Abstract
Eighteen of twenty-two patients who were having a tendon transfer to treat rupture of the posterior tibial tendon had evidence of injury to the spring ligament. The injury consisted of a longitudinal tear in the ligament in seven patients, a lax ligament without a gross tear in seven, and a complete rupture of the ligament in four. The ruptured posterior tibial tendon was treated with transfer of the flexor digitorum longus in twenty of the twenty-two patients. A variety of methods were used to repair the ligament. It is essential to determine the status of the spring ligament when patients are managed for rupture of the posterior tibial tendon. Patients who have a torn or lax spring ligament in addition to the ruptured posterior tibial tendon may have more severe abnormalities of the hindfoot than those who have only a ruptured tendon.
Rupture of the posterior tibial tendon, apparently first described13 in 1953, has received increased attention in the literature since the mid-1980's. Recent reports have described the clinical features of the deformity, the pathological anatomy, and the proposed operative procedures to treat the abnormal posterior tibial tendon4,5,9,11,12,15-17. However, the deformity of the hindfoot appears to result from damage to other anatomical structures3 in addition to the tendon.
The mechanical integrity of the medial longitudinal arch depends on the dynamic support of muscles, the static support of ligaments and joint capsules, and the manner in which the tarsal bones interlock. The posterior tibial muscle-tendon unit is the main dynamic stabilizer of the hindfoot. The tendon courses behind the medial malleolus under the flexor retinaculum, beyond which it changes direction acutely to insert on the navicular, the cuneiforms, and the bases of the metatarsal bones18. It is considered a plantar flexor and invertor of the hindfoot. After rupture of the posterior tibial tendon, the ligaments and joint capsules appear to tear, stretch, or rupture under increased stress. This may lead to an acquired flatfoot deformity characterized by valgus alignment of the calcaneus, plantar flexion of the talus, and abduction of the forefoot with a resultant pronation deformity of the foot. Despite the important role of the posterior tibial tendon, it may be the static structures of the arch that provide the most support, especially while the person is standing2. The support that these structures provide to the hindfoot and the arch has recently been evaluated in cadaver specimens3,8,19.
Although several methods have been reported for reconstruction of the posterior tibial tendon, usually with good clinical results, the integrity of the ligaments rarely has been described5,7,10,16. As far as we know, plication of the spring ligament with concomitant reconstruction of the tendon has only been mentioned in a small clinical study involving six patients6. Therefore, in order to evaluate the over-all deformity of the hindfoot in our patients who were having operative reconstruction of a ruptured posterior tibial tendon, we explored and evaluated the spring ligament complex to determine if we could detect gross evidence of tears and to see if it was possible to repair the ligament if it was damaged.
*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.
†Department of Orthopaedic Surgery, University of California at Los Angeles School of Medicine, Box 956902, Los Angeles, California 90095-6902.
Twenty-four patients who had a rupture of the posterior tibial tendon were operated on by the senior one of us (A. C., III) between 1991 and 1993; twenty-two were available for clinical follow-up. There were eighteen women and four men, and the ages ranged from thirty-five to seventy-eight years (average, fifty-eight years) (Table I). The right side was involved in nine patients and the left side, in thirteen. Only seven patients recalled a specific acute event or injury that precipitated the pain; these included a twisting injury while playing tennis, a misstep off a ladder, and a slip off a step. No patient had a history of associated systemic disease. The average time between the onset of the symptoms and the initial visit to our clinic was twenty-seven months (range, three to ninety-six months).
All patients had pain in the hindfoot that was severe enough for them to seek medical attention. Although most of the patients initially had pain on the medial side of the hindfoot, the majority also reported that, as the condition progressed, the pain was predominantly along the lateral side of the hindfoot in the region of the sinus tarsi or between the calcaneus and the distal end of the fibula. Most of the patients also noted that the longitudinal arch of the involved foot was flatter than that of the contralateral foot, although some patients did not realize this until they were questioned about it specifically. At the time of the preoperative clinical examination, the longitudinal arch of the involved foot was lower than that of the contralateral foot in sixteen patients; the two arches were at the same level in the remaining six patients. No patient had clear evidence of unilateral flatfoot with a normal-appearing contralateral foot. All involved feet had a flexible planovalgus deformity.
No patient was able to perform a single-heel-rise test, and each had tenderness as well as mild or moderate swelling along the course of the posterior tibial tendon. When viewed from behind with the heel in neutral or valgus, most of the patients had more toes visible on the involved foot (a positive too-many-toes sign) than is considered normal. However, nine of the twenty-two patients had the same number of toes (usually the third, fourth, and fifth toes) visible on the contralateral foot. We do not consider this sign to be effective for the evaluation of patients who have a rupture of the posterior tibial tendon.
Before they were seen by us or before an operation was performed by the referring surgeon, all of the patients had been managed non-operatively with various means, including anti-inflammatory medication (twenty patients), a cast (ten patients), and an arch support (ten patients). Two patients had taken corticosteroids, from a dose pack, orally, and one had received four injections of corticosteroids into the tendon sheath. Three patients had had a previous operation on the tendon, including a Kidner procedure, a limited tenosynovectomy, and a primary repair of the posterior tibial tendon to the navicular.
All of the procedures were performed with the patient under general, epidural, or spinal anesthesia, and a tourniquet was used around the proximal aspect of the thigh. A rolled towel was placed under the contralateral buttock so that the treated lower extremity was in external rotation, which facilitated exposure of the medial side of the hindfoot. A longitudinal skin incision was made just posterior and distal to the course of the posterior tibial tendon. The tendon sheath was opened in its entirety to explore the length of the tendon (Fig. 1). After the abnormal findings were assessed, the tendon was divided a few centimeters from its insertion into the navicular. The spring ligament could then be inspected from the sustentaculum tali to the medial and planter sides of the navicular.
The grossly abnormal portion of the posterior tibial tendon, which was frequently the entire tendon, was then resected. If it appeared uninvolved by the pathological process, the distal few centimeters of the tendon inserting into the navicular were retained to assist in repair of the spring ligament if it was grossly damaged; otherwise, that portion was also excised. Only two patients had a complete rupture of the posterior tibial tendon; both ruptures were just distal to the medial malleolus. In all but one patient, the tendon appeared grossly abnormal and was scarred and thickened to various degrees. All but two patients had a moderate-to-large amount of synovial fluid within the sheath. Fifteen patients had at least one longitudinal tear within the tendon, and some of the tears measured six to seven centimeters.
The spring ligament was then explored. Eighteen patients had various degrees of abnormal findings in this ligament, which were classified as one of three grades. Grade 1 (seven patients) indicated either a single longitudinal tear within the mid-substance (four patients) or two, three, or four small, partial tears at the insertion of the ligament at the sustentaculum tali (one patient) or the navicular (two patients), but the ligament did not appear stretched or lax. Grade 2 (seven patients) indicated an obviously loose ligament that appeared to be stretched when probed, with or without a visible tear in the mid-substance or at the insertion (Fig. 2). Grade 3 (four patients) indicated a complete rupture of the ligament. Two patients had had a previous operation involving the posterior tibial tendon: a Kidner procedure and an attempted primary repair of the tendon. The spring ligament appeared to be completely intact without evidence of stretching or laxity in four patients, including the patient who had had a previous tenosynovectomy.
In seventeen of the eighteen patients who had damage of the spring ligament, the posterior tibial tendon was reconstructed by transfer of the flexor digitorum longus tendon. The flexor digitorum longus tendon was sharply divided in the midfoot, where it crosses the flexor hallucis longus tendon, and was then rerouted through a vertical drill-hole made in the navicular from the plantar surface to the dorsal surface. In most patients, the flexor digitorum was not removed from its tendon sheath. The drill-hole was made as close to the naviculocuneiform joint as possible. The foot was held in maximum inversion and approximately 30 degrees of plantar flexion while the tendon transfer was sutured in place. In twelve patients the tendon was sewn back on itself with a 2-0 non-absorbable suture, and in one patient the tendon was sutured to the sustentaculum tali with a bone anchor. The transferred tendon was too short in three patients and had to be sewn within the navicular drill-hole with use of 2-0 non-absorbable sutures placed through the bone. In another patient, the transferred tendon was too short to be sewn within the navicular drill-hole so the tendon was sutured directly into the insertion of the posterior tibial tendon, which appeared normal and was not involved by the pathological process of the tendon. These four patients had been managed early in our series.
The remaining patient did not have a reconstruction because the posterior tibial tendon was intact with good excursion and only minimum scarring at the region of the insertion into the navicular. In this patient, the tendon was divided at its insertion, shortened, and reattached to the navicular with sutures through drill-holes; this patient had grade-1 findings in the spring ligament.
All eighteen patients who had involvement of the spring ligament had operative repair of the spring ligament just before suturing of the tendon transfer. The type of treatment depended on the type of abnormal findings. In the seven patients who had grade-1 findings, the tear was repaired with a few interrupted 2-0 non-absorbable sutures. Four of the eleven patients who had grade-2 or 3 findings had repair of the tears and augmentation of the ligament with use of the distal stump of the posterior tibial tendon, which appeared to be uninvolved by the pathological process. One patient had repair of the tears and augmentation with the superficial fibers of the deltoid ligament. This patient was the third in our series, and the fibers of the deltoid ligament in the foot of this patient and in the next two that we explored did not appear to be substantial enough for use in repairs. Therefore, we abandoned use of this tissue for future ligament repairs. Six patients had reconstruction of the ligament with use of one-half of the tendon of the anterior tibial muscle, which was detached proximally (such as in the Cobb procedure7), passed through the drill-hole in the navicular from dorsal to plantar, and sutured to the sustentaculum tali with use of a bone anchor.
Of the four patients in whom the spring ligament was intact, three had a transfer of the flexor digitorum longus tendon to substitute for the function of the posterior tibial tendon. One patient in whom the posterior tibial tendon did not appear grossly abnormal had primary repair of a five-centimeter longitudinal rent with 6-0 nylon sutures.
A below-the-knee non-weight-bearing cast was applied postoperatively with the foot in inversion and plantar flexion. The sutures were removed about two weeks postoperatively, after which another cast was applied with the foot in less inversion. The cast was removed six weeks postoperatively, and weight-bearing was allowed only after the shoe had been modified with use of a custom longitudinal arch support and a quarter-inch (0.64-centimeter) medial heel and sole wedge. Physical therapy was then started to regain muscle strength and range of motion.
The duration of follow-up ranged from twenty-four to forty-six months (average, thirty-two months). All of the patients returned to be examined by one of us (A. R. G.). The physical examination included an assessment of the longitudinal arch, the alignment of the heel, and the ability to perform a single-heel-rise test. The patient subjectively assessed the operative result with use of a modification of the clinical rating system developed by the American Orthopaedic Foot and Ankle Society14. The result was considered excellent if the patient was pain-free, able to walk more than six blocks, able to perform a single heel-rise, and pleased with the over-all improvement. The result was fair if the patient had mild pain or was not able to walk more than six blocks, and the result was poor if the patient had moderate pain or was not able to perform a single heel-rise. Anteroposterior and lateral radiographs of the treated foot and ankle were made with the patient bearing weight and were compared with the preoperative radiographs. The lateral talocalcaneal angle and the lateral talus-first metatarsal angle were measured. Evidence of osteoarthrosis in the joints was also recorded. Talar tilt was measured on the anteroposterior radiograph of the ankle by measuring the angle made by drawing a line parallel to the articular surface of the dome of the talus and a line drawn parallel to the articular surface of the distal end of the tibia. This was done to determine if malalignment of the ankle may have occurred secondary to chronic laxity of the ankle ligaments.
Patients Who Had Injury of the Spring Ligament
Fourteen of the eighteen patients had an excellent result, all with dramatic relief of pain (Table I). Eight did not have any limitations of daily or recreational activities. Six reported minor limitations of recreational activities, although this represented a marked improvement from the preoperative status. All patients were able to perform a single heel-rise, including four who could do so with active varus angulation of the heel. Eleven patients were able to walk an unlimited distance; the other three reported problems unrelated to the foot, including claudication related to the calf, pain in the contralateral foot, and shortness of breath, that limited the distance that they were able to walk. Over-all, these patients were completely satisfied and reported a marked improvement in the quality of life.
Two patients had a fair result; the major symptom was mild pain in the medial aspect of the hindfoot on exertion with a resultant inability to walk more than six blocks. However, these patients had no limitation of daily activities and were able to perform a single heel-rise easily.
Two patients had a poor result. Although one of the two had complete relief of the preoperative pain, reflex sympathetic dystrophy developed postoperatively, which limited the level of activity and the ability to do a single heel-rise. A series of sympathetic blocks gave only slight relief. This patient also had mild osteoarthrosis of the talonavicular joint, which did not appear to contribute to the pain. The second patient had pain in the medial aspect of the hindfoot that limited both recreational and daily activities. Despite good strength of the transferred tendon on examination, the patient was not able to perform a single heel-rise. Radiographs showed evidence of osteoarthrosis of the talonavicular and subtalar joints. This patient was the only one whom we thought would benefit from an arthrodesis of the osteoarthrotic joints, but she declined this option. At the time of the latest follow-up examination, she still worked as a sign artist and had adequate relief of the pain with non-steroidal anti-inflammatory medication.
Five patients who had grade-1 abnormal changes in the spring ligament had an excellent result, one had a fair result, and one had a poor result. Of those who had grade-2 changes, five had an excellent result, one had a fair result, and one had a poor result. All four patients who had grade-3 changes had an excellent result.
Fifteen patients, including nine who had an excellent result, needed comfort footwear or shoe inserts. Five patients were able to wear fashionable shoes, usually on a limited basis. Only five patients could walk without difficulty on any terrain, while the remainder, including nine who had an excellent result, had some difficulty on uneven terrain or inclines.
On examination, all patients had a low arch and valgus alignment (range, 2 to 12 degrees; average, 9 degrees) of the heel. Sixteen patients were able to perform a single heel-rise, although only four of them had varus alignment of the heel while performing the test.
Patients Who Had a Normal Spring Ligament
Two patients had an excellent result, one had a fair result, and one had a poor result (Table I). Both patients who had an excellent result had complete relief of pain, had no limitations of activities or walking distance, and were able to perform a single heel-rise. The patient who had a fair result had intermittent pain in the medial aspect of the hindfoot, most commonly on exertion. Interestingly, this patient had had the classic signs and symptoms of rupture of the posterior tibial tendon preoperatively, but exploration revealed an intact spring ligament and a minimally affected tendon with a five-centimeter longitudinal rent and copious synovial fluid. Because the integrity of the tendon was well preserved and it did not appear grossly abnormal, the rent was closed primarily with 6-0 nylon sutures and a tendon transfer was not done. Despite continued mild pain thirty-two months after the operation, the patient reported excellent subjective improvement and was very satisfied with the result. The patient who had a poor result had moderate pain, on a daily basis, over the lateral part of the hindfoot, which limited recreational activities but not walking distance. The patient was, however, unable to perform a single heel-rise. Radiographs showed mild osteoarthrosis of the talonavicular and subtalar joints. However, the patient did not think that the symptoms warranted additional operative treatment.
Complications
There were three complications, including a postoperative wound infection, which healed uneventfully with a ten-day course of orally administered antibiotics; numbness of the skin over the medial and plantar surfaces of the great toe, which occasionally limited recreational activities; and reflex sympathetic dystrophy, as noted.
Radiographic Results
The lateral talus-first metatarsal angle averaged 13 degrees (range, -5 to 34 degrees) preoperatively and 11 degrees (range, -5 to 32 degrees) postoperatively, as determined on the radiographs made with the patient bearing weight (Fig. 3). For the patients who had an excellent or fair result, the angle averaged 13 degrees (range, -5 to 34 degrees) preoperatively and 9 degrees (range, -5 to 30 degrees) postoperatively. For the three patients who had a poor result, the angle averaged 13 degrees (range, 12 to 15 degrees) preoperatively and 20 degrees (range, 12 to 32 degrees) postoperatively. However, only one of these three patients had an increase in the angle, from 12 degrees preoperatively to 32 degrees postoperatively.
The lateral talocalcaneal angle averaged 42 degrees (range, 28 to 60 degrees) preoperatively and 44 degrees (range, 30 to 64 degrees) postoperatively. For the patients who had an excellent or fair result, the angle averaged 41 degrees (range, 28 to 60 degrees) preoperatively and 42 degrees (range, 30 to 55 degrees) postoperatively. For the three patients who had a poor result, the angle averaged 51 degrees (range, 50 to 52 degrees) preoperatively and 55 degrees (range, 50 to 64 degrees) postoperatively.
Only eight patients had evidence of osteoarthrosis, which was detectable in the talonavicular joint or the subtalar joint, or both (one patient). These changes were seen in all three patients who had a poor result. Talar tilt, measured on the anteroposterior radiographs of the ankle made with the patient bearing weight, was seen in only two patients; one patient, who had an excellent result, had a tilt of 2 degrees and one, who had a poor result, had a tilt of 3 degrees.
In normal gait, the posterior tibial muscle produces inversion of the hindfoot. This locks the transverse tarsal joints, thereby allowing the powerful gastrocnemius-soleus complex to act through the foot at the metatarsal heads1. When the posterior tibial tendon is ruptured, stabilization of this normal inversion of the hindfoot is lost and the gastrocnemius-soleus complex acts at the talonavicular joint. This may be responsible for increased stress on the head of the talus resulting in damage to the spring ligament. The term spring ligament is misleading because this structure has recently been reported to represent two separate ligaments with different osseous insertions, strengths, and histological characteristics3. Studies of cadavera have indicated that the larger and stronger portion is the superomedial calcaneonavicular ligament, which originates from the superomedial aspect of the sustentaculum tali and the anterior facet of the calcaneus and inserts broadly on the edge of the navicular facet. The other portion, the inferior calcaneonavicular ligament, courses from the anterior aspect of the sustentaculum tali to the inferior surface of the middle of the navicular3. During the period when our patients were operated on, anatomical studies were in progress; these studies, which have recently been published, describe the two portions of the spring ligament3,4. However, in retrospect, it appears that almost all of the tears that we saw and sutured probably were within the superomedial calcaneonavicular ligament. Moreover, it is difficult to find or free a sufficient amount of normally appearing structures to perform a reconstruction in a patient who has a loose, torn, or completely ruptured ligament. The superficial deltoid ligament has also been described as inserting into the superomedial calcaneonavicular ligament, and this combination of ligaments has been referred to as the spring ligament complex3. We found it difficult to evaluate the superficial deltoid ligament at the time of the operations. The ligament is thin and broad and, although we used it to repair the spring ligament in one patient, it did not appear to be of sufficient strength or length to be useful. As all but two of our patients had no radiographic evidence of talar tilt, we assume that these fibers were intact.
Our patients reported a gradual onset of the discomfort, and, in most, the symptoms and signs were vague. Moreover, we found it difficult to measure precisely the arch and the other deformities that were present. The single-heel-rise test appeared to be the best way to evaluate the patient preoperatively, as almost all patients had moderate or severe pain on attempting to perform this test and were unable to proceed. Postoperatively, the results of the single-heel-rise test varied considerably. While almost all patients could perform a single heel-rise postoperatively without pain, not all of them could also bring the heel into varus as is usually seen with a normal foot. However, since manual testing is usually inaccurate because of an intact anterior tibial tendon that holds the foot inverted, we found the single-heel-rise test to be the most useful clinical test for this group of patients.
While it appears that all but two of our patients had a satisfactory clinical result of the tendon transfer and ligament repair, we observed that most of the patients had some residual deformity of the hindfoot. Almost all of our patients had positive physical findings, such as a low arch or increased valgus alignment of the contralateral hindfoot, which was usually asymptomatic. Thus, it may be that these patients, like many of our patients who have a ruptured posterior tibial tendon, probably did not have a normal hindfoot even before there was any evidence of clinical abnormality. The term dysfunction, rather than rupture, of the tendon was initially recommended because very few patients appeared to have sustained an acute rupture of the posterior tibial tendon12. Rather, the pathological process seemed to have been gradual and chronic, unlike an acute rupture of the Achilles tendon. However, since all of our patients had no function of the posterior tibial muscle preoperatively, we considered the tendon to be ruptured despite our finding of a grossly abnormal tendon in continuity at the time of the operation.
It is not possible for us to determine clinically if our attempts to repair the spring ligament will lead to an over-all better clinical result. However, it is important to assess whether there are any pathological changes in the ligament at the time of transfer as treatment for a damaged posterior tibial tendon. We believe that the damaged spring ligament should be repaired, if possible, in order to strengthen this important structure that gives static support to the hindfoot.
Ambagtsheer, J. B.: The function of the muscles of the lower leg in relation to movements of the tarsus. Acta Orthop. Scandinavica, Supplementum 172, 1978.
Basmajian, J. V., and Stecko, G.: The role of muscles in arch support of the foot. An electromyographic study. J. Bone and Joint Surg.,45-A: 1184-1190, Sept. 1963.45-A1184
1963
Davis, W. H.; Sobel, M.; DiCarlo, E. F.; Torzilli, P. A.; Deng, X.; Geppert, M. J.; Patel, M. B.; and Deland, J.: Gross, histological, and microvascular anatomy and biomechanical testing of the spring ligament complex. Foot and Ankle Internat.,17: 95-102, 1996.1795
1996
Deland, J. T.; Arnoczky, S. P.; and Thompson, F. M.: Adult acquired flatfoot deformity at the talonavicular joint: reconstruction of the spring ligament in an in vitro model. Foot and Ankle,13: 327-332, 1992.13327
1992
[PubMed]
Funk, D. A.; Cass, J. R.; and Johnson, K. A.: Acquired adult flat foot secondary to posterior tibial-tendon pathology. J. Bone and Joint Surg.,68-A: 95-102, Jan. 1986.68-A95
1986
Goldner, J. L.; Keats, P. K.; Bassett, F. H. III; and Clippinger, F. W.: Progressive talipes equinovalgus due to trauma or degeneration of the posterior tibial tendon and medial plantar ligaments. Orthop. Clin. North America,5: 39-51, 1974.539
1974
Helal, B.: Cobb repair for tibialis posterior tendon rupture. J. Foot Surg.,29: 349-352, 1990.29349
1990
[PubMed]
Huang, C.-K.; Kitaoka, H. B.; An, K.-N.; and Chao, E. Y. S.: Biomechanical evaluation of longitudinal arch stability. Foot and Ankle,14: 353-357, 1993.14353
1993
[PubMed]
Jahss, M. H.: Spontaneous rupture of the tibialis posterior tendon: clinical findings, tenographic studies, and a new technique of repair. Foot and Ankle,3: 158-166, 1982.3158
1982
[PubMed]
Janis, L. R.; Wagner, J. T.; Kravitz, R. D.; and Greenberg, J. J.: Posterior tibial tendon rupture: classification, modified surgical repair, and retrospective study. J. Foot and Ankle Surg.,32: 2-13, 1993.322
1993
Johnson, K. A.: Tibialis posterior tendon rupture. Clin. Orthop.,177: 140-147, 1983.177140
1983
[PubMed]
Johnson, K. A., and Strom, D. E.: Tibialis posterior tendon dysfunction. Clin. Orthop.,239: 196-206, 1989.239196
1989
[PubMed]
Key, J. A.: Partial rupture of the tendon of the posterior tibial muscle. J. Bone and Joint Surg.,35-A: 1006-1008, Oct. 1953.35-A1006
1953
Kitaoka, H. B.; Alexander, I. J.; Adelaar, R. S.; Nunley, J. A.; Myerson, M. S.; and Sanders, M.: Clinical rating systems for the ankle-hindfoot, midfoot, hallux, and lesser toes. Foot and Ankle Internat.,15: 349-353, 1994.15349
1994
Mann, R. A.: Acquired flatfoot in adults. Clin. Orthop.,181: 46-51, Dec. 1983.18146
1983
[PubMed]
Mann, R. A., and Thompson, F. M.: Rupture of the posterior tibial tendon causing flat foot. Surgical treatment. J. Bone and Joint Surg.,67-A: 556-561, April 1985.67-A556
1985
Myerson, M. S.; Corrigan, J.; Thompson, F.; and Schon, L. C.: Tendon transfer combined with calcaneal osteotomy for treatment of posterior tibial tendon insufficiency: a radiological investigation. Foot and Ankle Internat.,16: 712-718, 1995.16712
1995
Sarrafian, S. K.: Anatomy of the Foot and Ankle: Descriptive, Topographic, Functional, pp. 397-399. Philadelphia, J. B. Lippincott, 1983.
Thordarson, D. B.; Schmotzer, H.; Chon, J.; and Peters, J.: Dynamic support of the human longitudinal arch. A biomechanical evaluation. Clin. Orthop.,316: 165-172, 1995.316165
1995
[PubMed]