Purpose: To test the hypothesis that subjects with a known malignancy at the time of acute pulmonary embolism (PE) have different clinical characteristics and predictors of 30-day all-cause mortality when compared with subjects with no known malignancy.
Materials and methods: A retrospective (August 2003 to March 2010) cohort of 1596 consecutive positive (for acute PE) computed tomography pulmonary angiograms (CTPAs) performed at a single, large, urban teaching hospital was separated into those from subjects with (n=835) and those from subjects without (n=761) a known malignancy. Clinical characteristics were compared between groups, and a logistic regression model determined predictors of 30-day all-cause mortality for each group.
Results: Subjects with malignancy were older (60.8±13.9 vs. 54.5±18.8 y, P<0.001), had fewer risk factors for PE, and had a higher 30-day all-cause mortality (19.6% vs. 3.2%, P<0.001). The malignancy group had fewer predictors of death compared with the nonmalignancy group; advanced age, presence of coronary artery disease, history of stroke, and chronic obstructive lung disease were significantly more predictive of death in the nonmalignancy population. An enlarged right ventricle on CTPA (right to left ventricular diameter ratio >1.0) had a higher risk of 30-day death only among subjects with no known malignancy at the time of the CTPA (odds ratio=4.08, 95% confidence interval: 1.67-9.96).
Conclusions: Among subjects who present with acute PE, those with a malignancy had different clinical characteristics and predictors of mortality when compared with the cohort of subjects with no known malignancy. A computed tomography-derived right to left ventricular diameter ratio predicts 30-day all-cause mortality only for those subjects who do not have a malignancy.