Anticoagulation use and rate of bland thrombus
A total of 10 (20%) patients with tumor thrombus received anticoagulation in combination with cancer-directed therapies. Patient and tumor thrombus characteristics are summarized in Table 2. Six patients received therapeutic doses of anticoagulation and 4 received prophylactic doses of anticoagulation. The rationale for starting anticoagulation was symptomatic tumor thrombus in 4 patients: pulmonary embolism (PE, n=2), renal vein thrombus (RVT, n=1) and lower extremity pain and swelling (n=1) concomitant to tumor thrombus diagnosis. Patient 6 was a teenager who relapsed with a right atrial thrombus identified on routine surveillance imaging after completing therapy and achieving complete remission of osteosarcoma. Despite therapeutic anticoagulation, she developed a symptomatic PE 1 week later which required urgent mechanical thrombectomy. Tissue specimen confirmed relapsed osteosarcoma and the anticoagulant was discontinued. She subsequently died of another PE. Patient 7 had newly diagnosed Ewing sarcoma treated with upfront radical nephrectomy and IVC tumor thrombectomy, which was confirmed by pathology. Routine scans obtained 2 weeks later, showed possible residual tumor thrombus in IVC surgical site and pulmonary infarcts with confirmation of PE via CT angiography. She was started on chemotherapy following the surgery. The finding of thrombus on repeat imaging was concerning for tumor thrombus vs a component of bland thrombus that might have developed post thrombectomy. Hence, she was started therapeutic anticoagulation in addition to chemotherapy for Ewings sarcoma. After 6 months of therapeutic anticoagulation and chemotherapy, repeat imaging studies showed complete resolution of IVC thrombus and PE. She ultimately died due to relapsed disease. Patient 8 had clear cell sarcoma with extensive tumor thrombus extending from IVC to common iliac and femoral veins. Chemotherapy was initiated soon after diagnosis, but anticoagulation was not started at that time. However, he presented a few weeks later with left lower leg pain and swelling. Repeat imaging showed the same extensive tumor thrombus and a new separate small occlusive thrombus in the superficial femoral vein which could be tumor thrombus vs bland thrombus. Given the patient was symptomatic, he was started on therapeutic anticoagulation. Leg pain and swelling resolved after 4 days and anticoagulation was discontinued at that point. Patient underwent surgical resection and IVC thrombectomy later and pathology confirmed the finding of tumor thrombus. Finally, patient 10 was started on therapeutic anticoagulation for occlusive renal vein tumor thrombus causing elevated creatinine, concerning for acute kidney injury. He received three months of therapeutic anticoagulation along with chemotherapy for metastatic embryonal carcinoma and repeat imaging at the end of three months showed improved blood flow through the renal vein. The other six patients had an asymptomatic tumor thrombus. Rationale for initiating anticoagulation in these patients was to prevent further progression and possible embolism of the bland thrombus component of the tumor from high risk locations (IVC and/or right atrium).
The follow up period for the cohort was a median of 13.5 months (range 1 month to 7 years). Of the patients with or without anticoagulation, none developed a subsequent bland thrombus and/or embolization secondary to tumor thrombus, although there was uncertainty about one patient: Patient 7 developed another thrombus post-surgical resection, but without pathologic exam, it was unknown if this was bland thrombus or a remnant of the previous tumor thrombus. A single patient developed a clearly catheter-associated deep vein thrombus not related to tumor thrombus 1 month after his cancer diagnosis.