| Objective: To create the Nomogram prediction model and assess its effectiveness in foretelling the outcome of neoadjuvant treatment for breast cancer,ultrasonic image characteristics and clinical markers were used.Methods: The clinical and imaging data of patients who received complete neoadjuvant chemotherapy(neoadjuvant chemotherapy,NAC)in the breast Surgery Department of the Northern Theater Command General Hospital from October 2019 to February 2023 were retrospectively analyzed.Basic data like age,clinical data like menstruation status,immunohistochemistry results,postoperative pathological results,and ultrasonic imaging features like maximal diameter,boundary,and echo type of mass were collected in order to examine the prediction of NAC treatment effect on breast cancer.We subsequently enrolled 115 individuals after screening them based on pertinent criteria.The patients were split into two groups—pathological complete response(pathological complete response,p CR)group and non-PCR group—by comparing the postoperative pathological data.Then,to design the prediction model and create the ROC curve to assess the predictive power of the model,we first used LASSO regression analysis to screen the distinctive factors associated to p CR.After that,we utilized Logistic regression to screen independent predictors.To ensure that the calibration model was consistent,a calibration curve was constructed.With the help of this study,we hope to be able to more precisely anticipate how well breast cancer treatments will work and improve patient care.Results: 1.Efficacy evaluation of NAC:51(44.3%)of the 115 patients who received neoadjuvant chemotherapy were given p CR,while 64(55.7%)were not.2.After NAC,a single-variate p CR analysis was achieved:Between the two groups,there were significant differences in age(χ2=17.985,P<0.001),menstrual status(χ2=6.166,P=0.013),boundary(χ2=20.719,P<0.001),echo type(χ2=7.001,P=0.002),internal calcification(χ2=5.921,P=0.015),posterior echo(χ2=17.837,P<0.001),Ki-67 expression rate(χ2=32.718,P<0.001)Statistics showed that there were differences(all P <0.05).3.Results of LASSO regression analysis using filtered variables:Age,the largest diameter of the mass,the burr on the boundary and edge,the posterior echo,the rate of Ki-67 expression,the expression of the ER and HER-2genes—these eight factors are indicative of the likelihood that p CR will be achieved following neoadjuvant chemotherapy.4.Analysis of p CR following NAC forecast using multivariate logistic regression:The maximum tumor diameter(OR=0.033,95%CI=0.003～0.193),the presence of burrs at the margin(OR=0.073,95%CI=0.006～0.613),the expression rate of Ki-67(OR=85.256,95%CI=3.964～6405.428),and the expression of ER(OR=0.073,95%CI=0.008～0.442)were all independently associated with the likelihood that neoadjuvant chemotherapy.5.Construction of a nomogram prediction algorithm and diagnostic outcomes:The four factors chosen above were used to create a Nomogram prediction model(maximum diameter,margin,Ki-67 expression rate,ER expression).The Nomogram model predicted the probability of p CR for patients receiving neoadjuvant chemotherapy to be 0.951,showing good predictive efficacy,according to ROC curve analysis.It performed more accurately in terms of diagnosis than using just the maximal diameter,margin,Ki-67,and ER(AUC=0.582,0.668,0.756,and 0.742,respectively).This Nomogram prediction model demonstrates that the actual curve is fairly close to the45° straight line after internal validation and calibration.Conclusions:1.Following neoadjuvant treatment for breast cancer,the maximum diameter of the mass,the existence of a burr at the edge,and the expression of Ki-67 and ER are all independent predictors of p CR.2.In this study,we developed a Nomogram model to accurately predict the therapeutic benefit of neoadjuvant chemotherapy for breast cancer patients using ultrasonic image features and clinical signs. |
PDF Full Text Request