The treatment of chronic radiocapitellar dislocation after a Monteggia lesion in children has included observation5,17,26; closed16,27,29 or open1,3,4,6-9,11,13,15,16,18-22,27,30,31,33 reduction of the radial head; repair7,12,16,18 or reconstruction1,3,4,6,7,9,11,12,16,17,19-21,24,27-32 of the annular ligament; and osteotomy of the ulna4,7,9,11-13,15,18,19,21,22,27,29-31 or radius13,27,29, or both, to treat malunion.
We present our results of the operative treatment of chronic radiocapitellar dislocation after a Monteggia lesion in seven children. We found the results of such treatment to be unpredictable and associated with a number of serious complications.
*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, Harvard Medical School, Boston Children's Hospital, Hunnewell 2, 300 Longwood Avenue, Boston, Massachusetts 02115.
Seven patients, three boys and four girls, who had a chronic radiocapitellar dislocation following a Monteggia lesion were managed operatively at Boston Children's Hospital from 1981 to 1992. The mean age at the time of the injury was five years and ten months (range, seven months to nine years and ten months), and the mean age at the time of the reconstruction was six years and nine months (range, eleven months to twelve years). The right arm was injured in two patients and the left, in five (Table I).
The initial injury, an anterior radiocapitellar dislocation (Bado2 type I or the equivalent), had been missed in all seven patients. None of the patients had been seen initially at our hospital. The ulnar fracture, which was always diaphyseal, was oblique in the four patients for whom the initial radiographs were available for review, and it had healed by the time that the remaining three patients were seen by us. One patient (Case 6) had an associated diaphyseal fracture of the radius (so-called Bado type IV), and another patient (Case 7) had a dislocated distal radio-ulnar joint. Five patients had been managed with a cast, and two had not received any treatment.
For three patients (Cases 3, 6, and 7), the accurate diagnosis was made (two to five weeks after the injury) while the cast was still in place. One patient (Case 3), who was referred to us two weeks after the injury and the application of the cast, had two unsuccessful attempts at closed reduction at our institution.
On referral to our institution two weeks to thirty-nine months after the injury, four patients (Cases 1, 2, 4, and 5) were no longer wearing or had never worn a cast and could be examined. Two (Cases 4 and 5) had a palpably dislocated radial head (Fig. 1). Two (Cases 2 and 4) had an obvious valgus carrying angle of 35 and 31 degrees, respectively; the carrying angle of 31 degrees increased to 45 degrees with stress. The remaining patient (Case 1) had been managed in a cast for three weeks, and the diagnosis was made two months after the injury. He was noted to have a posterior-interosseous-nerve palsy at the time of referral. During the operation, the nerve was found to be draped around the radial neck and entrapped in the joint. It was partially lacerated during exposure of the dislocated radial head and was repaired primarily under loupe magnification.
The four patients for whom the range of motion could be evaluated had mean flexion of the elbow of the injured extremity of 120 degrees (range, 90 to 135 degrees), mean extension of -3 degrees (range, -40 to 30 degrees), mean pronation of 69 degrees (range, 5 to 90 degrees), and mean supination of 83 degrees (range, 70 to 90 degrees). The ranges of motion of the contralateral extremities were not available.
Preoperative radiographs were not standardized; therefore, meaningful calculations could not be made. The radiocapitellar joint was always dislocated anteriorly. Two patients did not have secondary centers of ossification in the radial head at the time of the injury. Callus was visible at the site of the ulnar fracture in four patients (Cases 1, 2, 6, and 7).
The type of index operation, performed a mean of twelve months (range, five weeks to thirty-nine months) after the injury, was based on the preference of the surgeon. The radial and ulnar nerves were not routinely identified during the operation. Six patients (Cases 1 through 6) had open reduction of the radiocapitellar joint through a Kocher approach. The annular ligament was repaired in three patients (Cases 1, 3, and 6) and was reconstructed with use of the triceps fascia in two (Cases 2 and 4) and with use of the ulnar periosteum in one (Case 5) (Table I). Four patients (Cases 1 through 4), who had instability of the radiocapitellar joint, also had an osteotomy of the ulna, and one patient (Case 7) had an osteotomy without repair of the annular ligament (Table I). A local cancellous graft from the ulna was used in three patients (Cases 2, 3, and 7), a tricortical autogenous graft from the iliac crest was used in one (Case 1), and a tricortical allograft from the iliac crest was used in one (Case 4). The osteotomy site was stabilized with an intramedullary Kirschner wire in one patient (Case 1), with an intramedullary Rush rod in another (Case 4), and with a five-hole 3.5-millimeter plate in two patients (Cases 3 and 7). One patient (Case 2) did not have any fixation. In one patient (Case 6), who had only repair of the annular ligament without an ulnar osteotomy, the radiocapitellar joint was stabilized with a transcapitellar wire for two weeks (Table I). An above-the-elbow cast was worn until there was union of the osteotomy site; the plates were removed after healing (seven to thirteen months after the operation).
The complications included ulnar nerve palsy in three patients (Cases 1, 2, and 4) who had had an opening-wedge osteotomy. A compartment syndrome developed in the forearm of one patient (Case 3) after an ulnar osteotomy and fixation with a plate, open reduction of the radiocapitellar joint, and direct repair of the annular ligament. An emergent fasciotomy was performed, with delayed closure of the wound forty-eight hours later. These complications may have been related to the distraction of the osteotomy site. One patient (Case 1) had a laceration of the posterior interosseous branch of the radial nerve as it was entrapped in the dislocated radiocapitellar joint. Preoperatively, he had had a radial nerve palsy, although the electrodiagnostic studies did not show a conduction block. The nerve was repaired after exposure of the annular ligament. The patient had a complete recovery of function by six months postoperatively.
The fixation of the osteotomy site was lost in one patient (Case 3) in the operating room after skin closure and in another patient (Case 1) in the recovery room. The fixation in the latter patient was changed from an intramedullary Kirschner wire to a five-hole plate, and the osteotomy site healed without additional problems. One patient (Case 4), who had grafting with allograft bone, had a non-union that was treated with an autogenous bone graft seven months later. The osteotomy healed uneventfully.
One patient (Case 1) refused to move the elbow postoperatively. An arthrogram, performed at the time of removal of the plate (seven months after the index procedure), showed normal findings. A psychiatrist thought that the refusal to move the elbow was a conversion reaction secondary to unresolved psychological trauma.
Six patients were evaluated by us at the latest follow-up evaluation. A questionnaire was completed, and an anteroposterior radiograph was made with the elbow in maximum extension and a lateral radiograph was made with the elbow flexed 90 degrees, the forearm pronated, and the shoulder abducted 90 degrees. One patient, who was unable to return to Boston for evaluation, was interviewed by telephone and completed the questionnaire, providing responses regarding pain, over-all functional activity, and any limitation of the activities of daily living. A local orthopaedic surgeon examined the patient according to our protocol and sent the findings of the physical examination as well as the radiographs.
The physical examination included evaluation of the range of motion of both elbows and wrists as well as a complete motor and sensory examination. Grip and two-point pinch strength were measured by averaging the results of three alternating trials performed with a dynamometer. Pain, stability, and crepitus of the proximal and distal radio-ulnar joints were measured quantitatively. The carrying angle was determined with the elbow in maximum extension, and stability to varus and valgus stress was measured with the elbow in maximum extension and 30 degrees of flexion. The stability of the distal radio-ulnar joint was tested in maximum pronation and supination.
Postoperatively, the carrying angle, the capitellum-radial head distance and percentage of subluxation, the diameter of the radial head, the ulnar bow, and the ulnar variance14 were measured for both the injured and the contralateral, uninjured extremity. Any deformity of the radial head, capitellum, or proximal radial or lateral condylar physis of the injured extremity was noted and these structures were compared with those on the contralateral side.
The questionnaire elicited information regarding hand dominance, occupation, athletic interests and avocations, and ability to perform activities of daily living as well as a subjective evaluation of the severity (mild, moderate, or severe), chronicity, and anatomical location of pain in the elbow and wrist joint, stiffness, weakness, or limitation of motion.
We used the rating system of Morrey et al.23, which is based on function, range of motion, and pain; the results were graded as good, fair, or poor. A good result indicates flexion, extension, and stability to varus and valgus stress that are equal to those of the contralateral extremity as well as at least 100 degrees of rotation of the forearm. The patient has no pain or functional disability, and the radiocapitellar joint is located on follow-up radiographs. A fair result indicates at least 100 degrees of flexion and extension; a flexion contracture of less than 30 degrees; and 100 degrees of rotation of the forearm, with a minimum of 30 degrees of pronation; and little instability to varus or valgus stress. Pain, if present, is only occasional, and the patient has no difficulties with activities of daily living. The radiocapitellar joint is reducible on follow-up radiographs. A poor result indicates less than 100 degrees of flexion and extension and of rotation of the forearm, frequent pain, difficulties with activities of daily living, neurological complications that had not been present preoperatively, or recurrent dislocations of the radiocapitellar joint evident on follow-up radiographs.
The mean duration of follow-up was four years and six months (range, two years to eleven years and three months). The mean age at the latest follow-up evaluation was eleven years and four months (range, five years and seven months to twenty years and six months) (Table I). Only one patient (Case 2) had reached skeletal maturity by the latest evaluation. Three patients (Cases 3, 5, and 6) had a good result; two (Cases 2 and 7), a fair result; and two (Cases 1 and 4), a poor result. All of the patients were pain-free.
All of the patients were right-hand dominant at the time of the latest evaluation. We mention this because one patient (Case 5) had not established hand dominance at the time of the injury (at the age of seven months), and another patient (Case 6) was left-hand dominant at the time of the injury (at the age of two years and eight months) but was right-hand dominant at the latest follow-up examination. We have no explanation for this change.
Two patients (Cases 2 and 4) had mildly restricted motion of the elbow or rotation of the forearm and one (Case 1) had moderately restricted motion. Two patients (Cases 1 and 4) who had perioperative ulnar-nerve palsy noted mild weakness of grip. Three patients (Cases 1, 3, and 4) had a mean loss of extension of 15 degrees (range, 7 to 28 degrees). Two of these patients also lost 8 degrees (Case 3) and 13 degrees (Case 1) of flexion. One patient (Case 6) had 19 degrees of hyperextension compared with 10 degrees on the contralateral side. All of the patients had limited rotation, with a mean pronation of 49 degrees (range, 5 to 85 degrees) and a mean supination of 64 degrees (range, 35 to 90 degrees). None of the patients had any instability of the elbow, although two patients (Cases 1 and 4) had a noticeably increased carrying angle. Both of these patients had a flexion contracture of the elbow and, as such, conclusions about the carrying angle may not be valid.
Of the four patients for whom the preoperative range of motion was known, three (Cases 1, 2, and 4) had 35, 10, and 10 degrees, respectively, of improvement in flexion, and one patient (Case 5) had no change. Two patients (Case 2, who had 30 degrees of hyperextension preoperatively and none at the latest follow-up examination, and Case 4) lost extension. One patient (Case 1) gained 14 degrees of extension. In these four patients, the mean arc of lost rotation of the forearm was 64 degrees: 30 degrees (range, 0 to 75 degrees) of pronation and 34 degrees (range, 0 to 50 degrees) of supination.
The motion of the wrist was equal to that on the contralateral side in six patients; the remaining patient (Case 1) lacked 30 degrees of dorsiflexion. The patient (Case 7) who had a dislocated distal radio-ulnar joint had no sequelae at the time of follow-up; however, there was a 16 per cent decrease in grip strength and a 23 per cent decrease in two-point pinch strength compared with those on the contralateral side. As all of the patients were right-hand dominant and five left arms were injured, the effects of handedness on these data cannot be discounted.
The patients who had sustained neurological damage at the time of the injury or the treatment all had residual findings. The patient (Case 1) who had a disruption of the posterior interosseous nerve at the time of the Monteggia injury of the left extremity, and in whom a complete ulnar-nerve palsy developed postoperatively, had no motor weakness at the latest evaluation but did have a Tinel sign over the cubital tunnel and a 20 per cent decrease in grip strength compared with that of the dominant right hand. Two other patients (Cases 2 and 4) in whom postoperative ulnar neuropathy developed noted subjectively decreased grip strength. On motor-testing, they were found to have mild (motor grade 4+ of 5) weakness of the intrinsic muscles and the flexor profundus tendon to the little finger and a 20 per cent decrease in grip strength. One patient (Case 3) had no sequelae from a dorsal compartment syndrome in the forearm.
Three patients had persistent subluxation of the radiocapitellar joint. One patient (Case 7) had anterior subluxation of the radial head in supination that reduced spontaneously (with a palpable click) in pronation. This patient had been managed with an ulnar osteotomy and closed reduction of the radial head without reconstruction of the annular ligament. Another patient (Case 1), who had had repair of the annular ligament, had recurrent anterior subluxation. The remaining patient (Case 4) had posterolateral subluxation of the radial head after reconstruction of the annular ligament with the triceps fascia. This patient had mildly painful crepitus in maximum supination.
Two patients (Cases 4 and 6) had premature closure of the proximal radial physis. One (Case 4) was nearly skeletally mature at the time of the latest follow-up examination. Early closure of the physis may be a potential problem in younger patients. The radial head was enlarged in four patients, and the enlargement was most noticeable one year after the operation in skeletally immature patients. We did not find any association between enlargement of the radial head and decreased rotation.
Two patients (Cases 2 and 5), who had had reconstruction of the annular ligament, had an hourglass constriction of the radial neck (Fig. 2-A). One (Case 2) appeared to have ossification of a portion of the reconstructed annular ligament (Fig. 2-B) and had a small exostosis at the level of a previous osteotomy, which had been performed through the same incision as the ligament reconstruction. This patient had only 85 degrees of rotation of the forearm, half that on the contralateral side, despite a concentric reduction and a radial head of nearly normal size. The patient may have had a partial synostosis of the forearm.
Bado, using the eponym of Monteggia lesion, classified dislocation of the radiocapitellar joint associated with fracture of the ulna2. The acute lesion is misdiagnosed frequently (16 to 33 per cent of the time11,12), especially when there is associated plastic deformation of the ulna or a greenstick fracture. Even when the correct diagnosis is made and a closed reduction is performed, the reduction is often lost (in approximately 20 per cent of patients), especially when the ulnar fracture is oblique11. Unreduced dislocations of the radiocapitellar joint that are still present more than four weeks after the injury are considered chronic.
The treatment of a chronic Monteggia lesion varies from watchful waiting5,17,26 to operative intervention in an attempt to restore congruity of the joint, to improve function of the extremity, and to prevent long-term complications and disability. The problems associated with chronic dislocation of the radial head include deformity3,21, pain21, instability21, and subluxation of the distal radio-ulnar joint21,22. In one series, median and ulnar neuropathies developed in two patients nearly two decades after the initial injury10.
The operative procedures for the treatment of chronic Monteggia lesions have included closed16,27,29 or open1,3,4,6-9,11,13,15,16,18-22,24,27-32 reduction of the radial head and repair7,12,16,18 or reconstruction1,3,4,6,7,9,11,12,16,17,19,21,24,27-32 of the annular ligament. Repair of the ligament has been performed with fascia lata1,28, palmaris longus20,32 or peroneus tertius tendon32, or deep fascia of the forearm6,7,12,20. Bell Tawse3 used the triceps tendon to repair the annular ligament in eight patients, six of whom had a chronic Monteggia lesion. This procedure was modified by several authors9,17,21,24,27 and was disparaged by others29, yet it remains the most common method of reconstruction of the annular ligament3,9,16,17,19,20,24,27. Fixation with wire has been recommended to maintain reduction of the radiocapitellar joint8,9,11,12,16,17,21,27; however, the problems of breakage and migration of the wire have led many to abandon this technique19,20.
Osteotomy of the ulna, either with or without repair of the annular ligament, is performed in order to correct malunion and in order to facilitate the reduction of the radial head4,7,9,11-13,15,18,19,21,22,27,29,30. Osteotomy of the radius has also been performed13,27,29. Excision of the radial head has been performed in adults, but it is not recommended for the treatment of children5,11,25,27.
Each of the studies reported in the literature have stressed the need for intervention when a patient has a chronic Monteggia lesion; however, there has been little clinical or radiographic documentation of the results. The treatment of nearly 100 lesions has been reported1,3,4,6-9,11-13,15-22,24,26-33, but none of the authors emphasized the difficulties associated with operative intervention. The largest series29, published only in abstract form, reported improvement in fourteen of nineteen patients, but all of the patients had a poor range of motion on the injured side. The few reports in which the range of motion was noted described a marked decrease in rotation, especially pronation, of the forearm3,9,12,15,21,24,27,29-31. Our patients all had decreased rotation of the forearm. Some reports have noted slight restriction of motion of the elbow but have not described a major loss of rotation3,12,16,21,24,27.
In the present study, there were complications that were more serious than loss of rotation. Transient motor and sensory ulnar-nerve palsies developed in three patients who had residual weakness at the latest follow-up examination. Only one other report mentioned a neurological complication, a transient tourniquet palsy12. There was one report of posterior-interosseus-nerve palsy preoperatively15, and one of our patients had such a palsy on the initial examination and had a partial laceration of the entrapped nerve during the operative exposure. In response to this finding, we now expose the radial nerve and decompress the ulnar nerve as part of the reconstruction and, additionally, we now routinely perform a volar and dorsal fasciotomy of the forearm.
Several of the complications in the present series were related to the ulnar osteotomy. We now use plate fixation and autogenous corticocancellous bone graft to provide rigid fixation and to facilitate osseous union of the osteotomy site. Hirayama et al.15 reported that the hardware failed in two of their patients. Despite difficulty in maintaining adequate alignment of the fragments after an osteotomy, Best4 did not recommend fixation of the ulna and advocated transcapitellar fixation of the radiocapitellar joint. Since one of our patients (Case 4) had non-union of the ulna and since Best had better results, his recommendations may be worth considering.
Three of our patients had recurrent subluxation of the radiocapitellar joint (Fig. 1). Two patients had hourglass constriction of the radial neck (Fig. 2-A). A partial cartilaginous radio-ulnar synostosis may have developed in one patient (Fig. 2-B). All of these complications have been reported previously12,15,16,19,21,24,27. Although the radiographic appearance improved, only four of our seven patients had concentric reduction of the radiocapitellar joint. To our knowledge, no one has previously noted the consistent enlargement of the radial head and the tendency for early closure of the proximal radial physis reported in this study. Whether these findings will affect the outcome in a skeletally immature patient is still unknown. In addition, the long-term sequelae of recurrent subluxation are unknown.
Stelling and Cote26 recommended that chronically dislocated radial heads were better left unreduced because the final result was "generally better than ... surgical reconstruction." However, we believe that the long-term sequelae of chronic Monteggia injuries3,10,21,22 warrant intervention in skeletally immature patients.
Our currently preferred method is open reduction through an extended Kocher approach. The radial nerve is identified and protected throughout its course, and the ulnar nerve is decompressed from proximal to the cubital tunnel through the flexor carpi ulnaris muscle. This exposure of the nerves is performed in an attempt to prevent neurological complications. If malunion of the ulna prevents reduction, an osteotomy should be performed. The osteotomy site should be fixed with a plate, and autogenous corticocancellous bone graft should be used. Repair or reconstruction of the annular ligament is performed depending on the stability of the reduction after the osteotomy and on the residual ligament. Our preference is the triceps fascia reconstruction described by Bell Tawse3. Before closure, we release the volar and dorsal fascia to lessen the risk of postoperative compartment syndrome. The extremity is immobilized in an above-the-elbow cast with the elbow in at least 90 degrees of flexion and full supination until radiographic union of the osteotomy site is achieved. Our protocol and results differ from those in two recent reports4,24. A prospective study of treatment options would help to determine the optimum type of intervention; however, such an investigation is unlikely to be done because of the rarity of the injury.
Our experience underscores the complex problem confronting a surgeon attempting to treat a chronic Monteggia lesion in a child. The patient and his or her family should be informed of all of the risks and complications. It must be emphasized that the operations for the treatment of chronic Monteggia lesions are salvage procedures; the injury is severe and in all likelihood the patient will be left with some residual functional limitation. The best treatment for this problem remains preventive. Proper treatment of an acute Monteggia fracture-dislocation is much simpler and more successful than any reconstructive procedure.