Comment on “Supplemental breast cancer-screening ultrasonography in women with dense breasts: a systematic review and meta-analysis”

We commend Yuan et al. for having discussed the PRISMA approach ostentatiously. In accordance with the PRISMA guidelines, Yuan et al. have performed a thorough systematic review and meta-analysis examining the diagnostic performance of mammography alone plus Ultrasound (US) for breast cancer in women with dense breasts, as well as that of follow-up US in women with dense breasts and negative mammography results. We also commend the authors who have comprehensively collated data regarding recall rate, biopsy rate and biopsy-referenced positive predictive value (PPV) across multiple other studies.¹ There are several critical points that may need to be elucidated so as to augment the validity of the authors’ multifactorial findings.

Digital mammograms save images of the breast that can be scrutinised for a detailed analysis and are more effective in tracing the dense breast tissue than the previous mammographic technique. Breast MRI, 3D mammogram, molecular breast imaging have their merits and limitations but unlike ultrasound imaging they are not operator dependent in interpreting the quality of images. The limited specificity of an Ultrasound warrants the need for an additional imaging or an invasive biopsy. The multiple confounding factors affect breast density, such as age, menopausal status, medications, pregnancy, and genetics eclipse the predictive value of imaging in detecting occult cancer.^2,3

A larger group of patients and more study-specific extensive follow-up may have yielded more representative data on both the quantitative and qualitative aspects of this meta-analysis and would have garnered evidence-based recommendations for the need for diagnostic imaging.

We recommend that Tau² parameter be used in addition to the I² metric for redundancy and increased robustness in analysis of between-study heterogeneity.⁴

The lack of publication bias assessment is also a major issue. Although the authors address selection bias in the limitations of the study, publication bias has not been addressed. Publication bias is inherent to the publication process, where large studies and positive results are given precedence.⁵ The extent of this issue’s effect on the results of the meta-analysis can be identified visually through “funnel plot” symmetry, via the Egger’s bias indicator test, which has not been performed by Yuan et al. Therefore, without analysis of publication bias and its key indicators, the clinical validity of the results can be called into question.

Two other issues related to conceptual interpretation observed as follows;

1. Comparison with previous similar studies is missing.⁶

2. We also noted that the authors fail to concisely define the estimated effect size. This is a major concern as for a meta-analysis, accuracy regarding the parameter between the studies is being compared, and whether the particular parameter is appropriate for the chosen type of study, is integral to its relevance.

We highlight these issues merely to promote the clinical utility and relevance of Yuan et al.’s study and recommend that the authors of similar such studies may consider replicating these additional points.

Author contributions

R.J. predominantly conceived this review and led the development of the letter to the editor. R.J., P.S., S.S.S. and C.K. wrote the first draft of the letter, and G.K.M., R.R.M. and S.B. critically revised and edited successive drafts of the manuscript. All authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

This study did not require any ethics approval or consent to participate.

Consent to publish

Not applicable.

Data availability

All data were sourced from published journal articles.

Competing interests

The authors declare no competing interests.

Funding information

This work was not funded by any external source/body.