Ultrasound and immunohistochemical predictors of neoadjuvant chemotherapy response in breast cancer

Introduction

Breast cancer is the most prevalent malignant tumor among women and remains a major public health concern worldwide (1). In 2020 alone, an estimated 2.1 million women were diagnosed with the disease, representing roughly one-quarter of all cancer cases in females (2). Immunohistochemical markers—including the proliferation index (Ki-67), estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2)—are integral to both the diagnosis and management of breast cancer, offering critical guidance for clinical decision-making (3,4).

Neoadjuvant chemotherapy (NAC), administered as a systemic treatment before surgery, plays a pivotal role in the management of patients with locally advanced or inoperable breast cancer (5). Achieving a pathologic complete response (pCR) after NAC has been consistently associated with improved long-term outcomes (6,7). However, the effectiveness of NAC varies considerably among patients, with reported pCR rates ranging from 6% to 26% (8). This variability underscores the need for accurate assessment of chemotherapy efficacy and early identification of patients most likely to benefit.

Although histopathological examination remains the gold standard for evaluating NAC response, its invasive nature and delayed turnaround limit its suitability for routine or repeated assessment. By contrast, ultrasound offers a non-invasive, widely accessible, and highly reproducible tool that has shown value in the initial diagnosis, treatment monitoring, and response evaluation of breast cancer (9,10). Several studies have explored whether ultrasound features can predict pCR in breast cancer (10-12). However, most have reported inconsistent results, failed to incorporate immunohistochemical marker data, or lacked rigorous validation of their predictive models. Additionally, research on the correlation between ultrasound features and immunohistochemical markers is still scarce.

In this study, we aim to investigate the correlation between pre- and post-NAC ultrasound features and immunohistochemical markers in breast cancer, and to evaluate their combined predictive value for treatment response. Our goal is to provide a more comprehensive and precise basis to support clinical decision-making and optimize therapeutic strategies for breast cancer. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-942/rc).

Methods

Subjects

This study included 101 female breast cancer patients who received NAC at The Second Hospital of Lanzhou University between October 2020 and December 2023. Eligibility criteria were: (I) breast cancer confirmed by preoperative biopsy with clinical staging indicating NAC; (II) no prior breast treatment before ultrasound examination; and (III) availability of routine breast ultrasound performed both before and after chemotherapy, along with complete imaging and pathological data. Exclusion criteria included: (I) presence of other malignancies; and (II) severe systemic diseases. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The Second Hospital of Lanzhou University (No. 2022A-256) and informed consent was taken from all the patients.

All patients received between four and eight 21-day cycles of neoadjuvant chemotherapy, primarily using regimens based on anthracyclines, taxanes, and alkylating agents. Patients with HER2-positive tumors were treated with anti-HER2 agents such as trastuzumab and pertuzumab, in accordance with current neoadjuvant treatment guidelines. pCR was defined as the complete absence of invasive cancer cells in both breast tissue and axillary lymph nodes following therapy.

Data collection

General clinical data were collected for all patients, including age, menopausal status, and key immunohistochemical markers: Ki-67, a proliferation index indicating tumor growth rate; ER and PR, which reflect hormone receptor status; and HER2, a protein associated with aggressive tumor behavior and targeted therapies. Two experienced ultrasound physicians evaluated and recorded lesion characteristics, such as location, largest cross-sectional size, morphology, margins, and echogenicity. Lesions were classified according to the Breast Imaging Reporting and Data System (BI-RADS), a standardized system used to assess breast imaging findings and estimate malignancy risk. All images were saved and reviewed by senior physicians to ensure consistent interpretation.

Following the 2017 St. Gallen International Breast Cancer Conference Expert Consensus (13), breast cancers were classified into four pathological subtypes: (I) Luminal A: ER(+), PR(+), HER2(−/+), Ki-67 <14%; (II) Luminal B: ER(+), PR(+), HER2(−/+), Ki-67 ≥14%; or ER(+), PR(+), HER2(+++), any level of Ki-67; (III) HER2 overexpression: ER(−), PR(−), HER2(+++); (IV) triple-negative: ER(−), PR(−), HER2(−), and if HER2(++) is present, FISH testing is performed.

Ultrasound examination method

Ultrasound evaluations were conducted at baseline (prior to NAC initiation) and after completion of the full NAC course but before surgery. The timing of ultrasound assessments aligned with routine clinical practice and aimed to monitor tumor response and guide subsequent treatment decisions.

The patients were placed in a supine position with both hands behind the head to fully expose the breast and axillary regions. A Siemens ACUSON Sequoia ultrasound diagnostic system, equipped with a 5–12 MHz high-frequency linear array transducer, was used for routine ultrasound scanning. Scanning was performed using radial, transverse, longitudinal, and stacking methods, starting from the nipple and extending outward.

Immunohistochemical examination method

Interpretation of Ki-67 positive expression (13): Ki-67 >14% is defined as high expression, and Ki-67 ≤14% as low expression. The threshold for positive ER and PR detection is 1%. HER2 overexpression (3+) or gene amplification is considered positive.

Outcome measures

The evaluation criteria for breast cancer treatment efficacy can be determined according to the Miller-Payne grading system for pathological response. According to this system, MP4 and MP5 grades are defined as pathologically effective, while MP1 to MP3 grades are considered pathologically ineffective (14).

Statistical analysis

Statistical analysis was performed using R 4.3.2. Categorical data were presented as counts (percentages) and compared using chi-squared tests for 2×2 tables. Variables that showed statistical significance in univariate analysis were considered for inclusion in the multivariate binary logistic regression model. Additionally, clinical relevance based on prior breast cancer research and expert judgment was taken into account to ensure that selected variables were both statistically robust and clinically meaningful. These selected variables were then used to construct a predictive nomogram for NAC efficacy.

Model performance was assessed using the receiver operating characteristic (ROC) curve and area under the ROC curve (AUC) to evaluate discriminative ability, and calibration plots to assess agreement between predicted and observed outcomes.

To identify the optimal cutoff probability for clinical decision-making, the maximum Youden index (sensitivity + specificity − 1) was calculated from the ROC curve. Sensitivity, specificity, and corresponding cutoff values were reported. A P value of less than 0.05 was considered statistically significant.

Results

Patients’ baseline characteristics and ultrasound features

This study included 101 breast cancer patients; all diagnosed with invasive ductal carcinoma confirmed by pathology. The baseline data showed that 58.4% of patients were aged 50 years or older, and 64.4% were postmenopausal. Immunohistochemical analysis revealed that 67.3% of patients had high Ki-67 expression, 67.3% were ER-positive, and 53.5% were PR-negative. Molecular subtyping classified 34.7% as HER2-positive and 33.7% as luminal B (Table 1). Ultrasound evaluation showed that most tumors were located in the upper outer quadrant (51.5%), measured less than 3 cm in diameter (58.4%), and demonstrated shrinkage after chemotherapy (90.1%). The majority were solitary (81.2%), hypoechoic (86.1%), had spiculated margins (59.4%), and showed little or no blood flow on Doppler imaging (66.3%). More than half of the patients (52.5%) were categorized as BI-RADS 5 or below (Table 2). Notably, PR status was significantly associated with ultrasound BI-RADS classification (P=0.03), with PR-positive tumors tending to have higher BI-RADS scores (Table 3). Figure 1 presents an example ultrasound image from a PR-positive patient.

Figure 1 Ultrasound features of a PR-positive breast cancer patient. (A) The lesion in the left breast appears hypoechoic with well-defined margins and an irregular, spiculated (“crab-foot”) shape. The internal echo pattern is heterogeneous, with scattered punctate hyperechoic spots. (B) Color Doppler imaging demonstrates dotted internal blood flow signals within the lesion. PR, progesterone receptor.

Pathological treatment effect

In this study, pathological response was assessed using the MP grading system. Grades MP4 and MP5 were considered indicative of a pathological response, while grades MP1 to MP3 were classified as non-responsive.

Out of the 101 patients, 56 (55.45%) achieved a pCR following NAC, representing the effective group. The remaining 45 patients (44.55%) were categorized as non-responders.

Comparison of baseline characteristics and ultrasound features between pathologically effective and ineffective patients

Comparison of baseline characteristics, immunohistochemical markers, and ultrasound features between the effective and ineffective groups revealed significant differences in Ki-67 expression (P=0.04), ER (P=0.008), PR (P=0.008), pathological type (P=0.005), tumor shrinkage after chemotherapy (P=0.04), and tumor number (single tumor vs. multiple tumors) (P=0.03) (Table 4). Specifically, patients with high Ki-67 expression, ER positivity, PR negativity, HER2-positive pathological type, tumor shrinkage after chemotherapy, and single tumor were more likely to achieve an effective response to NAC (Tables 4,5).

Logistic regression model results

Before constructing the multivariate logistic regression model, variables showing statistical significance in univariate analysis and those with established clinical relevance were selected to ensure a robust and meaningful predictive model. We included baseline characteristics, immunohistochemical results, and ultrasound features in the logistic regression model, which showed significant differences between patients with different treatment outcomes. The analysis revealed that patients with negative PR status [odds ratio (OR): 0.18, P=0.01], a single lesion (OR: 0.26, P=0.04), and a decrease in lesion size after chemotherapy (OR: 8.16, P=0.02) were more likely to respond favorably to NAC (Table 6). Based on these findings, a nomogram was constructed to predict the pathological response to NAC, providing individualized estimates of treatment effectiveness (Figure 2). For instance, the predicted probability of NAC effectiveness for the patient illustrated in Figure 1—who was ER-, PR-, and Ki-67-positive, with tumor shrinkage and a solitary lesion post-chemotherapy—was 63%.

Figure 2 Nomogram for predicting pathological response to neoadjuvant chemotherapy in breast cancer patients. The nomogram integrates key variables including tumor number (Number), tumor shrinkage status after chemotherapy (Lesion has shrunk), pathological subtype (Type), and immunohistochemical markers: PR, ER, and Ki-67 expression levels. Each factor contributes to a total score that estimates the probability of achieving a favorable pathological response. *, P<0.05. ER, estrogen receptor; HER2, human epidermal growth factor receptor 2; PR, progesterone receptor.

Model stability evaluation

The calibration curve (Figure 3) demonstrates close agreement between the predicted probabilities and the observed outcomes, aligning well with both the theoretical curve and the ideal reference line. The ROC curve (Figure 4) further validates the model’s performance, with an AUC of 0.80 [95% confidence interval (CI): 0.72–0.80]. At the optimal cutoff point, the model achieved a sensitivity of 0.70 and a specificity of 0.76, corresponding to the maximum Youden index of 0.46, indicating robust and reliable predictive accuracy.

Figure 3 Calibration curve for the pathological response prediction model. “Apparent” represents the internal calibration curve; “Bias-corrected” indicates the calibration curve adjusted for optimism via bootstrap validation; “Ideal” denotes the theoretical perfect calibration line. The close alignment of the curves demonstrates good agreement between predicted and observed outcomes.

Figure 4 ROC curve of the pathological response prediction model. The AUC demonstrates good discriminative ability. Sensitivity and specificity at the optimal cutoff point indicate robust performance in predicting pathological response. AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic.

A lower threshold would improve sensitivity to detect potential responders but may result in more false positives, which could be acceptable for patients with a strong desire for breast conservation to avoid missing effective treatment opportunities. In contrast, a higher threshold, such as the one corresponding to the maximum Youden index (1.0), would improve specificity but may lead to more missed responders.

Discussion

This study is based on a retrospective analysis of 101 breast cancer patients who underwent neoadjuvant chemotherapy. We integrated ultrasound features with immunohistochemical markers to investigate differences between effective and ineffective treatment response groups. Our analysis revealed a significant association between PR expression and BI-RADS classification (P=0.03), with PR-positive patients tending to exhibit higher BI-RADS categories. PR positivity is generally linked to lower tumor aggressiveness and better prognosis, highlighting an important imaging-pathology correlation.

While many previous studies have explored associations between breast cancer molecular subtypes and imaging features, few have focused specifically on the relationship between PR expression and BI-RADS classification (15,16). Our study fills this gap by demonstrating this significant correlation in a cohort predominantly composed of postmenopausal patients undergoing neoadjuvant chemotherapy. The high proportion of postmenopausal patients in our study may affect PR expression and imaging features, which could explain differences from studies with more balanced menopausal groups. These findings underscore the importance of considering menopausal status when evaluating the interplay between immunohistochemical markers and imaging features, and suggest that further stratified studies are warranted to validate and extend these observations.

Additionally, although one study reported inconsistent findings on the PR expression-BI-RADS correlation—and some investigations found no statistically significant relationship (17)—this discrepancy may stem from differences in patient populations and inclusion criteria. Notably, this study did not clearly stratify breast cancer molecular subtypes, which are known to affect imaging features.

We also found that PR-negative patients exhibited better efficacy from NAC, aligning with the results reported by Nakhlis et al. (18). We confirmed that breast cancer patients with a single lesion, PR-negative, and a reduction in the maximum tumor diameter following chemotherapy had more favorable NAC outcomes compared to the ineffective NAC group (MP1–MP3 stages). These findings are further supported by clinical observations, which suggest that NAC can effectively reduce tumor size and lower disease staging (19).

Breast cancer is a disease that relies on angiogenesis for tumor growth and progression (20). NAC can disrupt tumor angiogenesis, leading to tumor necrosis and fibrosis, which effectively suppresses tumor growth and progression. NAC promotes the closure of microvessels surrounding the tumor, inducing apoptosis in tumor cells and contributing to tumor shrinkage. Consequently, changes in the maximum tumor diameter after NAC treatment can serve as a reliable indicator of chemotherapy efficacy. In this study, we confirmed that ultrasound measurements of the maximum tumor diameter before and after NAC treatment can be used to assess NAC efficacy in breast cancer patients. This has also been consistently supported by various studies (21,22). Furthermore, the visual nomination map that we developed demonstrated effective performance in predicting the pathological efficacy of NAC in breast cancer.

Overall, this study analyzed the relationship between ultrasound features and immunohistochemical results, and developed a predictive model for the pathological efficacy of NAC in breast cancer by integrating both characteristics. Compared with existing studies, our work offers several advantages: (I) we examined the correlation between ultrasound features and immunohistochemical markers, enabling a more comprehensive assessment of treatment response; and (II) the model that we developed relies on readily available indicators, making it practical for routine clinical application.

Nevertheless, this study has several limitations. First, this was a retrospective study based on historical medical records. Some important information—such as details of adverse events during chemotherapy and patient adherence—was either incomplete or not recorded in a standardized way. For example, the measurement of maximum tumor diameter relied on ultrasound reports. Slight differences in measurement practices among sonographers could lead to small errors, which might in turn influence the analysis of the relationship between tumor shrinkage and treatment efficacy. Second, the decision to administer NAC and the choice of regimen were made according to the treating clinician’s prior judgment. This could introduce implicit bias, such as a tendency to prescribe more intensive chemotherapy regimens for patients with PR-negative tumors. Finally, the relatively small sample size and the absence of external validation of our predictive model may limit the generalizability of our findings. Future research should address these limitations by expanding the sample size, conducting multicenter prospective studies, and applying standardized imaging and measurement protocols. Additionally, incorporating more detailed histopathological and imaging analyses will help refine and validate the predictive value of models in assessing NAC response.

Conclusions

In summary, this study investigated the predictive value of preoperative ultrasound characteristics and immunohistochemical markers for the efficacy of NAC in breast cancer patients. Our findings demonstrated that a combined assessment of specific indicators—including PR negativity, single-lesion status, and post-chemotherapy tumor size reduction—significantly predicts favorable NAC outcomes. Among these, PR negativity emerged as a robust independent predictor, consistent with previous studies linking it to enhanced chemotherapy responsiveness.

Ultrasound features, such as BI-RADS classification and tumor morphology (e.g., solitary lesions and evidence of shrinkage following chemotherapy), further complemented immunohistochemical markers in refining predictive accuracy. The constructed nomogram, which integrates these key factors, exhibited good discriminative ability (AUC =0.80) and calibration. An optimal decision threshold of 0.46, determined by the Youden index, was identified to support clinical decision-making: patients with predicted probabilities above this threshold may be prioritized for NAC to maximize therapeutic benefit.

This study reinforces the value of combining ultrasound and immunohistochemical parameters, addressing limitations of previous research that often relied on single-modality predictors. Nonetheless, certain limitations—such as the single-center, retrospective design and potential selection bias—may restrict the generalizability of the findings. Future multi-center, prospective studies with larger cohorts and standardized protocols are warranted to validate these results. Overall, our findings provide a practical and comprehensive tool to predict NAC response, facilitating individualized treatment planning and improving outcomes in breast cancer management.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-942/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-942/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-942/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-942/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The Second Hospital of Lanzhou University (No. 2022A-256) and informed consent was taken from all the patients.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Wilkinson L, Gathani T. Understanding breast cancer as a global health concern. Br J Radiol 2022;95:20211033. [Crossref] [PubMed]
2. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
3. Ueno T. Biomarkers of neoadjuvant/adjuvant endocrine therapy for ER-positive/HER2-negative breast cancer. Chin Clin Oncol 2020;9:35. [Crossref] [PubMed]
4. Yan J, Liu XL, Han LZ, et al. Relation between Ki-67, ER, PR, Her2/neu, p21, EGFR, and TOP II-α expression in invasive ductal breast cancer patients and correlations with prognosis. Asian Pac J Cancer Prev 2015;16:823-9. [Crossref] [PubMed]
5. Bartsch R, Bergen E, Galid A. Current concepts and future directions in neoadjuvant chemotherapy of breast cancer. Memo 2018;11:199-203. [Crossref] [PubMed]
6. Cortazar P, Zhang L, Untch M, et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. Lancet 2014;384:164-72. [Crossref] [PubMed]
7. Asaoka M, Narui K, Suganuma N, et al. Clinical and pathological predictors of recurrence in breast cancer patients achieving pathological complete response to neoadjuvant chemotherapy. Eur J Surg Oncol 2019;45:2289-94. [Crossref] [PubMed]
8. Zou Y, Xue M, Hossain MI, et al. Ultrasound and diffuse optical tomography-transformer model for assessing pathological complete response to neoadjuvant chemotherapy in breast cancer. J Biomed Opt 2024;29:076007. [Crossref] [PubMed]
9. Tadesse GF, Tegaw EM, Abdisa EK. Diagnostic performance of mammography and ultrasound in breast cancer: a systematic review and meta-analysis. J Ultrasound 2023;26:355-67. [Crossref] [PubMed]
10. Dan Q, Zheng T, Liu L, et al. Ultrasound for Breast Cancer Screening in Resource-Limited Settings: Current Practice and Future Directions. Cancers (Basel) 2023;15:2112. [Crossref] [PubMed]
11. Clevert DA, Beyer G, Nieß H, et al. Ultrasound—New Techniques Are Extending the Applications. Dtsch Arztebl Int 2023;120:41-7. [Crossref] [PubMed]
12. Wang S, Hossack JA, Klibanov AL. From Anatomy to Functional and Molecular Biomarker Imaging and Therapy: Ultrasound Is Safe, Ultrafast, Portable, and Inexpensive. Invest Radiol 2020;55:559-72. [Crossref] [PubMed]
13. Untch M, Huober J, Jackisch C, et al. Initial Treatment of Patients with Primary Breast Cancer: Evidence, Controversies, Consensus: Spectrum of Opinion of German Specialists at the 15th International St. Gallen Breast Cancer Conference (Vienna 2017). Geburtshilfe Frauenheilkd 2017;77:633-44. [Crossref] [PubMed]
14. Hagenaars SC, de Groot S, Cohen D, et al. Tumor-stroma ratio is associated with Miller-Payne score and pathological response to neoadjuvant chemotherapy in HER2-negative early breast cancer. Int J Cancer 2021;149:1181-8. [Crossref] [PubMed]
15. Cen D, Xu L, Li N, et al. BI-RADS 3-5 microcalcifications can preoperatively predict breast cancer HER2 and Luminal a molecular subtype. Oncotarget 2017;8:13855-62. [Crossref] [PubMed]
16. Zhan D, Ren Y, Huang Y, et al. Relationship between degree of breast MRI background parenchymal enhancement and breast cancer subtype. Zhong Nan Da Xue Xue Bao Yi Xue Ban 2020;45:1291-7. [Crossref] [PubMed]
17. He T, Shi T, Luo W, et al. The diagnostic value of BI-RADS grade 3 to 5 for breast masses is correlated with the expression of estrogen receptor, progesterone receptor, human epidermal growth factor receptor-2. Medicine (Baltimore) 2023;102:e33208. [Crossref] [PubMed]
18. Nakhlis F, Regan MM, Warren LE, et al. The Impact of Residual Disease After Preoperative Systemic Therapy on Clinical Outcomes in Patients with Inflammatory Breast Cancer. Ann Surg Oncol 2017;24:2563-9. [Crossref] [PubMed]
19. Elsayed B, Alksas A, Shehata M, et al. Exploring Neoadjuvant Chemotherapy, Predictive Models, Radiomic, and Pathological Markers in Breast Cancer: A Comprehensive Review. Cancers (Basel) 2023;15:5288. [Crossref] [PubMed]
20. Mouron S, Bueno MJ, Muñoz M, et al. Monitoring vascular normalization: new opportunities for mitochondrial inhibitors in breast cancer. Oncoscience 2021;8:1-13. [Crossref] [PubMed]
21. Lim HF, Sharma A, Gallagher C, et al. Value of ultrasound in assessing response to neoadjuvant chemotherapy in breast cancer. Clin Radiol 2023;78:912-8. [Crossref] [PubMed]
22. Derouane F, van Marcke C, Berlière M, et al. Predictive Biomarkers of Response to Neoadjuvant Chemotherapy in Breast Cancer: Current and Future Perspectives for Precision Medicine. Cancers (Basel) 2022;14:3876. [Crossref] [PubMed]