Pulmonary embolism outcomes across five cancer types: A nationwide analysis.

e20548 Background: Pulmonary embolism (PE) is known to have high mortality and morbidity burden especially in cancer patients. While the association between cancer and thrombotic events is well established, the variations in outcomes and management strategies are not well studied. Our study aims to analyze a large cohort of patients with Lung, Pancreatic, Central nervous system (CNS), Pancreatic, and Ovarian cancers with PE from the National Inpatient Sample (NIS) database to compare outcomes among these 5 cancers with high risk of PE. Methods: Utilizing the NIS from 2017 to 2021, we identified patients aged 18 years and older with a primary or secondary diagnosis of PE and Lung/stomach/CNS/Pancreatic/Ovarian cancers using ICD-10 codes. We compared both cohorts and analyzed outcomes using logistic regression analysis. Results: Our analysis identified 94460 lung cancer, 9210 stomach cancer, 49930 CNS cancer, 30630 pancreatic cancer, and 13290 ovarian cancer patients with PE. Mortality was significantly (p<0.05) higher in patients with PE and Stomach cancer (OR 1.36) followed by Lung cancer (OR 1.31), followed by CNS malignancies (OR 1.14) when compared with each other. Odds of major bleeding was highest in patients with PE and stomach cancer (OR 4.32; p <0.05; 95% CI 3.35 - 5.57) followed by pancreatic cancer (OR 1.62; <p 0.05; 95% CI 1.27 - 2.06) and lower in Lung cancer (OR 0.77; p <0.05; 95% CI 0.62 - 0.95). Odds of hemorrhagic stroke was 10 times higher in the co-occurrence of PE with CNS malignancies (HR 10; p <0.05, 95% CI 7.96 - 12.72) and lowest in Stomach cancer (OR 0.32; p <0.05, 95% CI 0.15 - 0.70). The odds of IV thrombolytic use was least in patients with PE & CNS malignancies (OR 0.57; p <0.05; 95% CI 0.39-0.84). Odds of IVC filter use was highest in the co-occurrence of PE with CNS malignancies (OR 2.04; p <0.05; 95% CI 1.82 - 2.28) followed by Stomach cancers (OR 1.77; p <0.05, 95% CI 1.48 - 2.12). Although odds of catheter directed thrombolysis use in CNS cancers with PE was higher compared to others (OR 1.32; p 0.133, 95% CI 0.92-1.90), it did not achieve statistical significance. Mechanical ventilator use was less likely in PE with pancreatic (OR 0.49; p <0.05, 95% CI 0.41 - 0.60) and ovarian cancers (OR 0.61; <p 0.05; 95% CI 0.49 - 0.76). Mean length of stay and total charges during hospitalization were higher for patients with PE & ovarian cancer 8.8 days (8.3 - 9.3) days and $111,054 ($103230 - $118877) respectively. Conclusions: Our large-scale analysis of PE across five cancer types reveals significant variations in outcomes and management strategies. Stomach and lung cancers showed higher mortality rates, while CNS malignancies had unique challenges with hemorrhagic stroke risk and IVC filter use. Our study highlights the need for cancer-specific approaches to PE management, considering the distinct risk profiles and treatment outcomes observed. Further research aimed at improving PE outcomes in different cancer populations is warranted.