Traumatic Hemothorax Blood Contains Elevated Levels of Microparticles that are Prothrombotic but Inhibit Platelet Aggregation.

OBJECTIVES: Autotransfusion of shed blood from traumatic hemothorax is an attractive option for resuscitation of trauma patients in austere environments. However, previous analyses revealed that shed hemothorax (HX) blood is defibrinated, thrombocytopenic, and contains elevated levels of D-dimer. Mixing studies with normal pooled plasma demonstrated hypercoagulability, evoking concern for potentiation of acute traumatic coagulopathy. We hypothesized that induction of coagulopathic changes by shed HX blood may be due to increases in cellular microparticles (MP) and that these may also affect recipient platelet function. METHODS: Shed HX blood was obtained from 17 adult trauma patients under an Institutional Review Board approved prospective observational protocol. Blood samples were collected every hour up to 4 h after thoracostomy tube placement. The corresponding plasma was isolated and frozen for analysis. The effects of shed HX frozen plasma (HFP) and isolated HX microparticles (HMP) on coagulation and platelet function were assessed through mixing studies with platelet-rich plasma at various dilutions followed by analysis with thromboelastometry (ROTEM), platelet aggregometry (Multiplate), enzyme-linked immunosorbent assays, and flow cytometry. Furthermore, HFP was assessed for von Willebrand factor antigen levels and multimer content, and plasma-free hemoglobin. RESULTS: ROTEM analysis demonstrated that diluted HFP and isolated HMP samples decreased clotting time, clotting formation time, and increased α angle, irrespective of sample concentrations, when compared with diluted control plasma. Isolated HMP inhibited platelet aggregation in response to adenosine diphosphate, arachidonic acid, and collagen. HFP contained elevated levels of fibrin-degradation products and tissue factor compared with control fresh frozen plasma samples. MP concentrations in HFP were significantly increased and enriched in events positive for phosphatidylserine, tissue factor, CD235, CD45, CD41a, and CD14. von Willebrand factor (vWF) multimer analysis revealed significant loss of high molecular weight multimers in HFP samples. Plasma-free hemoglobin levels were 8-fold higher in HFP compared with fresh frozen plasma. CONCLUSION: HFP induces plasma hypercoagulability that is likely related to increased tissue factor and phosphatidylserine expression originating from cell-derived MP. In contrast, platelet dysfunction is induced by HMP, potentially aggravated by depletion of high molecular weight multimers of vWF. Thus, autologous transfusion of shed traumatic hemothorax blood may induce a range of undesirable effects in patients with acute traumatic coagulopathy.

Shed Pleural Blood from Traumatic Hemothorax Contains Elevated Levels of Pro-Inflammatory Cytokines.

PURPOSE: The autotransfusion of unwashed (or unprocessed) shed hemothorax blood (USHB) in trauma patients is widely assumed to be beneficial; however, the inflammatory potential of shed pleural blood has not been thoroughly studied. Since previous studies have documented marked changes in coagulation function of shed pleural blood, we hypothesized that its level of inflammatory cytokines would be elevated. METHODS: A prospective observational study of trauma patients in whom cytokine levels from USHB were compared to venous samples from healthy volunteers was conducted. Differences between the cytokine content of patient-derived samples were compared to those from healthy subjects. RESULTS: There was a statistically significant increase in pro-inflammatory cytokines (IL-6, IL-8, TNFα, GM-CSF), a pro-inflammatory Th-1 cytokine (IFNγ), and anti-inflammatory Th-2 cytokines (IL-4 and IL-10) in shed pleural blood over four hours when compared with samples from healthy controls (P <0.05). Cytokine levels in USHB are approximately 10- to 100-fold higher compared with healthy control venous samples. CONCLUSIONS: USHB, even collected within the accepted four-hour window, contains significantly elevated cytokine levels, suggesting the potential for deleterious effects from autotransfusion. Randomized trials are needed to determine the safety and efficacy of autotransfusion in trauma patients.

An experimental model of hemothorax autotransfusion: impact on coagulation.

BACKGROUND: Traumatic hemothorax (HTX) has been demonstrated to predictably contain low fibrinogen, low hematocrit, and low platelet counts. When analyzed on its own, shed HTX demonstrates coagulopathy. However, when mixed with normal pooled plasma (NPP) at physiologically relevant dilutions, HTX demonstrates accelerated coagulation. We hypothesize that when HTX is mixed with a patient's own plasma, the mixture will demonstrate hypercoagulability. The accelerated coagulation of this mixture would have important implications for the autotransfusion of HTX as a method of resuscitating a trauma patient. METHODS: Adult trauma patients from whom greater than 140 mL of HTX was evacuated within 1 hour of tube thoracostomy were included. HTX was sampled at 1 hour after evacuation, and a portion of the sample was centrifuged and stored as frozen plasma for later analysis. The remainder of the sample was analyzed (coagulation, hematology, electrolytes), and values were compared with concurrent venous values extracted via chart review. A citrate tube containing the patient's venous blood was additionally spun down and frozen for subsequent mixing study analysis. Coagulation was further evaluated by mixing serial dilutions of the previously frozen HTX with NPP. Additionally, the previously frozen HTX was mixed in serial dilutions with the previously frozen sample of patient plasma (PTP). RESULTS: Subjects (10) were enrolled based on inclusion criteria and collection of a discarded venous sample. In HTX samples analyzed alone, no thrombus was formed in any coagulation test (activated partial thromboplastin time [aPTT] > 180). The median aPTT value of PTP alone was 25.5. In 1-hour specimens mixed at a clinically relevant dilution of 1:4, HTX mixed with NPP had a mediana PTT value of 26.0, whereas HTX mixed with PTP had a median aPTT value of 21.7. Thus, the mixture of HTX + PTP demonstrated a statistically significantly lower aPTT than the mixture of HTX + NPP (P = 0.01). Additionally, the mixture of HTX and PTP shows a statistically significantly lower aPTT value than PTP alone (P = 0.03), indicating a hypercoagulable state. CONCLUSIONS: HTX demonstrates coagulopathy when analyzed independently, but is hypercoagulable when mixed with NPP or PTP. Furthermore, mixing studies show a statistically significantly lower aPTT when HTX is mixed with PTP versus HTX mixed with NPP. Thus, autotransfusion of HTX would likely produce a hypercoagulable state in vivo, and should not be used in place of other blood products to resuscitate a trauma patient. The autotransfusion of HTX may, however, be of use in a resource-limited environment where other blood products are not available.

A small amount can make a difference: a prospective human study of the paradoxical coagulation characteristics of hemothorax.

BACKGROUND: The evacuated hemothorax has been poorly described because it varies with time, it has been found to be incoagulable, and its potential effect on the coagulation cascade during autotransfusion is largely unknown. METHODS: This is a prospective descriptive study of adult patients with traumatic chest injury necessitating tube thoracostomy. Pleural and venous samples were analyzed for coagulation, hematology, and electrolytes at 1 to 4 hours after drainage. Pleural samples were also analyzed for their effect on the coagulation cascade via mixing studies. RESULTS: Thirty-four subjects were enrolled with a traumatic hemothorax. The following measured coagulation factors were significantly depleted compared with venous blood: international normalized ratio (>9 vs 1.1) (P < .001) and activated partial thromboplastin time (aPTT) (>180 vs 24.5 seconds) (P < .001). Mixing studies showed a dose-dependent increase in coagulation dilutions through 1:8 (P < .05). CONCLUSIONS: An evacuated hemothorax does not vary in composition significantly with time and is incoagulable alone. Mixing studies with hemothorax plasma increased coagulation, raising safety concerns.

The etiology of pneumoperitoneum in the 21st century.

BACKGROUND: We sought to determine the origin of free intraperitoneal air in this era of diminishing prevalence of peptic ulcer disease and imaging studies. In addition, we attempted to stratify the origin of free air by the size of the air collection. METHODS: We queried our hospital database for "pneumoperitoneum" from 2005 to 2007 and for proven gastrointestinal perforation from 2000 to 2007. Massive amount of free air was defined as any air pocket greater than 10.0 cm. RESULTS: Among patients with free air, the predominant causes were perforated viscus (41%) and postoperative (<8 days) residual air (37%). For patients with visceral perforation, only 45% had free air on imaging studies, and for these patients, the predominant cause was peptic ulcer (16%), diverticulitis (16%), trauma (14%), malignancy (14%), bowel ischemia (10%), appendicitis (6%), and endoscopy (4%). The likelihood that free air was identified on an imaging study by lesion was 72% for perforated peptic ulcer, 57% for perforated diverticulitis, but only 8% for perforated appendicitis. The origin of massive free air was equally likely to be gastroduodenal, small bowel, or colonic perforation. CONCLUSION: The cause of free air when surgical pathology is the source has substantially changed from previous reports. LEVEL OF EVIDENCE: Epidemiologic study, level IV.