Gait Biomechanics and Prosthetic Management of Children with Proximal Femoral Focal Deficiency (PFFD)

Stefania Fatone, PhD

Proximal Femoral Focal Deficiency, more commonly referred to as PFFD, is a rare, congenital lower limb deficiency of unknown cause affecting the formation of the hip joint (acetabulum and proximal femur) and surrounding musculature (Figure 1 ).

Figure 1

Physical examination alone is not diagnostic and cannot determine severity of the deficiency. Early in life, an x-ray will confirm that the femur is short with the proximal third of the femoral shaft, neck, trochanteric area and head seemingly absent. However, the area between the femoral head and shaft is usually occupied by a cartilage anlage in which ossification is delayed making it undetectable on x-ray. Serial x-rays taken over the first year or two of life will clarify the actual severity and detail the development of the deficiency.

As with many congenital limb deficiencies, the degree of deficiency may vary considerably among individuals. Many systems have been proposed to classify these variations. It has been observed that only the number of cases seen limits the potential number of categories (1)! Most systems attempt to classify PFFD and related femoral deficiencies based on anatomical features of the acetabulum and proximal femur identifiable on x-ray. While no single system of classification has achieved universal acceptance, the most commonly acknowledged system is that of Ait-ken (2). Aitken identifies four classes (A through D) based on serial x-ray examination of the formation of the femoral head and acetabulum, with A being the least severe form of deficiency and D the most severe.

PFFD results in severe limb length inequality and hip joint instability. Treatment can be both surgical and/or prosthetic and is aimed primarily at restoring function, in particular, the ability to walk. PFFD occurs in approximately 1 in 52,000 births (3). It most commonly occurs

in only one limb but may, in approximately 10 to 15% of cases, occur in both lower limbs (4). There is a high incidence of associated anomalies, with the most common being fibular hemimelia (absence of a portion of the fibula) (5). Associated anomalies complicate management of the PFFD child and in some instances may con-traindicate certain surgical options. Bilateral cases are often left untreated since, if the limbs are of approximately the same length, the child can usually ambulate effectively and painlessly without intervention.

Management of PFFD

The biomechanics of gait in unilateral PFFD depend greatly on the type of treatment undertaken and can be almost as variable as the deficiency itself. Where the decision is made to leave the limb surgically unrevised the issue of limb length inequality is addressed prosthetically using some form of extension prosthesis. Where the decision to revise the limb is made, there are two main options available (Figure 2 ).

Figure 2

Traditionally, the most common revision procedure has been the Syme's ankle disarticulation with knee fusion, which results in the functional equivalent of a knee disarticulation (through-knee) amputation. To ensure that knee joint axes are at approximately equal heights at maturity, ankle disarticulation and knee fusion may be accompanied by epiphysiodesis (halting the growth plate) of the opposite knee. The other, more controversial option is to convert the limb to a functional transtibial equivalent using the Van Nes (Tibial) Rota-tionplasty procedure (6). This procedure involves turning the ankle and foot 180° about a vertical axis, so that ankle dorsiflexion simulates knee flexion. This makes it possible for the rotated ankle to voluntarily control a mechanical knee during gait. Regardless of the choice of revision surgery, it is generally accepted that when the limb is surgically revised and, provided the femur is not fused to the pelvis, the knee should be fused.. This creates a single skeletal lever with which to control the prosthesis. Where the knee is not fused, stability of the proximal knee joint relies upon the prosthetic socket and the flexion-abduction orientation of the thigh persists. Regardless of the surgical procedure chosen, most authors prefer to operate early, such as when the toddler demonstrates the interest and ability to stand (7).

Biomechanics of Gait in Unilateral PFFD

We analyzed the gait of 9 children with unilateral PFFD (8). We believe that pelvic and hip kinematics are consistent within the PFFD population despite variability in the degree of severity and treatment (8,9,10). Regardless of surgical treatment, hip motion on the affected side, is characterized by a 'pause' in hip extension during mid-stance, absent stance phase hip abductor moment, and absent terminal stance power generation. On the sound side, there is increased power generation at the hip in early stance and mid-stance vaulting at the ankle to assist with toe-ground clearance of the prosthetic limb during swing.

In the sagittal plane, the mid-stance 'pause' in hip extension - actually a decrease in the rate of hip extension during mid-stance - coincided with increased anterior pelvic tilt (Figure 3 a)

Figure 3suggesting that hip extension was hampered by a flexion contracture and compensated for by changing the pelvic orientation. Our results suggested that abnormal hip and pelvic motion in mid-stance was more pronounced in the PFFD subjects who had moderate hip flexion contracture, and less severe in those who did not have a hip flexion contracture (8).

Visually, PFFD gait is marked by excessive lateral trunk flexion (often referred to as Trendelenburg gait), which is due to the inadequate abductor mechanism (11). Even if present, the abductor muscles cannot be effective where the hip joint is so unstable. This was reflected in the absence of coronal plane hip moments in our PFFD subjects (Figure 3 b).

Usually amputees use hip musculature to compensate for lack of knee and ankle muscle activity (12), however, the deficient nature of the PFFD hip hinders compensatory use on the affected side as demonstrated by reduced power generation at the hip (Figure 3 c) (8). There is also very litde power generation at the knee or ankle on the affected side. Our results suggest that the sound limb may be generating the power required for prosthetic swing phase, since at the time of terminal stance on the affected limb, the sound limb exhibited increased hip power generation. Simultaneous rapid internal rotation of the affected side pelvis with respect to the sound hip transferred power to the affected limb (8). This interpretation of the motion analysis data could be substantiated by EMG data, however EMG data was not acquired in this study.

Given the osseous defect and muscular abnormalities, motion between the pelvis and femur in PFFD is unique. This movement has been previously described as 'telescoping' (11). Telescoping refers to the displacement that occurs between the pelvis and residual femur during weight bearing. In order to measure motion of the hip, current motion analysis systems generally incorporate an anthropometric model of the pelvis based on measurements of normal anatomy. While these models prove reasonable in able-bodied subjects, they are likely to be inaccurate in cases such as PFFD where the pelvis and hip are malformed and there is no easily identifiable axis of rotation at the hip. We developed a method of predicting the hip joint center (HJC) in children with PFFD that did not rely on anthropometric data (8). Instead, an optimization model was developed based on dynamic data. This model was found to provide reasonable estimates of HJC location in able-bodied children compared to other HJC models (8). This was used to measure telescoping of the PFFD limb and a meaningful relationship between telescoping and axial loading of the limb was established (Figure 4 ).

Figure 4

We investigated the effect prosthetic socket design might have in limiting telescoping of the PFFD limb (8). Based on Lehneis' ideas (13) we hypothesized that the higher trim line and larger radius of the posterior brim of the Ischial Containment socket would provide better support of the pelvis through the loading response phase of gait in children with PFFD thus reducing telescoping and improving gait biomechanics. We compared the effects of the Quadrilateral and Ischial Containment sockets on the gait kinematics and energy expenditure of five children with unilateral PFFD and Syme's ankle disarticulation. The results suggested that Ischial Containment sockets provide better support of the pelvis. Pelvic and trunk range of motion, telescoping, and vertical sacral displacement were reduced and velocity, stance duration asymmetry and energy expenditure improved when walking in the Ischial Containment socket compared to the Quadrilateral socket. Energy expenditure was 20% less in the Ischial Containment socket compared to the Quadrilateral socket (8).

Preliminary gait studies have indicated that subjects with Tibial Rotationplasty have a degree of prosthetic control not possible with a Syme's ankle disarticulation and transfemoral prosthesis (8,9,14). Fowler et al. (9) reported that most Tibial Rotationplasty subjects demonstrated stance phase knee flexion and stance phase knee extensor moments during loading response. Sheil et al. (14) reported that the Tibial Rotationplasty procedure encouraged a longer stride length, shorter stride time and therefore, faster velocity. Comparing the two surgical procedures, they also found that subjects with Tibial Rotationplasty maintained a closer to normal stance/swing ratio than the subjects with Syme's ankle disarticulation. We undertook pre- and post-operative evaluation of the Tibial Rotationplasty procedure in 2 young subjects that suggested that the rotated ankle demonstrated adaptation by functioning as a knee, as demonstrated by the presence of stance and swing phase knee flexion, relatively early in the post-operative period (8).

Compared to able-bodied children, and regardless of surgical treatment, children with PFFD incur a greater metabolic cost when walking (8). A number of authors have reported that the Tibial Rotationplasty procedure results in a more energy efficient gait when compared to the Syme's ankle disarticulation procedure. In fact, the Tibial Rotationplasty procedure has been reported as being 10 to 25% more energy efficient (15,16).

Figure 5


It has been demonstrated that the gait of children with PFFD differs from that of able-bodied children and that the differences are consistent despite variability in the severity and treatment of the deficiency. Pelvic and hip kinematics and hip kinetics are abnormal and the sound limb is used to compensate for lack of power generation on the affected side. Compared to able-bodied children, children with PFFD incur a greater metabolic cost when walking regardless of the type of surgical revision, although the Tibial Rotationplasty procedure was less energy expensive compared to the Syme's ankle disarticulation and knee fusion procedure. Our results suggested that Ischial Containment sockets decrease energy expenditure by providing better support of the pelvis in children with unilateral PFFD and Syme's amputation.



    1. Torode LP. (1997) In: Broughton N.S. (Ed.) A Textbook of Paediatric Orthopaedics. W.B.Saunders Co. Lim.: London, U.K. Aitken G.T. (1969) In: Aitken G.T. (Ed.) Proximal Femoral Focal Deficiency: A Congenital Anomaly. A Symposium. National Academy of Sciences, Washington D.C., pp. 1-22. Rogala E.J., et al. (1974) Journal of Medical Genetics, 11:221-233.


    1. KrajbichJ.I. (1989) In: Kalamchi A. (Ed.) Congenital Lower Limb Deficiencies. Springer-Verlag: New York, U.S.A. Chapter 6, pp. 108-127.


    1. Tooms R.E. (1987) Journal of the Association of Children's Prosthetic-Orthotic Clinics. 22(2):22. Van Nes C.P. (1950) Journal of Bone and Joint Surgery. 32B(1):12-16.


    1. Epps C.H. (1983) Journal of Bone and Joint Surgery. 65A(6):867-870.


    1. Fatone S. (2000) Gait Biomechanics and Prosthetic Management of Children with Proximal Femoral Focal Deficiency (PFFD). PhD Thesis, La Trobe University, Australia. Fowler E.G., et al. (1999) Journal of Pediatric Orthopaedics. 19(6):720-731. Sutherland D.H. (1984) Gait Disorders in Childhood and Adolescence. Williams and Wilkins: Baltimore, U.S.A. Pp. 89-106. Pirani S., et al. (1991) Journal of Pediatric Orthopaedics. ll(5):563-570.


    1. Bagley A.M. and Skinner H.B. (1991) Critical Reviews in Physical and Rehabilitation Medicine. 3(2):101-120.


    1. Lehneis H.R. (1985) Clinical Prosthetics and Orthotics. 9(4): 6-8.


    1. Shell E.M.H., et al. (1995) Gait and Posture. 3:107.


    1. Alman B.A., et al. (1995) Journal of Bone and Joint Surgery. 77A(12):1876-1882. McClenaghan B.A., et al. (1989) Journal of Bone and Joint Surgery. 71A(8):1178-1182.





Financial support for part of the work described in this article was received from the Hugh Williamson Orthotic and Prosthetic Unit, Royal Children's Hospital, Melbourne Australia in the form of The Ray McKernan Studentship for Orthotic and Prosthetic Biomechanical Research. I am also indebted to my supervisors and mentors Dr Timothy M. Bach, Professor H. Kerr Graham and Mr Ian Torode for all their assistance during my PhD candidature.


ACPOC likes to thank "Capabilities" for their kind permission to reprint this article.