In 1926, Meyer21 was the first to describe “Z”-shaped osteotomy in the first metatarsal bone. Weil22 and Barouk23 also reported the use of this osteotomy, which was later named "Scarf osteotomy” by Weil. Generally, different Scarfs are selected according to the size of IMA, including short, medium, and long Scarf osteotomy22, respectively, for patients with mild, moderate, and severe IMA. Scarf osteotomy is multifunctional and is a widely used surgical method, which can correct IMA, improve DMAA, lengthen or shorten the length of the first metatarsal bone, and press the metatarsal head downward to increase the weight bearing2. Traditional Scarf osteotomy to correct IMA is limited by the width of the first metatarsal bone. Experienced doctors can only reserve 1/3 of the width, and some even increase the proportion to 1/46. When healing is not considered, the proximal medial cortex may encounter the distal lateral cortex at the extreme position. However, more than 50% of medial displacement could lead to instability, especially when the angle between metatarsals is high23. This instability is the limitation of traditional Scarf osteotomy. After translation, the cortex of two parts of the metatarsal bone will insert into opposite cancellous bone, leading to the troughing7, which is more common in elderly patients with osteoporosis and is likely to lead to postoperative metastatic metatarsal pain6. Traditional Scarf osteotomy is associated with many postoperative complications, including fracture, troughing, and stiffness, with rates of these complications reaching 10%, 35%, and 41.7%, respectively7; 23; 24. Many scholars have proposed improved surgical methods based on traditional Scarf osteotomy10; 25 to reduce complications after the procedure and improve the surgical indications.
In previous studies, the angle (HVA, IMA, and DMAA) used to evaluate imaging examinations depended on the bisection of the metatarsal axis. Since osteotomy destroys the long axis of the first metatarsal, the measurement results were inaccurate no matter whether the median or midpoint of the longitudinal axis was used for measurement after osteotomy26. Digital goniometer was recommended by many researchers to measure HVA8; 27. As mentioned in previous studies, we selected the line connecting the midpoint of the proximal and distal articular surfaces to calculate the angle28–30. Through finite element modeling and comparing the two methods of IMA correction on three-dimensional model, it was observed that rotary osteotomy could correct larger IMA. Under the same degree, the contact area of the broken end was more advantageous. With the increase of the correction angle, this advantage was more obvious, which guaranteed postoperative stability maintenance.
Although the relationship between the length of the first metatarsal bone and hallux valgus deformity is unclear, many scholars believe that the shortening of the first metatarsal bone is significantly positively related to the postoperative metastatic metatarsal pain, delayed healing, or nonunion of the broken end29–31. By measuring the absolute length of the first metatarsal bone, the difference between the two groups was not linear when correcting different IMA. The length of first metatarsal bone in the rotational group was significantly longer than in the translational group within the correction of 11°, especially in the correction of 3–8°. Furthermore, the length of the first metatarsal bone in the translational group was longer than that in the rotational group after the correction of 12°. It is an effective way to increase the length of the first metatarsal bone or press down the metatarsal bone when performing post osteotomy swing. Nevertheless, the rotational Scarf increases the DMAA because of an outward rotation with the center point of the osteotomy site. Our results also show that a bigger rotation angle was associated with a larger postoperative DMAA. For patients with a large preoperative DMAA or subluxation of the metatarsophalangeal joint, this mode of operation is not advantageous. Therefore, rotational Scarf is most effective for patients with a small preoperative DMAA.
Rotational Scarf can also reduce the extent of the troughing. The finite element model analysis demonstrated that, after rotating, the broken end of the rotational group was the contact between bone cortex to cortex, while that of the translational group was the direct contact between bone cortex to cancellous. The cancellous bone had a large strain concentration, which was a large deformation, indicating that it could cause troughing easily. On the contrary, the rotation group avoided this shortcoming successfully, avoiding the complications caused by troughing after operation from the source. In addition, due to troughing, the cortical bone of the proximal bone block was embedded into the cancellous bone of the distal bone block, resulting in a supination state of the distal metatarsal head relative to the proximal metatarsal bone, which may be a reason for the recurrence of hallux valgus. The stress at the contact part of the rotated metatarsal bone was mainly concentrated on the cortical bone, which could not only avoid the troughing to the maximum extent but also avoid the rotation deformity of the distal metatarsal head from the force line.
In addition, we also analyzed the stress distribution of plantar aponeurosis, plantar soft tissue, and ground. The results showed no significant difference between translational and rotational Scarf osteotomy. Unal's study32, through a three-point bending test, first observes the degree of separation at the broken end near the metatarsal wedge joint to judge the fixation strength. We took out the first metatarsal separately and performed a three-point bending test. It was found that after loading, the separation of broken ends in the rotational Scarf osteotomy model was slightly smaller than that in the translational Scarf, and the gap gradually increased with the increase of load, which also showed that the rotation group could more effectively maintain the stability of broken ends after surgery.
This study has some limitations that should be addressed. First, although the finite element results have verified the difference between rotiongal and translational Scarf osteotomy to some extent, clinical follow-up results are still required for confirmation. Secondly, since the finite element model did not fully reflect the true structure of the foot, it is necessary to conduct biomechanical experiments on cadaver specimens to confirm our findings. In addition, because rotational Scarf osteotomy will appropriately increase DMAA, which will have adverse effects on postoperative recurrence. It is currently advocated that rotational Scarf osteotomy applies to patients with moderate to severe hallux valgus with small DMAA.In the future, it is still necessary to follow up patients with moderate to severe hallux valgus with mismatched metatarsophalangeal joints who have undergone rotational Scarf osteotomy.