SS 310 - Tumours and inflammation
Purpose: To assess the utility of including sagittal STIR sequence of dorsolumbar spine and coronal STIR/PD fat saturated sequence through both hips, to routine SI joint MR imaging protocol, in patients clinically suspected to have SpA.
Methods and Materials: Observational study was conducted between February 2013 and October 2016 on clinically suspected SpA patients referred to our department for imaging. The images obtained using this new SI joint protocol were evaluated for findings suggesting SpA diagnosis as per the Assessment of SpondyloArthritis international Society criteria. Other differentials in symptomatic patients without radiologic SpA were also evaluated.
Results: Of the 308 patients (225M; 83F), 192 had features confirming the diagnosis of SpA and 116 had no radiological manifestations of SpA (29 were normal and 87 had other findings to suggest clinical symptoms e.g. degenerative spondyloarthropathy, Pott’s spine, skeletal metastases, early AVN of hip, cysticercus, iliofemoral impingement etc.). 18/192 patients had normal SI joints but other findings to suggest diagnosis of SpA, e.g. romanus lesions, costovertebritis/costotransversitis, pubic symphysitis, inflammatory hip arthropathy, enthesitis, iliofemoral/trochanteric bursitis. 27/49 patients with chronic sacroilitis had disease activity in spine/hip.
Conclusion: Inclusion of sections through dorsolumbar spine and both hips to routine SI joint protocol, helped identify: (a) early disease in 18 patients, who had normal SI joints and may have otherwise been missed with routine SI joint imaging protocol, (b) additional findings in SpA related sacroiliitis, (c) disease activity in chronic sacroilitis and (d) other causes of low back pain, and thus helped in further patient management.
Presence of bone marrow edema on MRI of the sacroiliac joints is predictive of subsequent (5 years) radiographic progression in early onset non-radiographic axial spondyloarthritis
Purpose: The aim of this study was to evaluate the rate of radiographic progression and its predisposing factors after a 5 years follow-up period in patients with recent onset axial SpA.
Methods and Materials: Patients: Recent onset (<3 years) axial SpA enrolled in the Cohort DESIR. Outcome measure: Radiographic SIJ status according to the mNY criteria after 5 years follow-up. Reading of the SIJ-X-Rays: 3 trained readers unaware of the chronology of baseline, 2-year and 5-year films and of the clinical and other imaging data.
Results: Of the 708 patients enrolled in the cohort 482 (68.1%) completed the 5 years follow-up. Complete X-Ray data was available in 379 patients. 326/ 379 patients were mNY negative at baseline and 23 (7.1%) of these were found positive at year 5. Conversely, 53/379 patients were mNY positive at baseline and 5 (5.7%) were found negative at year 5 (net progression of 1.4%). BME on MRI-SIJ was found in 15 (71.4%) of the patients who switched from mNY-negative at baseline to mNY-positive at year 5 and only in 51 (16.8%) patients who remained mNY-negative throughout the 5 years follow-up period. In the multivariable analysis, presence of BME on MRI was the strongest predictor of radiographic progression (OR=4.85 [95% CI: 2.95-7.97]).
Conclusion: These data suggest that 5-year radiographic progression in SIJ of patients with early onset SpA is modest and demonstrate that a positive MRI of the SIJ (presence of BME) is the strongest predictor of radiographic progression in the SIJ of these patients.
Purpose: To systematically document the whole spine MRI findings in a large cohort of synovitis-acne-pustulosis-hyperostosis-osteitis (SAPHO) syndrome.
Methods and Materials: 38 SAPHO patients (26 female, 46.3±9.3 years) were perspectively enrolled. MRI examinations of the whole spine were performed with standardised protocols. All the images were evaluated in a blinded manner, and discrepancies were reached in consensus after discussion.
Results: A total of 310 vertebrae were involved in 35 SAPHO patients. All the 35 patients had a multiple-vertebra involvement pattern (median: 7, range: 2-24), with a predilection on lumbosacral spine. 80.3%(249/310) affected vertebrae were continuously involved, while 19.7%(61/310) were separately involved. For the 310 affected vertebrae, a total of 567 lesions were observed. 35.8%(203/567) of the lesions were confined to vertebral corners, while 64.2% (364/567) also implicated adjacent endplate. Abnormal signal intensity (SI) of a mixture of bone marrow edema (BME) and fat deposition was found in 49.2%(279/567) lesions, while abnormal SI of solely BME or fat deposition was observed in 13.9%(79/567) and 36.9%(209/567) lesions, respectively. Cortical bone erosion was found in 50.6% (287/567) lesions. Vertebral collapse and bony bridge were observed in 0.6%(2/310) and 6.8%(21/310) affected vertebrae. Adjacent disk space was narrowed in 37.4%(116/310) affected vertebrae, but only 6.8%(21/310) were accompanied by hyperintensity on T2-weighted images. 1.3%(4/310) apophyseal joints were observed with hyperintensity on T2-weighted images. Thickening of anterior longitudinal ligaments was observed in 7.1%(22/310) affected vertebrae.
Conclusion: Imaging features of spinal disorders in SAPHO syndrome may facilitate better understanding of this rare disease and avoid misdiagnoses.
Initial experience with dual-phase hybrid 18F-Fluoride PET/MRI in patients with ankylosing spondylitis
Purpose: Increased 18F-F uptake during blood-pool phase PET represents regional hyperaemia, while indicating osteoblastic activity during mineralisation phase PET. This study aimed to assess the feasibility of dual-phase 18F-F PET/MRI of the sacroiliac joints (SIJ) in ankylosing spondylitis (AS) and find an association between regional hyperaemia and osteoblastic activity in different AS lesions.
Methods and Materials: Thirteen patients underwent dual-phase 18F-F PET/MRI. Blood-pool phase PET was acquired 6 and mineralisation phase PET 40 min after 18F-F injection. We analysed PET image quality (DIQ) and co-registration of blood-pool and mineralisation phase 18F-F PET/MRI. SIJ quadrants (SQ) were assessed for presence of bone marrow oedema (BME), erosion, fatty deposits (FD), sclerosis, or ankylosis on MRI and focal 18F-F uptake on PET.
Results: DIQ of both 18F-F PET datasets was high, albeit DIQ of mineralisation phase PET was better than that of blood-pool phase PET (p<0.001). Image co-registration was equally good. BME alone was associated with focal 18F-F uptake in 90.9% SQ on mineralisation phase and 63.6% SQ on blood-pool phase. FD, erosion, sclerosis, ankylosis were not associated with focal 18F-F uptake on blood-pool or mineralisation phase. SQ with BME had a significantly higher (p<0.001) percentage of focal 18F-F uptake on blood-pool phase (47.7% SQ) and mineralisation phase (86.4 % SQ) than SQ showing AS lesions without BME (6.7% and 11.7% SQ, respectively).
Conclusion: Dual-phase 18F-F PET/MRI of SIJ is feasible with high image quality. Inflammatory AS lesions presenting with BME rather than structural changes are associated with regional hyperaemia and osteoblastic activity.
Monitoring of tocilizumab treatment in patients with rheumatoid arthritis using selected imaging techniques
Purpose: The aim of the study was to assess the usefulness of imaging techniques vs. disease activity score (DAS28) for therapy monitoring in patients with rheumatoid arthritis (RA).
Methods and Materials: The study included 24 RA patients (mean (SD) age 53 (12) years; disease duration 9.8 (8.5) years; 88% anti-CCP positive, 96% rheumatoid factor positive). At baseline, 17 patients had high disease activity (DAS28 > 5.1), and 7 patients had moderate disease activity (3.2 < DAS28 ≤ 5.1). All patients received tocilizumab for a year. Hand X-ray and MRI scans, power Doppler ultrasound (PDUS) of the metacarpophalangeal joints (MCP) as well as DAS28 measurement were performed at enrolment and completion of the study. At month 6 of the study, PDUS was performed and DAS28 was calculated. PDUS (semiquantitative scoring (0-3)) scans were performed by one rheumatologist. X-ray and MRI scans were interpreted by one radiologist.
Results: At month 6, remission (DAS28 ≤ 2.6) was observed in 15 (63%) patients, and the absence of synovitis was confirmed by PDUS in 16 (67%) patients. At month 12, remission (DAS28 ≤ 2.6) was observed in 24 (100%) patients; the absence of synovitis was confirmed in 19 (79%) patients by PDUS and in 17 (71%) patients by MRI. All patients showed no radiographic progression.
Conclusion: PDUS and MRI allow for detection of subclinical lesions, which are not identified by DAS28. The use of imaging techniques may prevent premature treatment modification in patients with remission as evidenced by DAS28.
Ultrasound superb microvascular imaging in the evaluation of synovial vascularity: a preliminary study
Purpose: Superb microvascular imaging (SMI) is an advanced flow-detection ultrasound (US)-based modality able to detect very low-flow vessels with high detail and definition. Our aim was to evaluate the use of SMI in patients with early rheumatoid arthritis (RA) and RA under treatment with rituximab also comparing the diagnostic performance of SMI with that of power Doppler (PDI) and B-mode US.
Methods and Materials: 30 patients (18 women, mean age 45±11 years) affected by RA with remission-to-moderate disease activity were examined. Ulnar recess, metacarpophalangeal I-to-V and proximal interphalangeal II-to-V joints of both hands were evaluated using a high-end US system equipped with a 7-18MHz broadband linear-array transducer, PDI and SMI module. From each US examination, short video clips showing B-mode US, PDI and SMI of synovial vascularity were registered and exported for image analysis. Two radiologists reviewed video clips registered for all patients evaluating synovial vascularity intensity with a semi-quantitative scoring system.
Results: SMI demonstrated the presence of synovial vessel signals in a significantly higher number of patients than PDI (P=0.02). Inter-observer agreement for B-mode US, PDI and SMI semiquantitative scoring was moderate (k=0.59), almost perfect (k=0.87) and almost perfect (k=0.82), respectively.
Conclusion: SMI detects more vessels than B-mode US and PDI in assessing synovial abnormalities in RA patients. This may allow for early diagnosis of synovial inflammation as well as monitoring its dynamic changes under therapy.
Purpose: To assess diagnostic accuracy and inter-reader agreement of MRI texture analysis (TA) for grading of cartilaginous neoplasms of bone in comparison to visual MRI analysis.
Methods and Materials: MRI of 101 cartilaginous neoplasms (grade 0-4) were retrospectively included. T1w-, T2w-, STIR- and contrast-enhanced T1w fs-images were analysed qualitatively and quantitatively by four independent musculoskeletal radiologists. Qualitative visual MR-features (N=14) included tumour morphology as well as cortical bone, bone marrow and soft tissue abnormalities, whereas quantitative TA-parameters (N=19) included grey-level histogram, co-occurrence and run-length-matrix features. Inter-reader reliability, univariate, multivariate and ROC analysis were performed for MR- and TA-parameters separately and for combined models to determine independent predictors and diagnostic accuracy for grading of cartilaginous lesions.
Results: Mean inter-reader agreement of visual MR- and TA-features were 0.49 (range: 0.08-0.82) and 0.81 (range: 0.34-0.99) respectively. Diagnostic accuracies for differentiation of benign vs. malignant as well as for benign vs. low-grade cartilaginous lesions were 92.6% and 95.1% using a combined model of visual MR- and TA-predictors, 89.5% and 78.7% using visual MR-predictors, and 87.2% and 86.9% using TA-predictors exclusively. For differentiation of low-grade vs. high-grade chondrosarcoma no significant independent TA-predictors existed, whereas a model containing visual MR-predictors exclusively had a diagnostic accuracy of 87.4%.
Conclusion: TA showed substantial higher inter-reader reliability compared to visual MRI analysis. A model combining visual MR- and TA-features showed a higher diagnostic accuracy for differentiation of benign vs. malignant as well as for benign vs. low-grade cartilaginous lesions compared to models using exclusively visual MR- or TA-predictors.
Purpose: It is not easy to discriminate, thanks to imaging, benign lipomas and atypic lipomatous tumours (ALT). The objective was to discriminate benign and malignant fatty soft tissues tumours with shear wave elastography.
Methods and Materials: All patients were included consecutively and prospectively in our institution. The multidisciplinary staff decided a biopsy under ultrasonography guidance. All patients had ultrasonography with shear wave elastography mode (Toshiba Aplio 500). We selected fatty lesions. We assessed the size of the tumour, the elasticity (value and ratio under subcutaneous fat), and the histology of the tumour. We compared the group with benign soft tissue tumours (lipomas) and malignant tumours (Atypic lipomatous tumours).
Results: 33 tumours were included (24 lipomas and 9 ALT). The mean age was 51.8 ( +/- 15) in the benign group and 61.3 (+/- 15.1) in the malignant group (p=0.13) . The greater sizes were 82.2 mm (+/- 51) in the benign group vs 164 mm (+/- 89.2) in the malignant group (p= 0.03). The average volumes were 336.3 cc in the benign group vs 1463.3 cc in the malignant group (p=0.08). The average ratio between elasticity of the tumour and subcutaneous fat was 0.95 in malignant lesions vs 0.57 in benign lesion (p=0.05). The sensitivity and the specificity of this ratio to discriminate benign and malignant lesions were 77.8% and 75% with Area Under the Curve of 0.750 (p=0.01) with a threshold superior to 0.68.
Conclusion: We can discriminate benign and malignant fatty soft tissue tumours with shear wave elastography.
PETMR evaluation of the relationship between metabolic activity on PET and cell-density on DWI for bone metastases
Purpose: To assess the diagnostic performance of PETMR and DWI for bone metastases and evaluate whether, metabolic activity, on PET, and cell-density, on diffusion-weighted MRI, correlate in bone metastases.
Methods and Materials: Patients with histological proven cancer undergoing a PETMR (fully integrated system) for staging or restaging, with whole-body-DWI implementing the protocol, were evaluated.PETMR was rated positive when at least one hypointense lesion on T1w with high, non-inflammatory, uptake on PET (i.e.,FDG,PSMA,Dotanoc) was shown.DWI was separately evaluated.PETMR and PETMR-DWI agreement (k) with the reference standard (i.e.,histology and/or follow-up/previous examination) was assessed.The relationship between bone metastases’ metabolic activity (SUVmax,SUVmean) and cell-density (ADCmin,ADCmean) was investigated (Spearman correlation coefficient).
Results: One-hundred-three patients were evaluated (70 males;mean age±SD 58.29±16.76 years;47 staging and 56 restaging).Sixty-four patients had lymphoma, 23 prostate cancer, six melanoma, two breast-, three gastrointestinal cancer, soft tissue sarcoma, bladder, ovarian, thyroid and neuroendocrine tumour one patient each.Thirty-four patients were correctly rated as positive at PETMR and 33 at PETMR-DWI.PETMR and PETMR-DWI showed a high agreement with the reference standard (k=.915 and k=.852,respectively).SUVmax, SUVmean, ADCmin and ADCmean, mean±SD values, were 12.62±10.50, 5.82±5.58, 629±238.37 and 895.52±278.No correlation emerged between SUVmax and ADCmin (r=.013;p=.941) and between SUVmean and ADCmean (r=-0.001;p=.996) neither, separately, at staging (r=.289,p=.296;r=.475,p=.074) or restaging (r=-.023,p=.925;r=-.293,p=.211).
Conclusion: PETMR and DWI demonstrated a high diagnostic performance for bone metastases.SUV and ADC seem to be independent imaging biomarkers for bone metastases, but further studies focused on a larger population, based on cancer type and PET tracer, are necessary to fully assess this evidence.
Patient dose evaluation for whole-body skeletal CT using 100Sn filter for spectral shaping at 100kV in multiple myeloma
Purpose: To prospectively compare radiation dose parameters and image quality of an unenhanced whole-body CT protocol performed with 100kVp tube voltage in combination with a dedicated tin-filter (Sn-filter) for spectral shaping and a standard whole-body CT protocol performed with 120kVp tube voltage.
Methods and Materials: Twenty-five prospectively enrolled patients (18 men; median age 68 years) with proven multiple myeloma underwent a whole-body unenhanced 100kVp Sn-filter CT protocol. All scans were performed on a 3rd generation dual-source CT, using the following scan parameters: 100kVp tin filter, ref. mAs 280, 192 mm × 0.6 mm detector collimation, pitch 1.2. Each patient had undergone a previous CT whole-body standard protocol with 120 kVp within one year. Subjective and objective image quality was assessed in various anatomic regions and radiation dose was compared. Comparisons between the groups were analyzed with two-way ANOVA or Wilcoxon-Rank-Sum Test depending on the distribution of the data.
Results: All studies were rated as diagnostic. There was no significant difference between both protocols for subjective image quality (p>0.05). Objective image quality revealed significant higher attenuation values for the standard protocol, when compared to the 100 tin protocol for bone tissue, lung tissue and liver tissue (all p<0.05). Radiation dose was significantly less for the 100kVp tin protocol when compared to the standard protocol (mean dose-length-product 161±30 mGy*cm vs 1233±211 mGy*cm; p<0.0001).
Conclusion: 100 kVp spectral shaping whole-body CT by means of a tube-based tin-filter allows 88% dose reduction when compared to a standard 120 kVp.