Osteodesis for Hallux Valgus Correction: Is it Effective?

Abstract

Background

Although the etiology of hallux valgus is contested, in some patients it may be failure of the stabilizing soft tissue structures around the first ray of the foot. Because there is lack of effective soft tissue techniques, osteotomies have become the mainstream surgical approach to compensate for the underlying soft tissue deficiency; osteodesis, a soft tissue nonosteotomy technique, may be a third alternative, but its efficacy is unknown.

Questions/purposes

We asked: (1) Can an osteodesis, a distal soft tissue technique, correct hallux valgus satisfactorily in terms of deformity correction and improvement in American Orthopaedic Foot and Ankle Society (AOFAS) score? (2) Is the effectiveness of an osteodesis affected by the patient’s age or deformity severity? (3) What complications are associated with this procedure?

Methods

Between February and October 2010, we performed 126 operations to correct hallux valgus, of which 126 (100%) were osteodeses. Sixty-one patients (110 procedures) (87% of the total number of hallux valgus procedures) were available for followup at a minimum of 12 months (mean, 23 months; range, 12–38 months). This group formed our study cohort. During the study period, the general indications for this approach included failed conservative measures for pain relief and metatarsophalangeal angle greater than 20° or intermetatarsal angle greater than 9°. Intermetatarsal cerclage sutures were used to realign the first metatarsal and postoperative fibrosis was induced surgically between the first and second metatarsals to maintain its alignment. The radiologic first intermetatarsal angle, metatarsophalangeal angle, and medial sesamoid position were measured by Hardy and Clapham’s methods for deformity and correction evaluation. Clinical results were assessed by the AOFAS score.

Results

The intermetatarsal angle was improved from a preoperative mean of 14° to 7° (p < 0.001; Cohen’s d = 1.8) at followup, the metatarsophalangeal angle from 31° to 18° (p < 0.001; Cohen’s d = 3.1), the medial sesamoid position from position 6 to 3 (p < 0.001; Cohen’s d = 2.4), and AOFAS hallux score from 68 to 96 points (p < 0.001). Neither patient age nor deformity severity affected the effectiveness of the osteodesis in correcting all three radiologic parameters; however, the deformities treated in this series generally were mild to moderate (mean intermetatarsal angle, 14°; range, 9°–22°). There were six stress fractures of the second metatarsal (5%), five temporary metatarsophalangeal joint medial subluxations all resolved in one month by the taping-reduction method without surgery, and six metatarsophalangeal joints with reduced dorsiflexion less than 60°.

Conclusions

The osteodesis is a soft tissue nonosteotomy technique, and provided adequate deformity correction and improvement in AOFAS scores for patients with mild to moderate hallux valgus deformities, although a small number of the patients had postoperative stress fractures of the second ray develop. Future prospective studies should compare this technique with osteotomy techniques in terms of effectiveness of the correction, restoration of hallux function, complications, and long-term recurrence.

Level of Evidence

Level IV, therapeutic study. See the Instructions for Authors for a complete description of levels of evidence.

Electronic supplementary material

The online version of this article (doi:10.1007/s11999-014-3938-6) contains supplementary material, which is available to authorized users.

Introduction

Hallux valgus has been called a hallux valgus deformity complex because it is a condition primarily made up of the three deformities of hallux valgus, metatarsus primus varus, and metatarsosesamoid dissociation [26]. They are the result of the proximal phalanx, first metatarsal, and hallux sesamoid moving out of their normal anatomic positions owing to failure of their stabilizing ligaments such as the medial metatarsophalangeal collateral and medial metatarsosesamoid ligaments [7, 27] that were attenuated from wearing closed-toe shoes [13] and genetic propensity [21]. Hallux valgus is a progressive condition likely attributable to the majority of stresses being borne by the first ray in walking [10]. Generally, correction of metatarsus primus varus has been regarded as the primary surgical objective because its correction facilitates effective hallux valgus and metatarsosesamoid dissociation correction [5, 12, 22, 28].

There are numerous options available to correct metatarsus primus varus, and which to use is controversial [3]. Although the pathogenesis of metatarsus primus deformity lies in the soft tissues and osteotomy has its related complications [23], the osteotomy has been the more preferred approach to metatarsus primus varus correction [19, 20]. This is partly owing to variable results reported with the myriad of available soft tissue procedures [15]. However, physiologically, we believe it makes sense to correct hallux valgus by addressing the soft tissues [24] where the pathologic features of this disorder reside.

An osteodesis is a soft tissue nonosteotomy technique that was first reported in 1961 [2]. It has been described [11, 17, 30], with high levels of patient satisfaction and low complication rates. The osteodesis uses intermetatarsal cerclage sutures between the first and second metatarsals to realign the first metatarsal correcting the metatarsus primus varus. It also induces fibrous bonding between the first and second metatarsals to help stabilize the first metatarsal in preventing recurrence of metatarsus primus varus. However, compared with other approaches, little has been published regarding this approach, which we believe is a potentially appealing option. Osteodesis was the term used to describe the surgery by its original authors [2], and it may be misunderstood as a bone bonding procedure instead of a soft tissue procedure of fibrous connection between the first and second metatarsals.

We therefore sought to determine (1) if an osteodesis, a distal soft tissue technique, can correct hallux valgus satisfactorily in terms of deformity correction and improvement in American Orthopaedic Foot and Ankle Society (AOFAS) score, (2) if the effectiveness of osteodesis is affected by the patient’s age or deformity severity, and (3) complications associated with this procedure.

Patients and Methods

Study Design and Setting

This was a retrospective study of patients who had an osteodesis between February and October 2010. There were no other surgical methods used during the same period. All cases were performed by one surgeon (DYW). For followup assessment, patients were interviewed and examined by one of the authors (DYW). Radiographs and photographs were taken and measured by a nurse-technician (RT) to compare with the patients’ preoperative data.

Participants/Study Subjects

The primary reason given by patients for deciding on surgery was foot pain in the bunion, metatarsophalangeal joint, or midmetatarsal head area that was not alleviated satisfactorily after 6 months of conservative treatment consisting of wearing wide and soft shoes. The basic surgical indication was radiologic deformity of the first metatarsophalangeal angle greater than 20° or first intermetatarsal angle greater than 9° in a patient with pain over the site that had not responded to conservative treatment. None of the patients had rheumatoid arthritis, hallux rigidus, or previous surgery of their feet.

Description of Experiment, Treatment, or Surgery

The first intermetatarsal space was exposed through a 1-inch dorsal incision. After subcutaneous dissection and distal peroneal nerve protection, the interosseous muscle was retracted laterally and partially resected if necessary to expose the lateral capsule of the first metatarsophalangeal joint and the conjoint tendon of the lateral flexor hallucis brevis and adductor hallucis. A distal lateral soft tissue release was performed with an inverted T capsular incision to transect the lateral collateral and lateral metatarsosesamoid ligaments. The adductor hallucis tendon was not released. No fibular sesamoid was excised. The distal 1/3 of the first and second metatarsals then was exposed subperiosteally. Three 2-mm drill holes were made in the first metatarsal neck region approximately 5 mm apart in the dorsoplantar direction. Then double-stranded #1 PDS^® sutures (Ethicon Inc, Somerville, NJ, USA) were passed through the drill holes and around the neck of the second metatarsal displacing the interosseous muscle inferiorly. Opposing cortices then were fish-scaled with an osteotome (Fig. 1). The sutures were tied with four knots while the forefoot, being squeezed with one hand after intraoperative radiographic examination, showed satisfactory metatarsus primus varus correction by temporary suture fixation with a surgical clamp. Seven feet required additional surgery: extensor tendon lengthening, and pancapsular and collateral ligament releases for dorsal dislocation of the second metatarsophalangeal joint.

Fig. 1A–B .

(A) The opposing cortices of the distal 1/3 of the first (solid white arrow) and second metatarsals (solid black arrow) are fish-scaled for intermetatarsal fibrous bonding ingrowth, and intermetatarsal cerclage sutures (black-bordered white arrow) are seen deep in the wound after being passed through first metatarsal drill holes and around the distal shaft of the second metatarsal. (B) An elliptical excisional medial incision (black arrow) was made only after hallux valgus deformities were corrected by the intermetatarsal cerclage technique and the correction effects can be seen by comparing the right and left feet.

A medial horizontal elliptical excisional incision was made over the bunion to remove redundant skin, bursa, capsulotendinous tissues, and bony overgrowth (Fig. 2). The ligament, tendon, and capsule tissues then were approximated as one combined layer without plication. After skin closure, a compression dressing was applied before the thigh tourniquet was deflated.

Fig. 2.

The 10-day postoperative radiograph of the feet of a 70-year-old woman shows bilateral medial subluxation of the first metatarsophalangeal joints.

Aftercare

Full weightbearing was permitted according to the patient’s comfort, but no restrictions were imposed; all patients were fitted for a full-length 2.5-mm thick thermoplastic total-contact removable custom-made foot cast-brace. Walking was reduced and crutches were advised for long-distance walking for 3 months. Patients were instructed how to do metatarsophalangeal joint extension ROM exercises and physiotherapy usually was not prescribed. Patients could resume unprotected walking during the fourth month, but unrestricted activities and shoes were permitted only after 6 months.

Variables, Outcome Measures, Data Sources, and Bias

The AOFAS 100-point hallux score system was used for clinical evaluation and patients also were asked if they were satisfied with their results and the reasons.

All radiographic examinations were performed with the patient standing on the same pedal-specific digital x-ray platform made by 20/20 Imaging^® (Lake in the Hills, IL, USA). Hardy and Clapham’s middiaphyseal axial line method [8] was used for the first metatarsophalangeal angle and first intermetatarsal angle measurement preoperatively and postoperatively. The midaxial line was determined using a software program (Opal-RAD; Viztek, Garner, NC, USA) that joined the midpoints of the proximal and distal diaphyseal regions of the metatarsals. The medial sesamoid position was assessed by Hardy and Clapham’s seven-position method [8].

All feet were classified in one of the three deformity severity subgroups with mild metatarsus primus varus defined as a first intermetatarsal angle of 12° or less, moderate as greater than 12° and less than 16°, and severe as 16° or greater. Patients were divided in two age subgroups (younger than 60 years and 60 years or older) for further analysis.

Statistical Analysis, Study Size

The analyses were performed using JMP^®, Version 10.0.0 (SAS Institute Inc, Cary, NC, USA). The main outcome variables of interest were the levels of improvements, which were defined to be the differences between the preoperative and postoperative measurements at the final followup. A paired t-test was used to determine the overall effectiveness of the osteodesis. The levels of improvement were compared between the two age groups using an independent t-test and among the three deformity severity groups using one-way ANOVA. Pairwise comparisons of the mean improvements also were performed using the Tukey-Kramer honestly significant difference test. A p value less than 0.05 was considered statistically significant. Pearson correlations were computed to determine the associations of some outcome variables.

Demographics, Description of Study Population

Between February and October 2010, we had 70 consecutive patients with hallux valgus and a total of 126 operations to correct hallux valgus, of which 100% were performed using osteodesis. Of the 70 patients, 61 (110 procedures; 87% of the total number of hallux valgus procedures) were available for followup at a minimum of 12 months (mean, 23 months; range, 12–38 months). This group formed our study cohort. During the study period, the general indications for this approach included failed conservative measures for pain relief and metatarsophalangeal angle greater than 20° or intermetatarsal angle greater than 9°. Nine patients were excluded from this study for failing to return for the minimum 12 months followup radiographic and physical examinations. Four lived overseas, and five were too busy to return but had no complaints with their feet. Forty-nine patients had bilateral procedures and 12 had a unilateral procedure for a total of 110 procedures. The minimum followup was 12 months (mean, 23 months; range, 12–38 months). There were 56 women and five men with a mean age of 48 years (range, 20–71 years). Forty-eight patients (87 feet) were in the younger-age subgroup, with a mean age of 43 years (range, 20–59 years). Thirteen patients (23 feet) were in the older-age subgroup with a mean age of 65 years (range, 60–71 years).

Results

Alignment and Clinical Scores

All three radiologic parameters (first intermetatarsal angle, metatarsophalangeal angle, and medial sesamoid position) improved after surgery (p < 0.001; Table 1). The intermetatarsal angle was improved from a preoperative mean of 14° to 7° (p < 0.001; Cohen’s d = 1.8) at followup, the metatarsophalangeal angle from 31° to 18° (p < 0.001; Cohen’s = 3.1), and the medial sesamoid position from position 6 to 3 (p < 0.001; Cohen’s d = 2.4) (Appendices 1 and 2. Supplemental material is available with the online version of CORR^®).

Table 1.

Radiologic measurements summary

Classification	Number of feet	Intermetatarsal angle			Metatarsophalangeal angle			Medial sesamoid position
		Preoperative	Postoperative	Correction	Preoperative	Postoperative	Correction	Preoperative	Postoperative
Cohort	110	14° (9°–22°)	7° (3°–12°)	7° (3°– 22°)	31° (15°–61°)	18° (5°–36°)	12° (−7° to 32°)	6 (2–7)	3 (1–6)
Mild intermetatarsal angle (≤ 12°)	29	11° (9°–12°)	6° (3°–8°)	5° (3°–8°)	24° (15°–38°)	16° (9°–27°)	8° (−7° to 23°)	5 (2–7)	3 (2–4)
Moderate intermetatarsal angle (> 12°, < 16°)	60	14° (12°–16°)	7° (3°–10°)	7° (4°–11°)	31° (17°–61°)	19° (6°–33°)	12° (−2° to 32°)	6 (4–7)	3 (1–6)
Severe intermetatarsal angle (≥ 16°)	21	18° (16°–22°)	7° (3°–12°)	10° (7°–14°)	40° (27°–58°)	22° (5°–36°)	18° (7°–30°)	7 (6–7)	4 (1–6)
Younger age (< 60 years)	87	14° (9°–22°)	7° (3°–12°)	7° (3°–14°)	30° (15°–61°)	18° (5°–36°)	12° (−2° to 32°)	6 (2–7)	3 (1–6)
Older age (≥ 60 years)	23	14° (11°–20°)	7° (3°–9°)	8° (3°–12°)	34° (17°–52°)	20° (9°–33°)	14° (−7° to 26°)	6 (5–7)	4 (2–6)

The AOFAS score improved from the preoperative mean of 68 points (range, 49–83 points) to a mean of 96 points at the final followup (range, 72–100 points) (Table 2). Eleven feet had preoperative metatarsophalangeal joint dorsiflexion movement less than 60° and 15 feet postoperatively. Five feet with second metatarsophalangeal joint dorsal dislocations were corrected; however, one patient had incomplete correction of both feet but both were asymptomatic. No patients reported metatarsocuneiform joint or midfoot region problems.

Table 2.

Summary of AOFAS scores

Variable	Pain (40 points)	Function (45 points)	Alignment (15 points)	Total (100 points)
Preoperative	26	34	8	68
Followup	39	44	14	96
p value	0.001*	0.001*	0.001*	0.001*

AOFAS = American Orthopaedic Foot and Ankle Society; *significant difference (p < 0.05).

Influence of Patient Age and Deformity

With the numbers available, neither age nor severity of deformity appeared to influence the efficacy of the osteodesis. For patients in the younger and older age groups, the analyses revealed that the average changes before and after osteodesis in the intermetatarsal angle (7° versus 8°; p = 0.1730) and metatarsophalangeal angle (12° versus 14°; p = 0.3312) (Table 1) were not different. When comparing the levels of improvement among the three deformity severity groups, more severe preoperative deformities required more correction as expected. This was achieved initially and maintained at a minimum of 1 year. For the changes in intermetatarsal angle, the average correction increased from 5° in the mild group to 7° in the moderate group and to 10° in the severe group (p < 0.001). Pairwise comparisons revealed that all three means were different, indicating that the amount of correction was associated with the severity of the preoperative deformity. A similar pattern was observed for the changes in metatarsophalangeal angle: the average changes of the three levels were 8°, 12°, and 18°, respectively (p < 0.001; Table 1). Pairwise comparisons revealed that the three averages were different. Comparing the final outcomes of the patients receiving osteodesis based on the postoperative measurements indicated that the postoperative intermetatarsal and metatarsophalangeal angles were not different between the younger and older age groups (intermetatarsal angle, 7° versus 7°; p = 0.6393; metatarsophalangeal angle, 18° versus 20°; p = 0.1003). The mean postoperative intermetatarsal angles of the three severity groups (6° versus 7° versus 7°; p < 0.001) were different. The mild group was found to have the smallest mean postoperative intermetatarsal angle and metatarsophalangeal angle measurements.

The results indicated that osteodesis could provide an appropriate amount of correction for the patients and that the amount of correction was positively associated with the severity. We also observed a strong positive correlation between the preoperative intermetatarsal angle and the correction in intermetatarsal angle (Pearson correlation coefficient = 0.7578; p < 0.001) and metatarsophalangeal angle (Pearson correlation coefficient = 0.7304; p < 0.001).

Complications

Five patients sustained a unilateral stress fracture of the second metatarsal and one patient had bilateral stress fractures. Four fractures occurred during the fourth postoperative month when unprotected walking was resumed. One patient had an asymptomatic nonunion owing to lack of immobilization. Two patients (two feet) had early recurrence of partial metatarsus primus varus with an intermetatarsal angle of 10° or greater within 6 months postoperatively. The preoperative intermetatarsal angles were 15° and 22° and final angles were 10° and 12°, respectively. These patients were pleased with their reduced pain, improved balance, and shoe fit.

Five patients had medial subluxations of the metatarsophalangeal joint that appeared during the first postoperative month and all had overcorrection of the metatarsus primus varus (Fig. 2). All subluxations were resolved satisfactorily by wrapping with Micropore^TM Medical Tape (3M^TM, St Paul, MN, USA) (Fig. 3) for 1 month.

Fig. 3.

Surgical tape is wrapped around the base of proximal phalanx and over the fifth metatarsal head to help reduce medial subluxation of the metatarsophalangeal joint.

Five patients (six feet) had deteriorated metatarsophalangeal joint dorsiflexion movement from 60° or greater preoperatively to less than 60° postoperatively and two were the result of degenerative arthritis. There were no infections, hallux varus, or transfer metatarsalgia complications.

Discussion

The etiology of the hallux valgus complex is multifactorial. Failure of the stabilizing soft tissue structures around the first ray of the foot may be a common cause of hallux valgus [7, 27]. Physiologically, it seems that the ideal approach to realign these displaced normal bones would be by soft tissue means, not osteotomies. However, there is no effective soft tissue procedure and osteotomy approach that is the preferred choice of surgeons [19, 20]. Our retrospective study of 110 feet after soft tissue osteodesis confirmed the findings of previous studies [11, 17, 30] that the first metatarsal of a metatarsus primus varus deformity of hallux valgus feet could be realigned satisfactorily by the intermetatarsal cerclage suture technique without osteotomies and also maintained by the intermetatarsal fibrosis as described Botteri and Castellana [2], although their procedure name of osteodesis may be misleading as a bone bonding procedure instead of a soft tissue procedure. Consistent correction of metatarsus primus varus and hallux valgus without an osteotomy suggested to us that the hallux, first metatarsal, and sesamoids were all freely reducible after their deforming ligamentous contractures were released. The absence of any reports of discomfort in the midfoot region also showed that the metatarsocuneiform and one to two cuneiform joints were flexible and did not obstruct first metatarsal realignment (Fig. 4).

Fig. 4A–B.

(A) A preoperative standing radiograph revealed severe metatarsus primus varus, incongruent metatarsophalangeal (solid white arrow), metatarsocuneiform (black-bordered white arrows) joints, and metatarsosesamoid dissociation (solid black arrow) of both feet, and dorsal dislocation of the patient’s right second metatarsophalangeal joint. The patient, a 64-year-old woman, had a preoperative AOFAS hallux score of 55 points. (B) Her 2-year standing radiograph revealed decreased first intermetatarsal space, congruent metatarsophalangeal and metatarsocuneiform joints, restoration of normal metatarsosesamoid association, and reduced right second metatarsophalangeal joint dislocation. Her AOFAS score was 100 points for the left foot and 93 points for the right foot. Her metatarsophalangeal angles were 15.9° and 25.0° and intermetatarsal angles were 4.3° and 4.7°. The angle measurement were made by using a software program that provided computer-assisted calculations of the white midaxial lines and manually placed cortical reference points (white dots).

This study had numerous limitations. First, it would be difficult to do a comparative study because unlike the osteodesis, most osteotomies are recommended for specific types of deformities, whereas we used osteodesis unselectively for all types of deformities in this series. The surgeon (DYW) preferred this surgical technique for all his patients, although this exclusive preference also was reported by Irwin and Cape [11]. Efforts were made to minimize possible bias in patient assessment by controlling radiographic examinations and parameter measurements of the entire cohort. Deformity evaluation with the method of Hardy and Clapham [8] was used consistently throughout the study owing to the nonosteotomy nature of the procedure. Subgroup sample sizes also were adequate for statistical analysis. The osteodesis is a little-practiced technique, and there have been few reports of the same or similar technique to verify its results despite largely satisfactory results [11, 17, 30]. We are unaware of any reported evidence of inadequacy to account for its lack of popularity. There is no means at this time to verify the nature of proposed fibrous bonding; however, its validity will need to be shown in future studies. We consider contraindications of the osteodesis to be skeletally immature metatarsals, first metatarsal basal spurs, or a facet that may block first metatarsal realignment [3], symptomatic or radiologic evidence of metatarsophalangeal joint arthritis, and severe osteoporosis. Importantly, the surgeon should be prepared to convert the cerclage technique to an osteotomy if intraoperative radiographs show unsatisfactory metatarsus primus varus correction, although we did not have to do that in the current series. The main disadvantage of the osteodesis is that no technology is yet available to monitor the progress of intermetatarsal fibrous bonding formation or detect impending failure of the cerclage sutures from excessive walking during the early postoperative period. The nine excluded patients (13%) (16 feet; 13%) from the original total who had surgery could have caused bias in the final result of this study.

The McBride procedure [16], a popular soft tissue technique, has not been regarded as being effective because the intermetatarsal soft tissue sutures are not strong enough for large metatarsus primus varus correction [3, 15]. We used a much stronger bone-to-bone intermetatarsal cerclage technique during osteodesis which provided statistically significant correction for metatarsus primus varus in our patients (Fig. 5). The results of the first-ray deformity trio correction (hallux valgus, metatarsus primus varus, and metatarsosesamoid dislocation) by osteodesis were comparable to those seen with osteotomies [6, 14, 25]. The notion that the first metatarsal might not be realigned without an osteotomy might have stemmed from preoperative assessments when the displaced lateral sesamoid and contractures blocked proper first metatarsal realignment during manual examination. Simultaneous correction of the hallux valgus and medial sesamoid position in our patients also confirmed the importance of metatarsus primus varus correction as reported by others [5, 12, 22, 24, 28]. In a similar concept, bone-to-bone suture-cable mechanism, mini-tightrope, and suture-button techniques have been successful for first metatarsal realignment without osteotomies [9, 29]. For long-term maintenance of first metatarsal stability to prevent metatarsus primus varus recurrence, the osteodesis introduced the intermetatarsal fibrous bonding mechanism to replace the initial cerclage sutures. Longer-term studies are required to validate the concept, although a couple studies had followups longer than 5 years [11, 30].

Fig. 5A–B.

(A) A preoperative standing radiograph revealed a typical and uncomplicated first- ray deformity trio of hallux valgus, metatarsus primus varus, and metatarsosesamoid dissociation of moderate severity in a 54-year-old woman. Her preoperative AOFAS hallux scores for both feet were 70 points. (B) Her 3-year standing radiograph revealed satisfactory hallux, first metatarsal and sesamoid realignments. Both feet had AOFAS hallux scores of 100 points.

An armamentarium of surgical techniques has been recommended to correct metatarsus primus varus and hallux valgus in patients of different ages and with different severities [3]. However, our study showed that these variables did not appear to adversely influence either clinical results or radiographic correction achieved with the osteodesis in correcting the first-ray deformity trio of metatarsus primus varus, hallux valgus, and medial sesamoid position (Fig. 6). We believe the reasons were that all metatarsus primus varus deformities in our patients shared a common underlying disorder of ligamentous failure and therefore were correctible by the same technique of the same surgical principle. In addition, all metatarsocuneiform joints were flexible even in older patients and in patients with severe metatarsus primus varus. Therefore, we do not have recommendations for age and deformity severity restrictions for the osteodesis. We believe this is the first single surgical technique that has shown its effectiveness across a range of metatarsus primus varus and hallux valgus severities.

Fig. 6A–B .

(A) A 75-year-old woman had severe deformities of both feet, as seen on her preoperative standing radiograph. Her AOFAS hallux scores were 60 points for both feet. Her metatarsophalangeal angles were 45.4° and 50.8° and intermetatarsal angles were 17.1° and 20.1°. (B) Her 2-year standing radiograph revealed satisfactory metatarsus primus varus correction but residual hallux valgus owing to the possibly contracted lateral head of the flexor hallucis brevis tendon. She had a mildly increased distal metatarsal articular angle. Her AOFAS hallux scores were 83 for the right foot and 93 for the left foot. The angle measurement were made by using a software program that provided computer-assisted calculations of the white midaxial lines and manually placed cortical reference points (white dots).

Second metatarsal postoperative stress fracture in 6% of our patients was the most common complication and occurred mostly during the fourth postoperative month when patients returned to unprotected walking, and possibly the patients had postoperative disuse osteopenia. It is a complication unique to intermetatarsal binding techniques and was reported in 32% of patients who had a suture-button procedure [29]. Patients should be cautioned of this possibility and advised to return to normal walking activities gradually during the fourth postoperative month. We found the frequency of recurrence of metatarsus primus varus to be low, and the recurrences to be mild when compared with 10% to 33% reported after osteotomy [4, 18]. This could be attributed to the first metatarsal stability issue by intermetatarsal fibrous bonding during the osteodesis. Prospective studies may be necessary to identify the reasons for some of the recurrence of metatarsus primus varus and hallux valgus deformities. Postoperative joint stiffness mostly can be prevented by early and frequent exercises but some deformities may not be avoided completely owing to relative shortening of the hallucis tendons after nonosteotomy realignment of the first ray and for the same reason that compression force may be increased across stiff metatarsophalangeal joints to accelerate the degenerative process of some preoperatively undetected arthritis [1]. Medial subluxation of metatarsophalangeal joints can occur owing to metatarsus primus varus overcorrection but it is temporary and amenable by taping of the joint.

The advantages of the osteodesis are that it is less invasive than osteotomy-based techniques and therefore a bilateral procedure can be performed safely without additional restrictions on postoperative weightbearing, no special instruments and implants are required, and osteotomy-related complications can be avoided. Our study confirmed previous reports [11, 17] that the concepts of realigning the first metatarsals with intermetatarsal cerclage sutures and first metatarsal stabilization by intermetatarsal fibrous bonding are effective. The osteodesis can be considered an alternative to osseous procedures for hallux valgus correction for patients of all ages after bone maturation and without arthritis and osteoporosis. Although our study showed that adequate correction for severe metatarsus primus varus could be achieved, future studies should be conducted to determine if greater corrections can be achieved and to identify long-term deformity recurrence incidences.