Proximal Femoral Focal Deficiency

Generally, in individuals with PFFD, the proximal femur is partially absent, and the entire limb is shortened overall. A few main biomechanical abnormalities are present in children with PFFD, as well as in adults with limb deficiencies. These include limb-length discrepancies, malrotation, proximal joint instability, and inadequacy of the proximal musculature.

Vascular changes occur. Chomiak et al, using computed tomography (CT) angiography (CTA) in 21 patients to identify vascular changes associated with PFFD, found that in patients with Pappas type I-IV PFFD, the external iliac, femoral, and deep femoral arteries were substantially reduced in length and diameter, and the deep femoral artery arose more proximally than that in the contralateral extremity.[5]

In addition, two patients with type III disease in this study had an atypical anatomy of the vessels: the anterior part of the thigh and the pseudarthrosis were supplied through the femoral artery (the external iliac artery) as a terminal branch, whereas the remainder of the extremity was supplied from the internal iliac artery.[5]

Ligamentous changes also occur. In a study that used knee arthroscopy to identify changes in cruciate ligaments and their relation to the different types of PFFD in patients with Pappas type III, IV, VII, VIII, or IX deficiency, Chomiak et al found variable changes of the cruciate ligaments in all but one patient. Although these changes were not clinically relevant in most of the patients and were not related to the Pappas classification, the authors recommended imaging of cruciate ligaments before lengthening of the extremity in order to avoid knee dislocation.[6]

The etiology of PFFD is not known exactly, but certain theories have been proposed and certain agents implicated. Sclerotome subtraction is one such theory that has been offered to explain several different limb deficiencies. Specifically, this theory states that injury to the neural crest cells that form the precursors to the peripheral sensory nerves of L4 and L5 results in PFFD.[7]

Another theory, advanced by Boden et al, is that PFFD may result from a defect in proliferation and maturation of chondrocytes in the proximal growth plate.[8] Agents implicated in causing such injuries include anoxia, ischemia, irradiation, bacterial and viral infections and toxins, hormones, mechanical energy, and thermal injury.[7, 9] Thalidomide, when taken by the mother between the fourth and sixth weeks of gestation, has been shown to be a definite cause of PFFD in humans.[9] Evidence indicative of a genetic etiology has not been reported.[1, 10]

The incidence of PFFD is on the order of 0.5-2 per 100,000 live births. PFFD is the third most common longitudinal deficiency of the lower limb.[11]  It is predominantly (85-90%) unilateral,[12] with bilateral involvement noted in only 10-15% of cases.

Parents should receive a detailed explanation of the disease process. Individual parents will have differing ideas regarding the most appropriate goal of and approach to treatment. Some will prefer a one-time surgical procedure at a young age, whereas others will be more inclined to attempt to preserve the natural limb no matter what. Treatment must be individualized on the basis of the following factors[13] :

• Limb-length discrepancy
• Presence of associated malformations
• Adequacy of musculature
• Proximal joint stability

The clinical appearance of proximal femoral focal deficiency (PFFD) is not subtle, and the condition therefore is easily recognized. The femur is shortened, flexed, abducted, and externally rotated.[7, 9, 14] Gillespie noted that in his patients, the hips were never normal and the knees were dysfunctional.[3, 15] Flexion contractures of the hip and knee are also present.

The bulbous proximal thigh quickly tapers to the knee. Because of the short femur and bulbous thigh, examination of the hip can be difficult. As a result of hip instability, pistoning may be present. The knee is uniformly unstable in an anteroposterior plane secondary to absent cruciate ligaments. Additionally, generalized knee hypoplasia has been reported.[3]

A high incidence of fibular deficiency and valgus feet is associated with PFFD.[9] Fibular deficiencies are found in as many as 70-80% of persons with PFFD. Approximately 50% of patients with PFFD have other limb anomalies.[16] However, Aitken reported almost a 70% incidence of other anomalies.[17]  Cleft palate, clubfoot, congenital heart defects, and spinal anomalies, though rare, occur as well. PFFD is bilateral in 10-15% of cases.[10]  Patients with bilateral PFFD can present with upper-extremity involvement.[18]

Establishing an early antenatal diagnosis of proximal femoral focal deficiency (PFFD) facilitates the choice of the optimal management strategy. Three-dimensional (3D) and four-dimensional (4D) ultrasonographic options are helpful in explaining the nature of the syndrome to parents as early as the first trimester. They are also useful in planning future rehabilitation and determining appropriate use of a prosthesis.[11]

Plain radiographs of patients with PFFD show an apparent loss of continuity between the femoral shaft and the head/neck.[19] The fibrous or fibrocartilaginous tissue is responsible for this defect, which can ossify later on. Ultrasonography (US) and magnetic resonance imaging (MRI) are particularly useful for assessing the exact nature of the tissue at the loss of continuity. MRI is also useful for evaluating unossified structures, the hip joint, adjacent soft tissue, the knee, and other malformations of the limb.[20]

The Aitken classification divides PFFD into four categories on the basis of radiographic appearance (see Classification below).[17] It is important to remember that late ossification may occur, whereby the bone may be present but not visualized radiographically. Occasionally, push-pull comparison radiographs, as well as abduction-adduction views, are necessary to distinguish between Aitken class A and class B. Arthrography also can be helpful.[7, 9, 20, 21]

A study by Maldjian et al of the efficacy of magnetic resonance imaging (MRI) in classifying PFFD also compared MRI findings with radiographic classification.[20] The study, which used the Amstutz classification system, found that radiographic evaluation tended to overestimate the degree of deficiency and suggested that MRI was therefore the better modality.

There is a paucity of radiographic images of PFFD in mature skeletons; the appearance of PFFD evolves as the child grows and is different in an adult patient.[22]

US has been employed for antenatal diagnosis of PFFD.[23, 24]

Several classification systems have described congenital anomalies of the femur, but most have been based on radiographic appearances alone. The Amstutz and Pappas classifications provided detailed radiologic descriptions of the various forms of PFFD that these researchers encountered.[25, 26]  Hamanishi described a progressive reduction of the femur, ranging from simple shortening to total absence.[27]  Fixsen and Lloyd-Roberts divided their patients into stable and unstable categories.[28]

Gillespie and Torode reviewed their patients from both a radiographic and, more important, a clinical viewpoint and found that most could be divided into the following two groups:

• Group 1 - Persons with congenital short femurs
• Group 2 - Individuals with true PFFD

These two groups not only differed with respect to clinical and radiographic appearances but also were functionally unique and had different surgical and prosthetic requirements.[3]

The Aitken classification, which has been the most widely used system, divided PFFD into the following four classes on the basis of radiographic appearance[17] :

• Class A - A shortened femur is present proximally, ending at or slightly above the level of the acetabulum; the femoral head is often absent but later ossifies; femoral head presence is indicated by a well-developed acetabulum; additionally, there is a subtrochanteric defect, which eventually ossifies and thereby establishes bony continuity; after ossification, there is usually a residual subtrochanteric varus deformity
• Class B - There is a more severe defect or absence of the proximal femur, and the defect does not heal spontaneously; at skeletal maturity, there is no connection between the femoral head and proximal femur; the end of the proximal femur is above the acetabulum; the femoral head, though present, may have delayed ossification, and there is often a bony tuft on the proximal end of the shaft
• Class C - There is an absent femoral head that does not ossify and a markedly dysplastic acetabulum; the femoral shaft is shorter than in a person with class B, in whom the entire proximal femur, including the trochanters, does not develop
• Class D (the most severe form) - There is a severely shortened shaft, which often has only an irregularly ossified tuft of bone proximal to the distal femoral epiphysis; no acetabulum is present, because the lateral pelvic wall is flat

In a symposium, Gillespie proposed a more functional classification system, in which patients were divided into three treatment groups from a surgical and prosthetic viewpoint.[15]  The groups in this system were as follows:

• Group A - Possible candidates for limb-lengthening; this group included individuals who had congenitally short femurs but clinically stable hips, had no significant knee flexion contractures, and had the ipsilateral foot at or below the level of the middle of the contralateral tibia
• Group B - Patients classified by Aitken as classes A, B, and C and who required prosthetic treatment. Therefore, any surgical procedure is designed to maximize prosthetic function
• Group C - The same patients as Aitken class D, in that they had subtotal absence of the femur; Gillespie also recommended prosthetic treatment for his group C patients, but these patients did not require knee fusions before prosthetic fitting

Management of proximal femoral focal deficiency (PFFD) requires a multidisciplinary team, which includes the pediatric orthopedic surgeon, prosthetists, and physical therapists. No single treatment approach applies to all cases. Each person with PFFD must be assessed individually. 

The primary goal of treatment for children with PFFD is to compensate for the various functional deficits that may be present—more specifically, to enable patients to have as much function and walking ability as possible with the lowest possible consumption of oxygen per meter traveled. Cosmesis is much less of an issue.

The most apparent functional deficit in PFFD is the shortened limb. A less obvious one is the difficulty with hip function and stability. Because of the flexed and externally rotated position of the femur, the knee remains flexed, and the leg and foot are anterior to the body axis. There is a generalized deficiency of the hip musculature, even in patients with stable hips, resulting in a significant lurch to shift the center of gravity in the single-leg stance.[9]

The knee often has varying degrees of instability as well. Foot function generally depends on the severity of any associated leg deficiency, such as fibular hemimelia.

Treatment is planned on the basis of the following:

• Limb-length discrepancy
• Presence of foot and other deformities
• Adequacy and power of musculature
• Proximal joint stability

Nonoperative management options include the following:

• Observation
• Extension prosthesis

Operative management options for ambulation without a prosthesis include the following:

• Limb lengthening with or without contralateral epiphysiodesis

Operative management options for ambulation with a prosthesis include the following:

• Knee arthrodesis with foot ablation
• Femoral-pelvic fusion (Brown procedure)
• Van Nes rotationplasty ^[29]
• Amputation   

Multiple options and variations exist in the treatment of PFFD, probably more than in any other congenital limb deficiency. Most of the treatments should be delayed until the child is older than 3 years; it is difficult to determine the best option before this age. When the time comes for treatment, however, several important decisions must be made. The first such decision is whether limb lengthening is appropriate for the child.

Indications for lengthening include the following:

• Limb with a predicted discrepancy at maturity not exceeding 20 cm
• Hip that is or can be made stable
• Relatively good knee, ankle, and foot

In such an individual, multiple staged lengthenings can be planned that generally do not begin before the child's third birthday and optimally are completed by high-school age. The timing and staging of these procedures depend on physician choice. Griffith et al studied 11 children (12 limbs) who underwent limb lengthenings of the same bone twice.[30]

If the predicted discrepancy is greater than 20 cm, or if for any other reason the child is not suitable for limb lengthening, prostheses should be considered. Thus, a decision must be made about the best approach to fitting a prosthetic device. The patient with PFFD always requires a prosthesis for ambulation; therefore, any procedure that is attempted must be with the intent of improving the function or fit of a prosthesis.[3]

Another consideration is the type of prosthesis a patient will wear. Most procedures result in the use of an above-the-knee prosthesis with a mechanical knee. For example, the combination of a Syme amputation, knee fusion, and possibly an epiphysiodesis at the knee results in an above-the-knee stump. Rotationplasty procedures allow the use of a below-the-knee prosthesis with a biologic knee.[7, 31, 32]

The initial treatment of the child with unilateral PFFD should parallel normal development. This same principle applies to other congenital deficiencies as well. Therefore, regardless of planned future treatments, when the child is ready to stand, the child is fitted with a prosthesis to equalize leg lengths and allow standing and walking.[7]  This prosthesis, often termed a nonconventional or extension prosthesis, is designed to fit the limb without any surgical correction.[9]

The femoral deformity (short, flexed, abducted, and externally rotated) is accommodated in the hip's funnel proximal socket. Knee joints are often absent in these initial prostheses because the limb is often too long to permit the fitting of a prosthetic foot that also has a knee joint. However, having both is preferable if there is sufficient room. This allows the young child to have knee flexion during all developmental phases, so that he or she can half-kneel, squat, pull to stand, and climb on toys and furniture.

The nonconventional prosthesis is effective in permitting ambulation for the young child.[10, 33]  However, as the child grows, this prosthesis poses limitations. Specifically, the continued flexion, abduction, and external rotation of the femoral segment, the anterior limb alignment, the unstable hip, and the flexed knee create a poor lever arm with which to move the prosthesis. Additionally, the foot often lies at the level of the midcalf of the contralateral leg, making placement of the knee joint less than optimal.

Therefore, surgical correction of these problems is designed to make the limb a more efficient lever arm to move the prosthesis. The surgical options include knee arthrodesis, foot amputation, limb rotationplasty, hip reconstruction, limb lengthening, and iliofemoral arthrodesis (see Surgical Therapy).

Treatment of children with bilateral PFFD differs from that of children with unilateral PFFD.[3]  Some have argued against operative treatment on the grounds that most of these defects are Aitken class D defects, so that these children walk on their natural feet.[17]  The main difficulties with these children are limb-length discrepancies and foot deformities.

Surgical releases and orthotic use generally can provide a useful foot. Therefore, in persons with bilateral disease, the feet should be preserved.[9, 17]  Because these children will walk without the use of prostheses for most of their lives, knee fusions are not indicated. Additionally, Van Nes osteotomies should not be performed in patients with bilateral disease.[34, 35, 36]

When present, significant limb-length inequalities pose a difficult dilemma. This is because of the problem of shortening the child more versus the difficulty of lengthening these limbs. In one study of 91 patients with PFFD, 29 had bilateral involvement, and of these, males were affected nearly twice as often as females.[10]

Knee arthrodesis

Knee arthrodesis is performed commonly in children with PFFD. It allows for a longer and more efficient lever arm,[37] as well as a limb that is more easily contained within the prosthesis.[3] The knee should be fused in neutral or slight extension.[33] An additional advantage of knee fusions is that the muscles can act more efficiently across the hip joint.[9]

Generally, within 6 months of knee fusion and prosthetic fitting, the flexion and abduction deformity at the hip joint will correct, thereby aligning the extremity under the body axis.[3, 7] Depending on the femoral length, as well as that of the entire extremity, it is often beneficial to excise one or both of the physes about the knee at the time of fusion. These fusions are performed commonly in children aged 2-3 years.

In their series, Gillespie and Torode noted that most of their patients benefited from knee fusions.[3] The exceptions were those individuals who already had congenital knee fusions and those persons who had extremely short femoral segments (ie, the knee joints were so close to the pelvis that instability and flexion contractures were of little consequence). Traction injuries of the sciatic nerve have been reported after knee fusion due to sudden knee extension.

Foot amputation

Amputation, which often is best performed at the time of knee fusion, acts to shorten the newly created lever arm. This is often desirable because the new lever arm, which is composed of the fused femur and tibia, must be shortened to accommodate a mechanical knee when the child is older. Additionally, with growth, the foot becomes more difficult to accommodate in a cosmetically acceptable prosthesis.[3]

The Syme amputation has been widely performed and provides a superior end-bearing stump for immediate prosthetic fitting.[9] The Boyd amputation, which saves the entire calcaneus, is another option. The calcaneus is fused to the distal tibia, thereby creating a bulbous heel pad that facilitates prosthetic suspension. When only a portion of the calcaneus is retained and fused to the distal tibia, the heel pad remains centered, resisting the tendency for migration.[7]

Van Nes rotationplasty

The Van Nes tibial rotationplasty was first described by Borggreve in 1930 and subsequently popularized by Van Nes.[38, 39] In this procedure, the limb is rotated 180º; this is accomplished through the knee arthrodesis, the tibia, or both. The ultimate goal at maturity is to have the ankle of the short limb at the level of the contralateral knee, with the foot now acting like the residual tibia in a below-the-knee amputation.

Active control of the prosthetic knee is the most obvious advantage of rotationplasty. The function obtained after a Van Nes procedure is significantly greater than that of an above-the-knee prosthesis after knee fusion and Syme amputation.[34] Furthermore, the metabolic energy expenditure and oxygen consumption are lower than they are in patients who have undergone Syme amputation combined with knee arthrodesis.[34, 40]

Reasonable preoperative ankle and foot function are required for the ankle to serve as a knee.[35, 36] This is not always the case, because up to 70% of children with PFFD have an ipsilateral fibular deficiency.[17] Some valgus alignment of the foot and ankle can be accommodated in the prosthesis. However, severe valgus and equinus deformities with a deficient foot are a contraindication to the rotationplasty.

Preoperative toe and ankle strengthening is helpful because these structures power the new knee joint. Equinus position places the foot in a position of mechanical advantage and is, therefore, emphasized. Although hip instability and coxa vara are not contraindications to the Van Nes procedure, a stable hip allows for a more cosmetically acceptable prosthetic gait that is also less fatiguing.[36]

The two problems encountered with the Van Nes rotationplasty are (1) failure to achieve sufficient rotation at surgery and (2) subsequent derotation with continued growth.[35] Accordingly, some clinicians have advocated waiting until the child is aged 12 years before performing this procedure.[36]  The incidence of derotation can be decreased if a sufficient amount of bone is removed at the time of rotationplasty.[3]

An additional disadvantage is the unattractive appearance of a leg in which the foot has been reversed. In an attempt to improve the appearance, the toes often are removed; however, this removal often leads to a loss in sense of position as well as a loss in power. In contrast, there are reports of good patient acceptance with acceptable cosmesis by patients, parents, prosthetists, and surgeons.[34, 35, 36] It should be kept in mind that a Syme amputation serves as the salvage procedure for a failed Van Nes procedure.

In a controlled study of long-term (mean, 21.5 years; range, 11-45) results of Van Nes rotationplasty in 12 patients with congenital PFFD, Ackman et al found no differences between the PFFD group and the control group with regard to overall health and well-being on the Short Form (SF)-36.[41] The PFFD group showed significant differences in gait parameters. Computed dynamic posturography revealed reduced stance symmetry and reduced end-point and maximum excursions in the PFFD group.

Hip stabilization

Some controversy exists about the value of surgical procedures to stabilize the hip. Deficiencies in the osseous anatomy, as well as in the hip musculature, result in hip instability in most patients with PFFD. Aitken class A and B deficiencies have a femoral head within the acetabulum; therefore, some clinicians support surgical correction of the varus deformity and pseudarthrosis.

The bony stability of the femoral-pelvic articulation requires consideration. In those patients in whom lengthening is planned, achievement of good femoral head containment is necessary; this often requires acetabular procedures. In patients with class C or class D involvement, attempts at reconstruction are futile.[7]

Iliofemoral arthrodesis

Two different kinds of arthrodesis have been described in an attempt to address the hip instability. The fusion is performed to establish a stable union between the femur and pelvis, thereby eliminating pistoning. Additionally, prosthetic fitting is simplified and the gait is steadier. It is indicated in persons in whom a femoral deficiency with an unstable acetabulum is present, as in Aitken class C and class D deficiencies.

Steel described fusing the distal femoral segment to the pelvis; the femur is flexed 90º, making it perpendicular to the body axis and parallel to the floor. As a result, when these patients extend their anatomic knees, they are effecting hip flexion. Conversely, knee flexion functions as hip extension.[42, 43] In a study of 22 patients at a minimum 10-year follow-up, all had solid bony fusions and none had the fusions taken down or revised to an arthroplasty at skeletal maturity.[43]

Brown described a second procedure in which a rotationplasty is performed in conjunction with an iliofemoral arthrodesis.[44] In this case, the distal femur is rotated 180º prior to fusion, with its axis parallel with that of the body. Similar to the Van Nes rotationplasty, the knee functions as the hip joint and the ankle serves as the knee, thereby enabling these patients to function as below-knee amputees.

Lengthening procedures

These operations most commonly are performed in patients with congenital femoral deficiencies who do not have focal defects. The appropriate candidate has a congenital hypoplastic femur in which the hip and knee can be made functional. In a review of the experience at the Hospital for Sick Children in Toronto, successful femoral lengthening of up to 20% was possible with few complications. Beyond this, posterior knee subluxation and hip dislocation occurred.[3, 45]

Herzenberg et al reported on congenital femoral deficiency in children treated with femoral lengthening.[46] A common complication is fracture soon after removal of the external fixator, often despite prophylactic hip spica cast application. The authors treated nine such fractures, most of which were spontaneous events resulting in transverse fracture through regenerate bone or pin sites. All were stabilized with intramedullary Rush pins placed with special insertion techniques, and union was achieved without significant complications.

Aston et al performed a retrospective review of 30 femoral lengthenings in 27 patients with congenital short femurs using the Ilizarov technique.[47] On the basis of their findings, the authors recommended that in this group of patients, lengthening of the femur with an Ilizarov construct be carried out through a proximal osteotomy over a Rush nail. They also recommended that lengthening be limited to a maximum of 6 cm during one treatment, or 20% of the original length of the femur, to reduce the risk of complications.

Also a viable option is hip reconstruction osteotomy, which simultaneously improves gait mechanics and limb-length discrepancy while retaining joint range of motion. This procedure includes an acute valgus and extension osteotomy of the proximal femur combined with gradual varus and distraction for realignment and lengthening at a second, more distal, femoral osteotomy.[13]