Journal of the American Heart Association ORIGINAL RESEARCH Association Between Preexisting Versus Newly Identified Atrial Fibrillation and Outcomes of Patients With Acute Pulmonary Embolism Behnood Bikdeli , MD, MS*; David Jiménez , MD, PhD*; Jorge del Toro , MD; Gregory Piazza , MD, MS; Agustina Rivas, MD; José Luis Fernández-­Reyes, MD; Ángel Sampériz, MD, PhD; Remedios Otero, MD, PhD; José María Suriñach, MD, PhD; Carmine Siniscalchi , MD, PhD; Javier Miguel Martín-­Guerra, MD; Joaquín Castro , MD; Alfonso Muriel , PhD; Gregory Y. H. Lip , MD; Samuel Z. Goldhaber , MD; Manuel Monreal , MD, PhD; For the RIETE Investigators† BACKGROUND: Atrial fibrillation (AF) may exist before or occur early in the course of pulmonary embolism (PE). We determined the PE outcomes based on the presence and timing of AF. METHODS AND RESULTS: Using the data from a multicenter PE registry, we identified 3 groups: (1) those with preexisting AF, (2) patients with new AF within 2 days from acute PE (incident AF), and (3) patients without AF. We assessed the 90-­day and 1-­year risk of mortality and stroke in patients with AF, compared with those without AF (reference group). Among 16 497 pa- tients with PE, 792 had preexisting AF. These patients had increased odds of 90-­day all-­cause (odds ratio [OR], 2.81; 95% CI, 2.33–­3.38) and PE-­related mortality (OR, 2.38; 95% CI, 1.37–­4.14) and increased 1-­year hazard for ischemic stroke (hazard ratio, 5.48; 95% CI, 3.10–­9.69) compared with those without AF. After multivariable adjustment, preexisting AF was associated with significantly increased odds of all-­cause mortality (OR, 1.91; 95% CI, 1.57–­2.32) but not PE-­related mortality (OR, 1.50; 95% CI, 0.85–­2.66). Among 16 497 patients with PE, 445 developed new incident AF within 2 days of acute PE. Incident AF was associated with increased odds of 90-­day all-­cause (OR, 2.28; 95% CI, 1.75–­2.97) and PE-­related (OR, 3.64; 95% CI, 2.01–­6.59) mortality but not stroke. Findings were similar in multivariable analyses. CONCLUSIONS: In patients with acute symptomatic PE, both preexisting AF and incident AF predict adverse clinical outcomes. The type of adverse outcomes may differ depending on the timing of AF onset. Key Words: atrial fibrillation ■ mortality ■ outcome ■ pulmonary embolism P ulmonary embolism (PE) is a thromboembolic simplified Pulmonary Embolism Severity Index (sPESI).7 disease with potentially serious or fatal com- Nevertheless, risk assessment remains suboptimal for plications.1–­3 Understanding the risk profile for several patient subgroups. Therefore, finding additional patients with acute PE is critical, because timely as- ways to predict adverse outcomes is crucial. sessment may help with risk stratification.4 Several Atrial fibrillation (AF) is a common arrhythmia,8 as- tools exist for such risk assessment,5,6 including the sociated with increased rates of stroke, myocardial Correspondence to: Behnood Bikdeli, MD, MS, Cardiovascular Medicine Division, Brigham and Women’s Hospital, 75 Francis Street, Shapiro 5, Suite 5156. Boston, MA 02115. E-­mail: bbikdeli@bwh.harvard.edu; behnood.bikdeli@yale.edu *B. Bikdeli and D. Jiménez contributed equally. † A full list of the RIETE Investigators can be found in the Appendix at the end of the article. Supplementary Material for this article is available at https://www.ahajo​urnals.org/doi/suppl/​10.1161/JAHA.121.021467 For Sources of Funding and Disclosures, see page 11. © 2021 The Authors. Published on behalf of the American Heart Association, Inc., by Wiley. This is an open access article under the terms of the Creative Commons Attribution-­NonCommercial-­NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-­commercial and no modifications or adaptations are made. JAHA is available at: www.ahajournals.org/journal/jaha J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214671 Bikdeli et al Preexisting vs Incident AF in Patients With PE treatment patterns, and outcomes according to the CLINICAL PERSPECTIVE time course of AF. We focused on (1) preexisting AF diagnosed before the index PE event, and (2) new AF What Is New? diagnosed within the first 2 days of diagnosis of acute • This study demonstrated that atrial fibrilla- PE (incident AF). We considered patients with PE who tion (AF) may exist before or occur early in the did not have AF as referent. course of acute pulmonary embolism and lead to worse clinical outcomes, which may be de- pendent on the timing of the AF: preexisting METHODS AF predicts all-­cause mortality and 1-­year risk The methods implemented for the current study are of stroke whereas new (incident) AF predicts all-­cause mortality and pulmonary embolism-­ summarized. Further details about the study method- related mortality. ology or the underlying data will be made available to interested investigators, upon submitting a reasonable What Are the Clinical Implications? research request to the principal investigators (B.B., • The presence of coexisting AF in patients with D.J., and M.M.) and approval of the RIETE advisory pulmonary embolism is indicative of higher risk committee. for adverse events and should be assessed in future studies to determine whether it improves risk stratification and can direct more optimal Data Source use of treatment strategies. RIETE is a multicenter international registry of con- secutive patients with objectively confirmed venous thromboembolism (VTE; deep vein thrombosis, PE, or splanchnic vein thrombosis). The methodology of the Nonstandard Abbreviations and Acronyms registry has been described previously.21 The proto- col for patient enrollment was approved by the par- RIETE Registro Informatizado de Enfermedad ticipating centers. Data were stored and audited by TromboEmbólica the RIETE coordinating center (S&H Medical Sciences sPESI simplified Pulmonary Embolism Severity Services, Madrid, Spain). All study participants pro- Index vided informed consent, in accordance with local insti- tutional review board criteria. For the current study, we included patients with PE from April 2, 2014 to January infarction, and heart failure.9,10 The presence of AF 31, 2020. This choice was made to include patients may affect the thromboembolic and hemorrhagic risk for whom data elements related to AF were available, across a wide spectrum of cardiovascular conditions. while avoiding the inclusion of patients with VTE in the For example, investigations related to coexisting AF setting of the COVID-­19 pandemic.22,23 and coronary disease have led to tailored strategies to address the thrombotic risk.11–­13 Patients The association between PE and AF has received recent attention.14,15 Presentations of AF occurring be- Patients with acute symptomatic PE were considered for fore15–­17 or early in the course of18 acute PE have been inclusion in this study. We excluded patients with asymp- described. However, the majority of the prior studies tomatic PE. We hypothesized that patients with preex- of the AF-­PE association has been based on a small isting AF and patients with incident AF may each have number of patients or on administrative data.15,19 Little different characteristics and outcomes compared with is known about the comparative presentation and out- those without AF. However, because these differences comes according to the time of AF onset with respect may be epidemiologically and pathophysiologically dis- to PE. In addition, it remains unknown whether AF is tinct, we prespecified the analysis of these 2 subgroups merely a surrogate for cardiovascular risk, associated of AF separately. For all analyses, patients without pre- with increased all-­cause mortality, or if it is associated existing AF or incident AF were considered as referent. with PE-­related mortality. Further, it is unclear whether the associations, if present, are because of differences Patient Characteristics in baseline characteristics or whether they persist after We assessed the demographics and comorbidities multivariable adjustment.15,20 such as diabetes mellitus, hypertension, atheroscle- Using the data from the RIETE (Registro Informati­ rotic vascular disease, chronic pulmonary disease, zado de Enfermedad Tromboembólica) registry, we stroke, cancer, and recent surgery (ie, within 2 months sought to identify patients with PE and coexisting AF and before enrollment in RIETE). Treatment patterns for profile their PE presentation, coexisting characteristics, each group, including anticoagulant therapy and J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214672 Bikdeli et al Preexisting vs Incident AF in Patients With PE advanced therapies (including fibrinolytic therapy and regression models for adjustment and reported odds percutaneous or surgical thrombectomy) were evalu- ratio (OR) as the effect measure. ated. For patients who remained alive at follow-­up, we For 1-­year all-­cause mortality, we used Cox pro- explored within each group the proportion of patients portional hazard models and reported hazard ratios who remained on anticoagulation at 90 days and at (HR) as the effect measure. For outcomes other than 1-­year follow-­up. all-­cause mortality (ie, PE-­related mortality, thrombo- sis composite outcome, and major bleeding), we used Outcomes Fine-­Gray hazard models that accounted for the com- peting risk of death from events other than the out- The co-­primary outcomes were PE-­related mortality at come of interest. Survival estimates were presented 90 days from enrollment and all-­cause mortality at 1-­ with Kaplan-­Meier curves for all-­cause mortality and year follow-­up. The prespecified secondary end points competing risk regression (cumulative incidence func- were the 1-­year rate of ischemic stroke and the com- tion with Fine-­Gray method) for PE-­related mortality. posite of thrombotic events (consisting of myocardial Variables used for adjustment were prespecified and infarction, ischemic stroke, or subsequent VTE). Other included age (continuous), sex, sPESI, and history of study end points included 90-­day all-­cause mortality diabetes mellitus, heart failure, prior ischemic stroke, and 1-­year PE-­related mortality. Because we expected and arterial vascular disease (either coronary or pe- fewer short-­term events, in order to minimize the type ripheral artery disease). In all models, those without I error rate, we prespecified that we would assess history of AF were the reference group. stroke, recurrent VTE, and also composite thrombotic For the assessment of the co-­primary outcomes, end points only at 1-­year follow-­up. because 90-­day PE-­related mortality rates are likely to The main safety outcomes were major bleeding affect 1-­year mortality, we did not adjust the alpha level. at 90-­day and 1-­year follow-­up. Major bleeding was However, it was planned to compare each of the pa- defined as events that required transfusion of ≥2 units tient subgroups with AF (ie, those with preexisting AF, of blood or were retroperitoneal, spinal, intracranial, as well as those with new incident AF) against those intrapericardial, or fatal. This definition closely resem- patients who did not have AF. Therefore, it was pre- bles that of the International Society on Thrombosis specified to set the significance level for the co-­primary and Haemostasis.24 In all analyses, patients without AF outcomes at P<0.025 (ie, reducing the alpha to half). All were considered as reference group. other comparisons were considered exploratory. Statistical Analysis Categorical variables were reported as percentages. RESULTS Continuous variables were reported as mean and Of 92 258 patients in RIETE, 48 554 had PE. Based standard error of the mean (or median and interquar- on the inclusion and exclusion criteria, 16 497 patients tile range, where not normally distributed). To minimize with acute PE were included. Of those, 792 (4.8%) had type I error, it was prespecified to avoid unselected known preexisting AF, whereas 445 (2.7%) were diag- pairwise hypothesis testing for the baseline charac- nosed with new incident AF within 48 hours of the di- teristics between groups.25 We planned to report the agnosis of acute PE (Figure 1). relative frequencies for categorical variables with 99% CIs. Post hoc chi square tests were considered only if there were nonoverlapping 99% CIs of clinical rel- Baseline Characteristics and Presenting evance. For baseline numeric variables, to avoid infla- Features tion of the type I error, we prespecified to run tests of Patients with preexisting AF had differing baseline statistical comparisons only among the variables that characteristics compared with those without AF. reflected clinically meaningful differences (as decided Patients with preexisting AF were older (77.9±0.4 by the study steering committee based on the magni- versus 65.7±0.1 years), more likely to have diabetes tude of observed difference and expected difference mellitus (25.3% versus 15.9%), hypertension (77.0% based on clinical consensus), in addition to age and versus 50.2%), and history of heart failure (33.8% ver- the sPESI, which were prespecified.7 No a priori power sus 6.7%), (P<0.001 for pairwise all comparisons). The calculation was made for this study. proportion of patients with hypotension (3.6% versus In general, logistic regression models are more 3.0, P=0.29) or syncope (12.8% versus 14.1%, P=0.33) powerful and are less prone to statistical assump- at presentation was comparable between those with tions. As a unique strength of the registry, RIETE has preexisting AF and patients without AF. Patients with complete 100% follow-­up for all patients until 90 days preexisting AF had higher CHA2DS2-­VASC and sPESI from the index VTE event or until death. Accordingly, scores compared with those without AF (P<0.001 for for assessment of 90-­day outcomes, we used logistic both comparisons [Table 1]). J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214673 Bikdeli et al Preexisting vs Incident AF in Patients With PE Figure 1. Patient selection flow diagram. AF indicates atrial fibrillation; DVT, deep vein thrombosis; PE, pulmonary embolism; and RIETE, Registro Informatizado de Enfermedad Tromboembólica. Patients with incident AF also had important dif- follow-­up, a greater percentage of patients with preexist- ferences in presenting features, compared with pa- ing AF were on anticoagulation than those who did not tients without AF. Comorbidities such as hypertension have a history of AF (75.2% versus 63.2%, P<0.001). (63.9% versus 50.2%, P<0.001) and heart failure Similarly, compared with those without AF, a greater pro- (10.8% versus 6.7%, P=0.001) were more frequently portion of patients with incident AF were on anticoagula- observed in patients with incident AF than those with- tion at 1-­year follow-­up (82.9% versus 63.2%, P<0.001). out AF. Further, patients with incident AF were more likely to present with hypotension (7.6% versus 3.0%, P<0.001) or syncope (18.0% versus 14.1%, P=0.02). As Outcomes at 90-­Day Follow-­Up shown in Table 1, patients with incident AF had higher All-­Cause Mortality CHA2DS2-­ VASC and sPESI scores compared with In unadjusted analyses, patients with preexisting AF those without AF (P<0.001 for both comparisons). The had higher odds of 90-­day mortality compared with gradient of risk according to the European Society of those without AF (16.0% versus 6.1%; OR, 2.81; 95% Cardiology classification is summarized in Table S1. CI, 2.33–­3.38). Findings were attenuated but consist- ent after multivariable adjustment (OR, 1.91; 95% CI, Treatment Patterns 1.57–­2.32). Most patients in all 3 groups received initial anticoagula- Patients with incident AF, compared with patients tion. Use of percutaneous and surgical thrombectomy without AF, had higher odds of 90-­day mortality in un- was rare (Table 2). Thrombolytic therapy was used in adjusted analyses (13.0% versus 6.1%; OR, 2.28; 95% 1.1% of patients with preexisting AF, 4.3% of patients with CI, 1.75–­2.97). Results were similar after multivariable incident AF, and 2.8% of patients without AF. At 1-­year adjustment (OR, 1.61; 95% CI, 1.23–­2.10) (Figure 2A). J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214674 Bikdeli et al Preexisting vs Incident AF in Patients With PE Table 1. Cohort Characteristics of Patients With PE According to Presence and Time-­Course of AF Incident AF (diagnosed within Known preexisting AF 2 d after PE) No AF at all Number of patients (total N=16 497) N=792 N=445 N=15 260 Demographics Female sex, % 56.1% 52.8% 51.3% (99% CI, 51.4–­60.6) (99% CI, 46.5–­58.9%) (99% CI, 50.3–­52.3%) Age, y±SEM 77.9±0.4 75.2±0.6 65.7±0.1 Age, y, median±IQR 80 (72–­86) 78 (68–­84) 69 (55–­79) Prior history Current smoker 7.1% 9.3% 14.1% (99% CI, 4.9–­9.8%) (99% CI, 6.0–­13.4%) (99% CI, 13.4–­14.9%) Diabetes mellitus 25.3% 18.1% 15.9% (99% CI, 21.4–­29.5%) (99% CI, 13.5–­23.2%) (99% CI, 15.1–­16.7%) Hypertension 77.0% 63.9% 50.2% (99% CI, 72.9–­80.1%) (99% CI, 57.7–­69.8%) (99% CI, 4.9.1–­51.2%) Coronary artery disease 15.7% 9.1% 6.7% (99% CI, 12.5–­19.4%) (99% CI, 6.0%–­13.3%) (99% CI, 6.1–­7.2%) Peripheral arterial disease 8.5% 4.2% 3.5% (99% CI, 6.1–­114%) (99% CI, 2.1–­7.3%) (99% CI, 3.1–­3.9%) Coronary or peripheral arterial disease 22.0% 12.5% 9.4% (99% CI, 18.1–­26.1%) (99% CI, 8.7–­17.1%) (99% CI, 8.8–­10.0%) Heart failure 33.8% 10.8% 6.7% (99% CI, 29.6–­38.3%) (99% CI, 7.3–­15.1%) (99% CI, 6.1–­7.2%) Ischemic stroke 16.1% 11.7% 6.3% (99% CI, 12.9–­19.8%) (99% CI, 8.1–­16.2%) (99% CI, 5.8–­6.9) Prior (old) venous thromboembolism 13.4% 9.0% 14.1% (99% CI, 10.4–­16.8%) (99% CI, 5.8–­13.0%) (99% CI, 13.4–­14.8%) Coexisting deep vein thrombosis diagnosed 23.7% 26.1% 29.5% with the index PE event (99% CI, 19.9–­27.8%) (99% CI, 20.1–­31.8%) (99% CI, 28.6–­30.5%) Chronic lung disease 20.7% 12.4% 13.5% (99% CI, 17.1–­24.7%) (99% CI, 8.7–­16.9%) (99% CI, 12.8–­14.3%) Active cancer 17.3% 16.9% 16.7% (99% CI, 14.0–­21.0%) (99% CI, 12.5–­21.9%) (99% CI, 15.9–­17.5%) Anemia 39.3% 33.0% 30.9% (99% CI, 34.8–­43.9%) (99% CI, 27.4–­39.1%) (99% CI, 29.9–­31.9) Creatinine clearance levels, mL/min 73.3±247.3 66.4±31.5 87.1±64.3 Clinical factors Median (IQR) heart rate 90 (75–­110) 101 (85–­126) 90 (77–­103) Systolic blood pressure <90 mm Hg 3.6% 7.6% 3.0% (99% CI, 2.1–­5.7%) (99% CI, 4.8–­11.5%) (99% CI, 2.7–­3.4%) Syncope 12.8% 18.0% 14.1% (99% CI, 9.9–­16.2%) (99% CI, 13.5–­23.1%) (99% CI, 13.4–­14.8) Simplified Pulmonary Embolism Severity 2 (1–­2) 1 (1–­2) 1 (0–­2) Index score, median±IQR CHA 2DS2VASC Score, median±IQR 4 (3–­5) 3 (2–­4) 2 (1–­4) Use of echocardiography within the first 3 d 56.6 69.7% 58.5% of PE diagnosis (99% CI, 51.7–­61.5%) (99% CI, 63.7–­75.3%) (99% CI, 57.4–­59.6%) Aspirin use at baseline 35.7% 26.2% 18.9% Categorical variables have been reported with their respective 99% CIs. Pairwise statistical comparisons were intentionally avoided to minimize type I error. AF indicates atrial fibrillation; IQR, interquartile range; and PE, pulmonary embolism. Thrombotic Outcomes after multivariable adjustment (OR, 1.50; 95% CI, Patients with preexisting AF had higher odds of 90-­ 0.85–­2.66). day PE-­related mortality compared with those without When compared with patients without AF, patients AF (1.8% versus 0.76%; OR, 2.38; 95% CI, 1.37–­4.14, with incident AF had higher odds of 90-­day PE-­related P=0.002). However, the results were not significant mortality both in unadjusted (2.7% versus 0.76%; OR, J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214675 Bikdeli et al Preexisting vs Incident AF in Patients With PE Table 2. Treatment Patterns Incident AF (diagnosed within Known preexisting AF 2 d after PE) No AF at all N=792 N=445 N=15 260 Initial therapy Any anticoagulant therapy N=16 440 N=787 N=444 N=15 209 Low-­molecular-­weight heparin 676 (86%) 397 (89%) 12 696 (83%) Unfractionated heparin 42 (5.3%) 18 (4.1%) 879 (5.8%) Fondaparinux 14 (1.8%) 2 (0.45%) 227 (1.5%) Direct oral anticoagulants 23 (2.9%) 4 (0.90%) 902 (5.9%) Thrombolytic therapy 9 (1.1%) 19 (4.3%) 420 (2.8%) Surgical or percutaneous thrombectomy N=14 762 N=674 N=400 N=13 688 4 (0.59%) 3 (0.75%) 169 (1.2%) Vena cava filter use 18 (2.3%) 20 (4.5%) 377 (2.5%) No anticoagulant therapy 2 (0.25%) 1 (0.22%) 16 (0.10%) Long-­term treatment (among survivors with valid follow-­up information) Vitamin K antagonists 274 (35%) 201 (45%) 6942 (47%) Low-­molecular-­weight heparin 235 (30%) 128 (29%) 4214 (28%) Direct oral anticoagulants 210 (27%) 83 (19%) 3504 (24%) Fondaparinux 4 (0.51%) 3 (0.67%) 78 (0.53%) Unfractionated heparin 4 (0.51%) 0 32 (0.22%) Others 5 (0.63%) 2 (0.45%) 30 (0.20%) Anticoagulation at 90 d Patients alive and followed 664 386 14 306 Patients alive and still on anticoagulation 620 (93.4%) 372 (96.4%) 13 801 (96.5%) Anticoagulation at 180 d Patients alive and followed 434 252 9250 Patients alive and still on anticoagulation 369 (85.0%) 226 (89.7%) 7941 (85.8%) Anticoagulation at 365 d Patients alive and followed 258 164 5607 Patients alive and still on anticoagulation 194 (75.2%) 136 (82.9%) 3541 (63.2%) AF indicates atrial fibrillation; and PE, pulmonary embolism. 3.64; 95% CI, 2.01–­6.59, P<0.001) and multivariable ad- AF (HR, 2.62; 95% CI, 2.23–­3.07; P<0.001). Findings justed analyses (OR, 2.29; 95% CI, 1.25–­4.18). (Figure 2B). were attenuated but consistent after multivariable ad- justment (HR, 1.77; 95% CI, 1.50–­2.10) (Figure 3A). Bleeding Patients with incident AF, compared with patients with- out AF, had higher risk of 1-­year all-­cause mortality in Patients with preexisting AF had higher odds of 90-­day unadjusted analyses (HR, 1.86; 95% CI, 1.46–­ 2.36; major bleeding compared with those without AF (3.5% P<0.001). Results were attenuated but consistent in versus 2.0%; OR, 1.87; 95% CI, 1.27–­2.75). The associa- multivariable adjustment (HR, 1.34; 95% CI, 1.05–­1.71) tion was no longer significant after multivariable adjust- (Figure 3A and 3B). ment (OR, 1.40; 95% CI, 0.93–­2.10). Similarly, patients with incident AF, compared with patients without AF, had higher odds of 90-­day major bleeding (OR, 2.02; 95% CI, Thrombotic Outcomes 1.24–­3.29), which was not significant after multivariable Considering the competing risk of non-­PE death, pa- adjustment (OR, 1.48; 95% CI, 0.89–­2.45) (Figure 2C). tients with preexisting AF had a higher risk of PE death at 1-­year follow-­ up (HR, 2.27; 95% CI, 1.33–­ 3.86). However, the results were no longer significant after Outcomes at 1-­Year Follow-­Up multivariable adjustment (HR, 1.50; 95% CI, 0.88–­2.60). All-­Cause Mortality Compared with those without AF, patients with inci- Patients with preexisting AF were at higher risk for 1-­ dent AF had an increased hazard for 1-­year PE-­related year all-­cause mortality compared with those without mortality in analyses accounting for the competing J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214676 Bikdeli et al Preexisting vs Incident AF in Patients With PE Figure 2. Ninety-­day clinical outcomes based on presence of preexisting or incident AF. Patients with no AF were the reference group in all analyses. *Adjusted for age, sex, simplified Pulmonary Embolism Severity Index, and history of diabetes mellitus, heart failure, prior ischemic stroke, and arterial vascular disease (either coronary or peripheral artery disease). AF indicates atrial fibrillation; and PE, pulmonary embolism. risk of non-­PE death (HR, 3.24; 95% CI, 1.80–­5.85). 1.63–­3.26) as well as in multivariable adjusted analyses Findings persisted after multivariable adjustment (HR, (HR, 1.86; 95% CI, 1.29–­2.70, Figure 4E). 2.18; 95% CI, 1.18–­4.04) (Figure 4B). Among patients with PE and incident AF, the com- Accounting for the competing risk of nonthrombotic posite thrombotic outcome was rare at 1-­year follow death, patients with preexisting AF had a nonsignificant up, with only 6 thrombotic events, including 5 recur- increase in the risk of recurrent VTE (HR, 1.54; 95% CI, rent nonfatal VTEs but no ischemic strokes. As such, 0.96–­2.45) and a 5-­fold increased risk of stroke (HR, in models accounting for nonthrombotic death as the 5.48; 95% CI, 3.10–­9.69) (Figure 4C). Patients with pre- competing risk, patients with incident AF—­compared existing AF had a higher risk of the composite throm- with patients without AF—­did not have an increased botic outcome compared with those who did not have risk for recurrent VTE (HR, 0.68; 95% CI, 0.27–­1.65) AF, both in unadjusted analyses (HR, 2.31; 95% CI, or the composite thrombotic outcome (HR, 0.61; 95% J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214677 Bikdeli et al Preexisting vs Incident AF in Patients With PE Figure 3. Kaplan Meier curve for all-­cause death (A) and competing risk regression (Fine and Grey) for PE-­related death (B). AF indicates atrial fibrillation; and PE, pulmonary embolism. CI, 0.27–­1.38) (Figure 4E). Given the paucity of events, multivariable analysis was not performed. DISCUSSION Among patients with acute symptomatic PE, both pre- existing and newly incident AF were associated with Bleeding greater frequency of cardiovascular risk factors, and Patients with preexisting AF had increased 1-­year risk higher CHA2DS2VASC and sPESI scores, compared of major bleeding in unadjusted analyses compared with patients without AF. Both forms of AF were associ- with those without AF (Fine-­Gray HR, 1.60; 95% CI, ated with adverse outcomes, including all-­cause mor- 1.13–­ 2.27), but the findings were no longer signifi- tality, although the pattern of risk had distinct features cant after multivariable adjustment (HR, 1.16; 95% CI, based on the timing of AF onset. Preexisting AF was 0.79–­1.67). associated with increased risk of subsequent throm- Compared with patients without AF, patients with botic events, including stroke, but not a significantly in- incident AF had increased unadjusted 1-­year risk of creased risk for PE-­related mortality, once adjusted for major bleeding accounting for the competing risk of sPESI and comorbid conditions. In contrast, incident AF nonhemorrhagic death (HR, 1.65; 95% CI, 1.05–­2.58). was associated with increased risk of PE-­related mor- The results were no longer significant after multivari- tality but not with long-­term risk of stroke or the com- able adjustment. posite thrombosis end point. Although some findings were partially attributable to underlying risk factors, oth- ers persisted even in multivariable adjusted analyses. Sensitivity Analysis Findings from this study add novel information to Owing to clinicians’ discretionary decisions for hospital complement prior studies about concomitant AF in discharge, in-­hospital outcomes were not prespecified patients with PE. The proportion of patients with pre- for the main analyses. Sensitivity analyses showed that existing AF and incident AF reported in this study are in-­hospital outcomes were broadly similar to 30-­day in range with prior reports.15,19 Some prior studies,26,27 results (data not shown). including a meta-­ analysis,28 indicated an increased Figure 4. One-­year outcomes based on presence of preexisting or incident AF. Patients with no AF were the reference group in all analyses. (A) all-­cause mortality ((B)) PE-­related mortality, (C) ischemic stroke, (D) recurrent VTE (E) composite thrombotic outcome. In all models other than all-­cause mortality, competing risk of death was considered; see text for details. In models where the number of events were very few, multivariable adjustment was not pursued. *Adjusted for age, sex, simplified Pulmonary Embolism Severity Index, and history of diabetes mellitus, heart failure, prior ischemic stroke, and arterial vascular disease (either coronary or peripheral artery disease). †No patient with incident AF developed ischemic stroke during follow-­ up. AF indicates atrial fibrillation; PE, pulmonary embolism; and VTE, venous thromboembolism. J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214678 Bikdeli et al Preexisting vs Incident AF in Patients With PE risk for shock and mortality in patients with PE who and statistically significant association between AF had coexisting AF, whereas others did not show a sig- and clinical outcomes. Further, we were able to con- nificant association.19,29 Our investigation, based on a duct multivariable analyses, without concern for small large patient population showed a clinically meaningful number of events. J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.0214679 Bikdeli et al Preexisting vs Incident AF in Patients With PE We noted an increased risk of mortality and throm- This study has several limitations. First, RIETE does botic events in patients with preexisting AF. In this not include all patients with PE in the enrolling coun- sense, preexisting AF may be a thrombosis risk factor tries. However, it does include consecutive patients or an indicator of poor cardiopulmonary reserve, which in large and small regional and referral centers from increases the risk of death. The sPESI was higher in pa- 25 countries. Data from RIETE closely correlate with tients with preexisting AF compared with patients with- that of large administrative databases.32 Second, this out AF. However, the association between preexisting study was not designed to test to test the compara- AF and PE-­related mortality was no longer significant tive effectiveness of a more prolonged or more inten- after adjustment for sPESI and other comorbidities. sive antithrombotic regimen in patients with PE and In turn, newly incident AF may have been triggered coexisting AF. More prolonged use of anticoagulation by catecholamines or serotonin release in the setting in each subgroup may indicate regional practices, re- of acute PE or comorbidities such as thyrotoxicosis.15,19 sidual confounding, or mediation effect and is outside Worse outcomes in patients with PE and incident AF the scope of the current study. Rather, our goal was may be attributable to comorbidities, or the acute effect to determine the impact of preexisting versus incident of AF on right ventricular preload,15 as suggested by AF as risk markers for patients with acute PE. Of note, higher relative frequency of syncope and hypotension we identified an association with outcomes, even in in these patients. The association between incident short-­term adjusted models, in which most patients AF and PE-­related mortality attenuated but persisted, received anticoagulation. Third, newly identified AF in even after adjustment for sPESI and other clinical fac- a minority of patients may have been, indeed, preex- tors. Among survivors of acute PE, we did not identify isting but clinically silent and not otherwise perceived an increased risk of stroke in patients with PE and in- by the patient or captured by clinicians before the cident AF. We hypothesize that in many such patients index PE event. True ascertainment of such cases AF may have resolved after stabilization of acute PE. of silent AF would be possible only with long-­term In others, thrombotic risk may have been mitigated by population-­based rhythm monitoring, which is cum- anticoagulant therapy.15 bersome and outside the scope of this study. Lack Excess bleeding in patients with preexisting AF, or of any ischemic stroke events during short-­term and those with incident AF, compared with patients with- 1-­year follow-­up support the hypotheses that the vast out AF, is a clinical challenge. This excess risk is likely majority of newly identified AF events were, incident attributable to comorbidities that increase the risk of AF rather than preexisting silent AF (which has a no- not only thrombosis but also bleeding.30 More intensive ticeable risk for stroke).33 Fourth, although we present antithrombotic therapy (eg, combination of anticoagu- the results in unadjusted and multivariable adjusted lants and antiplatelet agents) may aggravate this risk. analyses that account for several of the comorbidities For example, in post hoc assessment, we noted that a and markers of disease severity, we cannot exclude larger proportion of patients with preexisting AF were the possibility of residual confounding. Fifth, although receiving aspirin at baseline, compared with patients RIETE is the largest ongoing VTE registry in the world, with AF (35.7% versus 18.9%). Therefore, the search we cannot exclude the possibility of type II error be- continues for antithrombotic regimens that may safely cause no a priori power calculation was performed. mitigate the residual thrombotic risk.31 Finally, we are unable to verify the termination or per- Findings of this study may have implications for risk sistence of AF among patients with PE and incident stratification and management. To our knowledge, this AF during long-­term follow-­up. This issue deservers is the largest investigation to address the presentation further investigation. and outcomes according to the timing of occurrence of AF. With respect to patients with preexisting AF, subsequent investigations are needed to understand CONCLUSIONS the long-­term treatment adherence and to find effec- In conclusion, AF is a relatively common and clinically tive strategies to mitigate the excess thrombotic risk. relevant coexisting condition in patients with acute In addition, dedicated studies are needed to assess symptomatic PE. Both preexisting AF and incident AF whether incorporation of AF (particularly incident AF) predict an adverse clinical course. The type of adverse can improve clinical risk stratification for patients with outcomes may differ, depending on the timing of AF acute PE, above and beyond sPESI and markers of onset. The clinical challenge will be to identify optimal right ventricular function. In addition, it should be de- strategies to mitigate the long-­term adverse outcomes termined whether acute management strategies, in- in patients with PE and preexisting AF. In addition, cluding level of care and use of advanced therapies, future research may lead to more sophisticated risk need to be modified in patients with acute PE and in- stratification algorithms that incorporate the presence cident AF. of incident AF. J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.02146710 Bikdeli et al Preexisting vs Incident AF in Patients With PE APPENDIX Consolación Rodríguez-­Matute, Jacobo Rogado, Sonia Roig, Vladimir Rosa, Pedro Ruiz-­Artacho, Nuria Ruiz-­ Coordinator of the RIETE Registry: Manuel Monreal. Giménez, Justo Ruiz-­Ruiz, Pablo Ruiz-­Sada, Giorgina RIETE Steering Committee Members: Paolo Salgueiro, Rosario Sánchez-­Martínez, Juan Francisco Prandoni, Benjamin Brenner and Dominique Sánchez-­Muñoz-­Torrero, Teresa Sancho, Merçe Sirisi, Farge-­Bancel. Silvia Soler, José María Suriñach, Raimundo Tirado, RIETE National Coordinators: Raquel Barba María Isabel Torres, Carles Tolosa, Javier Trujillo-­ (Spain), Pierpaolo Di Micco (Italy), Laurent Bertoletti Santos, Fernando Uresandi, Beatriz Valero, Reina Valle, (France), Sebastian Schellong (Germany), Inna Tzoran Jerónimo Ramón Vela, Gemma Vidal, Paula Villares, (Israel), Abilio Reis (Portugal), Marijan Bosevski Carles Zamora, ARGENTINA: Paula Gutiérrez, Fernando (R. Macedonia), Henri Bounameaux (Switzerland), Javier Vázquez, BELGIUM: Matthias Engelen, Thomas Radovan Malý (Czech Republic), Peter Verhamme Vanassche, Peter Verhamme, CZECH REPUBLIC: (Belgium), Joseph A. Caprini (USA), Hanh My Bui Jana Hirmerova, Radovan Malý, ECUADOR: Estuardo (Vietnam). Salgado, FRANCE: Nassim Ait Abdallah, Laurent SPAIN: María Dolores Adarraga, María Agud, Bertoletti, Alessandra Bura-­ Riviere, Benjamin Crichi, Jesús Aibar, Miguel Ángel Aibar, Cristina Amado, Juan Philippe Debourdeau, Espitia Olivier, Nicolas Falvo, Ignacio Arcelus, Carlos Baeza, Aitor Ballaz, Raquel Dominique Farge-­ Bancel, Silvia Galliazzo, Hélène Barba, Cristina Barbagelata, Manuel Barrón, Belén Helfer, Isabelle Mahé, Farès Moustafa, Geraldine Barrón-­ Andrés, Ángeles Blanco-­ Molina, Ernesto Poenou, GERMANY: Sebastian Schellong, ISRAEL: Botella, Ana María Camon, Inmaculada Cañas, Andrei Braester, Benjamin Brenner, Inna Tzoran, Ignacio Casado, Joaquín Castro, Leyre Chasco, Juan ITALY: Franca Bilora, Eugenio Bucherini, Maurizio Criado, Cristina de Ancos, Javier de Miguel, Jorge del Ciammaichella, Pierpaolo Di Micco, Egidio Imbalzano, Toro, Pablo Demelo-­ Rodríguez, José Antonio Díaz-­ Rosa Maida, Daniela Mastroiacovo, Federica Pace, Peromingo, Mario Virgilio Di Campli, Javier Díez-­Sierra, Raffaele Pesavento, Fulvio Pomero, Paolo Prandoni, Irene Milagros Domínguez, José Carlos Escribano, Roberto Quintavalla, Anna Rocci, Roberta Romualdi, Conxita Falgá, Ana Isabel Farfán-­ Sedano, Carmen Carmine Siniscalchi, Antonella Tufano, Adriana Fernández-­ Capitán, Jose Luis Fernández-­ Reyes, Visonà, Beniamino Zalunardo, LATVIA: Valdis Gibietis, María Ángeles Fidalgo, Katia Flores, Carme Font, Dana Kigitovica, Andris Skride, PORTUGAL: Samuel Llorenç Font, Iria Francisco, Cristina Gabara, Francisco Fonseca, Filipa Martins, Jose Meireles, REPUBLIC Galeano-­Valle, David Galindo, María Angelina García, OF MACEDONIA: Marijan Bosevski, SWITZERLAND: Ferrán García-­ Bragado, Marta García de Herreros, Henri Bounameaux, Lucia Mazzolai, USA: Behnood Raquel García-­ Hernáez, María Mar García-­ Mullor, Bikdeli, Joseph A. Caprini, Alfonso J. Tafur, Cassius Arantxa García-­ Raso, Olga Gavín-­ Sebastián, Aida Iyad Ochoa-­ Chaar, Ido Weinberg, Hannah Wilkins, Gil-­ Díaz, Covadonga Gómez-­ Cuervo, Enric Grau, VIETNAM: Hanh My Bui. Leticia Guirado, Javier Gutiérrez, Luis Hernández-­ Blasco, Elena Hernando, Miguel Herreros, Luis Jara-­ Palomares, María Jesús Jaras, David Jiménez, Rafael ARTICLE INFORMATION Jiménez, Maria Dolores Joya, José Manuel León, Received February 28, 2021; accepted June 24, 2021. Jorge Lima, Pilar Llamas, José Luis Lobo, Luciano Affiliations López-­ Jiménez, Patricia López-­ Miguel, Juan José Cardiovascular Medicine Division, Brigham and Women’s Hospital, Harvard López-­ Núñez, Raquel López-­ Reyes, Antonio López-­ Medical School, Boston, MA (B.B., G.P., S.Z.G.); Yale/YNHH Center for Ruiz, Juan Bosco López-­Sáez, Alicia Lorenzo, Olga Outcomes Research & Evaluation, New Haven, CT (B.B.); Cardiovascular Research Foundation (CRF), New York, NY (B.B.); Respiratory Department, Madridano, Ana Maestre, Pablo Javier Marchena, Hospital Ramón y Cajal (IRYCIS), Madrid, Spain (D.J.); Medicine Department, Mar Martín del Pozo, Javier Miguel Martín-­ Guerra, Universidad de Alcalá (IRYCIS), Madrid, Spain (D.J.); CIBER de Enfermedades Francisco Martín-­ Martos, Carmen Mella, Meritxell Respiratorias (CIBERES), Madrid, Spain (D.J.); Department of Internal Medicine, Hospital General Universitario Gregorio Marañón, Madrid, Spain (J.d.T.); Mellado, Maria Isabel Mercado, Jorge Moisés, Manuel Department of Pneumonology, Hospital Universitario Araba, Álava, Spain Monreal, María del Valle Morales, Arturo Muñoz-­ (A.R.); Department of Internal Medicine, Complejo Hospitalario de Jaén, Jaén, Blanco, Nuria Muñoz-­Rivas, María Sierra Navas, Jose Spain (J.L.F.); Department of Internal Medicine, Hospital Reina Sofía, Tudela, Navarra, Spain (Á.S.); Department of Pneumonology, Hospital Universitario Antonio Nieto, Manuel Jesús Núñez-­Fernández, Berta Virgen del Rocío, Seville, Spain (R.O.); Department of Internal Medicine, Obispo, Mónica Olid, María Carmen Olivares, José Hospital Universitario Vall d’Hebron, Barcelona, Spain (J.M.S.); Department of Luis Orcastegui, Clara Ortega-­Michel, Jeisson Osorio, Angiology, Azienda Ospedaliera Universitaria, Parma, Italy (C.S.); Department of Internal Medicine, Hospital Clínico Universitario de Valladolid, Valladolid, Sonia Otalora, Remedios Otero, Pedro Parra, Virginia Spain (J.M.M.); Department of Internal Medicine, Hospital Santa Bárbara, Parra, José María Pedrajas, Galadriel Pellejero, José Puertollano, Ciudad Real, Spain (J.C.); Clinical Biostatistics Unit, Hospital Antonio Porras, José Portillo, Antoni Riera-­ Mestre, Universitario Ramón y Cajal, IRYCIS, CIBERESP: Universidad de Alcalá, Madrid, Spain (A.M.); Liverpool Centre for Cardiovascular Science, University Agustina Rivas, Francisco Rivera-­ Civico, Diego of Liverpool and Liverpool Heart & Chest Hospital, Liverpool, United Kingdom Agustín Rodríguez-­ Chiaradía, Ana Rodríguez-­ Cobo, (G.Y.L.); Thrombosis Research Unit, Aalborg University, Aalborg, Denmark J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.02146711 Bikdeli et al Preexisting vs Incident AF in Patients With PE (G.Y.L.); Hospital Universitari Germans Trias i Pujol, Badalona, Barcelona, 6. Aujesky D, Obrosky DS, Stone RA, Auble TE, Perrier A, Cornuz J, Roy Spain (M.M.); and Universidad Autónoma de Barcelona, Spain (M.M.). PM, Fine MJ. Derivation and validation of a prognostic model for pulmo- nary embolism. Am J Respir Crit Care Med. 2005;172:1041–­1046. DOI: Acknowledgments 10.1164/rccm.20050​6-­862OC. We express our gratitude to Sanofi Spain for supporting this registry with 7. Jimenez D, Aujesky D, Moores L, Gomez V, Lobo JL, Uresandi F, Otero an unrestricted educational grant. The funders had no role in the design, R, Monreal M, Muriel A, Yusen RD; Investigators R. Simplification of analysis or interpretation of the data or the decision to submit. We also thank the pulmonary embolism severity index for prognostication in pa- the RIETE Registry Coordinating Center, S&H Medical Science Service, for tients with acute symptomatic pulmonary embolism. Arch Intern Med. their quality control of the data and logistic and administrative support and 2010;170:1383–­1389. DOI: 10.1001/archi​ntern​med.2010.199. Prof. Salvador Ortiz, Universidad Autónoma Madrid, and Silvia Galindo, both 8. Vitolo M, Lip GYH. Understanding the global burden of atrial fibrilla- statistical advisors in S&H Medical Science Service, for their help with the tion and regional variations: we need improvement. Cardiovasc Res. statistical analyses presented in this paper. 2021;117:1420–­1422. DOI: 10.1093/cvr/cvaa330. Author contributions: Study design: Bikdeli, Jiménez, Piazza, Lip, 9. Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, Goldhaber, Monreal; Data acquisition: Jiménez, Monreal; Statistical analy- Castella M, Diener H-­C, Heidbuchel H, Hendriks J, et al. 2016 ESC ses: Muriel, Jiménez, Bikdeli, Monreal; Critical revision of the article: all au- guidelines for the management of atrial fibrillation developed in collab- thors. Drs Bikdeli, Jimenez, and Monreal take responsibility for the integrity oration with EACTS. Eur Heart J. 2016;37:2893–­2962. DOI: 10.1093/ of the work as a whole, from inception to published article. eurhe​artj/ehw210. 10. Benjamin EJ, Muntner P, Alonso A, Bittencourt MS, Callaway CW, Sources of Funding Carson AP, Chamberlain AM, Chang AR, Cheng S, Das SR, et al. Heart None. disease and stroke statistics-­2019 update: a report from the American Heart Association. Circulation. 2019;139:e56–­ e528. DOI: 10.1161/ Disclosures CIR.00000​0 0000​0 00659. Dr Bikdeli reports that he is a consulting expert, on behalf of the plaintiff, 11. Dewilde WJM, Oirbans T, Verheugt FWA, Kelder JC, De Smet BJGL, for litigation related to 2 specific brand models of inferior vena cava filters. Herrman J-­P, Adriaenssens T, Vrolix M, Heestermans AACM, Vis MM, et Dr Jimenez has served as an advisor or consultant for Bayer HealthCare al. Use of clopidogrel with or without aspirin in patients taking oral anti- Pharmaceuticals, Boehringer Ingelheim, Bristol-­ Myers Squibb, Daiichi coagulant therapy and undergoing percutaneous coronary intervention: Sankyo, Leo Pharma, Pfizer, ROVI, and Sanofi; served as a speaker or a mem- an open-­ label, randomised, controlled trial. Lancet. 2013;381:1107–­ ber of a speakers’ bureau for Bayer HealthCare Pharmaceuticals, Boehringer 1115. DOI: 10.1016/S0140​- ­6736(12)62177​-­1. Ingelheim, Bristol-­Myers Squibb, Daiichi Sankyo, Leo Pharma, ROVI, and 12. Gibson CM, Mehran R, Bode C, Halperin J, Verheugt FW, Wildgoose Sanofi; and received grants for clinical research from Daiichi Sankyo, Sanofi, P, Birmingham M, Ianus J, Burton P, van Eickels M, et al. Prevention of and ROVI. Dr Piazza has received significant research grant support from bleeding in patients with atrial fibrillation undergoing PCI. N Engl J Med. BTG International, Bristol Myers Squibb, Daiichi-­Sankyo, Bayer, Portola, and 2016;375:2423–­2434. DOI: 10.1056/NEJMo​a1611594. Janssen and modest consulting fees from Pfizer and Thrombolex. Dr Lip 13. Cannon CP, Bhatt DL, Oldgren J, Lip GYH, Ellis SG, Kimura T, Maeng M, reports that he is a consultant for Bayer/Janssen, BMS/Pfizer, Medtronic, Merkely B, Zeymer U, Gropper S, et al. Dual antithrombotic therapy with Boehringer Ingelheim, Novartis, Verseon, and Daiichi-­Sankyo and a speaker dabigatran after PCI in atrial fibrillation. N Engl J Med. 2017;377:1513–­ for Bayer, BMS/Pfizer, Medtronic, Boehringer Ingelheim, and Daiichi-­Sankyo. 1524. DOI: 10.1056/NEJMo​a1708454. No fees are directly received personally. Dr Goldhaber reports grants from 14. Flegel KM. When atrial fibrillation occurs with pulmonary embolism, is it Bayer, Boehringer-­Ingelheim, BMS, BTG EKOS, Daiichi Sankyo, Janssen, US the chicken or the egg? CMAJ. 1999;160:1181–­1182. National Heart Lung and Blood Institute, and Thrombosis Research Institute; 15. Bikdeli B, Abou Ziki MD, Lip GY. Pulmonary embolism and atrial fibrilla- and personal fees from Bayer and Boehringer-­Ingelheim, outside the submit- tion: two sides of the same coin? A systematic review. Semin Thromb ted work. Dr Monreal has served as an advisor or consultant for Sanofi, Leo Hemost. 2017;43:849–­863. DOI: 10.1055/s-­0 036-­1598005. Pharma, and Daiichi-­Sankyo and has received a nonrestricted educational 16. Wang CC, Lin CL, Wang GJ, Chang CT, Sung FC, Kao CH. Atrial fibril- grant by Sanofi and Bayer to sponsor the RIETE registry. lation associated with increased risk of venous thromboembolism. A population-­based cohort study. Thromb Haemost. 2015;113:185–­192. Supplementary Material DOI: 10.1160/TH14-­05-­0405. Table S1 17. Enga KF, Rye-­Holmboe I, Hald EM, Lochen ML, Mathiesen EB, Njolstad I, Wilsgaard T, Braekkan SK, Hansen JB. Atrial fibrillation and future risk of venous thromboembolism:the Tromso study. J Thromb Haemost. 2015;13:10–­16. DOI: 10.1111/jth.12762. 18. Gex G, Gerstel E, Righini M, Le gal G, Aujesky D, Roy P-­M, Sanchez O, REFERENCES Verschuren F, Rutschmann OT, Perneger T, et al. Is atrial fibrillation as- 1. Barco S, Mahmoudpour SH, Valerio L, Klok FA, Münzel T, Middeldorp sociated with pulmonary embolism? J Thromb Haemost. 2012;10:347–­ S, Ageno W, Cohen AT, Hunt BJ, Konstantinides SV. Trends in mortality 351. DOI: 10.1111/j.1538-­7836.2011.04608.x. related to pulmonary embolism in the European Region, 2000–­15: anal- 19. Ebner M, Rogge NIJ, Parwani AS, Sentler C, Lerchbaumer M, Pieske B, ysis of vital registration data from the WHO Mortality Database. Lancet Konstantinides SV, Hasenfuss G, Wachter R, Lankeit M. Atrial fibrillation Respir Med. 2020;8:277–­287. DOI: 10.1016/S2213​-­2600(19)30354​- ­6. is frequent but does not affect risk stratification in pulmonary embolism. 2. Bikdeli B, Wang Y, Jimenez D, Parikh SA, Monreal M, Goldhaber SZ, J Intern Med. 2020;287:100–­113. DOI: 10.1111/joim.12985. Krumholz HM. Pulmonary embolism hospitalization, readmission, and 20. Bikdeli B, Jiménez D. Atrial fibrillation in the course of pulmonary embo- mortality rates in US older adults, 1999–­2015. JAMA. 2019;322:574–­ lism: just a little smoke, or fuel to the fire? J Intern Med. 2020;287:114–­ 576. DOI: 10.1001/jama.2019.8594. 116. DOI: 10.1111/joim.12999. 3. Kearon C, Akl EA, Ornelas J, Blaivas A, Jimenez D, Bounameaux H, 21. Bikdeli B, Jimenez D, Hawkins M, Ortíz S, Prandoni P, Brenner B, Huisman M, King CS, Morris TA, Sood N, et al. Antithrombotic ther- Decousus H, Masoudi F, Trujillo-­Santos J, Krumholz H, et al. Rationale, apy for VTE disease: CHEST guideline and expert panel report. Chest. design and methodology of the computerized registry of patients with 2016;149:315–­352. DOI: 10.1016/j.chest.2015.11.026. venous thromboembolism (RIETE). Thromb Haemost. 2018;118:214–­ 4. Konstantinides SV, Meyer G, Becattini C, Bueno H, Geersing G-­J, Harjola 224. DOI: 10.1160/TH17-­07-­0511. V-­P, Huisman MV, Humbert M, Jennings CS, Jiménez D, et al. 2019 ESC 22. Bikdeli B, Madhavan MV, Gupta A, Jimenez D, Burton JR, Der guidelines for the diagnosis and management of acute pulmonary em- Nigoghossian C, Chuich T, Nouri SN, Dreyfus I, Driggin E, et al. bolism developed in collaboration with the European Respiratory Society Pharmacological agents targeting thromboinflammation in COVID-­19: (ERS). Eur Heart J. 2020;41:543–­603. DOI: 10.1093/eurhe​artj/ehz405. review and implications for future research. Thromb Haemost. 2020;​ 5. Zondag W, Mos ICM, Creemers-­schild D, Hoogerbrugge ADM, Dekkers 120:1004–­1024. DOI: 10.1055/s-­0 040-­1713152. OM, Dolsma J, Eijsvogel M, Faber LM, Hofstee HMA, Hovens MMC, 23. Bikdeli B, Madhavan MV, Jimenez D, Chuich T, Dreyfus I, Driggin E, et al. Outpatient treatment in patients with acute pulmonary embo- Nigoghossian CD, Ageno W, Madjid M, Guo Y, et al. COVID-­19 and lism: the Hestia Study. J Thromb Haemost. 2011;9:1500–­1507. DOI: thrombotic or thromboembolic disease: implications for prevention, 10.1111/j.1538-­7836.2011.04388.x. antithrombotic therapy, and follow-­up: JACC state-­of-­the-­art review. J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.02146712 Bikdeli et al Preexisting vs Incident AF in Patients With PE J Am Coll Cardiol. 2020;75:2950–­ 2973. DOI: 10.1016/j.jacc.2020. embolism: systematic review and meta-­ analysis. Acad Emerg Med. 04.031. 2015;22:1127–­1137. DOI: 10.1111/acem.12769. 24. Schulman S, Angerås U, Bergqvist D, Eriksson B, Lassen MR, Fisher 29. Koracevic G, Atanaskovic V. Is atrial fibrillation a prognosticator in acute W. Definition of major bleeding in clinical investigations of antihemo- pulmonary thromboembolism? Med Princ Pract. 2010;19:166. DOI: static medicinal products in surgical patients. J Thromb Haemost. 10.1159/00027​3082. 2010;8:202–­204. DOI: 10.1111/j.1538-­7836.2009.03678.x. 30. Poli D, Antonucci E, Pengo V, Testa S, Palareti G. Comparison 25. Mehdipoor G, Jimenez D, Bertoletti L, Fidalgo Á, Sanchez Muñoz-­ of HAS-­ BLED and HAS-­ BED Versus CHADS(2) and CHA(2)DS(2) Torrero JF, Gonzalez-­Martinez JP, Blanco-­Molina Á, Ángel Aibar M, VASC stroke and bleeding scores in patients with atrial fibrillation. Bonnefoy PB, Khorasani R, et al. Patient-­level, institutional, and tem- Am J Cardiol. 2017;119:1012–­ 1016. DOI: 10.1016/j.amjca​ rd.2016. poral variations in use of imaging modalities to confirm pulmonary em- 12.007. bolism. Circ Cardiovasc Imaging. 2020;13:e010651. DOI: 10.1161/CIRCI​ 31. Bikdeli B, Chatterjee S, Kirtane AJ, Parikh SA, Andreozzi GM, Desai NR, MAGING.120.010651. Francese DP, Gibson CM, Piazza G, Goldhaber SZ, et al. Sulodexide 26. Escobar C, Jimenez D, Marti D, Lobo JL, Diaz G, Gallego P, Vidal R, versus control and the risk of thrombotic and hemorrhagic events: meta-­ Barrios V, Sueiro A. [Prognostic value of electrocardiographic findings analysis of randomized trials. Semin Thromb Hemost. 2020;46:908–­ in hemodynamically stable patients with acute symptomatic pulmonary 918. DOI: 10.1055/s-­0 040-­1716874. embolism]. Rev Esp Cardiol. 2008;61:244–­250. 32. Guijarro R, Montes J, Sanromán C, Monreal M. Venous thromboem- 27. Ng AC, Adikari D, Yuan D, Lau JK, Yong AS, Chow V, Kritharides L. The bolism in Spain. Comparison between an administrative database and prevalence and incidence of atrial fibrillation in patients with acute pul- the RIETE registry. Eur J Intern Med. 2008;19:443–­446. DOI: 10.1016/j. monary embolism. PLoS One. 2016;11:e0150448. DOI: 10.1371/journ​ ejim.2007.06.026. al.pone.0150448. 33. Glotzer TV, Ziegler PD. Silent atrial fibrillation as a stroke risk factor 28. Shopp JD, Stewart LK, Emmett TW, Kline JA. Findings from 12-­lead and anticoagulation indication. Can J Cardiol. 2013;29:S14–­S23. DOI: electrocardiography that predict circulatory shock from pulmonary 10.1016/j.cjca.2013.03.023. J Am Heart Assoc. 2021;10:e021467. DOI: 10.1161/JAHA.121.02146713 SUPPLEMENTAL MATERIAL Table S1. ESC Classification Risk Among the Study Participants* Known Pre-existing AF Incident AF (diagnosed within 2 No AF at all days after PE) Number of patients (total N=16,497) N=792 N=445 N=15,260 Low Risk 102 (12.9%) 64 (14.4%) 6,030 (39.5%) Intermediate-low risk PE 639 (80.7%) 322 (72.4 %) 8093 (53.0 %) Intermediate-high risk PE 22 (2.7 %) 25 (5.6 %) 681 (4.4 %) High Risk 2 (3.7%) 34 (7.4%) 458 (3.0%) * Cardiac biomarkers (or echocardiographic features of right ventricular dysfunction) are available in many RIETE participants, but are not mandatory fields in RIETE, with missing values in some participants. AF: Atrial fibrillation, ESC: European Society of cardiology, PE: pulmonary embolism.