Pulmonary embolism (PE) studies used direct oral anticoagulants (DOACs) with or without initial heparin. We aimed to (1) evaluate if PE patients benefit from initial heparin; (2) describe patient characteristics in the DOAC studies; and (3) investigate whether the anatomical extent of PE correlates with N-terminal pro-brain natriuretic peptide (NT-proBNP) levels, cause of PE, and recurrence rate. Our methods were (1) an indirect meta-analysis comparing the recurrence risk in DOAC-treated patients with or without initial heparin to those patients given heparin/vitamin K antagonist (VKA). (2) To compare the PE studies, information was extracted on baseline characteristics including anatomical extent. (3) The Hokusai-VTE study was used to correlate anatomical extent of PE with NT-proBNP levels, causes of PE, and recurrent venous thromboembolism (VTE). The meta-analysis included 11,539 PE patients. The relative risk of recurrent VTE with DOACs versus heparin/VKAs was 0.8 (95% confidence interval [CI]: 0.6-1.1) with heparin lead-in and 1.1 (95% CI: 0.8-1.5) without heparin. In the DOAC studies, the proportion of patients with extensive PE varied from 24 to 47%. In Hokusai-VTE, NT-proBNP was elevated in 4% of patients with limited and in over 60% of patients with extensive disease. Cause of PE and anatomical extent were not related. Recurrence rates increased from 1.6% with limited to 3.2% with extensive disease in heparin/edoxaban-treated patients, and from 2.4 to 3.9% in heparin/warfarin recipients. In conclusion, indirect evidence suggests a heparin lead-in before DOACs may be advantageous in PE. Anatomical extent was related to elevated NT-proBNP and outcome, but not to PE cause.