Is the selection of a long femoral prosthesis required in hip artroplasties after failed hip nailing? Biomechanical study

Purpose: Conversion to arthroplasty is a surgical option as a result of the failure of the proximal femoral nail (PFN) used in the treatment of unstable intertrochanteric fractures. The aim of our study is to biomechanically compare the cemented and uncemented femoral stems that bypass the distal locking screw hole defect in two cortex lengths and the uncemented femoral stems ending in the proximal of screw hole in terms of periprosthetic fracture risk. Material & Methods: Fifteen Sawbones were divided into 3 groups equally (Group I: cemented, bypass the screw hole; group II: cementless, standard stem; group III: cementless, long stem) in which unstable ITF fracture model was created in accordance with AO A2.1 classification. After standard PFN was applied to all models, all implants were removed and femoral stems were applied. Axial failure loading (15 mm/min) was applied to all models after axial (15 N - 350 N; 10000 cycle) and rotational (0,5 Nm - 10 Nm; 10000 cycle) loads were applied. In the axial and rotational mechanisms, the stiffness of the models before and after cyclic loading and the failure load of the models were calculated and compared between the groups. Results: Stiffness medians (Q1-Q3) of models with cyclic rotation at initial (first 1000 cycles) were 3903 (3430–4247) Nmm/° in group I, 3392 (2487–3586) Nmm/° in group II, 3512 (3295–3884) Nmm/° in group III. There was no significant difference in the initial stiffness of the first 1000 cycles between the groups (p= 0.208). Final stiffness (10,000 cycles) medians with cyclic rotation (Q1-Q3), 3817 (3699–4459) Nmm/° in group I, 3408 (2963–3546) Nmm/° in group II, 3476 (3246-3806) Nmm/° in group III. Group I was significantly better than group II (p=0.09). No significant difference was found between group I and group III (p=0.104), and between group II and group III (p=0.322). Stiffness medians (Q1-Q3) of the models with cyclic axial loading at initial (first 1000 cycles) were 930 (864–1034) N/mm in group I, 739 (682–807) N/mm in group II and 930 (902–1033) N/mm in group III. Between group I and group II, group I was significantly better than group II (p=0.011). Between group II and group III, group III was significantly better than group II (p=0.006). There was no statistically significant result between group I and group III (p=0.832). Final stiffness (after 10,000 cycles) medians (Q1-Q3), 956 (927–1074) N/mm in group I, 775 (701–830) N/mm in group II, 919 (910–830) in group III 1050) N/mm. Between groups I and II, group I was significantly better than group II (p=0.011). Between group II and group III, group III was significantly better than group II (p=0.006). There was no statistically significant result between group I and group III (p=0.832). The mean axial failure load medians (Q1-Q3) of the models after axial loading were 5537 (5229 – 5857) N in group I, 4101 (3938 – 4195) N in group II, 4851 (4432 –5752) N in group III. Between groups I and II, group I was significantly better than group II (p=0.002). Between group II and III, group III was significantly better than group II (p=0.024). There was no statistically significant result between group I and group III (p=0.437). Conclusion: Cemented or uncemented fixation did not significantly change the stiffness of the reconstruction when conversion to hip arthroplasty was required after proximal hip nailing for osteoporotic, unstable intertrochanteric fractures. Femoral stem lengths that bypass the distal locking screw hole with 2 cortical diameters after removal of the PFN reduced the risk of periprosthetic fracture.