Skeletal Radiol DOI 10.1007/s00256-017-2692-8

SCIENTIFIC ARTICLE

Evaluation of bone viability in patients after girdlestone arthroplasty: comparison of bone SPECT/CT and MRI G. Diederichs 1 & P. Hoppe 2 & F. Collettini 1 & G. Wassilew 3 & B. Hamm 1 & W. Brenner 2 & M. R. Makowski 1

Received: 15 March 2017 / Revised: 14 May 2017 / Accepted: 2 June 2017 # ISS 2017

Abstract Purpose To test the diagnostic performance of bone SPECT/ CT and MRI for the evaluation of bone viability in patients after girdlestone-arthroplasty with histopathology used as gold standard. Materials and methods In this cross-sectional study, patients after girdlestone-arthroplasty were imaged with single-photonemission-computed-tomography/computed-tomography (SPECT/CT) bone-scans using 99mTc-DPD. Additionally, 1.5 T MRI was performed with turbo-inversion-recoverymagnitude (TIRM), contrast-enhanced T1-fat sat (FS) and T1-mapping. All imaging was performed within 24 h prior to revision total-hip-arthroplasty in patients with a girdlestonearthroplasty. In each patient, four standardized bone-tissuebiopsies (14 patients) were taken intraoperatively at the remaining acetabulum superior/inferior and trochanter major/minor. Histopathological evaluation of bone samples regarding bone viability was used as gold standard. Results A total of 56 bone-segments were analysed and classified as vital (n = 39) or nonvital (n = 17) by histopathology. Mineral/late-phase SPECT/CT showed a high sensitivity (90%) and specificity (94%) to distinguish viable and nonviable bone tissue. TIRM (sensitivity 87%, specificity 88%) and contrast-enhanced T1-FS (sensitivity 90%, specificity 88%) also achieved a high sensitivity and specificity. T1-mapping achieved the lowest values (sensitivity 82%, specificity 82%).

* M. R. Makowski [email protected] 1

Department of Radiology, Charité, Charitéplatz 1, 10117 Berlin, Germany

2

Department of Nuclear Medicine, Charité, Berlin, Germany

3

Department of Orthopedic Surgery, Charité, Berlin, Germany

False positive results in SPECT/CT and MRI resulted from small bone fragments close to metal artefacts. Conclusions Both bone SPECT/CT and MRI allow a reliable differentiation between viable and nonviable bone tissue in patients after girdlestone arthroplasty. The findings of this study could also be relevant for the evaluation of bone viability in the context of avascular bone necrosis. Keywords MRI . SPECT/CT . Bone scan . Girdlestone arthroplasty . Bone viability

Abbreviations 99mTc-DPD 99mTc-3,3-diphosphono-1,2propanodicarboxylic-acid FS fat sat MRI magnetic resonance imaging PET positron emission tomography SPECT/CT single photon emission computed tomography/computed tomography TIRM turbo inversion recovery magnitude

Introduction In patients with advanced osteoarthritis, total hip arthroplasty is a common procedure. Worldwide more than 1 million arthroplasties are performed [1–4]. In most patients, hip arthroplasty leads to a restoration of function, a reduction in pain and an improved overall quality of life [5]; however, due to the large number of procedures, periprosthetic complications occur relatively frequently, if counted in absolute numbers. One of the most serious complication includes periprosthetic joint infection [2, 6]. Treatment options for this kind of infection depend on different factors, including the

Skeletal Radiol

patients’ physical condition and the type of bacteria involved in the inflammatory process. In recent decades, the girdlestone procedure became more and more popular as a salvage procedure following complications of total hip replacement surgery [7, 8]. If a substantial amount of nonviable bone of femur/acetabulum develops, these bony parts have to be removed surgically. Accordingly, the best fitting endoprosthesis has to be chosen for optimal stability and post-surgical outcome. Therefore, it is of clinical importance to determine whether and to what extent the remaining bone is viable and to clearly delineate the non-viable parts of bones, prior to the re-implantation of revision components. Different diagnostic modalities can be used to assess bone viability. These include planar bone scintigraphy in combination with single-photon emission computed tomography (SPECT) using 99mTC labelled bisphosphonates [9, 10]. A further alternative using radiotracers is Fluorine-18 positron emission tomography (PET). Additionally, MRI-based methods, such as fluid weighted MRI sequences (e.g. TIRM, turbo inversion recovery magnitude) or contrast-enhanced MRI (e.g. T1-FS) can be used. So far, SPECT/CT and contrast enhanced MRI have not been directly compared regarding the assessment of bone viability in patients after girdlestone arthroplasty. The aim of this study was to evaluate the diagnostic performance of bone SPECT/CT and MRI for the evaluation of bone viability in patients after girdlestone arthroplasty in comparison to histopathology as reference standard.

MRI imaging were performed. At the time of revision, including imaging and histopathology, all inflammatory laboratory parameters were within the normal range. Additionally, no signs of a clinical infection were found. Imaging protocol Combined bone SPECT/CT imaging

Material and methods

Planar bone and SPECT-CT imaging was performed using a single dedicated scanner type (Truepoint Symbia T6, Siemens Medical Solutions, Malvern, Pennsylvania, USA) in all patients combining the CT gantry and the dual-head gamma cameras in the same scanner. In this study, a three-phase scintigraphic examination of the pelvis/hip region was performed following the administration of 550–600 MBq of 99mTc-DPD (99mTc-3,3diphosphono-1,2-propanodicarboxylic acid). A dynamic first perfusion phase of the pelvis/hip region was acquired in anterior-posterior orientation using a 128 × 128 matrix with a 2-s frame rate for the first minute. Static blood-pool images were acquired in the following 3 min in the same camera position. Subsequently, planar whole-body and SPECT images of the hips were acquired in the delayed/mineralization phase 1.5– 2 h after the administration of the probe. Volumetric 3D acquisition of the pelvis/hip region was acquired using a doubleheaded gamma camera with a 360° rotation. SPECT scans of the pelvis/hip region were acquired using low-energy high-resolution collimators, a 128 × 128 matrix of 4.8-mm pixel size and a total of 120 projections over 360°. The dwell time was 15 s per angular view.

Study population

Contrast enhanced magnetic resonance imaging

This study was approved by and registered with the local ethics committee. Informed consent was obtained from all individual participants included in the study. Inclusion criteria for this study included patients with a girdlestone-arthroplasty prior to revision total-hip-arthroplasty. Exclusion criteria included age less than 18 years, not MR-compatible devices (e.g. pacemaker), claustrophobia, pregnancy, breast-feeding patients with mental disorders or unable to give consent, renal insufficiency (GFR < 30 ml/min/1.73 m2) measured less than 4 weeks prior to the MRI examination and known contrast agent allergy. The diagnosis was based on a combination of results from tissue culture, synovial fluid aspiration, tissue histology and serology markers [11]. Following the diagnosis of periprosthetic joint infection, all in vivo components of the prosthesis were removed, followed by a period without a prosthesis for the patient. During this period (6–8 weeks) patients were continuously treated with anti-inflammatory medication (antibiotics). Following the successful treatment of the infection, revision total hip arthroplasty was scheduled. Within 24 h prior to revision total hip arthroplasty SPECT/CT and

All MRI measurements were performed on a single Siemens 1.5 T scanner (Avanto, Siemens Medical Solutions, Erlangen, Germany) equipped with a standard body coil (Siemens Medical System). All MR imaging was performed in coronal orientation. Imaging parameters included: T1 TSE (Turbo Spin Echo): TR (repetition time) 560 ms, TE (echo time) 11 ms, flip angle 90°, number of slices 30, slice thickness 5 mm, FOV 400 mm, matrix 320 × 320, resolution 1.25 × 1.25 mm. Fluid-weighted TIRM (Turbo Inversion Recovery Magnitude): TR 5,000 ms, TE 61 ms, TI 160 ms, flip angle 150°, slice thickness and matrix same as T1 TSE. T1-FS VIBE: slice thickness 2.5 mm, TR 3.3 ms, TE 1.3 ms, FOV 350 × 350 mm, matrix 200 × 200, 1.75 × 1.75 mm. A looklocker inversion recovery (MOLLI) sequence was used for non-enhanced T1-mapping. Overall imaging time was 36 min. Imaging parameters included: slices 30, slice thickness 5 mm, TR 740 ms, TE 1.5 ms, field of view 350 × 350 mm, matrix 200 × 200, 1.75 × 1.75 mm. Following the native T1FS VIBE, 0.1 mmol/kg body weight of gadobutrol (Gadovist) was administered intravenously using a bolus injection. The

Skeletal Radiol

following MRI sequences were analysed: TIRM, T1-FS following the administration of contrast agent and T1 mapping. The T1-FS sequence was required for 4.5 min directly following the administration of contrast agent, and the T1 mapping sequence was started immediately following this sequence. Histopathology In all patients, biopsies from different bone tissues were taken intraoperatively by a the surgeon (GW) during revision total hip arthroplasty. An Ostycut bone biopsy needle (17G) was used to take systematic biopsies on the girdlestone site at the location of the remaining acetabulum superior, acetabulum inferior, trochanter major and trochanter minor. All samples were labelled accordingly and histological analysis was performed on standard haematoxylin–eosin stains. Vital tissue biopsies were defined as vital spongiose lamellar bone tissue with regular tribulation and/or with trilinear mature hemopoesis (vital bone marrow). Nonvital tissue biopsies were defined as bone sequesters including osteonecrosis and/or dystopic calcifications. Imaging analysis for the comparison of SPECT/CT and MRI To match region of interests (ROIs) between DPD SPECT/CT and MRI sequences, all SPECT/CT images were reformatted in a coronal orientation as all MRI sequences were acquired in a coronal orientation. To match the anatomical location of ROIs between DPD SPECT/CT and MRI sequences, data sets were co-registered based on the anatomical information from the os ilium, which was clearly visible in all sequences. All ROIs were placed during a consensus reading performed by two readers. The two readers analysed all images first independently in a blinded and random order. Disagreements were discussed before a final ROI was placed. All analysis of SPECT and MRI were blinded to the results from histopathology. Images were analysed using Osirix (version 5.8). Contrast-to-noise ratio In all participants, regions of interest (ROIs) were defined in the bone on the girdlestone location, in accordance to the points of biopsy, at the acetabulum superior, acetabulum inferior, trochanter major and trochanter minor and in reference tissues, represented by the Os ilium on the Bhealthy^ contralateral side. Signal ratio (SR) measurements were defined as described in Eq. 1. Signal ratio ¼ ½Signal ðBoneGirdlestoneÞ=Signal ðBoneContralateralSideÞ

ð1Þ

=Noiseðstandard deviationÞ

Noise was measured within an ROI placed in the air directly adjacent ventrally to the investigated body region.

Maximum sizes of nonvital bone fragments were measured on SPECT images, SPECT/CT images and MR sequences. Statistical analysis Variables are reported as mean ± standard deviation. Sensitivities and specificities of SPECT/CT compared to the different MRI sequences were computed by using receiveroperating characteristics analysis with calculation of the area under the ROC curve (AUC). Linear regression was applied to determine the relationship between signal measurements on SPECT/CT and MRI sequences. A p-value smaller than 0.05 was considered statistically significant.

Results Imaging of patients after girdlestone arthroplasty Fourteen consecutive patients (six male, eight female; mean age 65.4 ± 11.3 years) after girdlestone arthroplasty were included in this study from December 2013 to November 2015. In all patients, a two-stage revision was performed following the diagnosis of a periprosthetic joint infection. Clinical signs of periprosthetic joint infection included local inflammation, swelling, hyperthermia, hyperaemia, and tissue defects. Fourteen patients after girdlestone arthroplasty were successfully imaged using three-phase planar bone and SPECT/CT and contrast-enhanced MRI within 24 h prior to surgery (Figs. 1 and 2). A total of 56 bone areas were analysed histopathology and classified as vital (n = 39) or nonvital (n = 17). Assessment of bone viability using bone SPECT/CT and contrast-enhanced MRI Regarding bone viability relative signal measurements for bone SPECT/CT resulted in the highest value for vital bone tissue (1.76 ± 0.38) compared to nonvital bone tissue (0.82 ± 0.27, Fig. 3). Results differed significantly (p < 0.05) from each other. MR data sets based on the TIRM sequence resulted in a value of 1.6 ± 0.28 for vital bone tissue and 0.67 ± 0.32 for nonvital bone tissue. Results showed a significant difference (p < 0.05). MR data sets based on T1 fatsuppressed imaging (T1-FS) following the administration of the contrast agent resulted in a value of 1.31 ± 0.31 for vital bone tissue and 0.79 ± 0.36 for nonvital bone tissue. Results differed significantly (p < 0.05). MRI data sets based on T1 mapping following the administration of the contrast agent resulted in a value of 0.86 ± 0.13 for vital bone tissue and 1.21 ± 0.21 for nonvital bone tissue. Measurements showed a significant difference (p < 0.05).

Skeletal Radiol

Fig. 1 Example of viable bone tissue in a patient following a girdlestone procedure prior to revision total hip arthroplasty. A, B: Coronal fused bone CT and SPECT/CT images showing strong DPD uptake in the remaining vital acetabulum superior/inferior and trochanter major/minor (arrows). C, D: TIRM and T1-FS following contrast agent administration show an increased signal in the area of the vital acetabulum superior/inferior and

trochanter major/minor (arrows). The asterisk indicates imaging artefacts, resulting from remaining metal particles close to the bone following the removal of the metal components. CT computed tomography; SPECT/CT single photon emission computed tomography/ computed tomography; TIRM turbo inversion recovery magnitude; FS fat sat; CA contrast agent

Sensitivity and specificity for the differentiation of vital and nonvital bone tissue

specificity for a cutoff value of 1.51 (sensitivity 0.87 and specificity 0.88; 95% confidence interval 0.73 to 0.96 and 0.63 to 0.99). A high sensitivity and specificity was also achieved for T1-FS following the administration of the contrast agent with a cutoff value of 0.99 (sensitivity 90%, specificity 88%; 95% confidence interval 0.76–0.97 and 0.64– 0.98). For T1 mapping a moderate sensitivity and specificity was achieved for a cutoff value of 1.04 (sensitivity 82%, specificity 82%; 95% confidence interval 0.66–0.93 and 0.56– 0.96). In two locations, the signal analysis in MRI could not be reliably performed due to imaging artefacts resulting from

To determine the sensitivity and specificity for the differentiation of vital and nonvital bone tissue, histopathology was used as gold standard (Figs. 2 and 5). SPECT images from the delayed/mineralization phase showed the highest sensitivity and specificity with a cutoff value of 1.35 regarding the measured signal ratio (sensitivity: 90%, specificity 94%; 95% confidence interval 0.76–0.97 and 0.71–0.99). Regarding MRI measurements, TIRM showed a high sensitivity and

Skeletal Radiol

Fig. 2 A patient following the girdlestone procedure prior to the revision total hip arthroplasty with vital and nonvital bone tissue. A, B coronal bone CT and SPECT/CT image in a patient with reduced DPD uptake in the lateral part of the remaining proximal femoral bone (arrows). The red line indicates the location of the corresponding transversal magnified image. On the magnified transversal image (magnification B) the vital bone fragment can be clearly distinguished from the nonvital bone fragment (white arrows). On the upper and lower site to the magnification corresponding images from bone biopsies are shown. On

the left , nonvital bone tissue with fibrotic tissue and a bone sequester can be appreciated. C, D: corresponding TIRM and, T1-FS following contrast agent images are shown. Comparable to the SPECT/CT images, a reduced signal can be appreciated in the nonvital bone fragment. The asterisk indicates imaging artefacts, resulting from remaining metal particles close to the bone. CT computed tomography; SPECT/CT single photon emission computed tomography/computed tomography; TIRM turbo inversion recovery magnitude; FS fat sat; CA contrast agent

remaining metal particles following the girdlestone procedure. In these cases, the according ROIs were excluded from the analysis. Measurements are also summarized in Table 1.

observed (Fig. 4a). This overestimation most likely results from the limited spatial resolution of SPECT scans and is also depending on the chosen intensity level of the underlying colour scale (Fig. 5).

Assessment of the size of bone fragments using SPECT/CT and MRI

Discussion Size measurements revealed a close correlation for the maximum size of bone fragments between both imaging modalities for SPECT/CT versus TIRM (R2 = 0.91) and SPECT/CT versus T1-FS (R2 = 0.89). SPECT/CT versus T1 mapping (R2 = 0.76) showed a moderate correlation (Fig. 4b–e). If size measurements were performed on SPECT images alone, a slight overestimation of the size of bone fragments was

This is the first study to directly compare the clinical value of bone SPECT/CT and MRI for the assessment of bone viability in patients after girdlestone arthroplasty with histopathology as reference standard. Both techniques demonstrated a high sensitivity and specificity for the differentiation of vital and nonvital bone fragments.

Skeletal Radiol

Fig. 3 Comparison of signal measurements in bone SPECT, TIRM, T1FS following contrast agent administration and T1 mapping. A significant difference (p < 0.05) can be seen between vital and nonvital bone tissue in bone SPECT, TIRM and T1-FS including contrast agent. In T1 mapping, an increase in signal can be measured, as an increase in contrast agent

shortens the T1 relaxation time. Data are shown as mean ± standard deviation. DPD SPECT 3,3-diphosphono-1,2-propanodicarboxylic-acid single photon emission computed tomography; TIRM turbo inversion recovery magnitude; FS fat sat; CA contrast agent

Clinical relevance of bone viability in patients after girdlestone arthroplasty

treatment of the infection, revision components are reimplanted. The viability of the remaining bone fragments of the acetabulum and trochanter is dependent on different factors, including the degree and extent of the local infection, the surgical performance regarding the removal of infected components of the prosthesis, the extent of adjacent necrotic tissue and the response to antibiotic treatment. Whether local bone fragments remain vital or develop into nonvital bone fragments mainly depends on the local blood supply [13, 14]. This also plays a role in whether a clinically relevant concentration of the given antibiotic is reached in the inflamed area. If there is a substantial amount of nonviable bone, these bone parts have to be removed surgically and a different type of hip-arthroplasty has to be implanted. In clinical practice, this assessment is currently performed by surgeons intraoperatively during the revision total hip arthroplasty. Nevertheless, it could be clinically relevant to determine, prior to the revision total hip arthroplasty, whether the remaining bone of the femur and the hip after a girdlestone arthroplasty are viable or nonviable, in order to plan and perform the most appropriate surgical procedure. However, no randomized controlled or longitudinal follow-up studies reporting outcomes are currently available to support this assumption.

Total hip arthroplasty is a widespread surgical treatment method for patients with severe osteoarthritis, inflammatory osteoarthritis, osteonecrosis and developmental dysplasia. Total hip arthroplasty is associated with high rates of long-term success. However, in a limited percentage of cases it is associated with complications, including periprosthetic joint infections with reported incidents between 0.3 and 2.2% [4]. A two-stage revision technique has developed into the current gold standard [12]. This so-called girdlestone technique includes the removal of infected components of the prosthesis, followed by a time period in which the patient lives without a prosthesis. During this period, patients are treated with specific antibiotics for the local infection. Following the successful Table 1 Sensitivity and specificity for the differentiation of vital and nonvital bone tissue

SPECT TIRM T1-FS + CA T1 mapping + CA

Sensitivity

Specificity

0.90 (0.76–0.97) 0.87 (0.73–0.96) 0.9 (0.76–0.97) 0.82 (0.66–0.93)

0.94 (0.71–0.99) 0.88 (0.63–0.99) 0.88 (0.64–0.98) 0.82 (0.56–0.96)

This table demonstrates the sensitivity and specificity of bone SPECT and MRI with TIRM, T1-FS and T1 mapping sequences for the differentiation of vital and nonvital bone tissue. Both imaging modalities enabled a reliable differentiation of vital and nonvital bone tissue. SPECT single photon emission computed tomography; TIRM turbo inversion recovery magnitude; FS fat sat; CA contrast agent. Results are given as mean and 95% confidence intervals in parenthesis

Imaging-based evaluation of bone viability In general, bone viability is dependent on a sufficient blood/ oxygen supply and a balance between the different local osteoblastic/osteoclastic factors. Events that can lead to the development of nonviable bone fragments include avascular necrosis, osteonecrosis and the development of detached

Skeletal Radiol

Fig. 4 Signal and size measurements in bone SPECT, TIRM, T1-FS (including contrast agent) and T1 mapping. a Size measurements in SPECT/CT, SPECT alone and MRI. No significant difference can be found between SPECT/CT and TIRM, T1-FS. Measurements on SPECT alone however significantly overestimated the size of the bone tissue. This could be due to the lower resolution of SPECT compared to CT/MRI. The combined use of SPECT/CT enables the combination of the anatomical information from CT with the molecular information from

SPECT. b Accordingly, only a moderate correlation was found between SPECT alone and SPECT/CT. c–d A close correlation was found between TIRM/T1-FS MRI and SPECT/CT. e Only a moderate correlation was found between T1 mapping and SPECT/CT. Data are shown as mean ± standard deviation. SPECT single photon emission computed tomography; TIRM turbo inversion recovery magnitude; FS fat sat; CA contrast agent

fragments/loose bodies in case of osteochondrosis dissecans. Nuclear imaging techniques for the evaluation of bone viability are in most cases based on the application of 99mTcbisphosphonates in combination with 3-phase planar bone imaging, SPECT or SPECT/CT [15, 16]. The utility of this technique has been extensively reported and was shown to allow

the detection of avascular necrosis and osteonecrosis [17]. The underlying mechanism for the DPD signal in bone scintigraphy is the high affinity of 99mTc-labelled bisphosphonates to newly formed hydroxyapatite crystals by chemisorption during the process of bone remodelling and new bone formation. This process depends on a sufficient regional vascular supply

Fig. 5 ROC curves from SPECT/CT, TIRM and T1-FS imaging for the differentiation of vital and nonvital bone fragments. a Bone SPECT/CT showed the highest sensitivity and specificity (sensitivity: 90%, specificity 94%; 95% confidence interval 0.76–0.97 and 0.71–0.99) with a cutoff value of 1.35 for the differentiation of vital and nonvital bone tissue. Histopathology served as gold standard. b TIRM showed a high sensitivity and specificity for a cutoff value of 1.51 (sensitivity 87%

and specificity 88%; 95% confidence interval 0.73–0.96 and 0.63–0.99). c MRI achieve the highest sensitivity and specificity using a T1-FS sequence (sensitivity of 90%, a specificity 88% with a 95% confidence interval 0.76–0.97 and 0.64–0.98 was achieved). SPECT single photon emission computed tomography; TIRM turbo inversion recovery magnitude; FS fat sat; CA contrast agent

Skeletal Radiol

of the bone and sufficient osteoblastic activity [18]. A further technique is based on the use of 18F fluoride in combination with PET and PET/CT [19–24]. Imaging with 18F fluoride is based on a similar uptake mechanism as described for 99mTcbisphosphonates. Even though PET/CT with 18F fluoride can be performed with a higher image resolution and a higher sensitivity compared to bone scintigraphy, its use is not as widespread due to the associated high costs and the lower availability of PET/CT scanners [25, 26]. The most frequent technique used in clinical practice for the evaluation of bone viability is MRI. The signal in MRI is mainly based on the evaluation of the signal from fatty tissue/bone marrow on T1weighted images, the signal from bone edema on T2-weighted images and the vascularization using contrast-enhanced T1weighted MR sequences with or without fat suppression [27]. Relative signal changes in these sequences in relation to each other enable the evaluation of bone viability. A further technique which is used for the assessment of osseous changes is CT or x-ray radiography. CT or x-ray radiography are especially useful for the evaluation of reactive bone formation/ sclerosis or fractures/microfractures. The results of this study could be relevant, not only for the evaluation of patients with girdlestone arthroplasty, but also for the evaluation of bone viability in general. The evaluation of bone viability in MRI and CT is especially important in the context of avascular necrosis. The aetiology of avascular necrosis is multifactorial and has different known and unknown (idiopathic) ideologies. The known ideologies include trauma (e.g. minor recurrent trauma), gout, corticosteroid therapy, alcoholism, collagen vascular disorders and osteomyelitis [28]. Typical anatomical locations in which avascular necrosis can occur include not only the femoral head, but also the scaphoid, the capitate, the humeral head, vertebral bodies and the knee. In all these regions and especially in smaller joints, MRI is of high importance as it enables imaging with a higher spatial resolution compared to 99mTc-DPD bone scans and 18F fluoride PET. The time-course and features of pathological signal changes in MRI and bone scintigraphy at these anatomical locations are comparable to what can be observed in avascular necrosis of the femoral head, which is summarized in the ARCO (Association Research Circulation Osseous) criteria. Overall, avascular necrosis develops in typical stages. In the early stages the localized subchondral edema is a typical feature [29, 30]. An additional joint effusion and hyperemia can often be observed in this early phase of disease development. At this stage, the most useful imaging modalities include MRI and bone scintigraphy [31, 32]. In the affected bone regions MRI shows low signal intensities on T1-weighted imaging and relatively high-signal intensities on fatsuppressed T2-weighted images [33–35]. These signal changes are a result of the high amount of water in the

bone marrow. In one study, it was demonstrated that MR achieved a sensitivity of 100% compared to scintigraphy with 81% for the early detection of osteonecrosis [34]. Additional features of this early phase include hyperemia/hypervascularity, which are best visualized by a signal increase in contrast enhanced T1-weighted MR sequences and a higher tracer uptake in bone scintigraphy. On CT and x-ray images, these changes are reflected by osteopenia. In the later developmental stages of avascular necrosis, the flattening of the joint heads and the development of degenerative changes is best visualized by xray radiography or CT. Fractures/microfractures or reactive bone formation/sclerosis are also typical features which can be best detected by x-ray radiography and CT [36, 37]. The overall visualization of repair mechanisms, including the return of fatty bone marrow, is best visualized using contrast-enhanced MR techniques.

Multiphase SPECT/CT and contrast-enhanced MRI for the assessment of bone viability in patients following girdlestone arthroplasty MRI, including TIRM imaging and T1-FS imaging following the administration of a contrast agent, also demonstrated a high, however slightly lower sensitivity and specificity compared to late phase SPECT/CT. In this study collective, MRI was hampered by imaging artefacts resulting from tiny metal particles which are localized adjacent to the bone following the girdlestone procedure. Even though MRI enables imaging with a significantly higher spatial resolution compared to SPECT/CT, no nonvital bone fragments were missed by SPECT/CT. Signal measurements on SPECT/CT and MRI showed a close correlation especially with the TIRM and T1-FS sequence. Only a moderate correlation was measured using T1 mapping, which could be explained by the lower spatial resolution and the lower signal derived from this technique. The T1 mapping technique used in this study did not yield additional relevant information to the other used MR sequences and might not be required for a reliable evaluation of bone viability by MRI.

Limitations of this study Findings of this cross-sectional study were not correlated with the symptoms of patients. Results from imaging were not correlated to the short-term or long-term functional outcome of patients. Only a relatively small patient cohort was investigated in this study. Conclusions derived from MR data are not applicable to patients with contraindications to MR imaging.

Skeletal Radiol

Conclusion

13.

Both bone SPECT/CT and MRI enable a differentiation between viable and nonviable bone tissue in patients after girdlestone arthroplasty with a high sensitivity and specificity. The findings of this study could also be relevant for the evaluation of bone viability in the context of avascular bone necrosis.

14. 15.

16.

Acknowledgements The author MRM is grateful for the financial support from the Deutsche Forschungsgemeinschaft (DFG, 5943/31/41/91). 17. Compliance with ethical standards Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

18.

Conflict of interest The authors declare that they have no conflict of interest.

20.

References

21.

1. 2. 3.

4.

5.

6.

7.

8.

9.

10. 11. 12.

Siopack JS, Jergesen HE. Total hip arthroplasty. West J Med. 1995;162(3):243–9. Caton J, Prudhon JL. Over 25 years survival after Charnley’s total hip arthroplasty. Int Orthop. 2011;35(2):185–8. Berry DJ, Harmsen WS, Cabanela ME, Morrey BF. Twenty-fiveyear survivorship of two thousand consecutive primary Charnley total hip replacements: factors affecting survivorship of acetabular and femoral components. J Bone Joint Surg Am. 2002;84-A(2): 171–7. Zmistowski B, Parvizi J. Identification and treatment of infected total hip arthroplasty. Expert Rev Anti-Infect Ther. 2012;10(4): 509–18. Burge AJ. Total hip arthroplasty: MR imaging of complications unrelated to metal wear. Semin Musculoskelet Radiol. 2015;19(1):31–9. Pulido L, Ghanem E, Joshi A, Purtill JJ, Parvizi J. Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop Relat Res. 2008;466(7):1710–5. Pohlemann T, Stengel D, Tosounidis G, Reilmann H, Stuby F, Stockle U, et al. Survival trends and predictors of mortality in severe pelvic trauma: estimates from the German pelvic trauma registry initiative. Injury. 2011;42(10):997–1002. Sharma H, De Leeuw J, Rowley DI. Girdlestone resection arthroplasty following failed surgical procedures. Int Orthop. 2005;29(2):92–5. Velasco JG, Vega A, Leisorek A, Callejas F. The early detection of free bone graft viability with 99mTc: a preliminary report. Br J Plast Surg. 1976;29(4):344–6. Chang LY, Yang JY. The role of bone scans in electric burns. Burns. 1991;17(3):250–3. Zimmerli W, Trampuz A, Ochsner PE. Prosthetic-joint infections. N Engl J Med. 2004;351(16):1645–54. Matthews PC, Berendt AR, McNally MA, Byren I. Diagnosis and management of prosthetic joint infection. BMJ. 2009;338:b1773.

19.

22.

23.

24.

25.

26.

27.

28.

29.

30. 31.

32.

Atsumi T, Kuroki Y. Role of impairment of blood supply of the femoral head in the pathogenesis of idiopathic osteonecrosis. Clin Orthop Relat Res. 1992;277:22–30. Cruess RL. Osteonecrosis of bone: current concepts as to etiology and pathogenesis. Clin Orthop Relat Res. 1986;208:30–9. Bohuslavizki KH, Brenner W, Kerscher A, Fleiner B, Tinnemeyer S, Sippel C, et al. The value of bone scanning in pre-operative decision-making in patients with progressive facial asymmetry. Nucl Med Commun. 1996;17(7):562–7. Terheyden H, Warnke P, Dunsche A, Jepsen S, Brenner W, Palmie S, et al. Mandibular reconstruction with prefabricated vascularized bone grafts using recombinant human osteogenic protein-1: an experimental study in miniature pigs—part II, transplantation. Int J Oral Maxillofac Surg. 2001;30(6):469–78. Sarikaya I, Sarikaya A, Holder LE. The role of single photon emission computed tomography in bone imaging. Semin Nucl Med. 2001;31(1):3–16. Lisbona R, Rennie WR, Daniel RK. Radionuclide evaluation of free vascularized bone graft viability. AJR Am J Roentgenol. 1980;134(2):387–8. Even-Sapir E, Mishani E, Flusser G, Metser U. 18F-fluoride positron emission tomography and positron emission tomography/ computed tomography. Semin Nucl Med. 2007;37(6):462–9. Forrest N, Welch A, Murray AD, Schweiger L, Hutchison J, Ashcroft GP. Femoral head viability after Birmingham resurfacing hip arthroplasty: assessment with use of [18F] fluoride positron emission tomography. J Bone Joint Surg Am. 2006;88(Suppl 3): 84–9. Piert M, Winter E, Becker GA, Bilger K, Machulla H, MullerSchauenburg W, et al. Allogenic bone graft viability after hip revision arthroplasty assessed by dynamic [18F]fluoride ion positron emission tomography. Eur J Nucl Med. 1999;26(6):615–24. Brenner W, Vernon C, Conrad EU, Eary JF. Assessment of the metabolic activity of bone grafts with (18)F-fluoride PET. Eur J Nucl Med Mol Imaging. 2004;31(9):1291–8. Brenner W, Vernon C, Muzi M, Mankoff DA, Link JM, Conrad EU, et al. Comparison of different quantitative approaches to 18Ffluoride PET scans. J Nucl Med. 2004;45(9):1493–500. Pumberger M, Prasad V, Druschel C, Disch AC, Brenner W, Schaser KD. Quantitative in vivo fusion assessment by (18)F-fluoride PET/CT following en bloc spondylectomy. Eur Spine J. 2016;25(3):836–42. Grant FD, Fahey FH, Packard AB, Davis RT, Alavi A, Treves ST. Skeletal PET with 18F-fluoride: applying new technology to an old tracer. J Nucl Med. 2008;49(1):68–78. Blake GM, Moore AE, Fogelman I. Quantitative studies of bone using (99m)Tc-methylene diphosphonate skeletal plasma clearance. Semin Nucl Med. 2009;39(6):369–79. Manenti G, Altobelli S, Pugliese L, Tarantino U. The role of imaging in diagnosis and management of femoral head avascular necrosis. Clin Cases Miner Bone Metab. 2015;12(Suppl 1):31–8. Imhof H, Breitenseher M, Trattnig S, Kramer J, Hofmann S, Plenk H, et al. Imaging of avascular necrosis of bone. Eur Radiol. 1997;7(2):180–6. Roth A, Beckmann J, Bohndorf K, Fischer A, Heiss C, Kenn W, et al. S3-guideline non-traumatic adult femoral head necrosis. Arch Orthop Trauma Surg. 2016;136(2):165–74. Kramer J, Scheurecker G, Scheurecker A, Stoger A, Huber A, Hofmann S. Femoral head necrosis. Radiologe. 2009;49(5):410–8. Mitchell DG, Rao VM, Dalinka MK, Spritzer CE, Alavi A, Steinberg ME, et al. Femoral head avascular necrosis: correlation of MR imaging, radiographic staging, radionuclide imaging, and clinical findings. Radiology. 1987;162(3):709–15. Beltran J, Herman LJ, Burk JM, Zuelzer WA, Clark RN, Lucas JG, et al. Femoral head avascular necrosis: MR imaging with clinical-

Skeletal Radiol

33.

34.

pathologic and radionuclide correlation. Radiology. 1988;166(1 Pt 1):215–20. Genez BM, Wilson MR, Houk RW, Weiland FL, Unger HR Jr, Shields NN, et al. Early osteonecrosis of the femoral head: detection in high-risk patients with MR imaging. Radiology. 1988;168(2): 521–4. Markisz JA, Knowles RJ, Altchek DW, Schneider R, Whalen JP, Cahill PT. Segmental patterns of avascular necrosis of the femoral

35.

36. 37.

heads: early detection with MR imaging. Radiology. 1987;162(3): 717–20. Kramer J, Hofmann S, Imhof H. The non-traumatic femur head necrosis in the adult, II: radiologic diagnosis and staging. Radiologe. 1994;34(1):11–20. Ficat RP. Idiopathic bone necrosis of the femoral head: early diagnosis and treatment. J Bone Joint Surg Br. 1985;67(1):3–9. Mankin HJ. Nontraumatic necrosis of bone (osteonecrosis). N Engl J Med. 1992;326(22):1473–9.