The charts and radiographs of thirty-seven consecutive patients (forty-five
feet) who had been managed with correction of a hallux valgus deformity by a
single surgeon (W.-C.L.) between August 2003 and May 2004 were retrospectively
reviewed. Eight patients (ten feet) lacked adequate follow-up and were
excluded. The results of our operative procedures for the remaining
twenty-nine patients (thirty-five feet) were investigated. The indication for
surgery was chronic pain and deformity.
Radiographic criteria included a hallux valgus angle of >30°, a
first-second intermetatarsal angle of >12°, and lateral subluxation of
the sesamoids. Twenty-eight patients were female, and one was male. The
average age at the time of surgery was 51.0 years (range, thirty-three to
seventy-one years). The average duration of follow-up was 26.5 months (range,
twenty-four to thirty-two months).
An American Orthopaedic Foot and Ankle Society (AOFAS)
hallux-interphalangeal score was obtained both preoperatively and at the time
of the most recent follow-up visit. The patients were also questioned as to
whether they would undergo surgery again and, if so, with or without
reservations.
Radiographic Evaluation
All preoperative radiographs included weight-bearing dorsoplantar and
lateral views and a sesamoid view. The hallux valgus angle and the
first-second intermetatarsal angle and the extent of sesamoid subluxation were
assessed on the dorsoplantar radiograph. The talus-first metatarsal angle was
measured on the standing lateral radiograph. The hallux valgus angle and the
first-second intermetatarsal angle were measured with the method of Hardy and
Clapham15. Sesamoid
position was defined by the position of the medial sesamoid in relation to the
axis of the first metatarsal. A normal position was classified as grade 0, and
the most severe lateral subluxation, in which the medial sesamoid was
displaced laterally beyond the midaxis by >75% of its width, was classified
as grade 3.
Surgical Technique
The procedure is performed with the patient under spinal anesthesia. An
approximately 6-cm-long medial incision is made; the incision begins 1 cm
distal to the first metatarsophalangeal joint, curves slightly dorsally over
the bunion, continues proximally parallel to the plantar surface of the foot
(not along the first metatarsal), and ends 1 cm distal to the first
metatarsocuneiform joint (Fig.
2). If the incision is made along the shaft of the first
metatarsal, the scar ends up on the dorsum of the foot after correction, and
it can be seen when the patient stands.
A full-thickness dorsal flap including skin and subcutaneous tissue is then
elevated to the first web space (Fig.
3).
The dorsomedial cutaneous nerve over the first metatarsal head is exposed
and separated from the flap, and further dissection progresses laterally just
superficial to the extensor hallucis longus and brevis tendons
(Fig. 4). A vertical medial
capsular resection involving the removal of approximately 4 mm of capsule is
performed just proximal to the base of the proximal phalanx. A medial axial
longitudinal incision is made in the capsule, extending to the neck of the
first metatarsal. The medial eminence of the first metatarsal head is resected
with use of a micro-oscillating saw after dorsal and plantar capsular flaps
have been elevated from the medial eminence.
The deep fascia and areolar tissue between the first and second metatarsals
are then incised at about 2.5 cm proximal to the first metatarsophalangeal
joint, and the incision is extended distally. A lamina spreader is inserted
between the first and second metatarsal necks and is opened to widen the first
interspace. A curved hemostat is inserted under the deep transverse metatarsal
ligament, and the ligament is incised (Fig.
5). Both tendons of the adductor hallucis are released
(Fig. 6), and a longitudinal
incision in the capsule is made along the dorsal margin of the lateral
sesamoid (Fig. 7). The lateral
capsule of the first metatarsophalangeal joint, however, is not incised
vertically. In our series, we did not transfer the adductor tendon to the neck
of the first metatarsal, but we think that this procedure could be performed
through this approach.
At this point, the phalanx is abducted to test if the lateral soft-tissue
release is adequate. If the phalanx cannot be angulated 20° medially with
regard to the long axis of the first metatarsal
(Fig. 8), the longitudinal
incision along the dorsal margin of the lateral sesamoid is extended further
distally, and the capsular attachment laterally at the base of the proximal
phalanx is released.
A proximal chevron osteotomy is performed with use of a micro-oscillating
saw (Fig. 9). The apex of the
chevron is proximal and is located 7 mm distal to the first
metatarsalcuneiform joint. The angle of each arm of the chevron is made at
30° to the metatarsal axis.
After completion of the osteotomy, a small curette is placed on the lateral
side of the dorsum of the proximal fragment to lever the proximal fragment
medially as much as possible while the distal fragment is displaced and
angulated laterally.
Two 0.062-in (0.16-cm) Kirschner wires are inserted from proximal to distal
into the metatarsal head (Figs. 10-A and
10-B). An additional Kirschner wire is inserted if needed to
achieve secure fixation. The degree of correction should be checked with
fluoroscopy. If the first and second metatarsals are not parallel, the wires
are removed and the osteotomy site is remanipulated. Proximally, the medial
protruding bone is removed flush with the distal fragment. This bone and the
bone that has been resected from the medial eminence of the metatarsal head is
pushed into the osteotomy gap medially and applied to the lateral aspect of
the osteotomy site.
A medial capsulorrhaphy is performed with use of interrupted 2-0
nonabsorbable and absorbable sutures. The periosteum is approximated with 2-0
absorbable sutures, and closure is completed with a running 4-0 nonabsorbable
suture after about 3 mm of skin is excised from the plantar skin flap
(Fig. 11).
Postoperative Care
The foot is placed in a splint for about one week postoperatively.
Weight-bearing on the heel and the lateral border of the foot is permitted on
the day after surgery. The compressive dressings are changed on the third or
fourth postoperative day, and the sutures are removed two to three weeks
postoperatively. Weight-bearing on the first ray is not allowed until the
seventh postoperative week, and the Kirschner wires are removed during the
ninth postoperative week.
Statistical Analysis
The hallux valgus angle, first-second intermetatarsal angle, sesamoid
position, and AOFAS hallux-interphalangeal score at the time of the most
recent follow-up visit were compared with preoperative values, and paired
Student t tests were done. The level of significance was set at p <
0.05.
The hallux valgus angle improved by an average of 27.8°, from a
preoperative average of 35.8° ± 7.9° (range, 27.3° to
52.7°) to a postoperative average of 8.0° ± 7.6° (range,
—10.6° to 22.6°) (p < 0.001)
(Fig. 12). The first-second
intermetatarsal angle improved from a preoperative average of 16.8°
± 2.8° (range, 12.0° to 21.8°) to a postoperative average
of 5.0° ± 3.2° (range, —0.1° to 12.2°) (p <
0.001). The sesamoid position was reduced by an average of 2.5 grades, from a
preoperative average of 3.0 to a postoperative average of 0.5 ± 0.6
(range, 0 to 2). The mean lateral talus-first metatarsal angle was 6.5°
± 8.4° (range, —6.3° to 19.4°) preoperatively and
9.1° ± 7.9° (range, —3.7° to 33.7°)
postoperatively (p = 0.03). There was one hallux varus deformity of 5°,
but the patient had no discomfort.
The average AOFAS hallux-interphalangeal score improved from 54.8
preoperatively to 93.3 postoperatively (p < 0.01). With regard to pain,
patients reported that twenty-eight of thirty-five feet were completely
pain-free and that seven feet had only occasional or slight discomfort.
Postoperative motion (dorsiflexion plus plantar flexion) was regarded as
normal or mildly restricted (>75°) in thirty-one feet and moderately
restricted (30° to 74°) in four feet.
Twenty-three of the twenty-nine patients stated that they would undergo the
same procedure again without reservations. Five patients reported that they
would undergo the surgery with minor reservations because of the immediate
postoperative pain and the aggravation of Kirschner wire removal. One patient
stated that she would not undergo surgery again because she had a dorsiflexion
malunion that required revision surgery. There were three cases of marginal
wound necrosis, which healed secondarily without the need for any further
surgery.
One concern with this approach is its safety. The main blood supply over
the medial surface of the foot originates from the medial plantar artery, and
the dorsal area of the medial part of the foot is supplied primarily by the
dorsalis pedis
artery16,17.
The dorsal margin of the angiosome of the medial plantar artery is 2 to 3 cm
above the medial glabrous
junction16. Thus,
the main concern during exposure is the possibility of necrosis of the dorsal
flap. If the main blood supply to some part of the dorsal flap is from the
medial plantar artery, there would be a high risk of flap necrosis because the
skin incision crosses the borderline between the dorsal and plantar
angiosomes. We have used this approach to perform a proximal chevron osteotomy
in >200 feet since August 2003. Marginal wound necrosis developed in three
patients in that group, and in all three it healed uneventfully without
additional surgery. We believe that making the incision plantar to the
metatarsal shaft and parallel to the sole of the foot may protect the blood
supply to the dorsal flap. We think that this approach, with careful handling
of the dorsal flap, can be used safely, with only a slight possibility of
marginal wound necrosis.
Another concern with this approach is the possibility of an inadequate
lateral soft-tissue release, resulting in a less-than-satisfactory correction
of the hallux valgus deformity. The final hallux valgus angle, the
first-second intermetatarsal angle, and the sesamoid position were all
satisfactory in the present series, so we think that we can achieve a
sufficient lateral soft-tissue release through this approach.
In addition to the correction of alignment, we have obtained satisfactory
joint motion that is comparable with that described in any other report on
hallux valgus correction. We think that this approach creates less scar in the
first web space because the skin and subcutaneous tissue are not dissected and
the lateral capsule is not disrupted parallel to the joint line.
The apex of the proximal chevron osteotomy can be made
distally1,3,8,9
or
proximally6,18.
The potential risk associated with making the apex proximally is the creation
of a fracture of the proximal fragment, but, to our knowledge, this
complication has not been reported. The primary difference between making the
apex of the osteotomy proximally or distally is the location of the pivot
point for correction of the first metatarsal angulation. If the apex points
distally, the pivot point is more distal. Thus, if the apex is proximal, the
location of the pivot point is closer to the metatarsocuneiform joint, and we
believe that it is geometrically better to correct a deformity as close as
possible to the origin of the deformity.
With a proximal chevron osteotomy, the correction occurs through a
combination of lateral translation and angulation. Geometrically, more
translation is possible if the apex is made distally; however, in our
experience, we have been able to achieve a sufficient amount of translation
and therefore we prefer the proximal-apex osteotomy.
In conclusion, the present study suggests that we can release the lateral
soft tissues sufficiently by approaching the first web space dorsal to the
first metatarsal after elevating the skin and subcutaneous tissues. We believe
that satisfactory correction of moderate to severe hallux valgus deformities
can be obtained by combining this approach with a proximally apexed proximal
chevron osteotomy. ?