- Email: [email protected]
The oblique proximal phalangeal osteotomy in the correction of hallux valgus
The Oblique Proximal Phalangeal Osteotomy in the Correction of Hallux Valgus Michael M. Cohen, DPM, FACFAS An oblique proximal phalangeal osteotomy is introduced as an alternative to the Akin procedure for the treatment of hallux valgus. It consists of a single oblique osteotomy directed from proximal-dorsal to plantar-distal. A retrospective review of 32 patients who underwent 36 oblique proximal osteotomies is presented. The mean follow-up was 11 months (range, 3 to 21 months). Pre- and postoperative radiographic measurements of the distal articular set angles showed an average correction angle of 12° (range, 7° to 22°; SD ⫽ 2.99). Mean range of first metatarsophalangeal joint motion was 82% of the preoperative value when combined with distal metatarsal osteotomy, and 87% of the preoperative value without a distal metatarsal osteotomy. Time to clinical and radiographic healing averaged 5 weeks (range, 4 to 8 weeks). The average amount of phalangeal shortening was 1 mm (range, 0 to 2 mm). There were no delayed or nonunions. Subjective assessment conducted by retrospective review of 19 patients at 6 months postoperatively showed that 17 were completely satisfied (90%), 1 was satisfied (5%), and 1 fairly satisfied (5%). Nineteen would have the surgery again (1 with reservations). Eighteen were very satisfied with appearance and 1 was improved. Fourteen returned to shoe gear at weeks 6 to 8, and 5 returned at weeks 8 to 12. Eighteen were satisfied with pain relief and 1 was improved. The study indicates that the oblique proximal phalangeal osteotomy is an effective, reliable, and technically simple procedure for correction of deformities of the proximal phalanx. The advantages include minimal shortening, ease of adjustability, and a construct conducive to rigid fixation. ( The Journal of Foot & Ankle Surgery 42(5): 282–289, 2003) Key words: hallux valgus, phalangeal osteotomy, Akin procedure
T wo components of hallux valgus may be related to the structural aberration of the proximal phalanx. The degree of this deformity at the proximal pole of the phalanx can be quantified as the distal articular set angle (1). In the healthy foot, the acceptable range is believed to be 0° to 7.5° (2–5). However, the value of this angle may depend on the rotation of the great toe and the direction of the x-ray beam (2,3,7). The second abnormality is the hallux interphalangeus angle, which depicts deviation of the distal portion of the proximal phalanx (2,3,6). The accepted normal range is typically between 0° and 10° (2,3,6). The specific clinical indications for the proximal phalangeal osteotomy may not be clearly defined. Generally, the procedure is performed to correct a hallux interphalangeus or in conjunction with other procedures for the correction of hallux valgus (3,6 –10). It may also be used when treating a crossover second toe deformity (8). The Akin osteotomy From the Podiatry Section, Department of Surgical Services, Veterans Affairs Medical Center, Miami, FL. Address correspondence to: Michael M. Cohen, DPM, FACFAS, Chief Podiatry Section, Department of Surgical Services, Veterans Affairs Medical Center, 1201 NW 16th St, Miami, FL 33125. E-mail: [email protected] Copyright © 2003 by the American College of Foot and Ankle Surgeons 1067-2516/03/4205-0006$30.00/0 doi:10.1053/S1067-2516(03)00309-0
282
THE JOURNAL OF FOOT & ANKLE SURGERY
and its modifications have historically been used to correct said deformities in the proximal phalanx as an isolated procedure or in combination with others to correct a hallux valgus deformity. The technique consists of a transverse or an obliquely oriented, medially based wedge resectional osteotomy that is generally fixed with suture, steel wires, or screws (2,3,5,6,8 –12). Disadvantages of the Akin osteotomy include instability, shortening, under- or overcorrection, intraarticular fracture (2,3,5,8), plantar angulation caused by the pull of the extensor hallucis longus, and nonunion. Furthermore, the medial capsuloperiosteal tissue is often taken down with a proximal Akin procedure, compromising a secure plication of the capsule. An obliquely oriented osteotomy of the proximal phalanx (OPPO) is introduced, which allows adduction of the hallux with a single bone cut. The position of the toe is easily adjustable without shortening or the need for additional bone cuts and is easily rigidly fixated. The purpose of this article was to review the outcomes of an initial cohort of patients that underwent this procedure during correction of hallux valgus. Materials and Methods Between May 2000 and November 2001, the author performed a total of 36 OPPOs on 32 patients with hallux
TABLE 1
Patient data
Age (yr)
Sex
Indications
Adjunctive Procedures
60 55 43 78 62 51 63 49 57 58 37 44 44 66 66 63 47 53 42 39 28 32 38 38 34 34 49 71
M M M M F M M M M M F F F F F F M F M F F M M M M M M M
HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV/HL HV HV HV HV HV HV/HL
Chevron osteotomy Chevron osteotomy Lapidus, reverdin CSO CSO Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Lapidus, reverdin Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy CSO Chevron osteotomy CSO Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy CSO Cheilectomy
Time to Healing (weeks) 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 8 4 4 4 4 4 4 4 4
to to to to to to to to to to to to to to to to to to to
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6
to to to to to to to to
6 6 6 6 6 6 6 6
Time to Last Postoperative Follow-up (mos) 21 21 20 20 20 18 18 17 17 15 15 15 13 11 5 7 11 10 8 5 4 5 8 5 7 6 9 3
Abbreviations: CSO, crescentic shelf osteotomy; F, female; HL, hallux limitus; HV, hallux vulgus; M, male.
valgus. The study was approved by the Human Studies Subcommittee at the Veterans Affairs Medical Center in Miami, Florida, for retrospective review. Indications for the OPPO included hallux valgus, ranging from a mild deformity to a severe deformity, and crossover toe. In addition, patients with hallux interphalangeus and elevated distal articular set angles were also included. Exclusionary criteria included patients with hallux valgus who underwent a previous phalangeal osteotomy, symptomatic osteoarthrosis of the first metatarsophalangeal joint, and patients whose deformity was fully correctable at the time of surgery with proximal procedures. The minimum follow-up was 3 months. A variety of proximal first ray procedures were combined with the OPPO for the correction of the hallux valgus deformity. These procedures included a distal chevron osteotomy of the first metatarsal, cheilectomy, crescentic shelf osteotomies, Lapidus fusions, and distal derotational first metatarsal osteotomies. The specific procedures and their combinations are listed in Table 1. In addition, those patients with crossover toes and hammertoe deformities also had 1 or more of the following procedures: arthroplasty, lesser metatarsal phalangeal joint release, lesser metatarsal osteotomy, extensor lengthening, tendon transfers, and Kirschner-wire stabilization for 4 to 6 weeks.
Subjective Assessment Subjective assessments of patients who were more than 6 months postoperative were made using an anonymous questionnaire (Fig. 1) administered orally in person or by telephone by someone other than the surgeon (19 respondents). Pain, appearance, motion, time to return to shoe gear, and overall satisfaction was evaluated.
Objective Evaluation Clinical assessment of alignment and appearance was evaluated by the surgeon. Specific parameters included the correction of crossover toe (noted by relief of impingement and abutment), time to clinical and radiographic healing (noted by lack of motion or pain at the osteotomy site and by radiographic evidence of trabeculation), radiographic alignment, and hardware loosening. Pre- and postoperative range of motion of the first metatarsophalangeal joint at 6 months or greater was assessed to evaluate the specific influence of the osteotomies on joint motion. This was performed with the ankle and subtalar joint neutral and midtarsal locked, and included 17 patients with and 5 patients without distal metatarsal osteotomies.
VOLUME 42, NUMBER 5, SEPTEMBER/OCTOBER 2003
283
FIGURE 1
284
Subjective patient questionnaire for the oblique proximal phalangeal osteotomy.
THE JOURNAL OF FOOT & ANKLE SURGERY
Radiographic Assessment Preoperative lengths of the phalanges were compared with those measured postoperatively when the osteotomy was healed on standard weightbearing anteroposterior radiographs. In cases in which significant preoperative valgus rotation of the great toe was noted, the x-ray beam was rotated medially and perpendicular to the dorsal surface in an effort to capture a dorsoplantar image as closely as possible. A line was drawn from the midpoints of the base of the proximal phalanx to the distal articular surface and recorded in millimeters. The degree of angular correction was determined by measuring the distal articular set angle (9) preoperatively and postoperatively at the time of healing.
Surgical Technique A dorsomedial incisional approach was used to perform the index procedure. The OPPO was performed after the first metatarsal procedures had been completed. The capsuloperiosteal tissue was reflected off the base and shaft of the proximal phalanx. The medial capsular structures were kept intact to facilitate soft-tissue plication and preservation of the blood supply to the phalanx. The osteotomy was made 25° to 30° to the longitudinal bisection of the proximal phalanx (Fig. 2). This begins dorsally and proximally at the base and extends to exit distally and plantarly at a gradual angle. Plantarflexion of the hallux will facilitate this position and maneuver. One will often note a small dorsal tubercle of bone located at the base of the phalanx, which represents the anatomic attachment of the dorsal capsule and the extensor hallucis brevis. When present, the osteotomy should begin directly proximal to this tubercle. It is necessary to make the osteotomy as long as possible so that it exits as far distal as possible. This will avoid varus/valgus rotation (along the z axis) when adducting the distal fragment. When confronted with a short phalanx, a steeper cut is necessary. In this case, one may slide the hallux medially while combining the 2 movements to avoid varus rotation of the toe. With the osteotomy completed, the distal fragment is adducted to the desired position and clamped or temporarily fixed from dorsal to plantar with a 0.062-inch Kirschner wire (Fig. 2). If desired, intraoperative fluoroscopy may be used to confirm position. Permanent fixation is obtained with 1 or 2 dorsal-to-plantar– oriented 2.7-mm screws by using a standard AO technique (Fig. 3). Smaller screws may also be used. The screws should be placed away from the edge of the thin boney shelf to prevent stress risers. Standard closure is then performed. Postoperative care included immediate weightbearing in a postoperative shoe when the OPPO was combined with a distal metatarsal procedure. A loose-fitting, stable running shoe was allowed between weeks 6 and 8; otherwise, the patient was placed in a non-weightbearing short-leg cast for
4 to 6 weeks and allowed progressive weightbearing in a short-leg walker at weeks 4 to 8 when combined with a proximal metatarsal procedure. A return to shoe gear was permitted after radiographic bone healing. Results There were 36 OPPOs performed on 32 patients. The average postoperative follow-up was 11 months (range, 12 weeks to 21 months). The average age was 49.25 years (range, 27 to 78 years; SD ⫽ 13.25). Nineteen (59%) of the patients were men and 13 (41%) were women (Table 1). Radiographic Findings The average correction of the distal articular set angle was 12° (range, 8° to 22°; SD ⫽ 2.99) (Table 2). Postoperative interphalangeus angles were all within 5° of perpendicular (Fig. 3); 2 of the abnormal distal articular set angles were fully corrected and 1 was overcorrected. There was no evidence of nonunion, delayed union, excessive bone callus, or loss of correction in 35 of 36 osteotomies. All but 1 patient healed their osteotomies in 4 to 6 weeks. One patient was noted to be healed at 8 weeks. There was no hardware loosening. The average amount of shortening was 1 mm (range, 0 to 2 mm). Objective Findings The 22 patients who had a postoperative follow-up of 6 months or greater were evaluated for pre- and postoperative sagittal motion of the first metatarsophalangeal joint. Those patients who had distal metatarsal osteotomies (17 patients) were compared with those who did not (5 patients). The mean total range of motion was comparable between the 2 groups (82° and 87°, respectively). Postoperative range of motion of OPPOs performed with and without distal osteotomies measured 80% and 86% of preoperative dorsiflexory motion and 86% and 91% of plantarflexory motion, respectively. At 6-months postoperatively, there was no pain with joint motion or evidence of postoperative joint arthrosis in the metatarsophalangeal joint in any patient. All patients were ambulating in a tennis shoe at an average of 9 weeks (range, 8 to 12 weeks). At a mean of 11 months, all toes maintained their correction without recurrence of crossover. Subjective Evaluation Nineteen of 22 respondents (86%) who had 6 months or more of follow-up were evaluated by a questionnaire (Table 3). Eighteen patients were completely satisfied (95%) and 1 was satisfied and improved with respect to pain and appearance. Sixteen (84%) had no noticeable
VOLUME 42, NUMBER 5, SEPTEMBER/OCTOBER 2003
285
FIGURE 2 (A) Intraoperative view of saw position on proximal phalanx. Note the angle of the saw blade and the initial starting point proximal to the dorsal tubercle. (B) Illustration of osteotomy. (C) Dorsal view after osteotomy and with temporary fixation. (D) Dorsal illustration showing rotation of the osteotomy with temporary Kirschner wire and fixation screw. (E) Lateral illustration after 2.7-mm screw placement.
stiffness and 3 (16%) noted stiffness without pain during activity. Eighteen (95%) were very satisfied with the cosmetic result. Fourteen (74%) returned to normal shoe 286
THE JOURNAL OF FOOT & ANKLE SURGERY
gear at weeks 6 to 8. Eighteen (95%) stated that they would have the surgery again; 1 had reservations because of complaints related to anesthesia.
Complications Two osteotomies appeared overcorrected radiographically (but appeared rectus clinically) and 2 resulted in mild varus rotation of the great toe secondary to an excessively steep bone cut. None of these cases were clinically significant and were unnoticed by the patients. There were 2 minor intraoperative stress risers through the boney shelf. Both of these patients were allowed immediate weightbearing, and there was no loss of stability or fixation. There were no postoperative arthritic changes in the first metatarsophalangeal joint and 1 case of clinically significant malunion. This involved a 39-year-old woman who presented 3 days postoperatively with a subcutaneous abscess requiring incision and drainage. While performing the surgical preparation, the foot was inadvertently suspended by the hallux, destabilizing the phalangeal osteotomy. The screw was retightened after manual reduction. The osteotomy healed at 8 weeks with a 10° dorsal angulation. In this series, 2 stress risers developed after a 2.7-mm screw was placed too close to the edge of the bone. However, both went on to heal uneventfully. Screws appearing slightly long were removed at 10 and 18 weeks postoperatively in an effort to preclude flexor tendonitis in a competitive long-distance runner with a bilateral procedure. Otherwise, there were no cases of flexor or extensor tendinitis and palpable or symptomatic hardware (Table 1). Discussion
FIGURE 3 (A) Preoperative anteroposterior radiograph of hallux valgus with deviated proximal articular set angle (or distal metatarsal articular angle) and mild interphalangeus. (B) Postoperative anteroposterior radiograph with chevron/OPPO double osteotomy. Note correction of interphalangeus.
When evaluating hallux valgus, many criteria come into play. Specifically, one considers the intermetatarsal angle, the hallux valgus angle, the proximal articular set angle, the mobility of the first ray, and, perhaps, the distal articular set and interphalangeus angles of the proximal phalanx. Optimal results are obtained by using procedures geared to specifically correct the level of the deformity. In an ideal environment, patients with moderate or severe deformities are able to comply with weightbearing restrictions required of proximal metatarsal procedures. In an effort to accommodate their needs, the surgeon may attempt to stretch the indications for a distal metatarsal osteotomy to allow rapid resumption of their daily activities. In such instances, the proximal phalangeal osteotomy may be useful to enhance the outcome of a distal metatarsal procedure. A large number of patients in this study fell into this particular category (intermetatarsal angles ranging from 13° to 16°) and were effectively treated with a modified Chevron osteotomy combined with the OPPO. The combination of the Chevron-Akin double osteotomy has proven to be an effective method of treatment for hallux valgus (8,9). Mitchell and Baxter (9) reported a satisfaction rate of 95% in their retrospective study. Complications included malunion and degenerative changes in the metatarsophalangeal joint after intraarticular extension of the
VOLUME 42, NUMBER 5, SEPTEMBER/OCTOBER 2003
287
TABLE 2
Objective results (N ⴝ 34)
TABLE 3 more Mean
Correctional phalangeal rotation angle (°) Preoperative range of motion (°) Dorsiflexion Plantarflexion Postoperative range of motion at ⱖ6 months (°)a Dorsiflexion (with capital osteotomy) Dorsiflexion (without capital osteotomy) Plantarflexion (with capital osteotomy) Plantarflexion (without capital osteotomy) Total without capital osteotomy Total with capital osteotomy Time to clinical and radiographic healing (weeks) Hardware loosening Relief of crossover toe Recurrence of crossover Malunionb Overcorrection Undercorrection Varus rotation Valgus rotation Dorsal angulation Planter angulation
Range No. of Responses
12
7–22
65 30
45–85 15–40
50
38–70
55
45–65
30
5–40
32 87 82
10–38 — —
6 0 27 0
6–8 — — —
2 (radiographic) 0 2 0 1 0
—
a N ⫽ 22; 17 patients with capital osteotomy, 5 patients without capital osteotomy. b N ⫽ 14.
Akin osteotomy. The present investigation showed comparable satisfaction rates with the Akin osteotomy. Although the evaluation methods were different, the patient cohort seemed to fare as well as those with the double osteotomy reported by Mitchell and Baxter (9) and by Frey et al (8). One distinct difference in the comparisons was the method used to determine the range of motion. Mitchell and Baxter’s (9) study lacked information regarding preoperative joint motion and Frey et al (8) compared the postoperative motion with the contralateral foot, rather than the preoperative foot used in this study. Comparing the postoperative motion with the contralateral foot theoretically assesses a comparison to the norm for that individual, ie, with the presumption that the unoperated foot is normal. It does not accurately assess the range of motion gained or lost as a result of the procedure within that specific joint. Alternatively, in this study, there were also some limitations. Oftentimes, the first metatarsophalangeal joint was subluxed preoperatively, which caused pathologic motion and erroneous or out-of-axis measurements. Therefore, one has to appreciate that these estimations are crude at best. It was interesting, however, that the postoperative range of motion was quite similar to Mitchell and Baxter’s (9) findings (55.9° dorsiflexion and 35° plantarflexion vs. 52° dorsiflexion and 30° plantarflexion). Range of motion mea288
Postoperative subjective findings at 6 months or
THE JOURNAL OF FOOT & ANKLE SURGERY
Pain relief Completely satisfied Improved No changes Worse Cosmesis Very satisfied Improved No changes Worse Motion No noticeable stiffness Some stiffness painless during activity Noticeable stiffness with pain during activity Time to return to normal shoegear 6–8 weeks 8–12 weeks 12–16 weeks 16 weeks Overall satisfaction Completely satisfied Satisfied Fairly satisfied Dissatisfied Would have surgery again Yes Yes with reservations No
18 1 0 0 18 1 0 0 16 3 0 10 9 0 0 17 1 1 0 18 1 0
N ⫽ 19.
surements obtained at a minimum of 6 months allowed for the majority of the healing and rehabilitative period to have taken place, which allowed a more appropriate representation of the outcome. Furthermore, the patients in this study had the osteotomy rigidly fixated, which theoretically lends itself to rapid rehabilitation while enhancing overall joint motion and resisting displacement. Another limitation of this study was an accurate assessment of shortening. Valid measurements required that preand postoperative anteroposterior projections be taken in a reproducible manner. However, this was particularly difficult because of the preoperative sagittal and rotational displacement of the hallux. Nonetheless, the amount of shortening was still observed to be minimal by using our radiographic technique. Although attempts were made to obtain a true dorsal image of the preoperative phalanx in valgus it should be noted that the accuracy of the radiographic technique used in this study has not been validated. In their study, Frey et al (8) reported an average postoperative shortening of 4 mm with the Akin procedure. Alternatively, an average shortening of 1 mm with the OPPO occurs because correction is obtained by rotation without wedge resection. In this study, patients with moderate to severe hallux valgus and hammer or crossover toe of the second digit underwent the OPPO to medialize the great toe, provide room for the second toe to purchase the ground and enhance the cosmetic appear-
ance. This approach was a particularly effective alternative to a Keller arthroplasty or MTPJ fusion in a 71-year-old man with a combined asymptomatic hallux valgus/limitus and chief complaint of a painful crossover second toe. The phalanx was adducted 20 degrees and shortened by sliding the distal fragment proximally. The maneuver translated the hallux dorsally and shortened the lever arm to preclude symptoms in the joint after rectus alignment. The malunion rate of the OPPO is significantly less than that reported by Frey et al (8) who noted a 24% angulation deformity rate with the Akin procedure. One must take care not to place a screw too proximally because it may invade the first metatarsophalangeal joint or cause a stress fracture of the thin boney shelf. When using 2 screws, the author has found it preferable to place the 2.0-mm screw proximally to avoid the latter. Intraoperative observation has shown that the construct is quite stable and resistant to dislocation during normal loading. This feature was particularly evident when 2 noncompliant patients began immediate weightbearing without postoperative shoes. Both patients had uneventful outcomes. There may be other potential uses for this osteotomy. Recent studies indicate that osteotomies that correct the proximal articular set angle (PASA) (otherwise known as the distal metatarsal articular angle [DMAA]) may predispose the sesamoid complex to arthritic degeneration because of the displacement of the plantar metatarsal cristae (13,14). In light of this observation, perhaps the OPPO may be considered in lieu or in conjunction with a PASA(DMAA-) correcting osteotomy in the excessively deviated angle. This may preclude the problem by reducing the amount of necessary correction of the PASA. For severe hallux interphalangeus without hallux valgus, the author has performed the OPPO in reverse (the same osteotomy made from dorsal-distal to proximal-plantar, beginning at the distal phalangeal joint through the medial approach) in 3 patients. Although the correction was acceptable, the procedure was technically demanding because of difficulty in obtaining access for fixation without compromising the long extensor tendon. In cases of hallux interphalangeus angles of up to 25°, correction is easily obtained with the standard osteotomy. Another potential use of the OPPO is a phalangeal osteotomy to correct hallux limitus. As an alternative to the dorsiflexory wedge osteotomy, notable for its difficulty in allowing stable fixation, the OPPO has the added versatility of shortening with dorsiflexion and allows immediate range of motion because of stable fixation. The maneuver is technically easy and consists of sliding the phalanx proximally to the desired position after the osteotomy is performed. Excess bone that overhangs the joint is then pared off. This feature suggests its use as a sensible adjunct in the treatment of hallux limitus by elevating the phalanx, shortening the lever arm, and indirectly lengthening the long flexor. Al-
though it cannot provide the amount of hallux dorsiflexion obtained with a wedge osteotomy, the OPPO instead relieves dorsal impingement by translating the phalanx dorsally. Increased dorsiflexion can be attained by fashioning the osteotomy in a steeper or more acute angle. Lastly, the OPPO may be considered when correcting residual transverse plane malunion deformities of the phalanx after failed hallux valgus repair. The advantage of being able to correct the deformity precisely without additional shortening is realized. Conclusion The OPPO has been presented as an alternative to the Akin phalangeal osteotomy. The advantages include a single oblique bone cut that is less technically demanding, precise correction is attained, stable fixation is easily performed, and little or no shortening is realized. The study indicates that satisfactory outcomes can be achieved with this osteotomy when correction of a phalangeal deformity is required. References 1. Balding MG, Sorto LA. Distal articular set angle. Etiology and xray evaluation. J Am Podiatr Med Assoc 75:648 – 652, 1985. 2. Chang Thomas J. Hallux osteotomy procedures. In Textbook of Bunion Surgery, 3rd ed, pp 112–128, edited by J Gerbert, Saunders, Philadelphia, 2001. 3. Gerbert J, Spector E, Clark J. Osteotomy procedures on the proximal phalanx for correction of a hallux deformity. J Am Podiatr Assoc 64:617– 629, 1974. 4. Sorto LA Jr, Balding MG, Weil LS. Hallux abductus interphalangeus, etiology, x-ray evaluation and treatment. J Am Pod Assoc 66:384 – 396, 1976. 5. Springer KR. The role of the Akin osteotomy in the surgical management of hallux abductovalgus. In Clinics in Podiatric Medicine and Surgery, vol 6, no 1, pp 115–131, edited by KT Jules, Saunders, Philadelphia, 1989. 6. Barnett CH. Valgus deviation of the distal phalanx of the great toe. J Anat 96:171–177, 1962. 7. Seelenfreund M, Fried A. Correction of hallux valgus deformity by basal phalanx osteotomy of the big toe. J Bone Joint Surg 55A:1411– 1415, 1973. 8. Frey C, Jahss M, Kummer F. The Akin procedure an analysis of results. Foot Ankle 12:1– 6, 1991. 9. Mitchell LA, Baxter DE. A chevron Akin double osteotomy for correction of hallux valgus. Foot Ankle 12:7–14, 1991. 10. Myerson MS. Hallux valgus. In Foot and Ankle Disorders, vol 1, chapter 9, pp 234 –237, edited by MS Myerson, Saunders, Philadelphia, 2000. 11. Akin OF. The treatment of hallux valgus: a new operative procedure and its results. Med Sentinel 33:678 – 679, 1925. 12. Colloff B, Weitz E. Proximal phalangeal osteotomy in hallux valgus. Clin Orthop 54:105–113, 1967. 13. Breslauer C, Cohen M. Effect of proximal articular set angle-correcting osteotomies on the hallucal sesamoid apparatus: a cadaveric and radiographic investigation. J Foot Ankle Surg 40:366 –373, 2001. 14. Cohen M, Roman A, Ayres M, Freedline A. The crescentic shelf osteotomy. J Foot Ankle Surg 32:209 –226, 1993.
VOLUME 42, NUMBER 5, SEPTEMBER/OCTOBER 2003
289
T wo components of hallux valgus may be related to the structural aberration of the proximal phalanx. The degree of this deformity at the proximal pole of the phalanx can be quantified as the distal articular set angle (1). In the healthy foot, the acceptable range is believed to be 0° to 7.5° (2–5). However, the value of this angle may depend on the rotation of the great toe and the direction of the x-ray beam (2,3,7). The second abnormality is the hallux interphalangeus angle, which depicts deviation of the distal portion of the proximal phalanx (2,3,6). The accepted normal range is typically between 0° and 10° (2,3,6). The specific clinical indications for the proximal phalangeal osteotomy may not be clearly defined. Generally, the procedure is performed to correct a hallux interphalangeus or in conjunction with other procedures for the correction of hallux valgus (3,6 –10). It may also be used when treating a crossover second toe deformity (8). The Akin osteotomy From the Podiatry Section, Department of Surgical Services, Veterans Affairs Medical Center, Miami, FL. Address correspondence to: Michael M. Cohen, DPM, FACFAS, Chief Podiatry Section, Department of Surgical Services, Veterans Affairs Medical Center, 1201 NW 16th St, Miami, FL 33125. E-mail: [email protected] Copyright © 2003 by the American College of Foot and Ankle Surgeons 1067-2516/03/4205-0006$30.00/0 doi:10.1053/S1067-2516(03)00309-0
282
THE JOURNAL OF FOOT & ANKLE SURGERY
and its modifications have historically been used to correct said deformities in the proximal phalanx as an isolated procedure or in combination with others to correct a hallux valgus deformity. The technique consists of a transverse or an obliquely oriented, medially based wedge resectional osteotomy that is generally fixed with suture, steel wires, or screws (2,3,5,6,8 –12). Disadvantages of the Akin osteotomy include instability, shortening, under- or overcorrection, intraarticular fracture (2,3,5,8), plantar angulation caused by the pull of the extensor hallucis longus, and nonunion. Furthermore, the medial capsuloperiosteal tissue is often taken down with a proximal Akin procedure, compromising a secure plication of the capsule. An obliquely oriented osteotomy of the proximal phalanx (OPPO) is introduced, which allows adduction of the hallux with a single bone cut. The position of the toe is easily adjustable without shortening or the need for additional bone cuts and is easily rigidly fixated. The purpose of this article was to review the outcomes of an initial cohort of patients that underwent this procedure during correction of hallux valgus. Materials and Methods Between May 2000 and November 2001, the author performed a total of 36 OPPOs on 32 patients with hallux
TABLE 1
Patient data
Age (yr)
Sex
Indications
Adjunctive Procedures
60 55 43 78 62 51 63 49 57 58 37 44 44 66 66 63 47 53 42 39 28 32 38 38 34 34 49 71
M M M M F M M M M M F F F F F F M F M F F M M M M M M M
HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV HV/HL HV HV HV HV HV HV/HL
Chevron osteotomy Chevron osteotomy Lapidus, reverdin CSO CSO Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Lapidus, reverdin Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy CSO Chevron osteotomy CSO Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy Chevron osteotomy CSO Cheilectomy
Time to Healing (weeks) 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 8 4 4 4 4 4 4 4 4
to to to to to to to to to to to to to to to to to to to
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6
to to to to to to to to
6 6 6 6 6 6 6 6
Time to Last Postoperative Follow-up (mos) 21 21 20 20 20 18 18 17 17 15 15 15 13 11 5 7 11 10 8 5 4 5 8 5 7 6 9 3
Abbreviations: CSO, crescentic shelf osteotomy; F, female; HL, hallux limitus; HV, hallux vulgus; M, male.
valgus. The study was approved by the Human Studies Subcommittee at the Veterans Affairs Medical Center in Miami, Florida, for retrospective review. Indications for the OPPO included hallux valgus, ranging from a mild deformity to a severe deformity, and crossover toe. In addition, patients with hallux interphalangeus and elevated distal articular set angles were also included. Exclusionary criteria included patients with hallux valgus who underwent a previous phalangeal osteotomy, symptomatic osteoarthrosis of the first metatarsophalangeal joint, and patients whose deformity was fully correctable at the time of surgery with proximal procedures. The minimum follow-up was 3 months. A variety of proximal first ray procedures were combined with the OPPO for the correction of the hallux valgus deformity. These procedures included a distal chevron osteotomy of the first metatarsal, cheilectomy, crescentic shelf osteotomies, Lapidus fusions, and distal derotational first metatarsal osteotomies. The specific procedures and their combinations are listed in Table 1. In addition, those patients with crossover toes and hammertoe deformities also had 1 or more of the following procedures: arthroplasty, lesser metatarsal phalangeal joint release, lesser metatarsal osteotomy, extensor lengthening, tendon transfers, and Kirschner-wire stabilization for 4 to 6 weeks.
Subjective Assessment Subjective assessments of patients who were more than 6 months postoperative were made using an anonymous questionnaire (Fig. 1) administered orally in person or by telephone by someone other than the surgeon (19 respondents). Pain, appearance, motion, time to return to shoe gear, and overall satisfaction was evaluated.
Objective Evaluation Clinical assessment of alignment and appearance was evaluated by the surgeon. Specific parameters included the correction of crossover toe (noted by relief of impingement and abutment), time to clinical and radiographic healing (noted by lack of motion or pain at the osteotomy site and by radiographic evidence of trabeculation), radiographic alignment, and hardware loosening. Pre- and postoperative range of motion of the first metatarsophalangeal joint at 6 months or greater was assessed to evaluate the specific influence of the osteotomies on joint motion. This was performed with the ankle and subtalar joint neutral and midtarsal locked, and included 17 patients with and 5 patients without distal metatarsal osteotomies.
VOLUME 42, NUMBER 5, SEPTEMBER/OCTOBER 2003
283
FIGURE 1
284
Subjective patient questionnaire for the oblique proximal phalangeal osteotomy.
THE JOURNAL OF FOOT & ANKLE SURGERY
Radiographic Assessment Preoperative lengths of the phalanges were compared with those measured postoperatively when the osteotomy was healed on standard weightbearing anteroposterior radiographs. In cases in which significant preoperative valgus rotation of the great toe was noted, the x-ray beam was rotated medially and perpendicular to the dorsal surface in an effort to capture a dorsoplantar image as closely as possible. A line was drawn from the midpoints of the base of the proximal phalanx to the distal articular surface and recorded in millimeters. The degree of angular correction was determined by measuring the distal articular set angle (9) preoperatively and postoperatively at the time of healing.
Surgical Technique A dorsomedial incisional approach was used to perform the index procedure. The OPPO was performed after the first metatarsal procedures had been completed. The capsuloperiosteal tissue was reflected off the base and shaft of the proximal phalanx. The medial capsular structures were kept intact to facilitate soft-tissue plication and preservation of the blood supply to the phalanx. The osteotomy was made 25° to 30° to the longitudinal bisection of the proximal phalanx (Fig. 2). This begins dorsally and proximally at the base and extends to exit distally and plantarly at a gradual angle. Plantarflexion of the hallux will facilitate this position and maneuver. One will often note a small dorsal tubercle of bone located at the base of the phalanx, which represents the anatomic attachment of the dorsal capsule and the extensor hallucis brevis. When present, the osteotomy should begin directly proximal to this tubercle. It is necessary to make the osteotomy as long as possible so that it exits as far distal as possible. This will avoid varus/valgus rotation (along the z axis) when adducting the distal fragment. When confronted with a short phalanx, a steeper cut is necessary. In this case, one may slide the hallux medially while combining the 2 movements to avoid varus rotation of the toe. With the osteotomy completed, the distal fragment is adducted to the desired position and clamped or temporarily fixed from dorsal to plantar with a 0.062-inch Kirschner wire (Fig. 2). If desired, intraoperative fluoroscopy may be used to confirm position. Permanent fixation is obtained with 1 or 2 dorsal-to-plantar– oriented 2.7-mm screws by using a standard AO technique (Fig. 3). Smaller screws may also be used. The screws should be placed away from the edge of the thin boney shelf to prevent stress risers. Standard closure is then performed. Postoperative care included immediate weightbearing in a postoperative shoe when the OPPO was combined with a distal metatarsal procedure. A loose-fitting, stable running shoe was allowed between weeks 6 and 8; otherwise, the patient was placed in a non-weightbearing short-leg cast for
4 to 6 weeks and allowed progressive weightbearing in a short-leg walker at weeks 4 to 8 when combined with a proximal metatarsal procedure. A return to shoe gear was permitted after radiographic bone healing. Results There were 36 OPPOs performed on 32 patients. The average postoperative follow-up was 11 months (range, 12 weeks to 21 months). The average age was 49.25 years (range, 27 to 78 years; SD ⫽ 13.25). Nineteen (59%) of the patients were men and 13 (41%) were women (Table 1). Radiographic Findings The average correction of the distal articular set angle was 12° (range, 8° to 22°; SD ⫽ 2.99) (Table 2). Postoperative interphalangeus angles were all within 5° of perpendicular (Fig. 3); 2 of the abnormal distal articular set angles were fully corrected and 1 was overcorrected. There was no evidence of nonunion, delayed union, excessive bone callus, or loss of correction in 35 of 36 osteotomies. All but 1 patient healed their osteotomies in 4 to 6 weeks. One patient was noted to be healed at 8 weeks. There was no hardware loosening. The average amount of shortening was 1 mm (range, 0 to 2 mm). Objective Findings The 22 patients who had a postoperative follow-up of 6 months or greater were evaluated for pre- and postoperative sagittal motion of the first metatarsophalangeal joint. Those patients who had distal metatarsal osteotomies (17 patients) were compared with those who did not (5 patients). The mean total range of motion was comparable between the 2 groups (82° and 87°, respectively). Postoperative range of motion of OPPOs performed with and without distal osteotomies measured 80% and 86% of preoperative dorsiflexory motion and 86% and 91% of plantarflexory motion, respectively. At 6-months postoperatively, there was no pain with joint motion or evidence of postoperative joint arthrosis in the metatarsophalangeal joint in any patient. All patients were ambulating in a tennis shoe at an average of 9 weeks (range, 8 to 12 weeks). At a mean of 11 months, all toes maintained their correction without recurrence of crossover. Subjective Evaluation Nineteen of 22 respondents (86%) who had 6 months or more of follow-up were evaluated by a questionnaire (Table 3). Eighteen patients were completely satisfied (95%) and 1 was satisfied and improved with respect to pain and appearance. Sixteen (84%) had no noticeable
VOLUME 42, NUMBER 5, SEPTEMBER/OCTOBER 2003
285
FIGURE 2 (A) Intraoperative view of saw position on proximal phalanx. Note the angle of the saw blade and the initial starting point proximal to the dorsal tubercle. (B) Illustration of osteotomy. (C) Dorsal view after osteotomy and with temporary fixation. (D) Dorsal illustration showing rotation of the osteotomy with temporary Kirschner wire and fixation screw. (E) Lateral illustration after 2.7-mm screw placement.
stiffness and 3 (16%) noted stiffness without pain during activity. Eighteen (95%) were very satisfied with the cosmetic result. Fourteen (74%) returned to normal shoe 286
THE JOURNAL OF FOOT & ANKLE SURGERY
gear at weeks 6 to 8. Eighteen (95%) stated that they would have the surgery again; 1 had reservations because of complaints related to anesthesia.
Complications Two osteotomies appeared overcorrected radiographically (but appeared rectus clinically) and 2 resulted in mild varus rotation of the great toe secondary to an excessively steep bone cut. None of these cases were clinically significant and were unnoticed by the patients. There were 2 minor intraoperative stress risers through the boney shelf. Both of these patients were allowed immediate weightbearing, and there was no loss of stability or fixation. There were no postoperative arthritic changes in the first metatarsophalangeal joint and 1 case of clinically significant malunion. This involved a 39-year-old woman who presented 3 days postoperatively with a subcutaneous abscess requiring incision and drainage. While performing the surgical preparation, the foot was inadvertently suspended by the hallux, destabilizing the phalangeal osteotomy. The screw was retightened after manual reduction. The osteotomy healed at 8 weeks with a 10° dorsal angulation. In this series, 2 stress risers developed after a 2.7-mm screw was placed too close to the edge of the bone. However, both went on to heal uneventfully. Screws appearing slightly long were removed at 10 and 18 weeks postoperatively in an effort to preclude flexor tendonitis in a competitive long-distance runner with a bilateral procedure. Otherwise, there were no cases of flexor or extensor tendinitis and palpable or symptomatic hardware (Table 1). Discussion
FIGURE 3 (A) Preoperative anteroposterior radiograph of hallux valgus with deviated proximal articular set angle (or distal metatarsal articular angle) and mild interphalangeus. (B) Postoperative anteroposterior radiograph with chevron/OPPO double osteotomy. Note correction of interphalangeus.
When evaluating hallux valgus, many criteria come into play. Specifically, one considers the intermetatarsal angle, the hallux valgus angle, the proximal articular set angle, the mobility of the first ray, and, perhaps, the distal articular set and interphalangeus angles of the proximal phalanx. Optimal results are obtained by using procedures geared to specifically correct the level of the deformity. In an ideal environment, patients with moderate or severe deformities are able to comply with weightbearing restrictions required of proximal metatarsal procedures. In an effort to accommodate their needs, the surgeon may attempt to stretch the indications for a distal metatarsal osteotomy to allow rapid resumption of their daily activities. In such instances, the proximal phalangeal osteotomy may be useful to enhance the outcome of a distal metatarsal procedure. A large number of patients in this study fell into this particular category (intermetatarsal angles ranging from 13° to 16°) and were effectively treated with a modified Chevron osteotomy combined with the OPPO. The combination of the Chevron-Akin double osteotomy has proven to be an effective method of treatment for hallux valgus (8,9). Mitchell and Baxter (9) reported a satisfaction rate of 95% in their retrospective study. Complications included malunion and degenerative changes in the metatarsophalangeal joint after intraarticular extension of the
VOLUME 42, NUMBER 5, SEPTEMBER/OCTOBER 2003
287
TABLE 2
Objective results (N ⴝ 34)
TABLE 3 more Mean
Correctional phalangeal rotation angle (°) Preoperative range of motion (°) Dorsiflexion Plantarflexion Postoperative range of motion at ⱖ6 months (°)a Dorsiflexion (with capital osteotomy) Dorsiflexion (without capital osteotomy) Plantarflexion (with capital osteotomy) Plantarflexion (without capital osteotomy) Total without capital osteotomy Total with capital osteotomy Time to clinical and radiographic healing (weeks) Hardware loosening Relief of crossover toe Recurrence of crossover Malunionb Overcorrection Undercorrection Varus rotation Valgus rotation Dorsal angulation Planter angulation
Range No. of Responses
12
7–22
65 30
45–85 15–40
50
38–70
55
45–65
30
5–40
32 87 82
10–38 — —
6 0 27 0
6–8 — — —
2 (radiographic) 0 2 0 1 0
—
a N ⫽ 22; 17 patients with capital osteotomy, 5 patients without capital osteotomy. b N ⫽ 14.
Akin osteotomy. The present investigation showed comparable satisfaction rates with the Akin osteotomy. Although the evaluation methods were different, the patient cohort seemed to fare as well as those with the double osteotomy reported by Mitchell and Baxter (9) and by Frey et al (8). One distinct difference in the comparisons was the method used to determine the range of motion. Mitchell and Baxter’s (9) study lacked information regarding preoperative joint motion and Frey et al (8) compared the postoperative motion with the contralateral foot, rather than the preoperative foot used in this study. Comparing the postoperative motion with the contralateral foot theoretically assesses a comparison to the norm for that individual, ie, with the presumption that the unoperated foot is normal. It does not accurately assess the range of motion gained or lost as a result of the procedure within that specific joint. Alternatively, in this study, there were also some limitations. Oftentimes, the first metatarsophalangeal joint was subluxed preoperatively, which caused pathologic motion and erroneous or out-of-axis measurements. Therefore, one has to appreciate that these estimations are crude at best. It was interesting, however, that the postoperative range of motion was quite similar to Mitchell and Baxter’s (9) findings (55.9° dorsiflexion and 35° plantarflexion vs. 52° dorsiflexion and 30° plantarflexion). Range of motion mea288
Postoperative subjective findings at 6 months or
THE JOURNAL OF FOOT & ANKLE SURGERY
Pain relief Completely satisfied Improved No changes Worse Cosmesis Very satisfied Improved No changes Worse Motion No noticeable stiffness Some stiffness painless during activity Noticeable stiffness with pain during activity Time to return to normal shoegear 6–8 weeks 8–12 weeks 12–16 weeks 16 weeks Overall satisfaction Completely satisfied Satisfied Fairly satisfied Dissatisfied Would have surgery again Yes Yes with reservations No
18 1 0 0 18 1 0 0 16 3 0 10 9 0 0 17 1 1 0 18 1 0
N ⫽ 19.
surements obtained at a minimum of 6 months allowed for the majority of the healing and rehabilitative period to have taken place, which allowed a more appropriate representation of the outcome. Furthermore, the patients in this study had the osteotomy rigidly fixated, which theoretically lends itself to rapid rehabilitation while enhancing overall joint motion and resisting displacement. Another limitation of this study was an accurate assessment of shortening. Valid measurements required that preand postoperative anteroposterior projections be taken in a reproducible manner. However, this was particularly difficult because of the preoperative sagittal and rotational displacement of the hallux. Nonetheless, the amount of shortening was still observed to be minimal by using our radiographic technique. Although attempts were made to obtain a true dorsal image of the preoperative phalanx in valgus it should be noted that the accuracy of the radiographic technique used in this study has not been validated. In their study, Frey et al (8) reported an average postoperative shortening of 4 mm with the Akin procedure. Alternatively, an average shortening of 1 mm with the OPPO occurs because correction is obtained by rotation without wedge resection. In this study, patients with moderate to severe hallux valgus and hammer or crossover toe of the second digit underwent the OPPO to medialize the great toe, provide room for the second toe to purchase the ground and enhance the cosmetic appear-
ance. This approach was a particularly effective alternative to a Keller arthroplasty or MTPJ fusion in a 71-year-old man with a combined asymptomatic hallux valgus/limitus and chief complaint of a painful crossover second toe. The phalanx was adducted 20 degrees and shortened by sliding the distal fragment proximally. The maneuver translated the hallux dorsally and shortened the lever arm to preclude symptoms in the joint after rectus alignment. The malunion rate of the OPPO is significantly less than that reported by Frey et al (8) who noted a 24% angulation deformity rate with the Akin procedure. One must take care not to place a screw too proximally because it may invade the first metatarsophalangeal joint or cause a stress fracture of the thin boney shelf. When using 2 screws, the author has found it preferable to place the 2.0-mm screw proximally to avoid the latter. Intraoperative observation has shown that the construct is quite stable and resistant to dislocation during normal loading. This feature was particularly evident when 2 noncompliant patients began immediate weightbearing without postoperative shoes. Both patients had uneventful outcomes. There may be other potential uses for this osteotomy. Recent studies indicate that osteotomies that correct the proximal articular set angle (PASA) (otherwise known as the distal metatarsal articular angle [DMAA]) may predispose the sesamoid complex to arthritic degeneration because of the displacement of the plantar metatarsal cristae (13,14). In light of this observation, perhaps the OPPO may be considered in lieu or in conjunction with a PASA(DMAA-) correcting osteotomy in the excessively deviated angle. This may preclude the problem by reducing the amount of necessary correction of the PASA. For severe hallux interphalangeus without hallux valgus, the author has performed the OPPO in reverse (the same osteotomy made from dorsal-distal to proximal-plantar, beginning at the distal phalangeal joint through the medial approach) in 3 patients. Although the correction was acceptable, the procedure was technically demanding because of difficulty in obtaining access for fixation without compromising the long extensor tendon. In cases of hallux interphalangeus angles of up to 25°, correction is easily obtained with the standard osteotomy. Another potential use of the OPPO is a phalangeal osteotomy to correct hallux limitus. As an alternative to the dorsiflexory wedge osteotomy, notable for its difficulty in allowing stable fixation, the OPPO has the added versatility of shortening with dorsiflexion and allows immediate range of motion because of stable fixation. The maneuver is technically easy and consists of sliding the phalanx proximally to the desired position after the osteotomy is performed. Excess bone that overhangs the joint is then pared off. This feature suggests its use as a sensible adjunct in the treatment of hallux limitus by elevating the phalanx, shortening the lever arm, and indirectly lengthening the long flexor. Al-
though it cannot provide the amount of hallux dorsiflexion obtained with a wedge osteotomy, the OPPO instead relieves dorsal impingement by translating the phalanx dorsally. Increased dorsiflexion can be attained by fashioning the osteotomy in a steeper or more acute angle. Lastly, the OPPO may be considered when correcting residual transverse plane malunion deformities of the phalanx after failed hallux valgus repair. The advantage of being able to correct the deformity precisely without additional shortening is realized. Conclusion The OPPO has been presented as an alternative to the Akin phalangeal osteotomy. The advantages include a single oblique bone cut that is less technically demanding, precise correction is attained, stable fixation is easily performed, and little or no shortening is realized. The study indicates that satisfactory outcomes can be achieved with this osteotomy when correction of a phalangeal deformity is required. References 1. Balding MG, Sorto LA. Distal articular set angle. Etiology and xray evaluation. J Am Podiatr Med Assoc 75:648 – 652, 1985. 2. Chang Thomas J. Hallux osteotomy procedures. In Textbook of Bunion Surgery, 3rd ed, pp 112–128, edited by J Gerbert, Saunders, Philadelphia, 2001. 3. Gerbert J, Spector E, Clark J. Osteotomy procedures on the proximal phalanx for correction of a hallux deformity. J Am Podiatr Assoc 64:617– 629, 1974. 4. Sorto LA Jr, Balding MG, Weil LS. Hallux abductus interphalangeus, etiology, x-ray evaluation and treatment. J Am Pod Assoc 66:384 – 396, 1976. 5. Springer KR. The role of the Akin osteotomy in the surgical management of hallux abductovalgus. In Clinics in Podiatric Medicine and Surgery, vol 6, no 1, pp 115–131, edited by KT Jules, Saunders, Philadelphia, 1989. 6. Barnett CH. Valgus deviation of the distal phalanx of the great toe. J Anat 96:171–177, 1962. 7. Seelenfreund M, Fried A. Correction of hallux valgus deformity by basal phalanx osteotomy of the big toe. J Bone Joint Surg 55A:1411– 1415, 1973. 8. Frey C, Jahss M, Kummer F. The Akin procedure an analysis of results. Foot Ankle 12:1– 6, 1991. 9. Mitchell LA, Baxter DE. A chevron Akin double osteotomy for correction of hallux valgus. Foot Ankle 12:7–14, 1991. 10. Myerson MS. Hallux valgus. In Foot and Ankle Disorders, vol 1, chapter 9, pp 234 –237, edited by MS Myerson, Saunders, Philadelphia, 2000. 11. Akin OF. The treatment of hallux valgus: a new operative procedure and its results. Med Sentinel 33:678 – 679, 1925. 12. Colloff B, Weitz E. Proximal phalangeal osteotomy in hallux valgus. Clin Orthop 54:105–113, 1967. 13. Breslauer C, Cohen M. Effect of proximal articular set angle-correcting osteotomies on the hallucal sesamoid apparatus: a cadaveric and radiographic investigation. J Foot Ankle Surg 40:366 –373, 2001. 14. Cohen M, Roman A, Ayres M, Freedline A. The crescentic shelf osteotomy. J Foot Ankle Surg 32:209 –226, 1993.
VOLUME 42, NUMBER 5, SEPTEMBER/OCTOBER 2003
289