Computer-aided evaluation of the correlation between MRI morphology and immunohistochemical biomarkers or molecular subtypes in breast cancer

Computational analysis of MRIs predicts osteosarcoma chemoresponsiveness

Aim: This study aimed to improve osteosarcoma chemoresponsiveness prediction by optimization of computational analysis of MRIs. Patients & methods: Our retrospective predictive model involved osteosarcoma patients with MRI scans performed before OsteoSa MAP neoadjuvant cytotoxic chemotherapy. Results: We found that several monofractal and multifractal algorithms were able to classify tumors according to their chemoresponsiveness. The predictive clues were defined as morphological complexity, homogeneity and fractality. The monofractal feature CV for Λ'(G) provided the best predictive association (area under the ROC curve = 0.88; p <0.001), followed by   Y-axis intersection of the regression line  for  box fractal dimension, r²  for  FD_M and tumor circularity. Conclusion: This is the first full-scale study to indicate that computational analysis of pretreatment MRIs could provide imaging biomarkers for the classification of osteosarcoma according to their chemoresponsiveness.

Monitoring Treatment Efficacy of Antiangiogenic Therapy Combined With Hypoxia-Activated Prodrugs Online Using Functional MRI

Objective

This study aimed to investigate the effectiveness of intravoxel incoherent motion (IVIM) diffusion-weighted imaging (DWI) and blood oxygen level-dependent (BOLD) magnetic resonance imaging (MRI) in monitoring tumor responses to antiangiogenic therapy combined with hypoxia-activated prodrugs (HAPs).

Materials and methods

Establishing colon cancer xenograft model by subcutaneously injecting the HCT116 cell line into BALB/C nude mice. Twenty-four tumor-bearing mice were randomly divided into four groups and injected with bevacizumab combined with TH-302 (A), bevacizumab (B), TH-302 (C), or saline (D) on days 1, 4, 7, 10 and 13. Functional MRI was performed before and at 3, 6, 9, 12 and 15 days after treatment. Pathologic examinations, including HE staining, HIF-1α and CD31 immunohistochemical staining, and TUNEL and Ki-67 immunofluorescent staining, were performed after the last scan.

Results

At the end of the study, Group A showed the lowest tumor volume, followed by Groups B, C, and D (F=120.652, P<0.001). For pathologic examinations, Group A showed the lowest percentage of CD31 staining (F=73.211, P<0.001) and Ki-67 staining (F=231.170, P<0.001), as well as the highest percentage of TUNEL staining (F=74.012, P<0.001). Moreover, the D* and f values exhibited positive correlations with CD31 (r=0.868, P<0.001, and r=0.698, P=0.012, respectively). R2* values was positively correlated with HIF-1α (r=0.776, P=0.003). D values were positively correlated with TUNEL (r=0.737, P=0.006) and negatively correlated with Ki-67 (r=0.912, P<0.001). The standard ADC values were positive correlated with TUNEL (r=0.672, P=0.017) and negative correlated with Ki-67 (r=0.873, P<0.001).

Conclusion

Anti-angiogenic agents combined with HAP can inhibit tumor growth effectively. In addition, IVIM-DWI and BOLD-MRI can be used to monitor the tumor microenvironment, including perfusion, hypoxia, cell apoptosis and proliferation, in a noninvasive manner.

Evaluating the Treatment Efficacy of Nano-Drug in a Lung Cancer Model Using Advanced Functional Magnetic Resonance Imaging

Objectives

Nano-drug delivery system is an interesting field in precise cancer treatment, but few study has reported the microenvironmental changes after such treatment. This study aimed to detect the hemodynamic and microenvironmental changes in a lung cancer xenograft model after treated with doxorubicin (DOX) encapsulated by a cyclic arginine-glycine-aspartic acid polypeptide modified poly-(lactic-co-glycolic acid) nanosystem (cRGD-PLGA@DOX) using functional magnetic resonance imaging.

Materials and Methods

Thirty-two tumor-bearing mice were randomly divided into four groups. Group A was treated with 0.9% saline, Group B with 4 mg/kg of doxorubicin, Group C with 2 mg/kg of cRGD-PLGA@DOX, and Group D with 4 mg/kg of cRGD-PLGA@DOX. Intravoxel incoherent motion diffusion-weighed imaging (IVIM-DWI) and R2^∗ mapping were performed, and D^∗, f, D, and R2^∗ values were obtained before and1, 2, and 3 weeks after treatment. They were sacrificed for pathological examination after examinations.

Results

The reconstructed cRGD-PLGA@DOX was homogeneous, well-dispersed, and spherical in shape, with an average size of 180 nm. Group D demonstrated the smallest tumor volume and highest tumor inhibition rate in 3 weeks. D value of Group B, C, and D manifested an upward trend in 3 weeks with the highest increase in Group D. D^∗ values shared a similar increased trends with f values in Group A, B, and C in 3 weeks, except Group D. R2^∗ value of Group A gradually increased in 3 weeks, but the trends were reversed in the treatment groups. D value was significantly negative with Ki-67 expression (r = -0.757, P < 0.001) but positive with TUNEL (r = 0.621, P < 0.001), and phosphate and tension homology deleted on chromosome ten (PTEN) staining (r = 0.57, P = 0.004). R2^∗ value was closely correlated with HIF-1a (r = 0.721, P < 0.001).

Conclusion

The nano-drug demonstrated an enhanced anti-tumor effect without the need of increased chemotherapeutic dosage. The tumor microenvironment such as cellular and perfusion changes during treatment can be non-invasively detected by two functional MRI including IVIM-DWI and R2^∗ mapping.

Application of IVIM-DWI in Detecting the Tumor Vasculogenic Mimicry Under Antiangiogenesis Combined With Oxaliplatin Treatment

Objectives: This study aimed to detect the time window of vascular normalization during anti-vascular treatment using intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI). Simultaneously, we evaluated the tumor invasiveness and vasculogenic mimicry and performed synthetic assessment of treatment efficacy of angiogenesis inhibitor combined with conventional chemotherapy using IVIM-DWI.

Materials and Methods: HCT116 cells were subcutaneously administered into the right flank of BALB/C nude mice to build a colon cancer xenograft model. Thirty-two tumor-bearing mice were randomly divided into four groups and intraperitoneally administered with normal saline (Group A or control group), bevacizumab (Group B), oxaliplatin monotherapy (Group C), and oxaliplatin combined with bevacizumab (Group D). The IVIM-DWI was performed on days 0, 3, 6, 9, 12, and 15 after the treatments. Another 51 tumor-bearing mice were included in the pathological examinations. α-Smooth muscle actin (SMA) and CD31 double-staining, periodic acid-Schiff (PAS) and CD31 double-staining, hematoxylin and eosin (HE), Ki-67, and E-cadherin staining were performed. The tumor growth and dynamic change of each parameter were noted.

Results: The mice in Group D manifested the smallest tumor volume and highest tumor inhibition rate. Microvessel density was significantly decreased but accompanied by increased vasculogenic mimicry after antiangiogenic treatment. The trend was reversed by oxaliplatin treatment. Treated with bevacizumab, the vessel maturity index shared a similar trend with D^* and f-values during days 3–12, which slowly increased from days 0 to 9 and then decreased briefly. D-value significantly correlated with vasculogenic mimicry and Ki-67, while D^* and f-values showed positive correlations with microvessel density and E-cadherin, an indicator of epithelial–mesenchymal transition.

Conclusion: Oxaliplatin performed an inhibited effect on vasculogenic mimicry. Bevacizumab can enhance the tumor chemotherapy through vascular normalization within a transient time period, which can be detected by IVIM-DWI. D^* and f-values are able to predict the tumor invasiveness while D is superior in reflecting vasculogenic mimicry and Ki-67 expression during antitumor treatment.

Optimal acquisition time to discriminate between breast cancer subtypes with contrast-enhanced cone-beam CT

PURPOSE

To identify the optimal acquisition time to best discriminate between benign and malignant breast lesions on contrast-enhanced cone beam CT (CE-CBCT) and evaluate the potential of CE-CBCT to differentiate between breast cancer subtypes.

MATERIAL AND METHOD

A total of 98 women with a mean age of 49±10 (SD) years (range: 29-77 years) with 100 BI-RADS 4 or 5 breast lesions were prospectively included. CE-CBCT images were obtained at 1- and 2-min after intravenous administration of iodinated contrast material. Contrast enhancement of breast lesions on CE-CBCT were evaluated and compared between different subtypes. Cut-off values for best discriminating between benign and malignant breast lesions with CE-CBCT were obtained from receiver operating characteristic curves.

RESULTS

Malignant breast lesions showed greater enhancement than benign ones at 1-min (67.28±39.79 [SD] HU vs. 42.27±40.31 [SD] HU, respectively; P=0.007) and 2-min (70.93±38.05 [SD] HU vs. 48.94±41.83 [SD] HU, respectively; P=0.016) after intravenous administration of contrast material. At 1-min after intravenous administration of contrast material, an optimal cut-off value of 54.43 HU was found to best discriminate between malignant and benign breast lesions (AUC=0.681; 95%CI: 0.558-0.805; P=0.006) yielding 69.0% sensitivity (95%CI: 56.9-79.5%) and 69.2% specificity (95% CI: 48.2-85.7%). At 2-min, an optimal cut-off value of 72.65 HU was found to best discriminate between malignant and benign breast lesions (AUC=0.654; 95%CI: 0.535-0.774; P=0.020) yielding 50.7% sensitivity (95%CI: 38.6-62.8%) and 80.8% specificity (95%CI: 60.6-93.4%). CE-CBCT helped differentiate between immunohistochemical subtypes of breast lesions with lowest enhancement for triple negative lesions. No differences in enhancement were found among histopathological subtypes lesions at 1-min (P=0.478) and 2-min (P=0.625).

CONCLUSION

CE-CBCT helps discriminate between malignant and benign breast lesions, with best capabilities obtained at 1-min after intravenous administration of contrast material. For malignant lesions, quantitative analysis of enhancement on CE-CBCT helps differentiate between immunohistochemical subtypes.

Cone-beam breast CT features associated with HER2/neu overexpression in patients with primary breast cancer

OBJECTIVES

To identify the relationship between human epidermal growth factor receptor 2 (HER2) status and cone-beam breast CT (CBBCT) characteristics in surgically resected breast cancer.

METHODS

Preoperative CBBCT of patients with BI-RADS 4 or 5 lesions identified on mammography or ultrasound and dense or very dense breast tissue were retrospectively evaluated in 181 surgically resected breast cancer (triple-negative excluded) between May 2012 and November 2014. A set of CBBCT descriptors was semiquantitatively assessed by consensus double reading. Reader reproducibility was analyzed. Multivariable logistic regression analysis using backward elimination (BEA) with the Wald criterion was performed to identify independent predictive factors of harboring HER2/neu. Principle component analysis (PCA) was used to determine characteristics that might differentiate HER2 status. Receiver operating characteristic (ROC) curve analyses were conducted to determine the predictive capability.

RESULTS

HER2 positive was found in 101 (55.8%) of 181 patients. Inter-observer agreement was high for characteristics' assessment. Based on BEA, pathologic grade, maximum dimension, lobulation, ΔCT, and calcification morphology were confirmed as independent predictive factors of HER2/neu overexpression. PCA showed that calcification- and border-related characteristics were the most important for differentiation. ROC curve analyses showed that CBBCT features (AUC = 0.853) were superior to clinicopathologic features (AUC = 0.613, p < 0.001) and comparable with combination (AUC = 0.856, p = 0.866).

CONCLUSIONS

CBBCT features could be used to prognosticate HER2 status independently, which are potentially complementary to histopathologic result and helpful in guiding biopsy.

KEY POINTS

• Dmax, lobulation, ΔCT, and calcification morphology are independent predictors of HER2 status. • CBBCT features are superior to clinicopathologic features in HER2+/- discrimination. • CBBCT features are comparable with combination with clinicopathologic features in HER2+/- discrimination.

Overview of established and emerging immunohistochemical biomarkers and their role in correlative studies in MRI

Clinical practice in radiology and pathology requires professional expertise and many years of training to visually evaluate and interpret abnormal phenotypic features in medical images and tissue sections to generate diagnoses that guide patient management and treatment. Recent advances in digital image analysis methods and machine learning have led to significant interest in extracting additional information from medical and digital whole-slide images in radiology and pathology, respectively. This has led to significant interest and research in radiomics and pathomics to correlate phenotypic features of disease with image analytics in order to identify image-based biomarkers. The expanding role of big data in radiology and pathology parallels the development and role of immunohistochemistry (IHC) in the daily practice of pathology. IHC methods were initially developed to provide additional information to help classify tumors and then transformed into an indispensable tool to guide treatment in many types of cancer. IHC markers are used in daily practice to identify specific types of cells and highlight their distributions in tissues in order to distinguish benign from neoplastic cells, determine tumor origin, subclassify neoplasms, and support and confirm diagnoses. In this regard, radiomics, pathomics, and IHC methods are very similar since they enable the extraction of image-based features to characterize various properties of diseases. Due to the dramatic advancements in recent radiomics research, we provide a brief overview of the role of established and emerging IHC biomarkers in various tumor types that have been correlated with radiologic biomarkers to improve diagnostic accuracy, predict prognosis, guide patient management, and select treatment strategies. Level of Evidence: 5 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2020;51:341-354.

Diffusion Kurtosis at 3.0T as an in vivo Imaging Marker for Breast Cancer Characterization: Correlation With Prognostic Factors

BACKGROUND

Diffusion-kurtosis imaging (DKI) has preliminarily shown promise as a relatively new MRI technique to provide useful information regarding breast lesions, but the diagnostic performance of DKI has not been fully evaluated.

PURPOSE

To compare the diagnostic accuracy of DKI, diffusion-weighted imaging (DWI), dynamic contrast-enhanced (DCE)-MRI) and proton MR spectroscopy (¹ H-MRS) in differentiating malignant from benign breast lesions independently or jointly, and explore the correlation between DKI-derived parameters and prognostic factors.

STUDY TYPE

Prospective.

SUBJECTS

Seventy-one patients with breast lesions (50 malignant, 26 benign).

SEQUENCE

DKI, DWI, DCE-MRI, and ¹ H-MRS were performed at 3.0T.

ASSESSMENT

Mean kurtosis (MK), mean diffusivity (MD), apparent diffusion coefficient (ADC), BI-RADS category, and choline peaks were analyzed by two experienced radiologists.

STATISTICAL TESTS

Student's t-test was used for continuous variables; receiver operating characteristic (ROC) analysis for assessing the diagnostic accuracy of imaging parameters; Spearman or Pearson correlations for assessing the associations between imaging parameters and prognostic factors.

RESULTS

MK exhibited higher area under the curves (AUCs) for differentiating malignant from benign lesions than did MD, ADC, DCE, and tCho (0.979 vs. 0.928, 0.911, 0.777, and 0.833, respectively, P < 0.05). MK showed a positive association with Ki-67 expression (r = 0.508) and histologic grades (r = 0.551), whereas MD and ADC were negatively correlated with Ki-67 expression (r = -0.416 and r = -0.458) and histologic grades (r = -0.411 and r = -0.319). Moreover, MK showed relatively higher AUCs compared with MD and ADC in detecting breast cancers with lymph nodal involvement, histologic grades, and Ki-67 expression.

DATA CONCLUSION

MK has higher diagnostic accuracy compared with ADC, DCE, and tCho regarding detection of breast cancer. Moreover, DKI shows promise as a quantitative imaging technique for characterizing breast lesions, highlighting the potential utility of MK as a promising imaging marker for predicting tumor aggressiveness.

LEVEL OF EVIDENCE

2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;49:845-856.