Radiogenomic Associations Clear Cell Renal Cell Carcinoma: An Exploratory Study.

INTRODUCTION: This study investigates how quantitative texture analysis can be used to non-invasively identify novel radiogenomic correlations with clear cell renal cell carcinoma (ccRCC) biomarkers. METHODS: The Cancer Genome Atlas-Kidney Renal Clear Cell Carcinoma open-source database was used to identify 190 sets of patient genomic data that had corresponding multiphase contrast-enhanced CT images in The Cancer Imaging Archive. 2,824 radiomic features spanning fifteen texture families were extracted from CT images using a custom-built MATLAB software package. Robust radiomic features with strong inter-scanner reproducibility were selected. Random forest, AdaBoost, and elastic net machine learning (ML) algorithms evaluated the ability of the selected radiomic features to predict the presence of 12 clinically relevant molecular biomarkers identified from the literature. ML analysis was repeated with cases stratified by stage (I/II vs. III/IV) and grade (1/2 vs. 3/4). 10-fold cross validation was used to evaluate model performance. RESULTS: Before stratification by tumor grade and stage, radiomics predicted the presence of several biomarkers with weak discrimination (AUC 0.60-0.68). Once stratified, radiomics predicted KDM5C, SETD2, PBRM1, and mTOR mutation status with acceptable to excellent predictive discrimination (AUC ranges from 0.70 to 0.86). CONCLUSIONS: Radiomic texture analysis can potentially identify a variety of clinically relevant biomarkers in patients with ccRCC and may have a prognostic implication.

CT-based radiomics stratification of tumor grade and TNM stage of clear cell renal cell carcinoma.

OBJECTIVES: To evaluate the utility of CT-based radiomics signatures in discriminating low-grade (grades 1-2) clear cell renal cell carcinomas (ccRCC) from high-grade (grades 3-4) and low TNM stage (stages I-II) ccRCC from high TNM stage (stages III-IV). METHODS: A total of 587 subjects (mean age 60.2 years ± 12.2; range 22-88.7 years) with ccRCC were included. A total of 255 tumors were high grade and 153 were high stage. For each subject, one dominant tumor was delineated as the region of interest (ROI). Our institutional radiomics pipeline was then used to extract 2824 radiomics features across 12 texture families from the manually segmented volumes of interest. Separate iterations of the machine learning models using all extracted features (full model) as well as only a subset of previously identified robust metrics (robust model) were developed. Variable of importance (VOI) analysis was performed using the out-of-bag Gini index to identify the top 10 radiomics metrics driving each classifier. Model performance was reported using area under the receiver operating curve (AUC). RESULTS: The highest AUC to distinguish between low- and high-grade ccRCC was 0.70 (95% CI 0.62-0.78) and the highest AUC to distinguish between low- and high-stage ccRCC was 0.80 (95% CI 0.74-0.86). Comparable AUCs of 0.73 (95% CI 0.65-0.8) and 0.77 (95% CI 0.7-0.84) were reported using the robust model for grade and stage classification, respectively. VOI analysis revealed the importance of neighborhood operation-based methods, including GLCM, GLDM, and GLRLM, in driving the performance of the robust models for both grade and stage classification. CONCLUSION: Post-validation, CT-based radiomics signatures may prove to be useful tools to assess ccRCC grade and stage and could potentially add to current prognostic models. Multiphase CT-based radiomics signatures have potential to serve as a non-invasive stratification schema for distinguishing between low- and high-grade as well as low- and high-stage ccRCC. KEY POINTS: • Radiomics signatures derived from clinical multiphase CT images were able to stratify low- from high-grade ccRCC, with an AUC of 0.70 (95% CI 0.62-0.78). • Radiomics signatures derived from multiphase CT images yielded discriminative power to stratify low from high TNM stage in ccRCC, with an AUC of 0.80 (95% CI 0.74-0.86). • Models created using only robust radiomics features achieved comparable AUCs of 0.73 (95% CI 0.65-0.80) and 0.77 (95% CI 0.70-0.84) to the model with all radiomics features in classifying ccRCC grade and stage, respectively.

Shape and texture-based radiomics signature on CT effectively discriminates benign from malignant renal masses.

OBJECTIVES: Using a radiomics framework to quantitatively analyze tumor shape and texture features in three dimensions, we tested its ability to objectively and robustly distinguish between benign and malignant renal masses. We assessed the relative contributions of shape and texture metrics separately and together in the prediction model. MATERIALS AND METHODS: Computed tomography (CT) images of 735 patients with 539 malignant and 196 benign masses were segmented in this retrospective study. Thirty-three shape and 760 texture metrics were calculated per tumor. Tumor classification models using shape, texture, and both metrics were built using random forest and AdaBoost with tenfold cross-validation. Sensitivity analyses on five sub-cohorts with respect to the acquisition phase were conducted. Additional sensitivity analyses after multiple imputation were also conducted. Model performance was assessed using AUC. RESULTS: Random forest classifier showed shape metrics featuring within the top 10% performing metrics regardless of phase, attaining the highest variable importance in the corticomedullary phase. Convex hull perimeter ratio is a consistently high-performing shape feature. Shape metrics alone achieved an AUC ranging 0.64-0.68 across multiple classifiers, compared with 0.67-0.75 and 0.68-0.75 achieved by texture-only and combined models, respectively. CONCLUSION: Shape metrics alone attain high prediction performance and high variable importance in the combined model, while being independent of the acquisition phase (unlike texture). Shape analysis therefore should not be overlooked in its potential to distinguish benign from malignant tumors, and future radiomics platforms powered by machine learning should harness both shape and texture metrics. KEY POINTS: • Current radiomics research is heavily weighted towards texture analysis, but quantitative shape metrics should not be ignored in their potential to distinguish benign from malignant renal tumors. • Shape metrics alone can attain high prediction performance and demonstrate high variable importance in the combined shape and texture radiomics model. • Any future radiomics platform powered by machine learning should harness both shape and texture metrics, especially since tumor shape (unlike texture) is independent of the acquisition phase and more robust from the imaging variations.

Quantitative Contour Analysis as an Image-based Discriminator Between Benign and Malignant Renal Tumors.

OBJECTIVE: To investigate whether morphologic analysis can differentiate between benign and malignant renal tumors on clinically acquired imaging. MATERIALS AND METHODS: Between 2009 and 2014, 3-dimensional tumor volumes were manually segmented from contrast-enhanced computerized tomography (CT) images from 150 patients with predominantly solid, nonmacroscopic fat-containing renal tumors: 100 renal cell carcinomas and 50 benign lesions (eg, oncocytoma and lipid-poor angiomyolipoma). Tessellated 3-dimensional tumor models were created from segmented voxels using MATLAB code. Eleven shape descriptors were calculated: sphericity, compactness, mean radial distance, standard deviation of the radial distance, radial distance area ratio, zero crossing, entropy, Feret ratio, convex hull area and convex hull perimeter ratios, and elliptic compactness. Morphometric parameters were compared using the Wilcoxon rank-sum test to investigate whether malignant renal masses demonstrate more morphologic irregularity than benign ones. RESULTS: Only CHP in sagittal orientation (median 0.96 vs 0.97) and EC in coronal orientation (median 0.92 vs 0.93) differed significantly between malignant and benign masses (P = .04). When comparing these 2 metrics between coronal and sagittal orientations, similar but nonsignificant trends emerged (P = .07). Other metrics tested were not significantly different in any imaging plane. CONCLUSION: Computerized image analysis is feasible using shape descriptors that otherwise cannot be visually assessed and used without quantification. Shape analysis via the transverse orientation may be reasonable, but encompassing all 3 planar dimensions to characterize tumor contour can achieve a more comprehensive evaluation. Two shape metrics (CHP and EC) may help distinguish benign from malignant renal tumors, an often challenging goal to achieve on imaging and biopsy.

Hypertrophic Osteoarthropathy: Clinical and Imaging Features.

Hypertrophic osteoarthropathy (HOA) is a medical condition characterized by abnormal proliferation of skin and periosteal tissues involving the extremities and characterized by three clinical features: digital clubbing (also termed Hippocratic fingers), periostosis of tubular bones, and synovial effusions. HOA can be a primary entity, known as pachydermoperiostosis, or can be secondary to extraskeletal conditions, with different prognoses and management implications for each. There is a high association between secondary HOA and malignancy, especially non-small cell lung cancer. In such cases, it can be considered a form of paraneoplastic syndrome. The most prevalent secondary causes of HOA are pulmonary in origin, which is why this condition was formerly referred to as hypertrophic pulmonary osteoarthropathy. HOA can also be associated with pleural, mediastinal, and cardiovascular causes, as well as extrathoracic conditions such as gastrointestinal tumors and infections, cirrhosis, and inflammatory bowel disease. Although the skeletal manifestations of HOA are most commonly detected with radiography, abnormalities can also be identified with other modalities such as computed tomography, magnetic resonance imaging, and bone scintigraphy. The authors summarize the pathogenesis, classification, causes, and symptoms and signs of HOA, including the genetics underlying the primary form (pachydermoperiostosis); describe key findings of HOA found at various imaging modalities, with examples of underlying causative conditions; and discuss features differentiating HOA from other causes of multifocal periostitis, such as thyroid acropachy, hypervitaminosis A, chronic venous insufficiency, voriconazole-induced periostitis, progressive diaphyseal dysplasia, and neoplastic causes such as lymphoma. ©RSNA, 2016.

Torsion, infarction, and rupture of a nongravid uterus: a complication of a large ovarian cyst.

Torsion of a nongravid uterus is rare, as most cases of uterine torsion occur during pregnancy. We report a case of a large ovarian cyst causing uterine torsion, infarction, and rupture. A 57-year-old woman presented with acute-onset abdominal pain and increasing abdominal girth over the past year. Contrast-enhanced computed tomography (CT) demonstrated axial rotation and swirling of the uterus and the mesenteric fat, leading to a preoperative diagnosis of uterine torsion. Laparotomy confirmed that the uterine corpus had undergone a 180-degree axial rotation, and further revealed uterine wall infarction and rupture into the endometrium as well as partial decapitation of the uterus from the cervix. The swirled appearance of the uterus, radiologically similar to the "whirlpool sign" seen in bowel volvulus, is an important CT finding to recognize, especially in view of the risk of irreversible ischemic complications this uncommon entity may inflict on the uterus.

Aberrant systemic arterial supply to normal lung arising from the proper hepatic artery discovered during transarterial chemoembolization.

We report a rare case of dual arterial supply to an otherwise normal lung discovered incidentally during initial angiography performed with the intent of chemoembolization of hepatocellular carcinoma. In addition to normal hepatic arterial supply, the proper hepatic artery provided systemic arterial supply to the lower lobe of the left lung. Subsequent chest computed tomography angiography demonstrated a normal tracheobronchial tree and normal pulmonary arterial supply to the lung. Although other anatomic variants have been reported, there are no other reported cases of systemic arterial supply from the proper hepatic artery to the lung. Identifying systemic arterial supply to the lung during angiography is important while performing transcatheter chemoembolization or radioembolization in the liver in order to minimize non-target embolization of the lung.

Poly(iohexol) nanoparticles as contrast agents for in vivo X-ray computed tomography imaging.

Biocompatible poly(iohexol) nanoparticles, prepared through cross-linking of iohexol and hexamethylene diisocyanate followed by coprecipitation of the resulting cross-linked polymer with mPEG-polylactide, were utilized as contrast agents for in vivo X-ray computed tomography (CT) imaging. Compared to conventional small-molecule contrast agents, poly(iohexol) nanoparticles exhibited substantially protracted retention within the tumor bed and a 36-fold increase in CT contrast 4 h post injection, which makes it possible to acquire CT images with improved diagnosis accuracy over a broad time frame without multiple administrations.

Transarterial sorafenib chemoembolization: preliminary study of technical feasibility in a rabbit model.

PURPOSE: To test the feasibility of targeted intraarterial administration of the tyrosine kinase inhibitor chemotherapeutic agent sorafenib to inhibit embolotherapy-induced tumor angiogenesis and reduce systemic drug side effects. MATERIALS AND METHODS: The left hepatic lobes of five New Zealand White rabbits (mean weight, 2.7 kg±0.2) were treated with chemoembolization with sorafenib and ethiodized oil emulsion, followed by immediate euthanasia. Postprocedure noncontrast computed tomography (CT) was used to evaluate intrahepatic chemotherapy mixture distribution. Liquid chromatography/tandem mass spectrometry (LC-MS/MS) was then used to directly measure sorafenib concentration in the treated liver tissue. Histopathologic assessment of treated left lobes was performed to identify any immediate toxic effects of the sorafenib solution. RESULTS: Lobar sorafenib chemoembolization was successfully performed in all cases via the left hepatic artery. Sorafenib and ethiodized oil (mean, 6.4 mg±3.8 and 0.95 mL±0.7, respectively) were injected, and CT confirmed targeted left hepatic lobe sorafenib emulsion delivery in all cases. Corresponding LC-MS/MS analysis yielded a mean sorafenib concentration of 94.2 µg/mL±48.3 in treated left lobe samples (n = 5), significantly greater than typical therapeutic drug levels (2-10 µg/mL) achieved with oral sorafenib systemic therapy. Histopathologic assessment showed only mild or moderate nonspecific ballooning degeneration in zone 3 hepatocytes, without tissue necrosis. CONCLUSIONS: Targeted transarterial sorafenib delivery is feasible and results in higher tissue drug levels than reported for systemic sorafenib therapy, without immediate histopathologic tissue toxicity. Future studies should aim to determine the utility of sorafenib chemoembolization in reducing hypoxia-induced vasculogenesis in liver tumors.

Transcatheter embolotherapy for gastrointestinal bleeding: a single center review of safety, efficacy, and clinical outcomes.

PURPOSE: The purpose of this study was to assess safety, efficacy, and clinical outcomes following transcatheter arterial embolization (TAE) of acute gastrointestinal (GI) bleeding. MATERIALS AND METHODS: Ninety-five patients (male:female ratio = 53:42, mean age 62 years) that underwent 95 TAEs for GI hemorrhage between 2002 and 2010 were retrospectively studied. Seventy-six of 95 (80 %) patients had upper GI bleeds and 19/95 (20 %) patients had lower GI bleeds. A mean of 7 (range 0-27) packed red blood cell units were transfused pre-procedure, and 90/95 (95 %) procedures were urgent or emergent. Twenty-seven of 95 (28 %) patients were hemodynamically unstable. Measured outcomes included procedure technical success, adverse events, and 30-day rebleeding and mortality rates. RESULTS: Bleeding etiology included peptic ulcer disease (45/95, 47 %), cancer (14/95, 15 %), diverticulosis (13/95, 14 %), and other (23/95, 24 %). Vessels embolized (n = 109) included gastroduodenal (42/109, 39 %), pancreaticoduodenal (22/109, 20 %), gastric (21/109, 19 %), superior mesenteric (12/109, 11 %), inferior mesenteric (8/109, 7 %), and splenic (4/109, 4 %) artery branches. Technical success with immediate hemostasis was achieved in 93/95 (98 %) cases. Most common embolic agents included coils (66/109, 61 %) and/or gelatin sponge (19/109, 17 %). Targeted versus empiric embolization were performed in 57/95 (60 %) and 38/95 (40 %) cases, respectively. Complications included bowel ischemia (4/95, 4 %) and coil migration in 3/95 (3 %). 30-day rebleeding rate was 23 % (22/95). Overall 30-day mortality rate was 18 % (16/89). Empiric embolization resulted in similar rebleeding (23 vs 24 %) but higher mortality (31 vs 9 %) rates compared to embolization for active extravasation. CONCLUSIONS: TAE controlled GI bleeding with high technical success, safety, and efficacy, and should be considered when endoscopic therapy is not feasible or unsuccessful.

Splenic artery embolization: a single center experience on the safety, efficacy, and clinical outcomes.

PURPOSE: We aimed to assess the safety, efficacy, and clinical outcomes of splenic artery embolization (SAE). MATERIALS AND METHODS: A total of 50 patients (male:female, 33:17; mean age, 49 years) who underwent 50 SAEs between 1998 and 2011 were retrospectively studied. The procedure indications included aneurysm or pseudoaneurysm (n=15), gastric variceal hemorrhage (n=15), preoperative reduction of surgical blood loss (n=9), or other (n=11). In total, 22 procedures were elective, and 28 procedures were urgent or emergent. The embolic agents included coils (n=50), gelatin sponges (n=15), and particles (n=4). The measured outcomes were the technical success of the procedure, efficacy, side effects, and the 30-day morbidity and mortality rates. RESULTS: All embolizations were technically successful. The procedure efficacy was 90%; five patients (10%) had a recurrent hemorrhage requiring a secondary intervention. Side effects included hydrothorax (n=26, 52%), thrombocytosis (n=16, 32%), thrombocytopenia (n=13, 26%), and postembolization syndrome (n=11, 22%). Splenic infarcts occurred in 13 patients (26%). The overall and procedure-specific 30-day morbidity rates were 38% (19/50) and 14% (splenoportal thrombosis, 3/50; encapsulated bacterial infection, 1/50; splenic abscess, 1/50; femoral hematoma requiring surgery, 1/50; hydrothorax requiring drainage, 1/50). The overall and procedure-specific 30-day mortality rates were 8% (4/50) and 0%. The multivariate analysis showed that advanced patient age (P = 0.037), postprocedure thrombocytopenia (P = 0.008), postprocedure hydrothorax (P = 0.009), and the need for a secondary intervention (P = 0.004) predicted the 30-day morbidity, while renal insufficiency (P < 0.0001), preprocedure hemodynamic instability (P = 0.044), and preprocedure leukocytosis (P < 0.0001) were prognostic factors for the 30-day mortality. CONCLUSION: SAE was performed with high technical success and efficacy, but the outcomes showed nontrivial morbidity rates. Elderly patients with thrombocytopenia and hydrothorax after SAE, and patients who require secondary interventions, should be monitored for complications.

Quantitative morphometric analysis of hepatocellular carcinoma: development of a programmed algorithm and preliminary application.

PURPOSE: The quantitative relationship between tumor morphology and malignant potential has not been explored in liver tumors. We designed a computer algorithm to analyze shape features of hepatocellular carcinoma (HCC) and tested feasibility of morphologic analysis. MATERIALS AND METHODS: Cross-sectional images from 118 patients diagnosed with HCC between 2007 and 2010 were extracted at the widest index tumor diameter. The tumor margins were outlined, and point coordinates were input into a MATLAB (MathWorks Inc., Natick, Massachusetts, USA) algorithm. Twelve shape descriptors were calculated per tumor: the compactness, the mean radial distance (MRD), the RD standard deviation (RDSD), the RD area ratio (RDAR), the zero crossings, entropy, the mean Feret diameter (MFD), the Feret ratio, the convex hull area (CHA) and perimeter (CHP) ratios, the elliptic compactness (EC), and the elliptic irregularity (EI). The parameters were correlated with the levels of alpha-fetoprotein (AFP) as an indicator of tumor aggressiveness. RESULTS: The quantitative morphometric analysis was technically successful in all cases. The mean parameters were as follows: compactness 0.88±0.086, MRD 0.83±0.056, RDSD 0.087±0.037, RDAR 0.045±0.023, zero crossings 6±2.2, entropy 1.43±0.16, MFD 4.40±3.14 cm, Feret ratio 0.78±0.089, CHA 0.98±0.027, CHP 0.98±0.030, EC 0.95±0.043, and EI 0.95±0.023. MFD and RDAR provided the widest value range for the best shape discrimination. The larger tumors were less compact, more concave, and less ellipsoid than the smaller tumors (P < 0.0001). AFP-producing tumors displayed greater morphologic irregularity based on several parameters, including compactness, MRD, RDSD, RDAR, entropy, and EI (P < 0.05 for all). CONCLUSION: Computerized HCC image analysis using shape descriptors is technically feasible. Aggressively growing tumors have wider diameters and more irregular margins. Future studies will determine further clinical applications for this morphologic analysis.

Safety and efficacy of StarClose SE Vascular Closure System in high-risk liver interventional oncology patients.

PURPOSE: To assess the safety and efficacy of the StarClose SE Vascular Closure System (Abbott Vascular, Abbott Park IL, USA) in high-risk thrombocytopenic and coagulopathic interventional oncology (IO) patients. METHODS: In this single institution retrospective study, 63 high-risk thrombocytopenic or coagulopathic IO patients (M:F=51:12, mean age 58 years, range 31-88 years) who underwent 83 common femoral arteriotomy closures using the StarClose device were identified among all IO patients (n=131) undergoing StarClose closure (n=177) between 2008-2011. High-risk thrombocytopenia and coagulopathy were defined as platelet count ≤100 10(3)/mL and international normalized ratio (INR) ≥1.5. Procedures included chemoembolization (n=67), radioembolization (n=8), and hepatic arterial mapping with technetium-99m macroaggrated albumin administration (n=8) for treatment of hepatocellular carcinoma (n=79) or liver metastases (n=4). Measured outcomes included technical success of arterial closure and closure-related adverse events, graded according to the Society of Interventional Radiology classification. RESULTS: In all cases, 5 French common femoral arterial access was used. Platelet count was ≤100 10(3)/mL in 80/83 (96.4%) cases and INR was ≥1.5 in 35/83 (42.2%) cases. Mean pre-procedure platelet count was 71 (range 26-347) 10(3)/mL and mean INR was 1.4 (range 1.0-2.1). The StarClose device effectively sealed the arteriotomy in 83/83 (100%) cases, 60/83 (72.3%) cases were first-time closures, and 20/83 (24.1%) cases were repeat closures. Small groin hematomas, graded as class A minor complications, developed in 3/83 (3.6%) cases. No other complications were encountered. CONCLUSIONS: The StarClose SE Vascular Closure System confers high technical success and safety in common femoral arteriotomy closure in high-risk IO patients.

Current tumor ablation technologies: basic science and device review.

Image-guided tumor ablation is an increasingly utilized tool to treat focal malignancy. Tumor ablation can be divided into two large categories, thermal and chemical ablation. The authors provide an overview of the current methods used to achieve thermal and chemical ablation of tumors, specifically addressing the basic science behind the ablation methods as well as providing a brief synopsis of the commercial devices currently available for use in the United States.