 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1996 The Journal of Bone and Joint Surgery, Inc. Pain in the Posterior Aspect of the Ankle in Dancers. Differential Diagnosis and Operative Treatment*   , NEW YORK, N.Y.Investigation performed at St. Luke's-Roosevelt Hospital, New York City
A retrospective review was performed of the results of operative treatment of stenosing tenosynovitis of the flexor hallucis longus tendon or posterior impingement syndrome, or both, in thirty-seven dancers (forty-one operations). The average duration of follow-up was seven years (range, two to thirteen years). The results were assessed with use of a questionnaire for all patients, and a clinical evaluation was performed for twenty-one patients (twenty-two ankles). Twenty-six operations were performed for tendinitis and posterior impingement; nine, for isolated tendinitis; and six, for isolated posterior impingement syndrome. A medial incision was used in thirty-three procedures; a lateral incision, in six; an anterior and a medial incision, in one; and a lateral and a medial incision, in one. Thirty ankles had a good or excellent result; six, a fair result; and four, a poor result. (The result of the second procedure on an ankle that was operated on twice was not included.) The result was good or excellent for twenty-eight of the thirty-four ankles in professional dancers, compared with only two of the six ankles in amateur dancers.
The contorted, non-anatomical maneuvers of modern dance and the rigidly defined positions and steps of ballet necessitate placement of the feet and ankles in positions that are at the extremes of the ranges of motion of the joints and that require excessive demands on strength12,42,44. These functional demands can lead to inflammatory processes that are further aggravated by a strenuous rehearsal and performance schedule14. Chronic injuries are more common than acute injuries in dancers11. The maneuver of relevé (from the foot-flat to the demi-pointe or en pointe position) can predispose to stenosing tendinitis of the flexor hallucis longus tendon (dancer's tendinitis)9-11,45. Chronic inflammation and hypertrophy of the musculotendinous unit can lead to a painful stenosing tenosynovitis, similar to de Quervain disease. When dancer's tendinitis does not respond to non-operative treatment, operative release of the fibro-osseous tunnel may be needed31. The forceful plantar flexion of the ankle during the en pointe position can produce compression between the calcaneus and the posterior aspect of the tibia, leading to posterior impingement or the talar compression syndrome1,11,15,30. Dancer's tendinitis and posterior impingement syndrome are distinct entities, but they often coexist and usually respond to non-operative measures. However, operative treatment when indicated is usually successful in alleviating symptoms and in enabling patients to return to the extreme demands of ballet and modern dance. We reviewed the results of operative treatment of dancer's tendinitis and posterior impingement syndrome.
The performance outcomes of forty-one procedures in thirty-seven dancers were reviewed retrospectively with use of a questionnaire (Table I). Three patients had had a bilateral procedure and one had had two procedures on the same ankle on two separate occasions. The results for twenty-one patients (twenty-two ankles) were also assessed clinically. These thirty-seven dancers were part of a group of fifty-two dancers who had been followed for at least two years. The average duration of follow-up was seven years (range, two to thirteen years). The goal of treatment was to return the dancers to full functional capacity as performers, as we believe that this is the single best criterion for assessment of the efficacy of the operation. We accepted responses by mail and telephone and did not require that the patient have a physical examination to be included in the study, as many patients were performing ballet in other countries.
The average age of the thirty-seven dancers at the time of the index operation was twenty-three years (range, fifteen to thirty-nine years). The average age of the fifteen male dancers was twenty-four years (range, seventeen to thirty-four years), and that of the twenty-two female dancers was twenty-two years (range, fifteen to thirty-nine years). Twenty-two operations were performed on the left ankle and nineteen, on the right. Of the three patients who had bilateral involvement, two (Cases 16 and 17 and Cases 23 and 24) had exploration of both ankles during the same session of anesthesia. Another patient (Cases 30 and 33) had an operation on the same ankle on two separate occasions: first, because of posterior impingement due to an os trigonum and later, because of tendinitis of the flexor hallucis longus. Two patients had additional procedures: an open débridement for an anterior impingement syndrome in one (Case 18) and reconstruction of the lateral ligament complex for recurrent instability, in addition to treatment for posterior impingement and tendinitis, in the other (Case 31). Ballet was the principal form of dance for thirty patients and modern dance or jazz, for seven. Thirty-one patients were professional dancers, and six were amateur dancers. None of the professional dancers had a history of a known musculoskeletal or rheumatological condition or a collagen disorder. Two professional dancers had had a previous procedure involving the ankle. One of them (Case 5) had had curettage of an osteochondritis dissecans lesion of the talus four years before the index operation for posterior impingement syndrome; she had recovered fully from the earlier operation. The other patient (Case 35) had had reconstruction of the lateral ligaments of the ankle for instability three years before the index operation and had returned to professional dancing. The average duration from the onset of the symptoms to the time of the initial examination by the senior one of us (W. G. H.) was sixteen months (range, zero months to eight years). Twenty-six patients could not recall a precipitating event, and eleven recalled a specific injury. Three of the eleven patients had a history of multiple injuries of the same ankle. The symptoms had been present for one or two months before five operations and for at least three months before thirty-three. The onset had been acute before two operations, and the duration of symptoms before one operation was not recorded. A carefully recorded history and a thorough clinical evaluation allow the surgeon to differentiate between tendinitis and posterior impingement syndrome. Pain in the posterior aspect of the ankle when a dancer assumes the demi-pointe or en pointe position suggests a diagnosis of posterior impingement syndrome4,6. Forceful passive plantar flexion of the relaxed foot (the plantar-flexion test) reproduces the pain experienced while dancing, and this pain is the hallmark of posterior impingement syndrome. The diagnosis is confirmed if an injection of a local anesthetic and a steroid through the lateral approach in the region of the posterior process of the calcaneus relieves the pain. Dancer's tendinitis, or stenosing tenosynovitis of the flexor hallucis longus tendon, is characterized by recurrent pain, tenderness, and swelling posterior to the medial malleolus. It is often misdiagnosed as posterior tibial tendinitis. Triggering or crepitus of the tendon may be palpated, and tenderness along the sheath of the tendon when moving the great toe mimics the pain experienced while dancing and confirms the diagnosis. Non-operative treatment of tendinitis and posterior impingement syndrome involves modification of activities; use of non-steroidal anti-inflammatory medications; physical therapy modalities, such as contrast baths, ultrasound, massage, stretching, ice, and muscle-strengthening; and analysis and correction of the patient's dance technique. Injections of steroids are not used for the treatment of tendinitis because of the proximity of the posterior tibial nerve and the difficulty of entering the compartment of the flexor hallucis tendon accurately. Of the thirty-two ankles that had posterior impingement syndrome, thirteen had injection of a local anesthetic and a steroid in the posterolateral aspect. In five other ankles in that group, the injection was administered by other physicians with use of an unknown approach. In all thirty-seven patients, non-operative treatment provided by the senior one of us or by other physicians had failed. Twenty dancers were managed elsewhere before being seen by the senior one of us, and seventeen were managed by the senior one of us only. Most of the dancers had had several courses of non-operative treatment and had danced during periods when they were asymptomatic. Failure to respond to three months of non-operative therapy was an indication for operative treatment. The average duration of non-operative treatment recommended by the senior one of us was five months (range, zero [because some patients had already been managed elsewhere] to twenty-one months). Preoperatively, lateral radiographs were made for all patients. The average number of preoperative visits was three (range, one to twelve). Of the forty-one operations, twenty-six were for a combination of tendinitis and posterior impingement, nine were for isolated tendinitis, and six were for isolated posterior impingement. A tourniquet was used on all patients. General anesthesia was used in thirty-eight operations; local anesthesia, in two; and spinal anesthesia, in one. A posterolateral approach was used in the six procedures for the treatment of isolated posterior impingement, and a posteromedial approach was used in the remaining thirty-five procedures. Four patients had an additional procedure performed simultaneously with the index operation. Two of these patients (Cases 2 and 25) had a tarsal tunnel release in addition to a tendon release and posterior débridement, one patient (Case 31) had reconstruction of the lateral ligament complex in addition to a tendon release and posterior débridement, and one patient (Case 18) had anterior and posterior medial decompression for anterior and posterior impingement syndromes. An additional patient (Cases 30 and 33) had an operation on the same ankle on two separate occasions: a posterior débridement was done with use of a lateral approach and, subsequently, a tendon release was performed with use of a medial approach. She was graded as having a poor result, although she eventually had an excellent result after the second procedure. In retrospect, she probably would have done very well with a single operation that addressed both conditions with use of a medial approach. A questionnaire was used to determine the patient's level of satisfaction with the outcome of the operation. The questionnaire included a 0 to 100-point scale, with 0 points indicating that the patient considered the operation a failure and 100 points, that the patient was completely satisfied. The questionnaire was also used to assess the patient's ankle preference, level of performance, and willingness to have the procedure again if the circumstances were similar. The results were graded as excellent, good, fair, or poor. A result was considered excellent if the patient had made a full return to the pre-injury status, the severity of the symptoms was assigned 1, 2, or 3 points (on a 1 to 10-point scale), and the patient considered the ankle to have improved compared with the pre-injury status. A result was considered good if the patient had an 80 per cent level of satisfaction with the result (as determined on the questionnaire), had returned to the pre-injury status, and had only slight symptoms (with the severity rated as no more than 3 points). A fair result meant that more than 3 points was assigned for the severity of the symptoms with any activity and that the patient needed to take medication for pain and had made a slow return to full activity. A poor result indicated that there had been a major postoperative complication, that a second procedure was needed, or that the patient would be unwilling to have the same procedure again under similar circumstances. A follow-up clinical examination was performed for twenty-one patients (twenty-two ankles). Information was recorded with regard to gait, the cosmetic appearance of the scar, the circumference of the calf, the range of plantar flexion and dorsiflexion with the patient standing, muscle strength, laxity of the ankle, and the results on the plantar-flexion test and the Thomasen test40.
Operative Technique Medial approach: A four-centimeter curvilinear incision is made posterior to the malleolus at the level of the superior border of the calcaneus, following the course of the underlying neurovascular bundle. The neurovascular bundle is retracted posteriorly with a blunt retractor, and the underlying tunnel is identified by motion of the great toe. The tunnel is then released, proximal to distal, to the level of the sustentaculum tali, and débridement or repair is performed as needed (Fig. 1). Nodules on or within the flexor hallucis longus tendon are not usually excised because after the fibro-osseous tunnel has been released the excursion of the tendon is no longer restricted (Fig. 2). Functional hallux rigidus should be evaluated intraoperatively to ensure complete release and free excursion of the tendon. The released tendon is retracted posteriorly with the neurovascular bundle, and the os trigonum on the lateral side of the entrance of the fibro-osseous tunnel is removed. The ankle is examined in full plantar flexion, and any remaining bone or soft-tissue impingement is removed. The ankle should be in a neutral position of flexion and extension during closure so that the skin is properly aligned.
Lateral approach: Because of increased external rotation of the lower extremities of dancers, the patient should be placed in a lateral decubitus position on the operating table. A curvilinear incision is begun at the posterior aspect of the ankle mortise in line with the posterior border of the peroneal tendons, with care being taken to avoid the sural nerve. A capsulotomy is performed with the ankle in slight dorsiflexion, and the lateral tubercle or os trigonum is identified lateral to the tunnel. We do not believe that the fibro-osseous tunnel of the tendon can be released safely from the lateral side. Adequate osseous decompression is assessed by plantar-flexing the foot and palpating for any bone-on-bone impingement. Occasionally, loose bodies, calcaneal protuberances, or portions of the posterior part of the tibia need to be debrided.
Postoperative Treatment
There were no complete ruptures of the tendon; however, there were five longitudinal tears, and in one patient (Case 18) a diagnosis of a probable intratendinous partial rupture was made on the basis of a widened external appearance. Two of the longitudinal tears were treated with débridement and three, with tubularization of the flattened tendon performed with 5-0 non-absorbable running suture. Tenosynovitis of the flexor hallucis longus tendon was present in thirteen ankles; nodules, in two; distal insertion of the muscle fibers of the tendon that encroached on the fibro-osseous tunnel, in seven (Fig. 3); flexor hallucis accessorius26, in five; a ganglion, in three; and a bifid posterior tibial nerve, in one ankle. Early in the series, the flexor digitorum longus tendon was explored and released in five ankles; however, this step was abandoned as the tendon was normal in all five. The tunnel was released in all thirty-five procedures for dancer's tendinitis. No patient had subluxation of the tendon postoperatively.
Thirty-two operations were done for posterior impingement, which was due to several causes; in twenty-four of these procedures the os trigonum was removed and in four an apparent non-union of a fracture was seen. The facets on the superior surface of the calcaneus that articulated with the os trigonum were debrided. The os trigonum ranged from 0.2 to 2.0 centimeters. Other structures that may have contributed to the talar compression syndrome included a posterior slip of the tibial branch of the posterior talofibular ligament in five ankles, a pseudomeniscus syndrome in one, a posterior medial tubercle in two, and loose bodies from undetermined sources in five. The results were graded after forty of the procedures (thirty-seven patients). (The result of the second operation in the patient who had had two operations on the same ankle was not included.) Twenty-three procedures yielded an excellent result; seven, a good result; six, a fair result; and four, a poor result. Twenty-six patients (thirty procedures) reported that the ankle was not painful during non-dance activity, and eleven patients (eleven procedures) said that it was painful; the pain score ranged from 0 to 6 points. Twenty-nine patients (thirty-two procedures) noted no swelling during non-dance activity, and nine patients (nine procedures) reported some swelling; the score for swelling ranged from 0 to 3 points. (One patient noted swelling on one side but not on the other.) Twenty-one patients (twenty-five procedures) did not have stiffness during non-dance activity, and sixteen patients (sixteen procedures) had some stiffness; the score for stiffness ranged from 0 to 7 points. (The patient who had had two procedures on the same ankle had stiffness after the first procedure but no stiffness after the second procedure.) During dance activity, the pain score ranged from 0 to 9 points; the score for swelling, from 0 to 4 points; and the score for stiffness, from 0 to 7 points. The dancers returned to barre exercises at an average of eight weeks (range, two to twenty-four weeks) postoperatively and participated in class at an average of thirteen weeks (range, three to thirty-six weeks). Twenty-nine of the thirty-one professional dancers returned to full performance at an average of twenty-five weeks (range, six to ninety-six weeks). The return to a full dance schedule averaged six months after the combined procedure for posterior impingement and tendinitis, six months after the operation for isolated tendinitis, and five months after the operation for isolated posterior impingement. Only two of the six amateur dancers resumed a full dance schedule postoperatively. The postoperative complications consisted of two wound hematomas, two superficial draining wounds that responded to oral administration of antibiotics, and a minor suture abscess. There were no deep wound infections. One patient had a small area of dysesthesia around the scar. One year after the operation, two patients needed to take non-steroidal medication intermittently for mild pain in the ankle, and one patient noted painless crepitus in the flexor hallucis longus tendon. No patient reported locking or triggering, although fourteen patients had had these symptoms preoperatively. The average level of satisfaction with the result was 88 per cent (range, 0 to 100 per cent). Ten patients showed a preference for the ankle that had not had operative treatment, eleven had no preference, and fourteen showed a preference for the ankle that had had operative treatment. Of the two remaining patients, one (Cases 38 and 39) preferred the preoperative status of one ankle and the postoperative status of the other, and the other (Cases 23 and 24) did not record a preference. Thirty-four dancers stated that they would have the same procedure again if the circumstances were similar. Ideally, a physical examination would have been performed for all of the dancers; however, because they were dispersed around the world, only twenty-one patients (twenty-two ankles) had such an examination. Of the twenty-one patients, all had a normal gait, none had a problem with the cosmetic appearance of the scar, and one had a minor dysesthesia about the scar. In twenty patients, the strength of the peroneal muscles was equivalent to that on the side that had not been operated on; the exceptional patient (Case 9) had slight peroneal weakness and was dissatisfied with the result. Two patients had demonstrable instability of the ankle. One (Case 7), a professional male dancer, had persistent ligamentous laxity that resulted in instability and recurrent posterior impingement because of forward subluxation of the talus. The other patient (Case 18) had mild laxity of the ankle but was completely satisfied with the result and had functional stability. The average dorsiflexion was 30 degrees (range, 13 to 43 degrees) on the side that had had operative treatment and 28 degrees (range, 14 to 43 degrees) on the side that had not had operative treatment. The average plantar flexion was 59 degrees (range, 37 to 80 degrees) on the side that had had operative treatment and 55 degrees (range, 35 to 78 degrees) on the side that had not had operative treatment. The subsequent dance history and the patient's perception of the result were used to grade the operative result. Twenty of the twenty-one dancers who had a physical examination did not have findings suggestive of a persistent posterior impingement syndrome (on the plantar-flexion test) or tendinitis (on the Thomasen test40). Although there were only six amateur dancers, they accounted for a disproportionate number of the fair and poor results (Table II). Four of the amateur dancers discontinued their pursuit of a professional dance career, citing the result of the operation as a reason for their failure to achieve professional status. Twenty-eight of the thirty-four ankles in professional dancers had a good or excellent result. Two of the four professional dancers who had a fair result reported subjective stiffness and had a pain score of 5 points each. Of these four patients, one preferred the ankle that had been operated on, one preferred the ankle that had not been operated on, and two had no preference. One professional dancer (Cases 30 and 33) had an operation through a lateral approach for the treatment of posterior impingement. She later had another operation through a medial approach, for tendinitis. According to our criteria, she was graded as having a poor result. However, she returned to a professional career. Four of the six patients who had a fair result continued to perform professionally for an average of six years (range, two to ten years). Their average level of satisfaction with the result was 90 per cent (range, 70 to 100 per cent).
Tendinitis of the flexor hallucis longus and posterior impingement of the ankle are familiar to the orthopaedic surgeon who treats professional dancers10,11,15,30. Lack of familiarity with these conditions, a low index of suspicion with regard to patients who are not dancers, and the usual resolution of the symptoms after modifications of activity or rest alone probably contribute to the low reported prevalence13. The senior one of us (W. G. H.) operated on seventy-one ankles for either tendinitis or impingement during a fifteen-year period. The positions of en pointe and demi-pointe require the tendon to function through an extreme range of motion while passing through its fibro-osseous tunnel9-11. Frequent, prolonged repetition of dance movements can lead to irritation and swelling of the tendon with resultant crepitus22, and repeated strains can cause palpable nodules to develop on or within the tendon. Occasionally, the tendon enters the tunnel at an oblique angle, predisposing it to irritation. Thomasen described the limited excursion of the tendon caused by crowding of the muscles and the tendon at the entrance to the fibro-osseous tunnel40. Dorsiflexion of the great toe is reduced or absent when the ankle is in maximum dorsiflexion, and this is termed functional hallux rigidus. This finding is not always pathological as it may be present in asymptomatic dancers. Stenosing tenosynovitis has been described at the sesamoid area of the great toe7, but we believe that the posteromedial part of the fibro-osseous tunnel is the most common site of constriction3. We found only two reports of spontaneous complete rupture of the tendon32,41. In one of those reports, the tendon ruptured two centimeters from its insertion onto the proximal phalanx of the great toe32. In the other, the tendon ruptured within the fibro-osseous tunnel in a middle-aged non-athlete41. We found only one longitudinal tear in twenty-eight unselected cadaver ankles from elderly patients; the tear was one centimeter long and was located at the entrance of the fibro-osseous tunnel38. This finding suggests that the degenerative process is infrequent. In the current series, despite the extreme functional demands of ballet, only five tendons were found to have a degenerative tear. Tendinitis is characterized by pain in the posteromedial aspect of the ankle5; tenderness over the tunnel with active or passive motion; painful crepitus of the tendon with active motion, leading to a trigger hallux21,24; a pseudo-hallux rigidus9 or a functional hallux rigidus23,44; and partial laceration35. Injections of steroids into the sheath of the tendon should be avoided because of the risk of intratendinous damage. Injection of an anesthetic to confirm the diagnosis is not needed as the objective signs are usually diagnostic. When a patient has pain in the posteromedial aspect of the ankle, the differential diagnosis includes tendinitis, a posterior deltoid sprain, osteochondritis dissecans of the posteromedial aspect of the talar dome, soleus syndrome, posterior tibial tendinitis, and posteromedial tarsal coalition. When a patient has pain in the posterolateral aspect of the ankle, the differential diagnosis includes posterior impingement syndrome, fracture of a trigonal process (Shepherd fracture37), Achilles tendinitis, peroneal tendinitis or tear, retrocalcaneal bursitis, and pseudomeniscus syndrome38. The posterior impingement syndrome is rare in athletes9,11,15,30; it is most common in female ballet dancers25,34 who place the ankle in extreme equinus to assume the en pointe position. Talar compression or posterior impingement syndrome occasionally includes an enlarged talar process (a Stieda process39); a prominent posterior process of the calcaneus29; an avulsion fracture of the posterior tibiotalar ligament2,18; loose bodies43; and, most commonly, an os trigonum7,20,33,36. The posterior talofibular ligament and the posterior aspect of the capsule can be compressed, leading to symptoms. In our experience, neither the presence of an os trigonum nor its size corresponds to the symptoms, which usually subside without intervention. Acute sprains of the ankle can occasionally cause soft-tissue attachments to the os trigonum to loosen17; the os trigonum8 can then shift its position and become symptomatic. These tissues can be compressed when a dancer assumes the en pointe position. Laxity of the lateral ligaments of the ankle can lead to posterior impingement, and this diagnosis should be ruled out by an appropriate clinical evaluation. Bone-scanning19 or computerized tomography scanning is rarely indicated and should be used only to evaluate unexplained pain in the ankle. We managed a patient who had persistent symptoms despite a release of the tendon. Bone-scanning and computerized tomography revealed an osteoid osteoma of the posterior aspect of the talus. The lesion was excised and the symptoms decreased. This patient did not complete the questionnaire so she was not included in the current study. When non-operative treatment has failed and operative treatment is considered, an accurate diagnosis is essential so that the appropriate operative approach can be selected. The medial approach is more effective because both tendinitis and posterior impingement can be treated through this incision. The lateral approach involves less dissection, but only posterior impingement can be addressed. Iatrogenic injury to the nerve has been reported with use of a lateral approach13. The anatomical variations of nerves, tendons, and muscle bellies seen in the present series emphasize the importance of attention to operative detail27,28. The most frequent abnormality was a distal insertion of the muscle fibers of the flexor hallucis longus within the fibro-osseous tunnel. Both ankles are rarely affected equally; during the postoperative recovery period for one ankle, non-operative therapy can be provided for the ankle that has not been operated on. We therefore do not recommend operations on both ankles simultaneously. We reviewed the results of operative treatment of posterior impingement and dancer's tendinitis on the basis of the patients' assessments of the outcomes. Any patient who had a score of more than 3 points (on a 0 to 10-point scale) for pain, swelling, or stiffness did not receive a rating that was better than fair. The responses of the amateur dancers to the subjective questionnaire differed from the objective findings on physical examination. The operative results in amateur dancers, in our experience, usually do not meet their expectations. Professional ballet dancers who have an operation miss rehearsals, lose opportunities for roles, and carry the stigma associated with the injury. For these reasons, professional dancers usually avoid having an operation. The amateur dancer perhaps has the visible scar that precluded a professional career. The rigorous process of selection to become a professional dancer also produces a disciplined, resilient patient. Because of this, as well as the professional dancer's desire to perform and to be regarded as healthy, heavy reliance on their subjective assessment can skew the result so that it appears more favorable. For these reasons, we developed a very strict grading system that allowed the subjective reporting of multiple symptoms or signs (swelling, stiffness, and pain) to downgrade the result, even if the patient was completely satisfied with the outcome and had had a full return to professional dance.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Technorati What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |