The ratio between the screw distance and femoral neck width on lateral radiography is a reliable predictor of femoral head necrosis: a clinical review and corresponding numerical simulations Authors: Hong Li, Shengyu Wan, Jian Zhang, Fan Wu, Xiaozhong Luo, Chao Wu, Xu Lin, Jingchi Li Published: 2024 ABSTRACT AbstractBackgroundFixation instability serves as an initial trigger for femoral head necrosis in patients with femoral neck fractures undergoing internal fixation. The configuration of screw trajectories is critical in determining the stability of the fixation. Our previously published study indicates that abduction of screw trajectories may enhance postoperative angular stability. However, the effectiveness of a wider screw distribution on the sagittal plane in optimizing fixation stability and reducing the risk of femoral head necrosis remains uncertain.MethodsThis retrospective study reviewed imaging data from patients with femoral neck fractures who were treated with inverse triangle cannulated screw fixation. Postoperative lateral radiographs were used to evaluate the screw configuration strategy. The distance between the two cranial screws and the femoral neck width at its narrowest point were measured. The ratio of these two measurements was calculated to represent the screw spread grade. Demographic and imaging parameters were compared between patients who developed femoral head necrosis and those who did not. Regression analysis was conducted to identify potential risk factors associated with femoral head necrosis. Additionally, numerical models were employed to simulate the biomechanical impact of variations in screw spread grade.ResultsThe clinical analysis revealed that a lower screw distance-to-femoral width ratio was associated with a higher incidence of femoral head necrosis. This ratio was also identified as an independent risk factor for femoral head necrosis. Biomechanical simulations corroborated these findings, demonstrating that models with a lower screw-to-femoral width ratio exhibited reduced fixation stability and increased stress distribution.ConclusionsWider distribution of screws along the sagittal plane may significantly mitigate the risk of femoral head necrosis by enhancing fixation stability. The optimization of screw trajectories, particularly through broader screw spacing facilitated by a navigation system, emerges as a promising strategy for improving the local biomechanical environment and reducing the likelihood of femoral head necrosis.Keywords: Femoral neck fracture, Femoral head necrosis, Screw fixation stability, Finite element analysis, Screw spread grades Introduction Internal fixation operations have been widely employed in patients suffering from femoral neck fractures [1, 2]. However, femoral head necrosis remains a significant postoperative complication in these individuals. Although total hip arthroplasty can achieve excellent clinical outcomes for femoral head necrosis resulting from femoral neck fractures[3, 4], reducing the incidence of femoralhead necrosis remains crucial for optimizing patient prognosis and alleviating social and economic burdens. The potential risk factors associated with this complication have been extensively investigated [1, 5]. Among these, fixation instability has consistently been identified as an initial trigger for the progression of femoral head necrosis. Therefore, enhancing fixation stability is crucial for optimizing the prognosis of patients with femoral neck fractures undergoing internal fixation [6, 7]. Although recent years have seen the introduction of several innovative fixation devices, the inverse triangle cannulated screw fixation technique remains the most widely utilized approach for femoral neck fractures [1, 8]. The configuration of screw trajectories is critical, as it directly influences fixation stability. Among the various strategies, a wider screw spread is frequently employed in internal fixation procedures [9, 10]. Biomechanical studies have demonstrated that this configuration can enhance angular stability. Our previous research further established that variations in screw insertion angles on the coronal plane can significantly impact both fixation stability and the risk of femoral head necrosis, attributable to the wider distribution of screws within the femoral head [11, 12]. However, the biomechanical efficacy of optimizing screw configuration by increasing the screw distance to reduce the risk of femoral head necrosis remains to be fully elucidated. In this study, we conducted a comprehensive investigation encompassing both clinical review and numerical mechanical simulations. We quantified the ratio between the screw distance and femoral width on lateral radiographs, and employed statistical analyses to assess the significance of this parameter in relation to the risk of femoral head necrosis. Additionally, the biomechanical implications of this ratio on fixation stability were validated through numerical mechanical simulations. The findings of this study provide valuable theoretical insights for optimizing screw configuration strategies in the management of femoral neck fractures. Materials and methods Clinical review Patient data collection The study protocol was reviewed and approved by the ethics committee of our hospital (Approval No. 2023–008). We conducted a retrospective review of patients who sustained femoral neck fractures and subsequently underwent inverse triangle cannulated screw fixation in the orthopaedic department of our hospital. The exclusion criteria were meticulously defined to ensure the integrity and homogeneity of the study population: 1) Patients with femoral head necrosis prior to the occurrence of a femoral neck fracture; 2) patients with a history of alcohol dependence or corticosteroid abuse; 3) patients presenting with pathological fractures of the femoral neck due to conditions such as bone tumors or autoimmune diseases; 4) patients requiring revision surgery within a two-year follow-up period due to complications such as screw pull-out or breakage; 5) patients aged over 60 years, as this demographic is generally considered more suitable for total hip arthroplasty (THA) rather than triangle cannulated screw fixation; 6) patients who chose conservative treatment options; 7) patients lost to follow-up during the study period; and 8) patients lacking adequate lateral radiographs, which precluded feasible parameter measurements. Demographic data (patients age, sex, and body mass index (BMI) were recorded for all eligible patients according to their medical record. A highly experienced orthopedic surgeon performed all surgical procedures, ensuring consistency in technique and outcomes. The screw types and sizes used were standardized across the cohort. Furthermore, the tip-apex distance (TAD) for all enrolled patients was maintained at less than 25 mm, in accordance with best practices for optimizing fixation stability (Fig. 1). Inclusion and exclusion criteria for the current patient series, as well as the measurement of pre-operative Pauwle angles and the ratio between screw distance and femoral neck widths Radiographic parameter measurements Radiographic parameters were measured by an experienced senior orthopedic surgeon to ensure accuracy and consistency. Anteroposterior radiographs were obtained for all patients on two key occasions: immediately postoperatively and at the 2-year follow-up after internal fixation. The Pauwels angle and Garden classification were assessed using preoperative imaging data, with the Garden classification also serving to evaluate fracture displacement status. According to the Garden classification, patients with Garden I and II fractures were categorized as having "non-displaced fractures," while those with Garden III and IV fractures were categorized as having "displaced fractures" [6, 13]. During the 2-year follow-up period, patients exhibiting signs of femoral head collapse or subcortical cyst formation on radiographic imaging were diagnosed with femoral head necrosis. Additionally, the distance between the two cranial cannulated screws and the width of the femoral neck at its narrowest point were measured using immediate postoperative lateral radiographs or intraoperative C-arm radiographs. The ratio of these two measurements was calculated to determine the screw spread grade. Statistical analyses Parameter values are reported as means ± standard deviations for continuous variables and as percentages for categorical variables. Statistical analyses were conducted using IBM SPSS 26.0 software. One week after the initial measurement, parameters from 20 randomly selected patients were re-measured by both the orthopedic surgeon who initially measured the radiographic parameter values and another senior orthopedic surgeon. This study computed the value of intraclass correlation coefficient (ICC) to judge the repeatability of continuous variables. This study calculated the intraclass correlation coefficient (ICC) to judge the repeatability of Pawuls angles and the ratio between screw distance and femoral neck length. While the kappa value was computed to evaluate that of Garden calcification based fracture displaced status [14, 15]. An ICC and kappa ≥ 0.8 indicate excellent repeatability in measuring imaging-based parameters [16, 17]. When comparing the differences in demographic and radiographic parameters between patient series with and without femoral head necrosis, we utilized the independent samples Student's t-test for continuous and the chi-square test for categorical variables [3, 4]. We also conducted binary logistic regression to investigate potential independent risk factors for the complication [18, 19]. Univariate analyses of each potential risk factor were performed, variables that achieved a level of p < 0.1 were entered into multivariate analyses [20, 21]. In the multivariate analysis, variables were considered independent risk factors for femoral head necrosis when their P values were less than 0.05.Subsequently, ROC curve analyses were carried out to assess the predictive value of potential risk factors, and the area under the curve (AUC) was calculated as an indicator to evaluate the predictive performance of independent risk factors. A P value less than 0.05 indicated a significant difference (Fig. 2 and Fig. 3) [22, 23]. Significant Difference Computation: Based on the results of the significant difference analysis, patients with femoral head necrosis exhibit markedly higher Pauwls angles, a significantly greater incidence of displaced fractures as determined by Garden classification, and a notably lower ratio of screw distance to femoral neck width. Additionally, ROC curves were generated for predicting femoral head necrosis in these patients Typical cases supporting the primary research topic of this study are presented. One patient with a limited screw distance to femoral neck width ratio experienced femoral head necrosis. In contrast, another patient with an optimal screw distance to femoral neck width ratio achieved ideal bone healing without any signs of femoral head necrosis Surgical simulations and stress and deformation values computation under physiological loading condition Construction of the intact finite element (FE) model The proximal femur model was constructed using the outline of a syn-bone model rather than patient-specific anatomy. This approach was chosen to circumvent ethical concerns and to minimize the confounding effects of anatomical variability across different individuals [24, 25]. CT scan (slice thickness = 0.55 mm) was performed on this model and the proximal femur was defined as the region extending from the tip of the femoral head to 30 cm below the lesser trochanter. The outline of the proximal femur was then reconstructed in 3D-CAD software based on the CT scan data, consistent with methodologies employed in our previously published studies [12, 25]. This model construction strategy demonstrated superior computational efficiency and accuracy compared to traditional reverse modeling approaches [24, 25]. The cortical and cancellous bone structures were individually modeled, with their respective outlines derived directly from the CT imaging data. This method ensured a high level of anatomical fidelity while enhancing the robustness of the numerical simulations [26, 27]. Construction of triangle screw fixed femoral neck fracture models with different screw spread grades To simulate the fixation stability of various screw fixation strategies in an unstable fracture model, Pauwels type III femoral neck fractures were incorporated into the proximal femur model. The Pauwels angle was standardized at 55°, with the fracture positioned centrally within the femoral neck. A 1 mm gap was introduced to represent the fracture line. A numerical model of a semithreaded lag cannulated screw was also developed using 3D-CAD software [28, 29]. For the inverse triangle screw fixation simulation, a control group numerical model was created, wherein cannulated screws were inserted along a trajectory coaxial with the long axis of the femoral neck, maintaining a screw distance-to-femoral neck width ratio of 0.4. To explore the effects of different screw spread grades. The ratio between screw distance and femoral neck width was systematically varied to 0.6 and 0.2 by adjusting the spacing between the two cranial screws at the narrowest point of the femoral neck. These ratios were selected based on the cutoff values identified in clinical observations. For each screw distance-to-neck width ratio, separate numerical models were constructed to simulate parallel and crossed screw configurations, with trajectories set for abduction or adduction (Fig. 4). The construction of a femoral neck fracture model with a Pauwls angle of 55°. Additionally, models were developed utilizing various screw configuration strategies with differing screw distances Boundary and loading conditions When defining material properties in this study, all components—including bony structures (both cortical and cancellous) and internal fixation devices—were characterized as isotropic materials with varying elastic moduli and Poisson's ratios. The inferior of femoral models were fully constrained. Physiological compressive load was applied to the femoral head, oriented at 10° lateral to the coronal plane and 9° posterior to the sagittal plane [26, 30]. A compressive load of 2100 N, representative of forces encountered during daily activities, was applied to calculate the displacement and stress experienced by the femoral head (Fig. 4). Stress values across the femoral head and screws, and the failure load (fixation strength) in models with different screw trajectories were computed and recorded separately [30, 31]. Fixation strength was evaluated by determining the load at which the maximum displacement of the femoral head reached 10 mm. These measurements were recorded to assess the potential biomechanical risk of femoral head necrosis associated with each screw configuration strategy. Results Retrospective review of clinical data Patient collection and significant difference verification A total of 94 patients with femoral neck fractures treated with inverse triangle cannulated screw fixation were enrolled in this study (average follow-up period = 23.8 months). According to the clinical review, 27.66% screw fixed patients suffer femoral head necrosis during the follow-up period (2-years). Excellent interobserver and intraobserver reliability can be reached based on the computed ICC and kappa values, for which different parameters are shown in the Table 1. There were no significant differences in demographic covariates (age, sex, or BMI) between patients with and without femoral head necrosis. In contrast, a significantly larger parotid angle and a greater incidence of “displaced fracture” can be observed in patients with femoral head necrosis. Moreover, the ratio between the screw distance and femoral width was significantly lower in patients with femoral head necrosis. Table 1ICC and Kappa values of inter- and intraobserver reliability when measuring imaging based parametersInterobserverIntraobserverPauwls angle0.8770912Garden calcification based fracture displaced status (no displaced / displaced fracture)0.8981Ratio between screw distance and femoral neck length0.8840.865 ICC and Kappa values of inter- and intraobserver reliability when measuring imaging based parameters Independent risk factors and parameter prediction values for femoral head necrosis When identifying independent risk factors for femoral head necrosis, based on the results of univariate logistic regression analyses, the P values of a larger Pauwls angle, a greater incidence of “displaced fracture” and a lower ratio of the screw distance to the femoral neck width were less than 0.1 according to univariate regression analysis. Therefore, these factors were included in the multivariate analysis. The results showed that a greater incidence of “displaced fracture” and a lower ratio of the screw distance to the femoral width were important risk factors for femoral head necrosis in cannulated screw-fixed patients (Table 2 and Table 3). Finally, we performed receiver operating characteristic (ROC) curve analyses to assess the ability of a larger Pauwls angle, a greater incidence of “displaced fracture” and a lower screw distance–femoral neck width ratio to predict femoral head necrosis risk. The AUCs of these three parameters were 0.666, 0.689, and 0.705, respectively (Table 4, Fig. 2 and Fig. 3). Table 2Significant comparison between variates for patients with and without femoral head necrosisWith Femoral head necrosisWithout Femoral head necrosisP valueSex(Male/Female)11/1531/370.775Age51.23 ± 11.9350.59 ± 9.610.787BMI22.1 ± 3.3821.62 ± 2.790.491Pauwls angle56.12 ± 9.4550.97 ± 10.060.027*Garden calcification based fracture displaced status (no displaced/displaced fracture)2/2431/370.001**Ratio between screw distance and femoral neck length0.205 ± 0.090.271 ± 0.1060.006***, statistical significance (P < 0.05)**, statistical significance (P < 0.01) Significant comparison between variates for patients with and without femoral head necrosis *, statistical significance (P < 0.05) **, statistical significance (P < 0.01) Table 3Logistic regression analysis of AVFOR95% CIPUnivariate analysesSex0.8750.3512.180.775Age0.9940.951.040.785BMI0.9480.8141.1030.487Pauwls angle0.950.9060.9950.032#Garden calcification based fracture displaced status0.0990.0220.4540.003#Ratio between screw distance and femoral neck length5.9693.26410.9130.016#Multivariate analysisPauwls angle0.4060.9780.9281.031Garden calcification based fracture displaced0.019*0.1480.030.734Ratio between screw distance and femoral neck length0.036*3.5931.21910.59#,variables that achieved a significance level of p < 0.1 in the univariate analysis*, statistical significance (P < 0.05) Logistic regression analysis of AVF #,variables that achieved a significance level of p < 0.1 in the univariate analysis *, statistical significance (P < 0.05) Table 4The cut-off value, sensitivity and specificity for femoral head necrosis predictionCut-off valueSensitivitySpecificityAUCPLarger pauwls angle53.50.6920.7060.6660.007*Higher incidence of displaced fracture based on Garden calcification1.50.9230.4560.6890.001**Lower ratio between screw distance and femoral neck length0.2090.7210.6540.7050.001***, statistical significance (P < 0.05)**, statistical significance (P < 0.01) The cut-off value, sensitivity and specificity for femoral head necrosis prediction *, statistical significance (P < 0.05) **, statistical significance (P < 0.01) Numerical mechanical simulations An overall consistent variation tendency of computed parameter values was recorded in different models, which was consistent with the clinically observed phenomenon: wider screw spread can effectively optimize the instant postoperative local biomechanical environment. Specifically, although the differences between femoral head displacement, fixation failure load, and screw stress values were less than 10% between the different models, obviously lower stress values of the femoral head can be observed in models with a wide screw spread (0.6). Compared to that in the control group, the femoral head stress in the models with limited screw spread increased by nearly 100% in both the parallel and adduction trajectories. In contrast, compared to those in the limited screw spread models, the femoral head stress was reduced by more than 80% in the wide screw spread model with a parallel trajectory. This percentage was also reduced by 60% in the abduction screw trajectory model (Fig. 5 and Fig. 6). Stress and deformation values of the femoral head under physiological loading conditions: Compared to models with a wide screw distribution, those with a limited screw distribution exhibit stress concentration on the femoral head. This phenomenon may lead to an increased incidence of bony compaction and fixation instability Nephograms of stress distributions on cannulated screws in models with different screw-to-femoral neck width ratios and different screw trajectories. Obviously higher stress values can be observed in models with limit screw trajectories Discussion Internal fixation is the most widely employed surgical technique for the treatment of fractures [1, 32]. Biomechanical studies have established that variations in screw configuration strategies can significantly influence fixation stability and the associated risk of complications [33, 34]. However, whether optimizing screw spread grades can effectively mitigate the risk of femoral head necrosis in patients undergoing inverse triangle cannulated screw fixation remains unclear. In this study, we aimed to address this gap by measuring the distance between the two cranial screws and the width of the femoral neck at its narrowest point using immediate postoperative lateral radiographs. A lower ratio of these two parameters was identified as an independent risk factor for femoral head necrosis in patients with triangle screw fixation. The ratio between screw distance and femoral neck width, rather than the absolute distance, was selected to represent the screw spread grade, as this approach eliminates the confounding effects of individual variations in femoral size. To further elucidate the biomechanical mechanisms underlying this phenomenon, we constructed inverse triangle screw fixation models of femoral neck fractures. Consistent with our clinical observations, the ratio parameter was utilized for model construction, providing a more objective reference for numerical simulations [17, 35]. Biomechanical simulations were performed on these models, and displacement and stress values were computed to assess the potential risk of femoral head necrosis. The computational results revealed compromised instantaneous postoperative fixation stability and significant stress concentration on the femoral head in models with poor screw spread [29, 36]. Conversely, increasing the screw spread grade effectively optimized these outcomes by enhancing fixation stability and improving local stress distribution. Through a comprehensive combination of clinical and biomechanical research, this study demonstrates that increasing the screw spread grade can significantly reduce the risk of femoral head necrosis by optimizing fixation stability and local biomechanical conditions. Therefore, we recommend adopting a wider screw spread in inverse triangle screw fixation procedures to minimize the risk of femoral head necrosis. Although the Pauwels angle values was significantly larger in the femoral head necrosis patient series, this parameter did not emerge as an independent risk factor when the screw distance-to-femoral neck width ratio was considered. Based on clinical observations, we hypothesized that optimizing the screw configuration strategy could effectively reduce the risk of femoral head necrosis, even in patients with high Pauwels angle values. To test this hypothesis, we selected a Pauwels angle of 55°, rather than 50°, for constructing the femoral neck fracture model. While a 50° Pauwels angle is typically the threshold for assessing fracture stability, and a higher angle was observed in necrosis patients, the average Pauwels angle in patients without femoral head necrosis still exceeded 50°. Therefore, we chose a higher angle (55.12° ± 9.45, Table 2) to create a more unstable model that better reflects the conditions seen in necrosis patients. We believe that the biomechanical differences in fixation stability due to changes in screw spread grade can be more accurately identified using this model construction strategy [37, 38]. Consistent with our clinical observations, the fixation stability and local biomechanical environment improved significantly with an increasing ratio between screw distance and femoral neck width. This research protocol offers a key advantage: it combines clinical and biomechanical studies, providing a direct link between clinical outcomes and numerical simulations [17, 18]. By integrating these approaches, the mechanism behind clinically observed phenomena can be directly explained through computational results. This design enhances the credibility of the study and offers more reliable guidance for clinical practice, as well as a more rational explanation for the observed clinical outcomes [16, 17]. However, patients with Garden III and IV fractures face a significantly higher risk of femoral head necrosis, independent of the screw distance-to-femoral neck width ratio. Thus, the effect of optimizing fixation stability on reducing necrosis risk may be limited in these cases [5, 39]. One major limitation of our study is the inability to numerically simulate the effect of fracture fragment migration on blood supply, which is critical for assessing necrosis risk. Additionally, to simplify the numerical simulation process, we applied only a single directional load to the femur, ignoring the complex interactions of muscle forces [27, 40]. This limitation should be addressed in future studies by validating the numerical model against mechanical tests. Nonetheless, the loading protocol used in this study is widely accepted in similar studies and is unlikely to compromise the credibility of our conclusions. Furthermore, the screw distance-to-femoral width ratio was measured using immediate postoperative lateral radiography [38, 41]. Due to the absence of postoperative thin-slice CT imaging data, we could not assess the cortical penetration status in the current patient series. Future studies with more comprehensive imaging data are necessary to validate this aspect. Conclusions Compared to the trajectory of limit screws, a larger ratio of screw distance to femoral width in the sagittal plane can effectively optimize the instantaneous postoperative stress concentration on the femoral head. This optimization may significantly reduce the potential risk of femoral head necrosis in patients with femoral neck fractures. Acknowledgements We acknowledging Dr. Zhipeng Xi, for the guidance of figures preparation. Author contributions Conception and design: Hong Li, Shengyu Wan, and Jingchi Li; Acquisition of data: Jian Zhang, Xiaozhong Luo, and Chao Wu; Analysis and interpretation of imaging data: Hong Li, Shengyu Wan, Fan Wu, and Lin Xu; Statistical analysis: Shengyu Wan, and Jingchi Li; Manuscript Preparation: Hong Li, Shengyu Wan, and Jingchi Li; Manuscript revision and modification: Xu Lin, Jingchi Li. Availability of data and materials All the data of the manuscript are presented in the paper. Declarations Ethics approval and consent to participate Approval for this study was obtained from the Ethics committee of the Zigong Fourth People’s Hospital (2023–009). Competing interest The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Hong Li and Shengyu Wan contributed equally and should be considered co-first authors of this study. Contributor Information Xu Lin, Email: 596920149@qq.com. Jingchi Li, Email: Lijingchi9405@163.com. References Associated Data Data Availability Statement All the data of the manuscript are presented in the paper.