Reply to the Letter to the Editor: “Economic potential of abbreviated breast MRI for screening women with dense breast tissue for breast cancer”

by Fabian Tollens, Clemens Kaiser (clemens.kaiser@umm.de)

Economic potential of abbreviated breast MRI for screening women with dense breast tissue for breast cancer

Dear Editor in Chief,

we would like to express our gratitude for the response to our article about the economic evaluation of abbreviated breast MRI published in April 2022 in European Radiology [1]. The colleagues discuss the challenges of implementing abbreviated breast MRI protocols in developing countries, addressing both organizational and financial concerns.

Our cost-effectiveness analysis was designed to estimate costs and effects of abbreviated breast MRI for the United States healthcare system [1]. Transferability of the findings to other countries, and even to other settings within the U.S. healthcare system is limited due to differences in costs, diagnostic and therapeutic pathways, and potentially in utility levels.

We agree with the colleagues that financing and reimbursement of costly diagnostic tests such as MRI for screening requires a clear delineation of both the medical efficacy and the economic value, and needs to be employed in conformance with current guidelines.

The abbreviated protocols applied in the EA1411 ECOG-ACRIN study by Comstock et al comprised at least T2w images and T1w images before and after bolus contrast injection, and reached a level of specificity of 86.7% [2]. It has to be assumed that an abbreviated protocol without T2w imaging would achieve a lower level of specificity, which would result in smaller effectiveness and cost thresholds below 82% (see also Table 3 of our analysis reporting cost cut-offs for varying levels of specificity).

It illustrates the complexity of abbreviating breast MRI: There is no clear definition of abbreviated protocols, evidence on the diagnostic performance of abbreviated protocols is scarce and various protocols were applied in the different studies. Therefore, implementation of abbreviated protocols has to be addressed with caution.

While full diagnostic protocols are similarly undefined and may be abbreviated in individual settings, technical developments such as parallel imaging and deep learning-based reconstruction algorithms will further decrease acquisition times in the future [3–5]. The full diagnostic protocol at our institution for instance (T2, DWI and dynamic contrast enhanced series including pre- and 5 post-injection series) is acquired in 9:57 minutes.

The line between abbreviated and full diagnostic protocols will become even more blurry as acquisition times of full diagnostic protocols decrease. Nevertheless, research on abbreviating breast MRI is useful for further optimization and acceleration of MRI protocols.

When implementing optimized breast MRI protocols, quality assurance and monitoring are essential to maintain an acceptable level of specificity [6]. While the sensitivity of full and abbreviated protocols were similar in the most recent prospective multi-center trials, abbreviated protocols understandably suffered from low specificity [2, 7]. The argument concerning technologists and the heterogeneity of protocols underlines the necessity of high quality image acquisition and interpretation, especially in developing countries, as these factors directly affect diagnostic outcomes, potential follow-up and overall cost-effectiveness [8, 9].

In our opinion, workflow optimization, e.g. to reduce changeover times between consecutive examinations, and quality assurance and monitoring systems should be addressed more rigorously as opposed to protocol abbreviation, since they represent major determinants of cost-effectiveness. By today’s standards, full diagnostic protocols should be preferred to abbreviated protocols in screening, particularly in settings where high standards of quality cannot be guaranteed.