Our approach to reconstructing forefoot deformities in patients with rheumatoid arthritis was as follows. In the lateral toes with mild or moderate joint destruction, shortening oblique osteotomy of the metatarsals is performed. With severe joint destruction, metatarsal head is resected. Arthrodesis of the first MTP joint is performed as a rule with resection arthroplasty in the lateral toes. When shortening oblique osteotomy in the lateral toes is indicated, the great toe is managed as follows: in young patients with mild joint destruction in the great toe (Larsen grades I and II) and who are able to ambulate well, Mitchell’s osteotomy is done. In older patients, or in patients with moderate or severe joint destruction (Larsen grades III to V), flexible hinge toe prosthesis is implanted.

Between 1987 and 2000, Mitchell’s osteotomy was performed on 47 feet in 31 patients, whose mean age was 53 years, Larsen grade was 2.5 and hallux valgus angle (HVA) was 35.0 (SD11.9). Arthroplasty with flexible hinge toe prosthesis was performed on 31 feet in 23 patients, 58 years, Larsen grade was 3.7 and HVA was 45.3 (SD12.9). After 1995, grommets were used in 17 feet. In 2002, we studied clinical results of them. 40 feet of Mitchell’s osteotomy had no pain and 7 feet had some pain. 26 feet of arthroplasty with flexible hinge toe prosthesis had no pain and 5 feet had some pain. Radiologically, HVA was 17.2 (SD10.3) in Mitchell’s osteotomy and 12.1 (SD6.3) in arthroplasty with flexible hinge toe prosthesis. Maintenance of correction by arthroplasty with flexible hinge toe prosthesis was better than Mitchell’s osteotomy significantly, especially more than 30 degrees of HVA. Without grommets, grade 0 was 8 feet, grade I was 3, and grade II was 3 feet judged by Granberry’s grade. But no revision surgery was performed by silicone synovitis or fracture of implant. With grommets, there were no fractures.

We added degree of HVA to management of operation after 2002. More than 40 degrees of HVA was considered flexible hinge toe prosthesis. After 2002, Mitchell’s osteotomy was performed on 7 feet in 6 patients, 53.7 years, Larsen grade was 2.4 and HVA was 32.3 (SD6.8). Arthroplasty with flexible hinge toe prosthesis was performed on 14 feet in 10 patients, 60.7 years, Larsen grade was 3.9 and HVA was 42.5 (SD7.5). Radiological result in these patients at 2005, HVA was 14.6 (SD4.9) in Mitchell’s osteotomy and 14.9 (SD2.5) in arthroplasty with flexible hinge toe prosthesis.