1. Abstract 1.1. Background It is uncertain whether 1) patient’s characteristics (e.g., age, weight, height, and sex) influence anti-Xa heparin levels (hereafter referred to as "heparin levels"), or 2) if heparin levels influence recurrent venous thromboembolism (VTE) or bleeding events, in patients with acute VTE treated with weight-adjusted therapeutic-dose subcutaneous (SC) unfractionated heparin (UFH) or SC low-molecular weight heparin (LMWH). To determine if either association exist, we analyzed data from the Fixed-Dose Heparin (FIDO) study, in which patients were randomized to either SC UFH or SC LMWH, each given in fixed weight-adjusted doses and overlapped with 3 month of warfarin therapy for treatment of acute VTE. 1.2. Methods During the original study, 708 patients were asked to participate in a sub-study that would measure peak heparin levels while they were treated with heparin. 408 patients provided blood samples and met the eligibility criteria for the analyses in this thesis. Linear regression was used to examine the influence of patients’ baseline characteristics (e.g., age, weight, height, body mass index [BMI], sex) on heparin levels. The influence of other factors (e.g., type of heparin [UFH or LMWH]) on heparin levels was also assessed. Logistic regression was used to examine the association of heparin levels with the outcomes of 1) recurrent VTE during 3 months of follow up, and 2) bleeding events in the first 10 days of follow up. 1.3. Results: Mean heparin levels were 0.695 in patients treated with UFH, 0.698 in those treated with dalteparin and 1.034 in those treated with enoxaparin (p<0.001; R2=0.08 for variability accounted for by type of heparin). In a univariable analysis, heparin levels increase by 0.04 IU/ml (95% CI 0.02-0.07; p=0.001; R2=0.03) for every 10-kg increment in weight, by 0.02 IU/ml (95% CI 0.01-0.03; p<0.001; R2=0.04) for each unit of BMI, and by 0.03 IU/ml (95% CI 0.01-0.05; p=0.001; R2=0.03) for every 10 mol/l increment in creatinine. In a multivariable analysis, weight, BMI, and creatinine still influenced heparin levels, after adjusting for type of heparin and timing of blood sample withdrawal. Although heparin levels increased with weight, the magnitude was not large enough to suggest altering the current weight-based dosing method for LMWH. Other baseline factors such as age, height, type of VTE, creatinine clearance and hospitalization status did not influence heparin levels in patients treated with UFH or LMWH. In a univariable analysis, when heparin levels were treated as a continuous variable, higher heparin levels were associated with a lower risk of recurrent VTE at 90-days in patients treated with LMWH (OR 0.04, 95% CI 0.003-0.550, for each 1.0 IU/ml increase in heparin levels), but not in patients treated with UFH (OR 1.46, 95% CI 0.37-5.58, for each 1.0 IU/ml increase in heparin levels). In addition, higher heparin levels were associated with a higher risk of bleeding at 10-days in patients treated with UFH (OR 3.32, 95% CI 1.30-8.46 for each 1 IU/ml increase in heparin levels) but not in patients treated with LMWH (OR 3.77, 95% CI 0.42-33.92, for each 1.0 IU/ml increase in heparin levels). In a multivariable analysis, the association of heparin levels with VTE at 90-days in patients receiving LMWH (lower VTE events) and with bleeding events at 10-days in patients receiving UFH (higher bleeding events) persisted after adjusting for antiplatelet use at baseline and diagnosis of cancer at baseline. When heparin levels were treated as a dichotomous variable (subtherapeutic vs. non-subtherapeutic levels and supratherapeutic vs. non-supratherapeutic levels), the proportion of patient with recurrent VTE was significantly higher in patients with subtherapeutic levels compared with non-subtherapeutic levels in patients receiving LMWH (8.6% vs. 1.3%, p = 0.01). No significant difference was found in the proportion of patients with subtherapeutic levels and non-subtherapeutic levels in patients receiving UFH (0% vs. 3.4%, χ2=0.15, p= 0.70). The test of interaction supported the decision to analyze LMWH and UFH groups separately (p=0.02). Finally, the proportion of patient with bleeding was higher in patients with supratherapeutic compared with non-supratherapeutic heparin levels (6.5% vs. 1.5%, χ2=7.65, p=0.01). The test of interaction did not support the decision to analyze LMWH and UFH groups separately (p=0.13). 1.4. Conclusions Although it was possible to identify factors that were associated with heparin levels in patients who had been treated with weight-adjusted UFH or LMWH, none of these associations were strong enough to suggest that variables other than weight should influence SC heparin dosing. Subtherapeutic heparin levels were associated with a higher risk of recurrent VTE in patients treated with LWMH but not UFH, and supratherapeutic heparin levels were associated with a higher risk of bleeding in patients treated with UFH but not LMWH. Indirectly, these findings suggest that adjusting UFH or LMWH dose in response to heparin levels might improve clinical outcomes. / Thesis / Master of Science (MSc)