Diagnostic accuracy of 3.0T diffusion‐weighted MRI for patients with uterine carcinosarcoma: Assessment of tumor extent and lymphatic metastasis

2018 
BACKGROUND: Assessment of tumor extent and lymphatic metastasis of uterine carcinosarcomas is important for treatment planning. PURPOSE/HYPOTHESIS: To evaluate the diagnostic accuracy of 3.0T diffusion-weighted (DW) MRI for patients with uterine carcinosarcoma, in assessment of tumor extent and lymphatic metastasis. STUDY TYPE: Retrospective diagnostic accuracy study. POPULATION: A consecutive cohort of 68 patients with pathologically proved carcinosarcoma between January 2006 and July 2014. FIELD STRENGTH/SEQUENCE: 3T DW MRI. ASSESSMENT: Maximal tumor and uterus size, presence of deep myometrial invasion, cervical invasion, adnexal invasion, lymphadenopathy, and the apparent diffusion coefficient (ADC) values of each tumor were used. Histopathology was the gold standard. STATISTICAL TESTS: Diagnostic accuracy. Logistic regression. RESULTS: In all, 38 patients entered the final analysis, with median age of 58 years (range, 35-79 years). The sensitivity and specificity in detecting deep myometrial invasion, cervical stromal invasion, adnexal invasion, as well as pelvic and para-aortic lymph node metastases were 65% and 72%, 91% and 85%, 50% and 100%, 33% and 89%, and 33% and 100%, respectively. The largest tumor diameters predicted deep myometrium invasion (anteroposterior direction, P = 0.004) and cervical stroma invasion (craniocaudal direction, P = 0.008). Tumor ADCmin significantly predicted the lymphovascular permeation (P = 0.025; odds ratio = 0.96). DATA CONCLUSION: Preoperative DW MRI is useful to assess deep myometrial or cervical stromal invasion in uterine carcinosarcoma, yet the diagnostic performance for detecting adnexal invasion and lymphatic metastasis requires further improvement. LEVEL OF EVIDENCE: 3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    32
    References
    5
    Citations
    NaN
    KQI
    []