Volpara Health was delighted to be a silver sponsor at the UK’s biennial meeting of breast specialists, the Symposium Mammographicum, which celebrated its 40th anniversary in a virtual conference held from February 7-10. Reflecting on the sessions, we have summarized our highlights and key takeaways.
In a Volpara-sponsored knowledge share session, Professor Carla H. van Gils from the University Medical Center Utrecht in the Netherlands discussed the latest results from the DENSE trial. This trial uses Volpara’s Density Grade (VDG), an automatic assessment of volumetric breast density, to identify women in the Dutch screening program eligible for random selection for supplemental screening with MRI. Using Volpara’s VDG to triage women to MRI resulted in a significant reduction of interval cancers, or cancers found between a negative mammogram result and the next scheduled screening. In fact, just after the close of the conference, the Dutch government voted to approve the addition of MRI exams for women with dense breasts to the national screening program based on the success of this trial.
Many of the sessions centered around the impact of the COVID-19 pandemic on the UK Breast Screening Programme. Dr. Claire Mercer presented on the resilience of mammography educators and the ways in which they rapidly innovated to train and support their staff remotely – a place where Volpara believes both Live and Analytics can play important roles in our ‘new normal’.
Furthermore, Professor Gareth Evans from the University of Manchester gave an update on the UK-based PROCAS2-BCPredict trial which is recruiting again after a pause in UK Breast Screening Services in 2020 due to the COVID-19 pandemic. This trial uses Volpara’s Volumetric Breast Density Percentage (%VBD) as an input to the Tyrer-Cuzick risk model to stratify women to adapted breast screening protocols and/or cancer preventative drugs according to their risk assessment. Evidence from this trial is expected to represent an important advancement in knowledge for personalized screening and cancer prevention.
There was also discussion about the role of breast density and artificial intelligence (AI) in improving breast screening and breast radiology. Dr. Lucy Warren from the Royal Surrey NHS Foundation Trust, presented results on the development of an AI density assessment tool for processed images. Warren’s study used Volpara’s %VBD derived from raw images as the gold standard and demonstrated that the accuracy of AI-based density assessments can vary substantially depending on the vendor mix used in the training and test datasets. This finding sparked some interesting debate about the wider implications for AI development using processed images alone given the potential lack of robustness of the algorithms to new flavors of image processing developed by manufacturers.
Finally, discussions around objective, quantitative breast density assessments took center stage in a few talks. Dr. Laurie Fajardo, a professor at the University of Utah, discussed automated density measurement and advocated for objective measurements on a continuous scale, such as Volpara’s %VBD. This would provide finer gradations across populations than an ACR BI-RADS assessment and improve the accuracy of masking and cancer risk estimation. Finally, Professor Michael Webster from the University of Nevada highlighted the visual subjectivity of breast density assessments by radiologists. Webster reported that a radiologist’s visual density assessment could be heavily influenced by the breast composition of the previous studies they viewed.
Volpara looks forward to meeting with colleagues in person again at the next Symposium Mammographicum and hearing the outcomes from the exciting research currently underway.