Improving Breast Cancer Prognostic Models with Biomarkers
Author Information
Author(s): Fabio Parisi, Ana M González, Yasmine Nadler, Robert L Camp, David L Rimm, Harriet M Kluger, Yuval Kluger
Primary Institution: Yale University School of Medicine
Hypothesis
Can incorporating both molecular markers and clinico-pathological variables improve breast cancer survival predictions?
Conclusion
Prognostic models that include both molecular and clinico-pathological covariates can be more accurate than models based on either set of features alone.
Supporting Evidence
- A prognostic index derived from a multivariate Cox regression model incorporating molecular and clinico-pathological covariates is superior to models based on molecular studies alone.
- The mean cross-validated average time-dependent AUCROC of the 18-covariate model is 0.746 at 15 years.
- Feature selection can decrease the number of molecular variables needed to predict outcome.
Takeaway
This study shows that using both protein markers and clinical information can help doctors better predict how breast cancer will affect patients.
Methodology
The study used quantitative immunofluorescence to analyze protein expression of 14 markers in a cohort of breast cancer patients and developed multivariate Cox models.
Limitations
The study lacked a cohort where the Oncotype Dx test was performed for direct comparison.
Participant Demographics
Patients aged 24 to 88 years, with a mean age of 58 years; included both node-negative and node-positive breast cancer patients.
Statistical Information
P-Value
p<10-10
Statistical Significance
p<10-10
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website