Short-term mortality prediction in acute pulmonary embolism: Radiomics values of skeletal muscle and intramuscular adipose tissue.

BACKGROUND: Acute pulmonary embolism (APE) is a potentially life-threatening disorder, emphasizing the importance of accurate risk stratification and survival prognosis. The exploration of imaging biomarkers that can reflect patient survival holds the potential to further enhance the stratification of APE patients, enabling personalized treatment and early intervention. Therefore, in this study, we develop computed tomography pulmonary angiography (CTPA) radiomic signatures for the prognosis of 7- and 30-day all-cause mortality in patients with APE. METHODS: Diagnostic CTPA images from 829 patients with APE were collected. Two hundred thirty-four features from each skeletal muscle (SM), intramuscular adipose tissue (IMAT) and both tissues combined (SM + IMAT) were calculated at the level of thoracic vertebra 12. Radiomic signatures were derived using 10 times repeated three-fold cross-validation on the training data for SM, IMAT and SM + IMAT for predicting 7- and 30-day mortality independently. The performance of the radiomic signatures was then evaluated on held-out test data and compared with the simplified pulmonary embolism severity index (sPESI) score, a well-established biomarker for risk stratification in APE. Predictive accuracy was assessed by the area under the receiver operating characteristic curve (AUC) with a 95% confidence interval (CI), sensitivity and specificity. RESULTS: The radiomic signatures based on IMAT and a combination of SM and IMAT (SM + IMAT) achieved moderate performance for the prediction of 30-day mortality on test data (IMAT: AUC = 0.68, 95% CI [0.57-0.78], sensitivity = 0.57, specificity = 0.73; SM + IMAT: AUC = 0.70, 95% CI [0.60-0.79], sensitivity = 0.74, specificity = 0.54). Radiomic signatures developed for predicting 7-day all-cause mortality showed overall low performance. The clinical signature, that is, sPESI, achieved slightly better performance in terms of AUC on test data compared with the radiomic signatures for the prediction of both 7- and 30-day mortality on the test data (7 days: AUC = 0.73, 95% CI [0.67-0.79], sensitivity = 0.92, specificity = 0.16; 30 days: AUC = 0.74, 95% CI [0.66-0.82], sensitivity = 0.97, specificity = 0.16). CONCLUSIONS: We developed and tested radiomic signatures for predicting 7- and 30-day all-cause mortality in APE using a multicentric retrospective dataset. The present multicentre work shows that radiomics parameters extracted from SM and IMAT can predict 30-day all-cause mortality in patients with APE.

Primary renal sarcomas: imaging features and discrimination from non-sarcoma renal tumors.

OBJECTIVES: To assess imaging features of primary renal sarcomas in order to better discriminate them from non-sarcoma renal tumors. METHODS: Adult patients diagnosed with renal sarcomas from 1995 to 2018 were included from 11 European tertiary referral centers (Germany, Belgium, Turkey). Renal sarcomas were 1:4 compared to patients with non-sarcoma renal tumors. CT/MRI findings were assessed using 21 predefined imaging features. A random forest model was trained to predict "renal sarcoma vs. non-sarcoma renal tumors" based on demographics and imaging features. RESULTS: n = 34 renal sarcomas were included and compared to n = 136 non-sarcoma renal tumors. Renal sarcomas manifested in younger patients (median 55 vs. 67 years, p < 0.01) and were more complex (high RENAL score complexity 79.4% vs. 25.7%, p < 0.01). Renal sarcomas were larger (median diameter 108 vs. 43 mm, p < 0.01) with irregular shape and ill-defined margins, and more frequently demonstrated invasion of the renal vein or inferior vena cava, tumor necrosis, direct invasion of adjacent organs, and contact to renal artery or vein, compared to non-sarcoma renal tumors (p < 0.05, each). The random forest algorithm yielded a median AUC = 93.8% to predict renal sarcoma histology, with sensitivity, specificity, and positive predictive value of 90.4%, 76.5%, and 93.9%, respectively. Tumor diameter and RENAL score were the most relevant imaging features for renal sarcoma identification. CONCLUSION: Renal sarcomas are rare tumors commonly manifesting as large masses in young patients. A random forest model using demographics and imaging features shows good diagnostic accuracy for discrimination of renal sarcomas from non-sarcoma renal tumors, which might aid in clinical decision-making. KEY POINTS: • Renal sarcomas commonly manifest in younger patients as large, complex renal masses. • Compared to non-sarcoma renal tumors, renal sarcomas more frequently demonstrated invasion of the renal vein or inferior vena cava, tumor necrosis, direct invasion of adjacent organs, and contact to renal artery or vein. • Using demographics and standardized imaging features, a random forest showed excellent diagnostic performance for discrimination of sarcoma vs. non-sarcoma renal tumors (AUC = 93.8%, sensitivity = 90.4%, specificity = 76.5%, and PPV = 93.9%).

Dental Trauma on whole Body Trauma CT-An underreported finding.

BACKGROUND: The prevalence of dental injuries (DI) in polytrauma patients is unknown. The purpose of our study was to identify the frequency of dental injuries on whole body CTs acquired in a trauma setting and to estimate how often they are correctly reported by the radiologist. METHODS: In the time period between 2006 and 2018 the radiological database of one university hospital was screened for whole-body trauma CTs. A total of 994 CTs were identified and re-evaluated. RESULTS: Dental injuries were identified in 127 patients (12.8% of patients). There were 27 women (21.3%) and 100 men (78.7%) with a mean age of 51.0 ± 18.9 years (range 10-96 years). Regarding localization, most findings involved the molars (n = 107, 37.4%), followed by the incisors (n = 81, 28.3%), premolars (n = 59, 20.6%) and canines (n = 39, 13.7%). Most common findings were as follows: luxations (n = 49, 45.8%), followed by crown fractures (n = 46, 43%), root fractures (n = 10, 9.3%), extrusions (n = 1, 0.9%), and intrusions (n = 1, 0.9%). Only 15 findings (11.8% of all patients with dental injuries) were described in the original radiological reports. CONCLUSION: DI had a high occurrence in polytrauma patients. A high frequency of underreported dental trauma findings was identified. Radiologists reporting whole-body trauma CT should be aware of possible dental trauma to report the findings adequately.

Clinical and imaging features of myeloid sarcoma: a German multicenter study.

BACKGROUND: Myeloid sarcoma (MS), also known as chloroma, is an extramedullary manifestation of malignant primitive myeloid cells. Previously, only small studies investigated clinical and imaging features of MS. The purpose of this study was to elucidate clinical and imaging features of MS based upon a multicenter patient sample. METHODS: Patient records of radiological databases of 4 German university hospitals were retrospectively screened for MS in the time period 01/2001 and 06/2019. Overall, 151 cases/76 females (50.3%) with a mean age of 55.5 ± 15.1 years and 183 histopathological confirmation or clinically suspicious lesions of MS were included into this study. The underlying hematological disease, localizations, and clinical symptoms as well as imaging features on CT and MRI were investigated. RESULTS: In 15 patients (9.9% of all 151 cases) the manifestation of MS preceded the systemic hematological disease. In 43 cases (28.4%), first presentation of MS occurred simultaneously with the initial diagnosis of leukemia, and 92 (60.9%) patients presented MS after the initial diagnosis. In 37 patients (24.5%), the diagnosis was made incidentally by imaging. Clinically, cutaneous lesions were detected in 35 of 151 cases (23.2%). Other leading symptoms were pain (n = 28/151, 18.5%), neurological deficit (n = 27/151, 17.9%), swelling (n = 14/151, 9.3%) and dysfunction of the affected organ (n = 10/151, 6.0%). Most commonly, skin was affected (n = 30/151, 16.6%), followed by bone (n = 29/151, 16.0%) and lymphatic tissue (n = 21/151, 11.4%). Other localizations were rare. On CT, most lesions were homogenous. On T2-weighted imaging, most of the lesions were hyperintense. On T1-weighted images, MS was hypointense in n = 22/54 (40.7%) and isointense in n = 30/54 (55.6%). A diffusion restriction was identified in most cases with a mean ADC value of 0.76 ± 0.19 × 10- 3 mm2/s. CONCLUSIONS: The present study shows clinical and imaging features of MS based upon a large patient sample in a multicenter design. MS occurs in most cases meta-chronous to the hematological disease and most commonly affects the cutis. One fourth of cases were identified incidentally on imaging, which needs awareness of the radiologists for possible diagnosis of MS.

Radiological findings on irradiation-CT in patients with non metastatic breast cancer.

OBJECTIVE: The aim of this study was to evaluate the number of incidental findings on native CT treatment-planning scans for radiation in breast cancer patients. METHODS: The treatment-planning scans of 382 patients with non-metastatic breast cancer were retrospectively analyzed for additional findings. The planning scan area covered the entire thorax and the upper part of the abdomen. Incidental findings were classified according to their clinical relevance. RESULTS: Overall 892 incidental findings were detected in the CT treatment-planning scans (mean 2.34 findings per patient). Only a small proportion of patients (n = 63, 16.4%) had no finding. Most findings were located in the thorax (683, 76.57%), and 209 findings (23.43%) were abdominal. 79 findings (8.87%) were of major clinical relevance, 232 findings (26.01%) were of moderate clinical relevance and 580 findings (65.02%) were of minor clinical relevance. Most clinically relevant findings were in the thorax (p = 0.006). Abdominal findings were more of significantly minor clinical relevance (p < 0.0001). CONCLUSION: Radiological findings are frequent in native CT treatment-planning scans for radiation of breast cancer patients. Therefore, the radiologist should use this performed sectional image to obtain additional information of the patient. Advances in knowledge: Treatment-planning CT scans can show several radiological findings, namely 2.34 findings per patient. Major clinically relevant findings account for 8.87%.

ADC Histogram Analysis of Cervical Cancer Aids Detecting Lymphatic Metastases-a Preliminary Study.

PURPOSE: Apparent diffusion coefficient (ADC) histogram analysis has been used to some extent in cervical cancer (CC) to distinguish between low-grade and high-grade tumors. Although this differentiation is undoubtedly helpful, it would be even more crucial in the presurgical setting to determine whether a tumor already gained the potential to metastasize via the lymphatic system. So far, no studies investigated the potential of 3T ADC histogram analysis in CC to differentiate between nodal-positive and nodal-negative entities. Therefore, the principal aim of our study was to investigate the potential of 3T ADC histogram analysis to differentiate between CC with and without lymph node metastasis. The second aim was to elucidate possible differences in ADC histogram parameters between CC with limited vs. advanced tumor stages and well-differentiated vs. undifferentiated lesions. Finally, correlations of p53 expression and Ki-67 index with ADC parameters were analyzed. PROCEDURES: Eighteen female patients (mean age 55.4 years, range 32-79 years) with histopathologically confirmed cervical squamous cell carcinoma of the uterine cervix were prospectively enrolled. Tumor stages, tumor grading, status of metastatic dissemination, Ki67-index, and p53 expression were assessed in these patients. Diffusion weighted imaging (DWI) was obtained in a 3T scanner using the following b values: b0 and b1000 s/mm2. RESULTS: Group comparisons using Mann-Whitney U test revealed the following findings: nodal-positive CC had statistically significant lower ADC parameters (ADCmin, ADCmean, median ADC, Mode, p10, p25, p75, and p90) in comparison to nodal-negative CC (all p < 0.05). ADCentropy was significantly elevated (p = 0.046) in tumors with advanced T stages (T3/4) compared to tumors with limited T stage (T2). ADCmin values were different in a statistically significant manner comparing G1/G2 and G3 tumors (40.45 ± 18.63 vs. 65.0 ± 23.63 × 10-5 mm2 s-1, p = 0.035). Furthermore, Spearman Rho calculation identified an inverse correlation between ADCentropy and p53 expression (r = -0.472, p = 0.048). CONCLUSION: The main finding of our study is the discriminability of nodal-positive from nodal-negative CC using ADC histogram analysis in 3T DWI. This information is crucial for the gynecological surgeon to identify the optimal treatment strategy for patients suffering from CC. Furthermore, ADCentropy was identified as a potential imaging biomarker for tumor heterogeneity and might be able to indicate further molecular changes like loss of p53 expression, which is associated with EMT and consequentially indicates a poor prognosis in CC. Finally, our study confirmed the findings of previous works, which indicated that histogram analysis of ADC maps can distinguish between low-grade and high-grade CC. In conclusion, it can be stated that ADC histogram analysis provides additional, prognostically important information on tumor biology in CC.

Incidental findings in the skeletal musculature on computed tomography.

OBJECTIVE: The purpose of this study was to analyze intramuscular incidental findings identified on CT in a large patient cohort. METHODS: In the time period from 2010 to 2015, a total of 44,794 patients with several diagnoses were investigated by CT. Only those patients who underwent body CT including the neck, thorax, abdomen and pelvic regions after the i.v. application of a contrast medium were involved in the study. There were 4085 patients. On further analysis, only patients with intramuscular findings (IFs) incidentally detected on CT were included. Osseous or soft-tissue lesions with invasion into the musculature were excluded from the study. Patients with known or clinically suspicious muscle disorders were also excluded. Overall, 639 (15.64% of the 4085 analyzed cases) patients, 253 females and 386 males, with mean age 72.43 ± 12.02 years, were identified. Collected data were evaluated by means of descriptive statistics. RESULTS: In the 639 patients, 917 IFs were identified. More often, several benign disorders were diagnosed (n = 803, 87.6%). There were hernias (33.5%), atrophy of different muscles (22%), lipomas (21%), intramuscular calcifications (8.7%), bursitis (1.2%) and intramuscular bleeding (1.1%). Malignant IFs (n = 114, 12.4%) included intramuscular metastases (11.9%) and lymphomas (0.6%). Most frequently, the identified IFs were localized in the abdominal wall musculature, paravertebral and gluteal muscles. In 657 cases (71.6% of all IFs), the identified muscle findings were not diagnosed by the radiologist who initially assessed the investigation. CONCLUSION: CT can detect different incidental disorders within the skeletal musculature. Most of them were benign. However, malignant lesions can also occur. Therefore, skeletal muscles should be carefully evaluated on CT performed for other reasons. Advances in knowledge: IFs occur in 15.6% of CT investigations. Benign findings represent 87.6% and malignant lesions can be identified in 12.4%.

Thrombotic events as incidental finding on computed tomography in intensive care unit patients.

INTRODUCTION: Intensive care unit (ICU) patients are a risk group to develop thrombosis and/or thromboembolism. The purpose of this study was to analyze the frequency and localization of clinically silent thrombotic events (TE) detected on CT. MATERIALS AND METHODS: From 2006 to 2013 a total of 370 patients from the ICU of our university clinic were investigated by postcontrast CT. In all cases CT was performed for detecting septic foci. There were 135 women and 235 men. CT scans included cervical, thoracic, abdominal, and pelvic regions. CT images of all patients were re-interpreted by 2 radiologists by consensus. Only thromboses detected for the first time on CT were included into the analysis. Collected data were evaluated by means of descriptive statistics. Frequencies and localizations of TE in surgical and non surgical patients were analyzed by Chi-square test. Significance level was p<0.05. RESULTS: In 31.9% several TE were diagnosed. There were venous thrombosis (89.8%), cardiac thrombus (2.6%), and pulmonary embolism (7.6%). More often jugular veins were affected followed by brachiocephalic veins, and iliac veins. The frequency of TE in surgical patients was 31.1%, and 32.1% in non surgical patients. Patients after surgery had more often thrombosis of extremities veins in comparison to non surgical patients. In 61.9% of all TE the identified thrombotic complications were not diagnosed at the time of CT investigations. CONCLUSION: TE can be identified in 31.9% of ICU patients as incidental finding on CT. There were venous thromboses, pulmonary embolism, and cardiac thrombus. Most frequently neck and thoracic veins were affected. 61.9% of all TE were not diagnosed at the time of CT investigations. Radiologists should check carefully CT scans for presence of different TE.

Clinically Relevant Cardiovascular Findings Detected on Staging Computed Tomography in Patients with Several Malignancies.

We evaluated the frequency and subtypes of clinically relevant cardiovascular (CV) findings identified on staging computed tomography (CT) in a large sample. Patients (n = 5026) with different malignant diseases were staged by CT. Clinically relevant CV findings (CRCFs) were included into the study. The CRCFs were defined as cardiac aneurysm, cardiac thrombus, venous thrombosis, arterial thrombosis, arterial dissection, pulmonary thromboembolism, arterial dissection, and dislocation of venous ports/central venous catheters. The CRCFs were identified in 342 patients (6.8% of all patients). Overall, 491 CRCFs were identified in the patients (1.4 per patient). In 203 (59.4%) patients, 1 CRCF; in 129, 2 (37.7%) CRCFs; and in 10 (2.9%) cases, 3 CRCFs were detected. There were incidental venous thrombosis (n = 298, 60.7% of all CRCFs), pulmonary thromboembolism (n = 84, 17.1%), arterial aneurysms (n = 44, 8.9%), arterial thrombosis (n = 43, 8.8%), heart thrombus (n = 15, 3.1%), arterial dissection (n = 3, 0.6%), heart aneurysms (n = 2, 0.4%), and port catheter dislocation (n = 2, 0.4%). The identified CRCF can be associated with potential hazardous complications. The CV system should be carefully evaluated in staging CT investigations.

Costs associated with evaluation of incidental breast lesions identified on computed tomography.

OBJECTIVE: The purpose of this study was to calculate the costs associated with the evaluation of breast incidentalomas (BI) identified on CT. METHODS: All CT scans of the thorax performed at the radiological department of the University Hospital of the Martin-Luther-University Halle-Wittenberg between the years 2006 and 2014 were reanalysed retrospectively. 111 patients with BI were identified. The radiological and clinical reports of these cases were examined, and the costs of all further diagnostic procedures performed to clarify the detected BI were calculated. RESULTS: In 31 (27.9%) of the 111 patients, BI were not further investigated. Of the remaining 80 (72.1%) cases of BI, primary breast carcinoma was diagnosed in 10 (12.5%) cases, breast involvement by lymphoma in 7 (8.8%) cases, intramammary metastases in 19 (23.7%) cases and different benign findings in 44 (55.0%) cases. The total costs associated with clarification of BI in our institution (n = 53) amounted to €14,045.71. The costs per BI were €265.01 and per newly identified breast cancer €1560.63. CONCLUSION: Breast lesions incidentally detected on CT should be categorized as major incidental findings because of the high frequency of malignancy. The radiologist should carefully evaluate the breast on CT images and all identified BI should be clarified. The clarification of BI is associated with a low economic burden in comparison with screening programmes for primary breast cancer. ADVANCES IN KNOWLEDGE: The radiologist should carefully evaluate the breast on CT and all identified BI should be clarified. The evaluation of BI is associated with a low economic burden.

Imaging characteristics and embolus burden of unreported pulmonary embolism in oncologic patients.

The present study sought to estimate the frequency of overseen and unreported (i.e., false negative) pulmonary embolism (PE) events in oncologic patients. In a retrospective analysis, 3270 patients (6780 computed tomography examinations) were reviewed. Unreported PE was found in 74 patients (2.3%). It was particularly frequent in follow-up staging examinations in patients with metastasized malignancies of the lung and kidney. The present data support the thesis that the search error (thrombus was never fixed by the eyes of the reviewer) was the most common reason why PE was overseen.

Imaging findings of gastric diverticula.

INTRODUCTION: Gastric diverticula (GD) are very rare. Computer tomographic findings in GD have been reported only as case reports previously. The aim of this study was to estimate the prevalence of GD on computed tomography (CT) and to analyze their radiological appearances. MATERIALS AND METHODS: From 2006 to 2013, a total of 14,428 patients were examined by abdominal/thoracic CT at our institution. GD were diagnosed in 18 (0.12%) patients (13 women and 5 men, median age, 64 years). In 9 patients, additional endoscopy and in 7 patients upper gastrointestinal investigation with contrast medium were performed. Magnetic resonance imaging (MRI) was available for 3 cases. RESULTS: In all patients GD were diagnosed incidentally during CT examination. The diverticula were located at the posterior wall of the gastric fundus below the esophagogastric junction. On CT, GD presented as cystic lesions with a thin wall and an air fluid level, located behind the stomach between spleen, adrenal gland, and crus of the left diaphragm. CONCLUSION: The prevalence of GD encountered in our CT series is 0.12%. GD demonstrate typical CT appearances, namely, cystic lesions located in the left paravertebral region. The radiologist should be familiar with this finding to avoid possible misinterpretations.

PET/CT imaging of skeletal muscle metastases.

BACKGROUND: Skeletal muscle metastases (SMM) are very rare because of complex resistance of the musculature to metastatic invasion. Previously, positron emission tomography (PET) imaging of SMM has been reported only in few reports. A systematic analysis of SMM features in PET/CT has not been performed before. PURPOSE: To study PET/CT findings of SMM in a larger group of patients with known malignancies and to determine PET/CT patterns of SMM in different primary tumors. MATERIAL AND METHODS: Between January 2009 and December 2011 581 patients with lung cancer were investigated by PET with 18 F-fluordeoxyglucose (FDG PET) and computed tomography (CT) at the Center of Fusion Imaging, Halle. In five patients SMM were identified. Furthermore, PubMed database was screened for muscle metastases. Only articles containing SUV of SMM were considered in the study. Twenty-one articles with 33 patients could be included in this meta-analysis from the literature. RESULTS: At our center the prevalence of SMM was 0.9%. Our analysis comprised 38 patients with 67 muscle metastases. All identified SMM presented as intramuscular focal abnormal activity with SUV ranging from 2.4 to 25.9, median SUV 7.8. The median size of the muscle metastases was 2.5 cm (range, 0.6-6.5 cm). There were no significant differences between SUV and size of SMM arising from lung cancer, renal cell carcinoma, and esophageal cancer. Also, there was no correlation between SUV and size of SMM (r = 0.101, P = 0.558) and between SUV of SMM and primary tumors (r = 0.138, P = 0.686). In nine (23.7%) of the 38 patients, the identified SMM were isolated distant metastases or isolated tumor recurrence. CONCLUSION: SMM manifested on PET/CT as focal hypermetabolic intramuscular areas with different SUV. There were no significant differences between SUV or size of the identified SMM in esophageal cancer, renal cell carcinoma, and lung cancer.

Muscle metastases: comparison of features in different primary tumours.

BACKGROUND: Muscle metastases (MM) from solid tumours are rare. The aim of this study was to describe radiological features of MM, and to compare their patterns in different malignancies. METHODS: A retrospective search in the statistical database of our institution revealed 61 cases of MM. Additionally, a retrospective search in Pubmed database was performed. Together with our cases the present analysis comprises 461 patients (682 MM). RESULTS: MM derived from the following malignancies: lung cancer (25.1%), gastrointestinal tumours (21.0%), and urological tumours (13.2%). Other neoplasias with MM were rare. MM were localised most frequently in the thigh muscles, the extraocular musculature, and the gluteal and paravertebral muscles. The localisation of MM was different in several primary malignancies. CONCLUSION: MM present with a broad spectrum of radiological features. Different CT imaging findings of MM were observed in different primary tumours. The localisation of MM also varies with different primary malignancies.

Comparison between incidental malignant and benign breast lesions detected by computed tomography: a systematic review.

INTRODUCTION: With increasing use of CT for a variety of diagnostic pathways, breast incidentalomas will be identified more frequently. Therefore, it is important for radiologists to know typical CT appearances of malignant and benign breast lesions. The aim of this study was to compare radiological features of benign and malignant incidental breast findings on CT reported in the literature. METHODS: A search for appropriate studies published between 1988 and 2012 yielded five results with a total of 214 patients/257 lesions. Morphological features of the described breast masses, such as shape, margins, enhancement, axillary lymphadenopathy and calcifications, were analysed. RESULTS: The prevalence of incidental breast carcinoma varied from 9.2% to 60.9% of the lesions. The frequency of benign lesions was 30.4-71.8%. The remaining incidental findings were intramammary metastases and primary or secondary breast lymphomas. Malignant lesions were larger in size, were round in shape and demonstrated a higher density compared with benign lesions. There was no significant difference in the margin characteristics of the masses between the groups. The morphological features previously postulated as typical for malignancy turned out to be non-specific. CONCLUSIONS: Benign and malign breast incidentalomas are not safely distinguished from each other on standard chest CT. Therefore, all breast incidentalomas should be evaluated by further radiological/histological investigations.