Radiation therapy can produce osteopenia in children. More information is needed to elucidate all the parameters involved, but the risk is probably between 8 % and 23 % at doses above 20 Gy, primarily seen in patients with cranial irradiation and could be attributed to growth hormone deficit. However, bone mineral loss up to 7–8 % can be detectable after 30 weeks following 20–25 Gy irradiation in rat femurs. The relationship between osteopenia and pathologic fracture remains unclear [43].

The physical effects of radiation therapy are secondary to damage to the chondroblasts. This damage results in a slowing of, or arrested, physical growth, producing abnormal development of the involved bone, giving rise to (1) spinal abnormalities, (2) leg or arm length discrepancies, (3) angular deformities at joints, (4) slipped capitofemoral epiphyses, and (5) osteocartilaginous exostoses.

Spinal abnormalities most commonly include decreased stature or scoliosis, although lordosis or kyphosis can also occur [12, 67]. Except for decreased growth, severe spinal complications are less common after megavoltage RT, compared with the orthovoltage era, if the entire vertebral body is within the treatment field [83]. Although curvature can occur, it is usually a result of tethering that occurs with decreased muscle development as a result of irradiating just one side of the abdomen, as is often done in Wilms’ tumor.

On standard radiographs, vertebral bodies show subcortical lucent zones within 9–12 months after completing RT. The subcortical lucent zones progress over the next 1–2 years to form growth arrest lines, which parallel the epiphysis of the vertebral body [91]. Following the development of these arrest lines, little or no further growth occurs. The higher the dose, the quicker the growth arrest occurs. Other changes may also be seen on radiographic examination. During the first few months after irradiation, vertebral bodies may have a more bulbous contour [49]. Scalloping of the physical cartilage plates may also be observed, and not infrequently, the final appearance is that of rounded vertebral bodies with central beaking [91].

Clinically, the deleterious effect of radiation on growing bone has been known for a long time. Neuhauser and associates reported on 24 children who received orthovoltage irradiation to the spine during their treatment for Wilms’ tumor or neuroblastoma [67]. Doses below 10 Gy (using orthovoltage irradiation) caused no detectable vertebral abnormality regardless of the age at treatment. Higher doses caused severe late effects in the spine. Mayfield and coworkers reported on 28 children with neuroblastoma who received irradiation to their spine [60]. Scoliosis was the most common sequela and occurred most severely at doses above 30 Gy with orthovoltage irradiation. Megavoltage irradiation appears to cause fewer severe orthopedic complications. Rate et al. reported on a series of 31 Wilms’ tumor patients irradiated between 1970 and 1984. Three out of ten patients treated with orthovoltage irradiation developed late orthopedic abnormalities requiring intervention [82]. None of the 21 patients receiving megavoltage treatment developed such complications, even at doses of 4,000 cGy. Scoliosis was related to a higher median dose (2,890 cGy) and a larger field size (150 cm²); this sequela did not result for patients receiving smaller doses to a smaller field size (2,580 cGy–120 cm²). Paulino et al. report a lower incidence of scoliosis with doses <2,400 cGy than with higher doses [77]. Both orthovoltage and megavoltage irradiation were used. Probert and Parker were the first investigators to attempt to evaluate bone growth alterations secondary to megavoltage irradiation. They noted deficits sitting height for patients who received radiation to the spine for medulloblastoma, leukemia, and Hodgkin’s disease [79]. Doses greater than 35 Gy were found to produce a significantly greater deficit in sitting height, compared with doses less than 25 Gy.

Silber and coworkers, using data from 36 children whose spine or pelvis was irradiated, have developed a mathematical model for predicting stature loss [97]. The model is based on the radiation dose in Gy, the location of the therapy (including whether or not the capitofemoral epiphysis was treated), gender, and the ideal adult height. Figure 16.5 shows an example of the model, based on four different irradiation doses. The model agrees closely with a report by Hogeboom et al. who calculated the height deficit after flank RT for Wilms’ tumor patients at different ages and doses of radiation therapy [41]. In both reports, even a dose of 10 Gy at age 2 produced a height deficit of 2.3–2.4 cm by age 18 years. For infants less than 12 months of age receiving 10 Gy or more, the estimated height deficit was 7.0 cm. Although chemotherapy did not appear to increase the effect of radiation therapy in Hogeboom’s study, there was a deficit associated with doxorubicin that – although it could be explained, at least partially, by other factors – does not completely rule out a permanent doxorubicin effect on bone growth.

Laminectomy is required in selected cases of children presenting with intraspinal or paraspinal tumors. Children undergoing laminectomy at a younger age are prone to development of kyphoscoliosis [20, 42]. Deformity risk increases significantly when postoperative radiotherapy is added [21, 48].

Irradiation of the extremities or hip usually produces more long-term symptomatic sequelae than irradiation of the spine, particularly when an epiphysis is treated. Radiographic changes first reveal metaphyseal irregularities and epiphyseal widening. Later changes include sclerosis around the physis and eventually sclerosis and closure of the epiphyseal plate. Such premature closure can cause a discrepancy in ultimate length between the irradiated and unirradiated extremity. The degree of growth disturbance in long bones is dose and volume dependent; moreover, patient age plays a very important role [35]. In addition, if the entire physis is not included in the radiation port, then juxta-articular angular deformities can result [84]. It is possible to predict the effects of radiation on growth based on age and growth pattern. Altogether, about 65 % of leg growth occurs at the knee, 37 % from the distal femoral physis, and 28 % from the proximal tibial physis. Only 15 % occurs at the proximal femoral growth plate and 20 % from the distal tibial plate [5, 36]. This estimate may not be completely accurate because it assumes that no growth occurs after irradiation of the growth plate, but in reality, growth continues for at least a short period of time. Additionally, untreated physis in the same extremity can partially compensate for the affected one. Asymmetric dose delivery to the metaphysis can cause valgus and varus deformities [31].

Surgery may also have a significant effect on limb function. Limb preservation with expandable prosthesis has become a widely adopted method in the local management of bone malignancies in children [2, 10]. Repeat procedures for prosthesis, either lengthening or replacement, is often required [26], which potentially adversely affects the physical and emotional well-being of children [39, 102]. Rotationplasty is another durable reconstructive procedure which is associated with significant limb length discrepancy requiring prosthetics. Despite its complexity, children receiving this treatment appear to have reasonable physical and emotional resilience [32, 34]. Amputation is the most radical surgical procedure associated with the loss of a significant portion of the limb and can only be compensated with prosthetics. Despite significant differences between limb sparing, rotationplasty, and amputation, data suggests that survivors have fairly comparable physical and emotional outcomes with either approach [44].

Microcephaly, micrognathia, and other growth abnormalities are reported in children with retinoblastomas and rhabdomyosarcomas receiving external beam radiation to the craniofacial region [24, 62, 81].

Surgical management of tumors in the head and neck of young children may be associated with even more substantial cosmetic and functional disturbances. Given the anatomical challenges of these tumors, there is often concern about leaving microscopic or gross disease behind; thus, adjuvant radiotherapy is given [18]. This combined approach can result in more significant dysfunctions due to worse fibrosis developing over time. Reconstructive surgeries can correct these problems to an extent, but long-term outcomes are still far from satisfactory [16]. Anthropometric and descriptive methods are commonly used for documenting the extent of deformities, but they are hard to quantify. Severe hypoplasia or asymmetry of tissues causing cosmetic and functional problems may require reconstructive surgery.

Radiation therapy has been shown to increase the risk for slipped capital femoral epiphysis due to epiphyseal plate injury [9, 108, 112]. Patients with slipped capitofemoral epiphyses present with pain in the hip or knee (referred from the hip). The pain presents through either acute or chronic onset, since slippage of the capitofemoral epiphysis often proceeds slowly. Wolf and associates first reported cases of slipped capitofemoral epiphysis as a result of childhood irradiation [112]. These occurred 1–6 years after therapy with doses of 28.5–54 Gy and usually after the onset of puberty when there is a change in the angle of the femoral shaft in relation to the femoral neck and head (a change that increases the susceptibility to stress from excess weight and other factors). Children developing a slipped epiphysis after irradiation were generally 2–3 years younger than the average patient with idiopathic-slipped capitofemoral epiphysis, had twice the risk of bilateral involvement (20–50 %) if both proximal femurs were exposed to radiation, and, furthermore, did not usually fit the generally obese body habitus of the average patient presenting with the idiopathic variety. Paulino et al. reported on four infants (<6 months old) irradiated to the hip to doses of 20 Gy or higher for a neuroblastoma [75]. Two patients developed slipped capitofemoral epiphysis, at 25.5 Gy and 36 Gy. The other two patients received doses of 20 Gy and did not develop the problem. Slippage of the physis is thought to occur as a result of excess stress, either from obesity or from a weakening of the bone and physis secondary to radiation therapy. The threshold dose is thought to be around 25 Gy.

The use of chemotherapy has caused a reduction in the shear strength of the physis in an animal model [104]. Theoretically, chemotherapy would also appear to predispose patients to developing slipped capital femoral epiphysis, but this has not been studied in humans.

Osteocartilaginous exostoses are benign outgrowths of the physis. They have been reported to occur in up to 18 % of children treated with radiation therapy [3], although the incidence in the megavoltage era is much lower. The etiology is unknown, but is thought to be due to an injury to the periphery of the growth plate. The lesion is a combination of a radiolucent cartilaginous cap with areas of ossification and calcification. It has a typical cauliflower appearance and the base may be narrow or broad. On physical examination, a hard mass is palpable adjacent to a joint [94]. Lesions may continue to grow slowly throughout childhood, although by the third decade growth should cease.

There is a small incidence of malignant degeneration that presents with pain and occurs after skeletal maturation.

Irradiation of the metaphyseal portion of the bone may temporarily cause increased porosity and demineralization; radiographic examination will show cortical thinning and irregularities during this time. The irradiated segment of bone therefore acts as a stress concentrator and predisposes the bone to fracture [38]. This effect is particularly true in patients in whom the irradiated bone has been weakened by tumor involvement, such as in Ewing’s sarcoma of the humerus and femur. This effect also occurs if the cortex of the bone has been biopsied, particularly in the area of the femoral neck [107]. Pathologic fractures usually occur within 3 years after treatment [38]. The most common site is the upper third of the femur, and since fractures can be caused by tumor recurrence, a full reevaluation of the patient should be performed to ensure that it is not related to relapse [107].

Avascular necrosis (AVN) of the femoral head has been reported as a complication of hip irradiation in children who have received doses exceeding 30 Gy [53]. Older age (usually >10 years old) appears to increase the risk, indicating that the rapidly growing and maturing bones of adolescents are more susceptible to the development of AVN [59]. High cumulative corticosteroid doses administered during the therapy strongly predispose patients to AVN, and dexamethasone seems to be more toxic than prednisone. AVN may be asymptomatic or result in joint swelling, pain, limited range of motion, and even joint damage and articular collapse. MRI is the best technique to detect early stages of AVN. The typical radiographic picture is shown in Fig. 16.6a–d.

Osteoradionecrosis (ORN) of the mandible is a rare complication of head-and-neck cancer therapy, and its incidence is quite variable in the literature, ranging from 0.9 % to 35 % [100, 105].