Nancy Lin, Associate Chief of the Division of Breast Oncology at Dana-Farber Cancer Institute and Professor of Medicine at Harvard Medical School, highlights the prognostic potential of a multimodal artificial intelligence histopathology-based tool versus the 21-gene recurrence score in early-stage breast cancer, presented at ASCO 2026.
The following transcript has been edited for clarity.
Hello, my name is Dr Nancy Lin. I'm a medical oncologist at the Dana-Farber Cancer Institute in Boston, Massachusetts, and serve as the Associate Chief of the Division of Breast Oncology. I'm presenting the prognostic performance of a multimodal artificial intelligence histopathology-based tool versus the 21 gene recurrence score in early-stage breast cancer.
So, by way of background, the Artera breast model includes both digital pathology, as well as the clinical variables of tumor size, age, and nodal status. It is the only FDA-cleared and CE-marked risk stratification tool using digital pathology, and this tool was built and locked on phase III trial data, including over 12,000 patients in randomized controlled trials.
The purpose of this study was to compare the performance of the Artera AI tool versus the Oncotype 21 gene recurrence score in a contemporary real-world cohort of over 300 patients with ER positive, HER2 negative early-stage breast cancer. Patients with a recurrence score of 0-10 had mostly concurrent results, with 84 percent of patients also having an Artera low score.
Interestingly, if we compare the 10-year distant metastasis rates among those patients – all of whom had a recurrence score of 0-10 – who had a Artera low versus Artera high score, we see that those patients with concordantly low scores had no recurrences observed at 10 years or no distant recurrences. Whereas for those patients with discordant scores – Oncotype low, Artera high – the 10-year DM rate was 20 percent.
Looking at the intermediate risk category, 11-25, again, we see that about a third of patients have a discordant score with a high score on the Artera. And again, we see a significant difference in the outcomes with higher 10-year DM rates among those with discordant scores versus those with concordant scores. And together, these data really suggest that the Artera AI test adds additional prognostic value over and above what is provided by the 21 gene recurrence score.
I will note that for patients in these two groups, the vast majority did not receive chemotherapy, and so this is really a contemporarily treated population. So overall, it does look like the MMAI score demonstrates prognostic risk stratification in a contemporary modern treated patient population that provides additional prognostic information beyond traditional recurrence score categories.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
