Neutrophils and neutrophil extracellular traps enhance venous thrombosis in mice bearing human pancreatic tumors.

Pancreatic cancer is associated with a high incidence of venous thromboembolism. Neutrophils have been shown to contribute to thrombosis in part by releasing neutrophil extracellular traps (NET). A recent study showed that increased plasma levels of the NET biomarker, citrullinated histone H3 (H3Cit), are associated with venous thromboembolism in patients with pancreatic and lung cancer but not in those with other types of cancer, including breast cancer. In this study, we examined the contribution of neutrophils and NET to venous thrombosis in nude mice bearing human pancreatic tumors. We found that tumor-bearing mice had increased circulating neutrophil counts and levels of granulocyte-colony stimulating factor, neutrophil elastase, H3Cit and cell-free DNA compared with controls. In addition, thrombi from tumor-bearing mice contained increased levels of the neutrophil marker Ly6G, as well as higher levels of H3Cit and cell-free DNA. Thrombi from tumor-bearing mice also had denser fibrin with thinner fibers consistent with increased thrombin generation. Importantly, either neutrophil depletion or administration of DNase I reduced the thrombus size in tumor-bearing but not in control mice. Our results, together with clinical data, suggest that neutrophils and NET contribute to venous thrombosis in patients with pancreatic cancer.

Plasminogen activator inhibitor 1 and venous thrombosis in pancreatic cancer.

Pancreatic cancer patients have a high risk of venous thromboembolism (VTE). Plasminogen activator inhibitor 1 (PAI-1) inhibits plasminogen activators and increases the risk of thrombosis. PAI-1 is expressed by pancreatic tumors and human pancreatic cell lines. However, to date, there are no studies analyzing the association of active PAI-1 and VTE in pancreatic cancer patients. We investigated the association of active PAI-1 in plasma and VTE in pancreatic cancer patients. In addition, we determined if the presence of human pancreatic tumors expressing PAI-1 impairs venous thrombus resolution in mice. Plasma levels of active PAI-1 in patients with pancreatic cancer and mice bearing human tumors were determined by enzyme-linked immunosorbent assay. We measured PAI-1 expression in 5 different human pancreatic cancer cell lines and found that PANC-1 cells expressed the highest level. PANC-1 tumors were grown in nude mice. Venous thrombosis was induced by complete ligation of the inferior vena cava (IVC). Levels of active PAI-1 were independently associated with increased risk of VTE in patients with pancreatic cancer (subdistribution hazard ratio per doubling of levels: 1.39 [95% confidence interval, 1.09-1.78], P = .007). Mice bearing PANC-1 tumors had increased levels of both active human and active mouse PAI-1 and decreased levels of plasmin activity. Importantly, mice bearing PANC-1 tumors exhibited impaired venous thrombus resolution 8 days after IVC stasis compared with nontumor controls. Our results suggest that PAI-1 contributes to VTE in pancreatic cancer.

Replicated Risk Index of Patient Functional Status Prior to Allogeneic Hematopoietic Cell Transplantation Predicts Healthcare Utilization and Survival.

Poor physical functioning is associated with adverse outcomes after allogeneic hematopoietic cell transplantation (alloHCT). Analytic tools to predict mortality in alloHCT recipients include the HCT Comorbidity Index (HCT-CI) based on comorbidities and the Disease Risk Index (DRI) based on disease and disease status. We developed and replicated a risk model for overall survival (OS), early mortality (ie, death from any cause at or before day +100), initial hospital length of stay (LOS), and percentage of inpatient days within the first year post-alloHCT. In this study, we incorporated a physical therapy (PT) assessment with the HCT-CI and DRI to improve outcome predictions. The well-defined and feasible measure of functional status for assessing risk includes (1) the number of sit-to-stands performed in 30 seconds, (2) performance of 25 step-ups on the right/left side with (3) oxygen saturation recovery and (4) heart rate recovery, (5) weight-bearing ability, (6) assistance with ambulation, (7) motor and grip strength, (8) sensory and coordination impairment (eg, self-reported peripheral neuropathy, imbalance), (9) self-reported pain, and (10) limited endurance (ie, inability to complete step-ups and/or sit-to-stands). Our training cohort (TC) included 349 consecutive alloHCT recipients at Roswell Park treated between 2010 and 2016 and a subsequent replication cohort (RC; n = 163) treated between 2016 and 2019. Four of the 10 metrics-self-reported pain, limited endurance, self-reported neuropathy, and <10 sit-to-stands in 30 seconds-were identified as significant predictors and were included in the multivariable models with the HCT-CI and DRI to create a new risk index (HCT-PCDRI: HCT-physical, comorbidity, and DRI) for outcomes. Models were tested in the RC. Shorter OS was associated with self-reported pain, limited endurance, higher HCT-CI, and higher DRI. At a median follow-up of 34 months, the 3-year OS based on the HCT-PCDRI was 30% for the very-high-risk group, 54% for the high-risk group, 49% for the intermediate-risk group, and 80% for the low-risk group. The number of patients identified as very high risk increased from 55 using HCT-CI alone to 120 with the new HCT-PCDRI, whereas the number in the low-risk group decreased from 91 to 45. Early mortality and a higher percentage of inpatient days within the first year post-alloHCT (a proxy for poor quality of life and high healthcare utilization) were associated with self-reported pain, higher HCT-CI, and higher DRI. A shorter initial LOS (ie, initial low healthcare utilization) was associated with performance of >10 sit-to-stands in 30 seconds, no self-reported neuropathy, and lower HCT-CI. These PT metrics combined with the HCT-CI and DRI created the HCT-PCDRI, which resulted in more patients being categorized accurately as high risk versus low risk. The HCT-PCDRI results were replicated in an independent cohort. Pre-alloHCT PT metrics with self-reported symptoms (pain and neuropathy) were associated with survival post-alloHCT and prolonged hospital LOS. The HCT-PCDRI scoring system for risk stratification of alloHCT recipients more accurately identifies patients at potential risk of poor outcomes. The HCT-PCDRI can be tested in <15 minutes to identify patients for intervention before or during treatment to potentially improve outcomes.

Rivaroxaban does not affect growth of human pancreatic tumors in mice.

BACKGROUND: Some clinical studies have shown that low-molecular-weight heparins (LMWHs) prolong the survival of cancer patients. In addition, various anticoagulants have been shown to reduce growth of tumors in mice. However, there are no studies on the effect of the factor Xa inhibitor rivaroxaban on growth of human pancreatic tumors in nude mice. OBJECTIVES: To test the hypothesis that the factor Xa inhibitor rivaroxaban reduces the growth of tissue factor (TF)-positive pancreatic tumors but not TF-negative pancreatic tumors in mice. METHODS: The TF-positive human pancreatic cancer cell line BxPc-3 and the TF-negative human pancreatic cancer cell line MIA PaCa-2 were injected subcutaneously into nude mice and tumors grown to a mean volume of ~100 mm3 . Mice were then divided into two groups. One group was fed chow containing rivaroxaban (0.5 g/kg chow) whereas the other group was fed chow without rivaroxaban. RESULTS: Rivaroxaban significantly prolonged prothrombin time in tumor-bearing mice. Rivaroxaban did not affect cell proliferation or growth of either BxPc-3 or MIA PaCa-2 tumors grown subcutaneously in nude mice. CONCLUSION: Our results indicate that inhibition of factor Xa with rivaroxaban does not affect the growth of two human pancreatic tumors in nude mice.