exercise43.
Aprotinin is a potent MMP inhibitor and some studies demonstrated its success in treating tendinopathy of TA; however, it was withdrawn in May 2008 due to severe complications in its use during open heart surgery.
Sclerosants have been used to reduce neovascularization in the tendinopathic tendon. Early reports using polidocanol injected under Doppler US guidance into the abnormal vessels on the ventral aspect of the TA led to significant improvements in pain and function44. This was a small study with limited follow-up and has not been reproduced at other institutions.
Finally, electrocautery has been used to destroy the neovascularization, with one study reporting good results in a series of 11 patients followed up for six months45.