The Role of Valgus Osteotomy in Proximal Focal Deficiency

Jeffrey D Ackman, M.D., F.A.A.O.S.


Proximal femoral focal deficiency encompasses a spectrum of disorders characterized by deficiencies of bone and soft tissue of the proximal femur and hemipelvis. The patient's gait and function are determined by the bony and/or cartilaginous representation of the proximal femur and acetabulum as well as the musculature about the hip joint. Various classification systems have been proposed giving prognostic value and treatment recommendations based upon the grade of the deformity and femoral deficiency. The role of surgery, specifically valgus femoral osteotomy, however, is uncertain with some authors favoring intervention and others adopting a non-operative approach. This study was undertaken in an effort to assess the role of valgus osteotomy in the management of patients with PFFD to determine its effect on anatomy and functional improvement.


The charts of all patients with PFFD who had undergone proximal valgus osteotomy here at Shriners Hospitals for Crippled Children-Chicago Unit between 1960 and the present were reviewed. Other diagnostic categories included were hypoplasia of the lower extremity and congenital short femur as these cases may actually have been PFFD but were misdiagnosed at their initial presentation and their misdiagnosis perpetuated thereafter. Patients were classified according to Aitken as determined by radiographic criteria on serial X-rays. Only those patients with true PFFD who had undergone a valgus osteotomy of the proximal femur were selected for review. Information gathered from the charts included the Aitken class and affected side/sides, age at osteotomy/osteotomies, whether bone grafting and/or hip exploration was performed, if knee fusion was done prior to, at the time of, of after valgus osteotomy of the femur, use of internal fixation of the osteotomy, and observations on the patients gait before surgery and at the time of follow up. Radiographs were analyzed for Aitken class and neck shaft pre-op, post-op, and during the follow-up period. For those patients with bowing deformities of the proximal femur at the neck shaft junction, the angle calculated was that between the neck and the diaphysis of the femoral shaft excluding the deformity in the pertrochanteric region. Angles were calculated from standing AP radiographs of the pelvis and entire affected femur allowing the limb to hang freely with no attempt to control the rotational position during the X-rays. Neck shaft angles were determined serially until the time of final follow-up.


Twenty patients met the appropriate selection criteria and had sufficient X-rays and follow-up for inclusion in the study. There were 14 males and 6 females with 18 patients unilaterally affected and 2 patients bilaterally affected. Only Aitken A or B hips underwent proximal femoral valgus osteotomy, with 15 hips classified as Aitken A and 6 as Aitken B. One bilateral patient was classified as an a A on the left and a C on the right; giving 21 hips in these 20 patients that were treated with valgus osteotomy. Age at the time of initial osteotomy averaged 5.4 years, range 2 to 11 years, with 7 hips undergoing a second osteotomy, 3 hips a third and 2 hips a fourth , giving 33 osteotomies overall. For those patients with repeat osteotomies, the time interval from the initial to the repeat varied from 1 to 5 years, with the oldest undergoing an osteotomy at age 14. The 2 patients who had their initial osteotomy done before the age of 3 underwent repeat osteotomies but there was otherwise no correlation between age of initial osteotomy and number of osteotomies performed, with older patients having just as great a likelihood of recurrence as younger patient. Four hips, 3A and 1B, had had 6 bone grafts to the proximal femur, with 4 hips grafted prior to the valgus osteotomy and 2 hips grafted at the same time. All these hips had recurrent varus deformity in follow-up. One patient underwent knee fusion prior to the valgus osteotomy and 5 patients had a knee fused after the last valgus osteotomy with no apparent effect on the end result. Eight of the 33 osteotomies were performed without internal fixation, but varus recurred in all hips whether internal fixation was used or not. Four patients had acetabular procedures; 2 Salter osteotomies prior to femoral lengthening, 1 triple innominate and 1 shelf acetabuloplasty. All these acetabular procedures were done as the final procedure on the hip.

The initial neck shaft angles for the Class A patients averaged 80 degrees (range 35 to 90 degrees) and for Class B 69 degrees (range 55 to 80 degrees). Pre-operative neck shaft angles averaged 76 degrees (35 to 100 degrees) for Class A patients and 66 degrees (60 to 70 degrees) for Class B. Postoperative neck shaft angles averaged 115 degrees for Class A (60 to 135 degrees) and 83 degrees for Class B (70 to 100 degrees). Correction at the time of surgery averaged 30 degrees (14 to 45 degrees) for Class A and 22 degrees (10 to 40 degrees) for Class B.

Follow-up on these patients from the time of their osteotomy averaged 9 years, range 1 to 18 years. All patients except 2 lost correction from that obtained at the time of surgery. Comparing the preoperative neck shaft angle to that measured at final follow-up the Class A patients averaged 9 degree improvement (-25 to +60) and the Class B worsened 2.5 degree (-10 to +5). Excluding the high and low for Class A, the 13 remaining hips sustained an average of 3 degrees more valgus than measured preoperatively. For those patients who underwent repeat valgus osteotomies, they had no greater loss of correction than those patients who underwent only 1 procedure. Complications included 6 hardware failures in the 25 osteotomies performed with internal fixation.

On summary, 33 osteotomies were performed on 21 hips in 20 patients, with 2 patients maintaining the correction obtained at the time of surgery. The majority of patients yielded neck shaft angles within plus/minus 20 degrees of the pre-operative values. All patients continued to walk with a Trendelenburg gait.


Aitken was the first to introduce the term of proximal femoral focal deficiency and enumerated four basic problems in its management: 1) hip instability, 2) malrotation, 3) inadequate proximal musculature and 4) leg length inequality. Koman et al in 1982 carried this one step further listing the important factors affecting ambulation as 1) other skeletal anomalies, 2) leg length discrepancy, 3) foot position and 4) hip stability, with hip stability being the least important factor in ambulation.

Throughout the literature some authors have advocated proximal femoral valgus osteotomy for Aitken Class A and B patients, while others have opposed it. King recommended valgus osteotomy for Type A and metaphyseal synostosis with valgus alignment of the head for Type B, best done between two to three years of age to allow for the greatest amount of remodeling. However, he had few cases and the longest follow-up was seven years.

Fixsen and Lloyd Roberts in 1974 advocated abduction osteotomy of the proximal femur when their recommended grafting of the pseudarthrosis showed signs of incorporation and progressive ossification of the pre-existing cartilage. Three of 6 unstable hips were stabilized in this manner. They also recommended excising the pseudarthrosis and impacting the shaft into the femoral head if 1) pseudarthrosis existed both at the osteochondral junction and the femoral neck with the central fragment seemingly too mobile and devitalized to be worth preserving or 2) previous grafting operations had failed. They performed this four times in their series but no long-term follow-up was given, nor recommended age for treatment. Surgical intevention depended upon development of the hip joint as determined by serial X-rays.

In 1969 Amstutz reported on his experience with valgus osteotomies. Three Class 1 patients (Aitken A) had vlagus osteotomies with 2 lost to follow-up. The one patient had 5 corrections with delayed healing He also recommended valgus osteotomy for Class 2 and 3' (Aitken A and B) with progressive varus deformity or persistent subtrochanteric pseudarthrosis. However, no specific number of cases or follow-up were given.

Panting and Williams stated that coxa vara deformity for Aitken A and B equivalents can be improved by subtrochanteric valgus osteotomies but did not routinely perform this in their series. In 1982 Koman et al published their experience with 91 patients with PFFD. They felt that attempts to improve pelvifemoral stability for other than Class A patients was unwarranted. For selected Class A patients with neck shaft angles less than 110 degrees and adequate acetabulae according to their measurement criteria valgus osteotomy was indicated. No follow-up on their osteotomies was reported.

Gillespie and Torode in 1983 reviewed 69 patients and performed valgus osteotomies in 17. The best indications for valgus osteotomy for true PFFD's were those hips with a well formed head and neck and greater trochanter but a persistance of subtrochanteric pseudarthrosis. Other patients may have an improved radiographic appearance following surgery but continued to have abductor weakness and a marked Trendelenburg gait.

Kalamachi et al presented their classification system for congenital deficiencies of the femur in 1985. Sixteen Aitken A or B equivalents had femoral osteotomies, with no comment made on maintenance Or correction, need for reoperation, or change in gait pattern following surgery.

In 1989 Krajbich reported his treatment outline for PFFD. He recommended valgus osteotomy for Aitken A and B with excision of cartilage if necessary to fix the neck shaft angle at 155 to 170 degrees. However, patients may have persistent Trendelenburg gait depending upon their abductor function. This may be improved by taking down and reattaching the abductors where the greater trochanter would normally be. Surgery was recommended between 5 to 7 years of age when a significant portion of the proximal femur showed ossification.

Be van-Thomas and Millar, Westin and Gunderson, and Richardson and Grier, among others have advocated a nonoperative approach. Their reports have indicated high rates of failure, rapid loss of correction and no change in prosthetic fitting or ambulation following valgus osteotomies.

The results of our series would agree with those of others who feel that operations to improve the pelvic-femoral alignment in Aitken A or B PFFD is rarely indicated. For the 21 hips reviewed, the average correction and follow-up was 9 for Class A and 2-5 for Class B, with only 2 of 21 maintaining the correction obtained in surgery. Knee fusion did not appear to have an affect on our results. Only 1 patient in the series had the knee fused prior to valgus osteotomy, so the concept of knee fusion transferring stress to the proximal femur inhibiting development did not apply in this instance. Most authors believe that knee fusion has no consequence on development or ossification of the proximal femur.

Our age at the time of surgery averaged 5 + 4 years with a range of 2 to 11 years. The optimal timing of surgical intervention is not precisely defined in the literature However, in Our series those patients closer to skeletal maturity had as high a recurrence rate as those with more growth remaining.

We did not obtain the neck shaft angle of 155 to 170 degrees recommended by Krajbich in any of our patients. This was probably due to the surgeon's reluctance to excise the unossified cartilage or pseudarthrosis and/or shorten the femur enough to obtain that degree of correction, or release the abductor mechanism enough to rotate the trochanter down Perhaps this would have provided better long-term correction and stability, but even then, as Krajbich reported, the abductor lurch might not have been improved.

None of the internal fixation used crossed the growth plate of the femoral head and neck in an attempt to prevent recurrent varus with subsequent growth. However, I feel that the varus deformity recurs due to the muscle forces acting across the hip, and closing down the physis would have had little if any effect on the varus/valgus configuration of the proximal femur.

One criticism of our methods was that we made no attempt to control the rotation of the limb when obtaining X-rays, and thus may not have been measuring the true neck shaft angle. That is a flaw inherent in retrospective study and its bias on the end results is not known. A standardized method for obtaining radiographs would be an important element for future studies, We believe, as proposed by Pirani and Beauchamp et al, that the soft tissue envelope about the hip determines the ultimate gait and hip function of patients with PFFD. This muscular cuff plays a significant part in weight transfer across the hip joint and pelvic-femoral stability. The morphology of the proximal femur in Type A and B PFFD is dependent upon the muscle forces acting upon it, in adherence with the principle of form follows function. These patients have abnormalities of the glutei, adductors and short external rotators as well as the skeletal anomalies. Based upon the results of this study we cannot recommend valgus osteotomy to improve gait and function in patients with PFFD.

Shriners Hospital, Chicago Unit 2211 N. Oak Park Ave. Chicago, IL. 60635-3392