Preoperative radiographic and clinical factors associated with postoperative floating of the lesser toes after resection arthroplasty for rheumatoid forefoot deformity

Institutional review board approval of Hokkaido University Hospital was obtained prior to initiation of this study. Between 2007 and 2016, a total of 43 patients (47 ft) at a single center who were followed up for at least 12 months, and were diagnosed with rheumatoid forefoot deformities by the senior authors of this paper, that required arthrodesis of the first MTP joint, resection arthroplasty of the metatarsal heads of the lesser toes, and surgical repair of hammer toe deformity, were included in this study. One of the authors reviewed the medical records at our institution to identify patients with RA who underwent resection arthroplasty of the metatarsal heads of the lesser toes. Eligible patients were identified based on their diagnosis (RA) and the surgical procedure performed (resection arthroplasty of the metatarsal head of the lesser toes). All surgeries were performed by four senior surgeons. The indications for surgery in all patients were metatarsalgia associated with hallux valgus or joint destruction caused by RA. The patients who had no first MTP joint destruction and the hallux valgus deformity was less than 50 degree were performed joint-preserving surgery and were excluded in this study. The mean age of the subjects at the time of surgery was 61.8 (39–80) years. All the patients were women, and the follow-up period was 36.5 (12–114) months. All surgeries were performed by the same senior surgeon.

The operative methods were precisely described in our previous study [13, 14]. Arthrodesis was performed through a dorsal longitudinal incision centred over the first MTP joint. The inclination of the first metatarsal in preoperative X-ray was referred to achieve an optimal angle of dorsiflexion of the first MTP joint [14]. The first MTP joint was fused using cannulated cancerous screw or Kirschner wires (K-wires) (Fig. 1). An arthrodesis of the interphalangeal (IP) joints of the second to fourth toes or second to fifth toes was performed. The metatarsal heads of the lesser rays were resected at the anatomic neck of the involved metatarsal via Lipscomb’s dorsal approaches [15].

The patients were allowed full weight-bearing at 2 weeks postoperatively, using a special shoe that allowed them to bear weight on their hindfoot. The special shoe was worn until fusion of the arthrodesis was confirmed.

X-rays were taken during full weight-bearing, using lateral and standard dorsoplantar views before surgery and at the final follow up. Preoperative and postoperative evaluation was performed by one author. The valgus angle between the first metatarsal bone and the proximal phalanx (HVA: hallux valgus angle), the intermetatarsal angle between the first and second metatarsals (M1-2A: first-second intermetatarsal angle), and the intermetatarsal angle between the first and fifth metatarsals (M1-5A: first-fifth intermetatarsal angle), and the shortening distance of the metatarsal bone were measured on AP radiographs before surgery. The HVA were measured on AP radiographs after surgery. Tibial sesamoid position is measured on weight-bearing AP radiographs on a scale from 1 to 7 as described by Hardy and Clapham before and after surgery [16]. The joint destruction of the first MTP joint was evaluated by using the Larsen grading system [17]. The Larsen Grading Scale is used to determine the extent of radiographic changes due to rheumatoid arthritis. The scale score from 0 to 5 are based on the amount of spacing between joints, bony outlines, and level of joint erosion on radiographs. Total arc of first MTP joint was measured before surgery using a standard mechanical goniometer. Dislocation of MTP joints were graded from 0 (no dislocation) to 3 (complete dislocation) using preoperative weight-bearing lateral radiographs and the findings of inspection [18]. Since preoperative grading system of the dislocation of the MTP joints is based on the metatarsal head, it is impossible to grade if it is resected (especially grade 2 and 3). Therefore non-grounded toes while weight-bearing were defined as postoperative floating toes. All toes were divided into non-floating or postoperative floating groups. All the participating feet were also evaluated after categorizing them based on whether the medial, second and third toes, or lateral, fourth and fifth toes, lessor toes were affected. Regarding the foot with floating toes in both columns, we compared both columns by radiological and physical findings and divided into each group. Feet with absence of postoperative floating of both toes were classified as the non-floating group, while those with postoperative floating of one or both of the toes were defined as the postoperative floating group.

Before analysis of the X-ray films, intraobserver reproducibility was calculated by one author based on 5 consecutive measurements using the same X-ray film. Intraobserver consistency in all measurements was assessed by calculating the coefficient of variation (%, CV), which equals the standard deviation divided by the mean value. CVs were calculated for all the measurements. The CV of all measurements from single observer ranged from 1.5 to 3.1%. All of the variations were less than 10%, which was considered acceptable for comparison study purposes. To reduce data variation, a single investigator performed all study measurements.

Postoperative clinical outcomes were evaluated on the basis of data obtained during interviews and physical examinations. The Japanese Society for Surgery of the Foot (JSSF) hallux metatarsophalangeal-interphalangeal scale including hallux valgus deformity, pain, range of motion, activity of daily living, and function were used to assess clinical outcomes for the entire foot [19].

Differences between non-floating and postoperative-floating were analysed by applying the unpaired t-test or the Mann-Whitney U test for continuous variables and the chi-square test or Fisher’s exact probability test for categorical variables. Statistical analyses were performed using JMP Pro version 10.0 statistical software (SAS Institute, Cary, NC). Statistical differences were considered significant for values of p < 0.05.