| Background: Hormone receptor(HR)positive and human epidermal growth factor receptor 2(HER2)negative breast cancer subtypes account for the largest proportion of all breast cancer types,and the prognosis of this subtype Preferably,only systematic endocrine therapy is exempted from chemotherapy.However,there are still some patients who have undergone redundant treatment because the test cannot accurately distinguish the benefits of chemotherapy,and suffer the psychological and physical harm caused by adjuvant chemotherapy.Nowadays,the commonly used clinical methods to judge the prognosis risk of hormone receptor-positive and HER2-negative breast cancer patients to determine whether they should be chemotherapy are often Ki67 immunohistochemical detection and 21-gene detection based multi-gene detection methods,and a single Ki67 immune group The chemical method is limited by the experience and technical means of pathologists,and there is a large randomness.However,the high requirements of 21 gene testing and detection technology and facilities are difficult to popularize in hospitals.There is no unified testing standard,and the high cost puts a certain burden on patients.Both methods have their own advantages and disadvantages,but if they can combine the advantages of immunohistochemistry and multi-gene detection,detect the expression of multiple key index proteins on the basis of immunohistochemistry,and establish the risk prediction of hormone receptor-positive and HER2-negative early breast cancer patients The model is expected to improve the accuracy of judging such patients’ chemotherapy decisions under the most economical conditions,and is easy to standardize and popularize in various hospitals,which will bring good news to patients.Method: In the first part,we established a predictive model of hormone receptor-positive and HER2-negative early breast cancer based on immunohistochemistry and discussed its clinical significance.We collected 634 cases of early breast cancer patients who underwent surgery in the First Affiliated Hospital of China Medical University from April 2010 to November 2016.Clinicopathological information,follow-up information and tissue wax blocks were collected.Immunohistochemical detection of Ki67,Aurora A,Survivin,Cyclin B1,MMP11,MYBL2,MYBL2 and Cathepsin L2 protein expression and screening.The patients were randomly divided into training cohort and test cohort according to 7:3,the baseline data of the patients were described,and the differences in clinicopathological characteristics between the groups were analyzed by chi-square test and logistic regression.Through univariate and multivariate Cox regression analysis,the indicators with survival prediction significance were screened out,and the risk function of each protein expression was calculated as the weight and the risk prediction model was established.Substitute the immunohistochemical expression into the model formula to obtain the PI value corresponding to the patient,draw the respective ROC curves of the training cohort and the test cohort to verify the consistency between the groups,and use the Youden index to determine the cut-off value for predicting the high and low risks of the model.Draw the Kaplan-Meier curve of the training cohort and the test cohort to verify and study its predictive value for the patient’s prognosis again.Later,univariate and multivariate Cox regression analysis were used to evaluate whether the prognostic model can be used as an independent clinicopathological factor.The second part is a retrospective study based on the first part,exploring the predictive effect of the 6-IHC comprehensive score based on immunohistochemistry on the benefit of chemotherapy and the clinical value of comparison with Ki67.The patients were divided into 6-IHC prediction model high-risk and low-risk groups according to PI=2.16,and the patients were divided into Ki67 according to the high-risk and low-risk groups according to Ki67 IHC=14%.Logistic regression analysis was used to compare the differences in clinicopathological characteristics between 6-IHC prognostic model and Ki67 in different risk subgroups.Use SPSS and Med Calc statistical software to draw ROC curve and calculate the sensitivity and specificity of each curve,and calculate whether the difference between the groups is statistically significant.Kaplan-Meier analysis of the total population DFS was performed to draw a survival curve,and the log-rank test was used to compare the predictive value of 6-IHC score and Ki67 on the benefit of chemotherapy.In addition,lymph node metastasis was added as a stratification factor to compare the negative and positive lymph nodes.The benefit of chemotherapy was evaluated by the 6-IHC predictive model score and Ki67 expression.In the third part,a part of paraffin tissue specimens were selected from the specimens that have been stained and evaluated by 6-IHC immunohistochemistry in the aforementioned study,and the 21 genes detected by 21 genes were detected by RT-PCR,and the corresponding m RNA expression levels were obtained,and the 21 genes were substituted The corresponding RS value is obtained in the calculation formula.Calculate the skewness coefficient,kurtosis coefficient and draw a histogram to judge the distribution of RS values.Draw the ROC curve of Survivin,Aurora A,Cyclin B1,Cathepsin L2,MMP11 and Ki67 immunohistochemical protein expression and m RNA positive expression,analyze and evaluate the correlation between the respective m RNA expression and IHC results.The PI value of the 6-IHC scoring model and the RS value of21 gene detection were performed Kappa test to evaluate the consistency.The patients were divided into 6-IHC prediction model high and low risk groups according to PI=2.16,the patients were divided into 21 gene detection high and low risk groups according to RS=31,and the base distribution of clinicopathological characteristics of patients in the high and low risk groups was calculated.To analyze the difference between the 6-IHC prediction model score and the RS value obtained by 21 gene detection in patients who have undergone chemotherapy and endocrine therapy and those who have only undergone endocrine therapy,and perform Kaplan-Meier analysis on the DFS and OS of the high and low risk groups of the two methods.,Draw a survival curve.Result: In the first part,after screening,we included a total of 330 hormone receptor-positive and HER2-negative early breast cancer patients who met the inclusion criteria.Among them,231 were in the training cohort and 99 were in the test cohort.There was no overlap between the two groups.The baseline characteristics of the cohort are consistent.Six prognostic-related proteins were screened out,and the weight of each protein expression was calculated to obtain a 6-IHC score prognostic model: PI =(1.2 *Cathepsin L2 expression)+(1.3 * MMP11 expression)+(1.4 * Cyclin B1 expression)+(1.3 * Aurora A expression)+(1.2 * Survivin expression)+(1.4 * Ki67 expression),and the cut-off value of high risk and low risk is 2.16.The consistency between the groups is verified by comparing the KM curve and ROC curve of the training cohort and the test cohort.The ROC curve proves the effectiveness of the 6-IHC scoring model.The KM curve indicates that the prediction model is significantly different for both OS and DFS,and can effectively evaluate hormones.The prognostic risk of recurrence,metastasis and death in patients with receptor-positive and HER2-negative early breast cancer.The results of univariate and multivariate Cox regression analysis also show that the 6-IHC prediction model can be used as an independent risk factor.In the second part,the baseline characteristics of patients with 6-IHC prediction model score or single Ki67 IHC protein expression can be obtained.The high-risk group with Ki67 ≥ 14% has a greater proportion of patients with age ≤ 60,and the 6-IHC prediction model PI ≥ The 2.16 high-risk group has a larger proportion of patients with tumor size ≥ 2cm,lymph node positive,and histological grade II-III.Among the corresponding clinicopathological factors,there are statistical differences between the high and low risk groups.Comparing the area under the 6 ROC curve AUC,the 6-IHC prediction model obtained AUC=0.754(95% CI: 0.698-0.811),Ki67 obtained AUC=0.574(95% CI: 0.505-0.643),indicating that the 6-IHC score is predicting.The prognosis of breast cancer patients is more clinically significant than Ki67 IHC expression alone.In the Kaplan-Meier analysis,the prognosis of the low-risk group with6-IHC score and Ki67 showed no statistical difference,and the patients with the high-risk group with 6-IHC score who had undergone chemotherapy and endocrine therapy were compared with those who had only received endocrine therapy.There are significant differences in recurrence and metastasis,and patients can obtain survival benefits from chemotherapy.Neither 6-IHC score nor Ki67 showed a statistical difference in the benefit of chemotherapy when the axillary lymph nodes were negative,and the p value was insignificant.In patients with positive axillary lymph nodes,the 6-IHC score and Ki67 index were high-risk groups Both are statistically significant,with 6-IHC score p=0.003 and Ki67 index p=0.042.The third part of the study can be obtained,the RS value interval distribution range is4.833 ～ 58.25 points,the median RS value is 30.85 points,the average value is32.09±1.97 points,the kurtosis coefficient is 0.60,the skewness coefficient is 0.29,and the RS value is Normal distribution.The overall reliability and effectiveness of the ROC curve are good,confirming that there is a certain correlation between the protein expression and m RNA expression of each gene of Survivin,Aurora A,Cyclin B1,Cathepsin L2,MMP11 and Ki67.The Kappa test obtained a consistency coefficient of0.772,P<0.0001,which proved that the results of the 6-IHC scoring model are in good agreement with the results of the 21 gene test,and can represent the 21 gene test to a certain extent.In the baseline patients’ clinicopathological characteristics,several factors such as tumor diameter ≥ 2 cm,lymph node positive,and Ki67 expression ≥ 14%are in the 6-IHC scoring model high-risk group(PI ≥ 2.16)than the 21 gene detection high-risk group(RS ≥ 31)has a larger proportion in the middle,indicating that the6-IHC scoring model is a good indicator of overall survival risk.The results of the 6-IHC scoring model and the Kaplan-Meier analysis of 21 gene detection in DFS and OS showed statistical differences in DFS and OS in the high-risk groups of the two methods.Conclusion: A retrospective analysis of 330 hormone receptor-positive HER2-negative breast cancer patients was performed and a risk prognostic model 6-IHC was established on the basis of immunohistochemistry.This prediction model can be used to determine the risk of recurrence and the necessity of adjuvant chemotherapy in patients with hormone receptor-positive and HER2-negative early breast cancer.Its efficacy is higher than that of Ki67 alone.The 6-IHC prediction model is also related to tumor size and lymph node metastasis..Conclusion: In this study,a 6-IHC prediction model was successfully established on the basis of immunohistochemistry and retrospective analysis of 330 hormone receptor-positive HER2-negative breast cancer patients confirmed that the prediction model can be used as an independent prognostic factor,and its efficacy is higher than that of Ki67 protein alone.The expression is consistent with the Oncotype DX results,which can represent the Oncotype DX results to a certain extent,and can be a good assessment of the prognosis risk and chemotherapy of hormone receptor-positive and HER2-negative early breast cancer patients with no limit to the number of lymph node metastases Benefit. |
PDF Full Text Request