Patient population
This prospective institutional review board-approved study included 26 consecutive patients (14 males and 12 females; mean age of 32.3,
range 18–49 years) studied between January 2017 and January 2018.
All patients underwent DECT (80 kV and tin filter 150 kV) and MRA the same day.
The shoulder joint was distended by using a prefilled sterile syringe including a mixture of iodinate contrast material (Iopamiro 370,
Iopamidolo,
3,7 g of iodine,
Bracco imaging,
Milano),
saline (59.5%) and gadoteric acid (gadoteric acid 0,5 mmol/ml,
Dotarem,
Guerbet,
France; 0.5%).
The contrast material was injected with a 22-gauge spinal needle,
using an anterior approach under fluoroscopic guidance (mean fluoro-time 5 seconds).
No local anesthesia was performed before injection.
The mean interval between the imaging studies and surgery was 32 days (range,
3–65 days).
MRA imaging technique
A 1.5 Tesla scanner (Siemens MAGNETON Avanto,
Siemens Medical Systems,
Erlangen Germany; SQ-engine; gradient strength 45 mT/m; slew rate 200 T/m/s) was used.
Patients were imaged supine with the humerus in a neutral position and the thumb pointing upward,
by using a flexible four-channel body matrix phased-array surface coil (Siemens body matrix) and a six-channel spine matrix coil (Siemens Spine matrix).
Conventional turbo spin echo sequences were acquired first in all patients enrolled,
followed by isotropic 3D T1-weighted VIBE sequence (voxel size=0.8x0.8x0.8mm; imaging time 4 min and 28s).
MRA Imaging parameters are summarized in table 1 (fig 1).
DE-CTA imaging technique
The DECT examinations were performed with a third generation 384-slice dual source CT scanner (Somatom® Definition Force,
Siemens Healthcare,
Forchheim,
Germany).
The scanning parameters were as follows: tube A 80 kV and tube B 150 kV with a tin filter.
The predefined tube current–time product was set at a ratio of 1.6:1 (tube A,
220 quality reference mAs; tube B,
138 quality reference mAs).
Automated attenuation-based tube current modulation (CARE dose 4D; Siemens Healthcare) was used.
DECT post-processing
After each DECT,
soft-tissue kernel (Qr32) 80-kVp and 150-kVp set images were transferred to an offline workstation (SyngoVia® VB20; Siemens,
Erlangen,
Germany).
A three-material decomposition algorithm was applied and multiple look-up tables were available.
The DE-specific information was fused with the conventional gray-scale morphologic images (thickness,
1 mm; increment,
1 mm).
Each reader was free to adjust the kV values in the post-processing in order to reduce metal artifacts,
or to enhance the visualization of contrast material injected between articular cavity.
Image analysis
DECT images were reviewed by two experienced radiologists (21,
11 years of experience) and MRA images by a third radiologist (24 years of experience) blinded to clinical and surgical data. An additional reading session was performed by the Readers 1 and 2 for the calculation of intra-observer agreement,
with 8 weeks delay,
in order to reduce recall bias.
Each dataset was assessed for superior labral anterior posterior (SLAP) tears and anterior or posterior labral tears.
Labral lesion were diagnosed in case of partial-thickness tears (partial defect of the labrum,
enhanced by the passage of contrast material,
or by the presence of abnormal morphology or partial dislocation),
and in case of full thickness tears (represented by the visualization of a complete defect of the labrum or its absence),
and in case of detachment of labrum. A binary classification was used (0,
no labral tear; 1,
presence of labral tears),
without grading or sub-classifying the lesions.
For any disagreement on presence labral tear,
a consensus reading was appended,
and the consensual results were used for further analysis.
Surgery
Surgical data were considered as the reference standard.
Surgery was performed by 2 experienced surgeons (18 and 12 years of experience,
respectively) aware of MRA and CTA imaging findings.
Statistical analysis
Data were analyzed with the receiver operating characteristic (ROC) method and 95% confidence intervals (CIs).
The sensitivity,
specificity,
positive predictive values (PPV),
negative predictive values (NPV) and accuracy (Acc) for all types of labral tears were calculated for each dataset analyzed.
A value of p<0.05 was considered statistically significant.
Inter-observer agreement was calculated by using kappa statistics.