**4. Discussion**

In this study, we evaluated the mechanical properties of prophylactic femoral neck fixation using reconstruction interlocking screws for IMN in an osteoporotic femur model. Our composite bone model simulated a healed femoral shaft fracture fixed with an IM nail for osteoporosis. Our findings revealed that the reconstruction interlocking constructs showed a higher load-to-failure and prevented delayed femoral neck fractures compared with the standard interlocking construct in an osteoporotic femur model.

Surgeons prefer prophylactic neck fixation for treating femoral fractures. Several authors have advocated prophylactic femoral neck fixation for all femoral shaft fractures because of concerns regarding iatrogenic or missed femoral neck fractures [6,7]. Patton et al. [4] reported late femoral neck fractures after IM nail fixation of femoral shaft fractures in elderly patients. Fourteen patients (2.7%) developed a proximal femoral fracture adjacent to an IM implant in their series. Among them, 11 fractures occurred within months to years. Most patients were > 60 years of age and had osteoporosis and low-energy injuries. Therefore, we suggest that prophylactic neck fixation during the construction of osteoporotic femoral shaft fractures with a reconstruction nail is necessary. Bögl et al. [5] reviewed 897 patients treated for low-energy diaphyseal femoral fractures in Sweden. In their study, 640 patients were treated with IM nails with femoral neck fixation, whereas 257 patients were treated without femoral neck protection. The authors found a fivefold decrease in the risk of reoperation for peri-implant fractures and half the risk of major reoperation when treated with femoral neck fixation. Our study also demonstrated biomechanically that prophylactic neck fixation could effectively prevent femoral neck fractures after IMN for osteoporotic femoral shaft fractures.

Few biomechanical studies have been conducted on prophylactic neck fixation. Previous biomechanical studies using piriformis entry nails have reported that the load-to-failure was similar, regardless of femoral neck protection [13]. They postulated that this may have resulted from a large entry hole through the piriformis fossa, which created a significant bone defect at the base of the femoral neck. Another study using a GT entry nail with neck protection showed a higher load-to-failure than a piriformis entry nail, similar to that of an intact femur [9]. In the present study, higher load-to-failure values were observed in the RN group. This is the result of the difference between the nail entry portal and composite

bone used. For osteoporotic femoral shaft treatment, GT entry nail and prophylactic neck fixation may be considered appropriate.

The location of the entry site contributes to the strength of the femoral neck during fixation. In a cadaveric study, Miller et al. [14] showed that the entry site of the piriformis fossa could significantly weaken the neck. All the specimens in their study sustained basicervical fracture patterns under mechanical loading. Strand et al. [15] conducted a cadaveric study comparing entry portals at either the piriformis fossa or the GT of the cadaver. All femurs of the piriformis fossa entry portals sustained basicervical fractures at the entry site. The piriformis fossa group showed a lower load-to-failure than the GT group. As elderly patients with osteoporosis have a lower bone density around the femoral neck, the piriformis entry portal may pose a higher risk of femoral neck fractures in these patients. When performing nailing for femoral fractures, especially in elderly patients, avoiding the piriformis entry portal is recommended. Therefore, our experimental procedure was conducted using the trochanter entry portal.

Synthetic femurs have been widely used and accepted as a substitute for cadaveric specimens in biomechanics. Recently, a new osteoporotic synthetic femur has been introduced by the increasing osteoporosis population. The wall thickness and bone density were reduced to simulate osteoporotic bone. In biomechanical studies, the osteoporotic synthetic bone shows similar axial loading results compared with the osteoporotic bone [16].

Our study had several limitations. First, this was a biomechanical study using synthetic femur models and, thus, may not accurately represent human bone mechanics, especially in elderly patients. However, synthetic bone provides several advantages over cadaveric bone; thus, it is preferred in biomechanical studies [12,17]. Furthermore, synthetic bones can provide standard sizes and properties between specimens and guarantee implantation techniques' reproducibility. In addition, cadaveric specimens of varying dimensions, ages, and bone densities were excluded. Moreover, our approach is meaningful because we used a newly developed synthetic osteoporotic bone biomechanically similar to an osteoporotic cadaveric femur [16].

Second, the axial load provided by a mechanical testing machine may not accurately mimic the physiological load or vector experienced during a standard ground-level fall. We assumed that femur neck fracture is a result of axial load through the mechanical axis of the femur. We decided to use this assumption to enhance reproducibility, with the thought that weight bearing goes through the mechanical axis from a functional perspective. Other forces we did not simulate can cause femur neck fractures. However, the testing model used in this study has been validated and previously utilized in multiple biomechanical studies. Consequently, this is an appropriate and applicable method for isolating the strength of the proximal femur.

Third, we only tested load-to-failure rather than cyclic loading. Failure loading simulates a fall or other trauma. Moreover, some clinical reports had either no or minimal trauma due to the mechanism of the fracture, implying that femur neck fracture may be due to repetitive loading, leading to a stress fracture. However, our goal was to investigate the protection the fixation construct provides against a catastrophic event such as a fall.

Finally, this study was conducted only with an osteoporotic bone model, so the effect of prophylactic neck fixation is specific to the osteoporotic femur. Additional samples and comparison studies with non-osteoporotic bone may have provided a more accurate representation of the load-to-failure of the femur models.
