A ten-month-old boy was seen for a painless swelling of the left forearm that was not tender to palpation. He had been born to nonconsanguineous parents after a normal pregnancy and delivery. The family history was unremarkable, and the patient had a healthy older brother. After the mother had noticed that the patient had difficulty with swallowing, a barium study showed abnormal swallowing movements that may have resulted in bronchopulmonary aspiration. This in turn led to upper respiratory infection, and the infant was hospitalized twice, at the age of eight months and again at the age of nine months. The possibility of a neurological disorder was considered because of mild developmental delay.
When the patient was seen, at the age of ten months, because of swelling of the left forearm, there was no history of trauma or evidence of child abuse. The skin was intact, and radiographs showed a minimally displaced fracture of the shaft of the ulna (Figs. 1-A and 1-B). There were no signs of injury of the radiocapitellar or proximal radio-ulnar joint and no joint instability or subcutaneous hematoma. There were no associated osseous injuries. There was evidence of anhidrosis, normal lacrimation, hypoactive deep-tendon reflexes, and multiple well healed wounds involving the fingertips, tongue, lips, and cheeks. The patient appeared unable to discern hot from cold. There was no evidence of neurofibromas or café-au-lait spots.
The complete blood-cell count; the sedimentation rate; the levels of electrolytes, calcium, phosphate, glucose, and alkaline phosphatase; and the liver, kidney, and thyroid function tests were normal. The Treponema pallidum hemagglutination test for the child and his mother was negative, ruling out congenital syphilis. The results of an ultrasound of the abdomen were normal. Magnetic resonance imaging of the forearm showed a tumor without involvement of soft tissue or other evidence of malignancy. Bone-scanning revealed increased uptake at the site of the fracture. The results of electromyographic and nerve-conduction studies were normal. A diagnosis of congenital insensitivity to pain was made on the basis of a finding of anhidrosis and indifference to pain.
The extremity was placed in an above-the-elbow cast. Three weeks later, the site of the fracture was still mobile and immobilization was continued. An open biopsy was performed eight weeks after presentation. Histopathological studies showed fibrous tissue and chondrocytes compatible with a callus and no evidence of a benign or malignant tumor.
Radiographs made six weeks and eleven weeks after the initial visit showed resorption at the site of the fracture and massive periosteal reaction involving the ulna and, to a lesser degree, the radius (Figs. 1-C and 1-D). At eleven weeks, open reduction and internal fixation was performed with use of a five-hole 3.5-millimeter limited contact-dynamic compression plate (Stratec Medical, Oberdorf, Switzerland) through a posterior approach (Figs. 2-A and 2-B). The results of intraoperative cultures were negative for bacteria and fungi. The area of non-union was not resected, and no additional procedures were performed. The fracture was united eight weeks later (Figs. 2-C and 2-D), and the plate was removed four months after that. At the latest follow-up examination, when the child was twenty-eight months old, there was a symmetrical range of motion of the upper extremities.
Congenital insensitivity to pain with anhidrosis is a rare, autosomal recessive disorder of the peripheral nerves. We found reports of thirty-three patients who had this disorder3,9. The anhidrosis is believed to produce frequent hyperpyrexia. Related orthopaedic problems begin early in childhood and include self-mutilation of the fingertips, poor healing of fractures, Charcot-type destruction of joints, and osteomyelitis. Closed treatment of fractures in these children is frequently unsuccessful, and amputation of the extremity has been performed to treat persistent non-union8. The small size of the extremity and the abundance of subcutaneous adipose tissue make it difficult to apply a cast that can rigidly immobilize the fracture site. The inadequate immobilization leads to poor fracture-healing or Charcot-type destruction. In addition, these patients do not restrict the movements of the involved extremity as they lack the inhibitory pain reflex. Despite a normal sense of touch9, the indifference to pain makes these children extremely susceptible to breakdown of the skin over osseous prominences, leading to infection and osteomyelitis8.
Non-union of a fracture of a long bone is extremely rare in children. Lewallen and Peterson reported non-union of thirty diaphyseal fractures in thirty children. Twenty-eight of the children were between ten and sixteen years old; one was eight years and nine months old; and one, four years old. The authors did not mention the presence of any underlying neuropathy. They noted the occurrence of non-union after high-energy trauma, particularly when the fracture was open and there was severe loss of soft tissue or infection. Patients who had congenital pseudarthrosis, non-union after an osteotomy, bone tumors, acro-osteolysis, or hemangioma were excluded from the study. Open reduction, repeated manipulation, and inadequate internal fixation may also result in non-union of fractures of long bones in children.
In infants, non-union of the ulna is almost always associated with a pathological process such as neurofibromatosis1. In order to achieve osseous union, the area of non-union and affected tissue must be excised and treated with bone-grafting7. The differential diagnosis includes non-union secondary to osteomyelitis, congenital syphilis, or a benign or malignant tumor. In addition, battered-child syndrome must be considered and carefully ruled out whenever a patient has an unusual fracture with an unusual course at an unusual age.
Patients who have congenital insensitivity to pain do not have any intrinsic abnormality of the bone; the non-union is a result of the response to the injury. Because of the excessive periosteal reaction in our patient, an underlying bleeding disorder was suspected, but this was ruled out by coagulation studies. The excessive periosteal reaction was similar to the appearance of hypertrophic non-union and was interpreted as evidence of mechanical instability. This reaction is in contrast to the atrophic non-union associated with neurofibromatosis2. In adults, a hypertrophic non-union secondary to inadequate immobilization of a fracture resolves with appropriate immobilization, usually in the form of internal fixation4.
We had no guidelines for treatment as hypertrophic non-union in an infant does not appear to have been previously reported. In our patient, the bone healed rapidly with internal fixation. We chose to use a small-fragment 3.5-millimeter limited contact-dynamic compression plate instead of a one-third tubular plate because treatment with a cast had already proved unsuccessful, and we therefore believed it was safer and more appropriate to use an implant with better fatigue strength and higher bending and torque stiffness. It is true that the small-fragment limited contact-dynamic compression plate is thicker than the one-third tubular plate and there could have been problems with soft-tissue coverage over the plate. However, we were able to place the largest part of the plate within, not over, the callus after creating a slot in the callus. In order to maintain the length of the ulna, we did not resect the area of non-union.
The result in our patient encourages us to suggest that fractures in children who have congenital insensitivity to pain should be treated initially with non-operative methods and then, if the fracture does not heal in a timely fashion and there are signs of hypertrophic non-union, with internal fixation. Internal fixation should be performed before problems with the skin develop as a result of prolonged immobilization in the cast. The implants can be removed after the fracture has healed.