A nomogram for predicting the HER2 status of circulating tumor cells and survival analysis in HER2-negative breast cancer.

ABSTRACT

Background: In breast cancer patients with HER2-negative tumors (tHER2-), HER2-positive CTCs (cHER2+) were associated with promising efficacy of HER2-targeted therapy, but controversy has persisted over its prognostic effect. We developed a model including clinicopathologic parameters/blood test variables to predict cHER2 status and evaluated the prognostic value of cHER2+ in tHER2- patients. Methods: cHER2+ was detected, blood test results and clinicopathological characteristics were combined, and a nomogram was constructed to predict cHER2 status in tHER2- patients according to logistic regression analysis. The nomogram was evaluated by C-index values and calibration curve. Kaplan-Meier curves, log-rank tests, and Cox regression analyses were performed to evaluate the prognostic value of cHER2 status. Results: TNM stage, white blood cells (WBCs), neutrophils (NEUs), uric acid (UA), De Ritis ratio [aspartate transaminase (AST)/alanine transaminase (ALT)], and high-density lipoprotein (HDL) were found to be associated with cHER2 status in tHER2- patients in univariate logistic regression analysis, in which UA and De Ritis ratio remained significant in multivariate logistic regression analysis. A model combining these six variables was constructed, the C-index was 0.745 (95% CI 0.630-0.860), and the calibration curve presented a perfect predictive consistency. In survival analysis, patients of the subgroups "with cHER2+/UA-low" (p = 0.015) and "with cHER2+/De Ritis ratio - high" (p = 0.006) had a significantly decreased disease-free survival (DFS). Conclusions: Our nomogram, based on TNM stage, WBC, NEU, UA, De Ritis ratio, and HDL, may excellently predict the cHER2 status of tHER2- patients. Incorporation with UA and De Ritis ratio may enhance the prognostic value of cHER2 status.