Foot & Ankle International - July 2018 - 862

862
our null hypothesis that there would be no difference
between groups. Our results suggest that both the ISF and
PW fixation methods may have sufficient strength to resist
early postoperative protected weightbearing and active
ankle dorsiflexion. If supported by clinical experience, the
likely benefits of early rehabilitation compared with plaster
immobilization could include improved patient satisfaction,
less stiffness, reduced risk of VTE, and shorter duration of
chemical VTE prophylaxis.
Although early rehabilitation is feasible, it is not risk free.
The normal excursion of the TP muscle is only 1.75 cm,28
but even 4-mm elongation of the tendon is estimated to
cause dysfunction.14 For the purposes of the current study,
considering measurement error, we defined displacement of
more than 5 mm as failure. All 11 of the PW specimens survived 1000 cycles, with mean displacement of 2.9 mm.
Displacement of more than 5 mm occurred in only 1 specimen, measuring 8.9 mm after 1000 cycles. Given the short
excursion of the TP muscle, this amount of displacement is
probably undesirable. In comparison, 1 of 12 ISF specimens
failed prematurely, within 100 cycles. In the remaining 11
specimens, mean displacement was 2.4 mm and in all specimens was less than 4 mm. Displacement was observed to
occur earlier in the PW group. Mean displacement of 2 mm
had occurred by 300 cycles in the PW group but took 900
cycles to reach this threshold in the ISF group. We observed
greater variation in displacement in the PW group compared
with ISF (SD 2.5 and 1.1 mm, respectively). Screw fixation
may therefore be preferable over tendon-to-tendon repair as
it is more consistently reproducible.
Data on cyclical testing of tendon fixation in the foot are
limited, especially in comparison to the extensive literature
in the knee. The displacements we observed for TP tendon
transfer appear similar to that observed in some studies of
hamstring graft interference screw fixation for reconstruction of the anterior cruciate ligament (ACL).22,27 For example, Kousa et al22 showed that cyclic loading of ACL
hamstring grafts fixed with a 7-mm RCI screw displaced
3.9 mm after 1500 cycles at 200 N. Micucci et al27 recorded
hamstring graft slippage in a 9-mm tunnel at 250 N for
1500 cycles, noting mean displacements of 2.65 mm with a
9-mm diameter interference screw and 7.83 mm with an
11-mm screw. However, another study of ACL graft fixation found a mean displacement of 0.7 to 1.3 mm among 5
hamstring tendon fixation methods after 1000 cycles from
70 to 200 N and mean ultimate strengths of 490 to 945 N.4
In theory, tendon transfer directly to bone is preferable.
The single failure we observed in the ISF group is similar
to rates observed in previous biomechanical studies.9,22 In
the foot, for example, Drakos et al9 reported early failure in
3 of 10 specimens following ISF (5.5-mm diameter biotenodesis screw) for flexor hallucis longus (FHL) transfer
to the calcaneus, at 60 N loads. Simulating ACL reconstruction, Kousa et al22 observed that 1 of 10 specimens
failed early.

Foot & Ankle International 39(7)
The physiological force required for active dorsiflexion
is unclear. Although in an unloaded foot model, 50 N tension was required to achieve maximum dorsiflexion,17 modern estimates based on muscle cross-sectional area suggest
that greater physiological forces are encountered.6,7,20,37
During normal gait, Delp and Zajac7 estimated that the tibialis anterior and tibialis posterior muscles generate peak
forces of 600 N and 1270 N, respectively. Even with an
anticipated loss of 1 grade of strength following anterior
transfer of the TP, it is likely that more than 50 N tension
would occur at the fixation point.6,7,17,20 We therefore chose
to cyclically load the tendon at 150 N to simulate likely
physiological forces encountered during protected partial
weightbearing with the ankle immobilized in a boot and
also nonweightbearing active ankle dorsiflexion.
Both techniques demonstrated similarly high load-tofailure tests. In all specimens, the PW consistently failed at
the tendon-suture interface. In the ISF group, all specimens
failed at the tendon-screw interface, similar to biomechanical studies assessing flexor digitorum longus (FDL) and
flexor hallucis longus (FHL) tendon transfers.9,24 In the current study, the screw was placed inferior to the tendon
within the pilot hole. Upon loading, due to the direction of
tendon pull, the superior edge of the pilot hole was loaded,
fracturing in some cases (Figures 5-7). In future studies, it
may be useful to assess whether proximal placement of the
screw relative to the tendon would increase pullout strength.
In vivo management of an early tendon transfer failure
would pose particular difficulties for both techniques. For
the PW technique, after load-to-failure testing, the peroneus
brevis tendon consistently sustained damage due to the
sutures tearing out, which would make it hard to resuture
the TP tendon using the original technique. A side-to-side
tendon repair would be more feasible as a revision procedure or alternatively using a different transfer site. Failure
following the ISF method might be even more difficult to
revise. We consistently noted tunnel deformity or small
fractures through part of the tunnel wall. Furthermore, the
tendon lying within the pilot hole flattened due to compression by the screw. Overreaming to a larger pilot hole and
use of a larger diameter interference screw might be possible if the navicular had sufficient bone stock. If the pilot
hole deformity was too severe to revise, a different transfer
site would be required.
Interference screw fixation in the foot has gained popularity for several conditions.12 Screw fixation might be preferable to a tendon-to-tendon repair as it is less dependent
upon surgeon technique. The load-to-failure forces we report
are greater than for other tendon transfers in the foot. In 1
study, split anterior tibialis tendon transfer to the cuboid
using an 8-mm biotenodesis screw had a mean load to failure of 150 N.29 Sabonghy et al33 reported that FDL transfer
to the navicular using a 7-mm × 20-mm interference screw
with a 6.4-mm diameter pilot hole failed at a mean of 148 N.
Louden et al24 performed a similar study using FDL and

Table of Contents for the Digital Edition of Foot & Ankle International - July 2018