While an increasing number of recent reports in the literature have
tended to favor operative treatment of an acute rupture of the Achilles
tendon1-9, the exact type of operative
procedure as well as the postoperative regimen remains controversial.
Many surgeons favor a formal operative approach to secure the best
possible repair with the least chance of rerupture. Concerns about
the soft tissue have led others to plan such a procedure only for
professional or high-level athletes and to perform a percutaneous procedure
for other patients3,6. Ma and
Griffith10 developed such a percutaneous
technique and reported good results in eighteen patients. However,
there appear to be two problems with this approach. First, there
is a potential for sural nerve injury and, second, since there is
no open incision, the quality of the repair cannot be confirmed
visually as the tendon ends are brought into apposition. Our standard
treatment of an acute rupture of the Achilles tendon has long been an
open operative approach followed by cast immobilization for eight
weeks. However, we have noted a disturbing rate of problems with
wound-healing and infection. In addition, some of our patients were
dissatisfied with the unsightly and occasionally painful scars.
We performed a retrospective (unpublished) study of 169 consecutive
patients in whom an Achilles tendon rupture was treated surgically
at one level-I university hospital (Geneva) between 1983 and 1995.
The mean age of the patients was forty-two years (range,
fifteen to ninety-seven years). Although all patients had
the same operation, the operations were performed by a number of
surgeons with different levels of training; however, the less experienced
surgeons were supervised by an attending surgeon. We performed a
longitudinal paramedian incision with a mean length of 14 cm (range,
10 to 21 cm). The mean duration of follow-up was 5.4 years
(range, two to fourteen years), and 120 patients were available
for final follow-up. We noted the following complications: (1) anesthesia
or dysesthesia in the sural nerve distribution in eight patients
(7%); (2) deep infection in five patients (4%);
(3) superficial skin necrosis or superficial infection requiring
a prolonged period of wound care without additional surgery in seventeen
patients (14%); (4) scar sensitivity requiring a modification
of footwear in seven patients (6%); and (5) rerupture in
eight patients (7%), with all of the reruptures occurring
more than twelve weeks postoperatively.
On the basis of many recent reports in the literature1-9, we believe that ruptures of the
Achilles tendon are best treated operatively, and we were attracted
to the method described by Kakiuchi11,
which combines the advantages of open and percutaneous techniques.
In an attempt to improve on Kakiuchi’s technique, we performed
a study on cadavera to develop an instrument and a surgical technique.
We also instituted an early functional rehabilitation program. This
paper presents the results of our cadaveric research study, a description
of the instrument and the surgical technique, and the results of
a prospective multicenter study of the first eighty-seven
patients consecutively treated in this fashion.
Instrument
The main guiding instrument, the Achillon, was designed to be
appropriate for the shape of the Achilles tendon as determined in
our human cadaver study, which demonstrated a mean v-shaped
angle of 8° and a mean cross-sectional area of 81 mm2 at the thinnest part. The instrument,
made of 316L stainless steel, consists of a pair of internal branches
connected to a pair of external branches, with each branch having
a line of apertures at the same level to allow easy and accurate
passage of the sutures through all four branches (Fig. 1). There are
several holes in the instrument to allow the surgeon more freedom
to select the site of suture entry into the tendon. Ideally, sutures
should be placed as far from the area of rupture as possible to
ensure good fixation within undamaged tendon. A micrometric screw
allows the opening of the branches to be varied according to the
tendon morphology. Since we completed the study, the instrument
has been manufactured in rigid polymer as a single-use
device by Newdeal, Lyon, France). We used a straight needle, 1.6
mm in diameter and 12 cm in length, through which a number-1
slowly resorbable suture is threaded.
Cadaver Study
In sixteen fresh cadaver legs, the Achilles tendon was transected,
the guide instrument was inserted, and the sutures were introduced
proximal and distal to the laceration site. Three sutures were placed
percutaneously in the proximal part of the tendon (Fig. 2) and three were
placed in the distal tendon end, for a total of ninety-six
percutaneous introductions. In every case, the needle was passed
through the holes of the instrument without any misses or binding
and the sutures sustained no visible damage from the instrument.
It was found that the paratenon could be closed over the repair
site and there was sufficient room to bury the knots (six per tendon
repair) of the sutures beneath the paratenon. Following the procedure,
we dissected the leg and noted that all of the sutures had passed
through some part of the tendon, with no missed passes (Fig. 3).
Patient Population
From November 1996 through January 1999, all adult patients (ninety-three)
who presented to one of three level-I hospitals (in Geneva, Fribourg,
and Zurich) with an Achilles tendon injury (rupture or laceration)
were evaluated for inclusion in the study. Exclusion criteria consisted
of a chronic rupture of more than three weeks’ duration,
previous local surgery, present steroid use, an open rupture or
laceration of more than six hours’ duration, a complex
open rupture with a soft-tissue defect, and a rupture not occurring
between 2 and 8 cm proximal to the tuberosity of the calcaneus (the
site of >90% of ruptures of the Achilles tendon12). We believe that ruptures occurring >8
cm proximal to the tuberosity (muscular ruptures) can be treated
nonoperatively whereas those occurring <2 cm from the tuberosity
necessitate fixation directly to bone. We also excluded any patient who
was unable to cooperate with our treatment program because of dementia
or psychiatric illness. Six patients did not satisfy our entry criteria:
two had sustained the tendon rupture at six weeks and three months
before presentation, one had severe dementia, and three had had
a renal transplant and were taking high doses of corticosteroids.
We are reporting on eighty-seven consecutive patients (fifty-two
seen in Geneva; eighteen, in Fribourg; and seventeen, in Zurich),
with a male-to-female ratio of nine to one and a left-to-right ratio
of forty-five to forty-two. The average age of the patients was
36.5 years, with a range of 22.5 to eighty-two years. There
were eighty-five closed ruptures and two ruptures with
a simple laceration. Thirty-four patients had professional
occupations, twenty-eight were manual workers, ten were
students, ten were retired, and five were professional athletes.
At the time of the injury, fifty-eight patients were participating
in a pivot-sport activity (fourteen were playing squash;
ten, volleyball; eight, tennis; eight, basketball; seven, badminton;
six, soccer; and five, table tennis). Fifteen patients sustained
a high-energy load on the plantar flexed forefoot following a fall
on the stairs or from a bridge or while riding a bicycle. Twelve
ruptures occurred at the time of a strong push-off while
the person was sprinting. The two patients with a laceration were
cut by broken glass. Prior to the injury, three-quarters
of the patients were performing some type of sporting activity from
one to three or more times per week. Six patients had previous localized
achillodynia, and four had received at least one local injection
of corticosteroids. Three patients had recently been treated (less
than one month prior to the tendon rupture) with a quinolone antibiotic.
None of the patients had associated injuries.
On physical examination, all patients presented with a palpable
gap in the Achilles tendon associated with a positive Thompson test.
The location of the rupture (the palpable gap) was an average of
42 mm (range, 31 to 52 mm) proximal to the calcaneal tuberosity.
Anteroposterior and lateral radiographs were made of each ankle
in order to exclude the possibility of any associated fracture.
Magnetic resonance imaging and ultrasound examinations were not
done. All patients in the study were treated with exactly the same
operative technique, postoperative orthosis, and rehabilitation
program. All procedures were performed or supervised by attending
surgeons who had been trained in this technique by performing it on
a cadaver during a workshop.
The mean interval between the injury and the operation was three
days (range, twelve hours to thirteen days). Seventy-eight patients
underwent the surgery under a general anesthetic, and nine patients
had spinal anesthesia. The mean stay in the hospital following the
operation was one and one-half days (range, zero to five
days). All patients were operated on in the prone position, and
the mean operative time was twenty-seven minutes (range, eighteen
to fifty-two minutes). The tourniquet time averaged twenty-six
minutes (range, sixteen to fifty-two minutes).
Operative Technique
After induction of anesthesia, a tourniquet is placed around the
proximal part of the thigh and the patient is placed prone on the
operating table. After confirmation of the correct side of injury,
both legs are prepared and draped in a standard fashion so that
the tension of the Achilles tendons can be compared intraoperatively.
Plastic drapes are not used. A single dose of a second-generation
cephalosporin antibiotic is administered prophylactically to all
patients thirty minutes prior to the start of the procedure. After
inflation of the tourniquet, the incision is begun just medial to
the gap or soft spot in the tendon (Fig. 4) and is extended 1.5 to 2 cm proximally.
The skin and subcutaneous tissue are gently retracted with hooks,
and the paratenon is identified, carefully opened, and tagged with
stay sutures (Fig. 5). Both stumps of the ruptured tendon
are identified (Fig. 6), and the exact site of rupture
is carefully noted.
The Achillon instrument guide is introduced in the closed position,
under the paratenon, in a proximal direction (Fig. 7). The tendon
stump, held with a small clamp under the instrument, is located
between the two internal branches. As the instrument is introduced,
it is progressively opened while the tendon stump is held firmly
with the clamp. When the guide is fully introduced, its position
is confirmed by palpation. The surgeon should feel the tendon lying
between the two branches of the instrument. Three sutures are now
passed (Figs. 8-A and 8-B), and the end of
each is held with a small clamp to keep the sutures separate from
each other. The instrument is then slowly and carefully withdrawn
while the branches are progressively closed. This maneuver results
in the sutures sliding to a peritendinous position, and thus the
tendon itself is the only tissue held by the sutures (Fig. 9). Traction is
applied to the three suture pairs to ensure that they are firmly
anchored in the tendon, and then they are individually clamped to
avoid confusion. Exactly the same sequence is performed on the distal
stump, with the instrument introduced under the tendon sheath and
pushed until it touches the calcaneus. The ankle is placed in equinus
and all of the sutures are now organized for tightening (Fig. 10), which is
carried out with corresponding pairs, and the tendon reduction is
accomplished under direct visual control (Fig. 11). However,
if the tendon ends are too frayed for the surgeon to clearly establish
the correct length, tendon tension should be compared with that
in the contralateral leg. The tendon sheath and skin are carefully
closed, and no drain is used.
A commercially available below-the-knee leg orthosis
with an ankle hinge is applied and is locked with the ankle in 30° of
plantar flexion prior to moving or waking the patient. Patients
are usually discharged to home on the day of the operation. Low-molecular-weight
heparin is administered subcutaneously for prophylactic anticoagulation
in all patients for three weeks postoperatively.
Rehabilitation Protocol
We instituted an early functional rehabilitation program, carefully
supervised by a physical therapist, that is divided into four distinct
stages. For the first two weeks, patients are allowed partial weight-bearing
(15 to 20 kg) with the ankle orthosis locked in 30° of plantar flexion.
Beginning in the third week, gentle unloaded active motion of the
ankle (flexion-extension without extension of the ankle beyond neutral) is
begun, as are thigh-muscle-strengthening exercises and the use of
a stationary bicycle. The goal is to reach a neutral ankle position
by the end of the third week. After three weeks, full weight-bearing
is allowed with continuous use of the orthosis locked with the ankle
in neutral position. At the end of eight weeks, the orthosis is
no longer used. Patients are instructed to use two crutches during
the first six weeks and one crutch for an additional four weeks.
A more intensive program of ankle motion, stretching, isometric,
and proprioceptive exercises is instituted. Jogging is allowed at
three months, and more demanding sports activity is permitted at
six months.
Outcome Measurement
The clinical outcome of surgical treatment of Achilles tendon ruptures
has been assessed with many different scoring systems2,13,14; thus, comparison of results
is difficult. We decided to use the ankle-hindfoot scale
of the rating system developed by the American Foot and Ankle Society
(AOFAS)15. This scale assigns
50 points to function, 40 points to pain, and 10 points to alignment.
A perfect score of 100 points means that the patient has no pain,
a full range of ankle and hindfoot motion, no ankle or hindfoot
instability, good alignment, the ability to walk more than six blocks
and on any walking surface, no limp, no limitation of either occupational or
recreational activities, and no need of any assistive devices for
walking. As suggested by Kitaoka et al.15,
we examined other factors more specific to repair of an Achilles
tendon rupture—namely, the strength of ankle plantar flexion
with the patient standing on tiptoe, the ability to perform repeated
toe raises and single-limb hopping, and the neurological status
of the foot. For single-limb hopping, patients were asked to hop
as many times as possible until they could not lift the heel off
the floor. This global test allowed us to evaluate concentric and
eccentric muscle function of the lower limb.
In addition, all fifty patients who were followed for a minimum
of twenty-four months postoperatively were evaluated with
isokinetic dynamometry (Biodex System 3; Biodex Medical Systems,
Shirley, New York) of both limbs, after correction for gravity.
The patient sat with the thighs secured, the knee joints extended,
and the ankles in neutral position. The foot-plate was fixed to
the individual’s shoe to prevent additional movement. A
ten-minute warm-up period was performed on a bicycle ergometer
(rather than on the Biodex) at 120°/sec for seven cycles.
Each evaluation was preceded by a four-repetition test at the preset
speed (30°, 60°, and 120°/sec), with the uninvolved extremity
tested first. A period of rest was included after each measurement.
Concentric plantar flexion peak torque (in newton-meters) was registered
with five successive cycles at angular velocities of 30°/sec
and 60°/sec. Endurance was defined as the total work (in
joules) during the test time and was measured with thirty cycles
at 120°/sec. The percent differences in peak torque and
total work were calculated with use of the unaffected side as a
reference. The Student t test for paired and independent samples
was used for the statistical analysis.
The duration of follow-up ranged from eighteen to forty-two months,
with an average of twenty-six months. All patients were
personally examined by one of the treating surgeons, according to
a standardized assessment protocol, at ten days, three weeks, eight
weeks, twelve weeks, six months, and twelve months postoperatively
and again at the last assessment for this study. All sutures were
removed at ten days after the surgery. Four patients were followed
for only eight to twelve months postoperatively, at which time they
moved out of the country. Although the clinical results were excellent
at their last follow-up evaluation, these results were
not included with the final outcomes for this study. Another patient
died from a dissecting thoracic aortic aneurysm at eight months postoperatively.
According to the family, this patient did not have any difficulty
with walking before the time of death. Thus, eighty-two
patients were available for follow-up. Wound-healing was
uneventful in all patients, and there were no superficial or deep
infections, clinical deep venous thrombosis, or pulmonary embolism.
Also, no patient had sensory disturbance about the ankle or foot
or, in particular, in the sural nerve distribution.
Two patients had early failure of the Achilles tendon repair. Both
had been noncompliant with the postoperative regimen and had removed
the orthosis. The repairs failed after one patient fell at two weeks
and the other fell at three weeks after surgery. Both underwent
open surgical repair, which was performed by extending the original
incision. A third patient fell while riding a bicycle at twelve
weeks after surgery, sustained a rerupture, and also underwent an
open repair. All three patients had complete disruption of the original
repair. These three patients were excluded from the analysis of
the functional results of the group as a whole, leaving seventy-nine patients
with a full follow-up evaluation.
The mean AOFAS score at the time of the latest follow-up
of the seventy-nine patients was 96 points, with a range
of 85 to 100 points. All patients returned to their previous work
activities, and all who had been active in sports returned to their same
level of participation. The series included five high-level athletes
who had been members of the Swiss national team for fencing, martial
arts, and soccer, and all five returned to the previous level of
their particular sports activity. All patients succeeded in the
one-minute, unsupported toe-standing test. In
addition, after correction for dominance16,
there was no significant difference in the mean number of single-limb
hops between the injured (135 hops) and normal (144 hops) sides.
Isokinetic Results
The concentric peak torque was performed with the ankle in plantar
flexion at 30°/sec and 60°/sec of angular velocity,
after correction for dominance. There was no significant difference between
the injured and uninvolved sides (Table I). The endurance testing at 120°/sec
also revealed no difference between sides.
Rupture of the Achilles tendon is a common injury among high-level
athletes, recreational sports enthusiasts, or even sedentary individuals.
Much has been written about the Achilles tendon itself, including
its structure17-19, blood supply20-24, and biomechanics25-27, and there is an abundant amount
of literature concerning the epidemiology2,28-32 and
etiology13,19,33-35 of Achilles
tendon rupture.
The greatest amount of controversy concerns the treatment of an
acute rupture of the Achilles tendon. The original focus was on
managing these injuries nonoperatively, with plaster cast immobilization
for six to eight weeks. Initially, many investigators were convinced
that the results of nonoperative treatment were equal to those of
surgical repair1,32,36-39. Some
more recent reports on nonoperative management with functional bracing,
as opposed to prolonged plaster cast immobilization, have shown
good results9,40,41. The major
factor motivating surgeons to use such a nonoperative approach appears
to be the wish to avoid the wound complications that occur with
an operative repair.
Recent reports in the literature have favored operative treatment
of an acute rupture of the Achilles tendon1-9.
However, the exact type of operative procedure as well as the postoperative
regimen remains controversial. Most reports discuss either open
or percutaneous surgical techniques. Open repair usually requires
a long incision (the mean length was 14 cm in our retrospective
review) and possibly stripping of the paratenon, which should be
avoided if possible as the paratenon provides a valuable blood supply
to the damaged tendon22. Different
techniques of suture repair have been described, and it appears
that most surgeons generally favor an end-to-end technique. Although
others have advocated primary augmentation of the repair42,43, with
some preferring plantaris tendon44,45,
peroneal tendon46, or artificial
tendon implants47-49,
a study by Jessing and Hansen50 did
not show any evidence that such augmentation was superior to a nonaugmented
end-to-end repair.
Formal open procedures have been frequently associated with a
high rate of complications in the literature32,36-38.
These articles pertain to earlier treatment techniques, and the most
commonly reported complications were related to wound-healing, specifically
wound necrosis and infection. Such complications may be secondary
to the longitudinal incision commonly used for the surgical approach,
which has been shown to pass through an area of poor vascularity51. A report by Wills et al.52 mentioned that "the more
recent studies reported lower complication rates than the earlier
studies" but still noted a 15% wound complication
rate overall. In a retrospective study of 314 patients who had undergone
open repair between 1980 and 1991, Winter et al.53 noted
that nine patients had delayed wound-healing, ten had a deep infection
requiring additional operative treatment, and two had a sinus necessitating
débridement and closure. In the report by Cetti et al.1
on open repair in fifty-six of 111 patients, 4% had
deep wound infection, 2% had delayed healing, 10% had
adhesion of scar tissue, and 12% had disturbance of sensation.
Mandelbaum et al.7 reported that
superficial wound infection developed in two of twenty-nine
patients after open repair. In a study of end-to-end repair in twenty-three
consecutive patients, Soldatis et al.5 noted
two instances of delayed wound-healing, which resolved within three
months. To date, our limited open procedure has resulted in a 100% healing
rate, without any delayed wound-healing, skin necrosis, sinus formation,
or superficial or deep infection. Although it has been reported
that a transverse incision may result in fewer wound complications54, we prefer a small longitudinal
incision so that we have the ability to extend the approach proximally
or distally as the pathology necessitates.
Because of problems with wound-healing, a percutaneous approach
has been considered a compromise that avoids the soft-tissue problems
associated with an open repair. Ma and Griffith10 developed
such a technique, with sutures passed through small stab incisions
along the medial and lateral borders of the tendon, and reported
that, in eighteen patients treated with this method, there were
two minor wound complications but no infections or reruptures. However,
other authors have had less favorable results after percutaneous
repair, with major complications involving sural nerve entrapment
by the suture55-60. In
a recent report by Sutherland and Maffulli61,
thirty-one patients who had undergone repair of an acute rupture
through a "modified" percutaneous technique had
a total of five sural nerve injuries (16%), three of which resolved
in six to nine months. One patient underwent exploration, and the
sural nerve was found to be transfixed by a suture. Even an open
repair does not ensure that the sural nerve will escape injury,
as noted in a retrospective review by Winter et al.53. In that study, four sural nerve
injuries occurred in 314 patients treated with a formal open repair
between 1980 and 1991. On the basis of our cadaver study, it was
clear that by pulling the sutures down beneath the paratenon, from
an extracutaneous to a peritendinous position, we could perform this
procedure without entrapping the sural nerve in the suture loop.
To date, none of our patients have had a sural nerve complication.
While theoretically the nerve could be damaged by passage of the
needle with our method, we did not observe this in the cadaver study.
Most studies have shown a higher prevalence of rerupture in patients
treated with a percutaneous procedure. Bradley and Tibone6 reported two reruptures in a group
of twelve patients treated with the percutaneous technique and none
in a group treated with open repair. They recommended percutaneous
repair in recreational athletes and in patients concerned with appearance
and an open repair in all high-caliber athletes "who cannot
afford any chance of rerupture." In the recent study by Sutherland
and Maffulli61, there were two
reruptures, eleven and fifteen months postoperatively, in thirty-one
patients treated with a modified percutaneous technique. In the
study by Aracil et al.55, there
were two reruptures in six patients who had undergone repair with
the original percutaneous technique of Ma and Griffith10. The authors surmised that this "blind
technique" may result in inadequate apposition of the tendon
ends. Two of our three cases of tendon failure were directly related
to noncompliance by the patients, who removed the protective orthosis
and had failure of the repair at two and three weeks. Certainly
no tendon will be sufficiently healed and strong enough to withstand normal
loading without protection during the first six to eight weeks.
The one true case of rerupture occurred in a patient who sustained
a new injury at twelve weeks postoperatively.
There has been a recent interest in avoiding prolonged immobilization
following both nonoperative and operative treatment4,8,9,40,60. The goals, as outlined
by McComis et al.40, are to prevent
the musculoskeletal changes associated with immobilization, to reduce
the time needed for rehabilitation, and to facilitate an early return
to work and preinjury activities. We followed these principles;
all of our patients were allowed early partial weight-bearing
and range-of-motion exercises of the ankle and foot.
In 1995, Kakiuchi11 described
a combined open and percutaneous technique involving only a limited
incision at the site of the rupture and use of a suture guide to
introduce sutures in a percutaneous fashion proximal and distal
to the rupture site. Twenty patients were treated with this new
combined procedure. Compared with a group of fourteen patients who
had undergone a standard open repair, these patients had better
relief of symptoms during everyday activities, better single-limb
hopping, a greater chance to return to sports activity, and a better cosmetic
result. One patient had an increase in ankle dorsiflexion, representing
lengthening of the tendon, which was probably secondary to the passing
of sutures through injured parts of the tendon. There was also one
case of transient impairment of sural nerve function, which may
have been due to repeated piercing of the skin by the needle in
search of the suture guide. While we were intrigued with this new
combined procedure, we were not satisfied with the technique or the
suture guides. In his article, Kakiuchi stated that passing sutures
through the skin, the intact tendon, and the holes in the suture
guide is a "blind part of the procedure" and "requires the
surgeon to direct the needle repeatedly against the suture guide." For
this reason, we developed the new instrument and technique.
A major problem with attempting to compare our results with others
is the lack of a universally accepted scoring system for evaluating
the outcome of Achilles tendon ruptures. Our review of previously
reported series revealed many different outcome parameters, and
specific scoring systems have been developed by some investigators2,8,13,14,62. The most important factor
in the evaluation of the end result is whether the patient returned
to his or her preinjury status, including work, daily activities,
and sports. Most reports present the physical findings at the time
of follow-up, including muscle strength as evaluated with
such tests as standing on tiptoe, single-limb hopping, and repeated
toe raises. A number of authors have measured their results with
isokinetic dynamometry5-7,59.
In our series, all patients returned to their previous employment
and their previous level of sports activity. They had excellent
plantar flexion strength. There was no significant difference between
the injured and uninjured limbs with regard to the mean number of
single-limb hops. In addition, isokinetic dynamometry revealed no
significant difference between the two limbs with regard to concentric
peak torque and endurance. These results appear to be superior to
those in two recent reports in the literature5,6 and
similar to the experience of others7,59.
We performed isokinetic testing only for patients followed for at
least twenty-four months, and perhaps this is a deficiency
of our outcome assessment.
We acknowledge the problems inherent in multicenter trials—specifically,
the number of surgeons involved and their various levels of training
as well as the number of people who collect the data and assess
the patients at the time of follow-up. In our study, three
surgeons, one at each center, were responsible for the study. Each
was equally trained with use of cadaveric specimens, and each was
responsible for teaching the other surgeons at their institution
as well as for assisting during surgery. When the lead surgeon thought
that the other surgeons were skilled with this procedure, the close
assistance was discontinued. In addition, the surgeons who performed the
procedures were not necessarily the ones who evaluated the patients
at the time of follow-up. One particular surgeon in each
center was responsible for the follow-up evaluation of
all patients, and a trained physical therapist performed the isokinetic
dynamometry under the supervision of a physiatrist.
In summary, with our new instrument and limited open technique
we have been able to directly visualize the repair site and provide
precise apposition of the tendon ends. This has allowed us to optimize
the tension in the muscle-tendon unit while limiting the
surgical dissection and hopefully minimizing the disturbance of
the local blood supply. The results in our first eighty-two patients
justify the continued use of this technique.