European Journal of Radiology 84 (2015) 332–337
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
CT pulmonary angiography findings that predict 30-day mortality in patients with acute pulmonary embolism Andreas Gunter Bach a,∗ , Baasai Nansalmaa a , Johanna Kranz a , Bettina-Maria Taute b , Andreas Wienke c , Dominik Schramm a , Alexey Surov a a
Department of Radiology, Martin-Luther-University Halle-Wittenberg, Ernst-Grube-Str. 40, 06120 Halle , Germany Department of Internal Medicine, Martin-Luther-University Halle-Wittenberg, Ernst-Grube-Str. 40, 06120 Halle , Germany c Institute of Medical Epidemiology, Biostatistics and Informatics, Martin-Luther-University Halle-Wittenberg, Magdeburger-Str. 8, 06112 Halle, Germany b
a r t i c l e
i n f o
Article history: Received 21 October 2014 Received in revised form 7 November 2014 Accepted 11 November 2014 Keywords: Pulmonary embolism 30-day mortality Right ventricular dysfunction Computed tomography Contrast reflux Embolus burden
a b s t r a c t Purpose: Standard computed tomography pulmonary angiography (CTPA) can be used to diagnose acute pulmonary embolism. In addition, multiple findings at CTPA have been proposed as potential tools for risk stratification. Therefore, the aim of the present study is to examine the prognostic value of (I) thrombus distribution, (II) morphometric parameters of right ventricular dysfunction, and (III) contrast reflux in inferior vena cava on 30-day mortality. Material and methods: In a retrospective, single-center study from 06/2005 to 01/2010 365 consecutive patients were included. Inclusion criteria were: presence of acute pulmonary embolism, and availability of 30-day follow-up. A review of patient charts and images was performed. Results: There were no significant differences between the group of 326 survivors and 39 non-survivors in (I) thrombus distribution, and (II) morphometric measurements of right ventricular dysfunction. However, (III) contrast reflux in inferior vena cava was significantly stronger in non-survivors (odds ratio 3.29; p < 0.001). Results were independent from comorbidities like heart insufficiency and pulmonary hypertension. Conclusion: Measurement of contrast reflux is a new and robust method for predicting 30-day mortality in patients with acute pulmonary embolism. Obstruction scores and morphometric measurements of right ventricular dysfunction perform poor as risk stratification tools. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Acute pulmonary embolism (PE) is a common disease. It has a high mortality, especially in patients with right ventricular dysfunction [1]. Computed tomography pulmonary angiography is most commonly used for the detection of pulmonary embolism [2,3]. At the same time, multiple findings at CTPA such as thrombus characteristics and indicators of right ventricular dysfunction [4–6] have been proposed as potential tools for risk stratification [7]. Other tools that play a major role for risk stratification are clinical scores [8], biomarkers [9], and patient characteristics [3]. However, these were not considered in the present work. Both obstruction and signs of right ventricular dysfunction have the potential to support risk stratification. However, it is under
∗ Corresponding author. Tel.: +49 345 5572441. E-mail address: [email protected] (A.G. Bach). http://dx.doi.org/10.1016/j.ejrad.2014.11.023 0720-048X/© 2014 Elsevier Ireland Ltd. All rights reserved.
continued debate which radiologic signs have best prognostic value [4–6]. To quantify obstruction, the majority of previous studies [7] used the Qanadli score [10] or coarse descriptions of thrombus distribution. The Qanadli score is fast to assess. However, it is limited in its description of the thrombus distribution and obstruction grades. Therefore, the much more detailed Mastora score [11] was used in the present study. Right ventricular dysfunction leads to morphometric changes of the great thoracic vessels and of the heart chambers [12]. The ratio of the short axes of the right and left ventricle diameter is the measurement that is most commonly used to detect right ventricular dysfunction [4–6]. In addition, the diameter of the great thoracic vessels, and of the Inferior vena cava (IVC) is commonly measured [4–6]. Furthermore, the interventricular septal angle is associated with elevated pulmonary artery pressure [13]. Reflux of contrast medium into IVC was suggested to serve as an indicator of right heart dysfunction [14,15] and PE [16]. Therefore,
A.G. Bach et al. / European Journal of Radiology 84 (2015) 332–337
333
multifaceted measurement of IVC contrast reflux has the potential as a new method for predicting 30-day mortality. The aim of the present work is to examine (I) thrombus characteristics, (II) morphometric parameters of right ventricular dysfunction, and (III) contrast reflux in inferior vena cava (IVC) regarding their prognostic value for 30-day mortality. 2. Materials and methods 2.1. Study design This study was accepted by the institutional Ethics Committee. In a single-center, retrospective study at a hospital of maximum care, all patients with acute symptomatic PE were reviewed from 06/2005 to 01/2010. Patients were identified via analysis of patient records and using the radiological information system. Inclusion criteria were: performance of computed tomography pulmonary angiography positive for pulmonary embolism, availability of patient records and 30-day follow-up. The final study group included 365 patients (age median 65 years, range 18–91, 178 males). In the study group 39 patients (10.7%) died within 30 days due to PE. Survival was defined as surviving the following 30 days after the PE diagnosis. Patients who died from acute respiratory failure, cardiopulmonary arrest or shock in the absence of other cardiopulmonary diseases in this time were defined as non-survivors. Patients who died from other causes were excluded from the study group: one patient with intracranial hemorrhage, one patient with acute renal failure, and four patients with septic shock. 2.2. Computed tomography examination and review The computed tomography scanners used during the study period were two 64-multidetector systems (Somatom Sensation 64, Siemens, Erlangen, Germany; Aquillon 64, Toshiba, Neuss, Germany). In the pulmonary computed tomography angiography examination 60 ml of an iodinated intravenous contrast agent (Solutrast© 370 with 370 mg iodine/ml, Bracco Imaging Germany GmbH, Konstanz, Germany) were given at a rate of 2.0 ml/s via a central venous catheter or via peripheral venous line. Automatic bolus tracking was performed in the pulmonary trunk with a trigger of 120 Hounsfield units (HU). The mean density of the pulmonary trunk was 311 HU with a 95% confidence interval of 299 HU to 323 HU. Typical imaging parameters were 120 kVp, 150–300 mAs, slice thickness 1 mm, and a pitch of 0.6–1.2 depending on body size. It is known that the IVC diameter is dependent on breath cycle [17,18]. All CT examinations were routinely performed at full inspiration. The stored images of all included patients were reviewed by two radiologists (AGB, DS) in duplicate to perform all measurements. The reviewers worked in consensus and were blinded to the patient group. The reviewers were free to use any windows setting and both standard and lung kernel reconstructions. Multiplanar reformatting was available at a separate workstation (Vitrea; Vital Images, Minnetonka, Minnesota, USA). 2.3. Thrombus distribution The detailed Mastora score [11] was used in the present study. The obstruction of the mediastinal, lobar, and segmental arteries is quantified by a percentage score, and the sum of the percentages gives the global obstruction with a maximum of 300%. The use of the Mastora score allows for detailed analysis of thrombus distribution. The number of affected vessels was calculated. It included the number of vessels with thrombi,irrespective of vessel size. The
Fig. 1. Axial slice of a computed tomography pulmonary angiography at the level of the pulmonary trunk. The typical position of measurements is shown: diameter of the pulmonary trunk (A), diameter of the ascending aorta (B), diameter of the descending aorta (C) and the diameter of the superior vena cava (D).
central obstruction included the obstruction of the pulmonary trunk, the main pulmonary arteries, and the interlobar arteries (maximum 100%). The lobar obstruction and the segmental obstruction included the obstruction of the lobar and segmental arteries, respectively. The upper and lower lobe obstruction included the obstruction of the upper and lower lobar and segmental arteries, respectively. The middle lobe obstruction included the obstruction of the interlobar arteries, middle lobe arteries, and middle segmental arteries. 2.4. Morphometric parameters of right ventricular dysfunction The ratio of the short axes of the right and left ventricle was measured, as described before [12]. The pulmonary trunk was measured on an axial slice on which it showed its maximal diameter. On the same slice the diameter of the ascending aorta, the diameter of the descending aorta and the diameter of the superior vena cava were measured (Fig. 1). The distance from wall to wall was regarded as diameter. Thrombi were ignored when measuring diameter. The diameter of the right pulmonary artery was measured at the position shown in Fig. 2. The diameter of the IVC was measured on axial slices between the liver and the heart. In addition, the interventricular septal angle was measured on axial slices as the angle between the interventricular septum and the line joining the midpoint of the sternum to the thoracic vertebral spinous process [13]. 2.5. Contrast reflux in inferior vena cava In 17 patients contrast was applied via a femoral catheters. These patients were excluded from the analysis of reflux. According to the scale established by Groves [14] and simplified by Aviram [16] reflux was quantified in a 3-point scale: no reflux in IVC (Table 1A and B), subcardial reflux in IVC (Table 1C and D), intrahepatic reflux in IVC (Table 1E and F). To quantify the reflux, measurements of the density in the IVC were performed at subcardial level and at the level of the liver. The area from which the mean density in Hounsfield units was computed was at least 100 mm2 for each measurement. In addition, the craniocaudal length of reflux in the IVC was measured on coronal slices. Measurement was performed from the
334
A.G. Bach et al. / European Journal of Radiology 84 (2015) 332–337
Table 1 Examples of contrast reflux into the inferior vena cava according to a 3-point scale (computed tomography pulmonary angiography examinations in different patients).
base of the right atrium to the inferior margin of contrasted vessel volume (Fig. 3). When reflux was discontinuous (as shown in Fig. 3), still the inferior margin of contrast in the IVC was accepted. When the patient has more reflux and a slower pulmonary circulation the IVC gets progressively brighter while the aorta gets progressively darker. To quantify this effect, the ratio of the IVC (HU) density to the descending aorta (HU) density was calculated. Furthermore, it was recorded whether contrast was applied via central or peripheral venous catheter. This was done to estimate potential bias caused by different injection sites. Another potential bias is the presence of comorbidities that have impact on right heart function. Therefore, information on the presence of the following conditions were collected from patient records: heart insufficiency NYHA IV, pulmonary hypertension > 55 mm Hg, chronic lung disease, and organ insufficiency. Organ insufficiency was defined like in the Acute Physiology and Chronic Health Evaluation II (APACHE II) [19].
Fig. 2. Axial slice of a computed tomography pulmonary angiography at the level of the pulmonary trunk. The typical position of measurement of the diameter of the right pulmonary artery is shown (A).
2.6. Statistical analysis Collected data were evaluated by means of descriptive statistics: absolute and relative frequencies, median and 25–75% interquartile range. Sensitivity, specificity, negative predictive value, and positive predictive value were given to describe the prognostic power. The two-tailed Mann–Whitney U test was used to estimate statistical differences between groups for metric variables. The Chi-square-test was used to estimate statistical differences between groups for binary variables (SPSS Version 18, IBM, New York). The suggested cut-off points for measurements to predict 30-day mortality were selected so that the Youden index (sensitivity + specificity − 1) had a maximum value. A binary logistic multiple regression analysis for 30 days mortality was performed with comorbidities and reflux measurement as covariates. A pvalue of p < 0.05 was considered significant.
Fig. 3. Coronal slice of a computed tomography pulmonary angiography. The black line indicates the craniocaudal length of contrast agent reflux in the inferior vena cava. The cranial limit of the measurement is the right atrium. The caudal limit of the measurement is the inferior margin of the contrast column in the inferior vena cava.
A.G. Bach et al. / European Journal of Radiology 84 (2015) 332–337
335
Table 2 Thrombus distribution in survivors and non-survivors. Survivor* (n = 326)
Non-survivor* (n = 39)
p-Value
Thrombus mass Global obstruction (Mastora score) [%] Number of affected vessels [count]
30 (7–67) 10 (3–18)
19 (5–89) 7 (3–18)
0.810 0.447
Thrombus distribution by level Central obstruction [%] Lobar obstruction [%] Segmental obstruction [%]
0 (0–8) 13 (3–33) 13 (4–25)
0 (0–16) 10 (0–35) 7 (4–31)
0.238 0.846 0.954
Thrombus distribution by lobe Upper lobe obstruction [%] Middle lobe obstruction [%] Lower lobe obstruction [%]
10 (0–28) 7 (0–27) 17 (5–32)
7 (2–31) 7 (0–29) 11 (2–28)
0.962 0.668 0.676
*
Median (25–75% interquartile range).
Table 3 Morphometric parameters of right ventricular dysfunction in survivors and non-survivors. Measurements are sorted in order of ascending p-value.
Diameter of pulmonary trunk [mm] Ratio short axis right ventricle/left ventricle Short axis of the left ventricle [mm] Interventricular septal angle [◦ ] Short axis of the right ventricle [mm] Diameter of superior vena cava [mm] Diameter of ascending aorta [mm] Diameter of descending aorta [mm] Diameter of IVC [mm] Diameter of right pulmonary artery [mm]
Survivor* (n = 326)
Non-survivor* (n = 39)
p-Value
28.7 (26.3–32.2) 1.1 (0.9–1.4) 42 (37–48) 44 (39–52) 48 (42–54) 23 (21–26) 33 (30–36) 25 (23–27) 25 (25–31) 24 (21–27)
30.6 (27.8–32.4) 1.1 (1.0–1.7) 39 (32–48) 48 (38–56) 50 (43–56) 22 (19–25) 33 (30–36) 24 (22–26) 25 (25–31) 24 (22–27)
0.016 0.078 0.083 0.102 0.203 0.248 0.747 0.750 0.907 0.922
Abbreviation: IVC inferior vena cava. * Median (25–75% interquartile range).
3. Results
morphometric parameters showed no significant difference between groups.
The 30-day mortality was 10.7%. In the study period, 14 patients with acute PE died from respiratory failure and 25 patients died from cardiopulmonary arrest. 3.1. Thrombus distribution The thrombus distribution was calculated for each patient. The median of the results for the group of survivors and non-survivors is shown in Table 2. The thrombus mass was not significantly different between survivors (30%) and non-survivors (19%). In addition, there was also no significant difference in thrombus distribution by level or by lobe. 3.2. Morphometric parameters of right ventricular dysfunction Measurements were performed for each patient. The median of the results for the group of survivors and non-survivors is shown in Table 3. The diameter of the pulmonary trunk was significantly different between survivors and non-survivors. All other
3.3. Contrast reflux in inferior vena cava Measurements of contrast reflux were performed for each patient. The median of the results for the group of survivors nonsurvivors is shown in Table 4. All measurements of reflux were significantly different in the groups of survivors and non-survivors. The highest significance was found in the measurement of the craniocaudal length of reflux in IVC: median in survivors 7 mm versus median in non-survivors 33 mm. The optimal cut off to predict 30-day mortality according to the Youden index, as well as sensitivity, specificity, positive and negative predictive value were calculated for the measurements of contrast reflux that showed the most significant differences between survivors and non-survivors (Table 5). When the craniocaudal length of reflux in IVC was ≥31 mm about 28% of patients with acute PE died within 30 days. About
Table 4 Quantification of contrast reflux in inferior vena cava in survivors and non-survivors. Measurements are sorted in order of ascending p-value.
Craniocaudal length of reflux in IVC [mm] Contrast of IVC intrahepatic [HU] IVC reflux III (intrahepatic) [yes] Contrast of IVC subcardial [HU] Ratio contrast IVC/aorta descendes IVC reflux II–III (subcardial or intrahepatic) [yes] IVCI reflux I (no reflux) [yes]
Survivor* (n = 310)
Non-survivor* (n = 38)
p-Value
7 (0–23) 44 (40–99) 30% (93) 106 (45–187) 0.3 (0.2–0.6) 59% (183) 41% (127)
33 (12–57) 95 (55–192) 58% (22) 170 (108–240) 0.6 (0.2–1.8) 82% (31) 18% (7)
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
CT pulmonary angiography findings that predict 30-day mortality in patients with acute pulmonary embolism Andreas Gunter Bach a,∗ , Baasai Nansalmaa a , Johanna Kranz a , Bettina-Maria Taute b , Andreas Wienke c , Dominik Schramm a , Alexey Surov a a
Department of Radiology, Martin-Luther-University Halle-Wittenberg, Ernst-Grube-Str. 40, 06120 Halle , Germany Department of Internal Medicine, Martin-Luther-University Halle-Wittenberg, Ernst-Grube-Str. 40, 06120 Halle , Germany c Institute of Medical Epidemiology, Biostatistics and Informatics, Martin-Luther-University Halle-Wittenberg, Magdeburger-Str. 8, 06112 Halle, Germany b
a r t i c l e
i n f o
Article history: Received 21 October 2014 Received in revised form 7 November 2014 Accepted 11 November 2014 Keywords: Pulmonary embolism 30-day mortality Right ventricular dysfunction Computed tomography Contrast reflux Embolus burden
a b s t r a c t Purpose: Standard computed tomography pulmonary angiography (CTPA) can be used to diagnose acute pulmonary embolism. In addition, multiple findings at CTPA have been proposed as potential tools for risk stratification. Therefore, the aim of the present study is to examine the prognostic value of (I) thrombus distribution, (II) morphometric parameters of right ventricular dysfunction, and (III) contrast reflux in inferior vena cava on 30-day mortality. Material and methods: In a retrospective, single-center study from 06/2005 to 01/2010 365 consecutive patients were included. Inclusion criteria were: presence of acute pulmonary embolism, and availability of 30-day follow-up. A review of patient charts and images was performed. Results: There were no significant differences between the group of 326 survivors and 39 non-survivors in (I) thrombus distribution, and (II) morphometric measurements of right ventricular dysfunction. However, (III) contrast reflux in inferior vena cava was significantly stronger in non-survivors (odds ratio 3.29; p < 0.001). Results were independent from comorbidities like heart insufficiency and pulmonary hypertension. Conclusion: Measurement of contrast reflux is a new and robust method for predicting 30-day mortality in patients with acute pulmonary embolism. Obstruction scores and morphometric measurements of right ventricular dysfunction perform poor as risk stratification tools. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Acute pulmonary embolism (PE) is a common disease. It has a high mortality, especially in patients with right ventricular dysfunction [1]. Computed tomography pulmonary angiography is most commonly used for the detection of pulmonary embolism [2,3]. At the same time, multiple findings at CTPA such as thrombus characteristics and indicators of right ventricular dysfunction [4–6] have been proposed as potential tools for risk stratification [7]. Other tools that play a major role for risk stratification are clinical scores [8], biomarkers [9], and patient characteristics [3]. However, these were not considered in the present work. Both obstruction and signs of right ventricular dysfunction have the potential to support risk stratification. However, it is under
∗ Corresponding author. Tel.: +49 345 5572441. E-mail address: [email protected] (A.G. Bach). http://dx.doi.org/10.1016/j.ejrad.2014.11.023 0720-048X/© 2014 Elsevier Ireland Ltd. All rights reserved.
continued debate which radiologic signs have best prognostic value [4–6]. To quantify obstruction, the majority of previous studies [7] used the Qanadli score [10] or coarse descriptions of thrombus distribution. The Qanadli score is fast to assess. However, it is limited in its description of the thrombus distribution and obstruction grades. Therefore, the much more detailed Mastora score [11] was used in the present study. Right ventricular dysfunction leads to morphometric changes of the great thoracic vessels and of the heart chambers [12]. The ratio of the short axes of the right and left ventricle diameter is the measurement that is most commonly used to detect right ventricular dysfunction [4–6]. In addition, the diameter of the great thoracic vessels, and of the Inferior vena cava (IVC) is commonly measured [4–6]. Furthermore, the interventricular septal angle is associated with elevated pulmonary artery pressure [13]. Reflux of contrast medium into IVC was suggested to serve as an indicator of right heart dysfunction [14,15] and PE [16]. Therefore,
A.G. Bach et al. / European Journal of Radiology 84 (2015) 332–337
333
multifaceted measurement of IVC contrast reflux has the potential as a new method for predicting 30-day mortality. The aim of the present work is to examine (I) thrombus characteristics, (II) morphometric parameters of right ventricular dysfunction, and (III) contrast reflux in inferior vena cava (IVC) regarding their prognostic value for 30-day mortality. 2. Materials and methods 2.1. Study design This study was accepted by the institutional Ethics Committee. In a single-center, retrospective study at a hospital of maximum care, all patients with acute symptomatic PE were reviewed from 06/2005 to 01/2010. Patients were identified via analysis of patient records and using the radiological information system. Inclusion criteria were: performance of computed tomography pulmonary angiography positive for pulmonary embolism, availability of patient records and 30-day follow-up. The final study group included 365 patients (age median 65 years, range 18–91, 178 males). In the study group 39 patients (10.7%) died within 30 days due to PE. Survival was defined as surviving the following 30 days after the PE diagnosis. Patients who died from acute respiratory failure, cardiopulmonary arrest or shock in the absence of other cardiopulmonary diseases in this time were defined as non-survivors. Patients who died from other causes were excluded from the study group: one patient with intracranial hemorrhage, one patient with acute renal failure, and four patients with septic shock. 2.2. Computed tomography examination and review The computed tomography scanners used during the study period were two 64-multidetector systems (Somatom Sensation 64, Siemens, Erlangen, Germany; Aquillon 64, Toshiba, Neuss, Germany). In the pulmonary computed tomography angiography examination 60 ml of an iodinated intravenous contrast agent (Solutrast© 370 with 370 mg iodine/ml, Bracco Imaging Germany GmbH, Konstanz, Germany) were given at a rate of 2.0 ml/s via a central venous catheter or via peripheral venous line. Automatic bolus tracking was performed in the pulmonary trunk with a trigger of 120 Hounsfield units (HU). The mean density of the pulmonary trunk was 311 HU with a 95% confidence interval of 299 HU to 323 HU. Typical imaging parameters were 120 kVp, 150–300 mAs, slice thickness 1 mm, and a pitch of 0.6–1.2 depending on body size. It is known that the IVC diameter is dependent on breath cycle [17,18]. All CT examinations were routinely performed at full inspiration. The stored images of all included patients were reviewed by two radiologists (AGB, DS) in duplicate to perform all measurements. The reviewers worked in consensus and were blinded to the patient group. The reviewers were free to use any windows setting and both standard and lung kernel reconstructions. Multiplanar reformatting was available at a separate workstation (Vitrea; Vital Images, Minnetonka, Minnesota, USA). 2.3. Thrombus distribution The detailed Mastora score [11] was used in the present study. The obstruction of the mediastinal, lobar, and segmental arteries is quantified by a percentage score, and the sum of the percentages gives the global obstruction with a maximum of 300%. The use of the Mastora score allows for detailed analysis of thrombus distribution. The number of affected vessels was calculated. It included the number of vessels with thrombi,irrespective of vessel size. The
Fig. 1. Axial slice of a computed tomography pulmonary angiography at the level of the pulmonary trunk. The typical position of measurements is shown: diameter of the pulmonary trunk (A), diameter of the ascending aorta (B), diameter of the descending aorta (C) and the diameter of the superior vena cava (D).
central obstruction included the obstruction of the pulmonary trunk, the main pulmonary arteries, and the interlobar arteries (maximum 100%). The lobar obstruction and the segmental obstruction included the obstruction of the lobar and segmental arteries, respectively. The upper and lower lobe obstruction included the obstruction of the upper and lower lobar and segmental arteries, respectively. The middle lobe obstruction included the obstruction of the interlobar arteries, middle lobe arteries, and middle segmental arteries. 2.4. Morphometric parameters of right ventricular dysfunction The ratio of the short axes of the right and left ventricle was measured, as described before [12]. The pulmonary trunk was measured on an axial slice on which it showed its maximal diameter. On the same slice the diameter of the ascending aorta, the diameter of the descending aorta and the diameter of the superior vena cava were measured (Fig. 1). The distance from wall to wall was regarded as diameter. Thrombi were ignored when measuring diameter. The diameter of the right pulmonary artery was measured at the position shown in Fig. 2. The diameter of the IVC was measured on axial slices between the liver and the heart. In addition, the interventricular septal angle was measured on axial slices as the angle between the interventricular septum and the line joining the midpoint of the sternum to the thoracic vertebral spinous process [13]. 2.5. Contrast reflux in inferior vena cava In 17 patients contrast was applied via a femoral catheters. These patients were excluded from the analysis of reflux. According to the scale established by Groves [14] and simplified by Aviram [16] reflux was quantified in a 3-point scale: no reflux in IVC (Table 1A and B), subcardial reflux in IVC (Table 1C and D), intrahepatic reflux in IVC (Table 1E and F). To quantify the reflux, measurements of the density in the IVC were performed at subcardial level and at the level of the liver. The area from which the mean density in Hounsfield units was computed was at least 100 mm2 for each measurement. In addition, the craniocaudal length of reflux in the IVC was measured on coronal slices. Measurement was performed from the
334
A.G. Bach et al. / European Journal of Radiology 84 (2015) 332–337
Table 1 Examples of contrast reflux into the inferior vena cava according to a 3-point scale (computed tomography pulmonary angiography examinations in different patients).
base of the right atrium to the inferior margin of contrasted vessel volume (Fig. 3). When reflux was discontinuous (as shown in Fig. 3), still the inferior margin of contrast in the IVC was accepted. When the patient has more reflux and a slower pulmonary circulation the IVC gets progressively brighter while the aorta gets progressively darker. To quantify this effect, the ratio of the IVC (HU) density to the descending aorta (HU) density was calculated. Furthermore, it was recorded whether contrast was applied via central or peripheral venous catheter. This was done to estimate potential bias caused by different injection sites. Another potential bias is the presence of comorbidities that have impact on right heart function. Therefore, information on the presence of the following conditions were collected from patient records: heart insufficiency NYHA IV, pulmonary hypertension > 55 mm Hg, chronic lung disease, and organ insufficiency. Organ insufficiency was defined like in the Acute Physiology and Chronic Health Evaluation II (APACHE II) [19].
Fig. 2. Axial slice of a computed tomography pulmonary angiography at the level of the pulmonary trunk. The typical position of measurement of the diameter of the right pulmonary artery is shown (A).
2.6. Statistical analysis Collected data were evaluated by means of descriptive statistics: absolute and relative frequencies, median and 25–75% interquartile range. Sensitivity, specificity, negative predictive value, and positive predictive value were given to describe the prognostic power. The two-tailed Mann–Whitney U test was used to estimate statistical differences between groups for metric variables. The Chi-square-test was used to estimate statistical differences between groups for binary variables (SPSS Version 18, IBM, New York). The suggested cut-off points for measurements to predict 30-day mortality were selected so that the Youden index (sensitivity + specificity − 1) had a maximum value. A binary logistic multiple regression analysis for 30 days mortality was performed with comorbidities and reflux measurement as covariates. A pvalue of p < 0.05 was considered significant.
Fig. 3. Coronal slice of a computed tomography pulmonary angiography. The black line indicates the craniocaudal length of contrast agent reflux in the inferior vena cava. The cranial limit of the measurement is the right atrium. The caudal limit of the measurement is the inferior margin of the contrast column in the inferior vena cava.
A.G. Bach et al. / European Journal of Radiology 84 (2015) 332–337
335
Table 2 Thrombus distribution in survivors and non-survivors. Survivor* (n = 326)
Non-survivor* (n = 39)
p-Value
Thrombus mass Global obstruction (Mastora score) [%] Number of affected vessels [count]
30 (7–67) 10 (3–18)
19 (5–89) 7 (3–18)
0.810 0.447
Thrombus distribution by level Central obstruction [%] Lobar obstruction [%] Segmental obstruction [%]
0 (0–8) 13 (3–33) 13 (4–25)
0 (0–16) 10 (0–35) 7 (4–31)
0.238 0.846 0.954
Thrombus distribution by lobe Upper lobe obstruction [%] Middle lobe obstruction [%] Lower lobe obstruction [%]
10 (0–28) 7 (0–27) 17 (5–32)
7 (2–31) 7 (0–29) 11 (2–28)
0.962 0.668 0.676
*
Median (25–75% interquartile range).
Table 3 Morphometric parameters of right ventricular dysfunction in survivors and non-survivors. Measurements are sorted in order of ascending p-value.
Diameter of pulmonary trunk [mm] Ratio short axis right ventricle/left ventricle Short axis of the left ventricle [mm] Interventricular septal angle [◦ ] Short axis of the right ventricle [mm] Diameter of superior vena cava [mm] Diameter of ascending aorta [mm] Diameter of descending aorta [mm] Diameter of IVC [mm] Diameter of right pulmonary artery [mm]
Survivor* (n = 326)
Non-survivor* (n = 39)
p-Value
28.7 (26.3–32.2) 1.1 (0.9–1.4) 42 (37–48) 44 (39–52) 48 (42–54) 23 (21–26) 33 (30–36) 25 (23–27) 25 (25–31) 24 (21–27)
30.6 (27.8–32.4) 1.1 (1.0–1.7) 39 (32–48) 48 (38–56) 50 (43–56) 22 (19–25) 33 (30–36) 24 (22–26) 25 (25–31) 24 (22–27)
0.016 0.078 0.083 0.102 0.203 0.248 0.747 0.750 0.907 0.922
Abbreviation: IVC inferior vena cava. * Median (25–75% interquartile range).
3. Results
morphometric parameters showed no significant difference between groups.
The 30-day mortality was 10.7%. In the study period, 14 patients with acute PE died from respiratory failure and 25 patients died from cardiopulmonary arrest. 3.1. Thrombus distribution The thrombus distribution was calculated for each patient. The median of the results for the group of survivors and non-survivors is shown in Table 2. The thrombus mass was not significantly different between survivors (30%) and non-survivors (19%). In addition, there was also no significant difference in thrombus distribution by level or by lobe. 3.2. Morphometric parameters of right ventricular dysfunction Measurements were performed for each patient. The median of the results for the group of survivors and non-survivors is shown in Table 3. The diameter of the pulmonary trunk was significantly different between survivors and non-survivors. All other
3.3. Contrast reflux in inferior vena cava Measurements of contrast reflux were performed for each patient. The median of the results for the group of survivors nonsurvivors is shown in Table 4. All measurements of reflux were significantly different in the groups of survivors and non-survivors. The highest significance was found in the measurement of the craniocaudal length of reflux in IVC: median in survivors 7 mm versus median in non-survivors 33 mm. The optimal cut off to predict 30-day mortality according to the Youden index, as well as sensitivity, specificity, positive and negative predictive value were calculated for the measurements of contrast reflux that showed the most significant differences between survivors and non-survivors (Table 5). When the craniocaudal length of reflux in IVC was ≥31 mm about 28% of patients with acute PE died within 30 days. About
Table 4 Quantification of contrast reflux in inferior vena cava in survivors and non-survivors. Measurements are sorted in order of ascending p-value.
Craniocaudal length of reflux in IVC [mm] Contrast of IVC intrahepatic [HU] IVC reflux III (intrahepatic) [yes] Contrast of IVC subcardial [HU] Ratio contrast IVC/aorta descendes IVC reflux II–III (subcardial or intrahepatic) [yes] IVCI reflux I (no reflux) [yes]
Survivor* (n = 310)
Non-survivor* (n = 38)
p-Value
7 (0–23) 44 (40–99) 30% (93) 106 (45–187) 0.3 (0.2–0.6) 59% (183) 41% (127)
33 (12–57) 95 (55–192) 58% (22) 170 (108–240) 0.6 (0.2–1.8) 82% (31) 18% (7)
