Proximal Femoral Focal Deficiency

Robin C. Crandall, M.D.

I write this article at the request of Eugene Banzinger, who has been after me for some time to write an article about proximal femoral focal deficiency for the ACPOC Newsletter.

I have been involved with a large limb deficiency clinic for 26 years, and am always amazed at the vast differences and variability in treatment of patients with proximal femoral focal deficiency or PFFD. It has been my pleasure to be a member of ACPOC for a number of years. The purpose of this article is to familiarize ACPOC members and other interested parties with diagnosis and treatment options for patients with PFFD. The organization of this article will be as follows:

  1. Etiology of PFFD.

  2. Classification systems.

  3. Treatment options, both surgical and prosthetic.

I write this article as an orthopedic surgeon and much of the emphasis in the article will also be placed on surgical treatment options including doing nothing, foot ablation and prosthetic fitting, lengthening procedures, foot ablation and knee fusing procedures, as well as discussion on rotation-plasty. It should be noted that proximal femoral focal deficiency is one of the most difficult diagnoses to treat in a busy limb deficiency clinic and hopefully the article will clarify questions that the reader may have concerning this most interesting diagnosis. I have attempted to write this article somewhat informally so that it can more easily be understood by therapists and prosthetists and others without a surgical background.

Etiology of Proximal Femoral Focal Deficiency

Generally true PFFD is characterized by marked dysplasia/anaplasia of the proximal femur. This means that the near part or "proximal" part of the femur is poorly formed. It generally represents a sporadic mutation during the time of limb bud formation. Familial cases have been reported where siblings have been affected but to my knowledge no cases of DNA proven monozy-gous twins concordant for true PFFD have been published. A case of monozygous twins with congenital short femur has been reported in the literature. We also have a case in our Twin Cities clinic that is interesting with two children that are proven monozygous identical twins, one child with PFFD, the other not involved. (see also Figure 1 page 5) It is important to know that virtually all children with unilateral proximal femoral focal deficiency have no other organic or visceral malformation and this defect is generally considered an isolated structural or mechanical defect. These patients would normally be expected to carry out normal lifestyles, and other than the need for a prosthesis and possible surgery are very normal members of society. It is however important to know that certain syndromes can occur with PFFD and in particular bilateral PFFD. There is a genetic inheritability to the unusual faces/absent femur syndrome and this has been well described in the literature. A typical appearance of a child with PFFD can be seen in Figure 2 .

Classification of PFFD

Classifying PFFD has been a subject of much debate in the orthopedic literature. Part of the problem is that what is generally thought of as true PFFD is different radiographically from congenital short femur. Complicating this is that severe cases of congenital short femur are treated virtually identical to the various forms of PFFD. In this regard they tend to be "grouped" together as a "femoral reduction deformity". Many of you have probably heard heated discussions at ACPOC regarding the correctness of diagnosis of radiographs showing proximal femoral focal deficiency. It is important however to recognize that even though several radiographic systems of classification exist that one must also treat them based on a clinical presentation and this will also be discussed.

Radiographic Classification Systems of PFFD

Dr. George Aitken was the pioneer of radiographic classification systems for patients with true proximal femoral focal deficiency. The Aitken system is simple and generally very usable for surgeons. Several other classification systems also exist. Whenever I talk to another doctor however about PFFD over the telephone or at an informal meeting, we usually discuss the radiographic appearance classified as in the Aitken system. The Aitken system just seems easier and more user friendly to me.

In the Aitken system there are basically four different types. This includes patients with a hip and patients without a hip. Patients with a hip or a femoral head are classified as either Type A or B and patients without a femoral head are classified as either Types C or D.

Some Group "A" patients can be thought of as having congenital short femur and some of these cases one perhaps would not call true PFFD. Generally speaking, patients with true PFFD have significant varus with dysplasia of the proximal femur and femoral reduction of usually more than 50% of the normal femur. At birth Group A patients can be perplexing as the percentage of ultimate limb length loss is not as apparent as when the child ages. I have seen cases of overaggressive surgery done in patients with PFFD also when one anticipates more shortening that what one ultimately observes as the child ages. Giving the child a few years to mature will generally help in accurate classification both radiographically and clinically.

In Aitken Group B the hip is present within an acetabulum but there is no "firm" connection between the residuum of the distal shaft of the femur and the femoral head. Thus, the patient has a so-called "pseudoarthrosis". The length of the residual femur in this case is highly variable allowing a variety of different treatment options to be entertained. For example, just because a patient is Ait-ken B, this does not mean they are treated the same as another patient with Aitken B hip anatomy. This will be discussed in more detail when we go over the clinical classification systems.

Patients with Aitken Type C have no femoral head but do have what would be considered an acetabu-lum. Again, femoral length varies but in this case the patient does have a truly unstable hip.

In Aitken Type D there is no femoral head or acetabulum. Femoral length again is variable but often clinically significantly shortened due to proximal migration of the residual femur on the hemipelvis. Type D patients may still however have a long enough femur to be treated clinically as a longer type.

Clinical Classification Systems

The clinical classification system as proposed by Robert Gillespie makes a lot of sense to me in particular. In a personal conversation I had with Dr. Leon Krueger he had also proposed this back in the late 1950's. With the clinical classification system you basically react and treat PFFD by assessing how much residual limb is present, in other words, either "short, medium or long". Below is Gillespie's schematic of this, and I will also enclose photos of these different clinical types.


Clinical Type A "long":

Patients with clinical Type A have a significantly longer residual limb. The knee axis is proximal, but the leg length discrepancy may be little enough to realistically allow leg lengthening procedures or various shoe lifts. Generally the child is fitted at a young age with a shoe lift if the foot and ankle are relatively normal and occasionally a non conventional prosthetic fitting can be utilized. (see Figure 3 and Figure 4 ) The foot in clinical Type A may not be affected and usually has a relatively normal axis of motion. Occasionally longitudinal deficiency of the fibula can exist in addition causing an eversion deformity with the foot. The foot is not in the exaggerated external rotation position that is typical with the short forms of PFFD. The longer forms of Group A deficiency also includes patients with congenital short femur and not true PFFD. The treatment in both types however is identical. Young children again generally are fit with noncoventional pylon prostheses.

It may be difficult to tell if a patient is short, medium or long clinically in early age brackets. We generally hold off on surgical decisions until the pathology truly declares itself. I have seen patients for example that have had knee fusion foot ablation done at an early age when actually they probably would have been improved more with just simple foot ablation and then later carrying out leg lengthening procedures. Generally however by age 4 patients with PFFD can be accurately classified both clinically as well as radiographically.

Patients with the longer forms of PFFD, the so-called clinical Group A, generally do very well. Foot ablation with or without femoral lengthening is a frequent surgical option. By removing the foot, these children can readily be fit with a Syme prosthesis and regardless of the axis of the knee; they function essentially as Syme amputees should function with minimal or no limp. "Fine tuning" of the knee axis can be accomplished by physeal arrest of the distal femur on the contralateral side. It is extremely important that the surgeon and surgical team be vigilant as the child grows to not miss opportunities to correct angular deformity or limb length inequalities that may easily be handled by physeal surgical manipulation.

Figure 5

Clinical Group "B", medium length PFFD:

Medium length PFFD patients represent some of the most challenging patients for the orthopedic surgeon and the prosthetist. In medium type clinical PFFD patients the femur is decidedly short, usually more than 75% but the tibia is of relatively normal length. The foot tends to be minimally externally rotated, a distinguishing feature from the "short" PFFD patients. When the patient is asked to move the hip, the residual limb tends to flex and extend in an AP plane, whereas in short forms of PFFD the patient tends to move the residual limb in a lateral plane. The knee axis is near the ipsilateral hip with an ankle distal to the contralateral knee. Leg lengthening procedures are obviously much more involved due to the profound shortening and these procedures are often not a realistic option. Surgical and prosthetic options are generally as follows:

  1. Do nothing and fit a prosthesis alone. Some patients can be satisfactorily fit with Pylons or non conventional pylons with a distally based knee axis. I have also seen a patient in one particular case where the child is able to motor the "knee" of his prosthesis by using the ankle without rotationplasty. Prosthetic appearance however can be awkward and generally pylon prostheses are not well accepted as the child ages and requires knee mobility for normal activities of daily living.

  2. Foot ablation in conjunction with knee fusion: This is one of the most common treatments selected for "medium" PFFD patients. It is also the most common option for the short forms of PFFD. The child is essentially converted to an AK amputee. By fusing the knee the axis of motion is more stable, less laterally deviated and the length of the residual limb can be elegantly adjusted. It is extremely important however that the residual limb length be considered. If the child has a relatively normal tibia then any femur residuum may be too much femur. With a normal tibial length the femoral physis must be eliminated at the time of the knee fusion. One generally wants to end up with the residual limb a little shorter than the contralateral side to fit the prosthetic knee. I have seen several patients who have had foot ablation and later knee fusion and have unfortunately ended up with a residual "femur" that is decidedly too long for good prosthetic fitting. Surgeons need to use the Green-Anderson-Messner charts to keep track of formal residual length in these complex patients. As a ballpark estimate, try not to end up with an ultimate residual limb much greater than 40 cm. Surgeons need to communicate with their prosthetic team about residual length with different prosthetic component systems. As new component systems are available, the prosthetist should also confer with the surgeon as to how much residual limb they need in each particular case. A certain amount of deficit for example is required for consideration of fitting for appropriate C-leg device. Occasionally patients with "medium" clinical length PFFD can be treated effectively with foot ablation alone. This usually occurs when the tibia is particularly short and the knee can be incorporated into the prosthetic socket.

  3. Rotationplasty: Medium length PFFD patients can be nicely treated with rotationplasty. Rotationplasty is a surgical procedure where the ankle is "converted" into a "knee" by turning the foot backwards 180°. Initially described by Borgreve and later Van Nes, it can be accomplished in several ways. The most common technique is to accomplish this by rotating 180° through a knee fusion as described nicely by Dr. Ivan Krajbich. I will not however go into all of the various surgical options as there are several ways to accomplish this goal. I believe the best success however is done via knee fusion when all the muscle tendon units are detached and then reattached 180° opposite point. It is key however that the neurovas-cular structures be fully mobilized so that this can occur with ease. In a personal conversation I had with Dr. Robert Gillespie he felt that most of these rotationplasties should actually be rotated more than 180° anticipating that some derotation can occur. If done properly however derotation is less likely. It is very important that the surgeon present this in a positive way and that appropriate surgical and prosthetic experience be gained for these to be successful. If derotation or under rotation occurs the surgical and prosthetic team must recognize this and correct this. Advice by the prosthetic team here is important. Prosthetic fitting can be very difficult in particular since many of these patients have a "ball and socket" type ankle due to concomitant longitudinal deficiency of the fibula. A good rotationplasty however is a joy to experience and patients with successful rotationplasty are very happy indeed. It is important to realize that rotationplasty patients do consume less oxygen per meter traveled than AK amputee counterparts. Illustrated below is a patient before and after rota-tionplasty.

  4. Clinical Group C patients (short): Patients with "short" PFFD have a markedly shortened femur with the residual limb generally held in the position of external rotation. The ankle axis is at or above the contralateral knee. Rotation-plasty here is not usually a good option because of the ultimate leg length shortening that would be needed. These patients are generally treated by foot ablation with or without knee fusion. Knee fusion usually is needed to medial-ize the residual limb and eliminate the "ginglymus" motion that occurs when the knee and the hip are so close together. The goal here is a stable "AK" amputee. Various hip surgeries may be needed to stabilize the hip, in particular if they have hip anatomy of Aitken B. Prosthetic fitting with this and all forms of PFFD can be difficult and demanding. Soft tissue bulk around the thigh can be particularly vexing and difficulties in fitting the standard "ships funnel" type socket can be very difficult if patients gain weight. We definitely encourage our PFFD patients not to gain weight and this is tracked closely. Patients who have difficulty with bulk of the residual limb have been in the past treated successfully with suction lipectomy as a last resort. Below is patient successfully treated with foot ablation and subsequent knee fusion with short PFFD.

  5. Bilateral PFFD: Patients with bilateralproximal femoral focal deficiency generallyare treated without prosthetic fitting.In some clinics bilateral patients may represent10% of the PFFD patients. Wehave had several of these individuals atour clinic and virtually all of them areable to ambulate without prosthetic fitting.It is important to search for othercongenital deficits. These patients mayhave the unusual facies absent femursyndrome and other associated anomalies can be present. We presented severalcase studies at ACPOC of individualswho developed stress fractures withbilateral PFFD. This problem needs tobe considered in these individuals. Wehave had one patient who was successfullyfit with bilateral pylon prosthesesso as to increase overall height butagain, these were worn primarily in alimited setting. Below is a radiographof a patient with bilateral PFFD thatwe have treated. Most of the patientswith bilateral PFFD are of the shortclinical type, with hip anatomy highlyvariable.

Figure 6 , Figure 7 , Figure 9 , and Figure 10


PFFD is a demanding difficult condition for surgeonsas well as prosthetists and therapists. Of all the patientsin our busy limb deficiency clinic these patientsare our most challenging. Accurate diagnosis of thecondition both radiographically as well as clinically isparamount. It is important that these children betreated incorporating a team approach. Prostheticproblems need to be communicated to the surgeon.Surgeons also need to be highly aware of ultimategoals including adjustment of ultimate residual limblength, stability of gait and if rotationplasty is selectedthat it be accurately and carefully done. ACPOC is ofparticular importance in helping facilitate any neededhelp.

Robin Crandall is Director, Limb Deficiency Clinic Shrine Hospital for Children: Twin Cities Unit