A nomogram for predicting locoregional recurrence in primary breast cancer patients who received breast-conserving surgery after neoadjuvant chemotherapy

Predictive value of MRI-based deep learning model for lymphovascular invasion status in node-negative invasive breast cancer

To retrospectively assess the effectiveness of deep learning (DL) model, based on breast magnetic resonance imaging (MRI), in predicting preoperative lymphovascular invasion (LVI) status in patients diagnosed with invasive breast cancer who have negative axillary lymph nodes (LNs). Data was gathered from 280 patients, including 148 with LVI-positive and 141 with LVI-negative lesions. These patients had undergone preoperative breast MRI and were histopathologically confirmed to have invasive breast cancer without axillary LN metastasis. The cohort was randomly split into training and validation groups in a 7:3 ratio. Radiomics features for each lesion were extracted from the first post-contrast dynamic contrast-enhanced (DCE)-MRI. The Least Absolute Shrinkage and Selection Operator (LASSO) regression method and logistic regression analyses were employed to identify significant radiomic features and clinicoradiological variables. These models were established using four machine learning (ML) algorithms and one DL algorithm. The predictive performance of the models (radiomics, clinicoradiological, and combination) was assessed through discrimination and compared using the DeLong test. Four clinicoradiological parameters and 10 radiomic features were selected by LASSO for model development. The Multilayer Perceptron (MLP) model, constructed using both radiomic and clinicoradiological features, demonstrated excellent performance in predicting LVI, achieving a high area under the curve (AUC) of 0.835 for validation. The DL model (MLP-radiomic) achieved the highest accuracy (AUC = 0.896), followed by DL model (MLP-combination) with an AUC of 0.835. Both DL models were significantly superior to the ML model (RF-clinical) with an AUC of 0.720. The DL model (MLP), which integrates radiomic features with clinicoradiological information, effectively aids in the preoperative determination of LVI status in patients with invasive breast cancer and negative axillary LNs. This is beneficial for making informed clinical decisions.

Locoregional recurrence and survival of breast-conserving surgery compared to mastectomy following neoadjuvant chemotherapy in operable breast cancer

Background

The risk of locoregional recurrence (LRR) and the long-term prognosis of breast-conserving surgery (BCS) after neoadjuvant chemotherapy (NAC) are still controversial. This study aimed to evaluate oncological outcomes for patients undergoing BCS after NAC and determine LRR and survival predictors.

Methods

This study was a retrospective cohort study of patients with locally advanced breast cancer (LABC) who received NAC and underwent BCS or mastectomy from June 2011 to November 2020. LRR, disease-free survival (DFS), and overall survival (OS) were compared in patients undergoing BCS or mastectomy. Univariate and multivariate analyses were performed to determine LRR, DFS, and OS predictors.

Results

A total of 585 patients were included, of whom 106 (18.1%) underwent BCS and 479 (81.9%) underwent a mastectomy. The LRR rate was 11.3% in the BCS group and 16.3% in the mastectomy group, revealing no significant difference(p = 0.200). In patients who underwent BCS, clinical lymph node status, histological grade and pathological complete response (pCR) were independent factors to predict LRR. There was no significant difference in DFS and OS between the BCS and the mastectomy groups. Multivariable analysis showed that lymph node status, histological grade, molecular subtypes, pCR and Miller&Payne (M&P) classification were independent predictors of DFS. Lymph node status, molecular subtypes and pCR were independent predictors of OS. BCS or mastectomy was not an independent predictor of DFS or OS.

Conclusion

Compared with mastectomy, BCS after NAC may not increase the risk of local recurrence or mortality, BCS can be performed in selected patients with small tumor size and good response to NAC.

The GLIM criteria as an effective tool for survival prediction in gastric cancer patients

BACKGROUND

The Global Leadership Initiative on Malnutrition released a new version of the malnutrition criteria (GLIM criteria). To investigate the influence of the GLIM criteria on the long-term efficacy of radical gastric cancer surgery and establish a nomogram to predict the long-term prognosis of patients with gastric cancer.

METHODS

A retrospective analysis of 1121 patients with gastric cancer in our department from 2010 to 2013 was performed. A nomogram was established to predict overall survival (OS) based on the GLIM criteria. Patients were divided into the low-risk group (LRG) and high-risk group (HRG) based on the established nomogram.

RESULTS

Multivariate Cox regression analyses showed that GLIM criteria was an independent risk factor for the 5-year OS (HR = 1.768, Cl:1.341-2.329, p < 0.001). The C index, AUC and Time-ROC of the nomogram were significantly better than that of GLIM criteria and traditional criteria. The 5-year OS of patients receiving adjuvant chemotherapy in the high-risk group was significantly higher than that of patients without chemotherapy (45.77% vs. 24.73%,p < 0.001).

CONCLUSIONS

The GLIM criteria independently influence the long-term outcome of patients after radical gastric cancer surgery. The established nomogram can predict the long-term survival of patients with gastric cancer, and postoperative adjuvant chemotherapy for HRG can significantly improve the 5-year OS of patients.

Added value of systemic inflammation markers for monitoring response to neoadjuvant chemotherapy in breast cancer patients

BACKGROUND

Complete response after neoadjuvant chemotherapy (rNACT) elevates the surgical outcomes of patients with breast cancer, however, non-rNACT have a higher risk of death and recurrence.

AIM

To establish novel machine learning (ML)-based predictive models for predicting probability of rNACT in breast cancer patients who intends to receive NACT.

METHODS

A retrospective analysis of 487 breast cancer patients who underwent mastectomy or breast-conserving surgery and axillary lymph node dissection following neoadjuvant chemotherapy at the Hubei Cancer Hospital between January 1, 2013, and October 1, 2021. The study cohort was divided into internal training and testing datasets in a 70:30 ratio for further analysis. A total of twenty-four variables were included to develop predictive models for rNACT by multiple ML-based algorithms. A feature selection approach was used to identify optimal predictive factors. These models were evaluated by the receiver operating characteristic (ROC) curve for predictive performance.

RESULTS

Analysis identified several significant differences between the rNACT and non-rNACT groups, including total cholesterol, low-density lipoprotein, neutrophil-to-lymphocyte ratio, body mass index, platelet count, albumin-to-globulin ratio, platelet-to-lymphocyte ratio, and lymphocyte-to-monocyte ratio. The areas under the curve of the six models ranged from 0.81 to 0.96. Some ML-based models performed better than models using conventional statistical methods in both ROC curves. The support vector machine (SVM) model with twelve variables introduced was identified as the best predictive model.

CONCLUSION

By incorporating pretreatment serum lipids and serum inflammation markers, it is feasible to develop ML-based models for the preoperative prediction of rNACT and therefore facilitate the choice of treatment, particularly the SVM, which can improve the prediction of rNACT in patients with breast cancer.

Lympho-vascular invasion impacts the prognosis in breast-conserving surgery: a systematic review and meta-analysis

Background

It is estimated that breast cancer (BC) incidence, especially that of early-stage breast cancer cases continues to rise due to increased universal screening. Breast-conserving surgery (BCS) is the main intervention for early-stage BC. Lympho-vascular invasion (LVI) is reported to influence breast cancer prognosis but its prognostic value in breast-conserving treatment is controversial.

Methods

A search was conducted on the Cochrane library, PubMed, Web of Science, and EMBASE from inception to December 1^st, 2021, without language restrictions, to identify studies that explored the prognosis of lympho-vascular invasion in breast-conserving surgery. Reviews of each study were conducted, and data extracted. The meta-analysis was performed with StataSE 16. Study quality assessment was evaluated using the Newcastle–Ottawa Scale.

Results

Overall, 15 studies with 21,704 patients deemed eligible for this study. Event-free survival (EFS), disease-free survival (DFS), overall survival (OS), distant metastases (DM), loco-regional recurrence (LRR), local recurrence (LR), breast recurrence (BR), disease specific survival (DSS), and breast cancer specific survival (BCSS), were extracted from each study. We found that LVI leads to poor OS (HR = 1.46, 95% CI: 1.17–1.83), DM (HR = 2.08, 95% CI: 1.66–2.60) and LR (HR = 2.00, 95% CI: 1.54–2.61).

Conclusions

We confirmed that early-stage BC patients with LVI-positive have poorer OS, DFS, LRR, BCSS, DM and LR following receiving BCS than those LVI-negative patients. Mastectomy, in combination with radical systemic therapies could be considered, especially in those requiring second surgery. How to change the impact of LVI on the local recurrence rate and long-term survival in patients who undergo BCS may be a valuable research direction in the future.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-022-09193-0.

A Practical Predictive Model Based on Ultrasound Imaging and Clinical Indices for Estimation of Response to Neoadjuvant Chemotherapy in Patients with Breast Cancer

Purpose

Clinical responses of neoadjuvant chemotherapy (NACT) are associated with prognosis in patients with breast cancer. The selection of suitable variables for the prediction of clinical responses remains controversial. Herein, we developed a predictive model based on ultrasound imaging and clinical indices to identify patients most likely to benefit from NACT.

Patients and Methods

We recruited a total of 225 consecutive patients who underwent NACT followed by surgery and axillary lymph node dissection at the Sixth Hospital of Ning Bo City of Zhe Jiang Province between January 1, 2018, and March 31, 2021. All patients had been diagnosed with breast cancer following the clinical examination. First, we created a training cohort of patients who underwent NACT+surgery (N=180) to develop a nomogram. We then validated the performance of the nomogram in a validation cohort of patients who underwent NACT+ surgery (N=45). Multivariate logistic regression was then used to identify independent risk factors that were associated with the response to NACT; these were then incorporated into the nomogram.

Results

Multivariate logistic regression analysis identified several significant differences as to clinical responses of NACT, including neutrophil–lymphocyte ratio (NLR), body mass index (BMI), pulsatility index (PI), resistance index (RI), blood flow, Ki67, histological type, molecular subtyping, and tumor size. The performance of the nomogram score exhibited a robust C-index of 0.89 (95% confidence interval [CI]: 0.83 to 0.95) in the training cohort and a high C-index of 0.87 (95% CI: 0.81 to 0.93) in the validation cohort. Clinical impact curves showed that the nomogram had a good predictive ability.

Conclusion

We successfully established an accurate and optimized nomogram incorporated ultrasound imaging and clinical indices that could be used preoperatively to predict clinical responses of NACT. This model can be used to evaluate the risk of clinical responses to NACT and therefore facilitate the choice of personalized therapy.

Locoregional therapy in breast cancer patients treated with neoadjuvant chemotherapy

Introduction: Neoadjuvant chemotherapy (NAC) is increasingly used preoperatively in breast cancer patients to achieve disease downstaging, reduce distant dissemination, and assess chemosensitivity. While NAC indications are expanding, knowledge of its impact on subsequent locoregional treatment with surgery and radiation therapy (RT) decisions is evolving. Radiation oncologists are often called upon to estimate locoregional recurrence (LRR) risks and provide recommendations for adjuvant RT to the breast/chest wall and regional lymph nodes postoperatively. In the non-NAC setting, adjuvant RT decisions are guided by the pathology findings after definitive surgery. In the NAC setting, decisions for or against adjuvant RT are complex, particularly in patients who achieve complete pathologic response (pCR).Areas covered: This review will examine contemporary data on NAC in patients with breast cancer and discuss its impact on surgical and RT decisions. We will also evaluate controversies in the role of LRRT for these patients, focussing on prognostic factors that include biological subtypes and pCR after NAC.Expert opinion: Advances in personalized medicine and diagnostic techniques have shifted paradigms and increased complexities in locoregional treatment decisions, particularly in the setting of NAC. Despite the challenges, our goals while we await prospective data remain focused on improving survival, minimizing toxicity, and optimizing function and cosmesis.

Factors affecting locoregional recurrence in breast cancer patients undergoing surgery following neoadjuvant treatment

BACKGROUND

Neoadjuvant chemotherapy (NAC) has been the standard treatment for locally advanced breast cancer for the purpose of downstaging or for conversion from mastectomy to breast conservation surgery (BCS). Locoregional recurrence (LRR) rate is still high after NAC. The aim of this study was to determine predictive factors for LRR in breast cancer patients in association with the operation types after NAC.

METHODS

Between 2005 and 2017, 1047 breast cancer patients underwent BCS or mastectomy after NAC in Chang Gung Memorial Hospital, Linkou. We obtained data regarding patient and tumor characteristics, chemotherapy regimens, clinical tumor response, tumor subtypes and pathological complete response (pCR), type of surgery, and recurrence.

RESULTS

The median follow-up time was 59.2 months (range 3.13-186.75 months). The mean initial tumor size was 4.89 cm (SD ± 2.95 cm). Of the 1047 NAC patients, 232 (22.2%) achieved pCR. The BCS and mastectomy rates were 41.3% and 58.7%, respectively. One hundred four patients developed LRR (9.9%). Comparing between patients who underwent BCS and those who underwent mastectomy revealed no significant difference in the overall LRR rate of the two groups, 8.8% in BCS group vs 10.7% in mastectomy group (p = 0.303). Multivariate analysis indicated that independent factors for the prediction of LRR included clinical N2 status, negative estrogen receptor (ER), and failure to achieve pCR. In subgroups of multivariate analysis, only negative ER was the independent factor to predict LRR in mastectomy group (p = 0.025) and hormone receptor negative/human epidermal growth factor receptor 2 positive (HR-/HER2 +) subtype (p = 0.006) was an independent factor to predict LRR in BCS patients. Further investigation according to the molecular subtype showed that following BCS, non-pCR group had significantly increased LRR compared with the pCR group, in HR-/HER2 + subtype (25.0% vs 8.3%, p = 0.037), and HR-/HER2- subtype (20.4% vs 0%, p = 0.002).

CONCLUSION

Clinical N2 status, negative ER, and failure to achieve pCR after NAC were independently related to the risk of developing LRR. Operation type did not impact on the LRR. In addition, the LRR rate was higher in non-pCR hormone receptor-negative patients undergoing BCS comparing with pCR patients.

Dynamic contrast enhanced-MRI and diffusion-weighted image as predictors of lymphovascular invasion in node-negative invasive breast cancer

BACKGROUND

Lymphovascular invasion (LVI) is an important risk factor for prognosis of breast cancer and an unfavorable prognostic factor in node-negative invasive breast cancer patients. The purpose of this study was to evaluate the association between LVI and pre-operative features of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and diffusion-weighted imaging (DWI) in node-negative invasive breast cancer.

METHODS

Data were collected retrospectively from 132 cases who had undergone pre-operative MRI and had invasive breast carcinoma confirmed on the last surgical pathology report. MRI and DWI data were analyzed for the size of tumor, mass shape, margin, internal enhancement pattern, kinetic enhancement curve, high intratumoral T2-weighted signal intensity, peritumoral edema, DWI rim sign, and apparent diffusion coefficient (ADC) values. We calculated the relationship between presence of LVI and various prognostic factors and MRI features.

RESULTS

Pathologic tumor size, mass margin, internal enhancement pattern, kinetic enhancement curve, DWI rim sign, and the difference between maximum and minimum ADC were significantly correlated with LVI (p < 0.05).

CONCLUSIONS

We suggest that DCE-MRI with DWI would assist in predicting LVI status in node-negative invasive breast cancer patients.

Prognosis and Genomic Landscape of Liver Metastasis in Patients With Breast Cancer

Objective

The prognosis of breast cancer liver metastasis (BCLM) is poor, and its molecular mechanism is unclear. We aimed to determine the factors that affect the prognosis of patients with BCLM and investigate the genomic landscape of liver metastasis (LM).

Methods

We described the prognosis of patients with BCLM and focused on prognosis prediction for these patients based on clinicopathological factors. Nomogram models were constructed for progression-free survival (PFS) and overall survival (OS) by using a cohort of 231 patients with BCLM who underwent treatment at Shandong Cancer Hospital and Institute (SCHI). We explored the molecular mechanism of LM and constructed driver genes, mutation signatures by using a targeted sequencing dataset of 217 samples of LM and 479 unpaired samples of primary breast cancer (pBC) from Memorial Sloan Kettering Cancer Center (MSKCC).

Results

The median follow-up time for 231 patients with BCLM in the SCHI cohort was 46 months. The cumulative incidence of LM at 1, 2, and 5 years was 17.5%, 45.0%, and 86.8%, respectively. The median PFS and OS were 7 months (95% CI, 6-8) and 22 months (95% CI, 19-25), respectively. The independent factors that increased the progression risk of patients with LM were Karnofsky performance status (KPS) ≤ 80, TNBC subtype, grade III, increasing trend of CA153, and disease-free interval (DFS) ≤ 1 year. Simultaneously, the independent factors that increased the mortality risk of patients with LM were Ki-67 ≥ 30%, grade III, increasing trend of CA153, pain with initial LM, diabetes, and DFI ≤ 1 year. In the MSKCC dataset, the LM driver genes were ESR1, AKT1, ERBB2, and FGFR4, and LM matched three prominent mutation signatures: APOBEC cytidine deaminase, ultraviolet exposure, and defective DNA mismatch repair.

Conclusion

This study systematically describes the survival prognosis and characteristics of LM from the clinicopathological factors to the genetic level. These results not only enable clinicians to assess the risk of disease progression in patients with BCLM to optimize treatment options, but also help us better understand the underlying mechanisms of tumor metastasis and evolution and provide new therapeutic targets with potential benefits for drug-resistant patients.

Risk Factors for Ipsilateral Breast Tumor Recurrence in Triple-Negative or HER2-Positive Breast Cancer Patients Who Achieve a Pathologic Complete Response After Neoadjuvant Chemotherapy

BACKGROUND

Attention has been focused on attempts to eliminate breast surgery for breast cancer patients who achieve a pathologic complete response after neoadjuvant chemotherapy (NAC). However, there are few data on ipsilateral breast tumor recurrence (IBTR) among patients with triple-negative or epidermal growth factor receptor 2-positive (HER2+) tumors who achieve a pathologic complete response after NAC and breast-conserving treatment.

METHODS

Using a multi-institutional retrospective database, this study evaluated the risk factors for IBTR among patients with newly diagnosed stages 1 to 3 breast cancer involving triple-negative or HER2+ tumors who achieved ypT0 after NAC and breast-conserving treatment.

RESULTS

During a median follow-up period of 4.8 years (range, 0.1-15.5 years), the 5-year IBTR-free survival rate was 95.5%. The breast cancer subtype was not associated with IBTR-free survival. Patients younger than 40 years at diagnosis had significantly worse IBTR-free survival than those who were 40 years of age or older (5-year IBTR-free survival, 87.7 vs 96.9%; p = 0.002).

CONCLUSIONS

This retrospective study demonstrated that age at diagnosis was independently associated with IBTR-free survival. Special caution is needed when clinical trials analyzing omission of breast surgery after NAC are enrolling younger patients (UMIN-CTR No. UMIN000037067).

Triple-negative breast cancer: Pretreatment magnetic resonance imaging features and clinicopathological factors associated with recurrence

PURPOSE

We aimed to investigate the magnetic resonance imaging (MRI) features and clinicopathologic factors with recurrence of triple-negative breast cancer (TNBC).

PATIENTS AND METHODS

We identified 281 patients with 288 surgically confirmed TNBC lesions who underwent pretreatment MRI between 2009 and 2015. The presence of intratumoral high signal on T2-weighted images, high-signal rim on diffusion-weighted images (DWI), and rim enhancement on the dynamic contrast-enhanced MRI and clinicopathological data were collected. Cox proportional analysis was performed.

RESULTS

Of the 288 lesions, 36 (12.5%) recurred after a median follow-up of 18 months (range, 3.6-68.3 months). Rim enhancement (hazard ratio [HR] = 3.15; 95% confidence interval [CI] = 1.01, 9.88; p = .048), and lymphovascular invasion (HR = 2.73, 95% CI = 1.20, 6.23; p = .016) were independently associated with disease recurrence. While fibroglandular volume, background parenchymal enhancement, intratumoral T2 high signal, and high-signal rim on DWI, were not found to be risk factors for recurrence.

CONCLUSION

Pretreatment MRI features may help predict a high risk of recurrence in patients with TNBC.

Identifying the Prognosis Factors and Predicting the Survival Probability in Patients with Non-Metastatic Chondrosarcoma from the SEER Database

OBJECTIVE

To identify prognostic factors and establish nomograms for predicting overall survival (OS) and cause specific survival (CSS) of patients with non-metastatic chondrosarcoma.

METHODS

We collected information on patients with non-metastatic chondrosarcoma from the Surveillance, Epidemiology, and End Results (SEER) database between 2005 and 2014, together with data from the First Affiliated Hospital of Zhengzhou University from 2011 to 2016. Variables including patients' baseline demographics (age, race, and gender), tumor characteristics (tumor size and extension, histology subtype, primary site, and American Joint Committee on Cancer [AJCC] stage), therapy (surgery, chemotherapy, and radiotherapy), and socioeconomic status (SES) were extracted for further analysis. OS and CSS were retrieved as our researching endpoints. Patients from the database were regarded as the training set, and univariate analysis, Lasso regression and multivariate analysis as well as the random forest were used to explore the predictors and establish nomograms. To validate nomograms internally and externally, we applied bootstrapped validation internally with the training dataset, while the dataset for external validation was obtained from the First Affiliated Hospital of Zhengzhou University. We estimated the discriminative ability of nomograms based on Cox proportional hazard regression models by means of calibration curves and the concordance index (C-index) of internal and external validation.

RESULTS

After the implementation of exclusion criteria, there were 1267 patients in the training set and 72 patients in the testing set with non-metastatic chondrosarcomas. Age, gender, grade, histological subtype, primary site, surgery, radiation, chemotherapy, being employed/unemployed, tumor size, and tumor extension were significantly associated with prognosis in the univariate analysis. Age, gender, tumor size and extension, primary site, surgery, radiotherapy, chemotherapy, histological grade, and subtype were independent prognostic factors in the Cox models. The C-index of nomograms (internal: OS, 0.787; CSS, 0.821; external: OS, 0.777; CSS, 0.821) were higher than following conventional systems: AJCC sixth (OS, 0.640; CSS, 0.673) and seventh edition (OS, 0.675; CSS, 0.711).

CONCLUSIONS

Age, gender, tumor size and extension, surgery, histological grade, and subtype were independent prognostic factors for both OS and CSS. In addition, we revealed that chondrosarcomas in the trunk, radiotherapy, and chemotherapy were correlated with poor prognosis. Our nomograms based on significant clinicopathologic features can well predict the 3-year and 5-year survival probability of patients with non-metastatic chondrosarcoma and assist oncologists in making accurate survival evaluation.

Facilitating validation of prediction models: a comparison of manual and semi-automated validation using registry-based data of breast cancer patients in the Netherlands

Background

Clinical prediction models are not routinely validated. To facilitate validation procedures, the online Evidencio platform (https://www.evidencio.com) has developed a tool partly automating this process. This study aims to determine whether semi-automated validation can reliably substitute manual validation.

Methods

Four different models used in breast cancer care were selected: CancerMath, INFLUENCE, Predicted Probability of Axillary Metastasis, and PREDICT v.2.0. Data were obtained from the Netherlands Cancer Registry according to the inclusion criteria of the original development population. Calibration (intercepts and slopes) and discrimination (area under the curve (AUC)) were compared between semi-automated and manual validation.

Results

Differences between intercepts and slopes of all models using semi-automated validation ranged from 0 to 0.03 from manual validation, which was not clinically relevant. AUCs were identical for both validation methods.

Conclusions

This easy to use semi-automated validation option is a good substitute for manual validation and might increase the number of validations of prediction models used in clinical practice. In addition, the validation tool was considered to be user-friendly and to save a lot of time compared to manual validation. Semi-automated validation will contribute to more accurate outcome predictions and treatment recommendations in the target population.

Electronic supplementary material

The online version of this article (10.1186/s12874-019-0761-5) contains supplementary material, which is available to authorized users.

Preoperative prediction of lymphovascular invasion in invasive breast cancer with dynamic contrast-enhanced-MRI-based radiomics

BACKGROUND

Lymphovascular invasion (LVI) status facilitates the selection of optimal therapeutic strategy for breast cancer patients, but in clinical practice LVI status is determined in pathological specimens after resection.

PURPOSE

To explore the use of dynamic contrast-enhanced (DCE)-magnetic resonance imaging (MRI)-based radiomics for preoperative prediction of LVI in invasive breast cancer.

STUDY TYPE

Prospective.

POPULATION

Ninety training cohort patients (22 LVI-positive and 68 LVI-negative) and 59 validation cohort patients (22 LVI-positive and 37 LVI-negative) were enrolled.

FIELD STRENGTH/SEQUENCE

1.5 T and 3.0 T, T₁ -weighted DCE-MRI.

ASSESSMENT

Axillary lymph node (ALN) status for each patient was evaluated based on MR images (defined as MRI ALN status), and DCE semiquantitative parameters of lesions were calculated. Radiomic features were extracted from the first postcontrast DCE-MRI. A radiomics signature was constructed in the training cohort with 10-fold cross-validation. The independent risk factors for LVI were identified and prediction models for LVI were developed. Their prediction performances and clinical usefulness were evaluated in the validation cohort.

STATISTICAL TESTS

Mann-Whitney U-test, chi-square test, kappa statistics, least absolute shrinkage and selection operator (LASSO) regression, logistic regression, receiver operating characteristic (ROC) analysis, DeLong test, and decision curve analysis (DCA).

RESULTS

Two radiomic features were selected to construct the radiomics signature. MRI ALN status (odds ratio, 10.452; P < 0.001) and the radiomics signature (odds ratio, 2.895; P = 0.031) were identified as independent risk factors for LVI. The value of the area under the curve (AUC) for a model combining both (0.763) was higher than that for MRI ALN status alone (0.665; P = 0.029) and similar to that for the radiomics signature (0.752; P = 0.857). DCA showed that the combined model added more net benefit than either feature alone.

DATA CONCLUSION

The DCE-MRI-based radiomics signature in combination with MRI ALN status was effective in predicting the LVI status of patients with invasive breast cancer before surgery.

LEVEL OF EVIDENCE

1 Technical Efficacy Stage: 2 J. Magn. Reson. Imaging 2019;50:847-857.

Magnetic Resonance Imaging Combined With Second-look Ultrasonography in Predicting Pathologic Complete Response After Neoadjuvant Chemotherapy in Primary Breast Cancer Patients

BACKGROUND

Magnetic resonance imaging (MRI) or ultrasonography (US) alone is limited in the ability to predict the pathologic complete response (pCR) accurately after neoadjuvant chemotherapy (NAC). The aim of the present study was to predict the pCR using MRI combined with second-look US in primary breast cancer patients.

MATERIALS AND METHODS

A total of 1274 consecutive primary breast cancer patients who were examined by MRI and second-look US before and after NAC and had undergone breast-conserving surgery from 2004 to 2014 were included. The positive predictive value (PPV) of a clinical complete response (cCR) by MRI alone and MRI plus US was assessed. A CR was defined as no residual invasive carcinoma. The presence of a residual in situ component was also assessed (ypTis).

RESULTS

Of the 1274 patients, 333 (26.1%) had a pCR (ypT0/is), and 102 (8.0%) had a residual in situ component (ypTis). A cCR was found in 247 patients (19.4%) using MRI alone and in 182 patients (14.3%) using MRI plus US. The PPV for a cCR using MRI alone was 79.4% and the PPV for MRI plus US was 86.8%. The PPV for a cCR by MRI plus US was the greatest at 98.1% in the estrogen receptor-negative (ER^-)/human epidermal growth factor receptor-positive (HER2⁺) group (86.5% in the ER⁺/HER2⁺, 83.0% in the ER^-/HER2^-, and 64.7% in the ER⁺/HER2^- groups). The PPV for residual in situ component was as low as 72.2%.

CONCLUSION

Our results have shown that MRI combined with second-look US in predicting for a pCR was useful compared with MRI alone, especially for ER^-/HER2⁺. However, it was difficult to predict for the presence of a residual in situ component. Our ongoing prospective multi-institutional study has shown that adding vacuum-assisted biopsy to MRI plus second-look US is warranted to improve the prediction of pCR for omitting breast surgery.

Neoadjuvant Chemotherapy and Surgery for Breast Cancer: Preoperative MRI Features Associated with Local Recurrence

Purpose To investigate the MRI and clinical-pathologic features associated with local-regional recurrence (LRR) in patients who had undergone breast-conserving surgery (BCS) following neoadjuvant chemotherapy (NAC). Materials and Methods In this retrospective, single-institution study between October 2003 and September 2015, 548 consecutive women, consisting of 468 down-staged and 80 preplanned BCS patients (mean age, 45.7 years; range, 22-75 years), underwent preoperative MRI and BCS following NAC. The rate and site of LRR, preoperative MRI features including Breast Imaging Reporting and Data System lexicon, and clinical-pathologic features (age, stage, tumor subtype, histologic grade, lymphovascular invasion, adjuvant chemotherapy, and endocrine therapy) were analyzed with the Cox proportional hazards model to identify independent factors associated with LRR-free survival (LRFS). Results Of the 548 women, 23 (4.2%) had LRR at a median follow-up of 23.1 months. In Cox regression analysis, younger age (ie, ≤ 40 years) (hazard ratio = 2.932; 95% confidence interval [CI]: 1.233, 6.969; P = .015) or the presence of nonmass enhancement on preoperative MR images (hazard ratio = 3.220; 95% CI: 1.274, 8.140; P = .014) was associated with worse LRFS. LRR was more frequently observed in the same quadrant as the original tumor in the down-staged BCS group than in the preplanned BCS group (80.0% [16 of 20] vs 33.3% [one of three]; P = .021). Conclusion Age of 40 years or younger or the presence of nonmass enhancement on preoperative MR images tends to be associated with worse local-regional recurrence-free survival, and local-regional recurrence frequently occurs in the same quadrant as the original tumor in breast cancer patients who undergo breast-conserving surgery following neoadjuvant chemotherapy. © RSNA, 2018.

A Nomogram for Distinction and Potential Prediction of Liver Metastasis in Breast Cancer Patients

Liver metastasis from breast cancer has poor prognosis. We aimed at developing a reliable tool for making a distinction and prediction for liver metastasis in breast cancer patients, thus helping clinical diagnosis and treatment. In this study, totally 6238 patients from SEER database with known distant metastasis status and clinicopathologic variables were enrolled and divided randomly into training and validating groups. Logistic regression was used to screen variables and a nomogram was constructed. After multivariate logistic regression, sex, histology type, N stage, grade, age, ER, PR, HER2 status as significant variables for constructing the nomogram. The nomogram for distinguishing and predicting liver metastasis in breast cancer passed the calibration and validation steps and the areas under the receiver operating characteristic curve of the training set and the validation set were 0.6602 and 0.6511 respectively. Our nomogram is a reliable and robust tool for the distinction and prediction of liver metastasis in breast cancer patients, thus helping better choose medical examinations and optimize therapeutic regimen under the cooperation among medical oncologists and surgeons.

Discrepancies Between Pathological Tumor Responses and Estimations of Complete Response by Magnetic Resonance Imaging After Neoadjuvant Chemotherapy Differ by Breast Cancer Subtype

INTRODUCTION

The influence of breast cancer (BC) subtype in discrepancies between pathologic complete response (pCR) and complete response by magnetic resonance imaging (MRI-CR) after neoadjuvant chemotherapy (NAC) have not been discussed well. We evaluated the association between BC subtype and pCR or only residual in situ lesion without invasive cancer (pCR/in situ⁺) in patients with MRI-CR (positive predictive value [PPV]).

MATERIAL AND METHODS

From the data of 716 patients with primary BC who were diagnosed with invasive cancer and treated with NAC and then surgery from January 2009 to May 2014 at St. Luke's International Hospital, 180 patients were determined to have MRI-CR by retrospective chart review. BC subtypes at baseline were classified into 6 subtypes, as strong estrogen receptor (ER⁺⁺), moderately positive ER (ER⁺), negative ER (ER^-), and HER2 status expression.

RESULTS

Three subtypes had PPV (pCR) ≥ 50%: ER^-/HER2⁺ (56.3%, 27/48), ER^-/HER2^- (57.6%, 34/59), and ER⁺/HER2⁺ (56.2%, 9/16). However, PPV (pCR) for the ER⁺⁺/HER2^- and ER⁺⁺/HER2⁺ subtypes was < 30%; notably, only 12.0% (3/25) for the ER⁺⁺/HER2^- subtype, which was significantly low (P < .001) compared with ER⁺⁺/HER2^- and other subtypes. PPV (pCR/in situ⁺) was significantly low at 20.0% in the ER⁺⁺/HER2^- subtype (P < .001 compared with other subtypes). PPV (pCR/in situ⁺) in other subtypes was collectively greater than 60%, and was 91.7% in the ER^-/HER2⁺ subtype.

CONCLUSION

We should interpret carefully MRI-CR of NAC to evaluate residual disease for ER⁺⁺/HER2^- BC.

Preoperative MRI features associated with lymphovascular invasion in node-negative invasive breast cancer: A propensity-matched analysis

PURPOSE

In node-negative disease, the presence of lymphovascular invasion (LVI) is reported to be an unfavorable prognostic factor. Thus, the aim of this study was to evaluate whether preoperative breast MRI features are associated with LVI in patients with node-negative invasive breast cancer by a propensity-matched analysis.

MATERIALS AND METHODS

Among 389 patients with node-negative invasive ductal breast cancer who had preoperative breast 3.0 Tesla MRI with precontrast T2-weighted fat-suppressed, pre- and dynamic postcontrast T1-weighted fat-suppressed sequences, 61 patients with LVI (LVI group) were matched with 183 patients without LVI (no LVI group) at a ratio of 1:3 in terms of age, histologic grade, tumor size, and hormone receptor status. Two radiologists reviewed the MRI features, following profiles of focal breast edema (peritumoral, prepectoral, subcutaneous), intratumoral T2 signal intensity, adjacent vessel sign, and increased ipsilateral whole-breast vascularity, in addition to 2013 Breast Imaging Reporting and Data System lexicon.

RESULTS

The presence of peritumoral edema (45.9% [28/61] versus 30.6% [56/183], P = 0.030) and adjacent vessel sign (82.0% [50/61] versus 68.3% [125/183], P = 0.041) was significantly associated with LVI. Prepectoral edema was also more frequently observed in the LVI group than in the no LVI group with borderline significance (26.2% [16/61] versus 15.3% [28/183], P = 0.055). In cases of nonmass enhancement, regional enhancement was more frequently found in the LVI group than in the no LVI group (60.0% [3/4] versus 5.9% [1/4], P = 0.042).

CONCLUSION

Preoperative breast MRI features may be associated with LVI in patients with node-negative invasive breast cancer.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2017;46:1037-1044.

Nomogram for predicting recurrence in stage II colorectal cancer

BACKGROUND

There have been no established criteria to distinguish patients at high and low risk of recurrence in stage II colorectal cancer. Several risk factors have been identified but recurrence could not be fully predicted by each factor alone. This retrospective study sought to develop a nomogram for accurate prediction of recurrence in stage II colorectal cancer.

MATERIAL AND METHODS

We reviewed the data for 4167 patients with stage II colorectal cancer who underwent surgery between January 1997 and December 2006. The risk factors for recurrence were identified, and a nomogram for recurrence was created using the factors. The performance of the nomogram was assessed with a bootstrapped-concordance index and calibration plots.

RESULTS

Sex, carcinoembryonic antigen, tumor location, tumor depth, lymphatic invasion, venous invasion and number of lymph nodes studied were significantly associated with recurrence. A nomogram for five-year freedom from recurrence was created with these factors. The bootstrapped-concordance index of the nomogram was 0.64, and it was well calibrated.

CONCLUSIONS

Our nomogram can be a useful tool for accurate prediction of recurrence in stage II colorectal cancer.

Peritumoral apparent diffusion coefficients for prediction of lymphovascular invasion in clinically node-negative invasive breast cancer

OBJECTIVES

To evaluate whether visual assessment of T2-weighted imaging (T2WI) or an apparent diffusion coefficient (ADC) could predict lymphovascular invasion (LVI) status in cases with clinically node-negative invasive breast cancer.

MATERIALS AND METHODS

One hundred and thirty-six patients with 136 lesions underwent MRI. Visual assessment of T2WI, tumour-ADC, peritumoral maximum-ADC and the peritumour-tumour ADC ratio (the ratio between them) were compared with LVI status of surgical specimens.

RESULTS

No significant relationship was found between LVI and T2WI. Tumour-ADC was significantly lower in the LVI-positive (n = 77, 896 ± 148 × 10(-6) mm(2)/s) than the LVI-negative group (n = 59, 1002 ± 163 × 10(-6) mm(2)/s; p < 0.0001). Peritumoral maximum-ADC was significantly higher in the LVI-positive (1805 ± 355 × 10(-6) mm(2)/s) than the LVI-negative group (1625 ± 346 × 10(-6) mm(2)/s; p = 0.0003). Peritumour-tumour ADC ratio was significantly higher in the LVI-positive (2.05 ± 0.46) than the LVI-negative group (1.65 ± 0.40; p < 0.0001). Receiver operating characteristic curve analysis revealed that the area under the curve (AUC) of the peritumour-tumour ADC ratio was the highest (0.81). The most effective threshold for the peritumour-tumour ADC ratio was 1.84, and the sensitivity, specificity, positive predictive value and negative predictive value were 77% (59/77), 76% (45/59), 81% (59/73) and 71% (45/63), respectively.

CONCLUSIONS

We suggest that the peritumour-tumour ADC ratio can assist in predicting LVI status on preoperative imaging.

KEY POINTS

• Tumour ADC was significantly lower in LVI-positive than LVI-negative breast cancer. • Peritumoral maximum-ADC was significantly higher in LVI-positive than LVI-negative breast cancer. • Peritumour-tumour ADC ratio was significantly higher in LVI-positive breast cancer. • Diagnostic performance of the peritumour-tumour ADC ratio was highest for positive LVI. • Peritumour-tumour ADC ratio showed higher diagnostic ability in postmenopausal than premenopausal patients.

High expression of post-treatment Ki-67 status is a risk factor for locoregional recurrence following breast-conserving surgery after neoadjuvant chemotherapy

INTRODUCTION

During the past decade, neoadjuvant chemotherapy (NAC) has been increasingly used in patients to reduce large tumors to a size eligible for breast-conserving therapy (BCT). However, the association between NAC and Ki-67 has not yet been fully elucidated. The aim of this study was to evaluate the prognostic significance of Ki-67, specifically after NAC followed by BCT, particularly in terms of locoregional recurrence (LRR).

METHODS

A total 217 patients who received BCT after NAC were retrospectively analyzed. In these patients, immunohistochemistry analyses defined four tumor subtypes, Luminal A, Luminal B, Triple negative, and HER2 type. Ki-67 was examined by immunohistochemistry in both pretreatment core needle samples and post-treatment surgical excision specimens. High Ki-67 expression was defined as >20%. The prognostic factors LRR, locoregional relapse-free survival (LRRFS), and overall survival (OS) were analyzed.

RESULTS

In total, LRR developed in 14 patients, and the 5 year-LRRFS was 94.2%. Post-treatment high Ki-67 expression, triple negative, the presence of lymphovascular invasion, and histological grade 3 were significantly high in LRR for prognostic factors (P < 0.05). There were significant differences in Kaplan-Meier method for LRRFS curves according to these three factors for patients receiving BCT following NAC (P < 0.05). There was a significant difference in the 5 year-OS for patients with and without LRR (41.7% vs. 93.9%, P < 0.001).

CONCLUSION

Post-treatment high Ki-67 expression could be one of the important prognostic factors of LRR, and require careful follow-up on LRR at the time of surveillance.