Management of severe congenital femoral deficiency: Does surgical intervention enhance prosthetic function?

Abstract

Background & Objective:

The purpose of this study was to evaluate prosthetic outcome in patients with severe congenital femoral deficiency and the potential benefits of surgical intervention on prosthetic fitting and gait.

Methods:

A retrospective review identified 26 active case records with a proximal femoral focal deficiency using a prosthesis. Validated outcome measures evaluated comfort, function, and prosthetic use and quality-of-life assessment. Outcome compared age groups and surgical intervention. Gait analysis performed in 7 patients further evaluated hip and knee function.

Results:

Eleven male patients and 15 female patients, including 13 children (mean age 10 years, range 5–16) and 13 adults (mean age 36 years, range 23–63) were evaluated. Better prosthetic function and PedsQL scores were recorded in the pediatric group. There was a trend for better scores after surgery. Gait analysis demonstrated reduced hip extension compensated by knee flexion in 3 patients, 2 patients had hip extension with near normal kinematics, 1 untreated patient walked with an unsteady gait, and the remaining walked well using an ischial-bearing prosthesis with pelvic compensatory movements.

Conclusion:

The management strategy in severe proximal femoral focal deficiency remains a major challenge. Hip reconstruction seems to improve functional scores. Overall, the scores seem to decline into adulthood but not significantly. Gait analysis before further surgical intervention is recommended because compensatory knee flexion may improve step length in terminal stance. Limited numbers, with short follow-up, prevents clear guidance on the benefit of surgery.

Introduction

A failure of embryological development leads to a congenital femoral deficiency ranging from simple hypoplasia to an absent or rudimentary femur with a varied clinical and radiographic presentation.^1,2 The classical appearance in severe cases of a shortened, bulbous, externally rotated thigh³ became associated with the term proximal femoral focal deficiency (PFFD).^4,5 There is a lack of published literature to guide gold standard surgical intervention in these rare challenging cases. The choice of what surgery to do and when to undertake it is currently often a matter of surgeon’s preference and is often recommended in childhood. A lack of confidence in expected outcome goals may lead to parental confusion in deciding what treatment to accept, especially in those cases opting for long-term prosthetic use.

Understanding proximal femoral development helps predict future hip stability and surgical options to improve hip function.⁶ Grouping patients into those suitable for limb lengthening and reconstruction vs. prosthetic support offers further information for parents making difficult management decisions.^7,8 Recognition of constant anatomical patterns has enabled development of classification systems that offer surgical road maps to create a stable hip, under which limb lengthening can be considered to produce a leg of a length equivalent to the opposite normal side.^9,10 This in itself is not a guarantee of success because bone lengthening in congenital short femur remains a major surgical challenge with significant risk of complications, and often, multiple lengthening episodes are required.^11–13

Prosthetic use can be more appealing in certain scenarios, but surgical intervention may still be an option to improve both prosthetic fitting and overall limb function. The leg length discrepancy (LLD) can be managed by an extension prosthesis suspended on the tibia allowing knee function during gait, or extended above the knee acting as an ischial-bearing prosthesis. The primary surgical goal is to achieve hip stability by restoring normal proximal femoral anatomy and releasing the soft tissue contractures.¹⁴ A combination of proximal femoral valgus shortening de-rotation osteotomy and a pelvic osteotomy (Dega acetabuloplasty) can produce a radiographically “normal”-looking hip. The combination of these surgical procedures are often referred to as the Systematic Utilitarian Procedure for Extremity Reconstruction of the hip (SUPERhip) procedure.¹⁵ Further surgical options include a knee fusion with a Syme disarticulation, producing a single bone thigh segment, a rotationplasty resulting in the ankle acting as a knee,^16–20 or an isolated Syme disarticulation may simply lead to an improved cosmetic prosthetic design.^21,22 In cases of subtotal absence of the proximal femur, the question remains whether an iliofemoral fusion is beneficial.^23–25

This is an observational descriptive study evaluating prosthetic outcome in patients with severe congenital femoral deficiency. The aim was to provide a commentary on the potential benefits of surgical intervention, specifically looking at the influence on prosthetic fitting, gait, and impact on quality of life (QoL) for these patients with a severe lower limb longitudinal deficiency.

Methods

Patients with congenital limb deformity, associated with severe longitudinal deficiency, present to our institution either to a Consultant Pediatric Limb Reconstruction Orthopedic Surgeon or as a direct referral to the Limb Rehabilitation Unit for advice on prosthetic management. The timing of the referral can be variable, from newborn infants with an obvious lower limb shortening and deformity to adults with known PFFD, already using a prosthesis for advice on potential improvement in function. A full clinical and radiological evaluation confirms the diagnosis. The final LLD can be predicted in children²⁶ and surgical options described with the parents or mature patients following an agreed algorithm.²⁷ In cases of severe deformity and LLD, the information provided includes the timing and number of predicted surgical interventions, with associated complication risk, both for limb reconstruction (with the goal of achieving a limb equal in length to the opposite side with good hip, knee, and foot function) and surgery to optimize the use of a prosthesis.

After Research & Development approval (SE20.38), a retrospective review of the current limb rehabilitation database identified 146 active case records with a congenital limb deficiency. Eighty-one with no femoral deficiency, 27 had femoral shortening associated with another condition, and 12 with PFFD who had undergone lengthening were no longer using a prosthesis were excluded, leaving 26 patients with PFFD using a prosthesis as the study group. The extension prostheses included sockets manufactured using the hand casting technique and were attached using a pylon to a weight-suitable carbon fiber foot. A cross-sectional observational cohort study was undertaken. All parents of children younger than 18 years and adult patients agreed to partake in the study. Data were collected in a single episode per patient. It was accepted that preoperative and postoperative analyses would not be possible with the current study design.

Patients were contacted by telephone and validated outcome measures used to evaluate comfort, function, and prosthetic use, including the Special Interest Group in Amputee Medicine (SIGAM) scale,²⁸ the K activity levels,²⁹ and the locomotor capability index.³⁰ Quality of life was evaluated using the PedsQL inventory questionnaire.³¹ Radiographs were classified according to Paley.^9,10,15,25

The aim of this study was to see whether functional and QoL scores differed between different severities of PFFD and after surgical intervention. The outcome scores were compared according to Paley radiological classification, age at assessment (children under 18 vs. adult), and nonoperative vs. operative groups. The operative groups were subdivided further into those who had undergone the SUPERhip procedure and those who underwent other surgical procedures. The heterogenous group of 26 patients inevitably results in very small subgroups for analysis. For completeness, statistical analysis was undertaken using SPSS (v23; SPSS Inc); Mann-Whitney U tests were undertaken for comparison for 2 groups, and Kruskal-Wallis H test was undertaken for 3 groups, based on nonparametric data. Accepting the small numbers within each group, statistical analysis was generally not feasible, so a descriptive commentary on the scores recorded has been made, highlighting trends seen.

Patients were invited for gait analysis in an attempt to describe the gait patterns between different age and surgical groups. Seven patients (4 children and 3 adults) accepted, and analysis was undertaken using standardized Motor Learning Laboratory protocols. The patients walked barefoot on the Gait Real-time Analysis Interactive Laboratory system (Motek, Amsterdam, The Netherlands). This consists of an instrumented dual-belt treadmill, 10 infrared Bonita cameras (Vicon, Oxford Metric, United Kingdom) operating at 100 Hz, and an immersive environment projected on a semicylindrical screen. Surface markers were located in accordance with the conventional gait model.^32–36 Additional markers were fixed over the sacrum, patella, and tibia to improve tracking of the pelvis,³³ thigh,³⁴ and tibia/shank.³⁵ Markers were placed on the prosthetic surface to approximate the required landmark position. Tracking reference frames were related to anatomical reference frames with the subject in the midjoint range crouched position. Patients then walked on the treadmill, and the walking speed was controlled to cover a range of walking speeds. This included a slow walk, a midspeed walk, and a fast walk for approximately 1 min. The rates varied from 0.3 to 0.4 m/s in the slow walk and 0.7–1 m/s in the fast walk, compared with normal average adult walking speeds of 0.8 m/s (slow) and 1.6 m/s (fast). The marker locations were extracted from Vicon Nexus software (version 2.7.1) and reported into Visual3D (C-motion, Germantown, MD—version 2020.08.03). Raw data were filtered by a low-pass, fourth-order Butterworth filter with a cutoff frequency of 6 Hz, and 3D mean cycle kinematics were calculated using a standard coronal-transverse Euler rotation sequence, which was suitably reversed for the pelvis. The analysis focused on hip and knee function.

Results

Patient demographics are listed in Table 1 for the 26 study patients, including distribution by classification and surgical intervention. The mean age of surgical intervention was 49.3 months (range 18–88 months, confirming all surgeries were undertaken in childhood). Radiological examples demonstrating the anatomical differences from a normal hip of Paley type 1B using an extension prosthesis and type 3B PFFD using an ischial-bearing prosthesis are demonstrated in Figures 1(a) and 1(b) and 2(a) and 2(b).

Table 1.

Patient demographics.

Sex	11 male	15 female
Age	13 children	(mean age 10 y, range 5–16)
	13 adults	(Mean age 36 y, range 23–63)

Classification
Paley	Extension prosthesis	Ischial-bearing prosthesis
Type 1A	8	—
Type 1B	10	—
Type 2A	—	1
Type 2B	—	—
Type 3A	—	—
Type 3B	2	5

Surgical intervention	9 children	2 adults
SUPERhip		5
Proximal femoral valgus osteotomy		4 (including 1 patient with knee disarticulation)
Femoral lengthening		1
Knee fusion with Syme disarticulation		1

Figure 1.

(a) Untreated Paley 1B hip, aged 8 years. (b) Untreated Paley 1B hip, aged 22 years.

Figure 2.

(a) Paley 3B unstable hip with an ischial weight-bearing prosthesis in patient aged 12 years. (b) Paley 3B unstable hip in patient aged 23 years.

One adult patient’s scores were excluded; she no longer used a prosthesis because of hip osteoarthritis. In the remainder, there were no significant differences seen in any of the functional outcome or QoL scores when comparing Paley type 1, 2, and 3 femurs.

Comparing the age groups (children younger than 18 years against adults), there were significantly higher K scores seen in the pediatric group (p = 0.014) and better scores with the SIGAM scale, but these differences were not significant (Figure 3(a)). There was no significant difference in the locomotor capability index but again a trend for better scores in the pediatric group (Figure 3(b)). There was also no statistical difference in most of the PedsQL domains, but a trend for better scores in the pediatric group (Figure 3(c)). Significantly better scores were seen in the pediatric patients in the emotional domain (p = 0.039).

Figure 3.

(a) Special interest group in amputee medicine and K-level scores. Pediatric vs. adult. (b) Locomotor capability index score. Pediatric vs. adult. (c) QoL scores. Pediatric vs. adult.

To analyze the impact of surgery, the 25 patients were subdivided into 3 groups, those who underwent the SUPERhip procedure (5 patients) (Figures 4(a)–4(c)), those who underwent other surgical procedures (6 patients), and the remainder who did not undergo surgery (14 patients). The senior author (P.C.) performed surgery in 6 patients; the remaining 5 procedures were performed by other surgeons, 2 by an individual surgeon at a local children’s hospital, 2 undertaken at children’s hospitals abroad, and the final adult patient was operated historically as a child at another local hospital.

Figure 4.

(a) Systematic utilitarian procedure for extremity reconstruction of the hip procedure preoperative radiograph. (b) Systematic utilitarian procedure for extremity reconstruction of the hip procedure intraoperative radiograph. (c) Systematic utilitarian procedure for extremity reconstruction of the hip procedure postoperative radiograph.

There was no overall statistical difference in relation to the prosthetic function scores, K score and SIGAM, between surgery and no surgery groups, but better scores were recorded in the SUPERhip subgroup (Figure 5(a)). Similar scores were recorded in all groups in relation to the locomotor capability index (Figure 5(b)). Furthermore, there was a trend for better QoL scores after surgery, but this was not statistically significant (Figure 5(c)).

Figure 5.

(a) Special interest group in amputee medicine and K-level scores. No surgery vs. surgery. (b) Locomotor capability index score. No surgery vs. surgery. (c) QoL scores. No surgery vs. surgery.

Formal gait analysis was undertaken in 4 children with a mean age of 7.2 years (range 6–8 years) and 3 adults with a mean age of 22.3 years (22–23 years). Six were Paley 1B using an extension prosthesis below the knee, and 1 was an adult patient aged 23 years, with an unstable Paley 3B hip who used an ischial-bearing prosthesis. No surgical intervention had taken place in 3 patients (2 adults and 1 child.) Two patients had undergone a SUPERhip procedure; 1 adult patient a proximal femoral valgus osteotomy; the remaining patient had undergone a knee fusion and Syme disarticulation (Paley 1B with subtrochanteric delayed ossification) (Figures 6(a) and 6(b)).

Figure 6.

(a) Subtrochanteric pseudarthrosis preoperative radiograph. (b) Subtrochanteric pseudarthrosis treated with knee fusion and Syme disarticulation.

Similar kinematics were recorded in cases 1–3 (1 untreated adult, the adult proximal femoral valgus osteotomy, and 1 child who had undergone a SUPERhip). All 3 were characterized by reduced hip extension, with compensatory knee flexion, which enabled an increase in stride length. Case 4, the other child who had undergone a SUPERhip procedure, did demonstrate hip extension with only minor knee flexion compensation. The resulting orientation of the long shank segment of all 4 portrayed as graphically normal with a reasonable step length. Case 5was an untreated child who demonstrated an unsteady gait pattern, but he was able to produce an equal step length with compensatory movements at the pelvis, hip, and knee level. The patient who had undergone the knee fusion had near-normal kinematics of the hip, case 6.

Despite an unstable hip and ischial-bearing prosthesis, the remaining adult walked well, case 7. The pelvis, noted to be elevated and oscillated through approximately 10 degree of motion throughout the gait cycle, was posteriorly tilted through left (affected side) single-leg support and anteriorly through the left leg swing phase. These movements counter the action of the left hip, and the oscillations occur within a normal range of pelvic values; the average sagittal pelvic position was good and a similar range of hip movement recorded in comparison with the normal side. The prosthetic ankle kinematics demonstrated clear first and second rocker movements with a step length only slightly less in distance than that in the opposite side.

Discussion

This cohort study has confirmed good function using an extension prosthesis in these cases of severe deformity associated with congenital femoral deficiency and, furthermore, the potential benefit of surgical intervention to improve hip anatomy with associated improvement of prosthetic function. The wide spectrum of both clinical and radiological presentation seen in this uncommon condition raises the question of an optimum management strategy. While simple hypoplasia of the femur, with a small LLD, may be treated confidently with established surgical limb lengthening techniques, patients with PFFD and associated severe limb deficiency and proximal femoral deformity with acetabular dysplasia or absent proximal femora prove more challenging. Recommending surgical intervention when there are no long-term functional outcome studies regarding limb function in adulthood adds further difficulty in the decision-making process, especially when the surgery is offered in early childhood, as demonstrated by our study where the mean age during surgery was 49.3 months.

It seems logical to correct the severe proximal femoral deformity, namely coxa vara, recurvatum, and retroversion to improve hip stability, biomechanics, and gait. This avoids the need for an ischial weight-bearing bulky prosthesis and minimizes the potentially excessive pistonning of the prosthesis during gait. Fernandes et al have demonstrated improved hip stability in 31 children after a modified SUPERhip and subsequent lengthening, compared with a historical group in which hip reconstruction had not been undertaken.³⁷ In our study, the prosthetic scores were better after a SUPERhip procedure, and better QoL scores were recorded after any surgery. The gait analysis also demonstrated hip kinematics similar to normal population values in 1 patient assessed after the SUPERhip procedure. In the second, there were improved graphs compared with those of the child who had undergone no treatment, but compensatory knee flexion was observed to allow extension of the limb in terminal stance.

After the creation of a stable hip, the question of leg length equalization remains. A clear understanding of the surgical lengthening process, the number of operations, and risk of complications results in many parents opting for continued prosthetic management. This can be achieved by an extension prosthesis but without the ability to include a prosthetic knee joint, which presents biomechanical and practical challenges with growth. This raises questions as to whether further surgical intervention could lead to improved prosthetic fitting and limb function.

With a very short femur, the functional contribution of such a proximally sited knee joint may be challenged. Gait analysis confirmed compensatory knee flexion effecting thigh extension during the stance phase in half the patients. It therefore seems worthwhile undertaking a gait analysis before further surgical intervention, such as knee fusion, which theoretically may reduce stride length. In severe deficiency, a fusion combined with a Syme disarticulation can produce in effect an above-knee amputation,²¹ with an ability to bear weight through the heel pad. This hypothetically may enable the hip to gain more extension by creating a longer lever arm, with the single-bone thigh segment, stretching the tissues. The 1 patient in our study (Figure 6(b)), with a knee fusion had normal hip gait kinematics and excellent function, but it should be noted that the proximal femur was normal with a subtrochanteric delay in ossification and significant LLD, so further comment or conclusion cannot be made. The current length of the femoral segment is also too long to enable a knee socket to be incorporated within the prosthesis. However, because reduced growth of the affected limb continues until skeletal maturity, the final length of the thigh segment will be shorter than the opposite normal femur, enabling the patient to be managed as an above-knee amputee.

A Syme ankle disarticulation can be undertaken to allow for a more cosmetic extension prosthesis and provides space for the fitting of a prosthetic knee joint. Kant et al, however, in a retrospective review, suggested that patients who underwent no surgery and used an extension prosthesis had similar levels of mobility but reported fewer problems with the affected limb and the prosthesis itself, recommending a nonsurgical approach.²² Almost all the patients in our study used an extension prosthesis including their foot reporting no ankle pain and during writing, no cosmetic or function concern in relation to the foot. The only Syme disarticulation was performed after a knee fusion, so further comment on outcome cannot be made.

The concept of the Van Nes Rotationplasty³⁸ is popular in North America and parts of Europe^16–20; however, there is less enthusiasm in the United Kingdom. Anecdotally, at our Institution, no patients with congenital lower limb deficiency have accepted this procedure when offered it. Advocates of rotationplasty state that it promotes a more efficient gait pattern,^39–41 but the cosmetic appearance will remain a major factor in decision-making. Furthermore, Floccari et al demonstrated no measurable benefit in patients who had undergone rotationplasty compared with an extension prosthesis incorporating the foot in equinus or after Syme disarticulation.⁴² They demonstrated no benefit in walking speed, gait kinematics, power, oxygen consumption, or patient-reported outcomes. Revision surgery was also required in all cases undergoing rotationplasty, with 67% undergoing greater than 2 de-rotational osteotomies.

In the most severe cases with PFFD where the proximal femur is absent, the question remains as to whether surgery can be beneficial. We demonstrated no significant differences in the functional or QoL scores between the 5 patients with Paley type 3 femurs and the remaining cohort, 2 used an extension prosthesis below the knee and the remaining 3 an ischial-bearing socket. Gait analysis confirmed a stable functional gait pattern in the patient with an ischial-bearing prosthesis. The aim of any surgical intervention is to maintain preoperative limb function as a minimal requirement, with a goal for improvement. In this case, the challenge is to predict confidently an improvement in function after surgical intervention. Small published case studies of iliofemoral fusion conclude that a stable hip can be produced.^23–25 However, there are several significant complications reported including femoral artery thrombosis, varus deformity, or overgrowth due to continued growth of the femoral physis, wound necrosis, and dehiscence (occurring in 52% in a study of 19 patients²⁵). More information is therefore required, such as QoL analysis and gait analysis, to quantify the functional benefit and allow more confidence in recommending such complex procedures.

There are significant limitations of this study. This rare condition results in small numbers of a heterogeneous group of patients for analysis and short follow-up in the cases of children and multiple stakeholders involved with patient care and decision-making, including surgeons, rehabilitative consultants, and prosthetists. Firm conclusions cannot be made from our results, and we accept our interpretation may be biased. Data were collected at a single episode, and there is no preoperative data for comparison to judge the effect of surgery to the individual. The small numbers result in a questionable role for statistical analysis, with trends reported rather than statistical difference demonstrated by our results. There may also be difficulty in the differentiation of functional scores, for example, a highly active patient vs. a moderately active patient recording a similar functional score. The small subgroup of gait analysis patients may not truly represent the biomechanics of the prosthetic limb. The markers are placed in an approximated landmark position, which may be erroneous. Furthermore, the prosthetic foot is designed for shoe wear, and walking barefoot may produce further error in the biomechanical analysis. However, this study does identify which questions to ask and which data to collect to expand our knowledge in the treatment of this complex condition. Surgical restoration of hip anatomy may improve both prosthetic functional scores and gait, and we would recommend consideration of the SUPERhip procedure in appropriate cases. Limited hip extension, compensated for by knee flexion in terminal stance when using an extension prosthesis, may influence surgical strategy, and gait analysis can provide further information on hip and knee function before further surgery. The overall prosthetic and QoL scores do seem to decline in adulthood, but not significantly. Further collaboration is needed to collect sufficient data to discuss the influence of knee movement on resultant hip function and the benefit of undertaking an iliofemoral fusion.

In conclusion, an understanding of femoral development and prediction of LLD offers an ability to recommend surgical intervention and a long-term management strategy, with either a prosthesis or subsequent limb lengthening. Our approach is generally designed with a view to the following key goals: (1) to create a stable hip joint, (2) in cases of significant limb shortening, bring the foot to the ground using an extension prosthesis, (3) with further surgical options including knee fusion and Syme disarticulation to improve prosthetic cosmesis, allow the addition of a prosthetic knee or manage painful or unstable joints. While iliofemoral fusion is theoretically possible, there is currently no evidence of significant benefit of this over an ischial-bearing prosthesis.

Author Contributions

The authors disclosed the following roles as contributors to this article: P.C. Designing the study. Collecting and analyzing the data. Writing and revising the manuscript. A.E. Collecting and analyzing the data. G.C. Collecting and analyzing the data. M.T. Conducting and analyzing gait studies. J.W. Designing the study. Collecting and analyzing the data. Writing and revising the manuscript. D.E. Designing the study. Writing and revising the manuscript. I.S. Designing the study. Collecting the data. Writing and revising the manuscript.

Funding

The authors disclosed that they received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interest

The author disclosed no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article.

Supplemental material

No supplemental digital content is available in this article.