With institutional review board approval, we reviewed the clinical records and radiographs of all patients with a diagnosis of congenital fibular deficiency who were managed at Texas Scottish Rite Hospital for Children between 1971 and 2005. To be included in the study, patients had to have a predominant morphological deformity of fibular deficiency (tibial shortening with variable fibular hypoplasia and foot deformity) and radiographs of both lower extremities that were suitable for limb-length measurement prior to any treatment. Patients who had predominantly femoral congenital limb deficiency, either severe congenital deficiency (Gillespie Type B or C28, Pappas Classes I through V29 [Fig. 1]) or proximal femoral focal deficiency (as classified by Aitken30), were excluded. One hundred and four patients (including twenty-two patients who had bilateral fibular deficiency) with 126 affected extremities met the study criteria. Sixty-five of these patients (seventy-five affected extremities) had achieved skeletal maturity at the time of the review. Twelve patients (including seven with bilateral involvement) were not skeletally mature at the time of review but had been managed definitively with amputation. Thus, seventy-seven patients with ninety-three affected extremities had either achieved skeletal maturity or had been managed definitively with amputation at the time of the review. One of the seven patients with bilateral involvement had unilateral amputation but was not skeletally mature. Demographic information as well as clinical findings were obtained from the medical record. Details of all treatment rendered and clinical function as described in the medical chart were recorded.
Radiographs were assessed for relative femoral and tibial shortening and deformity, length of the fibula relative to the tibia, ankle morphology, position of the foot, and number of metatarsals present. Femoral, tibial, and overall limb lengths were measured from scanograms or full-length radiographs of the lower extremities. When available, serial radiographs were assessed prior to any surgical intervention in patients with preserved feet to evaluate the natural history of limb growth and limb-length inequality.
Source of Funding
There was no external funding for this study.
Overall
The demographic information, general morphological features, and surgical procedures for the 104 patients with 126 affected limbs are summarized in the Appendix. The study group included sixty-nine males and thirty-five females (male:female ratio, 2:1).
Limb-Length Inequality
Total limb-length inequality was determined by measuring the distance from the top of the femoral head to the distal tibial articular surface on scanograms or anteroposterior radiographs of the lower extremities. Absolute femoral, tibial, and total limb lengths were recorded in centimeters and as a percentage of the longer limb. Fifty-seven patients (including eleven with bilateral fibular deficiency) had radiographs made over a minimum two-year period (average, ten years and ten months; range, two years to fifteen years and six months) prior to any treatment. Limb-length inequality, expressed as a percentage of the longer limb, remained proportional (within 5% of the original percentage discrepancy) in forty-seven (82.5%) of fifty-seven patients during follow-up.
Associated Femoral Deformity
There was shortening of the femur in seventy (85.4%) of eighty-two patients with unilateral limb involvement. The presence and extent of femoral shortening increased as overall limb-length inequality increased. The ratio of femoral to tibial shortening had no apparent impact on the final treatment. Coxa valga was noted in twenty-seven limbs, whereas coxa vara was noted in nine.
Length of Fibula
The fibular remnant was >50% of the length of the ipsilateral tibia in sixty-five extremities (52.4%), <50% in six extremities (4.8%), and vestigial or absent in fifty-three extremities (42.7%). Radiographs were inadequate to accurately assess fibular length in two extremities.
All but two of the sixty-five extremities with a fibula measuring >50% of the length of the ipsilateral tibia had preservation of the foot. Three of the six extremities with a fibula measuring <50% of the length of the ipsilateral tibia had preservation of the foot. Twenty-two (41.5%) of the fifty-three extremities with an absent or vestigial fibula had preservation of the foot, including one foot that was preserved to substitute for severe bilateral upper extremity deficiency. Thirty-one (58.5%) of those with an absent or vestigial fibula were treated with amputation.
Foot Deformity
Most feet presented in a plantigrade or equinovalgus position. However, twelve feet had equinovarus deformity, which has been noted by others15.
The number of metatarsals or rays of the foot directly correlated with the rate of foot preservation. Thirty-eight (97.4%) of thirty-nine five or six-rayed feet, thirty (81.1%) of thirty-seven four-rayed feet, and twenty (48.8%) of forty-one three-rayed feet were preserved. Only one of nine feet with fewer than three rays was preserved (see Appendix). In that case, the patient had a two-rayed foot and no upper extremities and did not undergo amputation because the foot was used to substitute for upper extremity prehension. Of fifty-three limbs with an entirely absent fibula, twelve (22.6%) had a five-rayed foot, eleven (20.8%) had a four-rayed foot, twenty-one (39.6%) had a three-rayed foot, and nine (17.0%) had a foot with fewer than three rays. All feet with fewer than three rays in the present study were associated with an entirely absent fibula.
Unilateral and Bilateral Involvement
Of the eighty-two patients with unilateral fibular deficiency, forty-seven (57.3%) had involvement of the right extremity and thirty-five (42.7%) had involvement of the left. Twenty-two patients had bilateral fibular deficiency. Of these, twelve (54.5%) had preservation of both feet, three (13.6%) had preservation of one foot and amputation of one foot, and seven (31.8%) had amputation of both feet. Two of the twelve patients in whom both feet were preserved had at least one foot preserved to substitute for upper extremity prehension.
Upper Extremity Anomalies
Upper extremity anomalies ranging from syndactyly to extensive deficiency were noted in sixteen (15.4%) of the 104 patients. Upper extremity involvement was more frequent in patients with bilateral fibular deficiency (nine [40.9%] of twenty-two) than in those with unilateral fibular deficiency (seven [8.5%] of eighty-two).
Classification
Our review of this patient population validated our concerns regarding the applicability of previous classification systems in determining treatment ultimately adopted for these patients. The most important observations were that only 58.5% of patients with a vestigial or completely absent fibula had an amputation and that ten (45.5%) of twenty-two patients with bilateral congenital fibular deficiency had amputation of one or both feet because of recalcitrant or painful deformity. In addition, seventy-seven (85.6%) of ninety limbs that did not require amputation had associated femoral shortening that potentially required other treatment and therefore had to be taken into consideration during surgical planning. These findings encouraged us to propose a new clinical classification system for congenital fibular deficiency, based on the extent of foot deformity and total limb shortening.
Because we noted a direct correlation between the number of rays present and ultimate preservation of the foot, and because the presence of or the ability to provide a painless, plantigrade foot was key to the decision to preserve a foot (even in patients with bilateral involvement), we chose this morphological aspect of congenital fibular deficiency (a "preservable foot") as the preeminent feature of the classification system. The second consideration is the total limb shortening (femoral and tibial) compared with the contralateral side, expressed as a percentage. As there was little correlation between the extent of fibular preservation and ultimate treatment requirements, the classification system ignores the extent of preservation of the fibula or the radiographic morphology of the tibiotalar joint. Type-1 fibular deficiency is a limb with a foot that has at least three rays and that can provide a stable weight-bearing base for walking, with or without repositional reconstructive procedures. Type-1 congenital fibular deficiency is subclassified on the basis of the percentage of limb-length inequality. Type-2 fibular deficiency is a limb with a foot that is unsuitable for salvage (Fig. 2), irrespective of the extent of limb shortening. This group is subclassified based on the presence or absence of upper extremity deficiency requiring the use of the foot to substitute for upper extremity prehension.
Type-1 fibular deficiency is subdivided into four groups on the basis of the overall percentage limb-length inequality compared with the contralateral side: (1) Type 1A (0% to <6% overall shortening), (2) Type 1B (6% to 10% overall shortening), (3) Type 1C (11% to 30% overall shortening), and (4) Type 1D (>30% overall shortening). These four groups were devised on the basis of expected limb-length inequality treatment requirements. For patients with Type-1A deficiency, the expected range of treatment for limb-length inequality is no treatment, orthotic treatment, or appropriately timed contralateral epiphysiodesis because the expected inequality at maturity is ≤5 cm (6% shortening in a patient with fiftieth-percentile normal limb length would have a limb-length inequality of approximately 5 cm31,32 at skeletal maturity). Type-1B fibular deficiency is defined as 6% to 10% shortening. The greater discrepancy typically requires surgical intervention, either more aggressive contralateral epiphysiodesis or a single-stage lengthening of the affected limb. Type-1C fibular deficiency is defined as 11% to 30% shortening; the more severe shortening in this group mandates at least one lengthening in patients desiring limb preservation or reconstruction. With more extreme shortening, multiple lengthening sessions and possible contralateral epiphysiodesis may be required to equalize limb lengths. Type 1D fibular deficiency is defined as >30% overall lower extremity shortening. Treatment options for this group of patients include extensive reconstruction or, alternatively, the use of an extension orthosis, with or without amputation of the foot.
Patients with Type-2 deficiency are subdivided into two groups. Patients with Type-2A fibular deficiency have affected feet that are not appropriate for salvage and have intact upper extremity function. These patients are best managed with early amputation, irrespective of associated limb shortening or bilateral involvement1,21,33-37. Patients with Type-2B fibular deficiency have bilateral deficiency of the upper extremity and may potentially use a foot to substitute for hand function. In such rare cases, foot ablation must be deferred until substitution patterns for upper extremity function are clearly established.
This classification system provides no separate category for patients with bilateral involvement: each foot is assessed independently. In patients with bilateral involvement in whom the feet can be preserved, the longer limb is by definition a Type-1A limb (<6% shortening) and the shorter limb is classified according to the overall percentage shortening compared with the longer limb. This classification scheme is summarized in Table II.
Comparison of Predicted and Actual Treatment of the Study Population
A total of 101 extremities were classified as Type 1, and twenty-five were classified as Type 2.
Twenty-two (21.8%) of 101 Type-1 limbs changed their percentage shortening subclassification (A, B, C, or D) (average change, 4.9%; range, 0.5% to 15%) over the course of follow-up from the initial radiographs to the latest radiographs made before any surgical treatment (Table III). Forty-six (90.2%) of fifty-one limbs that were classified as Type 1A on initial radiographs remained Type 1A during follow-up, whereas five (9.8%) progressed to Type 1B. Fourteen (58.3%) of twenty-four extremities that were initially classified as Type 1B remained within the 6% to 10% shortening range, whereas nine became Type 1A and one became Type 1C. Of the twenty-four extremities that were initially classified as Type 1C, eighteen (75%) remained Type 1C, four became Type 1B, two became Type 1A, and none progressed to Type 1D. Of the two Type-1D extremities, one remained within the classification range and one became Type 1C. Stated another way, seventy-eight limbs (77.2%) remained in the initial category of percentage limb shortening, sixteen (15.8%) transitioned to a milder discrepancy category during the follow-up period, and seven (6.9%) transitioned to the next more severe shortening category.
Thirty-one (58.5%) of fifty-three extremities with an absent fibula were classified as Type 1 (see Appendix). The remaining twenty-two limbs (41.5%) were classified as Type 2 and ultimately were treated with foot amputation. In the case of one patient with a Type-2B limb in whom the foot was salvaged in order to substitute for upper extremity deficiency, a repositioning osteotomy of the involved lower extremity was performed at the age of six years.
Seventeen (38.6%) of forty-four feet in patients with bilateral involvement were deemed not amenable to salvage. Patients with bilateral involvement and preserved feet had undergone an average of 0.93 (range, zero to three) surgical procedure on extremities with preserved feet during an average duration of follow-up of 11.5 years (range, 7.5 to 15.5 years).
Seventy-seven patients with ninety-three affected extremities, including seventeen patients with bilateral involvement, had been followed to skeletal maturity and/or had been managed definitively with amputation. The predicted and actual treatment categories for this subset of patients are summarized in Table IV.
Of the thirty-three Type-1A extremities, twenty-eight were treated as predicted by our clinical classification (seventeen were treated with epiphysiodesis, whereas eleven had no surgical treatment). Five were treated other than as predicted (three were treated with amputation and two were treated with limb lengthening). Of the three patients with Type-1A involvement who were managed with amputation, one had a painful foot, one had a three-rayed equinovalgus foot that was considered to be too severely deformed to warrant preservation, and one had a four-rayed equinovalgus foot that was removed in an attempted foot-to-forearm transfer for the treatment of acheiria. All three of these patients had bilateral fibular deficiency. Two limbs that were initially classified as Type 1A progressed to Type 1B and were treated with lengthening.
Of the sixteen Type-1B extremities, seven were treated with epiphysiodesis, six were treated with lengthening, and three received no surgical treatment. Of the latter three extremities, one transitioned to a Type-1A deformity and two were in patients who refused treatment.
Of the twenty Type-1C extremities, seven were treated with lengthening. Of the thirteen Type-1C extremities that were not treated as predicted, ten were treated with amputation, two were in patients who refused amputation, and one transitioned to Type-1A deformity and was treated with epiphysiodesis. Of the two patients who refused definitive treatment, one had no surgical treatment and the other underwent epiphysiodesis to ameliorate limb-length inequality. Of the ten Type-1C extremities that were treated with amputation, eight were in patients who were advised to do so because of an absent fibula, one was in a patient who had failed foot centralization, and one was in a patient who elected to undergo amputation because of relatively severe shortening (11%) and a three-rayed valgus foot deformity. The one Type-1D extremity was treated with amputation.
Of the twenty-three Type-2A extremities, twenty-two were treated with amputation and one was in a patient who refused amputation.
The etiology of congenital fibular deficiency remains elusive. While the stereotypical deformity suggests a global lateral lower limb deficiency, the severity of each of these components varies widely from one patient to another. This complex deformity spectrum may be made even more confusing by its frequent association with major congenital femoral deficiency. We specifically excluded patients from our series with the severest forms of femoral deficiency.
The presence, nature, and severity of limb deformities associated with congenital fibular deficiency are important in both the evaluation and treatment of these patients. Specifically, patients for whom limb salvage reconstruction is deemed appropriate almost always have associated femoral shortening and deformity (prevalence, 85.4% in our series), which frequently has an impact on treatment strategies. Associated foot and ankle deformities include equinovalgus, planar or ball-and-socket ankle joint, and equinovarus. These deformities alone did not influence the management decision in the majority of patients but necessitated surgery if the foot was not plantigrade. The presence of tarsal coalition must be noted in feet requiring a surgical correction. The efficacy of the Ponseti method38 in correcting clubfoot deformity associated with congenital fibular deficiency has not been reported, to our knowledge. Reduced foot height may contribute to the overall limb-length inequality, and treating surgeons should be alert to this component of total limb discrepancy when planning limb-length equalization. Because limb-length inequality was assessed purely by means of radiographic review in our study, we were unable to specifically address the contribution of reduced foot height to overall limb-length inequality. Finally, the frequent association of upper extremity limb deficiency must be taken into account when determining lower extremity treatment because, when upper extremity deficiency is present, the deformed foot may be used to replace absent upper extremity function.
The review of the management of our patient population confirmed for us the prognostic deficiencies of the Coventry3, Achterman-Kalamchi1, and Letts18 classification systems. The implication that a vestigial or completely absent fibula was an indicator of an extremity in which the foot could not be salvaged proved not to be true in >40% of the extremities in our series. The presumption that bilateral fibular deficiency implicates a limb-salvage approach also proved to be true in only seven of seventeen skeletally mature patients in our series. Finally, none of these classification systems take into account femoral shortening or deformity, which are important components of the overall deformity and which may influence treatment requirements for limb salvage reconstruction. The Letts classification system does indicate that amputation may be appropriate for projected discrepancies of >10 cm, but the system does not assist in identifying such patients at an early age. Furthermore, advances in limb-lengthening techniques certainly permit the use of limb lengthening for the treatment of inequalities of >10 cm in selected cases.
The proposed clinical classification system for congenital fibular deficiency takes into account first the functionality of affected feet and second the total limb-length inequality. We chose to express limb-length inequality as a percentage shortening for several reasons. First, as the majority of patients in the present study maintained the same percentage shortening during growth prior to any treatment (as has been noted by others11), ultimate shortening is more easily extrapolated from radiographs of the infant or young child with this method. Second, a gait analysis study involving otherwise normal patients with limb-length inequality demonstrated that correlations between adaptive strategies and shortening magnitude could only be made by classifying the subjects according to percentage shortening39. Finally, several studies have demonstrated a notable increase in the rate of limb-lengthening complications when a limb segment is lengthened more than 15% to 20%22,40-48.
Our proposed classification system has several deficiencies. First, the definition of a "preservable foot" is arbitrary and subjective. We adopted the parameter of "three-rayed foot" for a simple reason: in our review of these patients, most five and four-rayed feet were preserved, whereas approximately half (twenty) of forty-one three-rayed feet and only one of nine feet with two or fewer rays was preserved. We do not suggest that all five-rayed feet must be preserved or that no two-rayed foot should be preserved. In this retrospective clinical series, three-rayed feet represented the boundary between foot preservation and foot ablation. We believe that the surgeon and family must decide on a course of treatment for the affected infant. In our institution, which is experienced in both limb preservation and amputation and prosthetics, foot ablation was more strongly considered and adopted when the foot had three rays or fewer given the complexity and duration of preservation reconstruction with severe deformity. Similarly, we cannot and should not establish absolute percentage shortening parameters above or below which lengthening is not indicated. Not long ago, limb lengthening was considered to be indicated when amputation was the only acceptable alternative49. Advances in limb-lengthening techniques and outcomes have allowed for modification of this approach. Future advances in limb lengthening undoubtedly will continue to extend the indications for limb lengthening at both ends of the spectrum of shortening. Even in our clinic currently, however, some parents of children who are predicted to require one or more lengthening procedures but who are within reasonable limb preservation and lengthening parameters elect foot amputation. The respective roles of limb preservation/reconstruction and foot ablation/prosthesis remain controversial as emphasized by the long-term amputation versus limb preservation outcome studies of Walker et al.27 and McCarthy et al.45 and likely will change with advancing reconstructive techniques. The purpose of this classification is to provide both the treating surgeon and the family with a sense of the total discrepancy to be overcome if limb-length equality is to be achieved and an appreciation that a substantial amount of the shortening to be overcome may reside in the femur. Although we were unable to assess final functional outcomes resulting from our treatment approach to fibular deficiency, our clinical classification system has proven to be useful for determining the approximate anticipated extent and nature of reconstructive treatment and helpful for guiding both the treating surgeon and the families toward a treatment strategy that is appropriate for the family and child.