Forty-four patients (forty-five feet) who had osteoarthrosis of the tarsometatarsal joints were managed with arthrodesis between 1976 and 1991. Four patients were lost to follow-up; thus, forty patients (forty-one feet) were included in the study. Twenty-one patients (twenty-two feet) had primary degenerative osteoarthrosis; these patients had an average age of sixty years (range, twenty-seven to seventy-five years) and an average duration of follow-up of seventy months (range, twenty-four to 186 months). Seventeen patients (seventeen feet) had post-traumatic osteoarthrosis; these patients had an average age of forty years (range, twenty-three to sixty-seven years) and were followed for an average of seventy-four months (range, twenty-four to 211 months). Two patients (two feet) had inflammatory arthritis: one was a seventy-year-old woman with rheumatoid arthritis who had been followed for twenty-four months, and the other was a forty-four-year-old man with lupus erythematosus who had been followed for sixty-nine months. Thirty-six patients returned to the office for a clinical and radiographic evaluation; the four patients who were unable to return were interviewed by telephone and arranged for radiographs to be made elsewhere and sent to us.
The patients were evaluated according to subjective and objective criteria both preoperatively and postoperatively. Data were collected by means of an interview in which questions about the ability of the patient to function were emphasized and areas of pain as well as the type of footwear that the patient preferred were identified. A physical examination was done to assess the posture of the foot as well as abnormal findings such as tenderness, evidence of an operative neuroma, malalignment of the metatarsal heads, and the presence of plantar callosities.
The indication for the operation was a severe loss of function due to painful osteoarthrosis, with or without deformity, that had failed to respond to non-operative treatment. (A severe loss of function was defined as the inability of the patient to return to his or her usual occupation or to perform the activities of daily living.) Approximately one-half of the patients were referred to us for operative treatment after an orthosis that had been prescribed elsewhere had failed to relieve the pain or the loss of function, or both, to their satisfaction. The rest of the patients were initially managed non-operatively by us; the devices that we had prescribed consisted of an ankle-foot orthosis (seven feet), a cast (six feet), a foot orthrosis (five feet), and a fracture-brace (two feet). When it was appropriate, we had prescribed modifications of the shoe (such as a stiffened sole or a rocker-bottom sole) or the use of a shoe of a different size in order to accommodate the orthotic device.
In addition, thirteen patients had had at least one previous operation to correct or stabilize the deformity, or both. Open reduction and internal fixation had been attempted for nine patients who had post-traumatic osteoarthrosis; two of these patients had had at least two operations. The other four patients who had had previous unsuccessful attempts to correct and stabilize the deformity had primary degenerative osteoarthrosis.
The index arthrodesis was to include all of the joints that had radiographic evidence of osteoarthrosis or that were considered to be part of the deformity.
Radiographic Evaluation
Anteroposterior, lateral, and oblique radiographs of the foot were made with the patient standing and bearing weight. The talus-first metatarsal angle, the talus-second metatarsal angle, the lateral talus-first metatarsal angle, the degree of coverage of the talar head by the navicular (Fig. 1), and the degree of hallux valgus were determined. A bone scan or a computerized tomography scan was made to help to define the extent of the osteoarthrosis in eight patients for whom the number of involved joints could not be determined on the basis of the radiographs.
Preoperatively, all patients had radiographic evidence of osteoarthrosis of the mid-tarsal or tarsometatarsal joints, or both. Only the extent of the osteoarthrosis and the deformity varied among the patients. The patients who had primary degenerative osteoarthrosis had more osseous proliferation about the tarsometatarsal joints than did those who had post-traumatic osteoarthrosis.
In order to quantify the correction that resulted from the arthrodesis, the preoperative and postoperative radiographs were evaluated with use of a system in which positive and negative values were assigned according to the direction of angulation. Specifically, a positive value was assigned for abduction of the forefoot and for dorsiflexion of the first metatarsal, and a negative value was assigned for adduction of the forefoot and for plantar flexion of the first metatarsal. The amount of correction then was determined by measurement of the actual change that had occurred in each foot. We believe that this method of evaluation provided a more accurate assessment of the degree of correction, as an adduction deformity (a negative value) was corrected in some feet and an abduction deformity (a positive value) was corrected in others. If the numbers simply had been averaged, the positive and negative values would have negated each other.
The postoperative radiographs were also analyzed for evidence of consolidation of the fusion mass, osteoarthrosis in adjacent joints, and failure of the implant.
Operative Technique
Longitudinal incisions were used to approach the mid-tarsal and tarsometatarsal joints. The first metatarsocuneiform joint was approached through an incision made along the medial aspect of the foot just dorsal to the midline; the second and third tarsometatarsal joints, through an incision centered between the two metatarsals; and the fourth and fifth metatarsocuboid joints, through an incision centered between the fourth and fifth metatarsals. The approach to these joints is difficult because many superficial nerves are encountered; these nerves may be quite fragile and may be damaged by being cut or stretched. The dorsalis pedis artery and the superficial portion of the deep peroneal nerve, which pass along the skeletal plane in the region between the first and second metatarsocuneiform joints, must be identified and preserved. The skin flaps were kept at full thickness, and dissection was carried down to the skeletal plane to minimize the possibility of a skin slough. The extent of the arthrodesis was determined by the degree of the osteoarthrosis and the severity of the deformity (Figs. 2-A, 2-B, 2-C through 2-D).
After the involved joints had been identified, all intervening fibrous tissue and articular cartilage was meticulously removed. This allowed the joints to move so that the foot could be manipulated into a plantigrade position. In most patients, a plantigrade foot was achieved. However, in a few patients—usually those who had a severe abduction deformity—resection of some bone was necessary, particularly in the area of the first and second metatarsocuneiform joints. Bone-grafting was performed in eleven patients (eleven feet): three patients (three feet) who had post-traumatic osteoarthrosis, six patients (seven feet) who had primary degenerative osteoarthrosis, and one patient (one foot) who had inflammatory arthritis. The bone grafts were obtained from local osteophytes (three feet), the iliac crest (seven feet), and the medial malleolus (one foot). In all instances, the graft was cut into small pieces and was placed about the site of the arthrodesis. After the bones to be included in the arthrodesis had been carefully scaled with a four-millimeter osteotome, the internal fixation device was inserted. Steinmann pins were used in the first few procedures, but fixation usually was achieved with use of interfragmentary screws in a variety of patterns depending on the specific deformity. Specifically, interfragmentary screws were used in seventeen feet, a plate and screws were used in fifteen feet, staples were used in five feet, and other types of fixation were used in four feet. A medial buttress plate was used in one patient who had a severe abduction deformity, as this made it easier to align the foot.
The wounds were closed meticulously by using a good subcutaneous closure and by providing wide margins for the skin sutures. This method of closure is used because severe swelling may occur; keeping tension off the edges of the skin helps to prevent subsequent breakdown of the wound. A delayed closure was not needed in any of the feet, and no drains were used.
Postoperative Care
After the application of a snug, well padded, below-the-knee compression dressing that incorporates plaster splints, the foot is elevated as much as possible for the first five to seven days. The dressing and the sutures were removed after ten to fourteen days, and a below-the-knee, non-weight-bearing cast was then worn for four weeks. Six weeks after the operation, the cast was removed and radiographs were made. If healing had progressed satisfactorily, the patient wore a below-the-knee weight-bearing cast for another six weeks. This cast was removed twelve weeks after the operation, and if satisfactory union had occurred the patient was permitted to walk as tolerated. If adequate healing had not occurred, then the immobilization was continued until it had.
Satisfaction Index
A so-called satisfaction index, which combined subjective and objective findings, was adopted because twelve patients (one of whom had post-traumatic osteoarthrosis and eleven of whom had primary degenerative osteoarthrosis) had functional results that were not consistent with the radiographic findings. The satisfaction index placed greater emphasis on the function of the patient than on the radiographic findings because improved function and diminished pain were the main goals of treatment (Table I).
The satisfaction index ranged from 0 (the least satisfactory result) to 5 (the most satisfactory result). A satisfaction index of 4 or 5 indicated that the patient had no pain and no impairment in the activities of daily living; an index of 4, that the patient had no pain but could not resume full pre-injury activities because of fatigue or weakness, or both; an index of 3, that the patient had some pain with activities such as hiking, golf, and doubles tennis but had no limitation in the activities of daily living despite some pain; an index of 2, that the patient was satisfied with the result of the operation despite pain that restricted activities such as walking, hiking, and golf and that limited the activities of daily living; and an index of 1 or 0, that the patient had noted little or no improvement after the operation and regretted that the operation had been performed.
Statistical Analysis
The data were analyzed with use of Excel 4.0 software (Microsoft, Bellevue, Washington) running on a Macintosh Centris 650 computer (Apple Computer, Cupertino, California). Significance was determined with the Student t test, analysis of variance, and Pearson correlation coefficients.
The series comprised forty patients (forty-one feet). Seventeen patients (seventeen feet), including nine women and eight men, had post-traumatic osteoarthrosis of the mid-tarsal or tarsometatarsal joints, or both; the mechanism of injury was a motor-vehicle accident (nine feet), a fall (three feet), an athletic injury (three feet), and a crush injury (two feet). Twenty-one patients (twenty-two feet), including eighteen women and three men, had primary degenerative osteoarthrosis. Two patients (two feet) had inflammatory arthritis; one was a woman who had rheumatoid arthritis and one was a man who had lupus erythematosus.
The time between the injury and the arthrodesis for the patients who had post-traumatic osteoarthrosis was 2.8 years (range, 0.5 to 14.0 years). The time between the approximate onset of symptoms and the arthrodesis for the patients who had primary degenerative osteoarthrosis was 10.9 years (range, 1.8 to 30.8 years). The time between the approximate onset of symptoms and the arthrodesis was ten years for one of the patients who had inflammatory arthritis and fourteen years for the other.
Preoperatively, all patients reported that they had pain that severely restricted the ability to walk and to perform the activities of daily living. None of the patients walked for pleasure. Thirty-one (78 per cent) of the forty patients reported problems with the posture of the foot, such as increased abduction or adduction of the foot or loss of the longitudinal arch, or both. Most of the patients had problems with wearing shoes secondary to the deformity. The patients who had primary degenerative osteoarthrosis noted proliferative bone on the dorsum of the foot and occasionally on the plantar aspect beneath the region of the medial cuneiform. Postoperatively, most patients were pleased with the absence of pain while walking and with the correction of the deformity.
Arthrodesis was done on an average of approximately six joints per foot in the patients who had post-traumatic osteoarthrosis and on an average of approximately four joints per foot in the patients who had primary degenerative osteoarthrosis (Table II). Arthrodesis was performed on seven joints in one of the patients who had inflammatory arthritis and on one joint in the other. Union was achieved after 176 (98 per cent) of the 179 attempted arthrodeses.
Postoperatively, one patient used an ankle-foot orthosis and eleven used an orthotic device such as a Hapad (Hapad, Bethel Park, Pennsylvania) or a Spenco liner (Spenco Medical, Waco, Texas), or both. All of the patients (except for one who had a non-union) could wear flat-heeled sport shoes comfortably, and most women could wear low-heeled dress shoes on occasion.
The average satisfaction index was approximately 4 for the seventeen patients (seventeen feet) who had post-traumatic osteoarthrosis as well as for the twenty-one patients (twenty-two feet) who had primary degenerative osteoarthrosis; the two patients who had inflammatory arthritis had satisfaction indices of 4 and 0. Over-all, thirty-eight (93 per cent) of the forty-one feet had a satisfactory result (an index of 2, 3, 4, or 5) and three (7 per cent) had an unsatisfactory result (an index of 0 or 1) (Table III).
Complications
A skin slough developed in two patients (Table IV); one (Case 33) was managed with operative débridement and the other (Case 39) received local wound care in the office.
A neuroma developed in three patients (Cases 3, 16, and 28) in association with the incision between the second and third metatarsals. None of these patients needed additional treatment.
Five patients (Cases 12, 15, 26, 33, and 34) noted at least one prominent metatarsal head. The second metatarsal head was prominent in two feet; the first and second metatarsal heads, in one foot; and the second and third metatarsal heads, in two feet. None of these patients needed débridement because of abnormal callus formation in the region of the prominent metatarsal head. Symptoms that were related to a prominent metatarsal head were successfully treated with the placement of a Hapad into the shoe just proximal to the metatarsal head or with the placement of a Spenco liner into the shoe.
A stress fracture of the second metatarsal developed in three patients (Cases 15, 26, and 38). All of these fractures healed after conservative treatment with a postoperative shoe or a cast.
Three patients had a non-union. One patient (Case 1) had no symptoms related to the non-union, one (Case 14) was managed with an additional arthrodesis in order to extend the fusion mass, and one (Case 33) had symptoms but stated that she had noted sufficient functional improvement that she did not desire an additional operation. This latter patient had a satisfaction index of 1. The joints that failed to unite were the lateral cuneiform-cuboid, the tarsal-first metatarsal, and the medial cuneiform-navicular. Fusion was achieved in the other joints on which arthrodesis had been performed.
Radiographic Findings
As mentioned previously, arthrodesis was done on an average of approximately six joints per foot in the patients who had post-traumatic osteoarthrosis and on an average of approximately four joints per foot in the patients who had primary degenerative osteoarthrosis (Table II). This difference was highly suggestive but not significant (p < 0.053).
A normal value as well as average preoperative and postoperative values were determined for each of the measured angles in each group (Tables IV and V). The unaffected extremity served as a control. (The average normal value for the group of patients served as the control for the patient who had an operation on both feet.) Preoperatively, the patients who had primary degenerative osteoarthrosis demonstrated a more pronated posture of the foot than the patients who had post-traumatic osteoarthrosis; the degree of deformity in the anteroposterior plane was almost twice as great and it was even greater in the lateral plane. Significant differences between the two groups were noted with regard to coverage of the talar head by the navicular and with regard to the talus-second metatarsal angle (p < 0.01 for both comparisons).
The true magnitude of the correction of the deformity is not obvious at first. For example, the average talus-first metatarsal angle on the anteroposterior radiograph was 11 degrees preoperatively and 9 degrees postoperatively for the patients who had post-traumatic osteoarthrosis, and it was 22 degrees preoperatively and 23 degrees postoperatively for the patients who had primary degenerative osteoarthrosis (Table V). The actual change was 9 degrees for the patients who had post-traumatic osteoarthrosis and 8 degrees for the patients who had primary degenerative osteoarthrosis, which reflects the degree of correction. The apparent discrepancy between the difference in the measured values and the absolute correction occurred because some feet had correction of an adduction deformity (a negative value) and others had correction of an abduction deformity (a positive value).
The joints adjacent to the site of the arthrodesis were assessed for evidence of osteoarthrosis that had not been present preoperatively. A mild degree of osteoarthrosis was observed in three patients (five joints) who had post-traumatic osteoarthrosis, in eight patients (twelve joints) who had primary degenerative osteoarthrosis, and in neither of the patients who had inflammatory arthritis. The changes occurred most frequently in the naviculocuneiform joints.
The purpose of the present study was to analyze the results of treatment of symptomatic osteoarthrosis of the mid-tarsal or tarsometatarsal joints, or both. Several authors have described the treatment of acute trauma1,2,5,12 or post-traumatic osteoarthrosis11,14 of the tarsometatarsal joints, but we know of only one report that has described the treatment of primary degenerative osteoarthrosis of the mid-tarsal or tarsometatarsal joints10. Moreover, only two of the nine feet in the latter report were in patients who had primary degenerative osteoarthrosis. Clearly, the literature provides little information on the anatomical presentation and treatment of primary degenerative osteoarthrosis compared with those of post-traumatic osteoarthrosis8,11-14.
Johnson and Johnson apparently were the first to report on the technique and results of arthrodesis for post-traumatic osteoarthrosis of the tarsometatarsal joints. They recommended use of a dowel graft but stated that they had made no attempt to correct any deformity. Sangeorzan et al. presented the concept of correction of the residual deformity followed by rigid internal fixation with screws14. They pointed out that a bone graft often was not needed for successful arthrodesis in such instances. Horton and Olney described the use of a medial plate to maintain the corrected alignment and to achieve rigid fixation.
Our patients were referred to us from a foot-and-ankle practice and were almost equally divided into two groups: those who had primary degenerative osteoarthrosis (twenty-one patients) and those who had post-traumatic osteoarthrosis (seventeen patients). The primary differences between the two groups were the average age of the patients (forty years for those who had post-traumatic osteoarthrosis compared with sixty years for those who had primary degenerative osteoarthrosis) and the extent of the deformity of the foot, which was much greater in the patients who had primary degenerative osteoarthrosis. The average age of our patients who had post-traumatic osteoarthrosis was similar to that reported in the literature for patients who have had operative treatment of this disorder11,14.
Clinically, the patients in both groups had similar symptoms (disabling pain and progressive deformity) when they were first seen. Abduction and dorsiflexion of the mid-tarsal area was the most common deformity in both groups, but it was more severe in the patients who had primary degenerative osteoarthrosis. This deformity leads to flattening of the longitudinal arch and occasionally to a painful osseous prominence in the area of the medial cuneiform. Adduction and plantar flexion deformities were seen in a few patients, most of whom had post-traumatic osteoarthrosis. Despite the difference in the anatomical presentation, the clinical results of both groups were essentially the same.
The average correction of the talus-first metatarsal angle in the anteroposterior plane in the present study was similar to the 10 degrees of correction reported by Horton and Olney, and the correction in the lateral plane that we observed was half the correction of 15.5 degrees reported by those authors. Other investigators either did not report the degree of correction11,14 or stated that they had made no effort to correct the deformity11. Sangeorzan et al. believed that the correction of a residual deformity was the most important predictor of a good outcome14. We agree that correction of the deformity plays a large role in the achievement of a satisfactory clinical result, as three of our four patients who had a satisfaction index of 2 or less had worsening (two patients) or persistence (one patient) of the deformity postoperatively.
The approach through longitudinal incisions and the creation of full-thickness flaps are important because wide exposure is necessary to carry out the operation adequately11,14. Although three patients in the present study had a clinically symptomatic neuroma, this number is relatively small when one considers that most patients had two or three incisions. All of the neuromas occurred in the incision over the dorsum of the foot, where branches of the superficial peroneal nerve fan out; such branches occasionally are difficult to observe, especially if a previous operation has been performed. Although it is preferable to avoid the problem altogether, this is not always possible. Fortunately, none of the neuromas in the series was sufficiently bothersome for the patients to need additional operative treatment.
Many different patterns of arthrodesis can be performed for osteoarthrosis of the tarsometatarsal joints. Johnson and Johnson performed an average of approximately four arthrodeses per foot; ten of their fifteen patients had an arthrodesis of the second and third tarsometatarsal joints, and the other five had an isolated arthrodesis of the first or second tarsometatarsal joint. Those authors stated that an arthrodesis that includes the second and third tarsometatarsal joints as well as the middle and lateral cuneiforms results in a more biomechanically stable foot than does an isolated tarsometatarsal arthrodesis. Sangeorzan et al., who performed arthrodesis on an average of three joints per foot, performed only one intercuneiform arthrodesis in sixteen patients14. Horton and Olney performed arthrodesis on an average of approximately five joints per foot; the arthrodesis included the first, second, and third tarsometatarsal joints along with the medial and middle cuneiforms in eight of the nine feet. Our patients who had primary degenerative osteoarthrosis had an average of approximately four arthrodeses per foot; an arthrodesis was done on the first tarsometatarsal joint in nineteen of the twenty-two feet in that group, and it included the second tarsometatarsal joint in twelve. The patients who had post-traumatic osteoarthrosis had an average of approximately six arthrodeses per foot. An arthrodesis was done on the first tarsometatarsal joint in sixteen of the seventeen feet in that group; it included the second tarsometatarsal joint in fifteen feet and the medial and middle cuneiforms in fourteen feet. We believe that if the viability of a joint is questionable, it is better to include it in the arthrodesis. We extended the arthrodesis into the naviculocuneiform joints in eight of forty-one feet. Horton and Olney also reported extending the arthrodesis site into this area in three of nine feet. The rate of failure in the present study (7 per cent; three of forty-one feet) was much lower than that reported by Sangeorzan et al.14 (five of sixteen feet) and by Johnson and Johnson (two of thirteen feet); this may have been due to the fact that we tended to create a larger fusion mass that incorporated the intertarsal or naviculocuneiform joints, or both.
No association was found between the age of the patient and the clinical result; this finding is in agreement with those of previous studies11,14. Although Sangeorzan et al. noted an association between early arthrodesis and a good result in patients who had post-traumatic osteoarthrosis14, this was not supported by the findings in our series.
After an arthrodesis of a tarsometatarsal or mid-tarsal joint, or both, the rigidity of the foot seems to be well tolerated by the patient. The three stress fractures that occurred were due to abnormal loading of the metatarsal heads, which was a result of our inability to realign them precisely. Although the stress fractures were a brief clinical problem for the patients, they had no lasting effect on the end result.
The best form of internal fixation involves interfragmentary screws14 or a medial plate10, or both. No specific screw construct was noted to be better than another for the repair of multiple joints. We found that each foot presented its own particular technical problems with regard to the placement of screws. Because the realignment of the foot is extremely important, we recommend that a buttress plate be used as a guide along the medial side of the foot when a severe abduction deformity is being corrected. Although such realignment can be achieved without a plate, we found this device to be useful. A bone graft was used in only eleven (27 per cent) of forty-one feet; this is consistent with the results of Sangeorzan et al., who found that bone-grafting was not necessary for most patients14. Other authors10,11, however, have used a bone graft for every patient. We believe that bone-grafting is not necessary if the reduction is adequate. The high rate of fusion in the present study (98 per cent; 176 of 179 joints) supports the effectiveness of our operative technique.
Although osteoarthrosis in adjacent joints has been observed after arthrodesis of the hindfoot9, we noted relatively few such changes after arthrodesis of the mid-tarsal or tarsometatarsal joints, or both. This is probably due to the limited motion of these articulations in the normal foot7.
In conclusion, we believe that realignment of the foot and arthrodesis of the mid-tarsal and tarsometatarsal joints can lead to a satisfactory result for patients who have primary degenerative osteoarthrosis and post-traumatic osteoarthrosis of the mid-tarsal or tarsometatarsal joints, or both. Despite a severe functional disability preoperatively, more than 90 per cent of our patients had a satisfactory result.