Back

The Prognostic Value of CanAssist Breast in Patients <=50 Years: A Retrospective Study

Sunder, M.; Durgekar, T. D.; Goutham, S.; Savitha, B. A.; Shrivastava, P.; Krishnamoorthy, N.; Shivashimpi, D. K.; S J, K. A.; Raghuram, A.; Bakre, M. M.

2026-08-06 oncology
10.64898/2026.08.04.26359653 medRxiv
Show abstract

Background: Patients aged [&le;]50 years with early-stage HR+/HER2- breast cancer are considered to have an aggressive disease biology and are treated with chemotherapy. However, a subset may still experience favourable outcomes without chemotherapy. Commercially available prognostic tests help guide such treatment decisions, but most are developed and validated predominantly in Western populations, with an underrepresentation of Asian patients. In this study, we explore the prognostic value of CanAssist Breast (CAB), a proteomic prognostic test, in optimal treatment management of patients aged [&le;]50 years. Methods: This study includes a previously published retrospective cohort. The performance of CAB was evaluated using Kaplan-Meier analysis, with 5-year Distant recurrence-free interval (DRFI) from diagnosis as the endpoint; the study also used multivariate analysis to evaluate the independent prognostic value of CAB. Results: In the retrospective cohort, CAB identified 70% as low-risk (LR) and 30% as high-risk (HR) with DRFI of 93.1% (P<0.0001); further classification showed 64% LR and 36% HR in the Asian and 75% LR and 25% HR in the Caucasian subgroup. In patients with N0 disease, CAB identified 85% as LR and 15% as HR. In N+ patients, CAB identified 49% as LR. All CAB LR patients have an acceptable DRFI of >90% at 5 years from diagnosis. Conclusions: Based on the results presented, CAB adds prognostic value for patients [&le;]50 years and can be used as a treatment guidance tool for these patients.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.