[MR venography in deep venous thromboses of the leg and pelvis. A comparison of 2D single layer images and 3D MIP reconstructions with phlebography]

Magnetic resonance imaging and ultrasonography in diagnosis of pelvic vein thrombosis during pregnancy

INTRODUCTION

Pelvic deep vein thrombosis (DVT) is difficult to diagnose during pregnancy. In a two-center trial, we evaluated the agreement between ultrasonography and magnetic resonance imaging (MRI) in diagnosing the extent of DVT into the pelvic veins during pregnancy.

MATERIALS AND METHODS

Pregnant women with proximal DVT were examined both with ultrasound and MRI as part of a study designed for treatment of DVT during pregnancy. Ultrasound was performed using color flow by specialist in vascular ultrasound with Doppler and compression techniques. The MRI sequences consisted of a 2D Time of Flight angiography with arterial flow suppression and maximum intensity projection reconstructions; a 3D, T1-w-prepared gradient echo sequence with fat saturation for thrombus imaging; a steady-state free precession sequence; and a Turbo-Spin-Echo. No contrast agent was used. Proportion of agreement (kappa) for detection of DVT in individual veins was measured for different ipsilateral veins and inferior vena cava.

RESULTS

All 27 patients were imaged with both techniques at an average gestational age of 29 weeks (range 23-39). Three cases (11.5%) of DVT in the pelvic veins were missed on ultrasound but detected by MRI. The upper limit of the DVT was always depicted at a higher (20 cases, 65.4%) or the same level (seven cases, 34.6%) on MRI than on ultrasound. Agreement expressed as kappa was 0.33 (95% CI 0.27-0.40) demonstrating only fair agreement. In one woman the thrombus had propagated into the inferior vena cava, shown only on MRI.

CONCLUSION

Our study suggests that in pregnant women there is only fair agreement between ultrasound and MRI for determination of extent of DVT into pelvic veins, with MRI showing consistently more detailed depiction of extension. Our results indicate that MRI has an important role as a complementary technique in the diagnosis of DVT during pregnancy.

Value of a blood pool contrast agent in MR venography of the lower extremities and pelvis: Preliminary results in 12 patients

The purpose of this study was to determine the value of a blood-pool contrast media (NC100150, Nycomed Imaging (now Amersham Health) Oslo, Norway) for evaluation of venous thrombosis of the deep veins of the pelvis and lower extremities. Twelve patients were prospectively evaluated with conventional X-ray venography (XRV) and MR venography (MRV) after injection of NC100150 (2 ml/kg body weight). The source images and 3D maximum intensity projection (MIP) were viewed on an independent workstation. Diagnosis was made in consensus from two radiologists. Diagnostic image quality was achieved in 87 veins with XRV and MRV. As determined by XRV, thrombus was present in 30 out of 87 veins (34.5%). There was agreement concerning absence or presence of thrombi in 83 out of 87 veins (95.4%; kappa = 0.9 +/- 0.05). Compared to XRV, overall sensitivity and specificity of blood-pool MRV were 93.3% and 96.5%, respectively. Two venous thromboses of the popliteal and posterior tibial vein were diagnosed in MRV, but not in XRV. Conversely, two venous thromboses below the knee had been missed by MRV. NC100150 allows prolonged and improved visualization of the peripheral vasculature and may overcome some limitations of gadolinium contrast media. A more complete examination of the proximal venous tree may be possible than with conventional XRV. Arterial and venous enhancement and motion artifacts can limit image interpretation.