Compliance with established guidelines for the radiological reporting of atypical femoral fractures

AFFnet - a deep convolutional neural network for the detection of atypical femur fractures from anteriorposterior radiographs

Despite well-defined criteria for radiographic diagnosis of atypical femur fractures (AFFs), missed and delayed diagnosis is common. An AFF diagnostic software could provide timely AFF detection to prevent progression of incomplete or development of contralateral AFFs. In this study, we investigated the ability for an artificial intelligence (AI)-based application, using deep learning models (DLMs), particularly convolutional neural networks (CNNs), to detect AFFs from femoral radiographs. A labelled Australian dataset of pre-operative complete AFF (cAFF), incomplete AFF (iAFF), typical femoral shaft fracture (TFF), and non-fractured femoral (NFF) X-ray images in anterior-posterior view were used for training (N = 213, 49, 394, 1359, respectively). An AFFnet model was developed using a pretrained (ImageNet dataset) ResNet-50 backbone, and a novel Box Attention Guide (BAG) module to guide the model's scanning patterns to enhance its learning. All images were used to train and internally test the model using a 5-fold cross validation approach, and further validated by an external dataset. External validation of the model's performance was conducted on a Sweden dataset comprising 733 TFF and 290 AFF images. Precision, sensitivity, specificity, F1-score and AUC were measured and compared between AFFnet and a global approach with ResNet-50. Excellent diagnostic performance was recorded in both models (all AUC >0.97), however AFFnet recorded lower number of prediction errors, and improved sensitivity, F1-score and precision compared to ResNet-50 in both internal and external testing. Sensitivity in the detection of iAFF was higher for AFFnet than ResNet-50 (82 % vs 56 %). In conclusion, AFFnet achieved excellent diagnostic performance on internal and external validation, which was superior to a pre-existing model. Accurate AI-based AFF diagnostic software has the potential to improve AFF diagnosis, reduce radiologist error, and allow urgent intervention, thus improving patient outcomes.

Fusion of electronic health records and radiographic images for a multimodal deep learning prediction model of atypical femur fractures

Atypical femur fractures (AFF) represent a very rare type of fracture that can be difficult to discriminate radiologically from normal femur fractures (NFF). AFFs are associated with drugs that are administered to prevent osteoporosis-related fragility fractures, which are highly prevalent in the elderly population. Given that these fractures are rare and the radiologic changes are subtle currently only 7% of AFFs are correctly identified, which hinders adequate treatment for most patients with AFF. Deep learning models could be trained to classify automatically a fracture as AFF or NFF, thereby assisting radiologists in detecting these rare fractures. Historically, for this classification task, only imaging data have been used, using convolutional neural networks (CNN) or vision transformers applied to radiographs. However, to mimic situations in which all available data are used to arrive at a diagnosis, we adopted an approach of deep learning that is based on the integration of image data and tabular data (from electronic health records) for 159 patients with AFF and 914 patients with NFF. We hypothesized that the combinatorial data, compiled from all the radiology departments of 72 hospitals in Sweden and the Swedish National Patient Register, would improve classification accuracy, as compared to using only one modality. At the patient level, the area under the ROC curve (AUC) increased from 0.966 to 0.987 when using the integrated set of imaging data and seven pre-selected variables, as compared to only using imaging data. More importantly, the sensitivity increased from 0.796 to 0.903. We found a greater impact of data fusion when only a randomly selected subset of available images was used to make the image and tabular data more balanced for each patient. The AUC then increased from 0.949 to 0.984, and the sensitivity increased from 0.727 to 0.849. These AUC improvements are not large, mainly because of the already excellent performance of the CNN (AUC of 0.966) when only images are used. However, the improvement is clinically highly relevant considering the importance of accuracy in medical diagnostics. We expect an even greater effect when imaging data from a clinical workflow, comprising a more diverse set of diagnostic images, are used.

Case Report - Short-term Bisphosphonate Use Associated Stress Fractures

Introduction

Long-term bisphosphonate use has been linked to an increased risk of pathological neck of femur fractures.

Case Report

We write concerning a patient who presented with the left hip pain following a low impact fall, which was diagnosed as a pathological left neck of femur fracture. This was a subtrochanteric stress fracture most frequently seen in patients who take bisphosphonate medications. A key point of difference in our patient is the length of time of bisphosphonate use. A further interesting point was the method of imaging used to diagnose this fracture whereby plain radiographs and computerized tomography imaging both did not show any acute fracture whereas only a magnetic resonance imaging (MRI) hip demonstrated this fracture. Surgical insertion of a prophylactic intramedullary nail was done to stabilize the fracture and reduce the risk of progression to a complete fracture.

Conclusion

This case brings up multiple key points not reviewed previously such as the fact a fracture developed only 1 month after bisphosphonate use rather than months or years. These points suggest that there should be a low threshold for investigation (including MRI scanning) into potential pathological fractures and that bisphosphonate use should be a red flag to initiate these investigations regardless of length of use.

Surveillance of atypical femoral fractures in a nationwide fracture register

Background and purpose - To continuously assess the incidence of atypical femoral fractures (AFFs) in the population is important, to allow the evaluation of the risks and benefits associated with osteoporosis treatment. Therefore, we investigated the possibility to use the Swedish Fracture Register (SFR) as a surveillance tool for AFFs in the population and to explore means of improvement. Patients and methods - All AFF registrations in the SFR from January 1, 2015 to December 31, 2018 were enrolled in the study. For these patients, radiographs were obtained and combined with radiographs from 176 patients with normal femoral fractures, to form the study cohort. All images were reviewed and classified into AFFs or normal femur fractures by 2 experts in the field (gold-standard classification) and 1 orthopedic resident educated on the specific radiographic features of AFF (educated-user classification). Furthermore, we estimated the incidence rate of AFFs in the population captured by the register through comparison with a previous cohort and calculated the positive predictive value (PPV) and, where possible, the inter-observer agreement (Cohen's kappa) between the different classifications. Results - Of the 178 available patients with AFF in the SFR, 104 patients were classified as AFF using the goldstandard classification, and 89 using the educated-user classification. The PPV increased from 0.58 in the SFR classification to 0.93 in the educated-user classification. The interobserver agreement between the gold-standard classification and the educated-user classification was 0.81. Interpretation - With a positive predictive value of 0.58 the Swedish Fracture Register outperforms radiology reports and reports to the Swedish Medical Products Agency on adverse drug reactions as a diagnostic tool to identify atypical femoral fractures.

Mid-term outcomes after the surgical treatment of atypical femoral fractures : minimum three-year follow-up

AIMS

The incidence of atypical femoral fractures (AFFs) continues to increase. However, there are currently few long-term studies on the complications of AFFs and factors affecting them. Therefore, we attempted to investigate the outcomes, complications, and risk factors for complication through mid-term follow-up of more than three years.

METHODS

From January 2003 to January 2016, 305 patients who underwent surgery for AFFs at six hospitals were enrolled. After exclusion, a total of 147 patients were included with a mean age of 71.6 years (48 to 89) and 146 of whom were female. We retrospectively evaluated medical records, and reviewed radiographs to investigate the fracture site, femur bowing angle, presence of delayed union or nonunion, contralateral AFFs, and peri-implant fracture. A statistical analysis was performed to identify the significance of associated factors.

RESULTS

The mean follow-up period was 70.2 months (36 to 191). There were 146 AFFs (99.3%) in female patients and the mean age was 71.6 years (48 to 89). The AFFs were located in the subtrochanter and shaft in 52 cases (35.4%) and 95 (64.6%), respectively. The preoperative mean anterior/lateral femoral bowing angles were 10.5° (SD 5.7°)/6.1° (SD 6.2°). The postoperative mean anterior/lateral bowing values were changed by 8.7° (SD 5.4°)/4.6° (SD 5.9°). Bisphosphonates had been used contemporarily in 115 AFFs (78.2%) for a mean of 52.4 months (1 to 204; SD 45.5) preoperatively. Nailing was performed in 133 AFFs (90.5%), and union was obtained at a mean of 23.6 weeks (7 to 85). Delayed union occurred in 41 (27.9%), and nonunion occurred in 13 (8.8%). Contralateral AFF occurred in 79 patients (53.7%), and the use of a bisphosphonate significantly influenced the occurrence of contralateral AFFs (p = 0.019). Peri-implant fractures occurred in a total of 13 patients (8.8%), and a significant increase was observed in cases with plating (p = 0.021) and high grade of postoperative anterolateral bowing (p = 0.044).

CONCLUSION

The use of a bisphosphonate was found to be a risk factor for contralateral AFF, and high-grade postoperative anterolateral bowing and plate fixation significantly increased the occurrence of peri-implant fractures. Long-term follow-up studies on the bilaterality of AFFs and peri-implant fractures are warranted. Cite this article: Bone Joint J 2021;103-B(11):1648-1655.

Deep neural networks with promising diagnostic accuracy for the classification of atypical femoral fractures

Background and purpose - A correct diagnosis is essential for the appropriate treatment of patients with atypical femoral fractures (AFFs). The diagnostic accuracy of radiographs with standard radiology reports is very poor. We derived a diagnostic algorithm that uses deep neural networks to enable clinicians to discriminate AFFs from normal femur fractures (NFFs) on conventional radiographs.Patients and methods - We entered 433 radiographs from 149 patients with complete AFF and 549 radiographs from 224 patients with NFF into a convolutional neural network (CNN) that acts as a core classifier in an automated pathway and a manual intervention pathway (manual improvement of image orientation). We tested several deep neural network structures (i.e., VGG19, InceptionV3, and ResNet) to identify the network with the highest diagnostic accuracy for distinguishing AFF from NFF. We applied a transfer learning technique and used 5-fold cross-validation and class activation mapping to evaluate the diagnostic accuracy.Results - In the automated pathway, ResNet50 had the highest diagnostic accuracy, with a mean of 91% (SD 1.3), as compared with 83% (SD 1.6) for VGG19, and 89% (SD 2.5) for InceptionV3. The corresponding accuracy levels for the intervention pathway were 94% (SD 2.0), 92% (2.7), and 93% (3.7), respectively. With regards to sensitivity and specificity, ResNet outperformed the other networks with a mean AUC (area under the curve) value of 0.94 (SD 0.01) and surpassed the accuracy of clinical diagnostics.Interpretation - Artificial intelligence systems show excellent diagnostic accuracies for the rare fracture type of AFF in an experimental setting.

Radiation-induced insufficiency fracture of the femur 18 years after radiation therapy

Advances in oncologic treatment have improved survival rates, allowing late effects of radiotherapy to become more prevalent. Our patient, an 82-year-old woman with a remote history of right thigh basal cell carcinoma treated with resection and radiation therapy 18 years prior, presented with severe right thigh pain and inability to bear weight as she had suffered a femur fracture after a fall from standing. Initial imaging was suspicious for pathologic fracture secondary to malignancy due to imaging findings and because radiation-induced fractures have rarely been reported beyond 44 months from treatment. However, upon further imaging, evidence pointed to radiation-induced osteonecrosis as the mechanism for her insufficiency fracture. This case highlights the permanent deleterious effects of radiation therapy on bone, and the prudence of considering radiation-induced osteonecrosis as a mechanism of injury in low-energy trauma even long after radiation therapy. In addition, the case serves to review the natural history of irradiated bone injury and pertinent imaging findings.

Current Understanding of Epidemiology, Pathophysiology, and Management of Atypical Femur Fractures

PURPOSE OF REVIEW

To summarize reports published since the 2013 American Society of Bone and Mineral Research Task Force Report on atypical femoral fractures (AFF).

RECENT FINDINGS

The absolute incidence of AFFs remains low. AFFs are primarily associated with prolonged bisphosphonate (BP) exposure, but have also been reported in unexposed patients and those receiving denosumab for osteoporosis and metastatic bone disease. Asians may be more susceptible to AFFs. Lateral femoral bowing and varus hip geometry, which increase loading forces on the lateral femoral cortex, may increase AFF risk. Altered bone material properties associated with BP therapy may predispose to AFFs by permitting initiation and increasing propagation of micro-cracks. Relevant genetic mutations have been reported in patients with AFFs. Single X-ray absorptiometry femur scans permit early detection of incomplete and/or asymptomatic AFFs. Orthopedists recommend intramedullary rods for complete AFFs and for incomplete, radiologically advanced AFFs associated with pain and/or marrow edema on MRI. Teriparatide may advance AFF healing but few data support its efficacy. Greater understanding of biological and genetic predisposition to AFF may allow characterization of individual risk prior to initiating osteoporosis therapy and help allay fear in those at low risk for this complication, which remains rare in comparison to the osteoporotic fractures prevented by antiresorptive therapy.

Atypical femoral fractures in Italy: a retrospective analysis in a large urban emergency department during a 7-year period (2007-2013)

The aim of this study was to determine the incidence of atypical femoral fractures (AFFs) seen in a large emergency department in Italy. It was a retrospective study of all men and women aged 40 years or older admitted to the Emergency Department of Parma University Hospital for a femoral fracture. Cases were identified in the hospital database with use of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 820 or 821 or text strings. All the radiographic images of fractures not clearly identified as proximal or condylar were retrieved and evaluated by three independent reviewers. Fractures were considered as atypical if all three reviewers agreed on at least four of five major features defined by the 2013 American Society for Bone and Mineral Research criteria. In the 7-year period (2007-2013), with a total follow-up of 1,383,154 patient-years, we found 22 AFFs in 21 patients, accounting for 7.1% of low-trauma subtrochanteric/femoral shaft fractures and 0.6% of all femoral fractures. The incidence was very low (1.6 in 100,000 patient-years in both sexes combined). In contrast, the incidence of classic fractures of the proximal end of the femur was at least two orders of magnitude higher (typical/atypical rate ratio 152). Bisphosphonate use was reported in 13 patients (62%; mean treatment duration 9 years; range 5-14 years). Among 286 patients with typical subtrochanteric/femoral shaft fractures, 20 were being treated with bisphosphonate (7%; odds ratio 22; 95% confidence interval 8-58; p < 0.001). This study confirms the very low incidence of AFFs in the largest Italian cohort of patients to date. Even though the risk is higher in patients treated with bisphosphonates, AFFs are very rare, and typical femoral fractures are at least 100-fold more frequent.

Bone Structure and Turnover Status in Postmenopausal Women with Atypical Femur Fracture After Prolonged Bisphosphonate Therapy

Atypical femur fracture (AFF), a serious complication of long-term bisphosphonate therapy, is usually preceded by an incomplete fracture appearing on the lateral femur. AFF is most likely the result of severely suppressed bone turnover (SSBT). However, the differences in bone structure and turnover between patients with incomplete and complete AFF remain unknown. We examined trans-iliac bone biopsies from 12 white postmenopausal women with AFF (incomplete = 5; complete = 7) on BP therapy of >5 years and 43 healthy white premenopausal women. Histomorphometric measurements were performed separately in cancellous, intracortical and endosteal envelopes. Of the 43 histomorphometric measurements on 3 difference bone surfaces (cancellous, intracortical and endosteal), only 2 bone resorption variables (Oc.S/BS and Oc.S/NOS) on the endosteal surface were significantly lower in patients with complete AFF than those with incomplete AFF. Compared to healthy premenopausal women, the trabecular bone volume, thickness and number were all significantly lower in patients with AFF. The dynamic bone formation variables in patients with AFF were significantly reduced on all bone surfaces. The likelihood of a biopsy with no tetracycline labeling was significantly higher in AFF patients than in healthy premenopausal women. Based on these results, we conclude that there are no significant differences in bone turnover between patients with incomplete and complete AFF, suggesting that the suppression of bone turnover had already existed in the femur with incomplete AFF. Compared to healthy premenopausal women, bone turnover is similarly suppressed in patients with either type of AFF.

Identifying Incomplete Atypical Femoral Fractures With Single-Energy Absorptiometry: Declining Prevalence

Background:

Atypical femur fractures (AFFs) are associated with long-term bisphosphonate (BP) therapy. Early identification of AFF prior to their completion provides an opportunity to intervene, potentially reducing morbidity associated with these fractures. Single-energy X-ray absorptiometry (SE) is an imaging method recently shown to detect incomplete AFF (iAFF) prior to fracture completion.

Methods:

Between May 2013 and September 2014, we assessed 173 patients who had been prescribed BP therapy for >5 years for iAFF using SE at their presentation for routine bone mineral density testing. We compared these findings with those of our previously published prospective study (n = 257) in which the femur was imaged for iAFF using dual-energy X-ray absorptiometry. In addition, we estimated the yearly prevalence of complete AFF among patients with subtrochanteric fracture at our institution from 2006 to 2014, and we evaluated prescribing trends for BP in Ireland from 2009 to 2014.

Results:

No patients had iAFF using SE femur compared with a prevalence of 2.7% in the earlier study. Between 2006 and 2014, we observed a rise and decline in AFFs at our hospital and a similar national trend in BP prescribing.

Conclusions:

AFFs appear to be decreasing. New customized scan modes of dual-energy X-ray absorptiometry systems, which visualize the entire femur at high image quality and take measurements, have the potential to identify iAFF prior to fracture completion and to ascertain those at highest risk of AFF.

Bilateral atypical femur fractures without bisphosphonate exposure

Atypical femur fractures have common radiographic features that set them apart from more typical higher-energy subtrochanteric femur fractures. They are noncomminuted, transverse fractures with medial spiking of the femoral cortex and increased lateral cortical thickness. These fractures have been associated in the literature with the use of bisphosphonate medications. This case describes bilateral atypical femur fractures in a patient with a medical history devoid of bisphosphonate use. We present his history, co-morbidities, and subsequent treatment. From this case, we call attention to bisphosphonate use as not the only cause of subtrochanteric femur fractures with atypical features and highlight that some patients may sustain these injuries even bilaterally without use of the medications. In addition, it is important to identify this fracture type and obtain imaging of the contralateral femur to facilitate prophylactic treatment if needed.