This study was approved by our institutional review board. It was carried out in accordance with the Declaration of Helsinki. Written informed consent was obtained from all patients prior to the study. Magnetic resonance imaging was performed on 828 hips in 588 patients with ONFH between January 2005 and February 2016. Among them, we retrospectively reviewed 72 consecutive hips in 55 patients who underwent CE magnetic resonance imaging because it was difficult to evaluate the limit of necrosis. BME injury after subchondral collapse (ARC stage III or higher, Fig. 1) or difficulty differentiating ONFH from subchondral insufficiency femoral head fracture or transient hip osteoporosis by non-contrast MRI19.
Thirty-two hips were treated with transtrochanteric anterior rotational osteotomy, 2 hips with transtrochanteric posterior rotational osteotomy, 10 hips with transtrochanteric curved varus osteotomy, and 26 hips with prosthetic replacement. Two hips without surgical treatment were continuously observed. Three of 44 (6.8%) cases treated with joint-sparing procedures underwent prosthetic replacement due to osteoarthritic change after osteotomy (2, 4, and 4 years, respectively). One case (2.2%) underwent prosthetic replacement due to nonunion at the osteotomy site due to deep infection after surgery. The survival rate was 91.0% (end point: prosthetic replacement), with a mean of 9.1 years after the osteotomies (range, 5 to 15 years). The ONFH stage (2019 revised ARCO classification) and the location of the affected lesion (Japanese Investigation Committee (JIC) classification) are shown in Fig. 1.16.20. Demarcating sclerosis was present in 51 hips (70.8%) on anteroposterior (AP) or lateral radiographtwenty-one.
MR images were obtained with a 1.5T MR unit (Achieva 1.5T; Philips Healthcare, Best, The Netherlands) in 52 hips and a 3.0T MR unit (Achieva 3.0T ; Philips Healthcare, Best, The Netherlands) in 20 hips. After obtaining fat-saturated (FS) T1-weighted noncontrast and T2-weighted spin echo images (coronal only) (repeat times/echo times[TR/TE]= 400–540/10–18ms, FST2; 3000–4709/80–100), CET1-weighted images with fat saturation (TR/TE = 620–700/10–18) were obtained by administering 0.2 mL/kg gadopentetate dimeglumine (Magnevist; Bayer Pharma , Berlin Germany) . All sequences had a slice thickness of 5 mm with an interslice spacing of 1 mm and a field of view of 360 × 360 mm. The duration of the MR examination was 30-40 min. T1-, FST2-, and CET1-weighted images were available in 68 of 72 hips (94.4%) in the coronal plane, and noncontrast and CET1-weighted images were available in 67 of 72 hips (93.1%) in the coronal plane. the oblique-axial plane. plane (parallel to the axis of the femoral neck) (Fig. 2). Based on medical records, the median duration from the time of hip pain onset to MR imaging was 4.7 months (range, 0.5 to 19 months).
In the coronal plane, the boundaries of the necrotic lesion were assessed by mediolateral necrotic angles using T1- and FST2-weighted MR images in the mid-coronal slice, and compared with those of the CET1-weighted MR images (Fig. 3). . In the oblique axial plane, the boundaries of the necrotic lesion were assessed by AP necrotic angles using T1-weighted MR images in an oblique-axial mid-section and compared with those of the CET1-weighted MR images (Fig. 3). . When the necrotic angles between T1- or FST2-weighted images and CET1 were different by more than 10°, the case was defined as difference group (Fig. 3). Degrees of BME extension (BME grade) were classified as grade I (within the femoral head), grade II (beyond the femoral head but within the femoral neck), and grade III (beyond the femoral neck) using all slices of T1, FST2 and CET1 in both axial, coronal and oblique planes (Fig. 4). The assessments were performed by two observers (SI and TU), who are orthopedic surgeons and have extensive experience in diagnostic imaging. To assess the intra- and interobserver reproducibility of the borderline assessments between the comparison of necrotic and live bone between non-contrast images and CET1, and the BME grade assessments, the reliability of the measurements was assessed using kappa statistics. A kappa value of 0.21 to 0.4 indicates fair agreement; 0.41–0.6, moderate agreement; and 0.61-0.8, substantial agreement. A value of > 0.81 is considered almost perfect22.
After comparison of the boundaries of necrotic lesions between the CE and non-contrast MR images, all hips were divided based on the differences between the two images. Statistical analyzes were performed using the chi-square test or Fisher’s exact probability to compare sex, history of steroid or alcohol use, stage, type, and presence of radiological bounding sclerosis between the two groups. Age, BMI, and duration from baseline to MR imaging examination were compared between the two groups using the you-Test. The degree of BME between the two groups and the relationship between the duration from baseline to MR imaging examination and the degree of BME were analyzed using the Mann-Whitney your Wilcoxon’s test and signed rank test, respectively. Multivariate analysis was performed to identify parameters associated with differences in necrotic lesion boundaries between CE and noncontrast MR images using stepwise logistic regression with variable selection (P <0.2). Simple logistic regression analysis was performed to calculate the relationship between the duration from baseline to MR imaging examination and the rate of differences in the borders of necrotic lesions between T1-weighted and CET1-weighted MR images without contrast. Statistical analyzes were performed using JMP Ver. Software 9.0.1 (SAS Institute Inc., Cary, NC, USA). Pvalues <0.05 were considered statistically significant.
Ethical approval and consent to participate
This retrospective study was approved by the Kyushu University Institutional Review Board for Clinical Research (NO. 2019-584). Written informed consent was obtained from all patients prior to the study.