Researchers on hip joint preservation surgeries of NONFH have tried to use finite element analysis to investigate the biomechanical performance of core decompression24, rod implantation27,28 and CTT19. However, there has been no such investigations of LBT. Although LBT has obtained relatively higher success rate in the treatment of NONFH10, there are still concerns about its risk of fractures and comparisons among different techniques without evident warrant14,8. In this study, we performed finite element analysis of LBT in comparison with CTT, in order to provide biomechanical references for the current debates. Both LBT and CTT were calculated in their biomechanical aspect corresponding with the JIC classification.
In terms of validation of finite elemental models in this study, we compared our results with references and the physical character of the hip. Moreover, one patient in Type C1 and one patient in Type C2 were employed to validate the biomechanical properties of LBT. The shape and location of stress transfer path in this study is relevant to previous studies, which showed that stress transfer path was consistent with the distribution of bone density19,29,30. In addition, the average stress of weight-bearing area after fibula implantation, ranging from 2.32 to 2.59 Mpa, were in accordance with a recent study28. Therefore, the simulation results could reflect the physical status of the hip and could be used to analyze the effects of LBT and CTT.
The necrosis of femoral head varies from cases to cases, thus appropriate classification is prerequisite for the modelling. The JIC classification is a direct evaluation method to reflect the relative position between necrotic lesions and the weight-bearing area31–33, and has been proved to be an appropriated classification to predict stress distribution of the hip joint in vivo19,30,34,35. In the JIC classification, Type C1 and C2 are recommended to undergo joint-preserving therapies, so both types were employed in this study. Kuroda et al16 reported that JIC classification could assist with the selection of therapeutic options before the collapse of the femoral head, particularly for patients with JIC Type C2 NONFH35. As Type C2 accounts for 53% of all NONFH36 and is related to higher collapse rate recently35, more emphasis should be placed on the treatment of this type.
In all postoperative finite element models, the average von Mises stresses were alleviated in the studied regions for both LBT and CTT. It is reasonable that the former necrotic areas were replaced by fibular grafts. Additionally, the stiffness increased in all models (Table 3), which indicated that both techniques could effectively improve structural stability of proximal femur and prevent the collapse of femoral head. These results were supported by the simulation of patient-specific models (Table 4 & Fig. 3). The stiffness of CTT was 4.01%-5.38% higher than that of LBT, which might be related to the different drilling ways and the length of grafted fibula. Note that, CTT needed a longer grafted fibula (Table 1).
The average stress of the weight-bearing area, indicating its collapse risk, is of the most importance in this study. Interestingly, different JIC groups showed opposite results (Table 3). In JIC C1 group, the average stress in the weight-bearing area of the femoral head after CTT was relatively lower, whereas LBT was relatively lower in JIC C2 group. The opposite result may be caused by the combination of several factors. Firstly, the tail of grafted fibula was partially supported by the lateral cortex of femur in LBT, while there was no such support by cortex in CTT. Secondly, in JIC C2 group, the angle formed between grafted fibula and stress transfer path was smaller than that in JIC C1 group, and the force transmitted to fibula was more, making the mechanical effect of LBT better than CTT in JIC C2 group. Moreover, cases of Type C2 patients account for the highest proportion (53%) in all types of NONFH, meaning that more patients received appropriate treatment in LBT. Maybe this is the reason why there were relative higher clinical success rates of LBT than CTT (94.6%-96% vs 60%-90%).
An assignable concern of surgeons regarding FVFG procedures is the structural defect at the proximal femur, which may cause the fracture of femoral neck and subtrochanteric regions potentially12,37. Aldridge et al15 considered that LBT raised the stress by opening a window in the femoral neck. However, our results showed that the average stresses of femoral neck, trochanteric and subtrochanteric regions for LBT was lower than that preoperatively. Also, no yielding units appeared in these regions, meaning that the risk of fracture in these regions was reduced in LBT. Gao et al.10 reported 578 hips with LBT procedure did not occur proximal femoral fractures postoperatively. However, as an comparative technique in this study, CTT was reported of 0.7%-1% fracture rate in previous studies11,15,37, and those fractures all occurred in the intertrochanteric and subtrochanteric region after a fall37. Since the integrity of lateral femoral cortex is crucial to the structural strength against intertrochanteric and subtrochanteric fracture38,39, the impact of CTT on this integrity needs further study. From our clinical experience, adequate postoperative rehabilitation and fall prevention may reduce the prevalence of fracture.
The limitation of this study should be clarified. Firstly, only one healthy hip joint model was used for simulation and two patient-specific models for validation. However, all comparisons were based on the healthy hip joint model, so the deviation caused by differences of models was eliminated. Secondly, our study used a simplified model, regardless of some reality details. For instance, we did not consider internal fixation for both LBT and CTT, instead, we considered that they had got bony union. However, our study focused on the structural changes of proximal femur as same as previous studies19,40, the simplification would not affect the results. Last but not least, multiple factors affecting graft survival were not taken into consideration, such as the accuracy of surgery, revascularization realization, unbalance of creeping substitution, etc41–43. Further studies on the clinical comparison between LBT and CTT should combine these factors with biomechanical characteristics.