Clinical Biomechanics of the Partial Foot

Gerald Stark, MSEM, CPO/L, FAAOP

Partial Foot Prosthetic Challenges

Partial foot amputation presents with a variety of unique prosthetic challenges for the clinician. One is the variety of amputation levels included under the same classification such as metatarsalphalangeal (MTP), transmetatarsal (TMA), tarsometatarsal or Lisfranc, and intertarsal or Chopart not to mention the various individual presentations of toe amputations, ray resections, and surgical variations. The informed clinician needs basic understanding of the functional differences and considerations for all these levels with regard to skeletal anatomy, muscular involvement, and load bearing capability.

Partial foot amputation is often the effect of diabetes mellitus which compromises not only ambulation, but also the general health and physical stamina of the patient. Patients with diabetes have 15-46 times greater risk for amputation than those without diabetes. 16 These patients walk 66% 15 as fast as normal and score poorly on a number tests that measure function and physical performance, 4,6 The average life span of a patient with TMA is 34 months as opposed to non-diabetics with a life span of 1, 2, and 11 years. 8 Diabetes mellitus compromises load bearing capability of the already reduced plantar surface increasing the risk of ulceration and subsequent amputation. Over a 5 year period 27% of TMA's developed skin breakdown and 28% required higher amputation level. 15 These patients appear to walk much more tentatively, loading the foot only to the distal end 40% the length of shoe 15 with a greatly shortened stance phase (32%GC vs. 60% GC) and increased double support that does not utilize a late stance push off. These patients exhibit an increased hip moment to pickup the foot and bring it forward. This shortened gait also results in increased load on the uninvolved limb, increasing the risk of ulceration and possibly amputation.

Prosthetists, orthotists, and pedorthists often approach management of the partial foot prosthesis very differently, often relying on clinical beliefs not proven in recent evidence that takes into question the common practice of allowing motion at the ankle and loading a forefoot deflection plate. 18,19,20 Evidence has shown that once the metatarsal heads have been compromised, regardless of level, the amputee greatly limits use of the ankle, even if the range of motion and plantarflexor strength is available. To load the forefoot and increase stance phase a more rigid clamshell design with an anterior pretibial shell is required that increases the coupling between the leg and the prosthesis and lengthening effective anterior proximal lever arm. It is ironic that the A.A. Marks Manual of Artificial Limbs in 1914 warns against flexible supramalleolar designs or articulated ankle designs as "ill advised" for amputations of the "instep". It refers to a flexible boot with a cork filler and steel shank as giving a natural appearance, but failing to support the user in a "helpful or substantial way". A semi-articulated design is criticized as "insecure" and "not effective in holding the appliance at all times and checking the action of the ankle at all times." Gait characteristics are negatively described, "When weight is applied... the wearer will walk flat footed and will press the delicate cicatrized surface against the attachment. These conditions will not only cause suffering but defeat the object of the artificial foot." The design that is recommended is a rigid anterior metal prosthesis with a flexible posterior leather lacer. This design allows "weight on the ball of the foot similar to the position taken by the natural foot when in the act of throwing the body forward." 21

Functional Surgical Presentation

The presentation of the residual limb of the partial foot can be fairly individual depending on the surgical involvement and the need for viable tissue for healing. Diabetes is the most common contributing factor with the chance of another amputation on the same extremity or contralateral limb in 4 years as 50%. 11 Surgeons may choose more proximal amputations because of dissatisfaction at the partial foot level with varied success rates at 93.3%, 68%, 83.3% and 50%. 11 Typically more proximal amputations exhibit less ROM especially in dorsiflexion and eversion with a greater degree of equinovarus unless special surgical procedures are employed. This is due to the muscular imbalance present because the major inverters and plantarflexors remain intact at the midfoot while the everters and dorsiflexors are compromised with more distal attachements. This may not be true equinas but the lack of dynamic dorsiflexion may be described as "functional equinas". 11 Often the anterior of the Chopart residuum is confused because muscular atrophy and the high degree of equinas.

Metatarsalphalangeal Amputation

In gait, the toes function to flex in late stance phase to broaden the area of support as the foot rises during heel off and toe off and carry support in rollover the length of the foot. Toes contribute to balance and weight bearing. Dysfunction or absence contributes to fatigue and proximal tissue breakdown. 8 A majority of the load is transferred to the great hallux as the foot externally rotates. If the hallux is amputated, there is a lateral shift of the of center of pressure from between the second metatarsal head to the third. 13 Peak pressures on the first metatarsal increased 27% more with the lesser metatarsals having 40% more pressure. 13 In one study 63% of diabetic patients with great toe amputation developed ulceration and 53% had subsequent amputation of the same foot. 13 The first metatarsal head should be retained to avoid disruption of the insertion of the plantar aponeurosis and intrinsic muscles that stabilize the metatarsal head. Perservation of the metatarsal heads appears to be crucial in maintaining the windlass effect for successful forward progression. Removal of the second toe may result in a hallux valgus deformity. A ray resection of the second toe brings the third toe into support of the first. Ray resections or metatarsal head fusion of surrounding metatarsals results in more normal load pattern although it is not commonly used. 3 Time of peak dorsiflexion for toe amputees was significantly delayed and terminal stance ankle moment and it timing is different. 4MTP amputations had little impact on power generation at the ankle. 18 Because these patients exhibit little change in stance phase duration and velocity, a simple prosthesis constructed of a stiff toe plate, custom molded insert and toe filler to help with medial support at toe off.

Transmetatarsal Amputation

Transmetatarsal amputation is a transection of the metatatarsal shafts. In 1991 there were approximately 10,000 TMAs. 5 It is considered when two or more medial or more then one interior ray must be removed. 22 As much of the metatarsal shafts are preserved with a 15° transverse angle paralleling the metatarsalpha-langeal toe break. 22 The success rates are varied from 50%, 53%, 91%, and 81% depending on folio w-up and patient group.22 Although it would seem "mechanically intuitive" to retain a longer lever arm, research shows major functional compromise at the transmetatarsal level.

Transmetatarsal amputation effectively removes the distal attachments of the dorsiflexors and partially disrupts the effectiveness of the peroneus longus, but maintains the attachments of the prime inverters of the tibialis anterior and posterior. At this and proximal levels a medial midfoot posting helps to support the remaining longitudinal arch to assist in the support of the transverse tarsal joint. 11 Absence of the metatarsal heads severs the plantar aponeurosis, intrinsic foot muscles and toe propriocepters making forefoot weight shift painful and difficult. 1,10The stability of the forefoot is diminished since the head of the fifth metatarsal which stabilizes the lateral column, has been removed. 8 This results in midfoot instability since distal stability cannot be achieved. The mid tarsal locking mechanism is compromised and abnormal shear increases hypermobility of weight bearing bones creating a poor lever for push off. 8 As a result the ability of the TMA to generate effective ankle moment, work, and power are compromised to the same level as more proximal levels of Lisfranc and Chopart. 18'19>2° The prosthetic solutions that endeavor to allow ankle motion are ineffective since the patient does not exhibit the ability to shift the load past the distal end to the prosthetic position of the metatarsal head. 18,19,20Because the patient does not utilize ankle motion, the prosthetic solution lies in rigid anterior pretibial shell that locks the ankle and utilizes the proximal anterior lever to transfer load anterior.

Tarsometatarsal Disarticulation (Lisfranc)

Tarsometatarsal disarticulation first described by the Napoleonic surgeon Jacque Lisfranc de St. Martin, represents significant loss of the forefoot lever. With unbalanced musculature to check the massive plantarflexors the TMT amputee can easily fall into an equinovarus position. Bowker describes surgical methods which address the equinovarus position and transverse stability by transferring the peroneous longus anterior tibialis to the medial cuneiform, preserving the base of the second metatarsal to preserve the transverse arch, and leaving the base of the fifth to maintain the peroneus brevis. 22 The equinas position can be rectified with Achilles tendon lengthening and cast immobilization. 22 These surgical correction are necessary to insure the more general loading of the plantar surface and bare foot walking capability to avoid increased localized loading latero-distal in late stance. The prosthetic solution is to lock the ankle into a clamshell design with a removable posterior door. The fit in the socket must be intimate insuring maximum coupling or connection to the prosthesis. The anterior trimline must not terminate at the tibial shafts otherwise localized loading will occur. The pressure gradient is least at midshaft and greatest at the trimline with the center of the force more proximal.

Midtarsal Distarticulation (Chopart)

The Chopart or midtarsal disarticulation is made through the talonavicular and calcaneocuboid joints normally utilized to lock and unlock the midfoot based on tibial rotation and subtalar position. It is primarily used for trauma and tumor, but rarely used for diabetic foot amputations because of its proximity to the heel pad. 22 At this level all dorsiflexors are transected and the plantarflexors can pull the foot into severe equinas. This can result in a painful limb as the load transfers anteriorly to the sharp articular surfaces of talus and calcaneous that should be trimmed and smoothed. Bowker and Marquardt describe methods to restore dorsiflexion by transferring the anterior tibial tendon to the talus with Achilles tendon release. 22 Baumgartner advocates a post-operative external fixation of the talus and calcaneus to avoid what he considers inevitable equinas. The bony profile shaped and contoured to be conducive to comfortable load bearing. 22 Several surgeons advocate arthordesis of the subtalar and talocrural joint to eliminate the severe equinas deformity. 22 Again a rigid prosthesis is advocated. Barefoot walking can be more comfortable and practical than the Symes's amputation which requires the prosthesis even for short distances. With the Chopart prosthesis the patient can fully load bear at heel strike, but depending on the surgical technique may have pain late in stance. The load must be transferred to the rigid anterior pretibial shell and be firmly coupled to prevent A-P migration during anterior foot progression.

Partial Foot Amputees are Tentative in Gait

Recent evidence shows portrays partial foot amputee as much more tentative gait and general physical health. Patients with diabetes and Transmetatarsal amputation in general performed consistently lower on simple measures of physical ability including the 40% less on the Functional Reach Test 4,6 22% less with the Physical Perfomance Test 4,6 higher with the Sickness Impact Profile (17.0±12.7 vs 3.7±4.7) 4,6 and lower gait speed of 51.5+/- 13.2 vs. 75.6+/-9.3 m/min or 68%. 1 Unilateral partial foot amputees walked at roughly 66% the velocity of normal with little difference between level. 20 Significant reductions in stride length, cadence, and step length contributed to the reduction of velocity with reduction in stride length being a major contributor.20 All groups of amputees, including toe amputees, demonstrated altered forefoot rocker mechanics with reduced gait velocity when walking barefoot compared to normal 61.8% for one or more toes, 56.8% for all toes, 62.3% or ray resections.4 Disruption of the forefoot substantially reduced the dorsiflexion moment and delayed the timing of peak ankle moment and motion contributes to sound side loading. 4 Mueller correlated this slowed walking speed for TMA at 68% 6,10 Dillon further describes the reduction TMT 1.18±.04 ms, Cho-part 1.18±.03 m/s as well as bilateral Lisfranc ,92±,02 cp, compared to normal 1.41-1.71 m/s with reductions in velocity correlated with reductions in stride length and cadence. 20 The amount of single limb support on the affected side decreased for TMT-38% GC, Chopart-37%GC, Lisfranc 37±0% GC compared to normal. 38%-41% GC. 20 Conversely duration of double support increased for unilateral Chopart increased 13% GC compared to normal 9%-12%GC as well as the unilateral TMT and Bilateral Lisfranc 14±1%, 20 This evidence would seem to suggest a very limited time spent on the involved residual limb eliminated heel to toe gait. The involved limb would appear to used only for balance in a step-to gait with very short stride lengths regardless of level. This gait pattern may be due to the compromised physical condition of the patient or a protective strategy for reduction of pain and pressure.

This limited stance phase also increases loading to the sound side also at risk for ulceration and amputation. Low ankle power absorption in complete loss of the toes may increase sound side loading. 4 Bumfield notes that higher vertical loading recorded to the sound side with TMA places greater demand secondary amputation this becomes a risk. This was less prevalent for patients with toe amputation who could generate greater plan-tarflexion torque. 9

Delayed Center of Pressure and Shortened Stance Phase

As mentioned the probability of ulceration and subsequent amputation is higher with TMA. Skin breakdown to high pressures are also found in various patient populations with peripheral neuropathy including diabetes, Hansen's, Charcot-Marie-Tooth, alcoholism, myelomeninocele. 15 Skin breakdown with wound failure or higher amputation can arise in 17-44% of patients with TMA with an average rate of ~30% 17 Another study by Kelly followed patient history over a 5 year period and found 27% developed skin breakdown and 28% required a higher level amputation.15 Another study observed that once a patient has had a amputation they were 36 times more likely to ulcerate than subjects without amputation history as a result of peak pressures being elevated. 16 This could be due to the limited dorsiflexion found with many amputees who exhibit a certain amount of "functional equinas" with a reduction in ankle dorsiflexion peak in TMA4 although Dillon and Garabolsa did measure static ankle ROM that was statistically different. 20,11 Most authors agree that there is a significant delay of the peak dorsiflexion angle compared to normal.

Peak plantar pressures were measured as 28% higher in one study (80.0±31.1 N/cm3 vs. 62.5±21.0 N/cm3) 16 and Kelly notes a 16% 15 increase although it did not seem to be statistically significant. After great toe amputation forefoot pressure was measures at 58+/-20 psi, 39%+/-8% higher on the amputated side. 7 Peak force is reached earlier in stance phase at 68.5%±6.5% vs 80.2%±5.1%. 15 Dillon notes that with toe amputees the CoP progressed relatively normally along the length until loading response then did not move distally along the length of the foot normally. 19,20 This was more pronounced when metatarsal heads have been amputated with the Transmetatarsal amputation. The ground reaction force did not continue to progress, but remained well behind the distal end 40% 20 or 58%-65% 19 of the shoe length. At 45% GC the largest GRF occurred with COP well behind end of residual limb and did not progress until after double support. CoP did extend entire length of shoe. 19,20 In Chopart amputees with a more rigid prosthesis the CoP was able to move beyond the distal end to apply a loading to the forefoot. 19,20 This implies that the patient using a toe filler or slipper type prosthesis is not effectively utilizing anterior deflection plates typically utilized for prosthetic pushoff. The superior coupling of the Chopart type prosthesis is able to transfer a moderate amount of force to the toe plate better approximating normal late stance rocker motion. The second vertical GRF peak was delayed on both limbs of the unilateral Chopart 35±1%GC and bilateral Chopart 33±1%GC compared to normal 23% GC to 31%GC. This delay of the second peak was more significant with the bilateral Lisfranc 31±2% GC and sound side of TMT 31%. Magnitude was significant on affected limb of unilateral TMT 9.95 -9.51 N/kg compared to normal 10.04 to 11.96N/ kg. The reduced magnitude of the GRF of third vertical force peak was significant in the affected limb of the TMT 10.2N/kh. In all there is a relatively little reduction in the vertical GRF except if adjusted for velocity. There is a delay in the peak value for TMT. The peak GRF was increased for the sound side for the MTP and TMT levels.

The patterns of the horizontal shear forces ground reaction force and center of pressure were similar with differences in timing and peak loads. The timing of the first horizonatal shear force was delay in the bilateral Lisfranc 14±1%GC as well as the unilateral Chopart 16% GC compared to normal 9%GC to 12%GC. The magnitude of the shear force was also smaller in the bilateral Lisfranc (-1.36±.12N/kg) compared to normal-1.56N/kg to -2.63N/kg). The bilateral chopart the magnitude was substantially smaller for left limb (-.80 N/kg) with no clear peak force for the right limb. 20 The second horizontal GRF peak was earlier thatn normal for the TMT (48%) and Lisfranc 48%±3% comared to normal 50%-54% GC. The magnitude of the second shear force was much smaller in the bilateral Lisfranc amputee 1.21±.03 N/Kg compared to normal 1.65-2.66 N/kg. 20 The real significance is how low these shear values are especially with the Lisfranc, TMT and how normalized they are for MTP and Chopart. This would indicate along with a shortened stride length that the GRF is more vertical without normal horizontal component. This would be present in a gait that is exhibits more "touch and go" ambulation and the foot is picked up by the hip musculature before a rollover is present.

Motion, Moments, Work, and Power

Ankle Motion and Moments

The ability of the patient to actively move the ankle and apply force has a direct impact on peak loads, CoP location, forefoot loading and timing of gait. Dillon found ankle plantarflexion was reduced in bilateral Lisfrance 21°+/-1°, Unilateral Chopart 20°, bilateral Chopart 30°+/-3° compared to normal 32° to 60°. Significant on TMT 36° Lisfranc 35°+/- 10° but these were not considered gait limitations. 20 Mueller records TMA with DM that showed decreased plantarflexion -3.7°+/-9.7° vs 5.9°+/-5.9° (or 9.6° less.) 6 Dorsiflexion range was also limited but not to significant values TMT 10°, Lisfranc 12° +/-5° and Chopart 8° compared to normal population of 6°-18°. Dorsiflexion range reduced on sound limb TMT 5° Lisfranc 12°+/-8° and Chopart 8°. 20

Ankle inversion/eversion was also measured by Dillon. This would suggest little functional compromise of plantar/dorsiflexion, but could be the result of superior surgical technique. Typically partial foot amputees exhibit increasing equinas with more proximal amputations and stiffened if not limited dorsiflexion that cannot achieve neutral positioning. This is due to the muscular imbalance of the intact primary inverters, tibialis anterior and posterior with large plantarflexor group unopposed by the primary dorsiflexors and everters which are often transected. Prothetists still observe breakdown in custom foot wear because the impression at made 90°, but neutral position of foot is probably not the same during peak plantar pressure during walking but dorsiflexed. 15 Ankle inversion range significantly compromised in bilateral Lisfranc 6+/-2° and bilateral Chopart 2±0° and unilateral Chopart 8° compared to normal 12°- 31°. Eversion range was more affected with bilateral Lisfranc 2±0° Chopart 2±0° TMT 5° Lisfranc 2°±2° and Chopart 2° compared to normal 7°-15°. 20 This presence of inversion and lack of functional eversion must be addressed in the forefoot with a medial posting to support the midfoot that easily falls into a varus presentation distally. If the medial posting is not provided the patient will fall medially and the load progresses anterior albeit just to the distal end of the limb. At initial contact the MTP (5.38±1.00°)and Lisfranc (5.85±3.39) amputees exhibited dorsiflexion as opposed to normal -13.15°-3.67°. The role of this excessive dorsiflexion is not clear, but it could serve to initiate the first rocker and stimulate knee flexion. The timing of plantarflexion was delayed on the unilateral and bilateral Chopart 12%GC compared to normal 5-9%GC. With a plantarflexion peak plantarflexion angle of -5.06±1.45° the bilateral MTP showed initial plantarflexion peak of - 3.28±3.14°. 20 The Chopart rigidity is due to the security of the prosthesis. The more neutral presentation of the TMT initially shows a slight tendency load the limb in the initial rocker.

During gait peak dorsiflexion was delayed on the affected limbs of all subjects except the bilateral MTP. 20 The unilateral TMT 51±0%GC, Lisfranc 51±2%, unilateral Chopart 50± 0% GC, bilateral Lisfranc 55±5% and Chopart 54±0%GC compared to normal 40-49%GC. 20 This correlates to the delay in the progression of the Cop in limiting the amount of horizontal shear. The magnitude of dorsiflexion was larger than normal, 4.17°-12.17° for the bilateral MTP 17.27±2.43° amputee and involved TMT 13.91° and Lisfranc 14.00±2.93°. 20 The relationship between the increase in dorsiflexion peak and delay is the advent of double support permitting shared loading with the sound side. It is evident that the residual limb is being unloaded prematurely before the CoP progresses anterior. The Chopart as unilateral and bilateral showed less peak dorsi or plantarflexion do to its locked down position in the prosthesis.

Significant reductions in plantarflexion angle at toeoff of TMT ( -4.20°), Lisfranc (-.93±.60°), Chopart (-4.89°), as well as bilateral MTP (-2.27±2.14°), bilateral Lisfranc (1.28±8.42°) and bilateral Chopart (1.76±1.17°). Compared to normal -15.95°to -32.35°. 20 Reductions in maximum plantarflexion were significant in the unilateral Chopart (-9.25), unilateral Lisfranc (-9.22±4.59°), and bilateral MTP (-9.65±1.88°) bilateral Lisfranc (-11.53±1.28°) and bilateral Chopart (-4.00±1.67°). 20 Plantarflexion delays were found in the bilateral Lisfranc, Chopart, and also TMT and unilateral Chopart. 20 The most significant is the MTP and TMT limiting maximum plantarflexion. Even though they are afforded the motion these amputees are not using this for forward progression and loading of the anterior toe. A dramatic reduction of the plantarflexion moment peak was reduced in magnitude from normal of 1.49 Nm/kg to 1.95 Nm/kg. Reductions of the Lisfranc (.85±.30Nm/kg), TMT (.85 Nm/kg) and both bilateral Lisfranc (.59±.22 Nm/kg) 20 were fairly consistent despite level, again indicating the amount of force applied in plantarflexion was less even when it was available.

Knee Motion and Moments Knee motion was reported as very similar to the normal motion. Excessive knee motion was found for the bilateral MTP at initial contact and at stance flexion. Substantial reductions in stance flexion were found in the TMT (6.55°) and bilateral Chopart (8.00±2.42°) compared to normal 10.82° to 22.71°. 20 Excessive stance phase flexion was found in the unilateral Lisfranc of 24.62°. 20 The timing of stance flexion was delayed slightly in the bialateral Chopart (19±1%GC compared to 13%GC to 17%GC). 20 The main comparison again is the lack of stance flexion in the TMT possibly limiting the first rocker at initial contact or the effect of slower walking.

Knee hyperextension was observed with the unilateral and bilateral Chopart prosthesis which delayed the initiation of knee flexion into swing phase. This was probably a result of the construction of the clamshell design placing a load gradient along the rigid anterior shell. Maximum knee flexion was delayed slightly in the bilateral Lisfranc (75±0% GC) and the sound limb of the TMT amputee (74% GC) compared to normal 70-74%GC although not significant. 20 Pathologically patients utilize knee hyperextension to lock out the knee for stability and eliminate the possibility for stumbling late in stance.

Normal knee moments were of the sound side remained normal for unilateral involvement. The maximum extension moment was slightly delayed for bilateral Lisfranc and bilateral Chopart. The magnitude of maximum extension was noted for the unilateral TMT and Chopart. The peak knee flexion of the unilateral TMT and bilateral Lisfranc was also affected without characteristic peaks adding to the contention that the patient is not using available motion to apply force.

Hip Motion and Moments Hip motion also resembled normal gait. For all affected limbs except bilateral MTP and unilateral Chopart the hip extended right after heel contact. Hip extension was seen as premature of the unilateral Lisfranc, bilateral Chopart. Hip extension was delayed on sound side of the TMT, unilateral Lisfranc. The hip flexion extension angle at toe off was less than the normal of (-10.63° to 4.39°) of the unilateral Chopart (-14.42±0°) and larger with Lisfranc amputee (7.07±1.85°). 20 Altered hip motion may be the result of stride length regulation to even the short stride of the affected side.

Hip moment peak was most notable in that many subjects maintained extension moments well into stance phase until late gait. Possibly preserving stability of the heel. The hip flexion peak was well defined for the normal and amputee groups but delayed for both the sound side and affected limbs of the unilateral TMT amputees and similar delays in the bilateral Lisfranc and Chopart. This could be the result of the hip flexion moment serving to bring the leg forward in late stance instead of the normal propulsive loading of the forefoot.

Lower Leg Power Absorption

Power represents the rate of change of work. It represents energy storage/absorption as negative, the energy return or push as positive at the ankle, knee, and hip. Power generation on the sound side was comparable to normal population. Peak ankle power absorption was not well defined in normal or amputee subjects, but was somewhat delayed on the affected limbs of TMT and Lisfranc amputees. The power generation peak associated with push off was delayed on the sound limb of the TMT 57%GC and both limbs of the Lisfranc amputee 59±4% GC compared to normal 52- 55%GC. When the metatarsal heads were amputated the magnitude of the propulsive peak was significantly reduced in the unilateral TMT the unilateral amputee (.72W/kg) and Lisfranc (.91±.39W/kg) as well as bilateral Lisfranc (.41±.41W/kg) compared to normal (2.56-5.06 W/kg). 20 The MTP amputee however traced the lower boundary of power generation but not dramatically evident. This relationship definitively shows the role of active pushoff with the metatarsal head and the lack of push off once the metatarsal heads have been compromised. The reduction of power does not explain the shorter stride length and walking velocity. Chopart amputees also should a decreased power factor, but this was to be expected because of the general locking of the ankle within the prosthesis.

In regard to power absorption at the knee there was not many noticeable changes. Significantly less power absorption was seen with the affected side of TMT and Chopart amputee. Reduction in initial power absorption was observed with the bilateral Lisfranc and Chopart amputees. Normal power requirements were seen late in stance. Normal power requirements were seen with respect to the sound side, but one would expect to see more noticeable power requirements during double support when loading should be increased.

Hip powers were substantially different and variable with both the sound and affected limbs. The first peak was not defined well for involved or sound limbs. The power generated on the sound limb of the unilateral TMT, Chopart, and one Lisfranc subject was substantially larger for propulsive requirements. Similar power generation was observed on both the sound and affected limbs by the hip flexors with slight delays of the sound limbs of unilateral TMT, unilateral Lisfranc, and bilateral Lisfranc. The magnitude of the hip flexors was slightly higher than average although comparable to sound and involved sides. Dillon observes, "The Hip Joint not the ankle becomes the primary source of power for walking increased work across the sound hip joint provided forward impulse for the pelvis commensurate with the limited power generation across the affected ankle." 18

Prosthetic Objectives

Friedmann outlines the purpose of the foot as:

  1. Support Body Weight,
  2. Assists in initial Shock absorption,
  3. Allows internal rotation of hip in closed chain kinetics,
  4. Acts to stabilize upper body forces during midstance,
  5. Provides rigid lever for effective and efficient propulsion,
  6. Adapts to supporting Surface. 8

Kulkarni admits that a "partial foot prosthesis that is structurally strong protects the extremity from shock at heel strike and toe off controls plantar flexion and dorsiflexion and rotation, is cosmetically acceptable and stimulates normal gait patterns is a near ideal but in reality is a difficult solution". 14 A Silicone boot type prosthesis emphasizes intimate fit with coupling the limb to the prosthesis by emphasizing total contact socket with good containment of soft/bony structures of residual limb. 14 Armstrong speaks of three main extrinsic mechanisms of tissue breakdown: "Very high force over small area for a short time interval, constant low force over prolonged period of time, and repetitive moderate force over a prolonged time leading to inflammation and enzymatic autolysis or ambulation leading to breakdown." 16

Dillon criticizes the use of "toe fillers, foot orthosis, slipper sockets to TMT and Lisfranc were unable to restore 'effective' length. These did not rigidly couple the residual limb and prosthesis as the clamshell did." 19 He further elaborates, "sockets, and AFO's that allow unrestricted ankle motion did not improve ankle function compared to clamshell patellar tendon bearing prosthesis. 18 AFO designs are also criticized by Mueller as not being tolerated well or enhancing function and by blocking ankle movement. He suggests that they may be used in instances of drop foot or coronal plane instability. 5

Based on the evidence the prosthetic objectives can be refined for partial foot emphasizing stability and protection:

  1. Provide intimate surface matching for total surface loading, socket coupling, and protection of the distal residuum from distal end vertical forces
  2. Utilize rigid ankle design and forefoot lever when metatarsal head has been compromised to encourage toe loading and lengthen stride length
  3. Accommodate varus presentation of midfoot and/or forefoot remnant with medial posting of longitudinal arch if necessary.
  4. Employ heel and toe designs that minimize horizontal shear forces at heel strike and heel off.
  5. Provide proximal coronal plane stability to aid in balance during single limb support when necessary.

With new evidence on the actual gait characteristics of the partial foot amputee, old beliefs can be updated, integrating a more complete biomechanic picture with a more defined prosthetic goal in mind.

Photo 1 | Photo 2 | Photo 3

The Fillauer Companies, Inc. Chattanooga, Tennessee


    1. Muller, M., Functional Limitations in Patients With Diabetes and Transmetatarsal Amputations, Physical Therapy , Volume 77, Number 9, September 1997.


    1. Pinzur, M., Gold, J., Schwarz, D., Gross, N., Energy Demands for Walking in Dysvscular Amputees as Related to the Level of Amputation, Orthopedics , Volume 15, Issue 9, 1992 pp. 1033-6 discussion 1036 -7.


    1. Ramseier, L.E., Jacob, H.A.C., Exner, G.U., Foot function After Ray Resection for Malignant Tumors of the Phalanges and Metatarsals, Foot & Ankle International, Volume 25, No. 2 February, 2004.


    1. Boyd, L., Rao, S., Burnfield, J., Mulroy, S., Perry, J., Forefoot Rocker Mechanics in Individuals with Partial Foot Amputation, Jaquelin Perry Pathokinesiology Laboratory, Rancho Los Amigos Medical Center, Downey California. Funded by the Department of Veteran's Affairs (#A861R)


    1. Mueller, M., Strube, M.,,Therapeutic Footwear: Enhanced Function in People with Diabetes and Transmetatarsal Amputation., Archives of Physical Medicine Rehabilitation , Vol. 78, September 1997.


    1. Mueller, M., Salsich, G., Bastian, A., Differences in the gait characteristics of people with diabetes and transmetatarsal amputation compared with age-matched controls., Gait and Posture , Vol. 7, 1998 pp. 200-2006.


    1. Randolph, A., Wynn, T., Abayev, B., Nelson, M., Alexandrescu, R., Hatta, J., Foot pressures After Great Toe Amputation in Diabetic Patients: A Pilot Study., Archives of Physical Medicine and Rehabilitation , Vol. 83, November 2002.


    1. Friedmann, L., Padula, P., Weiss, J., Root, B., Polchaninoff, M., Sapiro, D., Studies on the survival of Transmetatarsal Amputation Stumps, Symposium Proceedings, 34th Annual meeting, American College of Angiology, Paradise Island, Bahamas, October 1987.


    1. Burnfield, J., Boyd, L., Rao, S., Mulroy, S., Perry, J., The Effect of Partial foot Amputation on Sound Limb Loading Force during Barefoot Walking., Abstracts/ Gait & Posture 7 (1998) 144-190.


    1. Mueller, M., Salsich, G., Strube, M., Functional Limitations in Patients with Diabetes and Transmetatarsal Amputations, Physical Therapy , Volume 77, Number 9, September 1997.


    1. Garbolosa, J., Cavanaugh, P, Wu, G., Ulbrecht, J., Becker, M., Alexander, I., Campbell, J., Foot Function in Diabetic Patients after Partial Amputation, Foot & Ankle International , Volume 17 (1), January 1996, pp 43-48.


    1. Hirsch, G., McBride, M., Murray, D. D., Sanderson, D., Dukes, I., Menard, M., Chopart Prosthesis and Semirigid Foot Orthosis in Traumatic Forefoot Amputation., American Journal of Physical Medicine and Rehabilitation , Number 11, 1996.


    1. Lavery, L., Lavery, D., Quebedeax- Farnham, T., Increased Foot Pressures After Great Toe Amputation in Diabetes, Department of Orthopedics, Mexican American Medical Treatment Effectiveness Research Center, University of Texas Health Science Center, San Antonio, Texas.


    1. Kulkarni, J., Curran, B., Ebdon-Parry, M., Harrison, D., Total contact silicone partial foot prosthesis for partial foot amputations, The Foot , Vol. 5, 1995, pp. 32-35.


    1. Kelly, V. Mueller, M. Sinacore, D., Timing of Peak Plantar Pressure During the Stance Phase of Walking: A Study of Patients with Diabetes Mellitus and Transmetatarsal Amputation.,Journal of the American Podiatric Medical Association , Vol. 90, No. 1, January 2000.


    1. Armstrong, D., Lavery, L., Plantar Pressures are Higher in Diabetic Patients Following Partial Foot Amputation., Ostomy/Wound Management , Vol. 44, No. 3, March 1998.


    1. Mueller, M., Strube, M., Allen, B., Therapeutic Footwear can Reduce Plantar Pressures in Patients with diabetes and Transmetatarsal Amputation., Diabetic Care , Vol. 20, No 4, April 1997.


    1. Dillon, M., Barker, T., Preservation of Residual Foot Length in Partial Foot Amputation: A Biomechanical Analysis., Foot and Ankle International , Vol. 27, No. 2, February 2006.


    1. Dillon, M., Barker, T., Can partial foot prostheses effectively restore foot length?, Prosthetics and Orthotics International , Vol 30, No. 1, April 2006, pp.17-23.


    1. Dillon, M., Chapter 4: Biomechanical Models for the Analysis of Partial Foot Amputee Gait. PhD Thesis, Department of Mechanical, Manufacturing and Medical Engineering. 2001; Queensland University of Technology: Brisbane.


    1. Marks, A. A., Manual of Artificial Limbs: An Exhaustive Exposition of Prosthesis , Chapter 3: Partial Feet Amputations, pp. 27-44. New York, New York, 1918.


  1. Bowker, J., Chapter 34: Amputations and Disarticulations Within the Foot: Surgical Management, Atlas of Amputations and Limb Deficiencies , 3rd Edition, American Academy of Orthopedic Surgeons, Rosemont, Illinois 2004.