Impact of cancer in pulmonary embolism presentation and outcomes: A large academic center study.

Abstract

6613 Background: The risk of venous thromboembolism is increased 4- to 7-fold in patients with malignancy, emphasizing the need to identify and treat these patients early to improve outcomes. We aimed to study the clinical presentation and outcomes of pulmonary embolism (PE) in patients with and without cancer. Methods: We performed a retrospective analysis of consecutive patients diagnosed with PE via CT scan from 2014-2016 at Jefferson Hospital. We compared patient characteristics, presentation, PE characteristics and mortality of patients with and without cancer. Cox proportional regression hazards model was used for survival-time analysis. Results: Our study included 581 patients, of which 187 (32.1%) had active cancer. Cancer patients were less likely to have chest pain (18.2% vs 37.4% p < 0.01), syncope (2.7% vs 6.6% p = 0.05), bilateral PEs (50% vs 60% p = 0.025), and right heart strain (RHS) (48% vs 58% p = 0.024). Indwelling catheters (IC) were present in 41.2% (n = 77) of cancer patients. However, presence of IC was not associated finding of incidental PEs (26% vs 18.2% p = 0.201). There was no difference in hospital length of stay (8.9 vs 9.4 days p = 0.61) or intensive care unit admission (31.9% vs 33.3% p = 0.75). There were fewer massive PE (3.2% vs 7.1% p = 0.06) in patients with cancer, but this difference was not statistically significant. Cancer patients elected comfort care at higher rates (15.2% vs 5.4% p = 0.01). Cancer patients had higher 1-year mortality as compared to non-cancer (adj HR 6.9, 95% CI 3.3- 14.7, p < 0.01). Among cancer patients, 52.7% had metastasis with a higher 1-year mortality (adj HR 2.5, 95% CI 1.8- 4.9, p < 0.1) and 35.8% were on active chemotherapy with no difference in 1-year survival (adj HR 1.1, 95% CI 0.6-1.8, p = 0.79). The most represented cancers were genitourinary, lung and head and neck (35.3%, 23.0%, 13.4%, respectively). Conclusions: Cancer patients presented with less severe pulmonary emboli which may be due to increased health care contact and pre-clinical suspicion. The presence of IC did not affect the size, location of PE or incidental PEs among cancer patients. Although cancer patients have higher 1-year mortality, PE may not be as large as a contributor as previously perceived.