|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department for Intensive Internal Medicine, General Hospital Celje, Oblakova, Slovenia.
Correspondence to: Matej Podbregar, MD MSc, Department for Intensive Internal Medicine, General Hospital Celje, Oblakova 5, 3000 Celje, Slovenia; e-mail: Matej.Podbregar{at}guest.arnes.si
| Abstract |
|---|
|
|
|---|
Design: A prospective registry of consecutive patients.
Setting: An 11-bed closed medical ICU at a 860-bed community general hospital
Patients: Forty-seven patients with shock or hypotension due to MPE and central pulmonary thromboemboli detected by transesophageal echocardiography who were treated with thrombolysis between January 1994 and April 2000.
Procedures: Patients were divided into two groups according to the following characteristics of the detected thromboemboli: group 1, thrombi with one or more long, mobile parts; and group 2, immobile thrombi. Right heart catheterization was performed.
Results: The incidence of both types of thromboemboli was comparable. Groups 1 and 2 showed no differences in demographic data, risk factors for pulmonary embolism, length of preceding clinical symptoms, percentage of patients in shock, hemodynamic variables, serum lactate levels on hospital admission, and treatment. Seven fatal cases due to obstructive shock and right heart failure were present in group 2, but none were present in group 1 (7 of 23 patients vs 0 of 24 patients, respectively; p < 0.05). At 12 h, the cardiac index was lower in group 2 than in group 1 (2.6 ± 1.0 vs 3.1 ± 0.9 L/min/m2, respectively; p < 0.05), and the central venous pressure (15.0 ± 6.2 vs 12.5 ± 3.7 mm Hg, respectively; p < 0.05) and total pulmonary resistance (12.9 ± 5.9 vs 8.6 ± 2.7 mm Hg/L/min/m2, respectively; p < 0.001) were higher in group 2 compared to group 1. On hospital admission, inclusion in group 2 (p < 0.03; hazard ratio, 9.53; 95% confidence interval [CI], 1.19 to 76.47) and preexisting chronic medical or neurologic disease (p < 0.01; hazard ratio, 16.4; 95% CI, 1.97 to 136.3) were independent predictors of 30-day mortality.
Conclusion: On hospital admission, morphology of the thromboemboli and the presence of pre-existing chronic medical or neurologic disease are independent predictors of 30-day mortality. Patients with immobile central pulmonary thromboemboli have a worse short-term outcome than those with mobile central pulmonary thromboemboli.
Key Words: echocardiography embolism prognosis thrombus
| Introduction |
|---|
|
|
|---|
The composition and morphology of venous thrombi change with the age of the patient.3 4 Blood clots tend to retract shortly after being formed. As the fibrin strands are pulled closer together, serum is expelled from the clot. With retraction or mechanical compression, clots lose > 90% of the initial plasminogen content and become resistant to thrombolytic treatment.5
The following two types of central pulmonary thromboemboli can be detected: type A, highly mobile thromboemboli; and type B, immobile thromboemboli. Type A thromboemboli are more prevalent in a subgroup of patients who have experienced a first embolic event.6 However, the practical importance of thromboembolic diversity remains unclear. The aim of our study was to determine whether the type of central thromboemboli detected by TEE in patients presented with shock or hypotension as a result of MPE had an influence on short-term outcome.
| Materials and Methods |
|---|
|
|
|---|
TEE
In our ICU, TEE is available around the clock and is the first diagnostic procedure used in all patients with high central venous pressure (CVP) and concomitant hypotension or shock. In this study, TEE was performed after an initial assessment (ie, clinical history and physical examination) was performed within 15 min of admission into the ICU. We employed a 5-MHz monoplane probe (Sonos 1000; Hewlett-Packard; Andover, MA) in 20 patients and a 5-MHz multiplane probe (MPT74, HDI 3000; ATL; Bothell, WA) in 27 patients. The detailed methodology was described previously.2
Each TEE was recorded by a video recorder, and the tape was first analyzed at the bedside by an intensive care physician. Within 24 h after ICU admission, the tapes were analyzed off-line by the intensive care physician who was trained in echocardiography (M.P. or B.K.) and was blinded to the bedside interpretation of the morphology of the thrombi, treatment, and outcome. If there was disagreement between the bedside and off-line interpretations of the morphology of thrombi, another experienced intensive care physician (G.V.) who also was trained in echocardiography, and was not involved in the study data analysis, was consulted for independent review. Patients were divided into the following two groups according to the characteristics of the detected central pulmonary thromboemboli after agreement by at least two physicians: group 1, mobile thrombus or thrombi with one or more long, highly mobile parts; and group 2, immobile thrombus or thrombi. If both types of thrombi were present, the patient was classified as being in group 1.
Hemodynamic Measurements
Systemic arterial pressure and heart rate were noninvasively monitored (SC 960/961 or SC 6000; Siemens AG; Erlangen, Germany). Right heart catheterization was performed in all patients. Thermodilution and continuous cardiac output monitoring catheters (Swan Ganz catheter; Baxter HealthCare; Deerfield, IL) were used. Hemodynamic measurements included CVP, mean pulmonary artery pressure (MPAP), and cardiac output. Cardiac index was calculated as the cardiac output divided by body surface area. Total pulmonary resistance (TPR) was calculated as MPAP divided by the cardiac index. Arterial lactate level was measured using colorimetric enzyme method (Chronolab; Hitachi; San Jose, CA).2
Definition of Patient Subgroups Based on Clinical Presentation
Hypotension was defined as a systemic systolic arterial pressure of < 90 mm Hg or a pressure drop of at least 40 mm Hg for a time period of > 15 min without clinical signs of cardiogenic shock or need for catecholamine support. Cardiogenic shock was defined as systolic arterial pressure of < 90 mm Hg and a cardiac index < 2.1 L/min/m2 with elevated serum lactate of > 3.0 mmol/L.
Treatment
The amount of infused fluids, the inotropic support of circulation (ie, norepinephrine therapy [Arterenol; Hoechst AG; Frankfurt, Germany] and dobutamine therapy [Dobutrex; Eli Lilly; Indianapolis, IN]), mechanical ventilation, and the type of thrombolytic agent were not predefined. In all patients, an IV bolus of heparin, 100 IU/kg, was initially administered. The study did not influence the treatment chosen by the treating physician. The following three thrombolytic agents were used: streptokinase (SK) [Boehringer; Ridgefield, CT], 1.5 million-IU infusion over > 2 h; urokinase (UK) [Ukidan; Serono; Rome, Italy], 1 million-IU infusion over 15 min, followed by a 1 million-IU infusion within 2 h; and recombinant human tissue plasminogen activator (rt-PA) [alteplase; Actilyse; Boehringer; Ingelheim, Germany], 10-mg IV bolus, followed by a 90-mg infusion over > 2 h. IV heparin infusion was started concomitantly with that of rt-PA and at the end of SK or UK infusion, which was adapted to achieve an activated partial thromboplastin time ratio of two to three times the basal value. Coumarin therapy was started on the third day if there were no contraindications to maintain the international normalized ratio at between 2 and 3.
Assessment of Risk Factors
In addition to immobilization due to a recent (ie, < 30 days before the occurrence of MPE) surgical procedure or trauma, preexisting chronic disease (medical or neurologic disease) and malignancy were considered to be risk factors for pulmonary embolism (PE).
Confirmation of PE
In 44 patients, PE was confirmed by a high-probability lung ventilation-perfusion scan.7
In three patients, autopsies showed PE.
End Points of the Study
Mortality in the ICU (from acute MPE) and at 30 days were defined as the end points of the study. Causes of death were classified as related or unrelated to PE. Autopsies were performed in all patients who died.
Statistical Analysis
Data were expressed as the mean ± SD or median and range. We used the Student t test for the comparison of normally distributed data and the
2 test for the comparison of noncontiguous variables. A nonparametric Kruskal-Wallis test was used for the comparison of not normally distributed data. The Cox proportional hazards model was used to assess the association of variables with survival. (Hazard ratio and 95% confidence interval [CI] for risk factors as well as p values for
2 according to the likelihood ratio test are given.) The hazard ratio of continuous variables refers to the risk ratio per unit of the analyzed variables. A statistical program (Statistica for Windows, version 5.0; StatSoft Inc; Tulsa, OK) was employed in the data analysis. A p value < 0.05 was considered to be statistically significant.
| Results |
|---|
|
|
|---|
|
|
|
|
At 12 h, the cardiac index was lower in group 2 compared to group 1 (2.6 ± 1.0 vs 3.1 ± 0.9 L/min/m2, respectively; p < 0.05), and the CVP (15.0 ± 6.2 vs 12.5 ± 3.7 mm Hg, respectively; p < 0.01) and TPR (12.9 ± 5.9 vs 8.6 ± 2.7 mm Hg/L/min/m2, respectively; p < 0.05) were higher in group 2 than in group 1.
Seven fatal cases due to obstructive shock and right heart failure occurred in group 2, but no deaths directly related to MPE were reported in group 1 (7 of 23 patients vs 0 of 24 patients; p < 0.05) [Table 4 ]. There were two additional deaths that were not related to PE in a 30-day period, one in each group. The detailed presentation of time, cause of death, preexisting chronic disease, and type of thrombolysis used are shown in Table 4 .
|
|
|
We did not detect any serious major complications due to right heart catheterization and concomitant thrombolysis.
| Discussion |
|---|
|
|
|---|
The detected thrombi differed not only in their mobility, but also in their echogenicity. Mobile (type A) thromboemboli were hypoechoic and more heterogeneous compared to the hyperechoic and homogenous structure of nonmobile (type B) thromboemboli. It is speculated that the type A thrombus, which originates from acute deep venous thrombosis, could be in an earlier phase of organization and could be more susceptible to thrombolysis than the type B thrombus, which shows partial organization. In previous case reports,8 the higher resistance of type B thrombi to thrombolytic treatment was noticed. Accordingly in patients with deep venous thrombosis, the echo density of the thrombi seen on ultrasound examination correlates with the course of organization and the efficacy of later thrombolysis.4 9
It is also possible that a higher incidence of fatal obstructive shock in patients with type B thrombi was related to different potencies of thrombolytic agents. In our study, rt-PA was used in almost half of patients who died due to MPEs (three of seven patients). Several other studies of thrombolysis in patients with MPEs did not note the differences in 12-h hemodynamic response among rt-PA, SK, and UK. However, the studied populations included a small number of patients with obstructive shock.10 11 12 13 Yet, rt-PA was shown to produce the most rapid lysis in the first 2 h after thrombolytic treatment, which may be crucial in hemodynamically unstable patients.14
The transport of a thrombolytic agent into thrombi represents a rate-limiting step in thrombolytic therapy.15 It was also previously shown that pulmonary thrombolysis predominantly occurred in partially obstructed, rather than totally obstructed, vascular units.16 If there was a total occlusion of a pulmonary vessel with type B thromboemboli, the amount of the delivered thrombolytic agent was probably not sufficient for effective thrombolysis compared to type A thromboemboli, in which the moving parts of thromboemboli indicated flow around the thrombus.
All patients who died had preexisting chronic disease, indicating the importance of the simultaneous existence of type B thrombi and preexisting chronic disease. In these patients, many episodes of acute PE tend to go undetected.17 Despite the fact that in our study the duration of preceding clinical symptoms associated with PE was not associated with any thrombus type, it is speculated that type B thromboemboli were lodged in pulmonary arteries at least few days before the MPE presentation, if the time scale of organization is considered to be the same as that in patients with deep venous thrombosis, in which the stabilization period of the thrombus ranges between 2 and 33 days (average, 10 days).18 This process is slower in debilitated patients.19
In our study, pulmonary artery catheterization was performed to study whether the type of central thromboemboli had an influence on thrombolytic efficiency and to optimize volume, inotropic therapy, and vasopressor therapy. The pulmonary artery catheter can simultaneously be used for local thrombolytic therapy, which is thought to be accompanied by more rapid and/or more complete clot lysis.20 However, a randomized study comparing intrapulmonary or systemic rt-PA in a dose of 50 mg administered over > 2 h showed no difference in improvement in pulmonary artery pressure and pulmonary perfusion.21 The drawback of local thrombolysis is also the need to perform pulmonary artery catheterization, which increases the bleeding from vascular access sites. In our study, there were no serious bleeding complications such as hemothorax or hemorrhagic shock.
In stable patients with COPD, TEE often reveals central pulmonary artery lesions, even in the absence of a previous PE and significant pulmonary hypertension. Most of the lesions are wall-adherent, hyperechoic, located in the proximal third of the right pulmonary artery, and might be misinterpreted as central pulmonary thromboemboli.22 However, the lesions show morphologic similarities to atherosclerotic lesions of the aorta and may actually represent central pulmonary atherosclerosis.23
The short-term mortality rate (ie, death due to MPE) in our study (14.9%) compares well to those found in other studies (mortality rate range, 15 to 32%).24 25 The low mortality rate cannot be explained only by the type of thrombolytic agent used, or by the use of fluid loading or inotropic support.26 27 It is our opinion that the 24-h availability of bedside TEE (sensitivity and specificity for detecting central thromboemboli in MPE in patients with right heart dilatation and shock, 92% and 100%, respectively) allows immediate treatment and influences the outcome.2 28 We have not found any study comparing short-term mortality rates and the time needed to diagnose MPE and to initiate management using different diagnostic procedures in patients with suspected MPE who have shock or hypotension.
Our findings suggest that patients with type B thromboemboli or preexisting chronic disease have worse short-term outcomes compared to patients with type A thromboemboli or patients without preexisting chronic disease. The former patients should be monitored closely to detect prolonged right heart dysfunction and should be considered for additional therapeutic measures.
Study Limitations
Our study has at least three major limitations. First, we are unable to draw conclusions about the efficiency of thrombolysis, because thrombolytics were not randomly assigned to one group that was identical to the regimen in a control group. Second, the supportive treatment also was not predefined. Since the treatment was left to the discretion of the attending physician, selection bias was unavoidable. Third, no ex vivo analysis was performed with which to detect the difference in resistance to thrombolysis among different types of thromboemboli.
| Conclusion |
|---|
|
|
|---|
| Acknowledgements |
|---|
nik, Rafael Skale, and Lucija Gabr
ek for their help in the planning of the study. We would also like to thank Drs. Edvard Kralj and Andreja Podbregar-Mar
for their critical review of the manuscript. | Footnotes |
|---|
Received for publication October 3, 2001. Accepted for publication February 13, 2002.
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |