Imaging of puerperal septic thrombophlebitis: prospective comparison of MR imaging, CT, and sonography

• Published in: American journal of roentgenology (1976) Vol. 169; no. 4; pp. 1039 - 1043
• Main Authors: Twickler, DM, Setiawan, AT, Evans, RS, Erdman, WA, Stettler, RW, Brown, CE, Cunningham, FG
• Format: Journal Article
• Language: English
• Published: Leesburg, VA Am Roentgen Ray Soc 01.10.1997; American Roentgen Ray Society
• Subjects: Adult; Biological and medical sciences; Cardiovascular system; Female; Humans; Investigative techniques, diagnostic techniques (general aspects); Magnetic Resonance Imaging; Medical sciences; Ovary - blood supply; Pregnancy; Prospective Studies; Puerperal Disorders - complications; Puerperal Disorders - diagnosis; Puerperal Infection - complications; Radiodiagnosis. Nmr imagery. Nmr spectrometry; Sensitivity and Specificity; Thrombophlebitis - complications; Thrombophlebitis - diagnosis; Tomography, X-Ray Computed; Ultrasonography, Doppler, Color; Sonography; Human; Radiodiagnosis; Indication; Septicemia; Cardiovascular disease; Puerperal fever; Nuclear magnetic resonance imaging; Thrombosis; Venous disease; Vascular disease; Echography; Medical imagery; Complication; Female; Computerized axial tomography; Deep vein; Diagnosis
• ISSN: 0361-803X; 1546-3141
• DOI: 10.2214/ajr.169.4.9308461

Our objective was to compare prospectively the abilities of MR imaging, CT, and sonography to reveal puerperal septic thrombophlebitis in the pelvis. Seventy-six women with puerperal fever for 5 days refractory to antimicrobial therapy underwent MR imaging, CT, and sonography. We obtained unenhanced axial CT images followed by enhanced images after the administration of an oral contrast agent for which we followed a specific protocol. Axial T1- and T2-weighted spin-echo MR images with phase reconstruction and sagittal T1-weighted MR images were obtained. Real-time sonography was performed using Doppler color flow mapping and spectral waveform analysis. Of the 76 women, 64 completed studies with all three techniques. Ovarian vein thrombosis was diagnosed in 12 women. True-positive results were indicated when at least two of the three studies showed the presence of a clot; true-negative results were indicated when at least two of the three studies showed a lack of thrombosis. MR imaging and CT revealed both ovarian veins in all cases (64/64). Sonography revealed 33 (52%) of 64 right ovarian veins and 15 (23%) of 64 left ovarian veins. MR imaging (sensitivity, 92%; specificity, 100%) and CT (sensitivity, 100%; specificity 99%) were comparable in all but two cases. In one such case, MR imaging showed patency, CT revealed findings interpreted as thrombosis, and sonography showed flow in the partially occluded vein. In the second such case, bilateral thrombosis was seen on CT, but interpretations based on sonography and MR imaging were left-sided thrombosis only. Sonography correctly revealed six of the 12 cases of ovarian vein thrombosis. CT and MR imaging proved to be the studies of choice in the evaluation of ovarian vein thrombosis.