A nomogram based on clinicopathological features and serological indicators predicting breast pathologic complete response of neoadjuvant chemotherapy in breast cancer

Evaluation of prognostic risk factors of triple-negative breast cancer with 18F-FDG PET/CT parameters, clinical pathological features and biochemical indicators

Introduction

Breast cancer is a heterogeneous disease comprising various molecular subtypes, including Luminal A, Luminal B, human epidermal growth factor receptor-2 (HER2) positive, and triple negative types, each with distinct biological characteristics and behaviors. Triple negative breast cancer (TNBC) remains a particularly challenging subtype worldwide. Our study aims to evaluate whether Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography (18F-FDG PET/CT) parameters, clinical pathological features, and biochemical indicators serve as prognostic risk factors for TNBC. Additionally, we explore correlations between biochemical indicators and 18F-FDG PET/CT parameters.

Methods

We conducted a retrospective analysis of 95 TNBC patients who underwent preoperative 18F-FDG PET/CT examinations at Tianjin Medical University Cancer Institute and Hospital from 2013 to 2018. Collected data included 18F-FDG PET/CT parameters, clinical and pathological features, and biochemical indicators. We used Kaplan-Meier survival analysis and multivariate Cox regression analysis to evaluate associations between 18F-FDG PET/CT parameters/biochemical indicators and disease free survival (DFS)/overall survival (OS). The log-rank test determined significant differences in survival curves, and the Spearman correlation coefficient analyzed correlations between quantitative variables. Visualization and analysis were performed using R packages.

Results

Among 95 TNBC patients, mean standardized uptake value (SUVmean) was significantly correlated with DFS. Fasting blood glucose (FBG), α- L-fucosylase (AFU) and Creatine kinase (CK) were independent predictors of DFS, while Precursor albumin (PALB) and CK were independent predictors of OS. FBG showed correlations with SUVpeak and SUVmean, and CK was correlated with peak standardized uptake value (SUVpeak). Our results indicated that 18F-FDG PET/CT parameters and biochemical indicators may constitute a new prognostic model for TNBC patients post-surgery.

Discussion

We found that SUVmean, FBG, AFU and CK are predictive factors for DFS in TNBC patients post-surgery, while PALB and CK are predictive factors for OS, which prompts us to pay more attention to these indicators in clinical practice. Also 18F-FDG PET/CT parameters and biochemical indicators have potential utility in constituting a new prognostic model for TNBC patients post-surgery.

Development and Validation of an Interpretable Machine Learning Prediction Model for Total Pathological Complete Response after Neoadjuvant Chemotherapy in Locally Advanced Breast Cancer: Multicenter Retrospective Analysis

Objective: This study aims to develop an interpretable machine learning (ML) model to accurately predict the probability of achieving total pathological complete response (tpCR) in patients with locally advanced breast cancer (LABC) following neoadjuvant chemotherapy (NAC). Methods: This multi-center retrospective study included pre-NAC clinical pathology data from 698 LABC patients. Post-operative pathological outcomes divided patients into tpCR and non-tpCR groups. Data from 586 patients at Shanghai Ruijin Hospital were randomly assigned to a training set (80%) and a test set (20%). In comparison, data from our hospital's remaining 112 patients were used for external validation. Variable selection was performed using the Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis. Predictive models were constructed using six ML algorithms: decision trees, K-nearest neighbors (KNN), support vector machine, light gradient boosting machine, and extreme gradient boosting. Model efficacy was assessed through various metrics, including receiver operating characteristic (ROC) curves, precision-recall (PR) curves, confusion matrices, calibration plots, and decision curve analysis (DCA). The best-performing model was selected by comparing the performance of different algorithms. Moreover, variable relevance was ranked using the SHapley Additive exPlanations (SHAP) technique to improve the interpretability of the model and solve the "black box" problem. Results: A total of 191 patients (32.59%) achieved tpCR following NAC. Through LASSO regression analysis, five variables were identified as predictive factors for model construction, including tumor size, Ki-67, molecular subtype, targeted therapy, and chemotherapy regimen. The KNN model outperformed the other five classifier algorithms, achieving area under the curve (AUC) values of 0.847 (95% CI: 0.809-0.883) in the training set, 0.763 (95% CI: 0.670-0.856) in the test set, and 0.665 (95% CI: 0.555-0.776) in the external validation set. DCA demonstrated that the KNN model yielded the highest net advantage through a wide range of threshold probabilities in both the training and test sets. Furthermore, the analysis of the KNN model utilizing SHAP technology demonstrated that targeted therapy is the most crucial factor in predicting tpCR. Conclusion: An ML prediction model using clinical and pathological data collected before NAC was developed and verified. This model accurately predicted the probability of achieving a tpCR in patients with LABC after receiving NAC. SHAP technology enhanced the interpretability of the model and assisted in clinical decision-making and therapy optimization.

TP53 Gene Polymorphism at Codon 72 as a Response Predictor for Neoadjuvant Chemotherapy

Introduction

Breast cancer is the most prevalent cancer in women worldwide, and neoadjuvant chemotherapy is a favored method for achieving pathologic complete response (pCR). The TP53 gene is involved in inducing the response to chemotherapy drugs.

Objectives

The present study sought to correlate polymorphism variants at codon 72 with pCR to neoadjuvant chemotherapy.

Casuistry and Methods

The study was conducted in the state of Sergipe, in northeastern Brazil. A total of 206 patients with a histopathological diagnosis of breast cancer who underwent neoadjuvant chemotherapy from 2019 to 2022 were included. DNA samples were collected for the evaluation of TP53 polymorphism at codon 72. A prospective evaluation of the cases was conducted to verify the surgical pathologic response after chemotherapy; the Response Evaluation Criteria in Solid Tumors (RECIST) were used. The study was approved by the University of São Paulo Ethics and Research Committee.

Results

Of the 168 patients, 44.6% were Arg72Arg, 17.3% were Pro72Pro, and 38.0% were Arg72Pro; pCR was achieved in 21.4% of the patients; 10.1% had progressive disease, 13.7% had stable disease, and 54.2% had a partial pathologic response. The only predictor of pCR in multivariate regression was immunohistochemistry (p < 0.001). In the multivariate analysis, Arg72Pro and Pro72Pro increased the odds of the patient evolving with stable disease. This study was innovative in demonstrating a predictor of stable disease in response to neoadjuvant chemotherapy.

Conclusion

TP53 polymorphism at codon 72 is not a predictor of pCR, but it can be a predictor of stable disease.

Systematic Review of Nomograms Used for Predicting Pathological Complete Response in Early Breast Cancer

Pathological complete response (pCR) is an important surrogate outcome to assess the effects of neoadjuvant chemotherapy (NAC). Nomograms to predict pCR have been developed with local data to better select patients who are likely to benefit from NAC; however, they were never critically reviewed regarding their internal and external validity. The purpose of this systematic review was to critically appraise nomograms published in the last 20 years (2010-2022). Articles about nomograms were searched in databases, such as PubMed/MEDLINE, Embase and Cochrane. A total of 1120 hits were found, and seven studies were included for analyses. No meta-analysis could be performed due to heterogeneous reports on outcomes, including the definition of pCR and subtypes. Most nomograms were developed in Asian centers, and nonrandomized retrospective cohorts were the most common sources of data. The most common subtype included in the studies was triple negative (50%). There were articles that included HER2+ (>80%). In one study, scholars performed additional validation of the nomogram using DFS and OS as outcomes; however, there was a lack of clarity on how such endpoints were measured. Nomograms to predict pCR cannot be extrapolated to other settings due to local preferences/availability of NAC. The main gaps identified in this review are also opportunities for future nomogram research and development.

Predictive miRNAs Patterns in Blood of Breast Cancer Patients Demonstrating Resistance Towards Neoadjuvant Chemotherapy

Objective

The effect of chemotherapy in patients with breast cancer (BC) is uncertain. This study attempted to analyze serum microRNAs (miRNAs) in NAC resistant and sensitive BC patients and develop a miRNA-based nomogram model. To further help clinicians make treatment decisions for hormone receptor-positive patients.

Methods

A total of 110 BC patients with NAC were recruited and assigned in sensitive and resistant group, and 4 sensitive patients and 3 resistant patients were subjected to high-throughput sequencing. The functions of their target genes were analyzed by GO and KEGG. Five BC-related reported miRNAs were selected for expression pattern measurement by RT-qPCR and multivariate logistic analysis. The nomogram model was developed using R 4.0.1, and its predictive efficacy, consistency and clinical application value in development and validation groups were evaluated using ROC, calibration and decision curves.

Results

There were 44 differentially-expressed miRNAs in resistant BC patients. miR-3646, miR-4741, miR-6730-3p, miR-6831-5p and miR-8485 were candidate for resistance diagnosis in BC. Logistic multiple regression analysis showed that miR-4741 (or = 0.30, 95% CI = 0.08-0.63, P = 0.02) and miR-6831-5p (or = 0.48, 95% CI = 0.24-0.78, P = 0.01) were protective factors of BC resistance. The ROC curves showed a sensitivity of 0.884 and 0.750 for miR-4741 and miR-6831-5P as markers of resistance, suggesting that they can be used as independent risk factors for BC resistance. The other 3 miRNAs can be used as calibration factors to establish the risk prediction model of resistance in BC. In risk model, the prediction accuracy of resistance of BC is about 78%. 5-miRNA signature diagnostic models can help clinicians provide personalized treatment for NAC resistance BC patients to improve patient survival.

Conclusion

MiR-4741 and miR-6831-5p are independent risk factors for breast cancer resistance. This study constructed a nomogram model of NAC resistance in BC based on 5 differentially-expressed serum miRNAs.

Tumor volume reduction after induction chemotherapy with gemcitabine plus cisplatin in nasopharyngeal carcinoma

OBJECTIVES

To evaluate the tumor volume reduction after induction chemotherapy (IC) with gemcitabine plus cisplatin (GP) and to build prediction models for tumor volume reduction in nasopharyngeal carcinoma (NPC).

METHODS

NPC patients who received GP IC were retrospectively enrolled. The gross tumor volume of the nasopharynx and lymph nodes (GTVnx and GTVnd) were contoured before and after IC. Univariate and multivariate analyses were performed to identify associated factors. Nomogram models were constructed to predict the possibility of tumor volume reduction.

RESULTS

A total of 192 patients were enrolled. The mean relative volume reduction for GTVnx and GTVnd was 29.66% and 31.75%, respectively. The volume reduction of GTVnx and GTVnd had a weak association (r = 0.229, p < 0.001). For GTVnx volume reduction, pre-treatment neutrophil count (p = 0.043), lymphocyte count (p = 0.026), LDH level (p = 0.005), and BMI (p = 0.020) were independently associated factors. For GTVnd volume reduction, pre-treatment EBV-DNA (p = 0.029), GTVnd volume (p < 0.001), eosinophil count (p = 0.043), NLR (p = 0.039), LDH level (p = 0.026), and serum potassium level (p = 0.027) were independently associated factors. For the GTVnx nomogram model, areas under the receiver-operating characteristic curve (AUC) were 0.702 and 0.698 for the training and validation cohorts, respectively. For the GTVnd nomogram model, the AUC was 0.872 and 0.758 for the training and validation cohorts, respectively.

CONCLUSIONS

Tumor volumes reduce significantly after GP induction chemotherapy. Nomogram models for predicting the possibility of tumor volume reduction are built.