In our study, we explored the biomechanical outcomes of positive buttress and negative buttress of FNS internal fixation in the treatment of nonanatomically reduced femoral neck fracture based on finite element analysis. When the Pauwels angle was 30°, the positive 1-mm and 2-mm models had lower FNS stress than the negative buttress model. The positive 3- and 4-mm models showed FNS stress similar to that of the negative buttress model. But the four positive buttress models had similar stresses on the femur as the negative buttress model. When the Pauwels angle was 50°, the four positive buttress models had higher FNS stress than the negative buttress model, and the three positive buttress models (2 mm, 3 and 4 mm) displayed lower stress for the femur than the negative buttress model, which was not observed for the 1-mm model. Hence, positive and negative buttress in the treatment of femoral neck fracture with FNS will vary due to the Pauwels angle. When the fracture angle was 30°, the positive buttress group was superior to the negative buttress in terms of FNS stress, and the two groups were basically equal in terms of femoral stress. When the fracture angle was 50°, FNS internal implant bear more stress in the positive buttress group than negative buttress, resulting in less femoral stress in the positive buttress.
When the Pauwels angle was 30°, the positive buttress model had a lower displacement of the FNS than the negative buttress model, but the displacement of the femur similar to that of the negative buttress model. When the Pauwels angle was 50°, the positive buttress model had a higher displacement of both FNS and femur than the negative buttress model. This means that the positive buttress group was more stable than the negative buttress at a Pauwels angle of 30° but may not at a Pauwels angle of 50°.
Traditionally, “anatomical reduction” is a key factor in promoting fracture healing and avoiding postoperative complications , which has never been questioned. The real problem is that regardless of effort, there is still a high possibility of encountering a refractory femoral neck fracture, and it is difficult to achieve anatomical reduction under closed reduction in such cases. Therefore, we explored how to perform FNS internal fixation for femoral neck fractures in young patients without anatomic reduction. The concept of Gotfried reduction for femoral neck fracture has been proposed for almost a decade. Several studies have shown that Gotfried positive buttress reduction and fixation for femoral neck fracture result in similar clinical effects with anatomic reduction but are much better than Gotfried negative buttress reduction [27, 28].
The technique of Gotfried reduction is to stabilize unstable sub-cephalic fractures . In our study, Pauwels type I and type II femoral neck fractures were adopted as the fracture mode. Our results show that when the Pauwels angle was 30°, positive buttress was superior to negative buttress. However, when the Pauwels angle was 50°, this advantage will weaken. We also observed this with femoral displacement: when the angle was 30°, the effect of the positive buttress was more stable than negative buttress; this advantage is not seen in the case at Pauwels angles of 50°. As the Pauwels angle increased, the Von Mises stress and displacement of FNS fixation and the femur also increased. A retrospective clinical study from Zhao et al  found that positive buttress position reduction of femoral neck fractures in young patients showed a lower incidence of complications and reoperations compared with those of negative reduction using three parallel cannulated screws. Another retrospective study found that anatomic reduction and Gotfried positive buttress reduction group had higher Harris hip scores and lower femoral neck shortening than Gotfried negative buttress and suggested that achieving positive valgus reduction can also obtain satisfactory clinical results and should try to avoid negative buttress . Our findings are partial consistent with previous studies [16, 27, 28, 30], which reported that positive buttress is better than negative buttress. Possible explanations may be related to the following aspects. First, the Gotfried reduction method was first applied to sub-cephalic femoral neck fractures. In our study, when the Pauwels angle was 30°, it was considered a sub-cephalic fracture, consistent with the results of previous studies. When the Pauwels angle was 50°, it was considered a transcervical femoral neck fracture, which may be the source of the inconsistency. Second, when Gotfried et al. presented their concept, they established three pre-requisitives for sub capital fractures to heal: a positive buttress reduction, minimum neck-shaft angle of 135 degrees, and 180 degrees alignment in the lateral view or a minimum of 160 degrees. Our model only satisfies positive buttress reduction and does not incorporate two out of three major parameters. Therefore, our findings can only explain the stability of FNS in femoral neck fractures under non-anatomical reduction, but not under Gotfried positive support concept. Third, our model assumed movement on a smooth fracture surface, rather than interlocking the fracture ends, as in the original Gotfried reduction. Finally, all previous clinical studies used three cannulated screws for internal fixation, which is different from our FNS internal fixation, and the difference in internal fixation type is also one of the reasons for the inconsistent results. In our opinion, the clinical efficacy of FNS internal fixation with positive buttress may be related to the fracture angle, neck-shaft angle and alignment in the lateral view.
The limitations of our study are similar to those inherent to all finite element studies, whereby the model in this study was based on the femur being set as a homogeneous, continuous and isotropic elastic material. However, human bone is an isotropic heterogeneous material; thus, the material properties in the finite element experiment may have affected the results. Moreover, the model does not reflect the real relationships between bone fragments which are observed in real fracture site. Smooth ends of bone fragments in the model which are not observed in reality. However, as an initial biomechanics report, it can be considered reasonable. In the future, we need to construct more realistic bone fragments in real fracture site. In addition, our binding contact is placed at the junction of the internal fixation and the bone. However, under the loading force, a relative movement occurs between the bone and the FNS locking plate. But our contact settings is based on previous literature. It was acceptable since we recreated the optimum state of stable contact between bone and internal fixation. Finally, our results have not been verified by animal or clinical experiments. Our research setting is effective because it is based on the previous verified research . Nonetheless, our objective was to explore trends rather than absolute measurements. In this respect, the lack of experimental validation is rational.