Improved breast cancer prognosis through the combination of clinical and genetic markers

• DOI: 10.1093/bioinformatics/btl543
• Publication Date: 2006-11-28T03:34:16Z
• Authors: Yijun Sun; Steve Goodison; Jian Li; Li Liu; William Farmerie
• Keywords: Gene signature

Accurate prognosis of breast cancer can spare a significant number patients from receiving unnecessary adjuvant systemic treatment and its related expensive medical costs. Recent studies have demonstrated the potential value gene expression signatures in assessing risk post-surgical disease recurrence. However, these all attempt to develop genetic marker-based prognostic systems replace existing clinical criteria, while ignoring rich information contained established markers. Given complexity prognosis, more practical strategy would be utilize both marker that may complementary.A computational study is performed on publicly available microarray data, which has spawned 70-gene signature. The recently proposed I-RELIEF algorithm used identify hybrid signature through combination A rigorous experimental protocol estimate performance other approaches. Survival data analyses compare different approaches.The performs significantly better than methods, including signature, makers alone St. Gallen consensus criterion. At 90% sensitivity level, achieves 67% specificity, as compared 47% for 48% makers. odds ratio developing distant metastases within five years between with good bad 21.0 (95% CI:6.5-68.3), far higher either or markers alone.The dataset at www.nature.com Matlab codes are upon request.