Joughin, Brian A.

Objective: Trauma-induced coagulopathy (TIC) is provoked by multiple mechanisms and is perceived to be one driver of massive transfusions (MT). Single laboratory values using prothrombin time (INR) or thrombelastography (TEG) are used to clinically define this complex process. We used a proteomics approach to test whether current definitions of TIC (INR, TEG, or clinical judgment) are sufficient to capture the majority of protein changes associated with MT. Methods: Eight level I trauma centers contributed blood samples from patients available early after injury. TIC was defined as INR >1.5 (INR-TIC), TEG maximum amplitude <50 mm (TEG-TIC), or clinical judgment (Clin-TIC) by the trauma surgeon. MT was defined as >10 units of red blood cells in 24 hours or >4 units RBC/hour during the first 4 hours. SomaLogic proteomic analysis of 1305 proteins was performed. Pathways associated with proteins dysregulated in patients with each TIC definition and MT were identified. Results: Patients (n = 211) had a mean injury severity score of 24, with a MT and mortality rate of 22% and 12%, respectively. We identified 578 SOMAscan analytes dysregulated among MT patients, of which INR-TIC, TEG-TIC, and Clin-TIC patients showed dysregulation only in 25%, 3%, and 4% of these, respectively. TIC definitions jointly failed to show changes in 73% of the protein levels associated with MT, and failed to identify 26% of patients that received a massive transfusion. INR-TIC and TEG-TIC patients showed dysregulation of proteins significantly associated with complement activity. Proteins dysregulated in Clin-TIC or massive transfusion patients were not significantly associated with any pathway. Conclusion: These data indicate there are unexplored opportunities to identify patients at risk for massive bleeding. Only a small subset of proteins that are dysregulated in patients receiving MT are statistically significantly dysregulated among patients whose TIC is defined based solely on laboratory measurements or clinical assessment.

Abstract Cancer cells can resist the effects of DNA-damaging therapeutic agents via utilization of DNA repair pathways, suggesting that DNA repair capacity (DRC) measurements in cancer cells could be used to identify patients most likely to respond to treatment. However, the limitations of available technologies have so far precluded adoption of this approach in the clinic. We recently developed fluorescence-based multiplexed host cell reactivation (FM-HCR) assays to measure DRC in multiple pathways. Here we apply a mathematical model that uses DRC in multiple pathways to predict cellular resistance to killing by DNA-damaging agents. This model, developed using FM-HCR and drug sensitivity measurements in 24 human lymphoblastoid cell lines, was applied to a panel of 12 patient-derived xenograft (PDX) models of glioblastoma to predict glioblastoma response to treatment with the chemotherapeutic DNA-damaging agent temozolomide. This work showed that, in addition to changes in O6-methylguanine DNA methyltransferase (MGMT) activity, small changes in mismatch repair (MMR), nucleotide excision repair (NER), and homologous recombination (HR) capacity contributed to acquired temozolomide resistance in PDX models and led to reduced relative survival prolongation following temozolomide treatment of orthotopic mouse models in vivo. Our data indicate that measuring the combined status of MMR, HR, NER, and MGMT provided a more robust prediction of temozolomide resistance than assessments of MGMT activity alone. Cancer Res; 77(1); 198–206. ©2016 AACR.