Parents are concerned with both the cosmetic appearance of the bowing, in-toeing deformity, and the associated problems of excessive tripping. A family history of bowing is common (45, 50). Examination reveals bowing of both the lower extremities with in-toeing (Fig. 27-17). Although the bowing deformity is bilateral, it may be asymmetric. Despite their bowlegged, toed-in gait, these young children are characteristically very agile walkers. The child should be observed walking, both toward and away from the examiner, to assess the severity of the dynamic varus and associated internal rotation deformity. Internal rotation of the extremity permits contact of the foot with the floor as the child stands. This also maintains the body center over the midline as the child walks. The presence of a lateral thrust should be noted. This is a brief,
dynamic, lateral, knee movement that occurs during single-leg stance. It can represent a lateral subluxation or shifting of the femur on the depressed medial tibia or lateral ligamentous laxity. A lateral thrust is characteristic of pathologic bowing. The amount of varus deformity of the lower extremity can be measured with a goniometer, with the knees extended and patellae facing forward. The distance between the knees when the patient stands with the ankles together is noted. Photographs are helpful in documenting the deformity. Angular and rotational alignment and joint range of motion of the entire lower extremity are assessed. Medial-lateral knee joint laxity is not usually present in physiologic bowing.

In younger children, those <18 months of age, radiographs can document the degree and location of the varus deformity, but usually will not distinguish between physiologic bowing and early Blount disease. Radiographs are an essential part of the evaluation in the older child (over 18 months of age). Radiographs are also indicated for those with more pronounced deformities (clinical femoral-tibial angle >20 degrees), when a lateral thrust is observed, if the child is of short stature (below fifth percentile), or if a metabolic bone disease is suspected (Fig. 27-18). The radiograph is taken with the patient standing, if possible, and the patellae pointing straight ahead. The relative degree of varus deformity is noted by observing the shaft-to-shaft angle of the femur and tibia (51, 52). More importantly, the distribution of bowing deformity is noted (53). When physiologic, the bowing occurs throughout the distal femur, proximal tibia, and distal tibia. In contrast, in early Blount disease, the varus deformity is more focally limited to the proximal tibia.

The differential diagnosis of bowing in the young child includes physiologic bowing and pathologic bowing (Table 27.1). Pathologic bowing occurs in infantile tibia vara (Blount disease), metabolic bone disease, skeletal dysplasias (Fig. 27-18B-D), and focal fibrocartilaginous dysplasia. The clinical and radiographic features

associated with metabolic bone disease or skeletal dysplasia readily differentiate these pathologic conditions from physiologic bowing. Most often, the necessary differentiation is between physiologic bowing and infantile Blount disease. If the bowing is physiologic, the entire lower extremity will appear to be bowed. If the varus results from a relatively greater deformity of the proximal tibia, infantile tibia vara or Blount disease may be present (54, 55 and 56). Children with either physiologic bowing or Blount disease usually are early walkers and typically present for evaluation at 15 to 18 months of age. Often, there is a positive family history of bowing deformity (in siblings, parents, uncles, and aunts) (45, 54). Physiologic bowing and Blount disease are two points within the same spectrum, with Blount disease being the pathologic result of unresolved infantile bowing (51, 54). Frequently, one extremity will have physiologic bowing, with Blount disease affecting the contralateral tibia.

Radiographic distinction between physiologic bowing and Blount disease is not obvious in very young children. The Langenskiöld changes, diagnostic of Blount disease, are not always distinct before 2 to 3 years of age (47, 50, 51, 53, 56). Measuring the metaphyseal-diaphyseal (MD) angle of both the proximal tibia and distal femur helps to identify the location and relative severity of varus deformity (51, 55, 56). Although an absolute MD angle is not diagnostic, it does serve as an early guide in differentiating Langenskiöld stage I infantile Blount disease from physiologic bowing (44, 49, 50, 53) (Fig. 27-19). Measurement can be affected by limb position (56). A study of the proximal tibial MD angle in patients with bowing (physiologic bowing or Blount disease) identified two distinct populations with considerable overlap (54, 55) (Fig. 27-20). On the basis of this study by Feldman and Schoenecker, when the MD angle is 10 degrees or less, there is a 95% probability that the diagnosis is physiologic bowing. Conversely, if the MD angle is 16 degrees or more, there is a 95% probability that Blount disease is present. Bowen et al. (52) noted that all children with a tibial MD angle >16 degrees showed progression of the varus deformity. For those patients with an MD angle between 10 and 16 degrees, follow-up for at least 1 to 2 years is necessary to determine whether the bowing resolves (physiologic bowing) or progresses (Blount disease).

The ratio of the distal femoral MD angle relative to that of the proximal tibia can be helpful in differentiating physiologic bowing from early Blount disease. This is measured on the AP lower extremity x-ray film by constructing a distal femoral MD angle similar to that measured in the proximal tibia. This angle represents the contribution of varus in the distal femur to the overall measure of femoral-tibial varus in the limb. The distal femoral MD angle is divided by the proximal tibial MD angle. This quotient represents the proportion of varus found in the femur relative to the tibia. A ratio of >1 suggests that the bowing is physiologic, that is the femur contributes as much as the proximal tibia to the varus, and resolution is expected to occur (55). A ratio of <1 indicates that the bowing is predominantly within the tibia and is more likely to evolve into Blount disease.

Bowing can be associated with metabolic bone disease or skeletal dysplasia (Fig. 27-18C,D). Rickets [usually X-linked hypophosphatemic (XLH) rickets] is the most common metabolic bone disease to present as a bowed leg deformity in a toddler (57, 58). Nutritional rickets should be suspected if the child was breast-fed and did not receive supplemental vitamin D. Infants presenting with rickets are short; the measured
height is often below the tenth percentile. Radiographs of infants with bowing deformity secondary to rickets should not be mistaken for physiologic bowing. In patients with XLH, the physis appears abnormally wide, and the metaphysis is flared and curves like a trumpet around the physis. Similar changes are found within all growth plates. Bone density will be diminished overall, with thinning of the diaphyseal and metaphyseal cortices. The severity of changes in bone morphology and the degree of osteomalacia is variable. The diagnosis of XLH is made by analyzing calcium and phosphate in serum and in urine (57, 58). Patients suspected of having rickets should be referred to an endocrinologist for a thorough metabolic workup.

The child who presents with bowing deformity in association with a skeletal dysplasia will be short, typically below the 5th percentile. The radiographic changes for each skeletal dysplasia varies with the site of involvement which may be principally epiphyseal, physeal, metaphyseal, or diaphyseal or involve multiple sites and may also involve the spine. Achondroplasia, the most common of the skeletal dysplasias, typically presents with bowing deformity. On a standing AP x-ray, characteristic findings include a knee centered varus deformity with an elongated fibula. Patients with pseudoachondroplasia may present with a varus deformity in association with ligamentous laxity. Metaphyseal chondrodysplasia (Schmid’s or McKusick’s type) typically presents with persistent bowing and short stature in an otherwise normal-appearing child (Fig. 27-16D). While the radiographic changes about the physis and metaphysis include flaring and widening of the medial metaphysis, the changes are distinct from those of osteomalacia and bone density will appear normal (see Chapter 8).

Focal fibrocartilaginous dysplasia is a very rare, but progressive, unilateral, focal deformity that can occur either in the proximal tibial metaphysis or in the distal femoral metaphysis (59, 60, 61, 62, 63, 64 and 65). The clinical presentation is similar to unilateral infantile Blount disease. The lesion consists of a focus of fibroblastic and cartilaginous tissue, usually below the insertion of the pes anserinus on the tibia (Fig. 27-21). This produces an acute angulation below the metaphysis. The diagnosis is made radiographically. A characteristic indentation is noted in the medial cortex at the junction of the metaphysis and diaphysis, and a focal varus deformity is associated with it. The lesion is radiolucent and often well circumscribed, often with a rim of reactive bone. The physis and epiphysis are normal. An identical process has also been observed in a corresponding location in the distal medial femur (63). Some authors have reported spontaneous improvement in angulation. Simple curettage is recommended for persistent lesions or those in unusual locations (64).

Parents should be informed that spontaneous correction of physiologic bowing is anticipated as predicted by Salenius and Vankka and Sabharwal et al. (1, 46, 47) (Fig. 27-16). Bowing, although present in infants, is often not noticed until the child begins standing. The bowing resolves, typically by 2 years of age, and physiologic valgus develops between 3 and 4 years of age (Fig. 27-22). Nonoperative treatment (orthotics, shoe modification, or nighttime splinting) is both unnecessary and ineffective. In mild deformities, the bowing and its associated internal tibial torsion predictably resolve; follow-up visits may not be necessary. For those patients with more pronounced or persistent deformities, follow-up visits are scheduled at 3- to 6-month intervals. Resolution or progression of the varus that occurs with growth can be documented by subsequent physical examination. Serial photos can be compared with the initial image to determine the degree of resolution or progression of the bowing deformity. Follow-up standing AP radiographs should be obtained if the varus does not appear improved on clinical appearance or by serial goniometric measurement. Uncommonly, varus may persist into late childhood without progressive physeal changes. Varus which does not resolve warrants continued follow-up and may require treatment before skeletal maturity with growth modulation.

Early on, infants with physiologic bowed leg deformity cannot be clearly distinguished from those with infantile Blount disease. Like physiologic bowing, infantile Blount disease is usually bilateral (48, 53, 54). When unilateral infantile Blount disease is noted, often the contralateral extremity is bowed physiologically. These children typically are early walkers (<10 months of age) and often are overweight (>95th percentile). As with physiologic bowing, there may be a family history of bowing deformity.

Cook et al. (71), using finite element analysis, calculated that the compressive force resultant from weight-bearing stresses on a bowed leg was sufficient to produce a disturbance in growth. Obesity increases the potential for growth disturbance. In Blount disease, the focal pathologic changes in the proximal medial tibial growth plate cause varus to progress. This chronic growth disturbance results from disorganization of the physis and the osteochondrosis of the medial proximal tibial physis and adjacent epiphysis and metaphysis as described by Blount (44, 48, 65, 67, 68, 69 and 70, 72). The proximal medial tibia fails to grow normally, and tibia vara of increasing severity develops. This results in shortening of the extremity and, if left untreated, ultimately results in depression of the medial tibial condyle and intra-articular deformity.

In 1952, Langenskiöld identified six distinct radiographic stages of development of the proximal tibia depicting the natural progression of untreated infantile Blount disease (44, 70) (Fig. 27-19). The
stage and the age at which it occurs have prognostic significance (48, 50, 51 and 52, 73). In stages I and II, the irregular metaphyseal ossification changes are often indistinguishable from physiologic bowing. These changes may be reversible. Clear-cut stage I or stage II changes of Blount disease are typically not apparent until the patient is 2 to 21/2 years of age. Stage III shows definite deformity in the proximal tibial physis, often with some fragmentation. Stage IV lesions can be associated with early bar formation across the deformed physis, as it assumes a vertical orientation (44, 70). These physeal bars are often difficult to detect. There is profound disruption of the physeal cartilage and abnormal growth in the adjacent bone as stage V develops, usually in children older than 8 years. Eventually, this process results in severe depression of the articular surface and stage VI disease (74). Consistent positioning of the lower extremities is important to detect subtle changes in the physis. Standing AP radiographs with the patellae facing forward are recommended for serial evaluations.

In North America, advanced Langenskiöld stages often occur at a much younger age than in Finland. All of Langenskiöld’s patients were Caucasians from Finland, whereas a large proportion of patients in the United States are African American. North American children typically experience more rapid progression with more severe, irreversible changes at an earlier age than their European counterparts. Stage IV changes are often seen in 4- to 5-year-old children in the United States compared to 7- to 8-year-old children in Finland. The greater incidence and severity of disease has been attributed to a higher proportion of overweight children in North America (80).

The distal femur in infantile Blount disease is usually normal. Although distal femoral varus does not occur in infantile Blount disease, valgus does occasionally occur in children with more advanced tibial changes. It may be a response to the asymmetric load across the knee, allowing relative overgrowth of the distal medial femur (52, 66). More often, the valgus appearance of the femur is a consequence of severe tibial bowing and relative abduction of the hip, which creates the illusion of valgus at the distal femur.

Brace treatment should be considered either in all patients <3 years of age with unilateral Blount disease changes (Langenskiöld stages I, II) or in patients older than 18 months who have persistent bowing and risk factors for Blount disease. A tibial MD angle of >16 degrees is a radiographic sign of significant risk for Blount disease (49, 54, 55). A sign of relative risk is an MD angle between 10 and 16 degrees along with the clinical appearance of a varus deformity or progression of varus. Additional risk factors include obesity, ligamentous instability, or the presence of a lateral thrust, any of which may potentiate a varus deformity (51, 54). Improvement in the tibial MD angle should be apparent within 12 months of brace treatment.

Studies have demonstrated that brace treatment can correct both the varus deformity and the pathologic proximal-medial tibial growth disturbance (75, 76 and 77). In these three reports, 79 extremities were treated using a brace. The best results were obtained with unilateral deformity, where brace treatment was successful in 22 of 23 patients. In contrast, brace treatment was less successful for treating bilateral deformities, with only 18 of 28 patients noted to be successfully treated. Compliance was much more difficult to achieve for the child with bilateral deformity, as is understandable. It is also less effective in obese patients. Bracing should not be initiated after 3 years of age, nor should brace treatment be continued if Langenskiöld stage III changes develop (75, 76 and 77).

We have successfully treated unilateral Blount disease with a modified knee-ankle-foot orthosis (KAFO). The knee is held in extension with a single medial upright KAFO. Thigh and calf cuffs and a varus-correcting lateral knee pad provide three-point fixation. It is worn for 23 of 24 hours (50, 51, 75, 76 and 77) (Fig. 27.23). The locked KAFO counteracts the pathologic medial compressive forces, allowing resumption of more normal growth and correction of the genu varum. In those patients who are compliant, the clinical appearance of bowing typically improves over the ensuing months; however, the pathologic radiographic changes at the proximal medial tibial metaphysis, physis, and epiphysis are slower to remodel. Brace treatment is continued until the bony changes in the proximal medial tibia resolve; typically, this takes 11/2 to 2 years of brace treatment (75, 76 and 77). Brace treatment will be successful if by 4 years of age the mechanical axis of the lower extremity passes through the center of the knee and the depression of the medial epiphysis resolves. The radiographic appearance of the medial epiphysis and metaphysis should normalize by 5 years of age. If satisfactory correction does not occur, surgical correction should be recommended.

Children older than 21/2 to 3 years with worsening stage II or evolving stage III Blount disease, who are either noncompliant or not good candidates for brace treatment because of obesity or bilateral involvement, should be treated surgically with proximal tibia and fibular varus-correcting osteotomy. Alternatively, surgical treatment with growth modulation has been shown to be effective in selected cases. Early surgery to realign the leg by 4 to 5 years of age is necessary to prevent progression to stage IV disease, when the early formation of a physeal bar can occur. Surgical treatment unloads the medial compartment of the knee and facilitates the growth of the proximal medial physis. Restoration of normal growth in the medial tibial physis is less likely to occur if surgery is delayed such as until 5 years of age (78).

Young patients and/or those with less than stage IV Langenskiold changes may be considered for growth modulation treatment (79). Growth modulation is best accomplished in this age patient with the placement of a small-fragment (typically two-hole) plate extraperiosteally across the proximal lateral tibial physis. The plate is secured by cortical screws. This construct impedes lateral physeal growth. Over time, correction will occur. Frequent follow-up visits are essential to monitor correction. This is assessed by measuring the mechanical axis of the involved lower extremity. With correction of the varus deformity, the mechanical axis will shift from its previous location medial to the center of the knee to a more lateral position. A mechanical axis slightly lateral to the center of the knee joint is desirable. The medial growth plate often

continues to grow more slowly than the lateral growth plate. Slight overcorrection is desirable to compensate for the often occurring and variable differential growth of the proximal tibial physis. Continued follow-up is mandatory to assure that correction of varus is maintained and/or to address recurrent deformity. Repeat application of the plate may be needed to correct mild recurrent deformity.

Osteotomy is preferred for more severe deformity or if close, reliable follow up cannot be assured (Figs. 27-24, 27-25, 27-26, 27-27, 27-28, 27-29, 27-30, 27-31 and 27-32). The proximal tibial osteotomy is performed with attention to both the known inherent risks and the need for obtaining adequate correction of the deformity (78, 80, 81). The fibula is osteotomized through a separate lateral incision, taking care to avoid injury to the deep motor branches of the peroneal nerve (38, 82). The tibial osteotomy can be accomplished in a variety of methods (48, 51, 68, 78, 82, 83, 84, 85 and 86). A straight transverse osteotomy allows for necessary correction of frontal, sagittal, and rotational deformity. The fragments are stabilized with smooth K-wires (Fig. 27-24C,D). Alternatively, Price et al. have effectively utilized monolateral external fixation to stabilize the tibial osteotomy (86). A slight “overcorrection” into valgus with or without translation of the distal fragment laterally is desirable (48, 50, 51, 82). This places the mechanical axis of the leg within the lateral compartment of the knee, unloading the medial proximal tibia. The mechanical axis of the leg can be assessed intraoperatively using the bovie cord. The cord is stretched from the center of the hip and across the center of the ankle with the leg resting on a radiolucent table. The leg should not be held, but simply allowed to rest on the table. The wire within the bovie cord, which is visible on the C-arm, is used to verify the position of the mechanical axis (Fig. 27-24E,F). This simple technique provides a reproducible method to verify that the mechanical axis has actually been transferred lateral to the center of the knee joint (82). Alternatively, an intraoperative AP x-ray film of the entire lower extremity can be taken to assess the correction obtained. If a unilateral deformity is present, clinical comparison to the normal extremity is helpful in determining whether adequate correction has been obtained. A subcutaneous fasciotomy of the anterior compartment is performed prior to wound closure. Suction drains are routinely used.

Postoperatively, the extremity is immobilized in a non-weight-bearing, long-leg, bent-knee cast. Alternatively, a spica cast is used for the child with relatively short, fat extremities. On occasion, a KAFO is used following cast removal if the preoperative deformity was severe and associated with ligamentous laxity. Following corrective osteotomy, the pathologic changes in the proximal medial tibia must be carefully monitored. These bony changes are not always reversible, and further progression in their development may be associated with a recurrence of deformity (44, 67, 69, 70). It is essential to document that valgus alignment has been obtained and maintained by the use of long cassette radiographs (hip to floor) and comparison of serial examinations until skeletal maturity (Fig. 27-24G).

Recurrent or worsening varus deformity with persistent pathologic changes can occur despite restorative osteotomy at a young age (44, 50, 51). The risk for recurrent deformity is present in some children as early as Langenskiöld stage III and is much more common in those with stage IV disease, marked obesity (>95th percentile weight), or ligamentous laxity (48, 50, 51) (Fig. 27-33). If patients present with advanced pathologic changes or with recurrent deformity following brace or surgical treatment, an MRI or CT scan should be utilized to search for the presence of a physeal bar within the distorted proximal medial tibia. Identification of a bar may be difficult because of the serpentine course of this abnormal physis, which takes on a vertical slope as the deformity worsens. Early bridging across the physis typically occurs at the inferior aspect of the vertical limb of the distorted medial physis (Fig. 27-33B).

Progressive varus deformity may occur without an identifiable bony bar (Langenskiöld III or early IV) (48). This occurs rather subtly because of a decreased growth rate of the proximal medial tibial physis relative to the lateral physis and can occur any time prior to skeletal maturity. As the physis transitions from a normal horizontal orientation to a more vertical position, this differential growth becomes more evident. Careful observation is needed to detect this change early in the course of treatment. Premature closure of the medial tibial physis frequently occurs because of persistent deficient proximal medial physeal growth. If this occurs, the varus deformity can often be controlled with growth modulation, using an eight-plate or hemiepiphyseal staples (48, 79, 90, 91). If additional tibial physeal growth is anticipated, the internal fixation is removed after a slight overcorrection is obtained. Careful follow-up including radiographic assessment of recurrent varus deformity and/or leg-length inequality are mandatory until skeletal maturity is reached.