Bayesian modeling of pretransplant variables accurately predicts kidney graft survival.

INTRODUCTION: Machine learning can enable the development of predictive models that incorporate multiple variables for a systems approach to organ allocation. We explored the principle of Bayesian Belief Network (BBN) to determine whether a predictive model of graft survival can be derived using pretransplant variables. Our hypothesis was that pretransplant donor and recipient variables, when considered together as a network, add incremental value to the classification of graft survival. METHODS: We performed a retrospective analysis of 5,144 randomly selected patients (age ≥18, deceased donor kidney only, first-time recipients) from the United States Renal Data System database between 2000 and 2001. Using this dataset, we developed a machine-learned BBN that functions as a pretransplant organ-matching tool. RESULTS: A network of 48 clinical variables was constructed and externally validated using an additional 2,204 patients of matching demographic characteristics. This model was able to predict graft failure within the first year or within 3 years (sensitivity 40%; specificity 80%; area under the curve, AUC, 0.63). Recipient BMI, gender, race, and donor age were amongst the pretransplant variables with strongest association to outcome. A 10-fold internal cross-validation showed similar results for 1-year (sensitivity 24%; specificity 80%; AUC 0.59) and 3-year (sensitivity 31%; specificity 80%; AUC 0.60) graft failure. CONCLUSION: We found recipient BMI, gender, race, and donor age to be influential predictors of outcome, while wait time and human leukocyte antigen matching were much less associated with outcome. BBN enabled us to examine variables from a large database to develop a robust predictive model.

Incidence of pulmonary embolus in combat casualties with extremity amputations and fractures.

BACKGROUND: The objective of this retrospective study was to determine the incidence of pulmonary embolism (PE) in casualties of wartime extremity wounds and specifically in casualties with a trauma-associated amputation. METHODS: Records of all combat-wounded evacuated and admitted between March 1, 2003, and December 31, 2007, were retrospectively reviewed. Continuous and categorical variables were studied with the Student's t test, Fisher's exact test or χ² test; multivariate analysis was performed using a stepwise regression logistic model. RESULTS: A total of 1,213 records were reviewed; 263 casualties met the inclusion criteria. One hundred three (41.5%) had amputations and 145 (58.5%) had long-bone fractures not requiring amputation. The observed rate of PE in these 263 casualties was 5.7%. More casualties with amputations, 10 (3.7%), developed PE than those with long-bone fractures in the absence of amputation, 5 (1.9%) (p = 0.045). Casualties with bilateral lower extremity trauma-associated amputations had a significantly higher incidence of PE compared with those sustaining a single amputation (p = 0.023), and the presence of bilateral lower extremity amputations was an independent risk factor for development of a PE (p = 0.007, odds ratio 5.9) (univariate and multivariate analysis, respectively). CONCLUSION: The cumulative incidence of PE was 5.7%. The incidence of PE is significantly higher with trauma-associated amputation than with extremity long-bone fracture without amputation. Bilateral amputations, multiple long-bone fractures, and pelvic fractures are independent risk factors for the development of PE. The use of aggressive prophylaxis, deep venous thrombosis screening with ultrasound, and use of prophylactic inferior vena cava filters should be considered in this patient population.

Lymphocyte modulation with FTY720 improves hemorrhagic shock survival in swine.

The inflammatory response to severe traumatic injury results in significant morbidity and mortality. Lymphocytes have recently been identified as critical mediators of the early innate immune response to ischemia-reperfusion injury. Experimental manipulation of lymphocytes following hemorrhagic shock may prevent secondary immunologic injury in surgical and trauma patients. The objective of this study is to evaluate the lymphocyte sequestration agent FTY720 as an immunomodulator following experimental hemorrhagic shock in a swine liver injury model. Yorkshire swine were anesthetized and underwent a grade III liver injury with uncontrolled hemorrhage to induce hemorrhagic shock. Experimental groups were treated with a lymphocyte sequestration agent, FTY720, (n = 9) and compared to a vehicle control group (n = 9). Animals were observed over a 3 day survival period after hemorrhage. Circulating total leukocyte and neutrophil counts were measured. Central lymphocytes were evaluated with mesenteric lymph node and spleen immunohistochemistry (IHC) staining for CD3. Lung tissue infiltrating neutrophils were analyzed with myeloperoxidase (MPO) IHC staining. Relevant immune-related gene expression from liver tissue was quantified using RT-PCR. The overall survival was 22.2% in the vehicle control and 66.7% in the FTY720 groups (p = 0.081), and reperfusion survival (period after hemorrhage) was 25% in the vehicle control and 75% in the FTY720 groups (p = 0.047). CD3(+) lymphocytes were significantly increased in mesenteric lymph nodes and spleen in the FTY720 group compared to vehicle control, indicating central lymphocyte sequestration. Lymphocyte disruption significantly decreased circulating and lung tissue infiltrating neutrophils, and decreased expression of liver immune-related gene expression in the FTY720 treated group. There were no observed infectious or wound healing complications. Lymphocyte sequestration with FTY720 improves survival in experimental hemorrhagic shock using a porcine liver injury model. These results support a novel and clinically relevant lymphocyte immunomodulation strategy to ameliorate secondary immune injury in hemorrhagic shock.