A retrospective observational study investigating the factors associated with right heart failure in patients with primary acute pulmonary embolism and deep vein thrombosis

Abstract Background The relationship between the risk of right heart failure in primary acute pulmonary embolism after embolization and the residual thrombus sites in the pelvis and lower limbs is not clear. Methods This single‐center retrospective observational study examined the results of contrast‐enhanced computed tomography and venous ultrasonography of patients with primary acute PE and DVT. We assessed the association between the occurrence of right heart failure and age; gender; pulmonary thrombosis distribution; most proximal site of deep vein thrombosis in the soleal vein, inferior vena cava (IVC), or common iliac vein (CIV); DVT distribution; and malignancy using univariate and multivariate logistic regression. Results In all, 77 of 165 patients were male (mean age: 65.1 ± 13.7 years). Right heart failure occurred in 53 patients (32.1%). Multivariate analysis revealed that the odds ratio (OR) for right heart failure was significantly lower in patients with the most proximal site of DVT in the IVC/CIV (OR = 0.07, 95% confidence interval [CI] 0.01‐0.62, P = .017), while it was significantly higher in females (OR = 2.51, 95% CI 1.05‐6.01, P = .039), and in patients who exhibited the presence of bilateral venous thrombosis (OR = 3.89, 95% CI 1.60‐9.48, P = .003). Conclusion A significant factor involved in PE without right heart failure was the most proximal site of DVT in the IVC/CIV, and significant risk factors associated with PE with right heart failure were more prevalent in females and in patients who exhibited the presence of bilateral venous thrombosis.


| INTRODUC TI ON
Acute pulmonary embolism (PE) can be caused by the sudden blockage of a pulmonary vessel by a clot that has detached from a vein.
Severe PE may result in acute right heart failure, causing the patient to go into shock. In the majority of PE cases, the embolic source is the deep veins of the lower extremities. 1,2 PE usually forms when part of a thrombus detaches and makes its way through the vein toward the heart. 3 After an embolus has broken off, the venous location of the thrombus that remains in the leg that is closest to the heart is known as the most proximal site of DVT. 4 Acute PE is diagnosed and treated based on the severity grading system described in the guidelines for the diagnosis, treatment, and prevention of PE and DVT which comprises hemodynamic findings and signs of right heart overload on echocardiography. 5 According to this classification system, PE is graded on a four-point scale as (1) nonmassive, (2) submassive, (3) massive, or (4) cardiac arrest or circulatory collapse.
Right heart failure is included in the diagnostic criteria for submassive and massive PE, as well as PE leading to cardiac arrest or circulatory collapse since the prognosis and recurrence rate of these conditions depend on of the presence of right heart overload. 6,7 Risk factors for acute submassive PE with right heart overload have been reported in previous studies. 8 In addition, there are several reports retrospectively examining the presence of right heart overload in patients with acute PE after embolization and most proximal sites of residual DVT in the lower limbs. 3,9 Ohgi et al reported that the commonest most proximal site leading to severe acute PE is the femoral vein. 3 In contrast, Ro et al's 9 investigation of the most proximal site of residual thrombus in the lower extremities of 100 patients who died of acute PE found that 63% of sites were near the popliteal vein.
However, these studies did not specifically analyze using rigorous statistical analysis. In addition, the inferior vena cava (IVC) was not included in the investigation site of the deep vein in these studies. Therefore, the relationship between the risk of right heart failure in primary acute PE after embolization and the residual thrombus site in the lower limb is not clear. The objective of this study was to examine the factors associated with right heart failure in patients with primary acute PE and DVT after embolization, especially the location of residual thrombus in the lower limbs. Therefore, retrospective examination was performed based on the findings of contrast-enhanced computed tomography (CT) and venous ultrasonography of the lower limbs. In this study, a significant factor involved in PE without right heart failure was the most proximal site of DVT in the IVC/CIV, and significant risk factors associated with PE with right heart failure were female gender and the presence of bilateral venous thrombosis. If DVT is suspected from the symptoms, venous ultrasonography of the lower limb may be performed before contrast-enhanced CT. In a situation where the presence of PE and right heart failure has not been confirmed by testing, if venous thrombus on both legs is present on venous ultrasonography of the lower limb, the possibility of right heart failure because of PE complications can be considered. It could be used as a criterion to decide whether or not to perform contrast-enhanced CT and echocardiography.

| Participants and exclusion criteria
This was a single-center retrospective observational study. The study was conducted at Saitama Cardiovascular Respiratory Center, Saitama, Japan, a 319-bed tertiary emergency medical institution.
The study was approved by the Ethics Committee of our institution (approval number 2 014 010); the need for informed consent was waived because of the retrospective nature of the study.
All PE or DVT patients diagnosed at our hospital from March 2006 to May 2016 were eligible for inclusion in this study. The patients included outpatients and hospitalized patients (in our hospital or other hospitals). All patients underwent CT pulmonary angiography (CTPA) and indirect CT venography using a multidetector scanner, 10 in addition to venous ultrasonography of the lower limbs and echocardiography. All tests were conducted at roughly the same time (mean time between tests: 1.3 days).
If PE was suspected from the symptoms, contrast-enhanced CT was performed before venous ultrasonography of the lower limbs.
Immediately after PE was detected by contrast-enhanced CT, venous ultrasonography of the lower limbs confirmed the presence of DVT. On the other hand, if DVT was suspected from the symptoms, venous ultrasonography of the lower limb was performed before contrast-enhanced CT. Immediately after DVT was detected by venous ultrasonography of the lower limbs, contrast-enhanced CT confirmed the presence of PE, and echocardiography confirmed the presence or absence of right heart failure. Acute DVT refers to venous thrombosis for which symptoms have been present for two weeks or for which imaging studies indicate that venous thrombosis occurred within the last 2 weeks. 11 Nevertheless, the definition of acute PE is not described in Japanese and foreign guidelines and literature. Normally, an acute PE is due to lower limb DVT. 1,2 Based on the above definitions, in this study, the definition of acute DVT and PE refers to venous thrombosis for which symptoms have been present for two weeks or less or for which imaging studies indicate that acute DVT occurred within the last two weeks. 4,5 Conversely, there were patients who had symptoms but did not go to the hospital or were suspected of other diseases and had not been tested for more than 2 weeks. There were also excluded cases where imaging did not show that acute DVT had occurred within the last two weeks.
Patients diagnosed with primary acute PE and DVT were included in the study, while those with DVT only were excluded. In addition, we excluded patients with a history of PE/DVT, cases demonstrating more than 2 weeks after symptom onset, unknown onset dates, interstitial pneumonia, or emphysema, as these factors may affect the distribution of PE and DVT, and right ventricular (RV) pressure.

| Thrombus detection methods
Patients were scanned from the chest downwards using a Discovery 64 row multislice CT system (GE Healthcare, Boston, MA) for CTPA and CT venography. The pulmonary artery (from the apex to the inferior margin of the diaphragm) was scanned starting 25 s after the start of contrast agent injection (standard injection rate, 3 mL/s; concentration, 370 mg iodine/mL; injection volume, 100 mL), and the leg veins were scanned (from the inferior vena cava [IVC] to the ankle) 3.5 minutes later (Table 1). 12 The amount of contrast agent administered was set at a total of 600 mg of iodine per kg of body weight. The amount was reduced at the discretion of the attending physician depending on the patient's general condition, such as the presence of right heart overload. To ensure adequate visualization of the lower leg veins (particularly the soleal vein), an object, such as a pillow, was placed under the heels during scanning to avoid compression of this area. In patients with severe PE, there was a possibility that the pulmonary circulation may be delayed, and the veins may not be contrast-enhanced 3.5 minutes after the injection of the contrast agent. In such patients, additional scanning was performed if the initial enhancement was deemed to be poor.

| Echocardiography devices
Echocardiography was performed using a GE LOGIQ E9 XDclear 2.0 system (GE Healthcare, Milwaukee, WI, USA) fitted with a 1.5-4.6-MHz sector probe, a GE Vivid7 system (GE Healthcare) fitted with a 1.5-4.0-MHz sector probe, a Philips IE33 system (Philips Medical, Best, Netherlands) fitted with a 1-5-MHz sector probe, or a Philips SONOS 7500 system (Philips Medical) fitted with a 1-3-MHz sector probe.

| Sites assessed for the presence of thrombi
We assessed the pulmonary artery using CTPA; the IVC, common

| Severity grading system for acute PE
The severity of acute PE was graded in accordance with the guidelines for the diagnosis, treatment, and prevention of PE and DVT (Table 2). 5 "Cardiac arrest" was defined as cardiac arrest or circulatory ischemia in either case associated with signs of right heart failure on echocardiography; "massive" as hemodynamic instability with signs of right heart failure on echocardiography; "submassive" as hemodynamic stability with signs of right heart failure on echocardiography; and "nonmassive" as hemodynamic stability with no signs of right heart failure on echocardiography. Therefore, right

| Parameters investigated
The associations between right heart failure and age, gender (male/ female), PE distribution (one or both lungs), the most proximal site of DVT in the soleal vein (yes or no), DVT distribution (one or both legs), the most proximal site of DVT in the IVC or CIV (yes or no), and malignancy (present or absent) were investigated.

| Statistical analysis
A logistic model was used for statistical analysis, and univariate and multivariate analyses were performed. The independent variables included in the multivariate analysis were chosen based on previous study findings and clinical judgment. 3,9,15,16 A univariate analysis of these independent variables was performed before the multivariate analysis. P values <.05 were considered to indicate statistical significance. All analyses were performed using EZR software (Jichi Medical University Saitama Medical Center, Saitama, Japan).

| Patients
During the study period, 376 patients were diagnosed with PE or

| DVT in the IVC or CIV
Since we found that the IVC or CIV as the most proximal site of DVT

| D ISCUSS I ON
This was a single-center retrospective study of factors associated with right heart failure in patients with primary acute PE and DVT.

Nonmassive (n = 112) P-value
Age, mean ± SD 67. 6  We also found that a significant risk factor associated with PE with right heart failure was the presence of bilateral venous thrombi. When venous thrombi are present in both legs rather than only one, the volume of the thrombus and the area which it occupies increases. In this case, there is a significant possibility that the area of the pulmonary artery that is embolized may be increased.
If compensation is inadequate, right heart failure may occur. There were no deaths recorded in our study. Ro et al retrospectively examined 100 patients who died of PE, finding that 89% had bilateral lower extremity residual thrombi. 9 The cases mentioned in the report all resulted in death, which is different from the PE severity of our study. However, this previous study indicated that the presence of bilateral thrombi was involved in right heart failure, and it could even lead to death. We recognize that there are some limitations to this study. First, this was a single-center retrospective study with a limited sample size, which may contribute to bias. In addition, the effect of the amount of clots on the severity of PE could not be investigated since thrombi could not be quantified. This may limit the generalizability of the results. In order to confirm our findings, it is necessary to conduct larger multicenter collaborative studies. We clarified the relationship between the risk of right heart failure and the residual thrombus site in the lower extremities in primary acute PE. In the future, it is necessary to examine the relationship between the distribution of PE and DVT, identify predictors of the clinical course, and determine whether invasive examination and treatment can be avoided.

| CON CLUS ION
A significant factor involved in PE without right heart failure was the most proximal site of DVT in the IVC/CIV, and significant risk factors associated with PE with right heart failure were more prevalent in females and in patients who exhibited the presence of bilateral venous thrombosis.

ACK N OWLED G EM ENTS
The authors are grateful to the members of the Department of Radiology, Saitama Cardiovascular Respiratory Center, for supporting this work. This study was submitted to Suzuka University of Medical Science as a master's thesis. The master's thesis has not been published by Suzuka University of Medical Science. In addition, we developed the research content of the paper and have submitted it to the Journal of General and Family Medicine.

CO N FLI C T O F I NTE R E S T
The authors have stated explicitly that there are no conflict of interest in connection with this article.