Enchondromas are benign growths of cartilage that develop within the medullary cavity of bone. A patient may have a solitary lesion12,13 or a more diffuse form of the disease, such as multiple enchondromatosis (Ollier disease)26.
Metaphyseal and diaphyseal involvement in patients who have enchondromatosis has been well described2-4,8,12,14,20,25,30-33,37,38. Although metaphyseal lesions have been noted to cause irregularities of the adjacent epiphyses in skeletally immature patients who have Ollier disease18, it is generally believed that any extension of the cartilaginous inclusions into the epiphysis occurs after closure of the growth plate19.
In the current paper, we describe a previously unreported clinical form of enchondromatosis in which there is extensive involvement of the epiphyseal and metaphyseal regions of the long bones of the lower extremity. The enchondromas appear before closure of the growth plate and lead to severe limb-length discrepancy and angular deformity, premature asymmetrical physeal arrest, and joint incongruity.
*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, Alfred I. duPont Hospital for Children, P.O. Box 269, Wilmington, Delaware 19899. E-mail address for Dr. Gabos: pgabos@aidi.nemours.org.
Epiphyseal Involvement
Each patient had unilateral involvement of a mean of three (two, three, or four) epiphyses of the lower extremity at the time of presentation (Table I). Speckled islands of radiodensity and amorphous radiolucency consistent with the radiographic appearance of an enchondroma were present in the distal femoral and proximal tibial epiphyses of each patient. Five patients had an epiphyseal enchondroma of the distal part of the tibia as well. Lesions of the proximal fibular epiphysis and the greater trochanter were noted in one patient each, and three patients had involvement of the lesser trochanter. No patient had epiphyseal involvement bilaterally, although one patient (Case 1) had radiographic involvement of the metaphyseal region of the proximal part of the contralateral fibula that was not of clinical importance. Epiphyseal involvement was always associated with metaphyseal lesions. Additional enchondromatous areas were noted in two patients: one (Case 3) had lesions of the ipsilateral metacarpal and metatarsal shafts, and the other (Case 8) had lesions of the ipsilateral ilium and the proximal part of the humeral shaft. No patient had vertebral involvement or a deformity of the skull.
In six patients (Cases 1, 2, 3, 6, 7, and 8), the involvement of the epiphysis included lesions in the subchondral bone. One patient (Case 4), who had epiphyseal enchondromas but no radiographically evident lesions in the subchondral area, was found to have multiple cartilaginous protuberances in the distal femoral articular cartilage at the time of arthroscopy that was performed because of a symptomatic discoid lateral meniscus. Five patients (Cases 1, 2, 3, 6, and 8) had a longitudinally oriented radiolucent streak associated with enchondromas extending directly across the physis. This streak was centrally located in the distal part of the femur in four patients and in the distal part of the femur and the proximal part of the tibia in one patient (Case 3) (Fig. 3). One patient (Case 6) had enchondromas extending through the central portion of the distal femoral physis and the lateral one-half of the proximal tibial physis (Fig. 4-A), with premature physeal closure of these areas by the time that he was ten years old (Fig. 4-B).
Limb-Length Discrepancy
At the time of presentation, the limb-length discrepancy averaged 6.7 centimeters (range, 2.8 to 11.1 centimeters). The mean femoral discrepancy was 4.3 centimeters (range, 1.4 to 7.2 centimeters), and the mean tibial discrepancy was 2.4 centimeters (range, 0.0 to 5.2 centimeters) (Table I).
The limb-length discrepancy increased in all patients, and seven had an initial lengthening procedure at a mean age of 9.7 years (range, 8.7 to 10.7 years) (Table II). At the time of this procedure, the limb-length discrepancy averaged 11.7 centimeters (range, 8.6 to 18.2 centimeters) and the mean predicted limb-length discrepancy at maturity was 18.0 centimeters (range, 9.5 to 26.0 centimeters). Five patients had an additional lengthening procedure at a mean age of 12.7 years (range, 11.8 to 13.7 years), and four had an epiphyseodesis at a mean age of 13.5 years (range, 12.2 to 14.2 years).
One patient (Case 8) initially was managed with a modified Boyd amputation1 at the age of 5.8 years. The predicted limb-length discrepancy at maturity was 26.0 centimeters. Amputation was chosen because the parents wanted to avoid repeated lengthening procedures.
Angular Deformity
At the time of presentation, seven patients had a distal femoral varus deformity averaging 26 degrees (range, 12 to 38 degrees); one, a distal femoral valgus deformity of 7 degrees; three, a proximal tibial varus deformity averaging 18 degrees (15, 17, and 21 degrees); three, a proximal tibial valgus deformity averaging 21 degrees (19, 21, and 23 degrees); and four, a distal tibial valgus deformity averaging 8 degrees (range, 4 to 11 degrees). The tibiofemoral angle was varus in six patients, averaging 20 degrees (range, 10 to 40 degrees), and valgus in two (5 and 8 degrees). The angular deformity increased in every patient from the time of presentation.
No patient had coxa vara or valga, and there were no instances of subluxation, dislocation, or dysplasia of the hip.
Operative Treatment
Initial Lengthening Procedures
Concomitant lengthening of the femur and tibia was performed in five patients (Cases 1, 2, 3, 5, and 6) (Table II). The amount of femoral lengthening averaged 7.2 centimeters (range, 6.1 to 8.7 centimeters) and 26 per cent (range, 21 to 33 per cent) of the total length of the bone. The amount of tibial lengthening averaged 8.0 centimeters (range, 4.7 to 10.7 centimeters) and 34 per cent (range, 19 to 47 per cent) of the total length of the bone. The mean total gain in limb length at the time of removal of the fixator was 15.2 centimeters (range, 13.1 to 18.0 centimeters).
Of the three remaining patients, one (Case 4) had lengthening, of 7.9 centimeters, of the femur only. Another patient (Case 7) had lengthening, of 4.5 centimeters, of the tibia only; however, the lengthening procedure was terminated prematurely because a complete peroneal nerve palsy developed during the postoperative distraction process. The third patient (Case 8) had a Boyd amputation without lengthening.
Additional Lengthening Procedures
Five patients (Cases 1, 3, 4, 6, and 7) had a second lengthening procedure because of residual limb-length discrepancy (Table II). The limb-length discrepancy at the time of this additional procedure averaged 4.0 centimeters (range, 0.5 to 5.9 centimeters), and the predicted limb-length discrepancy at maturity averaged 6.5 centimeters (range, 5.0 to 11.0 centimeters).
One patient (Case 6) had concomitant lengthening of the femur and tibia, resulting in an additional increase in length of 4.8 centimeters. Lengthening of 5.0 centimeters had been planned, but the procedure was terminated prematurely because of psychological distress of the patient during the postoperative distraction process.
Two patients (Cases 3 and 7) had lengthening of the femur only, resulting in an additional increase in length of 7.5 and 6.5 centimeters, respectively.
Two patients had lengthening of the tibia only. At the time of the procedure, one of these patients (Case 1) had a limb-length discrepancy of only 0.5 centimeter but the predicted discrepancy at maturity was 5.0 centimeters and he had recurrent angular deformity. He was managed with application of femoral and tibial circular-ring external fixators, realignment of the limb through distal tibial varus-extension and distal femoral valgus closing-wedge osteotomies, and postoperative distraction of the tibia through a proximal tibial corticotomy. The primary reason for the second lengthening procedure was to compensate for shortening produced by the closing-wedge osteotomies. The lengthening was stopped after 1.4 centimeters had been gained, as the limb lengths had been equalized. In the other patient (Case 4), the second lengthening procedure resulted in an additional increase in length of 5.0 centimeters.
Overall Gain in Length
The total increase in length for the seven patients who had lengthening averaged 16.2 centimeters (range, 11.0 to 22.2 centimeters).
Epiphyseodeses
Four patients (Cases 1, 2, 6, and 7) had an epiphyseodesis of the contralateral lower extremity at a mean age of 13.5 years (range, 12.2 to 14.2 years).
Osteotomies
Thirteen femoral and fourteen tibial realignment osteotomies (range, one to five procedures per patient) were performed. Twelve osteotomies were performed during a lengthening procedure. In one patient (Case 6), femoral realignment and lengthening was performed through the site of a fresh pathological fracture sustained in a fall from a jungle gym. Six patients needed a repeat osteotomy because of recurrent deformity. One patient (Case 2) had a single osteotomy during a lengthening procedure and did not have recurrence of the deformity. Another patient (Case 8) had a distal femoral valgus osteotomy at the time of a Boyd amputation, with satisfactory maintenance of alignment.
No attempt was made to avoid bone with enchondromatous involvement during either realignment osteotomy or corticotomy. All femoral and tibial osteotomy sites healed regardless of the location. However, one patient (Case 2) was managed with autogenous iliac-crest bone graft because of a delayed union at the site of a proximal femoral lengthening that had been performed through an area of uninvolved bone.
Complications
Complications related to lengthening included a complete peroneal nerve palsy in two patients. One patient (Case 7) had partial recovery of the palsy over a twenty-day period after cessation of the distraction process. The remaining patient (Case 1) did not regain peroneal nerve function.
Eleven pin-track infections occurred in four patients (range, one to four infections per patient). Ten infections were superficial, and one (with Staphylococcus aureus) was deep, necessitating antibiotic therapy and curettage. The latter infection was treated with cefazolin sodium (500 milligrams every eight hours for four days), administered intravenously, followed by a six-week course of cephalexin capsules (500 milligrams every six hours), taken orally.
Eleven pathological fractures occurred in four patients (range, one to five fractures per patient). Nine fractures occurred through enchondromatous bone before lengthening and two, through either a pin track or a lengthened segment after lengthening. One patient (Case 2) had an incomplete fracture through a proximal femoral pin track and was managed with plate-and-screw fixation. The fracture healed, but a deep infection developed in the area fourteen months later, necessitating removal of the hardware. Another patient (Case 6) was managed with application of a Wagner device. This patient was awaiting the initial lengthening-and-realignment procedure, and the fracture afforded the opportunity to lengthen and realign the femur through the site of the fracture. An additional Wagner device was placed on the tibia, and lengthening and realignment was performed through a proximal tibial corticotomy. All of the remaining fractures healed with non-operative treatment.
Follow-up
The mean age at the time of the latest follow-up was 17.6 years (range, 13.8 to 24.9 years) (Table III). Of the six patients who had reached skeletal maturity, one (Case 8) had been managed with an early Boyd amputation and was functioning well with use of a below-the-knee prosthesis. At the age of twenty-four years, a low-grade chondrosarcoma developed in an enchondromatous area in the left scapula; the lesion was treated at another institution with a partial scapulectomy. Of the five remaining patients, four had a limb-length discrepancy of less than 2.5 centimeters and no functional limitations. In one of these patients (Case 5), an acinic-cell carcinoma of the parotid gland developed when the patient was fifteen years old, necessitating operative resection. The sixth patient (Case 2) had a residual limb-length discrepancy of 5.0 centimeters and functioned well with use of a shoe support; he had no interest in additional lengthening.
Of the two patients who had not yet reached skeletal maturity, one (Case 1) had a limb-length discrepancy of 3.1 centimeters and an abnormal gait secondary to a flexion contracture and an angular deformity at the knee and ankle. He wore a solid ankle-foot orthosis because of a complete peroneal nerve palsy caused by the first lengthening procedure. He subsequently had epiphyseodeses of the contralateral limb but refused to have realignment osteotomies on the involved side. The remaining patient (Case 3) was awaiting a third lengthening procedure at the time of this writing because of a limb-length discrepancy of 12.5 centimeters.
The tibiofemoral angle at the time of the latest follow-up was varus, measuring 10 and 12 degrees, in two patients. It was valgus, averaging 7 degrees (range, 2 to 15 degrees), in six.
Six patients (Cases 1, 2, 3, 5, 6, and 7) had radiographic evidence of incongruity of the knee joint by a mean age of 15.9 years (range, 13.8 to 17.6 years) (Fig. 5). Five of these patients had subchondral enchondromas of the distal part of the femur or the proximal part of the tibia, noted early in the course of the disease (Figs. 6-A, 6-B, and 6-C). Two patients (Cases 1 and 3) had incongruity of the ankle joint (Figs. 7-A and 7-B). No patient reported joint pain.
In the six patients who had reached skeletal maturity, the enchondromas appeared to have stopped enlarging, with no additional expansion of the area of involvement in either the epiphyseal or the metaphyseal regions of the long bones. One patient (Case 5) had regression of the epiphyseal lesions; no discrete enchondromas were visible radiographically despite an incongruent knee joint with a disorganized epiphyseal trabecular pattern.
The pathological lesion that is termed an enchondroma is characterized by aberrant rests of epiphyseal growth-plate chondrocytes in the metaphyseal and diaphyseal regions of bones. This variation from the process of normal chondrocyte hypertrophy and death probably does not represent a true neoplasia but rather a developmental anomaly leading to regionally confined or generalized dysplasias. Enchondromas thus may exist as solitary lesions12,13; as localized entities, such as distal enchondromas involving predominantly the short tubular bones of the hands and feet19; or as the distinguishing components of more generalized disorders such as Ollier disease26. Spranger et al.33 classified enchondromatosis into six distinct types: Ollier disease, Maffucci syndrome, metachondromatosis, spondyloenchondrodysplasia, enchondromatosis with irregular vertebral lesions, and generalized enchondromatosis. Weickert and Friedel40 reported on a new form of enchondromatosis associated with enchondromas originating in the patellae, with associated cold-sensitive exanthema, pigment spots, drumstick fingers, watch-glass nails, and disordered dentition. More recently, Freisinger et al.9 described an enchondromatosis associated with severe segmental abnormalities of the vertebral column, which they termed dysspondylochondromatosis.
There have been numerous reports characterizing metaphyseal and diaphyseal involvement in patients who have enchondromatosis2-4,8,12,14,20,25,30-33,37,38. Mainzer et al.18 noted that the epiphysis adjacent to an involved metaphysis may be irregular in patients who have Ollier disease, but it is generally believed that extension of a lesion from the metaphysis into the epiphyseal end of the bone occurs secondarily, after closure of the growth plate19.
Enchondromas in the epiphyseal region of skeletally immature bone have been rarely reported. In 1934, Hunter and Wiles11 described a seven-year-old boy who had Ollier disease; radiographs demonstrated scattered punctate spots in the proximal femoral capital epiphysis. Those authors attributed the speckling to irregularities of calcification. Shapiro32, in a report on twenty-one patients who had Ollier disease, stated that, after the middle of the first decade of life, enchondromas became evident in some epiphyses but severe epiphyseal deformity did not occur. The number of patients with this finding was not provided. Milgram22 analyzed histological sections of a resected distal part of the femur from a six-year-old boy and a resected proximal part of the fibula from a nine-year-old girl, both of whom had Ollier disease; the sections showed multiple enchondromas within the osseous epiphysis and the articular cartilage. Milgram thought that these enchondromas had formed from abnormal chondrocytes on the reverse surface of the articular cartilage that had not undergone the normal process of calcification and osseous remodeling. We believe that these two cases22 represent epiphyseal-metaphyseal enchondromatosis of the type described in the current report.
Unlike other types of enchondromatosis, epiphyseal-metaphyseal enchondromatosis is characterized by extensive development of enchondromas within the epiphysis before closure of the growth plate. In most patients, there is direct extension of the lesions across the epiphyseal growth plate into the metaphysis. Epiphyseal involvement is unilateral and confined to the long tubular bones of the lower extremity. The location of the lesions in both the epiphysis and the metaphysis is responsible for dramatic and progressive limb-length inequality and angular deformity, asymmetrical premature physeal arrest, and joint incongruity. Multiple limb-lengthening procedures, epiphyseodeses, and corrective osteotomies are necessary, with recurrence of deformity being the rule before skeletal maturity.
In some ways, this entity is reminiscent of other osteochondrodysplasias, including dysplasia epiphysealis hemimelica, in which there appears to be an abnormality of the growth of the postnatal centers of epiphyseal ossification. Dysplasia epiphysealis hemimelica also has a predilection for the epiphyses of the long tubular bones of the lower extremity as well as the tarsal bones; however, the epiphyses are enlarged asymmetrically, with no metaphyseal involvement. The characteristic lesion in dysplasia epiphysealis hemimelica more closely resembles an intra-articular osteochondroma, with the normal progression of cancellous bone formation extending into a cartilage cap21,36.
It is notable that one of our patients had a secondary low-grade chondrosarcoma in enchondromatous bone and that one had an acinic-cell carcinoma of the parotid gland. There have been numerous reports of malignant degeneration of enchondromas to chondrosarcomas5,7,10,27,31,37, dedifferentiated chondrosarcomas6,23, and osteosarcomas2,28. Non-skeletal malignant lesions of mesenchymal and non-mesenchymal origin have been noted in patients who have Maffucci syndrome15,16,29,31,34 and in those who have Ollier disease31,35,39. It remains unclear whether these non-skeletal tumors represent a coincidental finding or are evidence of a more generalized dysplasia of mesoectodermal origin.
As patients are referred to the Limb-Lengthening Clinic at our institution primarily because of inequality of the lengths of the lower extremities, the spectrum of involvement in this study may have been weighted toward lesions that predominantly involved the lower extremity. We found no areas of epiphyseal involvement of the upper extremity; however, only areas with clinically apparent deformity were examined radiographically.
In summary, we have described a group of patients who had epiphyseal-metaphyseal enchondromatosis. There is a lack of evidence regarding the etiological or pathogenetic basis of the various forms of enchondromatosis that have been classified thus far. It remains unclear whether these disorders represent distinct nosologic entities related to one another by the presence of a common histological lesion or whether they are simply different expressions of the same basic disorder. We believe that epiphyseal-metaphyseal enchondromatosis differs enough from the previously classified enchondromatoses to merit its description as a new clinical entity.