Which is the most effective modality for detecting adenomyosis in adult females: Magnetic resonance imaging or transvaginal ultrasound? A systematic literature review

Abstract

This systematic literature review's (SLR) aim was to determine the accuracy of transvaginal ultrasound (TVUS) and magnetic resonance imaging (MRI) in the diagnosis of adenomyosis in adult females. Since the symptoms of adenomyosis are non-specific, and since the timely and accurate detection of adenomyosis is essential for women's quality of life, health, and fertility, it is advantageous that non-invasive diagnostic imaging modalities, including TVUS and MRI, have been advancing continuously in recent years. Both modalities are frequently used in routine diagnostic practice to detect adenomyosis and to plan clinical pathways, but knowledge regarding each one's sensitivity and specificity has yet to be concretised with a reliable and transparent systematic review and meta-analysis. Therefore, it was the focus of this SLR to illuminate the relative strengths and weaknesses of each one, to identify the diagnostic test accuracy of the respective modalities, and to offer evidence-based recommendations for clinicians regarding the optimal strategy for diagnosing adenomyosis.

The applicability, reliability, and clinical relevance of the findings generated by this SLR were primary concerns for the author, since without these it would not be possible to contribute to evidence-based practice in the diagnosis of adenomyosis. Therefore, every opportunity was taken to systematise, to reinforce, and to increase the quality and transparency of the various elements of the SLR, and to ensure the broadest applicability and generalisability of the findings. For example, a quantitative SLR was undertaken rather than an SLR that utilised qualitative data, principally because this enabled a reproducible and generalisable conclusion to be drawn about the sensitivity and specificity of TVUS and MRI in the diagnosis of adenomyosis. Additionally, a robust and rigorous search strategy was devised that leveraged Boolean operators, as well as a range of electronic databases, in order to ensure the broadest possible engagement with the literature. Finally, well-validated and widely-recommended protocols for search result screening, QUADAS-2 quality appraisal, and data extraction were adopted. As a result of the systematic search procedure applied in the SLR, a total of 4 relevant and high-quality articles were considered eligible for inclusion.

Although this SLR suffers from clear limitations and methodological shortcomings, each of the studies included in the SLR (namely, Sun et al. (2010), Luciano et al. (2013), Tellum et al. (2019), and Rasmussen et al. (2019)) was evaluated as having either moderate or high methodological quality. This has promising implications for the relevance and usability of the SLR's findings in routine clinical practice. Additionally, the included studies were all cohort studies, and between them, they applied MRI, 2D-TVUS, or 3D-TVUS to a combined total of 384 female patients with suspected adenomyosis. Taken together, the key findings from the SLR after extracting data on diagnostic test accuracy from the included studies were the following: firstly, MRI had the highest sensitivity (84.0%), followed by 3D-TVUS (80.5%) and 2D-TVUS (75.0%); secondly, 3D-TVUS had the highest specificity (84.5%), followed by MRI (80.5%) and 2D-TVUS (67.0%); thirdly, 2D-TVUS had the lowest diagnostic test accuracy; and finally, clinically relevant factors other than the observed differences in accuracy between 3D-TVUS and MRI suggest that, in clinical diagnostic practice, TVUS – ideally 3D-TVUS – should be used as the front-line modality, with MRI as a secondary support modality.

While a comparison of this SLR's findings against the current knowledge base indicates that, for the most part, the findings reported elsewhere in the literature are consistent with those given here, there are notable disparities that should be drawn attenti