The first metatarsal rotation in the coronal plane is critical in hallux valgus. Direct measurement of first metatarsal rotation and sesamoid subluxation is difficult. This study only focused on female patient, excluded gender difference from male patients. It revealed that hallux valgus deformity had more first metatarsal coronal rotation, and the α-angle correlation to IMA, DMAA, TSP.
The α-angle in control group was important to define abnormality. As adequate sample of control included, the average α-angle of 11.33 degrees was calculated. The average α-angle of the study group was 23.75 degrees, which was 12.42 degrees larger than the control group. All feet in study group had first metatarsal rotating into pronation, and there was no supination. This result was comparable to Kim and Mortier’s report.12,13 However, the mean of α-angle was different with their reports. Kim reported that the α-angle in his control group was 13.8 ± 4.1 degrees, but in our study, this was 21.9 ± 6.0 degrees.12 There are several reasons of this difference. First, simulating WBCT provided pressure on supine with a compressing machine which was not normal stress to foot. Second, the control group had larger sample size and were comparable with the study group with age, therefore guaranteed enough statistical power. Third, weightbearing X-ray and simulating WBCT were taken separately in Kim’s study. In our study, all radiographs acquired in single scanning.
Mortier et al. used weightbearing tangential radiographs to measure the first metatarsal pronation and reported a mean pronation value of 12.7 ± 7.7 degrees.13 The measurement of rotation on a plane radiograph may be influenced by a dorsiflexed first metatarsophalangeal joint. Collan et al. utilized WBCT to study the rotation in hallux valgus patients but showed no difference of coronal rotation between hallux valgus feet and normal feet.3 In their study, the WBCT only scanned one side at a time, not real physiological standing status. And their study had a small sample, which included 10 and 19 feet in the control and hallux valgus groups. So, our study was closer to the real physical condition.
Using the upper boundary of the 95% CI of the control group as boundary, there were 208 feet (96.3%) of the study group had excessive first metatarsal pronation. This percentage was largest compare to other reports. Kim reported 87.3% of hallux valgus cases had first metatarsal pronation more than the upper bounder of 95% CI of control.12 And Dayton reported 81% of his cases had first metatarsal pronation.5
Most of patient had sesamoid subluxation, 88.9% of hallux valgus patients had subluxation of sesamoids, and on average had a TSP grade 5.7 and Smith grade 1.79 in the coronal plane. TSP grading distribution showed that 32 feet (32/216, 14.81%) had tibial sesamoid position smaller than grade five. in the coronal plane, only 24 feet (24/216, 11.11%) had no subluxation of sesamoids, as Smith grade zero, in which 23 feet (10.6%) defined as pseudo-subluxation.
It was supposed that pseudo-subluxation hallux valgus feet will not require distal soft tissue release. But only 11.11% of cases were pseudo-subluxation. Smith and colleagues reported an anatomic triplane hallux valgus classification.8,18 Accordingly, in our study group, 11.11% of cases were Class 1 or Class 2A, and others were classified as Class 2B and Class 3. Average age was older, and IMA were relatively larger than. Only Most patients required distal soft tissue release. And age affects the sesamoids subluxation on coronal plane.10 Dayton and colleagues reported in their study, average age 32.42 years,4 that rotational correction without distal soft release reduce tibial sesamoid into its position.4,5
Rotational correction required on average was 22.1 degrees.5 They found that greater preoperative TSP scores were associated with greater intraoperative varus rotation required.5 In another case series with 21 feet, they reported that 81% (17/21) of cases had metatarsal pronation preoperatively.4
IMA, DMAA, and TSP had positive correlation to α-angle. But HVA and Smith grading had no correlation to α-angle. This was the difference between standing WBCT and simulating WBCT, or plane radiograph measurement. Kim reported that the rotation had no correlation among the parameters, such as IMA, HVA. Thus, first metatarsal pronation was independent to other hallux valgus changes.12 Motier also reported that first and second metatarsal angle had no correlation to radiographic pronation.13
IMA and TSP could be used as predictors of α-angle. But linear regression model showed that only TSP was a statistically predictable factor. Tibial sesamoid position(TSP) was used to evaluate the deformity of hallux valgus.6,15,20 Given that TSP was related to hallux valgus recurrence and failed correction,1,15 TSP also related to the rotation of first metatarsal.1,11,15 Our study indicated the association between TSP and α-angle in hallux valgus group. TSP is very useful to predict coronal rotation of first metatarsal. In Table 3, α-angle increased along with TSP increased. This is the first time quantifying the changing of coronal rotation with TSP. Major difference was occurred even when TSP was relatively small. Statistically, α-angle was different as the control and study group had same TSP grading 3. This phenomenon may imply that first metatarsal rotation occurs in mild hallux valgus. In the study group, TSP grading increased in different subgroups but clinically minor in practice.
This study has several limitations. This study was a retrospective study of female hallux valgus patients. All parameters were measured by one doctor, however measurement of DMAA may be less variable.2 Furthermore, our measurements were carried from the 2-D definition. The first metatarsal pronation is a 3-D rotation and mix with torsion of first metatarsal shaft.14 The α-angle maybe a primitive way to understand the complexity of the deformity and real 3-D measurements could reveal more details.