ancer is a major risk factors for thrombotic pulmonary embolism (PE), [1][2][3][4] and thrombotic PE associated with cancer is known as Trousseau's syndrome. Cancer induces not only thrombotic PE but also tumor PE and tumor invasion into large veins. However, there is lack of epidemiological data from the general population to compare the incidence of thrombotic PE, tumor PE and tumor invasion to large veins according to the histopathology and the site of the cancer. Clinicians have given more attention to the prevention of venous thromboembolism (VTE) with chemotherapeutic and surgical treatment of cancer patients, but it is important to evaluate the risk of the development of VTE according to the histopathology and primary site of the cancer.The goal of this study was to investigate variations in the incidence of thrombotic PE, tumor PE and tumor invasion into large veins according to tumor histopathology and tumor site.
MethodsA total of 65,181 cancer patients (66.0%) were identified from 98,736 postmortem examinations. [5][6][7] The incidence of PE, as well as the tumor site and the histopathology, was recorded for each case. PE was defined as critical (critical PE) when it was the primary cause of death or main diagnosis, and "total PE" was used to indicate the total number of thrombotic PEs, tumor PEs, bacterial PEs, mycotic PEs and other emboli (eg, fat, amniotic fluid etc). The type of emboli was determined by the pathologist performing the autopsy. Cases with 2 different types of emboli were counted twice. All cases were also reviewed for the presence of tumor invasion into a large vein. Within each PE group, the incidence of PE was compared between specific tumor types (adenocarcinoma, mucinous carcinoma, transi- Takeshi Nakano, MD*; Kunio Shirato, MDBackground The specific incidence of thrombotic pulmonary embolism (PE), tumor PE and tumor invasion into large veins according to tumor type and tumor site remains unclear. Methods and Results A total of 65,181 cancer patients were identified from 98,736 postmortem examinations. Thrombotic PE occurred in 2.32% of all cancer patients and comprised 88.6% of the total number of all PE events. The incidence of thrombotic PE was high in those with adenocarcinoma, leukemia and large cell carcinoma, and was low in those with hepatic cell carcinoma. The incidence of PE was high when tumor was present in hematogenous tissue, lungs, ovaries, pancreas and the biliary system, and was low when tumor was present in the liver. The incidence of tumor PE was high with large cell carcinoma, hepatic cell carcinoma and adenocarcinoma, and was also high when tumor was present in the lungs, ovaries, kidneys and liver. There was a significant correlation between the incidence of tumor PE and the incidence of tumor invasion into large veins. Conclusion The incidence of thrombotic PE, tumor PE and tumor invasion into large veins varies significantly according to tumor histopathology and tumor site. (Circ J 2006; 70: 744 -749)