- Email: [email protected]
Usefulness of Combined Assessment With Computed Tomographic Signs of Right Ventricular Dysfunction and Cardiac Troponin T for Risk Stratification of Acute Pulmonary Embolism
Usefulness of Combined Assessment With Computed Tomographic Signs of Right Ventricular Dysfunction and Cardiac Troponin T for Risk Stratification of Acute Pulmonary Embolism Doo Kyoung Kang, MDa,*, Joo Sung Sun, MDa, Kyung Joo Park, MD, PhDa, and Hong Seok Lim, MD, PhDb The aim of this study was to evaluate the incremental value of combined assessment with computed tomographic (CT) signs of right ventricular (RV) dysfunction and cardiac troponin T level for predicting early death or adverse outcomes due to acute pulmonary embolism (PE). One hundred seventy-three non-high-risk patients with acute PE, confirmed by CT pulmonary angiography, were retrospectively evaluated. The area under the curve and hazard ratio of CT signs and troponin T levels were compared for predicting early death or adverse outcomes. Patients were classified into intermediate- and low-risk groups on the basis of CT signs and troponin T levels, and mortality was compared. Seventeen patients (9.8%) died within 3 months. Early mortality of intermediate-risk patients (14% to 19%) was higher than that of low-risk patents (2% to 6%). A ratio of RV volume to left ventricular volume > 1.5 had the highest area under the curve (0.709) and hazard ratio (5.402) for predicting early death. The combination of CT signs and elevated troponin T level had an increased area under the curve and hazard ratio for predicting early death and adverse outcomes compared to those of CT signs or elevated troponin T level alone. In conclusion, the combined assessment of the ratio of RV volume to left ventricular volume and an elevated troponin T level provided incrementally more prognostic information in non-high-risk patients with acute PE compared to the single predictor of CT signs or troponin T level. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;108:133–140) In the risk stratification of acute pulmonary embolism (PE), echocardiography is the reference standard for the diagnosis of right ventricular (RV) dysfunction.1 However, computed tomographic (CT) pulmonary angiography has the benefit of directly diagnosing PE as well as simultaneously providing indirect signs of RV dysfunction.2,3 The ventricles have a complex 3-dimensional shape; thus, volumetric measurements of the right and left ventricles using CT pulmonary angiography are expected to provide a more accurate method for examining RV dysfunction.4 – 6 As an alternative approach, combined assessment of echocardiographic findings and cardiac biomarkers has been reported to offer incremental prognostic value for identifying highrisk patients.7–9 Thus, we evaluated whether volumetric measurement of RV dysfunction was superior to other CT signs of RV dysfunction as well as the incremental value of combined assessment with CT signs of RV dysfunction and cardiac troponin T level for predicting early death or adverse outcomes due to acute PE.
Departments of aRadiology and bCardiology, Ajou University School of Medicine, Suwon, South Korea. Manuscript received January 18, 2011; revised manuscript received and accepted March 3, 2011. *Corresponding author: Tel: 82-31-219-5849; fax: 82-31-219-5862. E-mail address: [email protected] (D.K. Kang). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2011.03.009
Methods We retrospectively identified 223 consecutive patients hospitalized at our hospital from March 2005 through March 2010 with acute PE confirmed by CT pulmonary angiography. The medical records department of our institution searched 1,087 patients who underwent CT pulmonary angiography in this period; 266 of these had been registered as having acute PE. In total, 223 patients were positive for PE on CT pulmonary angiography, and 43 had negative PE findings on CT pulmonary angiography. One hundred seventy-three patients were finally eligible for the present study, after the exclusion of 18 patients with incomplete follow-up data for the end point, 15 patients without laboratory data for cardiac troponin, 12 patients with highrisk PE, and 5 patients with insufficient contrast enhancement of the ventricular chambers for reliable delineation of the endocardial borders. High-risk PE was defined as the presence of shock or persistent systemic hypotension (defined as systolic blood pressure ⬍90 mm Hg or a pressure decrease of ⬎40 mm Hg for 15 minutes) and represented an immediately life-threatening emergency requiring specific management.10,11 For clinical assessment of the probability of PE, we used the Wells score, which classifies patients as “PE unlikely” or “PE likely.”12 Co-morbidities that increased the risk for adverse outcome included advanced age, cancer, chronic lung disease, and heart failure, in accordance with the International Cooperative Pulmonary Embolism Registry (ICOPER).13 The Human Research Commitwww.ajconline.org
134
The American Journal of Cardiology (www.ajconline.org)
Figure 1. A 78-year-old man with an acute PE confirmed by CT pulmonary angiography died on the second day of his hospital stay. Axial (A) and volume-rendered technique (B) shows multiple filling defects in the lobar (long arrow) and segmental (short arrows) pulmonary arteries involving the left lower lobe. (C) CT image shows grade 4 reflux of contrast medium into the inferior vena cava (long arrow) and proximal hepatic veins (short arrow). (D) Axial view shows septal bowing (arrows), convex toward the left ventricle (LV). CT images of measurements of maximal right ventricular (E) and left ventricular (F) diameters on transverse sections show an RVD/LVD ratio of 1.3. The endocardial contours of the right ventricle (RV) (G) and the LV (H) were manually traced on the transverse sections, and the outlining was then automatically propagated to the neighboring sections. (I,J) The ventricles were semiautomatically segmented from the valvular plane down to the apex of each ventricle. (K,L) The volume of each ventricle was displayed using a volume-rendered technique and was automatically estimated. The RVV/LVV ratio was subsequently calculated to be 3.2.
tee at Ajou University Hospital approved the study design and waived the need for informed patient consent for this retrospective analysis.
All patients had undergone CT imaging, without electrocardiographic gating, using a 16-slice multidetector CT system (Somatom Sensation; Siemens Medical Systems,
Miscellaneous/Acute PE Risk Stratification Using CTPA and Troponin T
135
Fig. 1. (Cont’d)
Table 1 Clinical characteristics and relevant co-morbidities in non-high risk patients
Table 2 Computed tomographic signs of right ventricular dysfunction for predicting adverse outcomes and 30-day death
Variable
Sign of RV Dysfunction
Total (n ⫽ 173)
Survival (n ⫽ 156) Uncomplicated Clinical Course (n ⫽ 139)
Early Death (n ⫽ 17) Adverse Outcome (n ⫽ 34)
p Value
Age (years)
61 ⫾ 14
Men
76 (44%)
Co-morbidities
61 (35%)
Cancer
41 (24%)
Chronic lung disease Heart failure
15 (9%)
60 ⫾ 15 60 ⫾ 14 68 (44%) 65 (47%) 51 (33%) 46 (33%) 33 (21%) 29 (21%) 14 (9%) 12 (9%) 7 (4%) 7 (5%)
64 ⫾ 12 61 ⫾ 14 8 (47%) 11 (32%) 10 (58%) 15 (44%) 8 (47%) 12 (35%) 1 (6%) 3 (9%) 1 (6%) 1 (3%)
0.296 0.706 0.987 0.185 0.061 0.236 0.031* 0.113 0.999 0.999 0.570 0.999
8 (5%)
Data are expressed as mean ⫾ SD or as numbers (percentage). * Statistically significant.
Forchheim, Germany) or a 64-slice scanner (Brilliance; Phillips Medical Systems, Eindhoven, The Netherlands). The image acquisition parameters for 16-sclice CT imaging were as follows: gantry rotation time 500 ms; collimation 16 ⫻ 0.75 mm; pitch 1.0; tube voltage 120 kV; and effective tube current of 200 mA. The image acquisition parameters for 64-slice CT imaging were as follows: gantry rotation time 500 ms, detector collimation 64 ⫻ 0.625 mm, pitch 0.924, tube voltage 120 kV, and effective tube current 200 mA. Precontrast CT imaging was performed with 3-mm section thickness to evaluate calcification or degree of enhancement, because up to 2/3 of the patients for whom there was initial suspicion of PE received other diagnoses.2 In the next step, postcontrast CT pulmonary angiography was performed with 1-mm reconstruction section thickness. Contrast media (80 ml of Omnipaque 350; GE Healthcare, Dublin, Ireland) was injected intravenously at a rate of 3.2 ml/s, followed by a saline flush (30 ml). Automated bolus triggering was used with a region of interest in the main
IVC contrast reflux Abnormal septal shape RVD/LVD ratio RVV/LVV ratio Elevated troponin T level
Survival (n ⫽ 156) Uncomplicated Clinical Course (n ⫽ 139)
Early Death (n ⫽ 17) Adverse Outcome (n ⫽ 34)
p Value
54 (35%) 44 (32%) 49 (31%) 42 (30%) 58 (37%) 50 (36%) 45 (29%) 37 (27%) 38 (24%) 30 (22%)
8 (47%) 15 (44%) 9 (53%) 16 (47%) 12 (71%) 20 (59%) 12 (71%) 20 (59%) 11 (65%) 19 (56%)
0.453 0.241 0.130 0.097 0.016* 0.025* 0.001* ⬍0.001* 0.001* ⬍0.001*
* Statistically significant. IVC ⫽ inferior vena cava.
pulmonary artery and a threshold of 120 Hounsfield units for triggering data acquisition. Data were acquired in a craniocaudal direction in the supine position with full inspiration. The scanning range extended from the level of the sternal notch to the level of the adrenal gland. All CT studies were independently reviewed by 2 experienced observers (D.K.K., J.S.S.), with 7 and 5 years of experience, respectively. Because of the retrospective nature of the study, the observers were aware that patients had been diagnosed with acute PE, but they were blinded to patients’ other clinical characteristics and outcomes. The initial clinical diagnosis of PE was confirmed in the presence of ⱖ1 filling defect in the pulmonary artery tree (Figure 1), including the subsegmental level.2 The severity of reflux of contrast medium into the inferior vena cava or hepatic veins was graded according to a previously published scale14: 1 ⫽ no reflux; 2 ⫽ trace of reflux into the inferior vena cava only; 3 ⫽ reflux into the inferior vena cava, but not the hepatic veins; 4 ⫽ reflux into the inferior vena cava and proximal hepatic veins; 5 ⫽ reflux into the inferior vena cava and hepatic veins down to the midportion
136
The American Journal of Cardiology (www.ajconline.org)
Table 3 Receiver-operating characteristic curve analysis between signs of right ventricular dysfunction for predicting early death Diagnostic Criterion
IVC contrast reflux IVC contrast reflux and troponin T Abnormal septal shape Abnormal septal shape and troponin T RVD/LVD ratio RVD/LVD ratio and troponin T RVV/LVV ratio RVV/LVV ratio and troponin T Elevated troponin T level
For Predicting Early Death
For Predicting Adverse Outcome
AUC
95% CI
p Value
AUC
95% CI
p Value
0.562 0.606 0.608 0.661 0.667 0.675 0.709 0.733 0.702
0.485–0.637 0.529–0.679 0.531–0.681 0.585–0.731 0.592–0.737 0.599–0.744 0.635–0.775 0.660–0.797 0.628–0.769
0.340 0.094 0.098 0.004* 0.006* ⬍0.001* ⬍0.001* ⬍0.001* 0.006*
0.562 0.637 0.584 0.644 0.614 0.659 0.661 0.706 0.671
0.485–0.637 0.560–0.708 0.507–0.659 0.568–0.715 0.537–0.687 0.584–0.730 0.585–0.731 0.633–0.773 0.596–0.741
0.190 0.003* 0.077 0.001* 0.016* ⬍0.001* ⬍0.001* ⬍0.001* ⬍0.001*
* Statistically significant. AUC ⫽ area under the receiver-operating characteristic curve; CI ⫽ confidence interval. Other abbreviations as in Table 2.
Figure 2. Early death rate within 3 months in 173 patients with acute PE according to CT signs alone of RV dysfunction and risk stratification with CT signs and troponin T levels. (A) Patients with positive findings on the basis of RVD/LVD ratio ⬎1.1 or RVV/LVV ratio ⬎1.5 had statistically higher early death rates than those in patients with negative CT signs. (B) Intermediate-risk patients with CT signs of RV dysfunction (abnormal septal shape, RVD/LVD ratio ⬎1.1, or RVV/LVV ratio ⬎1.5) and/or elevated troponin T levels for risk markers had statistically higher early death rates than low-risk patients. *Statistically significant. IVC ⫽ inferior vena cava.
of the liver; and 6 ⫽ reflux into the inferior vena cava with opacification of distal hepatic veins (Figure 1). The degree of contrast medium reflux was grouped into nonsubstantial (grades 1 to 3) and substantial (grades 4 to 6) reflux. Deviation of the interventricular septum was evaluated as follows: normal (convex toward the right ventricle), flattened, and septal bowing (convex toward the left ventricle).15,16 Flattened septum and septal bowing were considered abnormal septal shapes, indicating RV strain (Figure 1). RV and left ventricular (LV) diameter were measured with calipers on transverse section by identifying the maximum distance between the ventricular endocardium and the interventricular septum, perpendicular to the long axis of the heart. The ratio of RV diameter (RVD) to LV diameter (LVD) was also calculated from these images (Figure 1). The volumetric analysis of the 2 ventricles was performed using a commercially available workstation (Vitrea 2 version 3.8.1; Vital Image, Minnetonka, Minnesota). On every 4 transverse sections (i.e., 4-mm intervals), the endocardial contours for the left and right ventricles were manually traced, and the outlining was then automatically propagated to the neighboring sections. Papillary muscles, moderator bands, and trabeculations were assigned to the intracavitary lumen of the ventricles. For RV volume (RVV), the areas containing a recognizable RV outflow tract below the pulmonary valve were included in the ventricular portion.6 The ventricles were semiautomatically segmented from the valvular plane down to the apex of each ventricle. The volume of each ventricle was automatically estimated, and the ratio of RVV to LV volume (LVV) was subsequently calculated (Figure 1). Laboratory tests for the evaluation of acute PE included measurements of D-dimer and troponin T. The upper limits of D-dimer using latex immunoassay were 300 ng/ml for women and 200 ng/ml for men. Troponin T levels were obtained on admission as well as 12 and 24 hours thereafter. Samples for cardiac markers were centrifuged, and the serum was frozen until used in a quantitative assay (electrochemiluminescence; Roche Diagnostics GmbH, Mannheim, Germany). The reference level of troponin T in our hospital ranged from 0.0 to 0.1 ng/ml. According to the guidelines of the European Heart Association,10 we performed further risk stratification of our
Miscellaneous/Acute PE Risk Stratification Using CTPA and Troponin T
137
Table 4 Univariate and multivariate analysis of signs of right ventricular dysfunction for predicting early death Diagnostic Criterion
IVC contrast reflux IVC contrast reflux and troponin T Abnormal septal shape Abnormal septal shape and troponin T RVD/LVD ratio RVD/LVD ratio and troponin T RVV/LVV ratio RVV/LVV ratio and troponin T Elevated troponin T level
Multivariate Analysis†
Univariate Analysis HR
95% CI
p Value
HR
95% CI
p Value
1.584 2.175 2.343 3.712 3.726 4.721 5.321 9.327 4.823
0.588–4.265 0.838–5.649 0.849–6.466 1.431–9.624 1.407–9.867 1.824–12.216 1.905–14.861 3.584–24.271 1.792–12.980
0.339 0.116 0.071 0.014* 0.008* 0.007* ⬍0.001* ⬍0.001* 0.002*
1.862 2.280 3.621 4.800 3.246 4.357 5.402 10.168 5.204
0.709–4.894 0.838–6.202 1.287–10.192 1.540–14.963 1.112–9.479 1.219–15.577 1.890–15.436 2.323–44.503 1.913–14.161
0.210 0.108 0.015* 0.007* 0.032* 0.024* 0.002* 0.002* 0.001*
* Statistically significant. † Adjusted for co-morbidities: age, cancer, chronic lung disease, and heart failure. HR ⫽ hazard ratio. Other abbreviations as in Tables 2 and 3.
Table 5 Univariate and multivariate analysis of signs of right ventricular dysfunction for predicting adverse outcome Diagnostic Criterion
IVC contrast reflux IVC contrast reflux and troponin T Abnormal septal shape Abnormal septal shape and troponin T RVD/LVD ratio RVD/LVD ratio and troponin T RVV/LVV ratio RVV/LVV ratio and troponin T Elevated troponin T level
Multivariate Analysis†
Univariate Analysis HR
95% CI
p Value
HR
95% CI
p Value
1.573 2.591 1.895 2.880 2.313 3.441 3.339 5.088 3.612
0.772–3.209 1.318–5.094 0.922–3.898 1.466–5.656 1.158–4.621 1.756–6.742 1.603–6.955 2.581–10.031 1.678–7.776
0.174 0.007* 0.052 0.003* 0.011* ⬍0.001* ⬍0.001* ⬍0.001* ⬍0.001*
1.748 2.746 2.396 4.800 2.206 3.449 3.507 5.634 3.903
0.877–3.484 1.327–5.683 1.174–4.888 1.540–14.963 1.097–4.438 1.528–7.785 1.758–6.996 2.434–13.044 1.964–7.754
0.326 0.046* 0.111 0.014* 0.137 0.018* 0.007* ⬍0.001* 0.003*
* Statistically significant. † Adjusted for co-morbidities: age, cancer, chronic lung disease, and heart failure. Abbreviations as in Tables 2 to 4.
non-high-risk PE patients using a combination of CT signs of RV dysfunction and cardiac troponin T levels. If the markers of RV dysfunction and/or myocardial injury were present, the patients were classified as intermediate risk.10,17 Hemodynamically stable patients without evidence of RV dysfunction or myocardial injury were classified as low risk. A fatal clinical outcome was defined as early death within 3 months with reference to previous reports.13,18 Nonfatal adverse clinical outcomes were defined as escalation of therapy, according to the Management Strategies and Prognosis in Pulmonary Embolism Trial 3 (MAPPET-3) criteria,19 including cardiopulmonary resuscitation, required ventilator support, and vasopressor therapy. We used MedCalc version 10.4.8 (MedCalc Software, Mariakerke, Belgium) for all statistical analyses. We used chi-square tests or Fisher’s exact tests for comparisons of categorical variables and independent-samples Student’s t tests or Mann-Whitney U tests for comparisons in the distributions of continuous variables. Using receiver-operating characteristic curve analysis, the area under the curve of the RVV/LVV ratio for predicting 3-month mortality was evaluated, and the optimal cut-off value was determined. The Kaplan-Meier method with a log-rank test was used to
estimate cumulative mortality according to risk stratification. Multivariate analysis with adjusting co-morbidities was used with Cox proportional-hazards regression. A p value ⬍0.05 was considered to indicate statistical significance. Results The clinical characteristics of the study population are listed in Table 1. Our study population included 109 patients (63%) classified as PE likely (score ⬎4) according to the Wells score and 64 patients (37%) with abnormal Ddimer results. Overall, 17 patients (9.8%) died within 3 months. Of these 17 deaths, 8 were judged to be directly related to PE; the remaining 9 were due to cancer (n ⫽ 6), heart failure (n ⫽ 2) due to myocardial infarction, and respiratory failure (n ⫽ 1) due to chronic obstructive lung disease. Of the remaining 156 patients who survived, 17 had nonfatal adverse outcomes: 12 who required ventilator support, 3 who required the use of vasopressors, and 2 who required cardiopulmonary resuscitation. Sixty-one patients (35%) had relevant co-morbidities, and cancer was more common (p ⫽ 0.031) in patients who died within 3 months.
138
The American Journal of Cardiology (www.ajconline.org)
The correlation coefficients of the RVD/LVD and RVV/ LVV ratios between the 2 observers were 0.891 (p ⬍0.001) and 0.928 (p ⬍0.001), respectively. Interobserver agreement was good for inferior vena cava contrast reflux ( ⫽ 0.723) and abnormal septal shape ( ⫽ 0.651) and excellent for RVD/LVD ⬎1.1 ( ⫽ 0.808) and RVV/LVV ⬎1.5 ( ⫽ 0.826). Using receiver-operating characteristic curve analysis, the optimal cut-off values for predicting early deaths or adverse outcomes were 1.1 and 1.5 for RVD/LVD and RVV/LVV, respectively, and 0.20 ng/ml for cardiac troponin T level. An RVD/LVD ratio ⬎1.1 and an RVV/LVV ratio ⬎1.5 and an elevated troponin T level were more common in patients with early death or adverse outcomes than in patients who survived or had uncomplicated clinical courses (Table 2). However, there was no statistical difference in the prevalence of inferior vena cava contrast reflux and abnormal septal shape between patients who died and those who survived or between patients with adverse outcomes and uncomplicated clinical courses. The RVD/LVD ratio or RVV/LVV ratio alone could predict early death or adverse outcome, whereas inferior vena cava contrast reflux or abnormal septal shape alone did not (Table 3). RVV/LVV ⬎1.5 had the highest area under the curve (0.709, p ⬍0.001) for predicting early death compared to other CT signs. When troponin T level was combined as a risk stratification criterion, there were increases in the area under the curve for predicting early death or adverse outcome. The combination of an RVV/LVV ratio ⬎1.5 and an elevated troponin T level revealed the highest areas under the curve (0.733 and 0.706, respectively) for predicting early death or adverse outcome. Early death within 3 months was higher (13% to 22%) in patients with CT signs of RV dysfunction or elevated troponin T levels than in patients with negative findings (4% to 8%; Figure 2). According to risk stratification,10 using a combination of CT signs of RV dysfunction and elevated troponin T level, early mortality in the intermediate-risk group was higher (14% to 19%) than in the low-risk group (2% to 6%) (Figure 2). On the basis of univariate analysis, an RVD/LVD ratio ⬎1.1 and an RVV/LVV ratio ⬎1.5 and an elevated troponin T level were predictors of early death. On multivariate analysis after adjusting for co-morbidities, an abnormal septal shape was added as a predictor of early death. The hazard ratios (5.321 and 5.402, respectively) of RVV/LVV ⬎1.5 in univariate and multivariate analyses were higher than those of other CT signs or elevated troponin T level (Table 4). When troponin T level was combined as a risk stratification criterion, the hazard ratios were increased compared to those of CT signs alone. The hazard ratios (9.327 and 10.168, respectively) in univariate and multivariate analyses were highest in the combination of RVV/LVV ratio and elevated troponin T level. On univariate analysis, RVD/LVD ⬎1.1 and RVV/LVV ⬎1.5 and elevated troponin T level were predictors of adverse outcomes. Among CT signs of RV dysfunction, however, only an RVV/LVV ratio ⬎1.5 was a predictor of adverse outcomes on multivariate analysis (Table 5). All combinations of CT signs and elevated troponin T levels were predictors of adverse outcomes on univariate and mul-
tivariate analyses. The hazard ratios (5.088 and 5.634, respectively) on univariate and multivariate analyses were highest in the combination of RVV/LVV ratio and an elevated troponin T level, while other combinations were not higher than the hazard ratio (3.612 and 3.903, respectively) of an elevated troponin level alone. Discussion Risk stratification for patients with PE starts with clinical assessment of hemodynamic status and classification into high- and non-high-risk PEs.10 High-risk patients who are hemodynamically unstable have a ⬎15% potential for early mortality, whereas non-high-risk patients generally have benign clinical courses. However, although hemodynamically stable, the presence of markers of RV dysfunction can provide indirect evidence of impending hemodynamic failure. In 1 trial,20 10% of normotensive patients with RV dysfunction developed PE-related shock. In comparison, patients with no RV dysfunction on echocardiography had excellent outcomes; the reported mortality was lower, at approximately 1% to 2%.10,11,20,21 In the present study, the early death rate was higher in patients with RVD/LVD ratios ⬎1.1 or RVV/LVV ratios ⬎1.5 compared to those with negative findings (Figure 2). Thus, the RVD/LVD ratio or RVV/LVV ratio could predict early death or adverse outcomes. In particular, the RVV/LVV ratio was an independent predictor of early death and adverse outcome on multivariate analysis after adjusting for co-morbidities and showed the highest area under the curve (0.709, p ⬍0.001) and hazard ratio (5.402, p ⫽ 0.002) for predicting early death. In contrast, troponins are released in the bloodstream in the presence of myocardial damage secondary to microinfarction,22 and the increase in troponins correlates with RV dysfunction.22,23 Becattini et al24 reported that elevation of troponin I and/or T was associated with higher mortality (17.9%) compared to patients with normal troponin levels (2.3%), even in a subgroup of hemodynamically stable patients. Our study also confirmed that elevated troponin T was associated with higher mortality in non-high-risk patients (Figure 2). However, the main limitation of cardiac biomarkers is that different biomarker thresholds and various outcome definitions have been used in previous studies. In 1 recent study, an increased RVD/LVD ratio showed higher sensitivity (90% vs 40%) and a higher odds ratio (9.0 vs 6.3) for predicting an adverse outcome, compared to those of elevated troponin T.25 In our study, the area under the curve and hazard ratio of elevated troponin T levels for predicting adverse outcome were higher than those of the RVD/LVD and RVV/LVV ratio. The difference between the 2 studies is that we measured the RVD/LVD ratio on axial images, while the previous investigators used 4-chamber views. An alternative approach for predicting PE risk is to combine markers of RV dysfunction with other laboratory findings, which would add incremental prognostic value. In 1 trial,8 the combination of RV enlargement and an elevated troponin level significantly increased all-cause 30-day mortality (38%) compared to patients with elevated troponin levels (23%) or RV dilatation alone (9%). In contrast, preserved RV function with negative cardiac biomarkers iden-
Miscellaneous/Acute PE Risk Stratification Using CTPA and Troponin T
tified patients with excellent prognoses.7,8 Thus, the combination of echocardiography and cardiac troponin level had incremental prognostic information compared to echocardiography or cardiac troponin level alone. However, no report has addressed the combined assessment of CT signs of RV dysfunction and cardiac biomarkers. Indeed, the present study is the first regarding the combined assessment of CT signs of RV dysfunction and cardiac biomarkers, although a comparison study of CT signs of RV enlargement and cardiac biomarkers has been published.25 When troponin T level was applied as a cardiac biomarker, the areas under the curve of the CT signs for predicting early death or adverse outcomes were increased. Moreover, the combination of RVV/LVV ratio and elevated troponin T level revealed the highest area under the curve and hazard ratio for predicting early death as well as adverse outcomes. Thus, non-highrisk PEs can be further divided into intermediate- and lowrisk groups according to the presence of markers of RV dysfunction or myocardial injury.10 The low-risk group with no CT signs of RV dysfunction and negative troponin T findings had a lower mortality rate than the intermediaterisk group (Figure 2). Our study had several limitations. The most important, as with all previous work on this topic, was its retrospective nature. Because only patients who underwent CT pulmonary angiography and cardiac troponin testing obtained at the time of PE diagnosis were included, we cannot exclude selection bias. Second, we applied allcause mortality as a clinical end point. Thus, the patients enrolled in our study had from a variety of severe diseases, and only 8 deaths were judged to be directly related to PE. Third, our CT pulmonary angiographic acquisition by multidetector computed tomography was not electrocardiographically gated. Nongated CT is inevitably inaccurate for measuring ventricular chamber size, because the images are acquired in different phases of the cardiac cycle. However, the use of electrocardiographically gated CT protocols over routine chest CT imaging has been shown to result in only limited incremental diagnostic improvement.26 More important, because of the additional radiation exposure involved with electrocardiographically gated techniques, this approach is not currently used for routine PE imaging.2 Fourth, volumetric measurements were carried out on transverse sections instead of images reconstructed along the short axis, because our workstation did not allow volumetric measurement on reconstructed images of nongated CT. We then followed the method used by Dog˘an et al.6 Finally, we did not include RVD/LVD measurements on 4-chamber views in our study. However, this study was focused on the incremental value of combined assessment rather than comparisons between CT signs. Moreover, 2 previous studies reported that the volumetric assessment of ventricles showed higher interobserver agreement4 and seemed to be superior for the identification of high-risk patients with adverse clinical outcomes.5 1. Goldhaber SZ. Echocardiography in the management of pulmonary embolism. Ann Intern Med 2002;136:691–700. 2. Schoepf UJ, Costello P. CT angiography for diagnosis of pulmonary embolism: state of the art. Radiology 2004;230:329 –337.
139
3. Ghaye B, Ghuysen A, Bruyere PJ, D’Orio V, Dondelinger RF. Can CT pulmonary angiography allow assessment of severity and prognosis in patients presenting with pulmonary embolism? What the radiologist needs to know. Radiographics 2006;26:23– 40. 4. Kang DK, Ramos-Duran L, Schoepf UJ, Armstrong AM, Abro JA, Ravenel JG, Thilo C. Reproducibility of CT signs of right ventricular dysfunction in acute pulmonary embolism. AJR Am J Roentgenol 2010;194:1500 –1506. 5. Henzler T, Krissak R, Reichert M, Sueselbeck T, Schoenberg SO, Fink C. Volumetric analysis of pulmonary CTA for the assessment of right ventricular dysfunction in patients with acute pulmonary embolism. Acad Radiol 2010;17:309 –315. 6. Dog˘an H, Kroft LJ, Huisman MV, van der Geest RJ, de Roos A. Right ventricular function in patients with acute pulmonary embolism: analysis with electrocardiography-synchronized multi-detector row CT. Radiology 2007;242:78 – 84. 7. Kucher N, Wallmann D, Carone A, Windecker S, Meier B, Hess OM. Incremental prognostic value of troponin I and echocardiography in patients with acute pulmonary embolism. Eur Heart J 2003;24:1651– 1656. 8. Scridon T, Scridon C, Skali H, Alvarez A, Goldhaber SZ, Solomon SD. Prognostic significance of troponin elevation and right ventricular enlargement in acute pulmonary embolism. Am J Cardiol 2005;96: 303–305. 9. Binder L, Pieske B, Olschewski M, Geibel A, Klostermann B, Reiner C, Konstantinides S. N-terminal pro-brain natriuretic peptide or troponin testing followed by echocardiography for risk stratification of acute pulmonary embolism. Circulation 2005;112:1573–1579. 10. Torbicki A, Perrier A, Konstantinides S, Agnelli G, Galiè N, Pruszczyk P, Bengel F, Brady AJ, Ferreira D, Janssens U, Klepetko W, Mayer E, Remy-Jardin M, Bassand JP; ESC Committee for Practice Guidelines (CPG). Guidelines on the diagnosis and management of acute pulmonary embolism of the European Society of Cardiology (ESC). Eur Heart J 2008;29:2276 –2315. 11. Jiménez D, Aujesky D, Yusen RD. Risk stratification of normotensive patients with acute symptomatic pulmonary embolism. Br J Haematol 2010;151:415– 424. 12. Wells PS, Anderson DR, Rodger M, Ginsberg JS, Kearon C, Gent M, Turpie AG, Bormanis J, Weitz J, Chamberlain M, Bowie D, Barnes D, Hirsh J. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost 2000;83:416 – 420. 13. Goldhaber SZ, Visani L, De Rosa M. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999;353:1386 –1389. 14. Aviram G, Rogowski O, Gotler Y, Bendler A, Steinvil A, Goldin Y, Graif M, Berliner S. Real-time risk stratification of patients with acute pulmonary embolism by grading the reflux of contrast medium into the inferior vena cava on computerized tomographic pulmonary angiography. J Thromb Haemost 2008;6:1488 –1493. 15. Lim KE, Chan CY, Chu PH, Hsu YY, Hsu WC. Right ventricular dysfunction secondary to acute massive pulmonary embolism detected by helical computed tomography pulmonary angiography. Clin Imaging 2005;29:16 –21. 16. Araoz PA, Gotway MB, Harrington JR, Harmsen WS, Mandrekar JN. Pulmonary embolism: prognostic CT findings. Radiology 2007;242: 889 – 897. 17. Kreit JW. The impact of right ventricular dysfunction on the prognosis and therapy of normotensive patients with pulmonary embolism. Chest 2004;125:1539 –1545. 18. van der Meer RW, Pattynama PM, van Strijen MJ, van den BergHuijsmans AA, Hartmann IJ, Putter H, de Roos A, Huisman MV. Right ventricular dysfunction and pulmonary obstruction index at helical CT: prediction of clinical outcome during 3-month follow-up in patients with acute pulmonary embolism. Radiology 2005;235:798 – 803. 19. Konstantinides S, Geibel A, Heusel G, Heinrich F, Kasper W. Heparin plus alteplase compared with heparin alone in patients with submassive pulmonary embolism. N Engl J Med 2002;347:1143–1150. 20. Grifoni S, Olivotto I, Cecchini P, Pieralli F, Camaiti A, Santoro G, Conti A, Agnelli G, Berni G. Short-term clinical outcome of patients with acute pulmonary embolism, normal blood pressure, and echocardiographic right ventricular dysfunction. Circulation 2000;101:2817–2822.
140
The American Journal of Cardiology (www.ajconline.org)
21. Kasper W, Konstantinides S, Geibel A, Tiede N, Krause T, Just H. Prognostic significance of right ventricular afterload stress detected by echocardiography in patients with clinically suspected pulmonary embolism. Heart 1997;77:346 –349. 22. Müller-Bardorff M, Weidtmann B, Giannitsis E, Kurowski V, Katus HA. Release kinetics of cardiac troponin T in survivors of confirmed severe pulmonary embolism. Clin Chem 2002;48:673– 675. 23. Tulevski II, ten Wolde M, van Veldhuisen DJ, Mulder JW, van der Wall EE, Büller HR, Mulder BJ. Combined utility of brain natriuretic peptide and cardiac troponin T may improve rapid triage and risk stratification in normotensive patients with pulmonary embolism. Int J Cardiol 2007;116:161–166.
24. Becattini C, Agnelli G. Predictors of mortality from pulmonary embolism and their influence on clinical management. Thromb Haemost 2008;100:747–751. 25. Klok FA, Van Der Bijl N, Eikenboom HC, Van Rooden CJ, De Roos A, Kroft LJ, Huisman MV. Comparison of CT assessed right ventricular size and cardiac biomarkers for predicting short-term clinical outcome in normotensive patients suspected of having acute pulmonary embolism. J Thromb Haemost 2010;8:853– 856. 26. Lu MT, Cai T, Ersoy H, Whitmore AG, Levit NA, Goldhaber SZ, Rybicki FJ. Comparison of ECG-gated versus non-gated CT ventricular measurements in thirty patients with acute pulmonary embolism. Int J Cardiovasc Imaging 2009;25:101–107.
Departments of aRadiology and bCardiology, Ajou University School of Medicine, Suwon, South Korea. Manuscript received January 18, 2011; revised manuscript received and accepted March 3, 2011. *Corresponding author: Tel: 82-31-219-5849; fax: 82-31-219-5862. E-mail address: [email protected] (D.K. Kang). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2011.03.009
Methods We retrospectively identified 223 consecutive patients hospitalized at our hospital from March 2005 through March 2010 with acute PE confirmed by CT pulmonary angiography. The medical records department of our institution searched 1,087 patients who underwent CT pulmonary angiography in this period; 266 of these had been registered as having acute PE. In total, 223 patients were positive for PE on CT pulmonary angiography, and 43 had negative PE findings on CT pulmonary angiography. One hundred seventy-three patients were finally eligible for the present study, after the exclusion of 18 patients with incomplete follow-up data for the end point, 15 patients without laboratory data for cardiac troponin, 12 patients with highrisk PE, and 5 patients with insufficient contrast enhancement of the ventricular chambers for reliable delineation of the endocardial borders. High-risk PE was defined as the presence of shock or persistent systemic hypotension (defined as systolic blood pressure ⬍90 mm Hg or a pressure decrease of ⬎40 mm Hg for 15 minutes) and represented an immediately life-threatening emergency requiring specific management.10,11 For clinical assessment of the probability of PE, we used the Wells score, which classifies patients as “PE unlikely” or “PE likely.”12 Co-morbidities that increased the risk for adverse outcome included advanced age, cancer, chronic lung disease, and heart failure, in accordance with the International Cooperative Pulmonary Embolism Registry (ICOPER).13 The Human Research Commitwww.ajconline.org
134
The American Journal of Cardiology (www.ajconline.org)
Figure 1. A 78-year-old man with an acute PE confirmed by CT pulmonary angiography died on the second day of his hospital stay. Axial (A) and volume-rendered technique (B) shows multiple filling defects in the lobar (long arrow) and segmental (short arrows) pulmonary arteries involving the left lower lobe. (C) CT image shows grade 4 reflux of contrast medium into the inferior vena cava (long arrow) and proximal hepatic veins (short arrow). (D) Axial view shows septal bowing (arrows), convex toward the left ventricle (LV). CT images of measurements of maximal right ventricular (E) and left ventricular (F) diameters on transverse sections show an RVD/LVD ratio of 1.3. The endocardial contours of the right ventricle (RV) (G) and the LV (H) were manually traced on the transverse sections, and the outlining was then automatically propagated to the neighboring sections. (I,J) The ventricles were semiautomatically segmented from the valvular plane down to the apex of each ventricle. (K,L) The volume of each ventricle was displayed using a volume-rendered technique and was automatically estimated. The RVV/LVV ratio was subsequently calculated to be 3.2.
tee at Ajou University Hospital approved the study design and waived the need for informed patient consent for this retrospective analysis.
All patients had undergone CT imaging, without electrocardiographic gating, using a 16-slice multidetector CT system (Somatom Sensation; Siemens Medical Systems,
Miscellaneous/Acute PE Risk Stratification Using CTPA and Troponin T
135
Fig. 1. (Cont’d)
Table 1 Clinical characteristics and relevant co-morbidities in non-high risk patients
Table 2 Computed tomographic signs of right ventricular dysfunction for predicting adverse outcomes and 30-day death
Variable
Sign of RV Dysfunction
Total (n ⫽ 173)
Survival (n ⫽ 156) Uncomplicated Clinical Course (n ⫽ 139)
Early Death (n ⫽ 17) Adverse Outcome (n ⫽ 34)
p Value
Age (years)
61 ⫾ 14
Men
76 (44%)
Co-morbidities
61 (35%)
Cancer
41 (24%)
Chronic lung disease Heart failure
15 (9%)
60 ⫾ 15 60 ⫾ 14 68 (44%) 65 (47%) 51 (33%) 46 (33%) 33 (21%) 29 (21%) 14 (9%) 12 (9%) 7 (4%) 7 (5%)
64 ⫾ 12 61 ⫾ 14 8 (47%) 11 (32%) 10 (58%) 15 (44%) 8 (47%) 12 (35%) 1 (6%) 3 (9%) 1 (6%) 1 (3%)
0.296 0.706 0.987 0.185 0.061 0.236 0.031* 0.113 0.999 0.999 0.570 0.999
8 (5%)
Data are expressed as mean ⫾ SD or as numbers (percentage). * Statistically significant.
Forchheim, Germany) or a 64-slice scanner (Brilliance; Phillips Medical Systems, Eindhoven, The Netherlands). The image acquisition parameters for 16-sclice CT imaging were as follows: gantry rotation time 500 ms; collimation 16 ⫻ 0.75 mm; pitch 1.0; tube voltage 120 kV; and effective tube current of 200 mA. The image acquisition parameters for 64-slice CT imaging were as follows: gantry rotation time 500 ms, detector collimation 64 ⫻ 0.625 mm, pitch 0.924, tube voltage 120 kV, and effective tube current 200 mA. Precontrast CT imaging was performed with 3-mm section thickness to evaluate calcification or degree of enhancement, because up to 2/3 of the patients for whom there was initial suspicion of PE received other diagnoses.2 In the next step, postcontrast CT pulmonary angiography was performed with 1-mm reconstruction section thickness. Contrast media (80 ml of Omnipaque 350; GE Healthcare, Dublin, Ireland) was injected intravenously at a rate of 3.2 ml/s, followed by a saline flush (30 ml). Automated bolus triggering was used with a region of interest in the main
IVC contrast reflux Abnormal septal shape RVD/LVD ratio RVV/LVV ratio Elevated troponin T level
Survival (n ⫽ 156) Uncomplicated Clinical Course (n ⫽ 139)
Early Death (n ⫽ 17) Adverse Outcome (n ⫽ 34)
p Value
54 (35%) 44 (32%) 49 (31%) 42 (30%) 58 (37%) 50 (36%) 45 (29%) 37 (27%) 38 (24%) 30 (22%)
8 (47%) 15 (44%) 9 (53%) 16 (47%) 12 (71%) 20 (59%) 12 (71%) 20 (59%) 11 (65%) 19 (56%)
0.453 0.241 0.130 0.097 0.016* 0.025* 0.001* ⬍0.001* 0.001* ⬍0.001*
* Statistically significant. IVC ⫽ inferior vena cava.
pulmonary artery and a threshold of 120 Hounsfield units for triggering data acquisition. Data were acquired in a craniocaudal direction in the supine position with full inspiration. The scanning range extended from the level of the sternal notch to the level of the adrenal gland. All CT studies were independently reviewed by 2 experienced observers (D.K.K., J.S.S.), with 7 and 5 years of experience, respectively. Because of the retrospective nature of the study, the observers were aware that patients had been diagnosed with acute PE, but they were blinded to patients’ other clinical characteristics and outcomes. The initial clinical diagnosis of PE was confirmed in the presence of ⱖ1 filling defect in the pulmonary artery tree (Figure 1), including the subsegmental level.2 The severity of reflux of contrast medium into the inferior vena cava or hepatic veins was graded according to a previously published scale14: 1 ⫽ no reflux; 2 ⫽ trace of reflux into the inferior vena cava only; 3 ⫽ reflux into the inferior vena cava, but not the hepatic veins; 4 ⫽ reflux into the inferior vena cava and proximal hepatic veins; 5 ⫽ reflux into the inferior vena cava and hepatic veins down to the midportion
136
The American Journal of Cardiology (www.ajconline.org)
Table 3 Receiver-operating characteristic curve analysis between signs of right ventricular dysfunction for predicting early death Diagnostic Criterion
IVC contrast reflux IVC contrast reflux and troponin T Abnormal septal shape Abnormal septal shape and troponin T RVD/LVD ratio RVD/LVD ratio and troponin T RVV/LVV ratio RVV/LVV ratio and troponin T Elevated troponin T level
For Predicting Early Death
For Predicting Adverse Outcome
AUC
95% CI
p Value
AUC
95% CI
p Value
0.562 0.606 0.608 0.661 0.667 0.675 0.709 0.733 0.702
0.485–0.637 0.529–0.679 0.531–0.681 0.585–0.731 0.592–0.737 0.599–0.744 0.635–0.775 0.660–0.797 0.628–0.769
0.340 0.094 0.098 0.004* 0.006* ⬍0.001* ⬍0.001* ⬍0.001* 0.006*
0.562 0.637 0.584 0.644 0.614 0.659 0.661 0.706 0.671
0.485–0.637 0.560–0.708 0.507–0.659 0.568–0.715 0.537–0.687 0.584–0.730 0.585–0.731 0.633–0.773 0.596–0.741
0.190 0.003* 0.077 0.001* 0.016* ⬍0.001* ⬍0.001* ⬍0.001* ⬍0.001*
* Statistically significant. AUC ⫽ area under the receiver-operating characteristic curve; CI ⫽ confidence interval. Other abbreviations as in Table 2.
Figure 2. Early death rate within 3 months in 173 patients with acute PE according to CT signs alone of RV dysfunction and risk stratification with CT signs and troponin T levels. (A) Patients with positive findings on the basis of RVD/LVD ratio ⬎1.1 or RVV/LVV ratio ⬎1.5 had statistically higher early death rates than those in patients with negative CT signs. (B) Intermediate-risk patients with CT signs of RV dysfunction (abnormal septal shape, RVD/LVD ratio ⬎1.1, or RVV/LVV ratio ⬎1.5) and/or elevated troponin T levels for risk markers had statistically higher early death rates than low-risk patients. *Statistically significant. IVC ⫽ inferior vena cava.
of the liver; and 6 ⫽ reflux into the inferior vena cava with opacification of distal hepatic veins (Figure 1). The degree of contrast medium reflux was grouped into nonsubstantial (grades 1 to 3) and substantial (grades 4 to 6) reflux. Deviation of the interventricular septum was evaluated as follows: normal (convex toward the right ventricle), flattened, and septal bowing (convex toward the left ventricle).15,16 Flattened septum and septal bowing were considered abnormal septal shapes, indicating RV strain (Figure 1). RV and left ventricular (LV) diameter were measured with calipers on transverse section by identifying the maximum distance between the ventricular endocardium and the interventricular septum, perpendicular to the long axis of the heart. The ratio of RV diameter (RVD) to LV diameter (LVD) was also calculated from these images (Figure 1). The volumetric analysis of the 2 ventricles was performed using a commercially available workstation (Vitrea 2 version 3.8.1; Vital Image, Minnetonka, Minnesota). On every 4 transverse sections (i.e., 4-mm intervals), the endocardial contours for the left and right ventricles were manually traced, and the outlining was then automatically propagated to the neighboring sections. Papillary muscles, moderator bands, and trabeculations were assigned to the intracavitary lumen of the ventricles. For RV volume (RVV), the areas containing a recognizable RV outflow tract below the pulmonary valve were included in the ventricular portion.6 The ventricles were semiautomatically segmented from the valvular plane down to the apex of each ventricle. The volume of each ventricle was automatically estimated, and the ratio of RVV to LV volume (LVV) was subsequently calculated (Figure 1). Laboratory tests for the evaluation of acute PE included measurements of D-dimer and troponin T. The upper limits of D-dimer using latex immunoassay were 300 ng/ml for women and 200 ng/ml for men. Troponin T levels were obtained on admission as well as 12 and 24 hours thereafter. Samples for cardiac markers were centrifuged, and the serum was frozen until used in a quantitative assay (electrochemiluminescence; Roche Diagnostics GmbH, Mannheim, Germany). The reference level of troponin T in our hospital ranged from 0.0 to 0.1 ng/ml. According to the guidelines of the European Heart Association,10 we performed further risk stratification of our
Miscellaneous/Acute PE Risk Stratification Using CTPA and Troponin T
137
Table 4 Univariate and multivariate analysis of signs of right ventricular dysfunction for predicting early death Diagnostic Criterion
IVC contrast reflux IVC contrast reflux and troponin T Abnormal septal shape Abnormal septal shape and troponin T RVD/LVD ratio RVD/LVD ratio and troponin T RVV/LVV ratio RVV/LVV ratio and troponin T Elevated troponin T level
Multivariate Analysis†
Univariate Analysis HR
95% CI
p Value
HR
95% CI
p Value
1.584 2.175 2.343 3.712 3.726 4.721 5.321 9.327 4.823
0.588–4.265 0.838–5.649 0.849–6.466 1.431–9.624 1.407–9.867 1.824–12.216 1.905–14.861 3.584–24.271 1.792–12.980
0.339 0.116 0.071 0.014* 0.008* 0.007* ⬍0.001* ⬍0.001* 0.002*
1.862 2.280 3.621 4.800 3.246 4.357 5.402 10.168 5.204
0.709–4.894 0.838–6.202 1.287–10.192 1.540–14.963 1.112–9.479 1.219–15.577 1.890–15.436 2.323–44.503 1.913–14.161
0.210 0.108 0.015* 0.007* 0.032* 0.024* 0.002* 0.002* 0.001*
* Statistically significant. † Adjusted for co-morbidities: age, cancer, chronic lung disease, and heart failure. HR ⫽ hazard ratio. Other abbreviations as in Tables 2 and 3.
Table 5 Univariate and multivariate analysis of signs of right ventricular dysfunction for predicting adverse outcome Diagnostic Criterion
IVC contrast reflux IVC contrast reflux and troponin T Abnormal septal shape Abnormal septal shape and troponin T RVD/LVD ratio RVD/LVD ratio and troponin T RVV/LVV ratio RVV/LVV ratio and troponin T Elevated troponin T level
Multivariate Analysis†
Univariate Analysis HR
95% CI
p Value
HR
95% CI
p Value
1.573 2.591 1.895 2.880 2.313 3.441 3.339 5.088 3.612
0.772–3.209 1.318–5.094 0.922–3.898 1.466–5.656 1.158–4.621 1.756–6.742 1.603–6.955 2.581–10.031 1.678–7.776
0.174 0.007* 0.052 0.003* 0.011* ⬍0.001* ⬍0.001* ⬍0.001* ⬍0.001*
1.748 2.746 2.396 4.800 2.206 3.449 3.507 5.634 3.903
0.877–3.484 1.327–5.683 1.174–4.888 1.540–14.963 1.097–4.438 1.528–7.785 1.758–6.996 2.434–13.044 1.964–7.754
0.326 0.046* 0.111 0.014* 0.137 0.018* 0.007* ⬍0.001* 0.003*
* Statistically significant. † Adjusted for co-morbidities: age, cancer, chronic lung disease, and heart failure. Abbreviations as in Tables 2 to 4.
non-high-risk PE patients using a combination of CT signs of RV dysfunction and cardiac troponin T levels. If the markers of RV dysfunction and/or myocardial injury were present, the patients were classified as intermediate risk.10,17 Hemodynamically stable patients without evidence of RV dysfunction or myocardial injury were classified as low risk. A fatal clinical outcome was defined as early death within 3 months with reference to previous reports.13,18 Nonfatal adverse clinical outcomes were defined as escalation of therapy, according to the Management Strategies and Prognosis in Pulmonary Embolism Trial 3 (MAPPET-3) criteria,19 including cardiopulmonary resuscitation, required ventilator support, and vasopressor therapy. We used MedCalc version 10.4.8 (MedCalc Software, Mariakerke, Belgium) for all statistical analyses. We used chi-square tests or Fisher’s exact tests for comparisons of categorical variables and independent-samples Student’s t tests or Mann-Whitney U tests for comparisons in the distributions of continuous variables. Using receiver-operating characteristic curve analysis, the area under the curve of the RVV/LVV ratio for predicting 3-month mortality was evaluated, and the optimal cut-off value was determined. The Kaplan-Meier method with a log-rank test was used to
estimate cumulative mortality according to risk stratification. Multivariate analysis with adjusting co-morbidities was used with Cox proportional-hazards regression. A p value ⬍0.05 was considered to indicate statistical significance. Results The clinical characteristics of the study population are listed in Table 1. Our study population included 109 patients (63%) classified as PE likely (score ⬎4) according to the Wells score and 64 patients (37%) with abnormal Ddimer results. Overall, 17 patients (9.8%) died within 3 months. Of these 17 deaths, 8 were judged to be directly related to PE; the remaining 9 were due to cancer (n ⫽ 6), heart failure (n ⫽ 2) due to myocardial infarction, and respiratory failure (n ⫽ 1) due to chronic obstructive lung disease. Of the remaining 156 patients who survived, 17 had nonfatal adverse outcomes: 12 who required ventilator support, 3 who required the use of vasopressors, and 2 who required cardiopulmonary resuscitation. Sixty-one patients (35%) had relevant co-morbidities, and cancer was more common (p ⫽ 0.031) in patients who died within 3 months.
138
The American Journal of Cardiology (www.ajconline.org)
The correlation coefficients of the RVD/LVD and RVV/ LVV ratios between the 2 observers were 0.891 (p ⬍0.001) and 0.928 (p ⬍0.001), respectively. Interobserver agreement was good for inferior vena cava contrast reflux ( ⫽ 0.723) and abnormal septal shape ( ⫽ 0.651) and excellent for RVD/LVD ⬎1.1 ( ⫽ 0.808) and RVV/LVV ⬎1.5 ( ⫽ 0.826). Using receiver-operating characteristic curve analysis, the optimal cut-off values for predicting early deaths or adverse outcomes were 1.1 and 1.5 for RVD/LVD and RVV/LVV, respectively, and 0.20 ng/ml for cardiac troponin T level. An RVD/LVD ratio ⬎1.1 and an RVV/LVV ratio ⬎1.5 and an elevated troponin T level were more common in patients with early death or adverse outcomes than in patients who survived or had uncomplicated clinical courses (Table 2). However, there was no statistical difference in the prevalence of inferior vena cava contrast reflux and abnormal septal shape between patients who died and those who survived or between patients with adverse outcomes and uncomplicated clinical courses. The RVD/LVD ratio or RVV/LVV ratio alone could predict early death or adverse outcome, whereas inferior vena cava contrast reflux or abnormal septal shape alone did not (Table 3). RVV/LVV ⬎1.5 had the highest area under the curve (0.709, p ⬍0.001) for predicting early death compared to other CT signs. When troponin T level was combined as a risk stratification criterion, there were increases in the area under the curve for predicting early death or adverse outcome. The combination of an RVV/LVV ratio ⬎1.5 and an elevated troponin T level revealed the highest areas under the curve (0.733 and 0.706, respectively) for predicting early death or adverse outcome. Early death within 3 months was higher (13% to 22%) in patients with CT signs of RV dysfunction or elevated troponin T levels than in patients with negative findings (4% to 8%; Figure 2). According to risk stratification,10 using a combination of CT signs of RV dysfunction and elevated troponin T level, early mortality in the intermediate-risk group was higher (14% to 19%) than in the low-risk group (2% to 6%) (Figure 2). On the basis of univariate analysis, an RVD/LVD ratio ⬎1.1 and an RVV/LVV ratio ⬎1.5 and an elevated troponin T level were predictors of early death. On multivariate analysis after adjusting for co-morbidities, an abnormal septal shape was added as a predictor of early death. The hazard ratios (5.321 and 5.402, respectively) of RVV/LVV ⬎1.5 in univariate and multivariate analyses were higher than those of other CT signs or elevated troponin T level (Table 4). When troponin T level was combined as a risk stratification criterion, the hazard ratios were increased compared to those of CT signs alone. The hazard ratios (9.327 and 10.168, respectively) in univariate and multivariate analyses were highest in the combination of RVV/LVV ratio and elevated troponin T level. On univariate analysis, RVD/LVD ⬎1.1 and RVV/LVV ⬎1.5 and elevated troponin T level were predictors of adverse outcomes. Among CT signs of RV dysfunction, however, only an RVV/LVV ratio ⬎1.5 was a predictor of adverse outcomes on multivariate analysis (Table 5). All combinations of CT signs and elevated troponin T levels were predictors of adverse outcomes on univariate and mul-
tivariate analyses. The hazard ratios (5.088 and 5.634, respectively) on univariate and multivariate analyses were highest in the combination of RVV/LVV ratio and an elevated troponin T level, while other combinations were not higher than the hazard ratio (3.612 and 3.903, respectively) of an elevated troponin level alone. Discussion Risk stratification for patients with PE starts with clinical assessment of hemodynamic status and classification into high- and non-high-risk PEs.10 High-risk patients who are hemodynamically unstable have a ⬎15% potential for early mortality, whereas non-high-risk patients generally have benign clinical courses. However, although hemodynamically stable, the presence of markers of RV dysfunction can provide indirect evidence of impending hemodynamic failure. In 1 trial,20 10% of normotensive patients with RV dysfunction developed PE-related shock. In comparison, patients with no RV dysfunction on echocardiography had excellent outcomes; the reported mortality was lower, at approximately 1% to 2%.10,11,20,21 In the present study, the early death rate was higher in patients with RVD/LVD ratios ⬎1.1 or RVV/LVV ratios ⬎1.5 compared to those with negative findings (Figure 2). Thus, the RVD/LVD ratio or RVV/LVV ratio could predict early death or adverse outcomes. In particular, the RVV/LVV ratio was an independent predictor of early death and adverse outcome on multivariate analysis after adjusting for co-morbidities and showed the highest area under the curve (0.709, p ⬍0.001) and hazard ratio (5.402, p ⫽ 0.002) for predicting early death. In contrast, troponins are released in the bloodstream in the presence of myocardial damage secondary to microinfarction,22 and the increase in troponins correlates with RV dysfunction.22,23 Becattini et al24 reported that elevation of troponin I and/or T was associated with higher mortality (17.9%) compared to patients with normal troponin levels (2.3%), even in a subgroup of hemodynamically stable patients. Our study also confirmed that elevated troponin T was associated with higher mortality in non-high-risk patients (Figure 2). However, the main limitation of cardiac biomarkers is that different biomarker thresholds and various outcome definitions have been used in previous studies. In 1 recent study, an increased RVD/LVD ratio showed higher sensitivity (90% vs 40%) and a higher odds ratio (9.0 vs 6.3) for predicting an adverse outcome, compared to those of elevated troponin T.25 In our study, the area under the curve and hazard ratio of elevated troponin T levels for predicting adverse outcome were higher than those of the RVD/LVD and RVV/LVV ratio. The difference between the 2 studies is that we measured the RVD/LVD ratio on axial images, while the previous investigators used 4-chamber views. An alternative approach for predicting PE risk is to combine markers of RV dysfunction with other laboratory findings, which would add incremental prognostic value. In 1 trial,8 the combination of RV enlargement and an elevated troponin level significantly increased all-cause 30-day mortality (38%) compared to patients with elevated troponin levels (23%) or RV dilatation alone (9%). In contrast, preserved RV function with negative cardiac biomarkers iden-
Miscellaneous/Acute PE Risk Stratification Using CTPA and Troponin T
tified patients with excellent prognoses.7,8 Thus, the combination of echocardiography and cardiac troponin level had incremental prognostic information compared to echocardiography or cardiac troponin level alone. However, no report has addressed the combined assessment of CT signs of RV dysfunction and cardiac biomarkers. Indeed, the present study is the first regarding the combined assessment of CT signs of RV dysfunction and cardiac biomarkers, although a comparison study of CT signs of RV enlargement and cardiac biomarkers has been published.25 When troponin T level was applied as a cardiac biomarker, the areas under the curve of the CT signs for predicting early death or adverse outcomes were increased. Moreover, the combination of RVV/LVV ratio and elevated troponin T level revealed the highest area under the curve and hazard ratio for predicting early death as well as adverse outcomes. Thus, non-highrisk PEs can be further divided into intermediate- and lowrisk groups according to the presence of markers of RV dysfunction or myocardial injury.10 The low-risk group with no CT signs of RV dysfunction and negative troponin T findings had a lower mortality rate than the intermediaterisk group (Figure 2). Our study had several limitations. The most important, as with all previous work on this topic, was its retrospective nature. Because only patients who underwent CT pulmonary angiography and cardiac troponin testing obtained at the time of PE diagnosis were included, we cannot exclude selection bias. Second, we applied allcause mortality as a clinical end point. Thus, the patients enrolled in our study had from a variety of severe diseases, and only 8 deaths were judged to be directly related to PE. Third, our CT pulmonary angiographic acquisition by multidetector computed tomography was not electrocardiographically gated. Nongated CT is inevitably inaccurate for measuring ventricular chamber size, because the images are acquired in different phases of the cardiac cycle. However, the use of electrocardiographically gated CT protocols over routine chest CT imaging has been shown to result in only limited incremental diagnostic improvement.26 More important, because of the additional radiation exposure involved with electrocardiographically gated techniques, this approach is not currently used for routine PE imaging.2 Fourth, volumetric measurements were carried out on transverse sections instead of images reconstructed along the short axis, because our workstation did not allow volumetric measurement on reconstructed images of nongated CT. We then followed the method used by Dog˘an et al.6 Finally, we did not include RVD/LVD measurements on 4-chamber views in our study. However, this study was focused on the incremental value of combined assessment rather than comparisons between CT signs. Moreover, 2 previous studies reported that the volumetric assessment of ventricles showed higher interobserver agreement4 and seemed to be superior for the identification of high-risk patients with adverse clinical outcomes.5 1. Goldhaber SZ. Echocardiography in the management of pulmonary embolism. Ann Intern Med 2002;136:691–700. 2. Schoepf UJ, Costello P. CT angiography for diagnosis of pulmonary embolism: state of the art. Radiology 2004;230:329 –337.
139
3. Ghaye B, Ghuysen A, Bruyere PJ, D’Orio V, Dondelinger RF. Can CT pulmonary angiography allow assessment of severity and prognosis in patients presenting with pulmonary embolism? What the radiologist needs to know. Radiographics 2006;26:23– 40. 4. Kang DK, Ramos-Duran L, Schoepf UJ, Armstrong AM, Abro JA, Ravenel JG, Thilo C. Reproducibility of CT signs of right ventricular dysfunction in acute pulmonary embolism. AJR Am J Roentgenol 2010;194:1500 –1506. 5. Henzler T, Krissak R, Reichert M, Sueselbeck T, Schoenberg SO, Fink C. Volumetric analysis of pulmonary CTA for the assessment of right ventricular dysfunction in patients with acute pulmonary embolism. Acad Radiol 2010;17:309 –315. 6. Dog˘an H, Kroft LJ, Huisman MV, van der Geest RJ, de Roos A. Right ventricular function in patients with acute pulmonary embolism: analysis with electrocardiography-synchronized multi-detector row CT. Radiology 2007;242:78 – 84. 7. Kucher N, Wallmann D, Carone A, Windecker S, Meier B, Hess OM. Incremental prognostic value of troponin I and echocardiography in patients with acute pulmonary embolism. Eur Heart J 2003;24:1651– 1656. 8. Scridon T, Scridon C, Skali H, Alvarez A, Goldhaber SZ, Solomon SD. Prognostic significance of troponin elevation and right ventricular enlargement in acute pulmonary embolism. Am J Cardiol 2005;96: 303–305. 9. Binder L, Pieske B, Olschewski M, Geibel A, Klostermann B, Reiner C, Konstantinides S. N-terminal pro-brain natriuretic peptide or troponin testing followed by echocardiography for risk stratification of acute pulmonary embolism. Circulation 2005;112:1573–1579. 10. Torbicki A, Perrier A, Konstantinides S, Agnelli G, Galiè N, Pruszczyk P, Bengel F, Brady AJ, Ferreira D, Janssens U, Klepetko W, Mayer E, Remy-Jardin M, Bassand JP; ESC Committee for Practice Guidelines (CPG). Guidelines on the diagnosis and management of acute pulmonary embolism of the European Society of Cardiology (ESC). Eur Heart J 2008;29:2276 –2315. 11. Jiménez D, Aujesky D, Yusen RD. Risk stratification of normotensive patients with acute symptomatic pulmonary embolism. Br J Haematol 2010;151:415– 424. 12. Wells PS, Anderson DR, Rodger M, Ginsberg JS, Kearon C, Gent M, Turpie AG, Bormanis J, Weitz J, Chamberlain M, Bowie D, Barnes D, Hirsh J. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost 2000;83:416 – 420. 13. Goldhaber SZ, Visani L, De Rosa M. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999;353:1386 –1389. 14. Aviram G, Rogowski O, Gotler Y, Bendler A, Steinvil A, Goldin Y, Graif M, Berliner S. Real-time risk stratification of patients with acute pulmonary embolism by grading the reflux of contrast medium into the inferior vena cava on computerized tomographic pulmonary angiography. J Thromb Haemost 2008;6:1488 –1493. 15. Lim KE, Chan CY, Chu PH, Hsu YY, Hsu WC. Right ventricular dysfunction secondary to acute massive pulmonary embolism detected by helical computed tomography pulmonary angiography. Clin Imaging 2005;29:16 –21. 16. Araoz PA, Gotway MB, Harrington JR, Harmsen WS, Mandrekar JN. Pulmonary embolism: prognostic CT findings. Radiology 2007;242: 889 – 897. 17. Kreit JW. The impact of right ventricular dysfunction on the prognosis and therapy of normotensive patients with pulmonary embolism. Chest 2004;125:1539 –1545. 18. van der Meer RW, Pattynama PM, van Strijen MJ, van den BergHuijsmans AA, Hartmann IJ, Putter H, de Roos A, Huisman MV. Right ventricular dysfunction and pulmonary obstruction index at helical CT: prediction of clinical outcome during 3-month follow-up in patients with acute pulmonary embolism. Radiology 2005;235:798 – 803. 19. Konstantinides S, Geibel A, Heusel G, Heinrich F, Kasper W. Heparin plus alteplase compared with heparin alone in patients with submassive pulmonary embolism. N Engl J Med 2002;347:1143–1150. 20. Grifoni S, Olivotto I, Cecchini P, Pieralli F, Camaiti A, Santoro G, Conti A, Agnelli G, Berni G. Short-term clinical outcome of patients with acute pulmonary embolism, normal blood pressure, and echocardiographic right ventricular dysfunction. Circulation 2000;101:2817–2822.
140
The American Journal of Cardiology (www.ajconline.org)
21. Kasper W, Konstantinides S, Geibel A, Tiede N, Krause T, Just H. Prognostic significance of right ventricular afterload stress detected by echocardiography in patients with clinically suspected pulmonary embolism. Heart 1997;77:346 –349. 22. Müller-Bardorff M, Weidtmann B, Giannitsis E, Kurowski V, Katus HA. Release kinetics of cardiac troponin T in survivors of confirmed severe pulmonary embolism. Clin Chem 2002;48:673– 675. 23. Tulevski II, ten Wolde M, van Veldhuisen DJ, Mulder JW, van der Wall EE, Büller HR, Mulder BJ. Combined utility of brain natriuretic peptide and cardiac troponin T may improve rapid triage and risk stratification in normotensive patients with pulmonary embolism. Int J Cardiol 2007;116:161–166.
24. Becattini C, Agnelli G. Predictors of mortality from pulmonary embolism and their influence on clinical management. Thromb Haemost 2008;100:747–751. 25. Klok FA, Van Der Bijl N, Eikenboom HC, Van Rooden CJ, De Roos A, Kroft LJ, Huisman MV. Comparison of CT assessed right ventricular size and cardiac biomarkers for predicting short-term clinical outcome in normotensive patients suspected of having acute pulmonary embolism. J Thromb Haemost 2010;8:853– 856. 26. Lu MT, Cai T, Ersoy H, Whitmore AG, Levit NA, Goldhaber SZ, Rybicki FJ. Comparison of ECG-gated versus non-gated CT ventricular measurements in thirty patients with acute pulmonary embolism. Int J Cardiovasc Imaging 2009;25:101–107.