A Risk-benefit Analysis of Prophylactic Anticoagulation for Patients with Metastatic Germ Cell Tumours Undergoing First-line Chemotherapy.

BACKGROUND: It remains unclear which patients with metastatic germ cell tumours (mGCTs) need prophylactic anticoagulation to prevent venous thromboembolic events (VTEs). OBJECTIVE: To assess the risk and onset of VTEs stratified by risk factors. DESIGN, SETTING, AND PARTICIPANTS: This multi-institutional retrospective dataset included mGCT patients treated with first-line platinum-based chemotherapy. INTERVENTION: Patients with prophylactic anticoagulation were excluded. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: A regression analysis was performed to select risk factors for VTEs. The simulated number needed to treat (NNT) and the number needed to harm (NNH) with prophylactic anticoagulation were calculated based on the cumulative incidences retrieved from this study and hazard rates of recently published trials describing the efficacy of prophylactic anticoagulation to prevent VTEs and the risk of bleeding events. RESULTS AND LIMITATIONS: From 1120 patients, 121 (11%) had a VTE, which occurred prior to chemotherapy in 49 (4%) and on or after chemotherapy in 72 (6%). Six patients (<1%) had a bleeding event without anticoagulation. After backward regression, the one risk factor for a VTE during or after chemotherapy was the use of a venous access device. The simulated cumulative VTE incidence from prophylactic anticoagulation for patients on or after chemotherapy would translate into an NNT of 45 (95% confidence interval [CI] 36-56) and an NNH of 186 (95% CI 87-506). Limitations are mainly related to the retrospective nature of the study. CONCLUSIONS: The mGCTs associated VTEs are most common before and during, but not after, chemotherapy. Avoiding venous access device and/or prophylactic anticoagulation with an acceptable risk-benefit profile may decrease VTE occurring on chemotherapy. PATIENT SUMMARY: We found that venous thromboembolic events (VTEs) occur rarely after chemotherapy. Based on experience of prophylactic anticoagulation in other cancers, we conclude that the risk of VTE in men undergoing chemotherapy for metastatic germ cell tumours can be decreased by thromboprophylaxis with a reasonable risk-benefit profile and by avoidance of venous access devices.

Distribution of Mediastinal Lesions Across Multi-Institutional, International, Radiology Databases.

INTRODUCTION: Mediastinal lesions are uncommon; studies on their distribution are, in general, small and from a single institution. Furthermore, these studies are usually based on pathology or surgical databases and, therefore, miss many lesions that did not undergo biopsy or resection. Our aim was to identify the distribution of lesions in the mediastinum in a large international, multi-institutional cohort. METHODS: At each participating institution, a standardized retrospective radiology database search was performed for interpretations of computed tomography, positron emission tomography-computed tomography, and magnetic resonance imaging scans including any of the following terms: "mediastinal nodule," "mediastinal lesion," "mediastinal mass," or "mediastinal abnormality" (2011-2014). Standardized data were collected. Statistical analysis was performed. RESULTS: Among 3308 cases, thymomas (27.8%), benign mediastinal cysts (20.0%), and lymphomas (16.1%) were most common. The distribution of lesions varied among mediastinal compartments; thymomas (38.3%), benign cysts (16.8%), and neurogenic tumors (53.9%) were the most common lesions in the prevascular, visceral, and paravertebral mediastinum, respectively (p < 0.001). Mediastinal compartment was associated with age; patients with paravertebral lesions were the youngest (p < 0.0001). Mediastinal lesions differed by continent or country, with benign cysts being the most common mediastinal lesions in the People's Republic of China, thymomas in Europe, and lymphomas in North America and Israel (p < 0.001). Benign cysts, thymic carcinomas, and metastases were more often seen in larger hospitals, whereas lymphomas and thymic hyperplasia occurred more often in smaller hospitals (p < 0.01). CONCLUSIONS: Our study confirmed that the spectrum and frequency of mediastinal lesions depend on mediastinal compartment and age. This information provides helpful demographic data and is important when considering the differential diagnosis of a mediastinal lesion.