Plateletcrit is predictive of clinical outcome and prognosis for early-stage breast cancer: A retrospective cohort study based on propensity score matching

Abstract

Purpose: Breast cancer (BC) is diagnosed as the most common cancer in women in 2022 according to the American Cancer Society. It is essential to detect early and treat early. Several studies have shown that some blood parameters have important predictive value for BC. In this study, we aim to explore whether some immune-associated blood parameters are relevant to disease-free survival (DFS) in early-stage BC. Methods: A single-center, regression cohort study of 1490 patients with early-stage BC in Shanghai Cancer Center was conducted from January 2008 to December 2016. The patients were matched according to the ratio of 1:1 based on Propensity Score Matching (PSM). All patients who experienced disease progression were matched successfully. Thus, 58 pairs of subjects were obtained. Matched blood parameters were evaluated by paired samples t-test or Wilcoxon signed-rank test. Factors with statistical difference were further evaluated by stratified COX regression model. Results: Univariate analysis showed differences in platelet-related parameters (PLT, PCT, and PLR) and NLR between the two matched groups. However, stratified COX regression analysis, which ruled out the confounding effects of multiple factors, found that only PCT had prognostic value in early BC patients at baseline and study endpoint. Meanwhile, platelet-related parameters (PLT, MPV) and NLR were different in the progressive group by self before and after comparison. However, the multiple-factor analysis showed that only the NLR had prognostic value. ROC curve analysis indicated that the best sensitivity (65.45%) and specificity (78.18%) were obtained when the baseline PCT was 0.225. The optimal sensitivity (70.91%) and specificity (65.45%) were obtained when the PCT of disease progression was 0.215. The Kaplan–Meier curve was used to calculate the DFS rate based on the critical values of the two groups. Conclusions: Some blood parameters have value to predict DFS in early-stage BC patients, especially platelet-associated parameters.