Assessing tumor angiogenesis using dynamic contrast-enhanced integrated magnetic resonance-positron emission tomography in patients with non-small-cell lung cancer

MRI-Based Kinetic Heterogeneity Evaluation in the Accurate Access of Axillary Lymph Node Status in Breast Cancer Using a Hybrid CNN-RNN Model

BACKGROUND

Accurate evaluation of the axillary lymph node (ALN) status is needed for determining the treatment protocol for breast cancer (BC). The value of magnetic resonance imaging (MRI)-based tumor heterogeneity in assessing ALN metastasis in BC is unclear.

PURPOSE

To assess the value of deep learning (DL)-derived kinetic heterogeneity parameters based on BC dynamic contrast-enhanced (DCE)-MRI to infer the ALN status.

STUDY TYPE

Retrospective.

SUBJECTS

1256/539/153/115 patients in the training cohort, internal validation cohort, and external validation cohorts I and II, respectively.

FIELD STRENGTH/SEQUENCE

1.5 T/3.0 T, non-contrast T1-weighted spin-echo sequence imaging (T1WI), DCE-T1WI, and diffusion-weighted imaging.

ASSESSMENT

Clinical pathological and MRI semantic features were obtained by reviewing histopathology and MRI reports. The segmentation of the tumor lesion on the first phase of T1WI DCE-MRI images was applied to other phases after registration. A DL architecture termed convolutional recurrent neural network (ConvRNN) was developed to generate the KHimage (kinetic heterogeneity of DCE-MRI image) score that indicated the ALN status in patients with BC. The model was trained and optimized on training and internal validation cohorts, tested on two external validation cohorts. We compared ConvRNN model with other 10 models and the subgroup analyses of tumor size, magnetic field strength, and molecular subtype were also evaluated.

STATISTICAL TESTS

Chi-squared, Fisher's exact, Student's t, Mann-Whitney U tests, and receiver operating characteristics (ROC) analysis were performed. P < 0.05 was considered significant.

RESULTS

The ConvRNN model achieved area under the curve (AUC) of 0.802 in the internal validation cohort and 0.785-0.806 in the external validation cohorts. The ConvRNN model could well evaluate the ALN status of the four molecular subtypes (AUC = 0.685-0.868). The patients with larger tumor sizes (>5 cm) were more susceptible to ALN metastasis with KHimage scores of 0.527-0.827.

DATA CONCLUSION

A ConvRNN model outperformed traditional models for determining the ALN status in patients with BC.

LEVEL OF EVIDENCE

3 TECHNICAL EFFICACY: Stage 2.

Registration on DCE-MRI images via multi-domain image-to-image translation

Registration of dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) is challenging as rapid intensity changes caused by a contrast agent lead to large registration errors. To address this problem, we propose a novel multi-domain image-to-image translation (MDIT) network based on image disentangling for separating motion from contrast changes before registration. In particular, the DCE images are disentangled into a domain-invariant content space (motion) and a domain-specific attribute space (contrast changes). The disentangled representations are then used to generate images, where the contrast changes have been removed from the motion. After that the resulting deformations can be directly derived from the generated images using an FFD registration. The method is tested on 10 lung DCE-MRI cases. The proposed method reaches an average root mean squared error of 0.3 ± 0.41 and the separation time is about 2.4 s for each case. Results show that the proposed method improves the registration efficiency without losing the registration accuracy compared with several state-of-the-art registration methods.

Serum cytokines predict efficacy and toxicity, but are not useful for disease monitoring in lung cancer treated with PD-(L)1 inhibitors

Introduction

PD-(L)1 inhibitors (IO) have improved the prognosis of non-small-cell lung cancer (NSCLC), but more reliable predictors of efficacy and immune-related adverse events (irAE) are urgently needed. Cytokines are important effector molecules of the immune system, whose potential clinical utility as biomarkers remains unclear.

Methods

Serum samples from patients with advanced NSCLC receiving IO either alone in the first (1L, n=46) and subsequent lines (n=50), or combined with chemotherapy (ICT, n=108) were analyzed along with age-matched healthy controls (n=15) at baseline, after 1 and 4 therapy cycles, and at disease progression (PD). Patients were stratified in rapid progressors (RP, progression-free survival [PFS] <120 days), and long-term responders (LR, PFS >200 days). Cytometric bead arrays were used for high-throughput quantification of 20 cytokines and other promising serum markers based on extensive search of the current literature.

Results

Untreated NSCLC patients had increased levels of various cytokines and chemokines, like IL-6, IL-8, IL-10, CCL5, G-CSF, ICAM-1, TNF-RI and VEGF (fold change [FC]=1.4-261, p=0.026-9x10^-7) compared to age-matched controls, many of which fell under ICT (FC=0.2-0.6, p=0.014-0.002), but not under IO monotherapy. Lower baseline levels of TNF-RI were associated with longer PFS (hazard ratio [HR]= 0.42-0.54; p=0.014-0.009) and overall survival (HR=0.28-0.34, p=0.004-0.001) after both ICT and IO monotherapy. Development of irAE was associated with higher baseline levels of several cytokines, in particular of IL-1β and angiogenin (FC=7-9, p=0.009-0.0002). In contrast, changes under treatment were very subtle, there were no serum correlates of radiologic PD, and no association between dynamic changes in cytokine concentrations and clinical outcome. No relationship was noted between the patients' serologic CMV status and serum cytokine levels.

Conclusions

Untreated NSCLC is characterized by increased blood levels of several pro-inflammatory and angiogenic effectors, which decrease under ICT. Baseline serum cytokine levels could be exploited for improved prediction of subsequent IO benefit (in particular TNF-RI) and development of irAE (e.g. IL-1β or angiogenin), but they are not suitable for longitudinal disease monitoring. The potential utility of IL-1/IL-1β inhibitors in the management and/or prevention of irAE in NSCLC warrants investigation.

Upregulation of CCNB2 and Its Perspective Mechanisms in Cerebral Ischemic Stroke and All Subtypes of Lung Cancer: A Comprehensive Study

Cyclin B2 (CCNB2) belongs to type B cell cycle family protein, which is located on chromosome 15q22, and it binds to cyclin-dependent kinases (CDKs) to regulate their activities. In this study, 103 high-throughput datasets related to all subtypes of lung cancer (LC) and cerebral ischemic stroke (CIS) with the data of CCNB2 expression were collected. The analysis of standard mean deviation (SMD) and summary receiver operating characteristic (SROC) reflecting expression status demonstrated significant up-regulation of CCNB2 in LC and CIS (Lung adenocarcinoma: SMD = 1.40, 95%CI [0.98–1.83], SROC = 0.92, 95%CI [0.89–0.94]. Lung squamous cell carcinoma: SMD = 2.56, 95%CI [1.64–3.48]. SROC = 0.97, 95%CI [0.95–0.98]. Lung small cell carcinoma: SMD = 3.01, 95%CI [2.01–4.01]. SROC = 0.98, 95%CI [0.97–0.99]. CIS: SMD = 0.29, 95%CI [0.05–0.53], SROC = 0.68, 95%CI [0.63–0.71]). Simultaneously, protein-protein interaction (PPI) analysis indicated that CCNB2 is the hub molecule of crossed high-expressed genes in CIS and LC. Through Multiscale embedded gene co-expression network analysis (MEGENA), a gene module of CIS including 76 genes was obtained and function enrichment analysis of the CCNB2 module genes implied that CCNB2 may participate in the processes in the formation of CIS and tissue damage caused by CIS, such as “cell cycle,” “protein kinase activity,” and “glycosphingolipid biosynthesis.” Afterward, via single-cell RNA-seq analysis, CCNB2 was found up-regulated on GABAergic neurons in brain organoids as well as T cells expressing proliferative molecules in LUAD. Concurrently, the expression of CCNB2 distributed similarly to TOP2A as a module marker of cell proliferation in cell cluster. These findings can help in the field of the pathogenesis of LC-related CIS and neuron repair after CIS damage.

PET/MR Imaging of Somatostatin Receptor Expression and Tumor Vascularity in Meningioma: Implications for Pathophysiology and Tumor Outcomes

Background and Purpose

Meningiomas, the most common primary intracranial tumor, are vascular neoplasms that express somatostatin receptor-2 (SSTR2). The purpose of this investigation was to evaluate if a relationship exists between tumor vascularity and SSTR2 expression, which may play a role in meningioma prognostication and clinical management.

Materials and Methods

Gallium-68-DOTATATE PET/MRI with dynamic contrast-enhanced (DCE) perfusion was prospectively performed. Clinical and demographic patient characteristics were recorded. Tumor volumes were segmented and superimposed onto parametric DCE maps including flux rate constant (Kep), transfer constant (Ktrans), extravascular volume fraction (Ve), and plasma volume fraction (Vp). Meningioma PET standardized uptake value (SUV) and SUV ratio to superior sagittal sinus (SUVR_SSS) were recorded. Pearson correlation analyses were performed. In a random subset, analysis was repeated by a second investigator, and intraclass correlation coefficients (ICCs) were determined.

Results

Thirty-six patients with 60 meningiomas (20 WHO-1, 27 WHO-2, and 13 WHO-3) were included. Mean Kep demonstrated a strong significant positive correlation with SUV (r = 0.84, p < 0.0001) and SUVR_SSS (r = 0.81, p < 0.0001). When stratifying by WHO grade, this correlation persisted in WHO-2 (r = 0.91, p < 0.0001) and WHO-3 (r = 0.92, p = 0.0029) but not WHO-1 (r = 0.26, p = 0.4, SUVR_SSS). ICC was excellent (0.97–0.99).

Conclusion

DOTATATE PET/MRI demonstrated a strong significant correlation between tumor vascularity and SSTR2 expression in WHO-2 and WHO-3, but not WHO-1 meningiomas, suggesting biological differences in the relationship between tumor vascularity and SSTR2 expression in higher-grade meningiomas, the predictive value of which will be tested in future work.