Quantitative in vivo proton MR spectroscopic assessment of lipid metabolism: Value for breast cancer diagnosis and prognosis.

BACKGROUND: Breast magnetic resonance spectroscopy (1 H-MRS) has been largely based on choline metabolites; however, other relevant metabolites can be detected and monitored. PURPOSE: To investigate whether lipid metabolite concentrations detected with 1 H-MRS can be used for the noninvasive differentiation of benign and malignant breast tumors, differentiation among molecular breast cancer subtypes, and prediction of long-term survival outcomes. STUDY TYPE: Retrospective. SUBJECTS: In all, 168 women, aged ≥18 years. FIELD STRENGTH/SEQUENCE: Dynamic contrast-enhanced MRI at 1.5 T: sagittal 3D spoiled gradient recalled sequence with fat saturation, flip angle = 10°, repetition time / echo time (TR/TE) = 7.4/4.2 msec, slice thickness = 3.0 mm, field of view (FOV) = 20 cm, and matrix size = 256 × 192. 1 H-MRS: PRESS with TR/TE = 2000/135 msec, water suppression, and 128 scan averages, in addition to 16 reference scans without water suppression. ASSESSMENT: MRS quantitative analysis of lipid resonances using the LCModel was performed. Histopathology was the reference standard. STATISTICAL TESTS: Categorical data were described using absolute numbers and percentages. For metric data, means (plus 95% confidence interval [CI]) and standard deviations as well as median, minimum, and maximum were calculated. Due to skewed data, the latter were more adequate; unpaired Mann-Whitney U-tests were performed to compare groups without and with Bonferroni correction. ROC analyses were also performed. RESULTS: There were 111 malignant and 57 benign lesions. Mean voxel size was 4.4 ± 4.6 cm3 . Six lipid metabolite peaks were quantified: L09, L13 + L16, L21 + L23, L28, L41 + L43, and L52 + L53. Malignant lesions showed lower L09, L21 + L23, and L52 + L53 than benign lesions (P = 0.022, 0.027, and 0.0006). Similar results were observed for Luminal A or Luminal A/B vs. other molecular subtypes. At follow-up, patients were split into two groups based on median values for the six peaks; recurrence-free survival was significantly different between groups for L09, L21 + L23, and L28 (P = 0.0173, 0.0024, and 0.0045). DATA CONCLUSION: Quantitative in vivo 1 H-MRS assessment of lipid metabolism may provide an additional noninvasive imaging biomarker to guide therapeutic decisions in breast cancer. LEVEL OF EVIDENCE: 3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:239-249.

Early posttreatment assessment of MRI perfusion biomarkers can predict long-term response of lung cancer brain metastases to stereotactic radiosurgery.

Background: Imaging criteria to evaluate the response of brain metastases to stereotactic radiosurgery (SRS) in the early posttreatment period remains a crucial unmet need. The aim of this study is to correlate early (within 12 wk) posttreatment perfusion MRI changes with long-term outcomes after treatment of lung cancer brain metastases with SRS. Methods: Pre- and posttreatment perfusion MRI scans were obtained in patients treated with SRS for intact non-small cell lung cancer brain metastases. Time-dependent leakage (Ktrans), blood plasma volume (Vp), and extracellular extravascular volume (Ve) were calculated for each lesion. Patients were followed longitudinally with serial MRI until death, progression, or intervention (whole brain radiation or surgery). Results: We included 53 lesions treated with SRS from 41 total patients. Median follow-up after treatment was 11 months. Actuarial local control at one year was 85%. Univariate analysis demonstrated a significant difference (P = 0.032) in posttreatment Ktrans SD between patients with progressive disease (mean = 0.0317) and without progressive disease (mean = 0.0219). A posttreatment Ktrans SD cutoff value of 0.017 was highly sensitive (89%) for predicting progressive disease and no progressive disease. Early posttreatment volume change was not associated with outcome (P = 0.941). Conclusion: Posttreatment Ktrans SD may be used as an early posttreatment imaging biomarker to help predict long-term response of lung cancer brain metastases to SRS. This can help identify patients who will ultimately fail SRS and allow for timelier adjustment in treatment approach. These data should be prospectively validated in larger patient cohorts and other histologies.

Genome differentiation of Drosophila melanogaster from a microclimate contrast in Evolution Canyon, Israel.

The opposite slopes of "Evolution Canyon" in Israel have served as a natural model system of adaptation to a microclimate contrast. Long-term studies of Drosophila melanogaster populations inhabiting the canyon have exhibited significant interslope divergence in thermal and drought stress resistance, candidate genes, mobile elements, habitat choice, mating discrimination, and wing-shape variation, all despite close physical proximity of the contrasting habitats, as well as substantial interslope migration. To examine patterns of genetic differentiation at the genome-wide level, we used high coverage sequencing of the flies' genomes. A total of 572 genes were significantly different in allele frequency between the slopes, 106 out of which were associated with 74 significantly overrepresented gene ontology (GO) terms, particularly so with response to stimulus and developmental and reproductive processes, thus corroborating previous observations of interslope divergence in stress response, life history, and mating functions. There were at least 37 chromosomal "islands" of interslope divergence and low sequence polymorphism, plausible signatures of selective sweeps, more abundant in flies derived from one (north-facing) of the slopes. Positive correlation between local recombination rate and the level of nucleotide polymorphism was also found.

A Bayesian network approach for modeling local failure in lung cancer.

Locally advanced non-small cell lung cancer (NSCLC) patients suffer from a high local failure rate following radiotherapy. Despite many efforts to develop new dose-volume models for early detection of tumor local failure, there was no reported significant improvement in their application prospectively. Based on recent studies of biomarker proteins' role in hypoxia and inflammation in predicting tumor response to radiotherapy, we hypothesize that combining physical and biological factors with a suitable framework could improve the overall prediction. To test this hypothesis, we propose a graphical Bayesian network framework for predicting local failure in lung cancer. The proposed approach was tested using two different datasets of locally advanced NSCLC patients treated with radiotherapy. The first dataset was collected retrospectively, which comprises clinical and dosimetric variables only. The second dataset was collected prospectively in which in addition to clinical and dosimetric information, blood was drawn from the patients at various time points to extract candidate biomarkers as well. Our preliminary results show that the proposed method can be used as an efficient method to develop predictive models of local failure in these patients and to interpret relationships among the different variables in the models. We also demonstrate the potential use of heterogeneous physical and biological variables to improve the model prediction. With the first dataset, we achieved better performance compared with competing Bayesian-based classifiers. With the second dataset, the combined model had a slightly higher performance compared to individual physical and biological models, with the biological variables making the largest contribution. Our preliminary results highlight the potential of the proposed integrated approach for predicting post-radiotherapy local failure in NSCLC patients.

An extended Markov blanket approach to proteomic biomarker detection from high-resolution mass spectrometry data.

High-resolution matrix-assisted laser desorption/ionization time-of-flight mass spectrometry has recently shown promise as a screening tool for detecting discriminatory peptide/protein patterns. The major computational obstacle in finding such patterns is the large number of mass/charge peaks (features, biomarkers, data points) in a spectrum. To tackle this problem, we have developed methods for data preprocessing and biomarker selection. The preprocessing consists of binning, baseline correction, and normalization. An algorithm, extended Markov blanket, is developed for biomarker detection, which combines redundant feature removal and discriminant feature selection. The biomarker selection couples with support vector machine to achieve sample prediction from high-resolution proteomic profiles. Our algorithm is applied to recurrent ovarian cancer study that contains platinum-sensitive and platinum-resistant samples after treatment. Experiments show that the proposed method performs better than other feature selection algorithms. In particular, our algorithm yields good performance in terms of both sensitivity and specificity as compared to other methods.

Diagnosis of early relapse in ovarian cancer using serum proteomic profiling.

Surface-enhanced laser desorption/ionization time-of-flight (SELDI-TOF) mass spectrometry data has been increasingly analyzed for identifying biomarkers to help early detection of the disease. Ovarian cancer commonly recurs at the rate of 75% within a few months or several years later after standard treatment. Since recurrent ovarian cancer is relatively difficult to be diagnosed and small tumors generally respond better to treatment, new methods for the detection of early relapse in ovarian cancer are urgently needed. Here, we propose a new algorithm SVM-MB/RFE (SVM-Markov Blanket/Recursive Feature Elimination) based on SVM-RFE, which identifies biomarkers for predicting the early recurrence of ovarian cancer. In this approach, we first apply t-test for feature pruning and then binning using 5-fold cross validation. Finally, 58 peaks are obtained from 27,000 of the raw data. Such dramatically reduced features relax the computational burden in the next step of our algorithm. We compare the performance of three feature selection algorithms and demonstrate that SVM-MB/RFE outperforms other methods.