From 2D to 3D: Understanding Hallux Valgus Deformity.

Abstract

Hallux valgus deformity (HVD) is usually considered a “bread and butter” problem for orthopaedic foot and ankle surgeons. There is an overall understanding that “we’ve got it covered.” But do we really have it covered? Do we really understand this extremely complex multifactorial, multifocal, and multiplanar foot deformity, its risk factors, pathophysiology, deformity components, treatment goals, and expected outcomes? In recent systematic reviews of the literature, Barg et al1 reported that around 10% of HVD patients treated surgically are dissatisfied with the results of the surgical treatment, and Lalevee et al6 demonstrated that the postoperative recurrence rate of the deformity is as high as 64% after a minimum follow-up of 5 years. The truth is that we cannot really treat accurately things that we do not completely understand. Hallux valgus is a 3-dimensional (3D) multifaceted deformity that can involve multiple tarsal joints in the hindfoot, midfoot, and forefoot, as well as a multitude of soft tissue imbalances. Currently, the interpretation, staging, and surgical treatment planning of HVD for most foot and ankle surgeons are performed using physical examination findings and 2-dimensional (2D) conventional radiographic assessment. This traditional assessment limits the 3D interpretation of the deformity and the multiple involved factors, such as the rotational profile of the first ray, metatarsal-sesamoid interaction, and anatomical characteristics of bones and joints. The study by Ji et al4 in the current issue of FAI supports the critical transition from a 2D to a 3D assessment of HVD. The authors compared anatomical features of the first tarsometatarsal joint (1stTMTJ), particularly the shape of the articular surface of the proximal first metatarsal (M1), between HVD patients and controls using weightbearing CT (WBCT) imaging and 3D bone modeling. They identified 4 distinct anatomical features for the proximal M1 articular surface: continuous-flat, separated-flat, continuous-protruded, and separated-protruded. The continuous-flat morphology was significantly more prevalent in HVD patients than in controls (74.4% vs 16.5%). In comparison, the separated-protruded shape was significantly more prevalent in the control population when compared to HVD patients (48.1% vs 4.3%). They also found that patients with a flat proximal M1 configuration also demonstrated significantly increased hallux valgus and intermetatarsal angles. Even though their study cannot guarantee a cause-effect relationship, the interpretation of their findings supports the theory that some people could possibly be predisposed to develop HVD by having a flat and potentially more hypermobile and unstable 1stTMTJ. These concepts are not new and have been proposed before, primarily based on cadaveric, anatomical, and 2D conventional radiographic studies. Doty et al3 also demonstrated in cadaveric specimens with and without HVD that increased 2D radiographic 1stTMTJ angulation significantly and positively correlated with increased hallux valgus angle. They also observed different M1 proximal articular surface morphologies in anatomical dissections (continuous, bilobed, and separated, with the continuous pattern being the most common). Koury et al5 have cautioned that 2D radiographic assessment of the 1stTMTJ morphology and inclination is highly influenced by foot positioning and angulation of the x-ray beam, which significantly limits the interpretation of 1stTMTJ shape and alignment based on 2D imaging. WBCT imaging has been transforming the assessment of complex foot and ankle deformities and has already contributed considerably to an improved and more accurate 3D understanding of multiple deformity patterns associated with HVD.2 The body of literature on this subject has grown substantially in the last several years based on WBCTderived data. One emerging example is the rotational profile of the first ray. Steadman et al9 reported that the normal rotation of the distal aspect of M1 in relation to the ground in control patients is around 2 (±4) degrees of pronation. Mansur et al7 more recently reported similar values of normality for distal M1 pronation using WBCT, of around 4.2 degrees in controls and 11.5 degrees in HVD patients. More importantly, the authors called attention to the inaccuracy of conventional 2D anteroposterior radiographic assessment in predicting M1 pronation based on the roundness of the lateral aspect of the M1 head. This concept was classically proposed by Okuda et al,8 and popularized by Wagner and Wagner.10 The authors showed that sesamoid station/positioning represented the main factor influencing the round appearance of the M1 head in anteroposterior conventional radiographs. The current study of Ji et al4 exemplifies well the opportunities provided by WBCT imaging for the utilization of both 2D (represented by the sagittal plane 1stTMT angle) and proper 3D assessments (represented by the shape modeling reconstruction of the anatomy of the proximal aspect of M1) in HVD patients. The authors should be commended for taking one step further in the long pathway to 1180573 FAIXXX10.1177/10711007231180573Foot & Ankle Internationalde Cesar Netto article-commentary2023