Management of Subcutaneous Tears of the Achilles Tendon:
How I Do It

Nicola Maffulli MD, MS, PhD, FRCS(Orth)
Clinical Senior Lecturer and Honorary Consultant in Orthopaedic Surgery
Department of Orthopaedic Surgery
University of Aberdeen Medical School
Foresterhill, Aberdeen, SCOTLAND

Published in the Bulletin of the Hospital for Joint Diseases.

The Achilles tendon connects the gastro-soleus unit to the calcaneum. Proximally, the aponeurosis of the gastrocnemius (ventrally) and of the soleus (dorsally) merge into each other, and rotate simultaneously to form a compact tendon, the Achilles tendon (White 1943). More rarely, they may form two separate tendons (White 1943). The aponeurosis of the gastrocnemius can join the soleal aponeurosis in two ways. In type 1, the most frequent, the two aponeuroses join 12 cm proximal to their calcaneal insertion. In Type 2, the aponeurosis of the gastrocnemius is ill-defined, and its fibres insert directly into the aponeurosis of the soleus (O'Brien 1984). The Achilles tendon is enveloped by a paratenon along its whole length (Last 1984). The paratenon can itself be a cause of disabling pathology (Kvist and Kvist 1980).

The Achilles tendon is the tendon in the human body which most frequently undergoes subcutaneous tear (Jozsa et al 1989b). Its tensile strength is in the order of 50-100 Newton per square mm (Viidik 1969), and tensoceptors protect the tendon from excessive stresses (Barfred 1971a). If the tendon is lengthened more than 3-4% of its normal length, it starts to disrupt (Williams 1985), and fails at 8% of its physiological length. Microscopic interruptions in the tendinous substance occur during physiological activity, but fibres remodel, and new collagen is continuously formed (Parry et al 1978).

The incidence of Achilles tendon rupture is increasing (Leppilahti et al 1996; Nillius et al 1976), with a peak in the fourth decade (Leppilahti et al 1996). Males are affected significantly more often than females (Leppilahti et al 1996; Nillius et al 1976). Although most ruptures occur during sports activities (Landvater and Renstrom 1992), intrinsic biological and biochemical factors may play a significant role, and an Achilles tendon rupture is significantly more frequent in patients with the 0 blood group (Jozsa et al 1989a). Controlled data are not available yet for the population in our catchment area, but the surgeons in our hospital perceive an increase in the absolute number of patients with the condition. In the past five years, we have operated, on average, on about one patient every ten to fourteen days, with a peak in spring and summer, and a drop in the early winter months. However, a true incidence rate of rupture of the Achilles tendon cannot be given, as prospective studies on the exposure to causative factors have not been performed.

The etiopathogenesis of Achilles tendon rupture is unknown. The tensile strength of healthy tendons is much greater than the force that their corresponding muscles are able to generate (McMaster 1933). It has been hypothesised that the Achilles tendon ruptures because chronic, subclinical degeneration, possibly related to ageing (Strocchi et al 1991), may have altered its biomechanical characteristics (Kannus et al 1991). A relative hypovascularity in the area where the Achilles tendon most frequently ruptures (2 to 6 cm proximal to its calcaneal insertion) (Langergren and Lindholm 1958/1959) may couple with degenerative intra-tendinous changes (Arner et al 1958/1959) to produce a tendon less able to withstand physiological forces. Also, eccentric loads may exceed the tensile strength of the tendon (Viidik 1969). Therefore, changes in the Achilles tendon coupled with muscular forces exceeding its tensile strength seem to contribute to the rupture (Waterston and Maffulli 1997).

Recently, we showed a statistically significant association between Achilles tendon rupture and sciatica in 102 patients who underwent repair of an Achilles tendon rupture, and in a group of 128 individuals nominated by the patients, matched for age, sex and occupation, previous back pain, and previous surgery for back pain. However, 35 of the 102 patients had experienced sciatic pain prior to Achilles tendon rupture. Pain of a similar nature had been experienced by only five individuals in the control group (3.5%) (p < 0.001). The association may be due to similar nutritional vascular impairment of the intervertebral disc and the Achilles tendon, or to abnormal collagen present in both structures. Alternatively, Achilles tendon rupture may be predisposed by previous sciatica when there is impairment of the afferent signals from tendons, muscle and joint structures carried by the sciatic nerve (Irwin et al 1994).

Pathological Anatomy
Although Williams (1986) believed that normal tendons do rupture, a tendon with areas of degeneration seems to be a relatively common finding when histochemistry is performed following repair of a torn tendon. For example, of 14 ruptured Achilles tendons, only 5 showed areas of degeneration (Kvist and Jarvinen 1982). In the study by Jacobs et al (1978), more than 50% of the Achilles tendons examined showed signs of degeneration. Fox et al (1975) found that, of the 22 patients with Achilles tendinopathy, 10 subsequently suffered a complete tear. Intratendinous degeneration may be a feature of aging, and more than * of patients operated for an Achilles tendon rupture within 48 hours from the injury showed a degenerative tendinopathy (Jozsa et al 1989b).

At operation, oedematous areas and disruption of the tendinous fascicles are found, with extravasation of plasma protein at the rupture site (Jozsa et al 1989c). Areas of normal tissue may alternate with degeneration and haemorrhage. Degeneration can be variable (Arner et al 1958/1959; Kvist and Jarvinen 1982), or totally absent (Barfred 1971b).

Arner et al (1958/1959) identified four stages in the reparative process of unoperated ruptured Achilles tendons:

  1. within 24 hours from the tear, there is some inflammation, without granulation tissue
  2. one to four days from the tear, diffuse leukocyte infiltration is seen
  3. four days to two weeks from the tear, there is abundant granulation tissue, with a less pronounced inflammatory reaction
  4. after two weeks from the tear, there is neoformation of connective tissue, at various phases of differentiation according to the time from the rupture.

At electron microscopy, three phases have been identified (Enwemeka 1989):

  1. an initial five day period of inflammation
  2. a period of fibrillogenesis
  3. a period of progressive realignment and reorganisation of the newly formed collagen fibrils in fascicles oriented along the longitudinal axis of the tendon.

Physical Examination
For a non specialist, the diagnosis of a torn Achilles tendon can be difficult, and patients often reach orthopaedic surgeons at different post-injury stages, after initial misdiagnosis (Maffulli 1996). When examined soon after the injury, a gap, indicative of a tear in the substance of the tendon, can be seen and palpated. With increased time after the tear, the gap can be obliterated by oedema, and palpation becomes unreliable. Also, in the early stages oedema and bruising may not be apparent, and plantar flexion and even standing on tiptoes can still be possible through the action of the tibialis posterior and the peronei muscles (Maffulli 1996).

Numerous clinical tests have been described to aid diagnosis (Copeland 1990; Matles 1975; O'Brien 1984; Simmonds 1957; Thompson 1962; Thompson 1962). My physical examination routine includes history taking, medication history, general physical examination, and local physical examination, with special reference to the conditions of the skin around the ankle. I enquire whether a patient is taking steroids or quinolones (Huston 1995), as these may cause the rupture of another tendon. With the patient prone and his/her feet hanging from the end of an examination couch, I gently palpate the course of the tendon to locate a possible gap. I then perform the calf squeeze and the Matles' tests. I examine patients on admission or in the immediate pre-operative period, with the patient awake, and repeat the examination on the operating table, with the patients under general, spinal, or local anaesthesia, with the tourniquet (if used) not inflated. I always test the contralateral uninjured tendon.

The calf squeeze test, originally described by Simmonds (1957), and subsequently by Thompson and Doherty (1962) and again by Thompson (1962), is probably the best known clinical test to identify a subcutaneous tear of the Achilles tendon. The examiner gently squeezes the patient's calf muscles with the palm of his/her hand. If the Achilles tendon is intact, the ankle plantarflexes. If the Achilles tendon is torn, the ankle remains still, or only minimal plantarflexion occurs. In the Matles test, patients are asked to actively flex their knees to 90°. When the patients are anaesthetised, the examiner passively flexes both the patients' knees to 90°. In both instances, the position of the ankles and feet is observed during flexion of the knee. If the foot on the affected side falls into neutral or in dorsiflexion, an Achilles tendon tear is diagnosed. On the uninjured side, the foot remains in slight plantar flexion when the knee is flexed to 90° (Matles 1975). I have used the Copeland test (Copeland 1990) and the O'Brien needle test (O'Brien 1984) for research purposes. These two tests can be uncomfortable for the patients, and I do not use them routinely. For clinical management, if at least two of the tests described above indicate a rupture, I diagnose an Achilles tendon rupture, and manage my patients according to the criteria described hereinafter.

Radiography (Arner, Lindholm Orell 1958/1959), ultrasonography (Maffulli et al 1989), magnetic resonance imaging (MRI) (Keene et al 1989), and other imaging modalities have all been used to diagnose an Achilles tendon tears. However, radiography can be of little value, ultrasonography is open to subjective interpretation, and MRI is expensive, not always readily available for such patients, and can be badly tolerated by some individuals. For research purposes, I have used high-resolution real time ultrasonography (Maffulli et al 1989) and MRI (Keene et al 1989). However, my diagnosis is clinical, not based on images. In the rare doubtful cases, I prefer my patients to undergo ultrasound scanning as it is cheap, fast, readily available, and, in my hands, extremely reliable.


Supervised neglect
Every year, I see two or three patients above 70 years of age who are noticed by their general practitioner to have a strange gait. On questioning, they generally report that, between three weeks and six months beforehand, they stumbled but did not fall, experienced pain and felt "something going" in their calf. They noticed a swelling and some bruising in the postero-medial aspect of their lower leg, accompanied by a decrease in strength of plantar flexion and a strange gait. However, they put up with these symptoms, and went on with their life. After a full explanation of the condition, and of the management options, they normally refuse surgery and/or a cast. I normally reassure these patients, managing them by physiotherapy alone, if they indeed want anything at all. I review them at six weeks to three-month intervals, and tell them that, should any problems arise, they may need a more complex operation. Since my first seeing such a patient, some five years ago, none of them ever required an operation.

Closed management
I reserve closed management to patients who are too old, or too unfit, or simply do not wish to undergo surgery, and they cannot managed by supervised neglect because they are still working, or, even though they are old, are engaged in leisure activities such as gardening. These patients are told of the risks of re-rupture, and of the significant loss in strength and power of the calf muscle resulting from such management regimen (Haggmark et al 1986). The patient receives a below knee synthetic cast with the ankle in gravity equinus. The cast is changed at two and four weeks, and at each change of cast the ankle is more dorsiflexed. By the change of cast at the fourth week, the foot is plantigrade. Patients are given crutches, but are encouraged to bear weight as tolerated, and are fully weight bearing by the fourth week. The cast is removed after a total of six weeks. At this stage, I add a 1.5 cm felt at the heel. Some patients prefer to have the heel of the shoe of the injured side built up. The felt (or the heel) is reduced 0.5 cm every second week, and patients are encouraged to mobilise as able during this period.

Percutaneous repair
Some patients do not want an open procedure, possibly for cosmetic reasons. Their sporting requirements are minimal, and accept the higher risk of re-rupture using this technique, as compared with an open repair (Bradley and Tibone 1990). Nevertheless, they wish to make sure that the ruptured Achilles tendon ends are re-approximated, and that they remain so during the period of immobilization. In these instances, I perform a percutaneous repair according to the technique described by Ma and Griffith (1978), using strong absorbable suture material. Most of these patients receive local infiltration anaesthesia with 15 ml of 1% lignocaine injected in the sites of the stab wounds, but some prefer general or regional anaesthesia. I do not use skin sutures or steristrips, and the stab wounds are just covered with gauzes (Maffulli et al 1991). The post-operative management in cast follows the protocol described above, with plaster changes at two and four weeks after the repair. Patients are discharged on the day of operation or on the following day, after assessment by a physiotherapist, who teaches them how to use crutches, and how to perform isometric exercises of the calf muscles in the cast. The cast is removed six weeks after the percutaneous repair. A 1.5 cm felt at the heel or a 1.5 cm heel raise at the shoe is added, and reduced 0.5 cm every second week. Patients are encouraged to mobilise as able during this period.

Open repair
I prefer to manage my young, fit, previously healthy, sporting patients operatively. After physical examination according to the lines described above, patients have a full explanation of the procedure, and of its risks and possible side effects, stressing the possibility of wound breakdown (Cetti et al 1993). They sign an informed consent form, and are operated as soon as possible thereafter, generally on the next available trauma list. I avoid operating on these patients in the middle of the night, when, in the best of cases, only reduced resources are available. Patients are assessed by an anaesthetist with a special interest in trauma surgery, and are given either a general or a regional anaesthetic. After induction, a thigh tourniquet is applied but is inflated only if a non-controllable haemorrhage develops (Maffulli et al 1993). In my hands, no patient has required intra-operative inflation of the tourniquet. The patient is positioned prone with both feet dangling from the end of the operating table. The table is tilted approximately 20° cranially, so as to elevate the feet, and decrease the amount of pooled blood in the lower extremities. If a modern fracture table is available, I abduct the hips, and separate the legs, for better access to the side to operate on. The foot, ankle and lower leg to be operated on are then prepared with standard antiseptic solutions. I take extreme care to clean the interdigital spaces, and do not use a glove to cover the toes, as it tends to get in the way, and is easily dislodged with the necessary manipulation of the foot and ankle. The affected limb is then draped in a standard fashion, and I use a marking pen to draw the proposed incision on the skin.

My standard approach is a through an 8 to 10 cm skin incision placed just medial to the medial border of the tendon. After the skin incision, I do not undermine the skin edges, and go through the subcutaneous fat by sharp dissection. The paratenon is visualised. It is generally oedematous and injected with blood, and I incise it longitudinally in the midline for the length of the skin incision. If the paratenon is viable, I suture it with temporary sutures to the subcutaneous fat so that it does not interfere with the tendon itself. The tendon is thus exposed. In fresh ruptures, it presents with the typical horse tail appearance. I do not freshen the stumps in tears operated on within 48-72 hours of the injury, but I am prepared to do so in patients presenting at later stages. The tendon ends are juxtaposed, and I use a single Kessler suture with strong reabsorbable suture material, taking a good bite in both ends of the tendon. If the tendon ends are very frayed, I use a running circumferential suture with finer reabsorbable suture material. Generally, the repair is significantly thicker than the original tendon, and the paratenon cannot be sutured above it. If this is the case, I just cut the temporary sutures anchoring the paratenon to the subcutaneous fat, and leave it be. I use interrupted fine reabsorbable suture material sutures for the subcutaneous fat, and close the skin edges, after juxtaposing them, with steristrips, forming a "net" over the wound extending medially and laterally from it for at least 2 cm.

The wound is covered with routine dressing, and a below knee plaster-of-Paris cast is applied. Although until the recent past I used to apply the below knee plaster-of-Paris cast with the ankle in gravity equinus, I now apply the cast with the foot plantigrade (Rantanen et al 1993). My clinical impression is that it does not make any difference, and it reduces the number of cast changes required (see below).

Post-operative care
Normally, patients are discharged the same day or the day after the operation, after having been instructed to use crutches by an orthopaedic physiotherapist. Patients are allowed to bear weight on the operated leg as tolerated, but are told to keep the affected leg elevated for as long as possible to prevent post-operative swelling. Patients are followed on an outpatients basis at two weeks intervals, and the cast is removed six weeks after the operation. When I used to apply the cast with the ankle in gravity equinus, I changed the cast, putting the ankle in gradually increasing dorsiflexion, up to plantigrade, after two and four weeks, removing the cast altogether after six weeks from the operation. Even in this instance, patients were allowed to bear weight as tolerated.

At the beginning of my career, I used to immobilise the patients in a full above-knee cast for four weeks before reducing the plaster to a below knee cast for a further two weeks (Maffulli et al 1990). Patients were allowed partial weight bearing and gradual stretching and strengthening exercises, increasing to tolerance, only after removal of the cast (Maffulli et al 1990). Gradually, patients proceeded to full weight bearing eight to ten weeks after surgery.

During the period in cast, patients are instructed to perform gentle isometric contractions of the gastro-soleus complex after weight bearing has become comfortable (Maffulli et al 1990). At present, after removal of the cast, patients mobilise the ankle under physiotherapist guidance. Cycling and swimming are started two weeks after removal of the cast, continuing the ankle mobilisation exercises. Patients are prompted to increase the frequency of their self-administered exercise programme, ideally exercising 5 to 10 minutes each hour. Patients progress to a gradually increasing programme, and are normally able to return to their sport on the third or fourth post-operative month.

Post-operative management in athletes
In athletes and well-motivated, reliable patients, I do not apply a full cast. Immediately after the operation, an anterior below knee plaster-of-Paris slab is applied with the ankle in gravity equinus. Patients are discharged the same day or the day after the operation, and are allowed to toe-touch weight bear on the operated limb as tolerated. They are to keep the operated leg elevated for as long as possible, and are seen 48-72 hours after the operation in the plaster room. By this time, the post-operative swelling, if any, has significantly decreased, and the anterior below knee plaster-of-Paris slab is changed to an anterior below knee synthetic slab with the ankle in gravity equinus. The slab is kept in place by an elastic bandage which allows plantar flexion of the ankle, while dorsiflexion is limited by the foot piece of the slab (Carter et al 1992). Patients are allowed weight bearing as able, using crutches. The slab is changed at the second and fourth post-operative week, so that the ankle can dorsiflex to neutral by the fourth post-operative week. The limitation of dorsiflexion is continued for a total of six weeks, when the slab is removed.

In addition to the exercise programme described above, athletes are encouraged to perform isometric exercises at least three times per day, performing three sets of ten 3 seconds contractions with 3 seconds rest between the contractions, and 1 minute rest between the sets. Every two weeks, another 3 seconds are added to each contraction, up to three sets of ten repetitions of 15 seconds contractions. The recovery time within the sets is increased to 10 seconds, while the recovery time between the sets remains constant. Patients are instructed to perform the isometric strength training at three different angles, namely at maximum dorsiflexion, maximum plantar flexion and at a point midway between the two (Maffulli et al 1990). For research purposes, I monitored the progression of the training programme using ultrasound scanning, and according to individual pain (Maffulli et al 1990). If signs of paratenonitis became evident [i.e. hyperechogenicity of the Kager's triangle (Maffulli et al 1987)], and/or pain was manifest either during or after a training and/or physiotherapy session, the patient was asked to abstain from them for three to five days, and to start again from two stages back. However, given costs and manpower limitations, I do not use ultrasound scanning routinely, and base the progression of the rehabilitation programme on the pain experienced by the patients.

As I reserve this management programme to high level, well-motivated athletes, they are compliant, and are normally able to return to their sport six to eight weeks after the removal of the anterior slab. I have tried the regimen recommended by Solveborn and Moberg (1994) in four patients, but the patients found it cumbersome, and prefer the one described above. A hinged orthosis could be used as an alternative to the anterior slab (Mandelbaum et al 1995). However, it is more expensive than simple synthetic cast, although it could be re-usable. As the present regimen is serving me well, I am not planning to change.

Surgical management of older tears
I have always been able to suture the tendon stumps in an end-to-end fashion in all patients who were operated on within 72 hours of the injury (Maffulli 1995). The fact that patients diagnosed after this time are less than 10% of my personal series probably reflects the good referral system that we have developed, and the continuous stress that we put at all levels of medical education to the accurate and prompt diagnosis of the condition. After freshening, when the tendon stumps cannot be approximated without undue tension, I bridge the gap with a single central (Bosworth 1956) or two (one medial, one lateral) gastrocnemius fascial turndown flap(s) (Lindholm 1959). If available, I may use plantaris longus as a reinforcing membrane (Lynn 1966). I am willingly more conservative in the post-operative management of these patients, and I apply a full below knee cast, with the ankle in gravity equinus. Further post-operative management is along the lines described above.

Using the above described principles of management in the 126 patients treated surgically, I have had a relatively low number of complications. Of the 24 patients treated by percutaenous suture, two developed paresthesiae in the territory of innervation of the sural nerve. Neither of them required a re-exploration. A further one, a fireman on active duty, re-ruptured his Achilles tendon three months after removal of the cast. I performed an open repair augmented by two (one medial, one lateral) gastrocnemius fascial turn-down flaps (Lindholm 1959). Revovery was uneventful.

Of the 102 patients treated primarily by open repair, only one developed a superficial wound infection, which was managed by oral antibiotics and elevation. Another patient, a policeman, was so well after three weeks that he removed the below-knee cast by himself. Two weeks later, attempting to play volley ball, the skin in the surgical wound are gave way, creating a raw area of 1.5 by 2 cm. The tendon, however, was in continuity. The patient was kept in hospital with the leg elevated for two weeks, after the area had been skin grafted. The patient remained for two weeks with a front slab preventing dorsiflexion of the ankle above neutral, and the skin graft healed without significant problems. He returned to his normal duties six months after the skin graft.

I operated on six patients on systemic cortico-steroid therapy for a variety of medical conditions. In all of them, the tendon at operation was thinner, anelastic, with a lesser "horse tail" appearance than usual. In all of them I performed a primary non-augmented repair. None suffered from a wound problem, but two had a re-rupture, which was treated by two gastrocnemius fascial turn-down flaps (Lindholm 1959) in one patient, and by using plantaris longus as a reinforcing membrane (Lynn 1966). I plan to reinforce the repair primarily should I need to operate on such patients. Given this experience, I discourage patients with a ruptured Achilles tendon on systemic steroids from having a percutaneous repair. However, I would consider it, as I would consider managing them conservatively, should the conditions of the skin of the ankle be less than ideal.

The future
The field of surgery of Achilles tendon rupture is challenging, and on the increase, given the rising number of adults and middle aged individuals engaged in sports activities. However, surprisingly few basic sciences studies have been published beyond the description of the pathological changes associated with its tearing (Arner, Lindholm and Orell 1958/1960; Kannus and Jozsa 1991).

We have used synthetic adhesives to treat experimental fractures in the rabbit (Capasso et al 1991), and other groups have used similar substances in experimental tendon repair (Bonetti et al 1988; Trail et al 1992). Although I have not used yet adhesives in tendon surgery, I think that the cyanoacrylate derivatives presently on the market offer enough experimental evidence to at least augment the usual reabsorbable sutures used in the open management of Achilles tendon tears. However, any advantage would have to be shown in such a large prospective randomised study that I doubt whether we shall ever have the definitive proof. I suspect that it will come down to the surgeon's personal preference.

My research efforts are now concentrating on rehabilitation, trying to plan more aggressive programmes, especially in young sportspeople, and on the role of growth factors in tendon healing. In an in vitro model, we have shown that basic fibroblast growth factor is able to stimulate tendon healing (Chan et al 1997). However, the clinical application of these findings is still far away.

Gene transfer could improve the management of Achilles tendon tears, particularly when used as vehicles for the targeted delivery of growth factors. It is likely that genes encoding the structural proteins of the matrix could be delivered directly to the tendon, and enhance its healing in a programmed manner (Gerich et al 1996). However, I doubt whether they will be introduced in routine clinical practice in the immediate future.