Computed tomography-based automated 3D measurement of femoral version: Validation against standard 2D measurements in symptomatic patients

To validate 3D methods for femoral version measurement, we asked: (1) Can a fully automated segmentation of the entire femur and 3D measurement of femoral version using a neck based method and a head-shaft based method be performed? (2) How do automatic 3D-based computed tomography (CT) measurements of femoral version compare to the most commonly used 2D-based measurements utilizing four different landmarks? Retrospective study (May 2017 to June 2018) evaluating 45 symptomatic patients (57 hips, mean age 18.7 ± 5.1 years) undergoing pelvic and femoral CT. Femoral version was assessed using four previously described methods (Lee, Reikeras, Tomczak, and Murphy). Fully-automated segmentation yielded 3D femur models used to measure femoral version via femoral neck- and head-shaft approaches. Mean femoral version with 95% confidence intervals, and intraclass correlation coefficients were calculated, and Bland-Altman analysis was performed. Automatic 3D segmentation was highly accurate, with mean dice coefficients of 0.98 ± 0.03 and 0.97 ± 0.02 for femur/pelvis, respectively. Mean difference between 3D head-shaft- (27.4 ± 16.6°) and 3D neck methods (12.9 ± 13.7°) was 14.5 ± 10.7° (p < 0.001). The 3D neck method was closer to the proximal Lee (-2.4 ± 5.9°, -4.4 to 0.5°, p = 0.009) and Reikeras (2 ± 5.6°, 95% CI: 0.2 to 3.8°, p = 0.03) methods. The 3D head-shaft method was closer to the distal Tomczak (-1.3 ± 7.5°, 95% CI: -3.8 to 1.1°, p = 0.57) and Murphy (1.5 ± 5.4°, -0.3 to 3.3°, p = 0.12) methods. Automatic 3D neck-based-/head-shaft methods yielded femoral version angles comparable to the proximal/distal 2D-based methods, when applying fully-automated segmentations.

Hip Morphology on Post-Reduction MRI Predicts Residual Dysplasia 10 Years After Open or Closed Reduction

BACKGROUND

There is limited evidence supporting the value of morphological parameters on post-reduction magnetic resonance imaging (MRI) to predict long-term residual acetabular dysplasia (RAD) after closed or open reduction for the treatment of developmental dysplasia of the hip (DDH).

METHODS

We performed a retrospective study of 42 patients (47 hips) undergoing open or closed reduction with a minimum 10 years of follow-up; 39 (83%) of the hips were in female patients, and the median age at reduction was 6.3 months (interquartile range [IQR], 3.3 to 8.9 months). RAD was defined as additional surgery with an acetabular index >2 standard deviations above the age- and sex-specific population-based mean value or Severin classification grade of >2 at last follow-up. Acetabular version and depth-width ratio, coronal and axial femoroacetabular distance, cartilaginous and osseous acetabular indices, transverse ligament thickness, and the thickness of the medial and lateral (limbus) acetabular cartilage were measured on post-reduction MRI.

RESULTS

At the time of final follow-up, 24 (51%) of the hips had no RAD; 23 (49%) reached a failure end point at a median of 11.4 years (IQR, 7.6 to 15.4 years). Most post-reduction MRI measurements, with the exception of the cartilaginous acetabular index, revealed a significant distinction between the group with RAD and the group with no RAD when mean values were compared. The coronal femoroacetabular distance (area under the receiver operating characteristic curve [AUC], 0.95; 95% confidence interval [CI], 0.90 to 1.00), with a 5-mm cutoff, and limbus thickness (AUC, 0.91; 95% CI, 0.83 to 0.99), with a 4-mm cutoff, had the highest discriminatory ability. A 5-mm cutoff for the coronal femoroacetabular distance produced 96% sensitivity (95% CI, 78% to 100%), 83% specificity (95% CI, 63% to 95%), 85% positive predictive value (95% CI, 65% to 96%), and 95% negative predictive value (95% CI, 76% to 100%). A 4-mm cutoff for limbus thickness had 96% sensitivity (95% CI, 78% to 100%), 63% specificity (95% CI, 41% to 81%), 71% positive predictive value (95% CI, 52% to 86%), and 94% negative predictive value (95% CI, 70% to 100%).

CONCLUSIONS

Coronal femoroacetabular distance, a quantitative metric assessing a reduction's concentricity, and limbus thickness, a quantitative metric assessing the acetabulum's cartilaginous component, help to predict hips that will have RAD in the long term after closed or open reduction.

LEVEL OF EVIDENCE

Diagnostic Level IV . See Instructions for Authors for a complete description of levels of evidence.

Osteoporosis: Molecular Pathology, Diagnostics, and Therapeutics

Osteoporosis is a major public health concern affecting millions of people worldwide and resulting in significant economic costs. The condition is characterized by changes in bone homeostasis, which lead to reduced bone mass, impaired bone quality, and an increased risk of fractures. The pathophysiology of osteoporosis is complex and multifactorial, involving imbalances in hormones, cytokines, and growth factors. Understanding the cellular and molecular mechanisms underlying osteoporosis is essential for appropriate diagnosis and management of the condition. This paper provides a comprehensive review of the normal cellular and molecular mechanisms of bone homeostasis, followed by an in-depth discussion of the proposed pathophysiology of osteoporosis through the osteoimmunological, gut microbiome, and cellular senescence models. Furthermore, the diagnostic tools used to assess osteoporosis, including bone mineral density measurements, biochemical markers of bone turnover, and diagnostic imaging modalities, are also discussed. Finally, both the current pharmacological and non-pharmacological treatment algorithms and management options for osteoporosis, including an exploration of the management of osteoporotic fragility fractures, are highlighted. This review reveals the need for further research to fully elucidate the molecular mechanisms underlying the condition and to develop more effective therapeutic strategies.

Total Knee Arthroplasty In Patients With Juvenile Idiopathic Arthritis

Medical and surgical management of knee pain in juvenile idiopathic arthritis (JIA) is rapidly evolving. It is important for the orthopaedic surgeon to remain informed. In this review, we discuss the recent trends in surgical management of JIA in light of recent medical advances for the disease. The purpose of this article is to summarize current recommendations for TKA in patients with JIA.

MRI hip morphology is abnormal in unilateral DDH and increased lateral limbus thickness is associated with residual DDH at minimum 10-year follow-up

Purpose:

The purpose of the study was to compare the post-reduction magnetic resonance imaging morphology for hips that developed residual acetabular dysplasia, hips without residual dysplasia, and uninvolved contralateral hips in patients with unilateral developmental dysplasia of the hip undergoing closed or open reduction and had a minimum 10-year follow-up.

Methods:

Retrospective study of patients with unilateral dysplasia of the hip who underwent open/closed hip reduction followed by post-reduction magnetic resonance imaging. Twenty-eight patients with a mean follow-up of 13 ± 3 years were included. In the treated hips, residual dysplasia was defined as subsequent surgery for residual acetabular dysplasia or for Severin grade > 2 at latest follow-up. On post-reduction, magnetic resonance imaging measurements were performed by two readers and compared between the hips with/without residual dysplasia and the contralateral uninvolved side. Magnetic resonance imaging measurements included acetabular version, coronal/ axial femoroacetabular distance, acetabular depth–width ratio, osseous/cartilaginous acetabular indices, and medial/lateral (limbus) cartilage thickness.

Results:

Fifteen (54%) and 13 (46%) hips were allocated to the “no residual dysplasia” group and to the “residual dysplasia” group, respectively. All eight magnetic resonance imaging parameters differed between hips with residual dysplasia and contralateral uninvolved hips (all p < 0.05). Six of eight parameters differed (all p < 0.05) between hips with and without residual dysplasia. Among these, increased limbus thickness had the largest effect (odds ratio = 12.5; p < 0.001) for increased likelihood of residual dysplasia.

Conclusions:

We identified acetabular morphology and reduction quality parameters that can be reliably measured on the post-reduction magnetic resonance imaging to facilitate the differentiation between hips that develop with/without residual acetabular dysplasia at 10 years postoperatively.

Level of evidence:

level III, prognostic case-control study.

Correlation of Hounsfield Unit Measurements on Computed Tomography of the Shoulder With Dual-Energy X-ray Absorptiometry Scans and Fracture Risk Assessment Tool Scores: A Potential for Opportunistic Screening

BACKGROUND

The diagnosis of bone mineral density (BMD) abnormalities involves dual-energy x-ray absorptiometry (DXA), but few patients complete this after a fragility fracture. The assessment of BMD using Hounsfield unit (HU) measurements from computed tomography (CT) scans has been correlated with DXA results in previous studies. We aimed to evaluate the correlation between shoulder CT HU and DXA scores.

METHODS

Billing databases of 3 academic institutions were queried for patients who underwent both DXA and CT scan of the upper extremity within 1 year of each other. DXA T-scores for spine, hip, and femoral neck were recorded. BMD status was defined based on composite T-scores, using the lowest T-score in the spine, hip, or proximal femur. CT scans were measured for HU over 4 slices, recorded to create a bone column, then averaged. The patients' risks of major osteoporosis-related fracture and hip fracture were calculated using the Fracture Risk Assessment Tool.

RESULTS

In total, 300 patients were included. A positive correlation was found between composite T-scores and HU for glenoid and proximal humerus (0.36; 0.17). The proximal humerus HU was significantly associated with the BMD T-score for the hip (P = 0.01); the glenoid HU was significantly associated with BMD T-scores for the hip, spine, and femoral neck (P = 0.002; P = 0.001; P = 0.002). A 10-year risk of hip fracture >3% was associated with lower proximal humerus HU.

CONCLUSIONS

Our study is the first to discover significant correlations between HU at the glenoid and proximal humerus and risk factors as established by the Fracture Risk Assessment Tool scoring system.

LEVEL OF EVIDENCE

Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.

How Common Is Femoral Retroversion and How Is it Affected by Different Measurement Methods in Unilateral Slipped Capital Femoral Epiphysis?

BACKGROUND

Although femoral retroversion has been linked to the onset of slipped capital femoral epiphysis (SCFE), and may result from a rotation of the femoral epiphysis around the epiphyseal tubercle leading to femoral retroversion, femoral version has rarely been described in patients with SCFE. Furthermore, the prevalence of actual femoral retroversion and the effect of different measurement methods on femoral version angles has yet to be studied in SCFE.

QUESTIONS/PURPOSES

(1) Do femoral version and the prevalence of femoral retroversion differ between hips with SCFE and the asymptomatic contralateral side? (2) How do the mean femoral version angles and the prevalence of femoral retroversion change depending on the measurement method used? (3) What is the interobserver reliability and intraobserver reproducibility of these measurement methods?

METHODS

For this retrospective, controlled, single-center study, we reviewed our institutional database for patients who were treated for unilateral SCFE and who had undergone a pelvic CT scan. During the period in question, the general indication for obtaining a CT scan was to define the surgical strategy based on the assessment of deformity severity in patients with newly diagnosed SCFE or with previous in situ fixation. After applying prespecified inclusion and exclusion criteria, we included 79 patients. The mean age was 15 ± 4 years, 48% (38 of 79) of the patients were male, and 56% (44 of 79) were obese (defined as a BMI > 95th percentile (mean BMI 34 ± 9 kg/m2). One radiology resident (6 years of experience) measured femoral version of the entire study group using five different methods. Femoral neck version was measured as the orientation of the femoral neck. Further measurement methods included the femoral head's center and differed regarding the level of landmarks for the proximal femoral reference axis. From proximal to distal, this included the most-proximal methods (Lee et al. and Reikerås et al.) and most-distal methods (Tomczak et al. and Murphy et al.). Most proximally (Lee et al. method), we used the most cephalic junction of the greater trochanter as the landmark and, most distally, we used the center base of the femoral neck superior to the lesser trochanter (Murphy et al.). The orientation of the distal femoral condyles served as the distal reference axis for all five measurement methods. All five methods were compared side-by-side (involved versus uninvolved hip), and comparisons among all five methods were performed using paired t-tests. The prevalence of femoral retroversion (< 0°) was compared using a chi-square test. A subset of patients was measured twice by the first observer and by a second orthopaedic resident (2 years of experience) to assess intraobserver reproducibility and interobserver reliability; for this assessment, we used intraclass correlation coefficients.

RESULTS

The mean femoral neck version was lower in hips with SCFE than in the contralateral side (-2° ± 13° versus 7° ± 11°; p < 0.001). This yielded a mean side-by side difference of -8° ± 11° (95% CI -11° to -6°; p < 0.001) and a higher prevalence of femoral retroversion in hips with SCFE (58% [95% CI 47% to 69%]; p < 0.001) than on the contralateral side (29% [95% CI 19% to 39%]). These differences between hips with SCFE and the contralateral side were higher and ranged from -17° ± 11° (95% CI -20° to -15°; p < 0.001) based on the method of Tomczak et al. to -22° ± 13° (95% CI -25° to -19°; p < 0.001) according to the method of Murphy et al. The mean overall femoral version angles increased for hips with SCFE using more-distal landmarks compared with more-proximal landmarks. The prevalence of femoral retroversion was higher in hips with SCFE for the proximal methods of Lee et al. and Reikerås et al. (91% [95% CI 85% to 97%] and 84% [95% CI 76% to 92%], respectively) than for the distal measurement methods of Tomczak et al. and Murphy et al. (47% [95% CI 36% to 58%] and 60% [95% CI 49% to 71%], respectively [all p < 0.001]). We detected mean differences ranging from -19° to 4° (all p < 0.005) for 8 of 10 pairwise comparisons in hips with SCFE. Among these, the greatest differences were between the most-proximal methods and the more-distal methods, with a mean difference of -19° ± 7° (95% CI -21° to -18°; p < 0.001), comparing the methods of Lee et al. and Tomczak et al. In hips with SCFE, we found excellent agreement (intraclass correlation coefficient [ICC] > 0.80) for intraobserver reproducibility (reader 1, ICC 0.93 to 0.96) and interobserver reliability (ICC 0.95 to 0.98) for all five measurement methods. Analogously, we found excellent agreement (ICC > 0.80) for intraobserver reproducibility (reader 1, range 0.91 to 0.96) and interobserver reliability (range 0.89 to 0.98) for all five measurement methods in healthy contralateral hips.

CONCLUSION

We showed that femoral neck version is asymmetrically decreased in unilateral SCFE, and that differences increase when including the femoral head's center. Thus, to assess the full extent of an SCFE deformity, femoral version measurements should consider the position of the displaced epiphysis. The prevalence of femoral retroversion was high in patients with SCFE and increased when using proximal anatomic landmarks. Since the range of femoral version angles was wide, femoral version cannot be predicted in a given hip and must be assessed individually. Based on these findings, we believe it is worthwhile to add evaluation of femoral version to the diagnostic workup of children with SCFE. Doing so may better inform surgeons as they contemplate when to use isolated offset correction or to perform an additional femoral osteotomy for SCFE correction based on the severity of the slip and the rotational deformity. To facilitate communication among physicians and for the design of future studies, we recommend consistently reporting the applied measurement technique.

LEVEL OF EVIDENCE

Level III, prognostic study.

The Effect of Modality and Landmark Selection on MRI and CT Femoral Torsion Angles

Background Assessment of femoral torsion at preoperative hip imaging is commonly recommended. However, it is unclear whether MRI is as accurate as CT and how different methods affect femoral torsion measurements. Purpose To compare MRI- and CT-based assessment of femoral torsion by using four commonly used measurement methods in terms of agreement, reproducibility, and reliability and to compare femoral torsion angles between the four different measurement methods. Materials and Methods This retrospective study evaluated patients with hip pain who underwent CT and 3-T MRI of the hip including sequences of the pelvis and distal condyles between May 2017 and June 2018. The four measurement methods differed regarding the landmark levels for the proximal femoral reference axis and included measurements at the level of the greater trochanter, femoral neck, base of the femoral neck, and level of the lesser trochanter. Intraclass correlation coefficients (ICCs) were calculated, and Bland-Altman analysis was performed. Results Forty-five patients (mean age ± standard deviation, 19 years ± 5; 27 female) and 57 hips were evaluated. Inter- and intrarater reliability were excellent for each of the four CT- and MRI-based measurement methods (ICC range, 0.97-0.99). Mean difference between CT- and MRI-based measurement ranged from 0.3° ± 3.4 (P = .58) to 2.1° ± 4.1 (P < .001). Differences between CT and MRI were within the corresponding ICC variation for all four measurement methods. Mean torsion angles were greater by 17.6° for CT and 18.7° for MRI (all P < .001) between the most proximal to the most distal measurement methods. Conclusion MRI- and CT-based femoral torsion measurements showed high agreement and comparable reliability and reproducibility but were dependent on the level of selected landmarks used to define the proximal reference axis. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Zoga in this issue.