Magnetic resonance imaging and molecular features associated with tumor-infiltrating lymphocytes in breast cancer

Radiological Features for Predicting the Status of CD8-Positive Lymphocytes in HER2 Positive Breast Cancer

Background

The level of tumor-infiltrating lymphocytes (TILs) in human epidermal growth factor receptor type 2 (HER2)-positive breast cancer (BC) is positively correlated with pathological complete response.

Aims

To investigate the relationship between ultrasound (US) and magnetic resonance imaging (MRI) features and the level of CD8-positive TILs (CD8+-TILs) in patients with HER2-positive BC.

Study Design

Retrospective cohort study.

Methods

This retrospective study included 155 consecutive women with HER2-positive BC. Patients were divided into two groups: CD8+-TILlow (< 35%) and CD8+-TILhigh (≥ 35%) groups. US and MRI features were evaluated using the BI-RADS lexicon, and the apparent diffusion coefficient (ADC) value was calculated using RadiAnt software. Univariate and multivariate analyses revealed the optimal US and MRI features for predicting CD8+-TIL levels. Receiver operating characteristic analysis and the Delong test were used to compare the diagnostic performance of US and MRI features. Furthermore, implementing a nomogram will increase clinical utility.

Results

Univariate analysis of US features showed significant differences in shape, orientation, and posterior echo between the two groups; however, there were no significant differences in margins, internal echo, and microcalcification. Multifactorial analysis revealed that shape, orientation, and posterior echo were independent risk factors, with odds ratios of 11.62, 2.70, and 0.16, respectively. In terms of MRI features, ADC was an independent predictor of CD8+-TIL levels. These three US features and the ADC performed well, with area under the curve (AUC) values of 0.802 and 0.705, respectively. The combination of US and ADC values had higher predictive efficacy (AUC = 0.888) than either US or ADC alone (p = 0.009, US_ADC vs. US; p < 0.001, US_ADC vs. ADC).

Conclusion

US features (shape, orientation, and posterior echo) and ADC value may be a valuable tool for estimating CD8+-TIL levels in HER2-positive BC. The nomogram may help clinicians in making decisions.

Comprehensive assessment of immune context and immunotherapy response via noninvasive imaging in gastric cancer

BACKGROUNDThe tumor immune microenvironment can provide prognostic and therapeutic information. We aimed to develop noninvasive imaging biomarkers from computed tomography (CT) for comprehensive evaluation of immune context and investigate their associations with prognosis and immunotherapy response in gastric cancer (GC).METHODSThis study involved 2,600 patients with GC from 9 independent cohorts. We developed and validated 2 CT imaging biomarkers (lymphoid radiomics score [LRS] and myeloid radiomics score [MRS]) for evaluating the IHC-derived lymphoid and myeloid immune context respectively, and integrated them into a combined imaging biomarker [LRS/MRS: low(-) or high(+)] with 4 radiomics immune subtypes: 1 (-/-), 2 (+/-), 3 (-/+), and 4 (+/+). We further evaluated the imaging biomarkers' predictive values on prognosis and immunotherapy response.RESULTSThe developed imaging biomarkers (LRS and MRS) had a high accuracy in predicting lymphoid (AUC range: 0.765-0.773) and myeloid (AUC range: 0.736-0.750) immune context. Further, similar to the IHC-derived immune context, 2 imaging biomarkers (HR range: 0.240-0.761 for LRS; 1.301-4.012 for MRS) and the combined biomarker were independent predictors for disease-free and overall survival in the training and all validation cohorts (all P < 0.05). Additionally, patients with high LRS or low MRS may benefit more from immunotherapy (P < 0.001). Further, a highly heterogeneous outcome on objective response ​rate was observed in 4 imaging subtypes: 1 (-/-) with 27.3%, 2 (+/-) with 53.3%, 3 (-/+) with 10.2%, and 4 (+/+) with 30.0% (P < 0.0001).CONCLUSIONThe noninvasive imaging biomarkers could accurately evaluate the immune context and provide information regarding prognosis and immunotherapy for GC.

Magnetic Resonance Imaging Features Associated with a High and Low Expression of Tumor-Infiltrating Lymphocytes: A Stratified Analysis According to Molecular Subtypes

A total of 457 patients, including 241 HR+/HER2- patients, 134 HER2+ patients, and 82 TN patients, were studied. The percentage of TILs in the stroma adjacent to the tumor cells was assessed using a 10% cutoff. The low TIL percentages were 82% in the HR+ patients, 63% in the HER2+ patients, and 56% in the TN patients (p < 0.001). MRI features such as morphology as mass or non-mass enhancement (NME), shape, margin, internal enhancement, presence of peritumoral edema, and the DCE kinetic pattern were assessed. Tumor sizes were smaller in the HR+/HER2- group (p < 0.001); HER2+ was more likely to present as NME (p = 0.031); homogeneous enhancement was mostly seen in HR+ (p < 0.001); and the peritumoral edema was present in 45% HR+, 71% HER2+, and 80% TN (p < 0.001). In each subtype, the MR features between the high- vs. low-TIL groups were compared. In HR+/HER2-, peritumoral edema was more likely to be present in those with high TILs (70%) than in those with low TILs (40%, p < 0.001). In TN, those with high TILs were more likely to present a regular shape (33%) than those with low TILs (13%, p = 0.029) and more likely to present the circumscribed margin (19%) than those with low TILs (2%, p = 0.009).

Radiogenomic analysis of cellular tumor-stroma heterogeneity as a prognostic predictor in breast cancer

BACKGROUND

The tumor microenvironment and intercellular communication between solid tumors and the surrounding stroma play crucial roles in cancer initiation, progression, and prognosis. Radiomics provides clinically relevant information from radiological images; however, its biological implications in uncovering tumor pathophysiology driven by cellular heterogeneity between the tumor and stroma are largely unknown. We aimed to identify radiogenomic signatures of cellular tumor-stroma heterogeneity (TSH) to improve breast cancer management and prognosis analysis.

METHODS

This retrospective multicohort study included five datasets. Cell subpopulations were estimated using bulk gene expression data, and the relative difference in cell subpopulations between the tumor and stroma was used as a biomarker to categorize patients into good- and poor-survival groups. A radiogenomic signature-based model utilizing dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) was developed to target TSH, and its clinical significance in relation to survival outcomes was independently validated.

RESULTS

The final cohorts of 1330 women were included for cellular TSH biomarker identification (n = 112, mean age, 57.3 years ± 14.6) and validation (n = 886, mean age, 58.9 years ± 13.1), radiogenomic signature of TSH identification (n = 91, mean age, 55.5 years ± 11.4), and prognostic (n = 241) assessments. The cytotoxic lymphocyte biomarker differentiated patients into good- and poor-survival groups (p < 0.0001) and was independently validated (p = 0.014). The good survival group exhibited denser cell interconnections. The radiogenomic signature of TSH was identified and showed a positive association with overall survival (p = 0.038) and recurrence-free survival (p = 3 × 10-4).

CONCLUSION

Radiogenomic signatures provide insights into prognostic factors that reflect the imbalanced tumor-stroma environment, thereby presenting breast cancer-specific biological implications and prognostic significance.

Exploring the added value of pretherapeutic MR descriptors in predicting breast cancer pathologic complete response to neoadjuvant chemotherapy

OBJECTIVES

To evaluate the association between pretreatment MRI descriptors and breast cancer (BC) pathological complete response (pCR) to neoadjuvant chemotherapy (NAC).

MATERIALS AND METHODS

Patients with BC treated by NAC with a breast MRI between 2016 and 2020 were included in this retrospective observational single-center study. MR studies were described using the standardized BI-RADS and breast edema score on T2-weighted MRI. Univariable and multivariable logistic regression analyses were performed to assess variables association with pCR according to residual cancer burden. Random forest classifiers were trained to predict pCR on a random split including 70% of the database and were validated on the remaining cases.

RESULTS

Among 129 BC, 59 (46%) achieved pCR after NAC (luminal (n = 7/37, 19%), triple negative (n = 30/55, 55%), HER2 + (n = 22/37, 59%)). Clinical and biological items associated with pCR were BC subtype (p < 0.001), T stage 0/I/II (p = 0.008), higher Ki67 (p = 0.005), and higher tumor-infiltrating lymphocytes levels (p = 0.016). Univariate analysis showed that the following MRI features, oval or round shape (p = 0.047), unifocality (p = 0.026), non-spiculated margins (p = 0.018), no associated non-mass enhancement (p = 0.024), and a lower MRI size (p = 0.031), were significantly associated with pCR. Unifocality and non-spiculated margins remained independently associated with pCR at multivariable analysis. Adding significant MRI features to clinicobiological variables in random forest classifiers significantly increased sensitivity (0.67 versus 0.62), specificity (0.69 versus 0.67), and precision (0.71 versus 0.67) for pCR prediction.

CONCLUSION

Non-spiculated margins and unifocality are independently associated with pCR and can increase models performance to predict BC response to NAC.

CLINICAL RELEVANCE STATEMENT

A multimodal approach integrating pretreatment MRI features with clinicobiological predictors, including tumor-infiltrating lymphocytes, could be employed to develop machine learning models for identifying patients at risk of non-response. This may enable consideration of alternative therapeutic strategies to optimize treatment outcomes.

KEY POINTS

• Unifocality and non-spiculated margins are independently associated with pCR at multivariable logistic regression analysis. • Breast edema score is associated with MR tumor size and TIL expression, not only in TN BC as previously reported, but also in luminal BC. • Adding significant MRI features to clinicobiological variables in machine learning classifiers significantly increased sensitivity, specificity, and precision for pCR prediction.

Evaluation of Breast Cancer Tumor-Infiltrating Lymphocytes on Ultrasound Images Based on a Novel Multi-Cascade Residual U-Shaped Network

OBJECTIVE

Breast cancer has become the leading cancer of the 21st century. Tumor-infiltrating lymphocytes (TILs) have emerged as effective biomarkers for predicting treatment response and prognosis in breast cancer. The work described here was aimed at designing a novel deep learning network to assess the levels of TILs in breast ultrasound images.

METHODS

We propose the Multi-Cascade Residual U-Shaped Network (MCRUNet), which incorporates a gray feature enhancement (GFE) module for image reconstruction and normalization to achieve data synergy. Additionally, multiple residual U-shaped (RSU) modules are cascaded as the backbone network to maximize the fusion of global and local features, with a focus on the tumor's location and surrounding regions. The development of MCRUNet is based on data from two hospitals and uses a publicly available ultrasound data set for transfer learning.

RESULTS

MCRUNet exhibits excellent performance in assessing TILs levels, achieving an area under the receiver operating characteristic curve of 0.8931, an accuracy of 85.71%, a sensitivity of 83.33%, a specificity of 88.64% and an F1 score of 86.54% in the test group. It outperforms six state-of-the-art networks in terms of performance.

CONCLUSION

The MCRUNet network based on breast ultrasound images of breast cancer patients holds promise for non-invasively predicting TILs levels and aiding personalized treatment decisions.

Associations of Multiparametric Breast MRI Features, Tumor-Infiltrating Lymphocytes, and Immune Gene Signature Scores Following a Single Dose of Trastuzumab in HER2-Positive Early-Stage Breast Cancer

Dynamic biomarkers that permit the real-time monitoring of the tumor microenvironment response to therapy are an unmet need in breast cancer. Breast magnetic resonance imaging (MRI) has demonstrated value as a predictor of pathologic complete response and may reflect immune cell changes in the tumor microenvironment. The purpose of this pilot study was to investigate the value of breast MRI features as early markers of treatment-induced immune response. Fourteen patients with early HER2+ breast cancer were enrolled in a window-of-opportunity study where a single dose of trastuzumab was administered and both tissue and MRIs were obtained at the pre- and post-treatment stages. Functional diffusion-weighted and dynamic contrast-enhanced MRI tumor measures were compared with tumor-infiltrating lymphocytes (TILs) and RNA immune signature scores. Both the pre-treatment apparent diffusion coefficient (ADC) and the change in peak percent enhancement (DPE) were associated with increased tumor-infiltrating lymphocytes with trastuzumab therapy (r = -0.67 and -0.69, p < 0.01 and p < 0.01, respectively). Low pre-treatment ADC and a greater decrease in PE in response to treatment were also associated with immune-activated tumor microenvironments as defined by RNA immune signatures. Breast MRI features hold promise as biomarkers of early immune response to treatment in HER2+ breast cancer.

Non-invasive tumor microenvironment evaluation and treatment response prediction in gastric cancer using deep learning radiomics

The tumor microenvironment (TME) plays a critical role in disease progression and is a key determinant of therapeutic response in cancer patients. Here, we propose a noninvasive approach to predict the TME status from radiological images by combining radiomics and deep learning analyses. Using multi-institution cohorts of 2,686 patients with gastric cancer, we show that the radiological model accurately predicted the TME status and is an independent prognostic factor beyond clinicopathologic variables. The model further predicts the benefit from adjuvant chemotherapy for patients with localized disease. In patients treated with checkpoint blockade immunotherapy, the model predicts clinical response and further improves predictive accuracy when combined with existing biomarkers. Our approach enables noninvasive assessment of the TME, which opens the door for longitudinal monitoring and tracking response to cancer therapy. Given the routine use of radiologic imaging in oncology, our approach can be extended to many other solid tumor types.

Association between radiomics features of DCE-MRI and CD8+ and CD4+ TILs in advanced gastric cancer

Objective: The aim of this investigation was to explore the correlation between the levels of tumor-infiltrating CD8⁺ and CD4⁺ T cells and the quantitative pharmacokinetic parameters of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in patients with advanced gastric cancer. Methods: We retrospectively analyzed the data of 103 patients with histopathologically confirmed advanced gastric cancer (AGC). Three pharmacokinetic parameters, K_ep, K^trans, and V_e, and their radiomics characteristics were obtained by Omni Kinetics software. Immunohistochemical staining was used to determine CD4⁺ and CD8⁺ TILs. Statistical analysis was subsequently performed to assess the correlation between radiomics characteristics and CD4⁺ and CD8⁺ TIL density. Results: All patients included in this study were finally divided into either a CD8⁺ TILs low-density group (n = 51) (CD8⁺ TILs < 138) or a high-density group (n = 52) (CD8⁺ TILs ≥ 138), and a CD4⁺ TILs low-density group (n = 51) (CD4⁺ TILs < 87) or a high-density group (n = 52) (CD4⁺ TILs ≥ 87). ClusterShade and Skewness based on K_ep and Skewness based on K^trans both showed moderate negative correlation with CD8⁺ TIL levels (r = 0.630-0.349, p < 0.001), with ClusterShade based on K_ep having the highest negative correlation (r = -0.630, p < 0.001). Inertia-based K_ep showed a moderate positive correlation with the CD4⁺ TIL level (r = 0.549, p < 0.001), and the Correlation based on K_ep showed a moderate negative correlation with the CD4⁺ TIL level, which also had the highest correlation coefficient (r = -0.616, p < 0.001). The diagnostic efficacy of the above features was assessed by ROC curves. For CD8⁺ TILs, ClusterShade of K_ep had the highest mean area under the curve (AUC) (0.863). For CD4⁺ TILs, the Correlation of K_ep had the highest mean AUC (0.856). Conclusion: The radiomics features of DCE-MRI are associated with the expression of tumor-infiltrating CD8⁺ and CD4⁺ T cells in AGC, which have the potential to noninvasively evaluate the expression of CD8⁺ and CD4⁺ TILs in AGC patients.

Evaluating Tumor-Infiltrating Lymphocytes in Breast Cancer: The Role of Conventional Ultrasound and Contrast-Enhanced Ultrasound

OBJECTIVES

Tumor-infiltrating lymphocytes (TILs) have emerged as an efficient biomarker predicting treatment response and prognosis of breast cancer (BC). This study aimed to evaluate the association between conventional ultrasound and contrast-enhanced ultrasound (CEUS) imaging features with TIL levels in invasive BC patients.

METHODS

We retrospectively included 267 women with invasive BC who had undergone conventional ultrasound and CEUS. Patients were divided into low (≤10%) and high (>10%) TIL groups. Conventional ultrasound and CEUS features were analyzed by two sonographers. The associations between the TIL levels and imaging features were evaluated.

RESULTS

Of the 267 patients, 122 with high TILs and 145 with low TIL levels. High TIL tumors were more likely to have a circumscribed margin, oval or round shape, and enhanced posterior echoes on ultrasonography (p < 0.05). In contrast, low TIL tumors were more likely to have an irregular shape, un-circumscribed, indistinct and spiculated margin (p < 0.05). In CEUS, high TIL tumors showed a more regular shape, clearer margin, more homogeneous enhancement and higher peak intensity (PI) value (p < 0.05). Logistic analysis indicated that shape, posterior features, PI, and enhanced homogeneity were independent predictors for high TIL tumors. The model combined the four independent predictors have a moderate performance in predicting high TIL tumors with AUC 0.79, sensitivity 0.72, and specificity 0.78.

CONCLUSIONS

Conventional ultrasound and CEUS features were associated with TIL levels in invasive BC. Consequently, the results suggested that preoperative conventional ultrasound and CEUS may be a useful noninvasive imaging biomarker for individualized treatment decisions.

Deep Learning with Transformer or Convolutional Neural Network in the Assessment of Tumor-Infiltrating Lymphocytes (TILs) in Breast Cancer Based on US Images: A Dual-Center Retrospective Study

This study aimed to explore the feasibility of using a deep-learning (DL) approach to predict TIL levels in breast cancer (BC) from ultrasound (US) images. A total of 494 breast cancer patients with pathologically confirmed invasive BC from two hospitals were retrospectively enrolled. Of these, 396 patients from hospital 1 were divided into the training cohort (n = 298) and internal validation (IV) cohort (n = 98). Patients from hospital 2 (n = 98) were in the external validation (EV) cohort. TIL levels were confirmed by pathological results. Five different DL models were trained for predicting TIL levels in BC using US images from the training cohort and validated on the IV and EV cohorts. The overall best-performing DL model, the attention-based DenseNet121, achieved an AUC of 0.873, an accuracy of 79.5%, a sensitivity of 90.7%, a specificity of 65.9%, and an F1 score of 0.830 in the EV cohort. In addition, the stratified analysis showed that the DL models had good discrimination performance of TIL levels in each of the molecular subgroups. The DL models based on US images of BC patients hold promise for non-invasively predicting TIL levels and helping with individualized treatment decision-making.

Analysis of Breast Cancer Differences between China and Western Countries Based on Radiogenomics

Using radiogenomics methods, the differences between tumor imaging data and genetic data in Chinese and Western breast cancer (BC) patients were analyzed, and the correlation between phenotypic data and genetic data was explored. In this paper, we analyzed BC patients' image characteristics and transcriptome data separately, then correlated the magnetic resonance imaging (MRI) phenotype with the transcriptome data through a computational method to develop a radiogenomics feature. The data was fed into the designed random forest (RF) model, which used the area under the receiver operating curve (AUC) as the evaluation index. Next, we analyzed the hub genes in the differentially expressed genes (DEGs) and obtained seven hub genes, which may cause Chinese and Western BC patients to behave differently in the clinic. We demonstrated that combining relevant genetic data and imaging features could better classify Chinese and Western patients than using genes or imaging characteristics alone. The AUC values of 0.74, 0.81, and 0.95 were obtained separately using the image characteristics, DEGs, and radiogenomics features. We screened SYT4, GABRG2, CHGA, SLC6A17, NEUROG2, COL2A1, and MATN4 and found that these genes were positively or negatively correlated with certain imaging characteristics. In addition, we found that the SLC6A17, NEUROG2, CHGA, and MATN4 genes were associated with clinical features.

Non-invasive measurement of tumor immune microenvironment and prediction of survival and chemotherapeutic benefits from 18F fluorodeoxyglucose PET/CT images in gastric cancer

BACKGROUND

The tumor immune microenvironment could provide prognostic and predictive information. It is necessary to develop a noninvasive radiomics-based biomarker of a previously validated tumor immune microenvironment signature of gastric cancer (GC) with immunohistochemistry staining.

METHODS

A total of 230 patients (training (n = 153) or validation (n = 77) cohort) with gastric cancer were subjected to (Positron Emission Tomography-Computed Tomography) radiomics feature extraction (80 features). A radiomics tumor immune microenvironment score (RTIMS) was developed to predict the tumor immune microenvironment signature with LASSO logistic regression. Furthermore, we evaluated its relation with prognosis and chemotherapy benefits.

RESULTS

A 8-feature radiomics signature was established and validated (area under the curve=0.692 and 0.713). The RTIMS signature was significantly associated with disease-free survival and overall survival both in the training and validation cohort (all P<0.001). RTIMS was an independent prognostic factor in the Multivariate analysis. Further analysis revealed that high RTIMS patients benefitted from adjuvant chemotherapy (for DFS, stage II: HR 0.208(95% CI 0.061-0.711), p=0.012; stage III: HR 0.321(0.180-0.570), p<0.001, respectively); while there were no benefits from chemotherapy in a low RTIMS patients.

CONCLUSION

This PET/CT radiomics model provided a promising way to assess the tumor immune microenvironment and to predict clinical outcomes and chemotherapy response. The RTIMS signature could be useful in estimating tumor immune microenvironment and predicting survival and chemotherapy benefit for patients with gastric cancer, when validated by further prospective randomized trials.

Artificial intelligence in oncologic imaging

Radiology is integral to cancer care. Compared to molecular assays, imaging has its advantages. Imaging as a noninvasive tool can assess the entirety of tumor unbiased by sampling error and is routinely acquired at multiple time points in oncological practice. Imaging data can be digitally post-processed for quantitative assessment. The ever-increasing application of Artificial intelligence (AI) to clinical imaging is challenging radiology to become a discipline with competence in data science, which plays an important role in modern oncology. Beyond streamlining certain clinical tasks, the power of AI lies in its ability to reveal previously undetected or even imperceptible radiographic patterns that may be difficult to ascertain by the human sensory system. Here, we provide a narrative review of the emerging AI applications relevant to the oncological imaging spectrum and elaborate on emerging paradigms and opportunities. We envision that these technical advances will change radiology in the coming years, leading to the optimization of imaging acquisition and discovery of clinically relevant biomarkers for cancer diagnosis, staging, and treatment monitoring. Together, they pave the road for future clinical translation in precision oncology.

Integrated imaging and molecular analysis to decipher tumor microenvironment in the era of immunotherapy

Radiological imaging is an integral component of cancer care, including diagnosis, staging, and treatment response monitoring. It contains rich information about tumor phenotypes that are governed not only by cancer cellintrinsic biological processes but also by the tumor microenvironment, such as the composition and function of tumor-infiltrating immune cells. By analyzing the radiological scans using a quantitative radiomics approach, robust relations between specific imaging and molecular phenotypes can be established. Indeed, a number of studies have demonstrated the feasibility of radiogenomics for predicting intrinsic molecular subtypes and gene expression signatures in breast cancer based on MRI. In parallel, promising results have been shown for inferring the amount of tumor-infiltrating lymphocytes, a key factor for the efficacy of cancer immunotherapy, from standard-of-care radiological images. Compared with the biopsy-based approach, radiogenomics offers a unique avenue to profile the molecular makeup of the tumor and immune microenvironment as well as its evolution in a noninvasive and holistic manner through longitudinal imaging scans. Here, we provide a systematic review of the state of the art radiogenomics studies in the era of immunotherapy and discuss emerging paradigms and opportunities in AI and deep learning approaches. These technical advances are expected to transform the radiogenomics field, leading to the discovery of reliable imaging biomarkers. This will pave the way for their clinical translation to guide precision cancer therapy.

Machine Learning for Computed Tomography Radiomics: Prediction of Tumor-Infiltrating Lymphocytes in Patients With Pancreatic Ductal Adenocarcinoma

OBJECTIVES

The aims of the study were to develop and validate a machine learning classifier for preoperative prediction of tumor-infiltrating lymphocytes (TILs) in patients with pancreatic ductal adenocarcinoma (PDAC).

METHODS

In this retrospective study of 183 PDAC patients who underwent multidetector computed tomography and surgical resection, CD4 + , CD8 + , and CD20 + expression was evaluated using immunohistochemistry, and TIL scores were calculated using the Cox regression model. The patients were divided into TIL-low and TIL-high groups. An extreme gradient boosting (XGBoost) classifier was developed using a training set consisting of 136 consecutive patients, and the model was validated in 47 consecutive patients. The discriminative ability, calibration, and clinical utility of the XGBoost classifier were evaluated.

RESULTS

The prediction model showed good discrimination in the training (area under the curve, 0.93; 95% confidence interval, 0.89-0.97) and validation (area under the curve, 0.79; 95% confidence interval, 0.65-0.92) sets with good calibration. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value for the training set were 0.93, 0.85, 0.90, 0.89, and 0.91, respectively, while those for the validation set were 0.63, 0.91, 0.77, 0.88, and 0.70, respectively.

CONCLUSIONS

The XGBoost-based model could predict PDAC TILs and may facilitate clinical decision making for immune therapy.

Preoperative Radiomics Approach to Evaluating Tumor-Infiltrating CD8+ T Cells in Patients With Pancreatic Ductal Adenocarcinoma Using Noncontrast Magnetic Resonance Imaging

BACKGROUND

CD8+ T cell in pancreatic ductal adenocarcinoma (PDAC) is closely related to the prognosis and treatment response of patients. Accurate preoperative CD8+ T-cell expression can better identify the population benefitting from immunotherapy.

PURPOSE

To develop and validate a machine learning classifier based on noncontrast magnetic resonance imaging (MRI) for the preoperative prediction of CD8+ T-cell expression in patients with PDAC.

STUDY TYPE

Retrospective cohort study.

POPULATION

Overall, 114 patients with PDAC undergoing MR scan and surgical resection; 97 and 47 patients in the training and validation cohorts. FIELD STRENGTH/SEQUENCE/3 T: Breath-hold single-shot fast-spin echo T2-weighted sequence and noncontrast T1-weighted fat-suppressed sequences.

ASSESSMENT

CD8+ T-cell expression was quantified using immunohistochemistry. For each patient, 2232 radiomics features were extracted from noncontrast T1- and T2-weighted images and reduced using the Wilcoxon rank-sum test and least absolute shrinkage and selection operator method. Linear discriminative analysis was used to construct radiomics and mixed models. Model performance was determined by its discriminative ability, calibration, and clinical utility.

STATISTICAL TESTS

Kaplan-Meier estimates, Student's t-test, the Kruskal-Wallis H test, and the chi-square test, receiver operating characteristic curve, and decision curve analysis.

RESULTS

A log-rank test showed that the survival duration in the CD8-high group (25.51 months) was significantly longer than that in the CD8-low group (22.92 months). The mixed model included all MRI characteristics and 13 selected radiomics features, and the area under the curve (AUC) was 0.89 (95% confidence interval [CI], 0.77-0.92) and 0.69 (95% CI, 0.53-0.82) in the training and validation cohorts. The radiomics model included 13 radiomics features, which showed good discrimination in the training cohort (AUC, 0.85; 95% CI, 0.77-0.92) and the validation cohort (AUC, 0.76; 95% CI, 0.61-0.87).

DATA CONCLUSIONS

This study developed a noncontrast MRI-based radiomics model that can preoperatively determine CD8+ T-cell expression in patients with PDAC and potentially immunotherapy planning.

EVIDENCE LEVEL

5 TECHNICAL EFFICACY: Stage 2.

Predicting cancer outcomes with radiomics and artificial intelligence in radiology

The successful use of artificial intelligence (AI) for diagnostic purposes has prompted the application of AI-based cancer imaging analysis to address other, more complex, clinical needs. In this Perspective, we discuss the next generation of challenges in clinical decision-making that AI tools can solve using radiology images, such as prognostication of outcome across multiple cancers, prediction of response to various treatment modalities, discrimination of benign treatment confounders from true progression, identification of unusual response patterns and prediction of the mutational and molecular profile of tumours. We describe the evolution of and opportunities for AI in oncology imaging, focusing on hand-crafted radiomic approaches and deep learning-derived representations, with examples of their application for decision support. We also address the challenges faced on the path to clinical adoption, including data curation and annotation, interpretability, and regulatory and reimbursement issues. We hope to demystify AI in radiology for clinicians by helping them to understand its limitations and challenges, as well as the opportunities it provides as a decision-support tool in cancer management.

Machine learning for MRI radiomics: a study predicting tumor-infiltrating lymphocytes in patients with pancreatic ductal adenocarcinoma

OBJECTIVE

To develop and validate a machine learning classifier based on magnetic resonance imaging (MRI), for the preoperative prediction of tumor-infiltrating lymphocytes (TILs) in patients with pancreatic ductal adenocarcinoma (PDAC).

MATERIALS AND METHODS

In this retrospective study, 156 patients with PDAC underwent MR scan and surgical resection. The expression of CD4, CD8 and CD20 was detected and quantified using immunohistochemistry, and TILs score was achieved by Cox regression model. All patients were divided into TILs score-low and TILs score-high groups. The least absolute shrinkage and selection operator method and the extreme gradient boosting (XGBoost) were used to select the features and to construct a prediction model. The performance of the models was assessed using the training cohort (116 patients) and the validation cohort (40 patients), and decision curve analysis (DCA) was applied for clinical use.

RESULTS

The XGBoost prediction model showed good discrimination in the training (AUC 0.86; 95% CI 0.79-0.93) and validation sets (AUC 0.79; 95% CI 0.64-0.93). The sensitivity, specificity, and accuracy for the training set were 86.67%, 75.00%, and 0.81, respectively, whereas those for the validation set were 84.21%, 66.67%, and 0.75, respectively. Decision curve analysis indicated the clinical usefulness of the XGBoost classifier.

CONCLUSION

The model constructed by XGBoost could predict PDAC TILs and may aid clinical decision making for immune therapy.

Radiological tumor classification across imaging modality and histology

Radiomics refers to the high-throughput extraction of quantitative features from radiological scans and is widely used to search for imaging biomarkers for prediction of clinical outcomes. Current radiomic signatures suffer from limited reproducibility and generalizability, because most features are dependent on imaging modality and tumor histology, making them sensitive to variations in scan protocol. Here, we propose novel radiological features that are specially designed to ensure compatibility across diverse tissues and imaging contrast. These features provide systematic characterization of tumor morphology and spatial heterogeneity. In an international multi-institution study of 1,682 patients, we discover and validate four unifying imaging subtypes across three malignancies and two major imaging modalities. These tumor subtypes demonstrate distinct molecular characteristics and prognoses after conventional therapies. In advanced lung cancer treated with immunotherapy, one subtype is associated with improved survival and increased tumor-infiltrating lymphocytes compared with the others. Deep learning enables automatic tumor segmentation and reproducible subtype identification, which can facilitate practical implementation. The unifying radiological tumor classification may inform prognosis and treatment response for precision medicine.

Human Epidermal Growth Factor Receptor Type 2-Positive Breast Cancer: Association of MRI and Clinicopathologic Features With Tumor-Infiltrating Lymphocytes

Background: Tumor-infiltrating lymphocytes (TILs) are associated with therapeutic outcomes and prognosis in patients with human epidermal growth factor receptor type 2-positive (HER2+) breast cancer. Identification of TIL levels is clinically relevant. Objective: To explore associations of clinicopathologic and MRI features with TIL levels in patients with HER2+ breast cancer. Methods: A total of 212 consecutive women (mean age, 54 years) diagnosed with HER2+ breast cancer between January 2017 and December 2019 were included in this retrospective study. Patients were divided into low (<10%) and high (≥10%) TIL groups. Three breast radiologists independently reviewed images; interreader agreement was assessed, and the first readers' findings were used for further analysis. Associations of clinicopathologic and MRI features with TIL levels were evaluated using multivariable logistic regression analysis. Subanalysis of TIL levels by hormone receptor (HR) status was also performed. Results: A total of 115 (54.2%) patients had low, and 97 (45.8%) had high, TIL levels. High TIL level was associated (all p<.05) with histologic grade 3 (odds ratio [OR]=3.98; frequency of 78.4% vs 52.2% in high vs low TIL groups, respectively), high tumor cellularity (OR=4.59; median cellularity of 60% vs 50%), lower frequency of associated ductal carcinoma in situ (OR=0.16; frequency of 86.6% vs 94.8%), and higher frequency of peritumoral edema on T2-weighted images (OR=2.83; 71.1% vs 50.4%). In subgroup analysis by HR status, histologic grade 3 (OR=5.03, p=.002) was a significant independent predictor of high TIL in the HR+/HER2+ group, while high tumor cellularity (OR=9.06, p=.002), peritumoral edema (OR=5.23, p=.03), and low ADC (OR=11.69, p=.047) were independent predictors of high TIL in the HR-/HER2+ group. Interreader agreement for peritumoral edema was moderate among the three radiologists (к, range 0.432-0.539). Conclusion: Peritumoral edema on MRI and histopathologic feature of tumor aggressiveness help predict high TIL levels in patients with HER2+ breast cancer. Clinical Impact: Pretreatment MRI features may serve as a useful tool for assessing TIL levels in patients with HER2+ breast cancer, helping to classify patients with variable clinical outcomes related to immune activity and to guide selection among neoadjuvant chemotherapy (NAC) or HER2-targeted therapy or immunotherapy.

A radiomic signature model to predict the chemoradiation-induced alteration in tumor-infiltrating CD8+ cells in locally advanced rectal cancer

BACKGROUND AND PURPOSE

Regarding the altered tumor immune status following cytotoxic treatment, this study aims to develop a radiomic signature to predict CD8⁺ tumor-infiltrating lymphocyte (TIL) density changes in chemoradiotherapy (CRT) of rectal cancer.

MATERIALS AND METHODS

We used the magnetic resonance imaging (MRI) and immunohistochemistry data before and after neoadjuvant CRT. The discovery datasets consisted of pre-CRT dataset A1 (n = 113), post-CRT datasets A2 (n = 32; predominance of tumor) and A3 (n = 20; pure fibrosis). The developed model was validated in dataset B (n = 28). Thirty-eight radiomic features from T2-weighted MRI scans were incorporated into the least absolute shrinkage and selection operator method.

RESULTS

In pre-CRT dataset A1, the area under the receiver operating characteristic curve (AUC) values of radiomic score for predicting CD8⁺ TILs were 0.760 and 0.729 for training and validation subsets, respectively. A significant correlation was observed between the signature and CD8⁺ TIL density in the post-CRT dataset A2 (Pearson's R = -0.372, P = 0.036), whereas no association was found in dataset A3 (Pearson's R = -0.069, P = 0.77). The association was also observed in the validation dataset B (Pearson's R = -0.374, P = 0.049). In dataset A2, the radiomic score difference predicted changes in CD8⁺ TIL density (AUC = 0.824).

CONCLUSION

We established the MRI-derived radiomic signature for predicting CRT-induced alterations in CD8⁺ TILs. This study suggests the clinical utility of radiomics-immunophenotype modeling to evaluate tumor immune status following neoadjuvant chemoradiation in rectal cancer.

Radiographical assessment of tumour stroma and treatment outcomes using deep learning: a retrospective, multicohort study

BACKGROUND

The tumour stroma microenvironment plays an important part in disease progression and its composition can influence treatment response and outcomes. Histological evaluation of tumour stroma is limited by access to tissue, spatial heterogeneity, and temporal evolution. We aimed to develop a radiological signature for non-invasive assessment of tumour stroma and treatment outcomes.

METHODS

In this multicentre, retrospective study, we analysed CT images and outcome data of 2209 patients with resected gastric cancer from five independent cohorts recruited from two centres (Nanfang Hospital of Southern Medical University [Guangzhou, China] and Sun Yat-sen University Cancer Center [Guangzhou, China]). Patients with histologically confirmed gastric cancer, at least 15 lymph nodes harvested, preoperative abdominal CT available, and complete clinicopathological and follow-up data were eligible for inclusion. Tumour tissue was collected for patients in the training cohort (321 patients), internal validation cohort one (246 patients), and external validation cohort one (128 patients). Four stroma classes were defined according to the protein expression of α-smooth muscle actin and periostin assessed by immunohistochemistry. The primary objective was to predict the histologically based stroma classes by using preoperative CT images. We trained a deep convolutional neural network model using the training cohort and tested the model in the internal and external validation cohort one. We evaluated the model's association with prognosis in the training cohort, two internal, and two external validation cohorts and compared outcomes of patients who received or did not receive adjuvant chemotherapy.

FINDINGS

The deep-learning model achieved a high diagnostic accuracy for assessing tumour stroma in both internal validation cohort one (area under the receiver operating characteristic curve [AUC] 0·96-0·98]) and external validation cohort one (AUC 0·89-0·94). The stromal imaging signature was significantly associated with disease-free survival and overall survival in all cohorts (p<0·0001). The predicted stroma classes remained an independent prognostic factor adjusting for clinicopathological variables including tumour size, stage, differentiation, and Lauren histology. In patients with stage II or III disease in predicted stroma classes one and two subgroups, patients who received adjuvant chemotherapy had improved survival compared with those who did not (in those with stage II disease hazard ratio [HR] 0·48 [95% CI 0·29-0·77], p=0·0021; and in those with stage III disease HR 0·70 [0·57-0·85], p=0·00042). However, in the other two subgroups adjuvant chemotherapy was not associated with survival and might even be detrimental in the predicted stroma class 4 subgroup (HR 1·48 [1·08-2·03], p=0·013).

INTERPRETATION

The deep-learning model could allow for accurate and non-invasive evaluation of tumour stroma from CT images in gastric cancer. The radiographical model predicted chemotherapy outcomes and could be used in combination with clinicopathological criteria to refine prognosis and inform treatment decisions of patients with gastric cancer.

FUNDING

None.

High-background parenchymal enhancement in the contralateral breast is an imaging biomarker for favorable prognosis in patients with triple-negative breast cancer treated with chemotherapy

This study aimed to analyze the association between background parenchymal enhancement (BPE) in the contralateral breast tissue on magnetic resonance imaging (MRI) and clinicopathologic parameters in patients with unilateral breast carcinoma and to investigate its potential prognostic significance. A total of 467 patients who were pathologically confirmed to have unilateral breast cancer and underwent breast MRI were recruited to participate in this cohort study. BPE was assessed in the healthy contralateral breast. Minimal and mild levels were classified as low BPE, whereas moderate and marked levels were classified as high BPE. The effects of BPE on clinicopathologic parameters, overall survival (OS), and invasive disease-free survival (IDFS) were determined. Among the 467 patients, 327 cases were classified into the low-BPE group, whereas 140 cases were classified into the high-BPE group. The high-BPE pattern markedly correlated with age at diagnosis, menopausal status, histologic grading, and estrogen receptor status. BPE pattern did not correlate with OS and IDFS in the entire breast cancer cohort, regardless of whether adjuvant chemotherapy was received. Notably, BPE in the healthy contralateral breast on MRI is markedly related to OS and IDFS in triple-negative breast cancer (TNBC) cases who received chemotherapy. High BPE is related to chemotherapeutic benefits and can be an independent favorable prognostic factor for TNBC patients. Thus, our observations suggest that high BPE pattern can potentially be used as an imaging biomarker for relatively favorable prognosis in TNBC cases receiving chemotherapy. However, the findings need to be verified in a large-scale study.

Radiomic biomarkers of tumor immune biology and immunotherapy response

Immunotherapies are leading to improved outcomes for many cancers, including those with devastating prognoses. As therapies like immune checkpoint inhibitors (ICI) become a mainstay in treatment regimens, many concurrent challenges have arisen - for instance, delineating clinical responders from non-responders. Predicting response has proven to be difficult given a lack of consistent and accurate biomarkers, heterogeneity of the tumor microenvironment (TME), and a poor understanding of resistance mechanisms. For the most part, imaging data have remained an untapped, yet abundant, resource to address these challenges. In recent years, quantitative image analyses have highlighted the utility of medical imaging in predicting tumor phenotypes, prognosis, and therapeutic response. These studies have been fueled by an explosion of resources in high-throughput mining of image features (i.e. radiomics) and artificial intelligence. In this review, we highlight current progress in radiomics to understand tumor immune biology and predict clinical responses to immunotherapies. We also discuss limitations in these studies and future directions for the field, particularly if high-dimensional imaging data are to play a larger role in precision medicine.

Dynamic contrast-enhanced breast MRI features correlate with invasive breast cancer angiogenesis

Angiogenesis is a critical component of breast cancer development, and identification of imaging-based angiogenesis assays has prognostic and treatment implications. We evaluated the association of semi-quantitative kinetic and radiomic breast cancer features on dynamic contrast-enhanced (DCE)-MRI with microvessel density (MVD), a marker for angiogenesis. Invasive breast cancer kinetic features (initial peak percent enhancement [PE], signal enhancement ratio [SER], functional tumor volume [FTV], and washout fraction [WF]), radiomics features (108 total features reflecting tumor morphology, signal intensity, and texture), and MVD (by histologic CD31 immunostaining) were measured in 27 patients (1/2016-7/2017). Lesions with high MVD levels demonstrated higher peak SER than lesions with low MVD (mean: 1.94 vs. 1.61, area under the receiver operating characteristic curve [AUC] = 0.79, p = 0.009) and higher WF (mean: 50.6% vs. 22.5%, AUC = 0.87, p = 0.001). Several radiomics texture features were also promising for predicting increased MVD (maximum AUC = 0.84, p = 0.002). Our study suggests DCE-MRI can non-invasively assess breast cancer angiogenesis, which could stratify biology and optimize treatments.

Predicting the Level of Tumor-Infiltrating Lymphocytes in Patients With Breast Cancer: Usefulness of Mammographic Radiomics Features

Objectives

This study aimed to investigate whether radiomics classifiers from mammography can help predict tumor-infiltrating lymphocyte (TIL) levels in breast cancer.

Methods

Data from 121 consecutive patients with pathologically-proven breast cancer who underwent preoperative mammography from February 2018 to May 2019 were retrospectively analyzed. Patients were randomly divided into a training dataset (n = 85) and a validation dataset (n = 36). A total of 612 quantitative radiomics features were extracted from mammograms using the Pyradiomics software. Radiomics feature selection and radiomics classifier were generated through recursive feature elimination and logistic regression analysis model. The relationship between radiomics features and TIL levels in breast cancer patients was explored. The predictive capacity of the radiomics classifiers for the TIL levels was investigated through receiver operating characteristic curves in the training and validation groups. A radiomics score (Rad score) was generated using a logistic regression analysis method to compute the training and validation datasets, and combining the Mann–Whitney U test to evaluate the level of TILs in the low and high groups.

Results

Among the 121 patients, 32 (26.44%) exhibited high TIL levels, and 89 (73.56%) showed low TIL levels. The ER negativity (p = 0.01) and the Ki-67 negative threshold level (p = 0.03) in the low TIL group was higher than that in the high TIL group. Through the radiomics feature selection, six top-class features [Wavelet GLDM low gray-level emphasis (mediolateral oblique, MLO), GLRLM short-run low gray-level emphasis (craniocaudal, CC), LBP2D GLRLM short-run high gray-level emphasis (CC), LBP2D GLDM dependence entropy (MLO), wavelet interquartile range (MLO), and LBP2D median (MLO)] were selected to constitute the radiomics classifiers. The radiomics classifier had an excellent predictive performance for TIL levels both in the training and validation sets [area under the curve (AUC): 0.83, 95% confidence interval (CI), 0.738–0.917, with positive predictive value (PPV) of 0.913; AUC: 0.79, 95% CI, 0.615–0.964, with PPV of 0.889, respectively]. Moreover, the Rad score in the training dataset was higher than that in the validation dataset (p = 0.007 and p = 0.001, respectively).

Conclusion

Radiomics from digital mammograms not only predicts the TIL levels in breast cancer patients, but can also serve as non-invasive biomarkers in precision medicine, allowing for the development of treatment plans.

Relationship between MRI findings and invasive breast cancer with podoplanin-positive cancer-associated fibroblasts

OBJECTIVE

Purpose of our study is to assess the relationship between MRI findings and invasive breast cancer (IBC) with cancer-associated fibroblasts (CAFs) that are positive for podoplanin.

METHODS

We retrospectively analyzed the consecutive 109 IBCs. The IBCs were dichotomized as with (+) or without (-) podoplanin-positive CAFs. In MRI analyses, the dichotomized IBCs were compared the lesion to muscle ratio (L/M ratio) in STIR images, the ADC value, the distribution of kinetic parameters, and morphological findings.

RESULTS

Of the 109 IBCs, 28 (26%) IBCs had podoplanin(+) CAFs. Compared to the podoplanin(-) group, the podoplanin(+) group tended to have a more malignant pathological status. In the STIR images, the podoplanin(+) group had significantly higher L/M ratio (7.59 vs. 6.55, p = 0.040). In a dynamic study, the podoplanin(+) group had a significantly higher percentage of the washout pattern (42.21% vs. 29.43%, p = 0.045). There were 23 mass lesions and 5 non-mass enhancement (NME) lesions in the podoplanin(+) group, and 69 mass lesions and 12 NME lesions in the podoplanin(-) group. The mass lesions of the podoplanin(-) group had a significantly higher likelihood of showing an irregular shape (n = 47 vs. 8, p = 0.035). The podoplanin(+) group's lesions had a significantly higher likelihood of showing a circumscribed margin (n = 14 vs. 6, p < 0.001) and a rim enhancement (n = 10 vs. 13, p = 0.047). In multivariate analyses, only high nuclear grade was significant predictive value of podoplanin(+) CAFs.

CONCLUSION

Although not significant in multivariate analyses, MRI findings may be used to determine the podoplanin-positive CAF status of invasive breast cancer.

Radiogenomic signatures reveal multiscale intratumour heterogeneity associated with biological functions and survival in breast cancer

Advanced tumours are often heterogeneous, consisting of subclones with various genetic alterations and functional roles. The precise molecular features that characterize the contributions of multiscale intratumour heterogeneity to malignant progression, metastasis, and poor survival are largely unknown. Here, we address these challenges in breast cancer by defining the landscape of heterogeneous tumour subclones and their biological functions using radiogenomic signatures. Molecular heterogeneity is identified by a fully unsupervised deconvolution of gene expression data. Relative prevalence of two subclones associated with cell cycle and primary immunodeficiency pathways identifies patients with significantly different survival outcomes. Radiogenomic signatures of imaging scale heterogeneity are extracted and used to classify patients into groups with distinct subclone compositions. Prognostic value is confirmed by survival analysis accounting for clinical variables. These findings provide insight into how a radiogenomic analysis can identify the biological activities of specific subclones that predict prognosis in a noninvasive and clinically relevant manner.

Tumours are made up of heterogeneous subclones. Here, the authors show using breast cancer imaging and gene expression datasets that these subclones can be inferred by the deconvolution of gene expression data, mapped to MRI derived radiogenomic signatures and used to estimate prognosis.

Early response evaluation using primary tumor and nodal imaging features to predict progression-free survival of locally advanced non-small cell lung cancer

Prognostic biomarkers that can reliably predict early disease progression of non-small cell lung cancer (NSCLC) are needed for identifying those patients at high risk for progression, who may benefit from more intensive treatment. In this work, we aimed to identify an imaging signature for predicting progression-free survival (PFS) of locally advanced NSCLC. Methods: This retrospective study included 82 patients with stage III NSCLC treated with definitive chemoradiotherapy for whom both baseline and mid-treatment PET/CT scans were performed. They were randomly placed into two groups: training cohort (n=41) and testing cohort (n=41). All primary tumors and involved lymph nodes were delineated. Forty-five quantitative imaging features were extracted to characterize the tumors and involved nodes at baseline and mid-treatment as well as differences between two scans performed at these two points. An imaging signature was developed to predict PFS by fitting an L1-regularized Cox regression model. Results: The final imaging signature consisted of three imaging features: the baseline tumor volume, the baseline maximum distance between involved nodes, and the change in maximum distance between the primary tumor and involved nodes measured at two time points. According to multivariate analysis, the imaging model was an independent prognostic factor for PFS in both the training (hazard ratio [HR], 1.14, 95% confidence interval [CI], 1.04-1.24; P = 0.003), and testing (HR, 1.21, 95% CI, 1.10-1.33; P = 0.048) cohorts. The imaging signature stratified patients into low- and high-risk groups, with 2-year PFS rates of 61.9% and 33.2%, respectively (P = 0.004 [log-rank test]; HR, 4.13, 95% CI, 1.42-11.70) in the training cohort, as well as 43.8% and 22.6%, respectively (P = 0.006 [log-rank test]; HR, 3.45, 95% CI, 1.35-8.83) in the testing cohort. In both cohorts, the imaging signature significantly outperformed conventional imaging metrics, including tumor volume and SUV_max value (C-indices: 0.77-0.79 for imaging signature, and 0.53-0.73 for conventional metrics). Conclusions: Evaluation of early treatment response by combining primary tumor and nodal imaging characteristics may improve the prediction of PFS of locally advanced NSCLC patients.

Noninvasive imaging evaluation of tumor immune microenvironment to predict outcomes in gastric cancer

BACKGROUND

The tumor immune microenvironment can provide prognostic and predictive information. A previously validated ImmunoScore of Gastric Cancer (IS_GC) evaluates both lymphoid and myeloid cells in the tumor core and invasive margin with immunohistochemical staining of surgical specimens. We aimed to develop a noninvasive radiomics-based predictor of IS_GC.

PATIENTS AND METHODS

In this retrospective study including four independent cohorts of 1778 patients, we extracted 584 quantitative features from the intratumoral and peritumoral regions on contrast-enhanced computed tomography images. A radiomic signature [radiomics ImmunoScore (RIS)] was constructed to predict IS_GC using regularized logistic regression. We further evaluated its association with prognosis and chemotherapy response.

RESULTS

A 13-feature radiomic signature for IS_GC was developed and validated in three independent cohorts (area under the curve = 0.786, 0.745, and 0.766). The RIS signature was significantly associated with both disease-free and overall survival in the training and all validation cohorts [hazard ratio (HR) range: 0.296-0.487, all P < 0.001]. In multivariable analysis, the RIS remained an independent prognostic factor adjusting for clinicopathologic variables (adjusted HR range: 0.339-0.605, all P < 0.003). For stage II and stage III disease, patients with a high RIS derived survival benefit from adjuvant chemotherapy {HR = 0.436 [95% confidence interval (CI) 0.253-0.753], P = 0.002; HR = 0.591 (95% CI 0.428-0.818), P < 0.001, respectively}, whereas those with a low RIS did not.

CONCLUSION

The RIS is a reliable tool for evaluation of immunoscore and retains the prognostic significance in gastric cancer. Future prospective studies are required to confirm its potential to predict treatment response and select patients who will benefit from chemotherapy.

AUNet: attention-guided dense-upsampling networks for breast mass segmentation in whole mammograms

Mammography is one of the most commonly applied tools for early breast cancer screening. Automatic segmentation of breast masses in mammograms is essential but challenging due to the low signal-to-noise ratio and the wide variety of mass shapes and sizes. Existing methods deal with these challenges mainly by extracting mass-centered image patches manually or automatically. However, manual patch extraction is time-consuming and automatic patch extraction brings errors that could not be compensated in the following segmentation step. In this study, we propose a novel attention-guided dense-upsampling network (AUNet) for accurate breast mass segmentation in whole mammograms directly. In AUNet, we employ an asymmetrical encoder-decoder structure and propose an effective upsampling block, attention-guided dense-upsampling block (AU block). Especially, the AU block is designed to have three merits. Firstly, it compensates the information loss of bilinear upsampling by dense upsampling. Secondly, it designs a more effective method to fuse high- and low-level features. Thirdly, it includes a channel-attention function to highlight rich-information channels. We evaluated the proposed method on two publicly available datasets, CBIS-DDSM and INbreast. Compared to three state-of-the-art fully convolutional networks, AUNet achieved the best performances with an average Dice similarity coefficient of 81.8% for CBIS-DDSM and 79.1% for INbreast.

Correlation between magnetic resonance imaging and the level of tumor-infiltrating lymphocytes in patients with estrogen receptor-negative HER2-positive breast cancer

Background

High levels of tumor-infiltrating lymphocytes (TILs) are associated with improved prognosis and response to therapy in patients with human epidermal growth factor receptor 2 (HER2)-positive breast cancer.

Purpose

This study investigated the associations between TIL levels and magnetic resonance imaging (MRI) findings in patients with estrogen receptor (ER)-negative HER-2 positive breast cancer.

Material and Methods

This study included 110 consecutive patients with surgically confirmed ER-negative HER2-positive breast cancers who underwent preoperative MRI from January to December 2015. Images of all lesions were reviewed in accordance with the BI-RADS lexicon by radiologists blinded to clinicopathologic findings. Tumor kinetic features were acquired by computer-aided diagnosis (CAD). Patients were divided into three TIL groups: low (<10%); intermediate (10–50%); and high (>50%). Associations between TIL levels and clinicopathologic and imaging features were evaluated; independent predictors of high and low TIL were identified by multiple logistic regression analysis.

Results

The 110 patients included 29 (26.4%) with low, 45 (40.9%) with intermediate, and 36 (32.7%) with high TIL levels. Multiple logistic regression analysis showed that older age (odds ratio [OR] = 1.08; P = 0.017), high peak enhancement (OR = 1.01; P = 0.019), positive CK5/6 (OR = 4.36; P = 0.024), and low Ki-67 (OR = 14.29; P = 0.037) were significantly associated with low TILs; low peak enhancement (OR = 1.01; P = 0.020) was significantly associated with high TILs.

Conclusion

MRI features may predict TIL levels in patients with ER-negative HER-2 positive breast cancer, enhancing the ability to diagnose and treat these patients.

Usefulness of imaging findings in predicting tumor-infiltrating lymphocytes in patients with breast cancer

OBJECTIVES

Tumor-infiltrating lymphocytes (TILs) have been determined as a new prognostic indicator of immunotherapy response in breast cancer (BC). The aim of this study is to investigate the effectiveness of imaging features in predicting the TIL levels in invasive BC patients.

METHODS

A total of 158 patients with invasive BC were included in our study. All lesions were evaluated based on the BIRADS lexicon. US was performed for all the patients and 89 of them underwent MRI. The histologic stromal TIL (sTIL) levels were assessed and associations between the sTIL levels and imaging features were evaluated.

RESULTS

Tumors with high sTIL levels had more circumscribed margins, round shape, heterogeneous echogenicity, and larger size on ultrasonography (p < 0.005). There was a statistically significant positive correlation between the sTIL levels and ADC value (p < 0.001). Tumors with high sTIL levels had significantly more homogeneous enhancement than the tumors with low sTIL levels (p = 0.001). Logistic regression analysis showed that the ADC was the most statistically significant parameter in predicting the sTIL levels (the odds ratio was 90.952; p = 0.002). The optimal cutoff value for ADC in predicting low and high sTIL levels was found to be 0.87 × 10^-3 mm² s^-1 (AUC = 0.726, 73% specificity, and 60% sensitivity).

CONCLUSIONS

Imaging findings, especially the ADC, may play an important role as an adjunct tool in cases of uncertain situations and may improve the accuracy of biopsy results. The prediction of sTIL levels using imaging findings may give an opportunity to predict prognosis.

KEY POINTS

• Preoperative assessment of TILs is an important biomarker of prognosis and treatment efficacy. • ADC value can be a useful tool in distinguishing high and low sTIL levels as a non-invasive method. • The prediction of sTIL levels using imaging findings may give an opportunity to predict prognosis and an optimal treatment for the BC patients.

Predictive Power of a Radiomic Signature Based on 18F-FDG PET/CT Images for EGFR Mutational Status in NSCLC

Radiomics has become an area of interest for tumor characterization in ¹⁸F-Fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT) imaging. The aim of the present study was to demonstrate how imaging phenotypes was connected to somatic mutations through an integrated analysis of 115 non-small cell lung cancer (NSCLC) patients with somatic mutation testings and engineered computed PET/CT image analytics. A total of 38 radiomic features quantifying tumor morphological, grayscale statistic, and texture features were extracted from the segmented entire-tumor region of interest (ROI) of the primary PET/CT images. The ensembles for boosting machine learning scheme were employed for classification, and the least absolute shrink age and selection operator (LASSO) method was used to select the most predictive radiomic features for the classifiers. A radiomic signature based on both PET and CT radiomic features outperformed individual radiomic features, the PET or CT radiomic signature, and the conventional PET parameters including the maximum standardized uptake value (SUVmax), SUVmean, SUVpeak, metabolic tumor volume (MTV), and total lesion glycolysis (TLG), in discriminating between mutant-type of epidermal growth factor receptor (EGFR) and wild-type of EGFR- cases with an AUC of 0.805, an accuracy of 80.798%, a sensitivity of 0.826 and a specificity of 0.783. Consistently, a combined radiomic signature with clinical factors exhibited a further improved performance in EGFR mutation differentiation in NSCLC. In conclusion, tumor imaging phenotypes that are driven by somatic mutations may be predicted by radiomics based on PET/CT images.

Integrating Tumor and Nodal Imaging Characteristics at Baseline and Mid-Treatment Computed Tomography Scans to Predict Distant Metastasis in Oropharyngeal Cancer Treated With Concurrent Chemoradiotherapy

PURPOSE

Prognostic biomarkers of disease relapse are needed for risk-adaptive therapy of oropharyngeal cancer (OPC). This work aims to identify an imaging signature to predict distant metastasis in OPC.

METHODS AND MATERIALS

This single-institution retrospective study included 140 patients treated with definitive concurrent chemoradiotherapy, for whom both pre- and midtreatment contrast-enhanced computed tomography (CT) scans were available. Patients were divided into separate training and testing cohorts. Forty-five quantitative image features were extracted to characterize tumor and involved lymph nodes at both time points. By incorporating both imaging and clinicopathological features, a random survival forest (RSF) model was built to predict distant metastasis-free survival (DMFS). The model was optimized via repeated cross-validation in the training cohort and then independently validated in the testing cohort.

RESULTS

The most important features for predicting DMFS were the maximum distance among nodes, maximum distance between tumor and nodes at mid-treatment, and pretreatment tumor sphericity. In the testing cohort, the RSF model achieved good discriminability for DMFS (C-index = 0.73, P = .008), and further divided patients into 2 risk groups with different 2-year DMFS rates: 96.7% versus 67.6%. Similar trends were observed for patients with p16+ tumors and smoking ≤10 pack-years. The RSF model based on pretreatment CT features alone achieved lower performance (concordance index = 0.68, P = .03).

CONCLUSIONS

Integrating tumor and nodal imaging characteristics at baseline and mid-treatment CT allows prediction of distant metastasis in OPC. The proposed imaging signature requires prospective validation and, if successful, may help identify high-risk human papillomavirus-positive patients who should not be considered for deintensification therapy.

Machine learning in breast MRI

Machine-learning techniques have led to remarkable advances in data extraction and analysis of medical imaging. Applications of machine learning to breast MRI continue to expand rapidly as increasingly accurate 3D breast and lesion segmentation allows the combination of radiologist-level interpretation (eg, BI-RADS lexicon), data from advanced multiparametric imaging techniques, and patient-level data such as genetic risk markers. Advances in breast MRI feature extraction have led to rapid dataset analysis, which offers promise in large pooled multiinstitutional data analysis. The object of this review is to provide an overview of machine-learning and deep-learning techniques for breast MRI, including supervised and unsupervised methods, anatomic breast segmentation, and lesion segmentation. Finally, it explores the role of machine learning, current limitations, and future applications to texture analysis, radiomics, and radiogenomics. Level of Evidence: 3 Technical Efficacy Stage: 2 J. Magn. Reson. Imaging 2019. J. Magn. Reson. Imaging 2020;52:998-1018.

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.

Association of Peritumoral Radiomics With Tumor Biology and Pathologic Response to Preoperative Targeted Therapy for HER2 (ERBB2)-Positive Breast Cancer

IMPORTANCE

There has been significant recent interest in understanding the utility of quantitative imaging to delineate breast cancer intrinsic biological factors and therapeutic response. No clinically accepted biomarkers are as yet available for estimation of response to human epidermal growth factor receptor 2 (currently known as ERBB2, but referred to as HER2 in this study)-targeted therapy in breast cancer.

OBJECTIVE

To determine whether imaging signatures on clinical breast magnetic resonance imaging (MRI) could noninvasively characterize HER2-positive tumor biological factors and estimate response to HER2-targeted neoadjuvant therapy.

DESIGN, SETTING, AND PARTICIPANTS

In a retrospective diagnostic study encompassing 209 patients with breast cancer, textural imaging features extracted within the tumor and annular peritumoral tissue regions on MRI were examined as a means to identify increasingly granular breast cancer subgroups relevant to therapeutic approach and response. First, among a cohort of 117 patients who received an MRI prior to neoadjuvant chemotherapy (NAC) at a single institution from April 27, 2012, through September 4, 2015, imaging features that distinguished HER2+ tumors from other receptor subtypes were identified. Next, among a cohort of 42 patients with HER2+ breast cancers with available MRI and RNaseq data accumulated from a multicenter, preoperative clinical trial (BrUOG 211B), a signature of the response-associated HER2-enriched (HER2-E) molecular subtype within HER2+ tumors (n = 42) was identified. The association of this signature with pathologic complete response was explored in 2 patient cohorts from different institutions, where all patients received HER2-targeted NAC (n = 28, n = 50). Finally, the association between significant peritumoral features and lymphocyte distribution was explored in patients within the BrUOG 211B trial who had corresponding biopsy hematoxylin-eosin-stained slide images. Data analysis was conducted from January 15, 2017, to February 14, 2019.

MAIN OUTCOMES AND MEASURES

Evaluation of imaging signatures by the area under the receiver operating characteristic curve (AUC) in identifying HER2+ molecular subtypes and distinguishing pathologic complete response (ypT0/is) to NAC with HER2-targeting.

RESULTS

In the 209 patients included (mean [SD] age, 51.1 [11.7] years), features from the peritumoral regions better discriminated HER2-E tumors (maximum AUC, 0.85; 95% CI, 0.79-0.90; 9-12 mm from the tumor) compared with intratumoral features (AUC, 0.76; 95% CI, 0.69-0.84). A classifier combining peritumoral and intratumoral features identified the HER2-E subtype (AUC, 0.89; 95% CI, 0.84-0.93) and was significantly associated with response to HER2-targeted therapy in both validation cohorts (AUC, 0.80; 95% CI, 0.61-0.98 and AUC, 0.69; 95% CI, 0.53-0.84). Features from the 0- to 3-mm peritumoral region were significantly associated with the density of tumor-infiltrating lymphocytes (R2 = 0.57; 95% CI, 0.39-0.75; P = .002).

CONCLUSIONS AND RELEVANCE

A combination of peritumoral and intratumoral characteristics appears to identify intrinsic molecular subtypes of HER2+ breast cancers from imaging, offering insights into immune response within the peritumoral environment and suggesting potential benefit for treatment guidance.