Predictors of Radiation-Induced Liver Disease in Eastern and Western Patients With Hepatocellular Carcinoma Undergoing Proton Beam Therapy.

PURPOSE: To identify predictors of radiation-induced liver disease (RILD) in patients with hepatocellular carcinoma (HCC) treated with proton beam therapy (PBT). METHODS: This multicenter study included 136 patients with HCC (eastern, n = 102; western, n = 34) without evidence of intrahepatic tumor progression after PBT. The RILD was defined as ascites with alkaline-phosphatase abnormality, grade ≥3 hepatic toxicity, or Child-Pugh score worsening by ≥2 within 4 months after PBT completion. The proton doses were converted to equivalent doses in 2-GyE fractions. The unirradiated liver volume (ULV) was defined as the absolute liver volume (LV) receiving <1 GyE; the standard liver volume (SLV) was calculated using body surface area. Possible correlations of clinicodosimetric parameters with RILD were examined. RESULTS: The mean pretreatment LV was 85% of SLV, and patients with a history of hepatectomy (P < .001) or hepatitis B virus infection (P = .035) had significantly smaller LV/SLV. Nineteen (14%) patients developed RILD. Multivariate logistic regression analysis identified ULV/SLV (P = .001), gross tumor volume (P = .001), and Child-Pugh classification (P = .002) as independent RILD predictors, and mean liver dose and target-delivered dose were not associated with RILD occurrence. A "volume-response" relationship between ULV/SLV and RILD was consistently observed in both eastern and western cohorts. In Child-Pugh class-A patients whose ULV/SLV were ≥50%, 49.9%-40%, 39.9%-30% and <30%, the RILD incidences were 0%, 6%, 16%, and 39% (P < .001), respectively. For the Child-Pugh class-B group, the RILD incidences in patients with ≥60%, 59.9%-40%, and <40% of ULV/SLV were 0%, 14%, and 83% (P = .006), respectively. CONCLUSIONS: The ULV/SLV, not mean liver dose, independently predicts RILD in patients with HCC undergoing PBT. The relative and absolute contraindications for Child-Pugh class-A patient's ULV/SLV are <50% and <30%, and <60% and <40% for Child-Pugh class-B patients, respectively. Our results indicate that the likelihood of hepatic complications for PBT is dictated by similar metrics as that for surgery.

High metastatic node number, not extranodal extension, as a node-related prognosticator in surgically treated patients with nodal metastatic salivary gland carcinoma.

BACKGROUND: The prognostic relevance of extranodal extension (ENE) for salivary gland carcinoma (SGC) remains unclear. The present study is undertaken to investigate the predictive significance of pathological nodal parameters in surgically treated patients with nodal metastatic SGC. METHODS: This multicenter cohort included 114 patients with pathologically proven node-positive SGC between 2000 and 2014. Possible correlations of clinicopathological parameters and outcomes were examined. RESULTS: The median follow-up was 69 months (range, 11-173 months). The multivariate analysis identified metastatic node number (1-2 vs 3-6; 1-2 vs ≥7) as an independent predictor for regional control (P = 0.005; P = 0.02), locoregional control (P = 0.008; P = 0.04), distant metastasis-free survival (P = 0.17; P = 0.006), disease-free survival (P = 0.05; P = 0.002), and overall survival (P = 0.18; P = 0.009), whereas ENE was not associated with survival outcomes. CONCLUSIONS: Metastatic node number, not ENE, is an independent node-related prognosticator for SGC. Integration of ENE into the American Joint Committee on Cancer 8th edition staging criteria may not improve prognostic performance.

High-dimensional regression analysis links magnetic resonance imaging features and protein expression and signaling pathway alterations in breast invasive carcinoma.

BACKGROUND: Imaging features derived from MRI scans can be used for not only breast cancer detection and measuring disease extent, but can also determine gene expression and patient outcomes. The relationships between imaging features, gene/protein expression, and response to therapy hold potential to guide personalized medicine. We aim to characterize the relationship between radiologist-annotated tumor phenotypic features (based on MRI) and the underlying biological processes (based on proteomic profiling) in the tumor. METHODS: Multiple-response regression of the image-derived, radiologist-scored features with reverse-phase protein array expression levels generated association coefficients for each combination of image-feature and protein in the RPPA dataset. Significantly-associated proteins for features were analyzed with Ingenuity Pathway Analysis software. Hierarchical clustering of the results of the pathway analysis determined which features were most strongly correlated with pathway activity and cellular functions. RESULTS: Each of the twenty-nine imaging features was found to have a set of significantly correlated molecules, associated biological functions, and pathways. CONCLUSIONS: We interrogated the pathway alterations represented by the protein expression associated with each imaging feature. Our study demonstrates the relationships between biological processes (via proteomic measurements) and MRI features within breast tumors.

Tumor image-derived texture features are associated with CD3 T-cell infiltration status in glioblastoma.

This study analyzed magnetic resonance imaging (MRI) scans of Glioblastoma (GB) patients to develop an imaging-derived predictive model for assessing the extent of intratumoral CD3 T-cell infiltration. Pre-surgical T1-weighted post-contrast and T2-weighted Fluid-Attenuated-Inversion-Recovery (FLAIR) MRI scans, with corresponding mRNA expression of CD3D/E/G were obtained through The Cancer Genome Atlas (TCGA) for 79 GB patients. The tumor region was contoured and 86 image-derived features were extracted across the T1-post contrast and FLAIR images. Six imaging features-kurtosis, contrast, small zone size emphasis, low gray level zone size emphasis, high gray level zone size emphasis, small zone high gray level emphasis-were found associated with CD3 activity and used to build a predictive model for CD3 infiltration in an independent data set of 69 GB patients (using a 50-50 split for training and testing). For the training set, the image-based prediction model for CD3 infiltration achieved accuracy of 97.1% and area under the curve (AUC) of 0.993. For the test set, the model achieved accuracy of 76.5% and AUC of 0.847. This suggests a relationship between image-derived textural features and CD3 T-cell infiltration enabling the non-invasive inference of intratumoral CD3 T-cell infiltration in GB patients, with potential value for the radiological assessment of response to immune therapeutics.