| ObjectiveThe purpose of our study was to establish the optimal model and screen out bile acids biomarkers for the differential diagnosis of benign and malignant breast lesions by analyzing the profilings of serum bile acids in patients with breast cancer(BC)and benign breast disease(BBD).Methods1.Twenty-nine patients with BBD and 47 patients diagnosed with BC were recruited from November 2017 to January 2018.Clinical information was collected,including age,height,weight,age of menarche and comorbidities information.2.Twenty-four known bile acids and 15 possible bile acids in serum of BC and BBD patients were determined by ultra-high performance liquid chromatography-triple quadrupole time-of-flight mass spectrometry(UPLC-Triple TOF-MS/MS)by combining with targeted and non-targeted strategies to find the difference of serum bile acids contents between BBD and BC patients without comorbidities.3.Partial least squares discriminant analysis(PLS-DA)was used to establish the differential diagnostic model of benign and malignant breast lesions based on data from patients without comorbidities.The receiver operating characteristic(ROC)curve analysis and Logistic regression analysis were used to screen bile acids as combined biomarker for differential diagnosis of benign and malignant breast lesions.4.The differential diagnostic accuracies for benign and malignant breast lesions of the established model,the differential diagnostic biomarker and clinical biochemical markers were compared in patients with comorbidities.Results1.No significant differences were found in age,body mass index(BMI),age of menarche and liver function between BBD group and BC group without comorbidities.There were 16 patients with comorbidities in the BBD group(55.2%)and 27 patients with comorbidities in the BC group(57.4%).2.In patients without comorbidities,the BC group had higher level of chenodeoxycholic acid(CDCA)(P<0.05)and lower levels of two possible bile acids,a dihydroxy taurine-conjugated bile acid(Tdi-1)and a sulfated dihydroxy glyco-conjugated bile acid(Gdi-S-1)(P<0.05)compared to the BBD group.On the other hand,due to the elevated free BAs and the decreased level of conjugated BAs,the ratios of free to conjugated BAs among targeted and possible BAs in the BC group showed 1.92-fold and1.30-fold elevated(P<0.05),respectively,compared to the BBD group.Attributed to the decreased T-conjugated BAs,G-to T-conjugated BAs ratio in possible BAs of the BC group was also 2.62 folds higher than that in the BBD group(P<0.05).3.There were differences in the profiles of serum bile acids between the two groups of BBD and BC without comorbidities.PLS-DA model showed that serum bile acids profiles could distinguish the two groups.The model showed high fitting degree(R~2Y=0.791),good prediction ability(Q~2=0.683),and no over-fitting.The sensitivity and specificity of the established model were 100.0%and 92.3%,respectively.By combining with ROC curve analysis and regression analysis,CDCA and Tdi-1 were selected as combined biomarker for differential diagnosis of benign and malignant breast lesions.The area under the curve(AUC)was 0.954(95%CI:0.880~1.000),and the sensitivity and specificity were 95.0%and 92.3%,respectively.4.In patients with comorbidities,the differential diagnostic accuracies of the established model were 66.7%~100.0%.The accuracies of differential diagnostic biomarkers were 54.2%~66.7%.The differential accuracies of clinical biochemical markers were only 33.1%~39.3%.ConclusionThere were differences in the profilings of serum bile acids in patients with BBD and BC.Serum bile acids profilings and combined bile acids biomarker could be used in the differential diagnosis of benign and malignant breast lesions,and their differential diagnostic accuracies were superior to clinical biochemical markers. |