ALBERT

Description: Background:The relationship between metastatic colorectal cancer (CRC) and venous thromboembolism (VTE) is not well defined in the modern treatment era. Previous population-based studies date back to a period marked by inpatient intravenous heparin therapy and inferior survival due to paucity of therapeutic anti-cancer options, and before the advent of newer therapies including oxaliplatin, irinotecan, and anti-angiogenic treatment with bevacizumab. The objectives of this retrospective study were to examine the impact of multiple putative risk factors on VTE incidence in a large representative modern cohort of older patients with metastatic CRC. Methods:We performed a retrospective analysis of SEER-Medicare data on elderly patients with metastatic CRC diagnosed in 2004-2011. VTE and associated risk factors were analyzed using multivariate Cox proportional hazards models adjusted for sex, age at diagnosis, race, ethnicity, tumor anatomy (left/right/unknown), calendar year of diagnosis, Charlson comorbidity score, location of SEER registry and urban residence, with time-varying covariates for use of cancer therapies. Results:Of 339,778 records, 11,086 metastatic colon cancer cases were identified. 1,338 cases had VTE with a cumulative incidence of 13% at 1 year and 19% at 3 years. The mean age was 77.9 years (range 65-106). 49.7% were women and 83.5% white. 60.5% had a Charlson comorbidity score of zero at diagnosis; 6% had scores of 6-18. Significant predictors of VTE included female sex (Hazard Ratio (HR) 1.22; 95% Confidence Interval (CI) 1.10, 1.36; P

Description: Background: The cumulative incidence of symptomatic venous thromboembolism (VTE) among patients with malignant glioma (MG) is high, with estimates ranging from 20-35% during the course of therapy. VTE is the second leading cause of death among ambulatory patients receiving cancer chemotherapy, with a 2-fold or greater increase in mortality compared to those without VTE, even after adjusting for stage. In patients with MG and VTE the chance of two-year survival is significantly reduced. The development of VTE is also associated with increased morbidity, including increased risk of hospitalization, delays in cancer treatment, and an increased risk of complications including pain, bleeding, and bruising from high dose therapeutic anticoagulation. The economic burden of VTE in cancer patients is estimated at a 2-fold increase in cost. Data regarding VTE burden encompasses all cancer types and additional data is needed to characterize VTE burden in specific cancer types, particularly rarer cancers. The aim of this study was to assess the healthcare burden associated with the development of VTE in patients with MG. Methods: A retrospective chart review of patients with MG at an academic medical center was conducted from 2009-2017. Sixty-seven patients, age 18 years and older, with a histologic diagnosis of MG (WHO grade III-IV) were assessed. The number of office visits, emergency room visits, and inpatient hospitalizations with the associated costs of VTE management and its complications were collected. Using SPSS, linear regression models and descriptive statistics were used to determine the relationship between the development of VTE and healthcare consumption. Results: Of the 67 patients in the study, 18 developed VTE (27%). All patients who developed a VTE were placed on therapeutic anticoagulation. Fifty percent of patients developed complications related to anticoagulation, including gastrointestinal, retroperitoneal and intracranial bleeding events. Two patients required placement of an IVC filter after experiencing a bleeding complication. Patients that developed VTE had an increase in inpatient days (16.6 inpatient days) as compared to their non-VTE counterparts (8.8 inpatient days), (p=0.012). There was increased utilization of the emergency room with an average of 3.94 visits in those patients with VTE as compared to those without (1.84), (p=0.003). A full cost analysis found that the average primary total cost for ED visits and inpatient hospitalizations in patients with VTE was $48,863 while those without VTE averaged $35,948. This represented a 26% increase in the average primary total cost in those patients with VTE. Discussion: The development of VTE in patients with MG increases inpatient admissions days and incurs additional pharmaceutical costs related to anticoagulation. This study represents the first assessment of VTE-associated health care burden specific to primary brain cancer. We confirm that VTE affects a large number of patients with MG (27%) and limits the time spent at home due to increased hospitalizations and emergency room visits. VTE may be a preventable complication and further studies are needed to investigate safe prevention strategies for patients with MG. Disclosures No relevant conflicts of interest to declare.

Description: The storage and release by platelets of the pro-angiogenic protein, vascular endothelial growth factor (VEGF) and the anti-angiogenic protein, endostatin may influence not only wound healing but also tumor angiogenesis and metastasis. The intra-platelet mechanisms and platelet activators that control angiogenic protein release are incompletely elucidated. Previous work demonstrated that the proteinase-activated receptors (PARs), PAR-1 and PAR-4 differentially impact platelet release of these angiogenic proteins (Ma et al., PNAS 102:216). We determined the influence of the adenosine diphosphate (ADP) receptors, P2Y1 and P2Y12 on platelet release of VEGF and endostatin. Minimally altered whole blood from healthy volunteers was incubated with ADP alone [12.5 μM] or in combination with a reversible P2Y1 antagonist [MRS2179 (1.25 mM)] or a reversible P2Y12 antagonist [cangrelor (1mM)]. VEGF and endostatin release was measured with the use of an ELISA assay. ADP-induced activation of platelets increased the concentration of VEGF released by an average of 19.7 +/− 14 ng/mL (p= 0.019) but did not change the concentration of endostatin. Inhibition of the P2Y12 receptor prevented the ADP mediated release of VEGF and decreased VEGF release by 27.9 +/− 14 ng/ml (p=0.005) whereas inhibition of the P2Y1 receptor did not significantly reduce the release of VEGF (p=0.3, mean decrease 14.3 +/− 30 ng/ml) relative to ADP stimulated control. In summary, ADP-induced activation of platelets leads to the release of VEGF but not endostatin. Release of VEGF is attenuated by a P2Y12 but not a P2Y1 receptor antagonist. Accordingly, activation of platelets by ADP appears to be pro-angiogenic and inhibition by a P2Y12 antagonist should attenuate this effect.

Description: Abstract 3590 Genome-wide array or sequencing analyses are powerful tools for identifying genetic aberrations in cancers including leukaemias. However, the majority of these aberrations are likely to be random passenger events that do not drive clonal expansion. Currently, it is unknown whether cancers are maintained by a finite set of recurrent mutations similar for each patient or whether and to what extent malignancies are ‘personalised’, and also how molecular disease drivers evolve over time in the context of clinical intervention. The answers to these questions will determine whether future treatment modalities must be tailored according to individual and dynamic cancer characteristics. We hypothesized that differential quantitative high-resolution genome-wide array analysis of sequential samples from the same patients before treatment and at subsequent relapse would have the potential to identify emerging structural abnormalities with relevance to disease progression and/or response to treatment in a given patient. In order to test this hypothesis, we chose B-cell chronic lymphocytic leukaemia (CLL), because of its unique clinical characteristics, as our model to begin to evaluate the potential role of ‘companion diagnostics’ for this condition. We analysed DNA samples of 80 patients with CLL using a 1 million high resolution SNP array. On 34 of them, sequential pre-treatment and relapse samples were available. The raw data was analysed using the OncoSNP analysis tool designed in-house specifically for cancer samples as it enables quantification of copy number alterations (CNA) and copy neutral loss of heterozygosity (cnLOH) based on B-allele frequency plots in complex mixtures. This allows low levels of aberrations to be detected and for mosaic samples to be identified. Results were compared against the data from the Wellcome Trust Case Control Consortium, the DGA and germline DNA in selected cases. Large CNAs (〉1Mb) and cnLOH (〉5Mb) without deletions of 11q22.3 or 17p13.1 were identified in a third of patients. These patients had an intermediate clinical risk score that increased with the number of large CNAs. SNP array demonstrated clonal evolution in 32% of patients in the sequential sample cohort. These consisted of extension of the 13q abnormality (2), loss of the 13q deletion (1), a 10q23.1-q25.1 deletion (1), gain of 2pter-p14 (1), deletion of 2q33.1-q36.3, (1) a heterozygous deletion of 2q37.1 (1), gain of 8q22.2-qter (1), deletion of 8p (1), amplification of 8q (1), deletion of 8q (1), loss of 16p13.3 (2), mosaic deletion of 17q11.2 (1), an expansion of chromosomes carrying a 19p13.2-p13.11 gain and a 19p13.11 loss (1), deletions within 3p (1), conversion of a gain of 12p12.2-q21.31 to a copy neutral loss of heterozygosity (cnLOH) (1), deletion of 17pter-13.1 (1), increased proportion of chromosomes with the 7q33–34 deletion (1), expansion of cnLOH for 20q11.22-qter (1) and an increased number of cells with a deletion of 2q22.2-q24.1 (1) found at relapse. Importantly, most CNAs occurring at relapse were recurrent in the entire cohort implying that these are non-random events that are important in disease progression. Analysis of the minimal deleted region (MDR) of these recurrent and relapse associated CNAs revealed genes important in lymphoid development, such as NFκB2 and TRAP1 found in the alternative NFκB pathway and BLIMP1 involved in B-cell differentiation. There were also recurrent abnormalities in the region coding for SP140 which has been implicated in familial CLL. Known cancer genes were also affected by these recurrent and relapse-associated CNAs such as RND3, RIF1, RFXANK, and RHOT1, which are all members of the RAS pathway family. Using the OncoSNP program it was possible to determine that most of these emerging abnormalities were present in low numbers at diagnosis (fig. 1) suggesting that treatment does not induce the genetic alterations but may select for them. Fig .1 CNA and B-allele frequency plots of pre-treatment and relapse samples of a patient demonstrating clonal evolution. A low level of cnLOH can be seen in the B-allele frequency plot at diagnosis (a) at the end of chromosome 8 (dashed box) which becomes more apparent at relapse (b). Fig .1. CNA and B-allele frequency plots of pre-treatment and relapse samples of a patient demonstrating clonal evolution. A low level of cnLOH can be seen in the B-allele frequency plot at diagnosis (a) at the end of chromosome 8 (dashed box) which becomes more apparent at relapse (b). In conclusion, this is the first attempt to quantify CNAs in sequential leukaemia samples. The results demonstrate that recurrent and relapse associated CNAs affect genes important in B-cell development and cancer progression. Disclosures: No relevant conflicts of interest to declare.

Description: Background: Major progress has been made in understanding disease biology and therapeutic options for patients with chronic lymphocytic leukaemia (CLL). Recurrent mutations have been discovered using next generation sequencing, but with the exception of TP53 disruption their potential impact on response to treatment is unknown. In order to address this question, we characterised the genomic landscape of 250 first-line chemo-immunotherapy treated CLL patients within UK clinical trials using targeted resequencing and whole-genome SNP array. Methods: We studied patients from two UK-based Phase II randomised controlled trials (AdMIRe and ARCTIC) receiving FCR-based treatment in a first-line treatment setting. A TruSeq Custom Amplicon panel (TSCA, Illumina) was designed targeting 10 genes recurrently mutated in CLL based on recent publications.Average sequencing depth was 2260X. The cumulated length of targets sequenced was 7.87 kb from 330 amplicons covering 160 exons. Alignment and variant calling included a combination of three pipelines to confidently detect SNVs, indels and low level frequency mutations. SNP array testing was performed using HumanOmni2.5-8 BeadChips, (Illumina) and data analysed using Nexus 6.1 Discovery Edition, Biodiscovery. We performed targeted resequencing and genome-wide SNP arrays using selected samples’ germline material to confirm somatic mutations (n=40). Univariate and multivariate analyses using minimal residual disease (MRD) as the outcome measure were performed for 220 of the 250 patients. Results: Pathogenic mutations were identified in 165 (66%) patients, totalling 268 mutations in 10 genes. ATM was the most frequently mutated gene affecting 67 patients (29%) followed by SF3B1 (n=56, 24%), NOTCH1 (n= 32, 14%), TP53 (n= 21, 9%), BIRC3 (n= 17, 7%) and XPO1 (n=14, 6%). Less frequently recurrent mutations were seen in SAMHD1 (n=8, 3%), MYD88 (n= 4, 2%), MED12 (n=7, 3%) and ZFPM2 (n=5, 2%). Integrating sequencing and array results increased the patients with one or more CLL driver mutation from 66% to 94%. As previously reported del17p and TP53 mutations are co-occurring and associate with MRD positivity in all cases (n=15, p=0.0002). We report on minor TP53 subclones in 11 patients (VAF 1-5%), 8 of whom have MRD data available and were also associated with MRD positivity. Deletions of 11q were present in 44 patients. These lesions always included ATM but not always BIRC3. Bialleleic disruption was present in ATM for 27 patients (significantly associated with MRD positivity) and in BIRC3 for 4 patients. Rather surprisingly, trisomy 12 (n=33) and NOTCH1 mutations (n=28) were associated with MRD negativity (p=0.006 and 0.097, respectively). Analysing clonal and subclonal mutations per gene revealed the majority of mutations in SF3B1 and BIRC3 were subclonal (65% and 87% respectively). In contrast almost all SAMHD1 and MYD88 mutations were clonally distributed. There was an association between NOTCH1 subclonal mutations and MRD negativity, compared to clonal mutations, but this difference was not seen in the remaining mutated genes. From our copy number data, the presence of subclones was associated with MRD positivity (p=0.05). Combining important lesions in a multiple logistic regression analysis to predict MRD positivity, bialleleic ATM disruption, together with TP53 disruption, were the strongest predictors, followed by SAMHD1, whereas BIRC3 monoalleleic mutations were a medium predictor for MRD negativity. Conclusion: This is the first integrated genome-wide analysis of the distribution and associations of CLL drivers, using targeted deep resequencing and whole genome SNP arrays in an FCR-based first-line treatment setting. We have shown subclonal and clonal mutation profiles in all patients. For patients with two or more CLL-associated mutations we have begun to unravel clonal hierarchies. We have developed a comprehensive model using MRD as an outcome measure and have found bialleleic ATM mutations and SAMHD1 disruption to strongly predict for MRD positivity. Using MRD status as a robust proxy for PFS not only enables us to confirm results of previous studies, but is advantageous also in considerably reducing the timeframe for results. Indeed, we suggest that MRD status should be assessed routinely in future studies to complement modern integrated genomics approaches. Disclosures Hillmen: Pharmacyclics, Janssen, Gilead, Roche: Honoraria, Research Funding.

Description: Background:Historically, the identification of minimal deleted regions (MDRs) has been a useful approach for pinpointing genes involved in the pathogenesis of human malignancies and constitutional disorders. Microarray technology has offered increased capability for newly identifying or refining existing MDRs and minimal overlapping regions (MORs) in cancer. Despite this, in chronic lymphocytic leukemia (CLL), published MORs that pinpoint only a few candidate genes have been limited and with the advent of NGS, the utility of high resolution array work as a discovery tool has become uncertain. Here, we show that profiling copy number abnormalities (CNAs) and cnLOH using arrays in a large patient series can still be a valuable approach for the identification of genes that are disrupted or mutated in CLL and have a role in CLL development and/or progression. Methods: 250 CLL patient DNAs from individuals enrolled in two UK-based Phase II randomised controlled trials (AdMIRe and ARCTIC trials) were tested using Infinium HumanOmni2.5-8 v1.1 according to manufacturer’s guidelines (Illumina Inc, San Diego, CA). Data were processed using GenomeStudioV2009.2 (Illumina Inc.) and analysed using Nexus Discovery Edition v6.1 (BioDiscovery, Hawthorne, CA). All Nexus plots were inspected visually to verify calls made, identify uncalled events and exclude likely false positives. To exclude common germline CNVs, the Database of Genomic Variants (DGV), a comprehensive catalog of structural variation in control data, was used. Copy number (CN) changes that encompassed fully changes noted in the DGV were excluded from further analysis. Regions of copy neutral loss of heterozygosity (cnLOH) were recorded if 〉1Mb in size, but were not used to define or refine MORs. Data from 1275 age-appropriate control samples minimised the reporting of common cnLOH events. All genomic coordinates were noted with reference to the GRCh37, hg19 assembly. MORs were investigated using Microsoft Excel filtering functions. A subset of genes (n=91) selected from MORs mainly on the basis of event frequency and/or number of genes within the MOR and/or literature interest were taken forward for targeted sequencing (exons only) of appropriate samples with/without CN Losses or cnLOH (Set 1 n=124; Set 2 n=126). These were tested using custom designed TruSeq Custom Amplicon panels (Illumina Inc) and processed according to manufacturer’s instructions. SAMHD1 was excluded from these panels since it had been studied separately within our laboratory. The data were analysed using an in-house bioinformatics pipeline that uses the sequence aligners MSR and Stampy and the variant callers GATK and Platypus, followed by stringent filtering. Results: Using our datasets we have identified 〉50 MORs previously unreported in the literature. Six of these showed copy number (CN) losses in 〉3% of patients studied. Furthermore, we have refined 14 MORs that overlapped with regions described previously and that had also a CN loss frequency of 〉3%. Thirteen MORs involved only a single reference gene, often a gene implicated previously in cancer (eg. SAMHD1, MTSS1, DCC and RFC1). Of the 91 genes taken forward for targeted sequencing, stringent data filtering led to a subset of 19 genes of interest harbouring exonic mutations. Genes with mutations identified include DCC, BAP1 and FBXW7, also implicated previously in cancer. Conclusion: We have generated high resolution CNA and cnLOH profiles for 250 first-line chemo-immunotherapy treated CLL patients and used this information to document newly identified MORs, to refine MORs reported previously and to identify mutation harbouring genes using targeted NGS. Functional knowledge supports our hypothesis that these genes may have a contributory role in CLL. For two genes, SAMHD1 and FBXW7, relevance in CLL has been established already. Taken together, our data validate the utility of high resolution arrays studies for the identification of candidate genes that may be involved in CLL development or progression when disrupted. Further studies are required to confirm a role for these genes in CLL and to elucidate the nature of the underlying biological mechanisms. Disclosures No relevant conflicts of interest to declare.

Description: Background. Cancer patients often display procoagulant activity and are at a risk for developing venous thromboembolism (VTE). Multiple mechanisms may explain this procoagulant state, such as tumor cell expression of tissue factor (TF), host tissue response to tumor formation and also contributions from comorbid factors and treatment. We previously observed elevated levels of traditional biomarkers of coagulation (such as D-dimer, CRP, etc.) in patients with advanced cancer on therapy enrolled in a randomized trial comparing rosuvastatin versus placebo. In this study, we evaluated the frequency and concentration of endogenous plasma TF, Factor (F)XIa and FIXa in the cancer patient population analyzed for traditional coagulation biomarkers. The choice of new analytes was based on previous studies in which these 3 proteins were detected and quantitated in patients with cardiovascular diseases, inflammation, and trauma - often present over the course of several days to even weeks. By our knowledge, there are no published studies of the presence of FXIa and/or FIXa in cancer patients. Methods. Thirty-seven adult cancer patients receiving systemic therapy were originally enrolled in a randomized crossover trial comparing effects of rosuvastatin versus placebo on biomarkers of VTE. Inclusion criteria included locally-advanced or metastatic cancer, estimated survival time of greater than 6 months, and anticipated duration of therapy of at least 3 months. Patients on antithrombotic or statin therapy were excluded. Blood was collected on up to 4 occasions; at the start and end of each of 2 treatment periods corresponding to 2 cycles of systemic therapy with a 3-4 week washout period. Citrate plasma was prepared, frozen and stored at -80C. The assay, performed in previously unthawed and contact-pathway inhibited plasma, is based on a response of the lag phase of thrombin generation to corresponding monoclonal inhibitory antibodies (αFXIa-2, αFIXa-91 and αTF-5; all at 0.1 mg/mL). Concentrations of TF, FXIa and FIXa were calculated from corresponding calibration curves built by titrating purified proteins into multi-donor pooled plasma from healthy individuals. Results. A total of 116 blood samples were collected (not all patients were able to complete their regime) and analyzed for endogenous levels of TF, FXIa and FIXa (see table). Overall, rosuvastatin treatment did not have any significant impact on the levels and frequency of these 3 proteins. For the samples obtained from each patient's first blood draw (n = 37) there was a weak correlation between FXIa and FIXa levels (R = 0.37, P = 0.02) and a moderate correlation between FXIa and TF levels (R = 0.57, P 〈 0.001), while no correlation between FIXa and TF was observed (R = 0.02, P = 0.91). Conclusions. 1) A portion of patients with advanced cancer had active TF in their plasma while the majority, for the first time reported, had active FXIa and FIXa; 2) No pronounced differences were observed between the frequency and concentrations of these proteins at baseline versus later time-points; 3) There was a weak correlation between FXIa and FIXa and a moderate correlation between FXIa and TF, suggesting the TF-pathway is driving the majority of FXIa generation, although contribution from the contact pathway cannot be excluded; 4) Rosuvastatin treatment did not lead to any significant changes in the levels of TF, FXIa and/or FIXa. Table Table. Disclosures No relevant conflicts of interest to declare.

Description: Background: Recent clinical trials have evaluated the safety and efficacy of direct oral anticoagulant (DOAC) therapy for venous thromboembolism (VTE) prophylaxis in cancer outpatients at high risk for thrombosis. Bleeding risk in these trials were approximately 2% over the first 6 months of therapy. For individual patients, the utility of prophylactic anticoagulation (AC) depends on an acceptable safety profile between bleeding and thrombosis. We investigated whether ambulatory cancer patients on contemporary cancer-directed therapies and prophylactic AC had an increased risk of major bleeds over the first 6 months of therapy. Methods: As part of a single-center prospective cohort study, we assessed consecutive ambulatory patients initiating cancer-directed treatment, risk-stratified these patients for VTE using the Khorana score and educated them about VTE. High risk patients (Khorana score ≥3) were offered prophylactic AC. Major bleeding events and minor bleeding events (based on ISTH standard definitions) were prospectively captured via billing code screening and confirmed by physician review of the medical record. Logistic regression was used to compare the odds of developing a major bleed within 6 months in those who received prophylactic AC compared to those that did not. Results: A total of 1,210 patients were enrolled from October 2015 - June 2018, of which, 640 were women (52.9%). The most common cancers were gastrointestinal 270 (22%), lung 213 (18%), and breast 198 (16%). There were 393 patients (32%) with a Khorana score of 0, 706 (58%) with a Khorana score of 1-2, and 111 (9%) with a score of ≥3. A total of 421 patients received any AC (LMWH or DOAC). Of these, 282 received a prophylactic dose anticoagulant and 139 were receiving a full dose anticoagulant prior to enrolling for other medical reasons. Prophylactic dose anticoagulants prescribed included apixaban in 107 (41%), rivaroxaban in 6 (2.3%), enoxaparin in 119 (45.7%), and other heparin products in 50 (19.2%). A total of 27 (2.33%) major bleeds and 22 (1.81%) minor bleeds occurred within the first 6 months of starting therapy. Of these bleeding events, 8 (2.8%) occurred in those on prophylactic AC, and 6 (4.3%) occurred in those on full dose AC. The odds ratio (OR) of developing a major bleed on any type of AC was 1.78 [CI 0.817-3.88]. The OR of major bleeding on prophylactic AC was 1.49 [CI 0.64-3.479]. The OR of major bleed was highest in lung cancer patients on prophylactic AC (OR 2.81, CI 1.27-6.25). Men, when compared to women, were more likely to bleed on prophylactic AC in the first six month (OR 0.2; CI 0.07-0.52). The OR for major bleed with each 1 year increase in age was 1.02 (CI 0.99, 1.06). The OR of bleeding with a high risk Khorana score (≥3) compared to a lower score was 1.04 (CI 0.73-1.50). Conclusion: During the first six months of therapy, prophylactic AC was associated with an increased risk of major bleeding events in patients on cancer-directed therapy. In this study, the rate of major bleeding was similar as compared to published clinical trials. Neither age nor higher Khorana score were associated with an increased risk of major bleeds in patients on prophylactic AC. The finding that men, when compared to women, and patients with lung cancer may have an increased risk of major bleeding while on prophylactic AC and cancer-directed chemotherapy suggests these groups may warrant both increased education and monitoring to ensure safety while on prophylactic AC. Disclosures No relevant conflicts of interest to declare.

Description: Introduction: Among all venous thromboembolism (VTE) events, 20% occur during or in the three months following a hospitalization. Inpatient VTE prevention strategies incorporate pharmacological interventions that do not commonly continue after discharge. Recent trials of post-discharge prophylaxis have not identified benefit among recently discharged patients, probably due to lack of understanding of the relative and absolute risks of VTE after discharge. In order to better clarify the scope of the problem, we quantified the risk of VTE during hospitalizations, and up to 3 months post-discharge relative to patients without a recent hospitalization. Methods: We followed all primary care patients aged ≥18, at the University of Vermont Medical Center's primary care clinics for hospitalization and VTE using the electronic medical record. Patients entered the cohort at their first contact with the health system after 06/30/2010 and were followed until their last contact, change of primary care provider to a non-University of Vermont provider, or December 2016, whichever occurred first. First VTE during follow-up was defined as having two outpatient VTE codes 7-185 days apart or one inpatient code. Patients with a VTE code in the first 3 months of enrollment were excluded as having pre-baseline VTE. Hospital-acquired VTE was defined as VTE occurring after day 1 of hospitalization (to correctly categorize people with an outpatient VTE and admitted with VTE) and post-discharge VTE was defined as VTE occurring during successive 1-month time periods after discharge from the hospital. Age- and sex-adjusted Cox proportional hazard models were used to calculate hazard ratios (HRs) of VTE during and after hospitalization. Hospitalization and time periods after hospitalization were modeled as time-varying covariates and included the time in the hospital, and 1-30 (month 1), 31-60 (month 2), and 61-90 (month 3) days post-discharge. The reference group for this analysis was outpatients who were not within 90 days of a hospitalization. Results: From 2010-16, 87,821 patients (49,468 women, 56%) were included, with a mean age of 46 years. With 371,429 person-years of follow-up (mean follow-up 4.2 years), 749 first VTE events occurred for a rate of 2.0 per 1,000 person-years. There were 57,837 hospitalizations among 21,963 individuals for a rate 155.7 hospitalizations per 1,000 person-years. The Table presents the person-years of follow-up, the number, rate, and age- and sex-adjusted hazard of VTE for hospitalization and successive months after hospitalization. The VTE rate was 1.5 per 1,000 person-years outpatients not within 3 months of hospitalization. During hospitalization, the rate was 75.9 per 1,000 person-years, and slowly decreased over time after discharge, but remained elevated at 5.0 per 1,000 person-years event in month 3 after discharge. The age- and sex-adjusted HR for VTE was 40.3 during hospitalization and declined to 18.1, 6.0 and 2.9 over successive 1-month intervals after discharge (Table). Conclusion: In a primary care population in northern Vermont, hospitalization and time periods after hospitalization were associated with a dramatically increased incidence of and risk for VTE compared with patients not hospitalized in the past 3 months. The rate of VTE (34.4 per 1,000 person-years) in the first month post-discharge might not warrant universal post-discharge VTE prophylaxis, but suggests that if we can identify high-risk patients at low risk for bleeding, pharmacologic VTE prophylaxis may be appropriate in some patients. Further study of risk factors for post-discharge VTE is warranted. Table Disclosures No relevant conflicts of interest to declare.

Description: Introduction: Anemia by the WHO criteria is more common in American blacks than whites. There are few data examining differential associations of demographic, socio-economic, and co-morbid conditions with anemia by race. Methods: The Reasons for Geographic and Racial Differences in Stroke (REGARDS) study is a national longitudinal cohort currently enrolling 30,000 blacks and whites aged ≥ 45 years to assess stroke risk. Half the cohort will be black and half will live in the southeast. A phone interview for health history was followed by an in-home visit for phlebotomy and physical exam. After enrolling 8400 subjects, a complete blood count was added to the baseline exam. As of March 30th 2006 this was available in 12,060 participants. Anemia was defined by the WHO criteria and categorized into 3 non-exclusive groups: (1) low glomerular filtration rate (GFR) (