Tarsal coalition is a common condition that may be associated with a painful, rigid flatfoot deformity5. The suggested treatment for a talocalcaneal coalition for which non-operative treatment has failed include resection11,13,14,17,18, subtalar arthrodesis7,15,16, triple arthrodesis6, and calcaneal osteotomy2. Many authors have considered triple arthrodesis to be the standard treatment, but rates of success ranging from 50 per cent (ten of twenty feet18) to 94 per cent (thirty-one of thirty-three17) have been reported with resection11,13,14,17,18. Takakura et al. reported no poor results after thirty-three resections. Wilde et al. reported that only ten of twenty resections were considered successful.
The extent to which function of the subtalar joint is restored after resection of a coalition is unclear. Cowell and Elener stated that talocalcaneal coalition seldom is treated with resection because secondary degenerative changes that prevent a satisfactory result usually occur before the diagnosis has been made and the weight-bearing function of the talocalcaneal joint may be disturbed6. Other authors have thought that resection consistently improves motion of the hindfoot and reduces the symptoms associated with the coalition14,17.
The purpose of the current study was to evaluate the clinical and radiographic results and the functional results on gait analysis after resection of a coalition of the talocalcaneal joint.
*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 Orthopedics (H. B. K., M. A. W., and W. J. S.) and Biomechanics Laboratory (K.-N. A.), Mayo Clinic and Mayo Foundation, 200 First Street, S.W., Rochester, Minnesota 55905. Please address requests for reprints to Dr. Kitaoka.
Patients
We examined the records of fifty-three patients who had been evaluated for a talocalcaneal coalition at our institution between 1976 and 1992. One patient was excluded from the study because he had had arthrodesis on one side and resection on the other side. Eleven patients (fourteen feet) had had resection of a talocalcaneal coalition, and they formed the basis of this study.
Nine patients were male and two were female. The mean age at the time of the resection was seventeen years (range, thirteen to thirty-two years). Seven patients had a coalition in the right foot; seven, in the left foot; and three, in both feet. All eleven patients (fourteen feet) returned for follow-up examination by one of us (H. B. K.), and all had radiographic and gait analyses.
The indications for the operation were pain in the hindfoot and ankle region, talocalcaneal coalition, limited motion of the subtalar joint, and failure of non-operative treatment. No patient had severe degenerative osteoarthrosis of the hindfoot. The mean duration of the symptoms, for the nine feet for which it was known, was four years (range, 0.5 to nine years). The patients had been managed non-operatively with immobilization, which decreased the symptoms, but the pain had recurred. Other treatments had included the use of arch supports, modified shoes, and injections.
Anteroposterior and lateral radiographs, with the patient standing, as well as oblique radiographs, were made preoperatively. In order to assess the coalition, axial ski-jump radiographs, as described by Harris and Beath8, were made for two feet (two patients); plain tomograms, for three feet (two patients); and computed tomography scans, for seven feet (five patients)9. The patients did not routinely have bone scintigraphy for diagnostic purposes. None of the patients had other, coexisting coalitions; had had a previous operative procedure for treatment of the coalition; or had a positive family history of tarsal coalition.
Operative Procedure
A pneumatic tourniquet was used, and a medial oblique incision was made in the hindfoot. The flexor hallucis longus tendon was retracted plantarward to expose the sustentaculum tali. The anterior and posterior margins of the coalition of the middle facet were identified. The coalition, which was osseous in all patients, was resected with an osteotome and a rongeur. The resection was continued until the subtalar joint surfaces of the talus and the calcaneus were identified.
Motion of the hindfoot was assessed before and after the resection, and the improvement in motion was noted. Fat graft was interposed into the area of the resection in three feet, and a portion of the flexor hallucis longus tendon was used in two; no material was used in nine feet. The incision was closed, the foot was immobilized in a below-the-knee cast, and partial weight-bearing was continued for three weeks. After removal of the cast, range-of-motion exercises were initiated and weight-bearing was increased. Full weight-bearing was allowed at three weeks.
Evaluation of the Results
The range of motion of the ankle and the foot in all eleven patients was measured in the sagittal plane with a goniometer. The range of motion in the inversion-eversion plane was measured while the patient was kneeling on a chair and the ankle was in neutral flexion-extension. The neutral position of the heel with the talonavicular joint in a neutral, or aligned, position (the tibiocalcaneal angle) was recorded. The angular deviations of the calcaneal axis from the neutral position, with the foot placed in maximum inversion and maximum eversion, were measured with the goniometer.
The clinical results were graded with use of a 100-point hindfoot-ankle score, as described previously10. This score included factors such as the level of pain (maximum [best] score, 45 points); any limitation of daily and recreational activities and any use of a support for walking (maximum score, 10 points); maximum walking distance (maximum score, 10 points); any gait abnormality (maximum score, 10 points); any restriction of motion of the hindfoot (maximum score, 10 points); and alignment of the hindfoot (maximum score, 15 points). The results were graded as excellent (90 to 100 points), good (80 to 89 points), fair (70 to 79 points), or poor (less than 70 points).
The initial radiographs were reviewed to assess the extent of the coalition, the presence of degenerative osteoarthrosis in the joints of the hindfoot, and the presence of secondary radiographic signs such as talar beaking. Anteroposterior, lateral, and oblique radiographs of the foot were made to assess the presence of osteoarthrosis of the joints of the hindfoot and the secondary radiographic findings.
Analysis of Gait
The patients walked at a free-walking velocity across a level ten-meter-long walkway with a flush-mounted piezoelectric force-plate (Kistler, Amherst, New York) located at its center. Dynamic floor-reaction force in the vertical, fore-aft, and medial-lateral directions was determined as a percentage of body weight, as described in previous reports1,3,4. Peak values of the vertical component of the resulting force between the foot and the floor (F1, F2, and F3) and the associated time of occurrence of the peaks (T1, T2, and T3) were determined. Peak values of the fore-aft component of the resulting force between the foot and the floor were represented by F4, F5, and F6 and the associated time of occurrence of the peaks, by T4, T5, and T6. The medial-lateral component of the resulting force was represented by F7, F8, and F9 and the associated time of occurrence, by T7, T8, and T9. We compared the results with established normal values from our laboratory and, for the eight patients who had a unilateral coalition, with the values for the uninvolved foot1,3,4.
Optical switches were placed in the walkway to measure the walking velocity and to determine the cadence. Foot-switches were mounted on the shoes at the heel, at the heads of the first and fifth metatarsals, and at the great toe to determine step length and to provide foot-floor contact sequence during walking and thus subdivide the stance-phase gait. Temporal and distance factors—such as stride length, step length, and the percentage of the gait cycle that consisted of the stance phase, double-limb support, and single-limb support on the involved side—were compared with those factors on the uninvolved side and with those in a population of twenty-six normal, asymptomatic subjects1,3,4.
Triaxial electrogoniometers were applied to the lower extremities to monitor motion in the sagittal (extension-flexion), coronal (inversion-eversion), and transverse (abduction-adduction) planes during walking under different ground conditions, in order to assess the patient's ability to walk on a level surface and on a walkway that sloped 10 degrees to the side (both with the involved limb higher and with it lower). Because patients who have a disorder of the subtalar joint often have difficulty walking on an uneven terrain, the side-sloping walkway was designed to stress the subtalar joint into eversion or inversion. Motion was compared with that on the contralateral side and with established normal values. This electrogoniometric system and the rationale for its use have been described previously3.
A PDP 11/34 computer (Digital Equipment, Maynard, Massachusetts) was used to digitize, average, and reconstruct the three-dimensional motions for the different phases of the gait cycle and the characteristics of stride of the ankle and the hindfoot. Total (peak-to-peak) motion in the sagittal, coronal, and transverse planes was determined for the entire gait cycle. The gait cycle was divided into stance phase and swing phase with use of information obtained from the foot-switches. Discrete measurements of peak-to-peak motion at various subphases of the gait cycle were also obtained.
Characteristic patterns of motion were determined during the gait cycle for the ankle and the hindfoot under different ground conditions. Motions of the ankle and the hindfoot while the patient walked on a side slope were analyzed separately to determine the effects of ground conditions on the gait of the involved and the uninvolved extremity.
The patterns were averaged for all cycles of gait for each group of feet under different ground conditions with use of the Fourier series approximation technique4.
The statistical methods included a two-tailed paired t test of differences in various clinical factors of the involved compared with the uninvolved side and of the involved side compared with the data for normal subjects. A one-sample t test was used to compare various gait factors with those for normal subjects. Differences were considered significant when the p value was less than 0.05.
Clinical Results
The mean duration of follow-up was six years (range, two to thirteen years). The over-all result was classified as excellent for five feet (four patients), good for four (three patients), fair for three (two patients), and poor for two (two patients). There was no pain in three feet (two patients), mild pain in seven (six patients [one patient had moderate pain in one foot and mild pain in the other foot]), moderate in three (three patients), and severe in one. Seven patients (nine feet) had no limitation of activity, three (four feet) had limitation of recreational but not daily activity, and one patient (one foot) had limitation of daily activity. All patients could walk more than six blocks, and none had restricted motion of the ankle. There was no restriction of motion of the hindfoot in two feet (two patients), moderate restriction (from 25 to 74 per cent of normal motion) in eight (seven patients), and severe restriction (less than 25 per cent of normal motion) in four (two patients). Nine patients (eleven feet) did not limp, one patient (two feet) had a mild limp, and one patient (one foot) had an obvious limp. The patients who had a poor result had degenerative changes of the hindfoot, including stiffness and painful motion. None of these patients needed additional operative treatment. Complications did not occur in any patient.
The results did not show a consistent relationship with the age of the patient at the time of the operation. The result was excellent or good for five of the feet of the patients who were fifteen years old or less and fair or poor for three. The result was excellent or good for four of the feet of the patients who were more than fifteen years old and fair or poor for two.
There was no association between the clinical result and the duration of the symptoms. Of the six feet that had had the symptoms for four years or more, four had a good or excellent result. Of the five feet that had had the symptoms for less than four years, four had a good or excellent result. For the three remaining feet, the duration of the symptoms was not recorded.
The results also were assessed according to the duration of follow-up. The result was successful (excellent or good) for three of the seven feet that had been followed for five years or more and for six of the seven feet that had been followed for less than five years. This suggests that the longer the patients were followed, the more likely they were to have symptoms related to degenerative changes.
In addition, the results were assessed according to whether there had been any interposition of tissue after resection of the coalition. Of the five feet in which fat or tendon had been interposed, one had a successful result. Of the nine feet in which material had not been interposed, eight had a successful result.
Of the six feet in which motion of the subtalar joint was less than 50 per cent of normal, two had a successful result. Of the eight feet in which motion of the subtalar joint was 50 per cent of normal or more, seven had a successful result.
The range of motion (mean and standard deviation) was 8 ± 3.1 degrees (range, 4 to 13 degrees) of dorsiflexion and 47 ± 11.3 degrees (range, 30 to 66 degrees) of plantar flexion for the involved ankle and 11 ± 4.2 degrees (range, 4 to 15 degrees) of dorsiflexion (p = 0.129) and 54 ± 8.5 degrees (range, 40 to 70 degrees) of plantar flexion for the uninvolved ankle (p = 0.078).
Inversion was 8 ± 4.5 degrees (range, 2 to 18 degrees) for the involved foot and 17 ± 6.2 degrees (range, 9 to 26 degrees) for the uninvolved foot (p = 0.002). Eversion was 6 ± 2.3 degrees (range, 2 to 8 degrees) for the involved foot and 2 ± 4.0 degrees (range, -5 to 6 degrees) for the uninvolved foot (p = 0.017). We are unable to explain the lower mean eversion in the uninvolved feet.
Analysis of Gait
The mean cadence on level ground was 50 ± 2.8 strides per minute, and the mean velocity was 69 ± 7.4 meters per minute; with the numbers available, we could not detect a significant difference compared with the normal values. We were also unable to demonstrate significant differences with regard to the percentages of the gait cycle that consisted of the stance phase, single-limb support, or double-limb support, on level ground, between the involved side and the contralateral, uninvolved extremity or between the involved side and the laboratory values for twenty-six normal, asymptomatic subjects (thirteen male and thirteen female). The percentage of the cycle that consisted of the stance phase on the side slope with the involved foot higher was significantly lower than that for the normal subjects (p = 0.004).
The force-plate data of ground-reaction forces (F1 through F9) and the temporal force factors (T1 through T9) indicated abnormalities after resection compared with the values for the normal subjects. In the analysis of dynamic floor-reaction force components (Figs. 1-A, 1-B, and 1-C), F8 and F9 were significantly different between the patients and the normal subjects (p = 0.05 and 0.003, respectively). This indicated abnormally high medial-lateral shear forces during walking by the patients who had had resection.
The temporal force factor T7 was significantly different between the involved and the uninvolved side (p < 0.05). When the involved feet were compared with the feet of the normal subjects, significant differences were found in dynamic floor-reaction force components T2 (p = 0.028) and T3 (p = 0.013). These differences with regard to T2 and T3 are related to the increased loading rate in patients who have osteoarthrosis; the body center of mass peaks earlier in stance phase because the foot is loaded more rapidly.
Many motions of the hindfoot and the ankle were decreased on the involved side compared with those on the contralateral side or those in a group of ten normal subjects (Table I). The total motion in the sagittal plane on the involved side was significantly decreased compared with that of the normal subjects, both during walking on the level surface (p < 0.0001) and during walking on the side-sloping walkway with the involved foot higher (p = 0.001). In the coronal plane, the total motion on the involved side was significantly decreased compared with that on the uninvolved side during walking on a side slope with the involved foot lower (p = 0.0002), and it was significantly decreased compared with that of the normal subjects during walking on a level surface (p = 0.0003) and on the side slope with the involved foot higher (p = 0.006). The total motion in the transverse plane on the involved side was significantly decreased compared with that of the normal subjects during walking on a side slope with the involved foot higher (p = 0.026). The motion data indicated generalized stiffness of the hindfoot-ankle complex on the involved side, particularly in the coronal and sagittal planes.
Radiographic Analysis
Adequate follow-up radiographs were available for all eleven patients (fourteen feet). Osteoarthrosis was not consistently associated with a poor clinical result. Three feet (three patients) had osteoarthrosis of the subtalar joint, and the result was excellent or good in two of the three. Similarly, talar beaking was not consistently associated with a worse clinical result. Talar beaking was found in six patients (six feet), and the clinical result was excellent or good in five of the six.
There is no uniformity of opinion regarding the efficacy of resection of a talocalcaneal coalition12. In reports with follow-up of less than five years, most patients have had a successful result11,13,14. A review of these reports suggests that patients have improved function after resection and that neither the duration of the symptoms nor the age of the patient at the time of the operation is associated with a worse result. Our study, however, indicates that there are continuing difficulties with function of the hindfoot and the ankle after resection. Patients who were able to walk in an essentially normal fashion during routine physical examination had abnormalities on gait analysis, such as limitations of motion, particularly in the sagittal and coronal planes, and abnormalities in temporal distance factors, temporal force factors, and ground-reaction forces. Stiffness of the hindfoot and the ankle was demonstrated while patients were walking on the side-sloping walkway.
The clinical results in the current study were less favorable than those in previous reports11,13,14,17. Although most of our patients had improvement, few were completely asymptomatic. One explanation for this discrepancy may be the longer follow-up interval in this series (mean, six years) than in any previously reported series11,13,14,17,18. With longer-term follow-up, symptoms and signs of degenerative osteoarthrosis will likely develop in some patients. In our series, the clinical results were worse for the patients who had been followed for five years or more. The lower rate of success may also be a reflection of our more critical method of grading the results, with multiple subjective and objective factors being considered in the determination of the clinical score.
The results in this series differed from those reported by Takakura et al. Of thirty-three feet that had resection of talocalcaneal coalition in their study, none had a poor result, complications, or radiographic evidence of degenerative changes. The patients in the series of Takakura et al. were younger (mean, twelve years) than those in other reports11,13,14,18, and the indication for operative treatment was tarsal coalition in association with tarsal tunnel syndrome. Preoperatively, some of their patients did not have pain in the hindfoot and some had normal motion of the subtalar joint. In our series, all patients had pain in the hindfoot and restricted motion of the subtalar joint, and no patient had evidence of tarsal tunnel syndrome.
Many patients seen during the sixteen-year period of our study were managed successfully without an operation, and we recommend non-operative methods of treatment, including a period of immobilization. Although the resection was successful in most patients, it did not restore normal function.
NOTE: The authors acknowledge the assistance of Thomas Cahalan, P.T., Iowa Orthopedics and Sports Therapy.