Magnetic resonance imaging and ultrasound for assessing parametrial infiltration in cervical cancer. A systematic review and meta-analysis

Abstract

Aims: To provide information on the current evidence regarding the diagnostic performance of ultrasound and MRI for assessing parametrial involvement in cervical cancer using the histological report as the reference standard. Material and methods: Meta-analysis. An extensive search of papers comparing ultrasound and MRI in assessing parametrial infiltration in cervical cancer using pathologic analysis as a reference standard was performed in Medline (Pubmed) and Web of Science from January 1990 to September 2019. Quality was assessed using the QUADAS-2 tool. Results: Our extended search identified 205 citations but after exclusions we finally included 9 articles in the meta-analysis. The risk of bias for most studies was low for four domains were assessed in QUADAS-2. Overall, for ultrasound pooled estimated sensitivity and specificity for diagnosing parametrial infiltration was 78% (95% confidence interval [CI]:48%–93%) and 96% (95% CI=89%–99%), respectively. For MRI these figures were 68% (95% CI=54%–80%) and 91% (95% CI=84%–95%), respectively. No statistical differences were found when comparing both methods (p=0.548). Heterogeneity was low/moderate for MRI and high for ultrasound. Conclusion: Ultrasound and MRI have similar diagnostic performance for detecting parametrial infiltration in women with cervical cancer. This might have relevance from the clinical point of view, since ultrasound is cheaper than MRI.