HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Engelke, C. Articles by Marten, K. Search for Related Content PubMed PubMed Citation Articles by Engelke, C. Articles by Marten, K. Social Bookmarking                      What's this?

Acute Pulmonary Embolism on MDCT of the Chest: Prediction of Cor Pulmonale and Short-Term Patient Survival from Morphologic Embolus Burden

¹ All authors: Department of Radiology, Klinikum rechts der Isar, Technical University Munich, Ismaningerstrasse 22, Munich 81675, Germany.

Received April 15, 2005; accepted after revision June 17, 2005.

 
Address correspondence to C. Engelke (cengelke{at}roe.med.tum.de).

Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
OBJECTIVE. To predict cor pulmonale and short-term outcome in patients with pulmonary embolism (PE), we retrospectively investigated three morphology-based MDCT systems for scoring pulmonary artery obstruction.

MATERIALS AND METHODS. Eighty-nine consecutive patients (51 men and 38 women; age range, 23-83 years; median, 63.3 years) with an MDCT diagnosis of acute PE were included in the study. Sixty-four patients had a coexisting malignancy. PE severity was assessed by two masked observers using three percentage arterial obstruction indexes: two severity scores adapted from conventional angiography (excluding and including arterial branch obstruction grading: scores A and B, respectively) and a CT-derived severity score (index C). Echocardiographic reports were reviewed for elevation of right ventricular pressure. Obstruction index results were analyzed for correlation with pulmonary artery pressures and for prediction of cor pulmonale and 30-day survival. Statistical analysis included kappa, analysis of variance, linear correlation, chi-square, and logistic regression tests.

RESULTS. Kappa values of 0.89, 0.82, and 0.78 were obtained for interobserver agreement on PE severity for indexes A, B, and C, respectively. PE severity was moderate but varied significantly between the scores (for index A: median, 25.0%; range, 6.3-100; for index B: median, 12.5%; range, 3.1-65.6; for index C: median, 7.1%; range, 0.65-65.8; p < 0.0001 [analysis of variance]). Index C correlated best with pulmonary artery pressures (r = 0.69; p < 0.0016) and the presence of cor pulmonale (p = 0.0051; odds ratio [OR], 1.20/percentage increase [95% confidence interval, 1.05-1.35]; for an index C cutoff of 21.3%: p = 0.0001; positive predictive value, 1; negative predictive value, 0.87). Eight patients died within 30 days after CT. The PE severity of indexes A and B was not associated with patient outcome (p > 0.05). With score C, PE severity was a significant predictor of early death (p = 0.018; OR, 1.03/percentage increase [95% confidence interval, 1.00-1.06]; for an index C cutoff of 21.3%: p = 0.018; overall OR, 6.77; positive predictive value, 0.24; negative predictive value, 0.96).

CONCLUSION. Mastora score was a significant predictor of cor pulmonale and short-term outcome and may therefore allow therapy and risk stratification in patients with acute PE.

Keywords: angiography • chest • embolism • lungs • MDCT

Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Since the introduction of MDCT, many reports, in addition to previous helical CT data, have shown that this technique is useful in the primary evaluation of patients with suspected pulmonary embolism (PE) and has a value equivalent to that of conventional angiography [1-10]. Consequently, in many centers MDCT is now the first diagnostic choice for these patients. However, most previous studies have focused on the performance of MDCT angiography in the diagnosis of PE, and evaluation of its ability to determine the degree of embolic pulmonary artery obstruction and its relevance for stratification of patient treatment and short-term outcome remain underinvestigated [11]. To date, only a few studies have addressed this issue, either by adapting the angiographic scoring system introduced by Miller and coworkers to CT requirements or by quantifying the pulmonary embolus burden using a dedicated CT-derived scoring system introduced by Mastora et al. [11-15]. Yet, only a minority of CT scans in these reports were obtained using MDCT technique, only one group investigated the correlation of PE severity to pulmonary artery pressures [15], and the correlation of the CT-scored morphologic embolus burden to short-term patient outcome remains unknown.

Therefore, one objective of our study was to compare the correlations of three CT quantification systems of embolic pulmonary artery obstruction to pulmonary artery pressures and the presence of cor pulmonale. Another objective was to determine which index was the strongest predictor of 30-day patient outcome with regard to preexisting impairment of cardiorespiratory reserve and to systemic anticoagulant therapy.

View larger version (65K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1A —80-year-old woman with clinically suspected pulmonary embolism and evidence of major pulmonary embolism involving left main and intermediate arteries, left lower lobe artery, and left segmental branches 3 and 8-10 (arrows). Miller I and II scores and Mastora score were 56.3%, 25%, and 21.3%, respectively, for reviewer 1 and 50%, 25%, and 24.5%, respectively, for reviewer 2. Cor pulmonale was evident on transthoracic echocardiography, and patient underwent systemic thrombolysis. However, she died on third day after diagnosis from cardiogenic shock.

View larger version (71K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1B —80-year-old woman with clinically suspected pulmonary embolism and evidence of major pulmonary embolism involving left main and intermediate arteries, left lower lobe artery, and left segmental branches 3 and 8-10 (arrows). Miller I and II scores and Mastora score were 56.3%, 25%, and 21.3%, respectively, for reviewer 1 and 50%, 25%, and 24.5%, respectively, for reviewer 2. Cor pulmonale was evident on transthoracic echocardiography, and patient underwent systemic thrombolysis. However, she died on third day after diagnosis from cardiogenic shock.

View larger version (72K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1C —80-year-old woman with clinically suspected pulmonary embolism and evidence of major pulmonary embolism involving left main and intermediate arteries, left lower lobe artery, and left segmental branches 3 and 8-10 (arrows). Miller I and II scores and Mastora score were 56.3%, 25%, and 21.3%, respectively, for reviewer 1 and 50%, 25%, and 24.5%, respectively, for reviewer 2. Cor pulmonale was evident on transthoracic echocardiography, and patient underwent systemic thrombolysis. However, she died on third day after diagnosis from cardiogenic shock.

View larger version (83K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1D —80-year-old woman with clinically suspected pulmonary embolism and evidence of major pulmonary embolism involving left main and intermediate arteries, left lower lobe artery, and left segmental branches 3 and 8-10 (arrows). Miller I and II scores and Mastora score were 56.3%, 25%, and 21.3%, respectively, for reviewer 1 and 50%, 25%, and 24.5%, respectively, for reviewer 2. Cor pulmonale was evident on transthoracic echocardiography, and patient underwent systemic thrombolysis. However, she died on third day after diagnosis from cardiogenic shock.

View larger version (81K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1E —80-year-old woman with clinically suspected pulmonary embolism and evidence of major pulmonary embolism involving left main and intermediate arteries, left lower lobe artery, and left segmental branches 3 and 8-10 (arrows). Miller I and II scores and Mastora score were 56.3%, 25%, and 21.3%, respectively, for reviewer 1 and 50%, 25%, and 24.5%, respectively, for reviewer 2. Cor pulmonale was evident on transthoracic echocardiography, and patient underwent systemic thrombolysis. However, she died on third day after diagnosis from cardiogenic shock.

Top Abstract Introduction Materials and Methods Results Discussion References

CT Acquisition and Review
The scanner used between November 2001 and May 2002 was a 4-MDCT system (Volume Zoom, Siemens Medical Solutions); a 16-MDCT scanner (Sensation 16, Siemens Medical Solutions) was used thereafter. The acquisition parameters were standardized. Patients with clinically suspected PE underwent pulmonary CT angiography, and patients without clinically suspected PE underwent CT angiography of the thoracic aorta, thin-collimation CT for esophageal tumor staging, and standard chest CT. CT was performed using 120 kV and 90-200 mAs. For the 4-MDCT scanner, 4 x 1.0 mm and 4 x 2.5 mm slice collimations were used for thin-collimation and standard chest CT, respectively, and for the 16-MDCT scanner, 16 x 0.75 mm slice collimation was used. The table feed was 5-15 mm per rotation. The reconstruction slice thickness was 0.7-1.25 mm for thin-collimation CT and 3-5 mm for other scans. Contrast bolus injection used 120 mL of a 300 mg I/mL concentration of iomeprol (Imeron 300, Altana) at flow rates of 3-5 mL/sec and was routinely followed by a 30-mL normal saline "chaser." For CT angiography scans, automatic bolus timing was used with effective delays of 12-25 sec. For other scans, the delays were 40-50 sec.

After retrospective confirmation of the PE diagnosis, the MDCT scans were reviewed in a blinded fashion by two independent chest radiologists with 5 and 10 years of experience in clinical chest CT. The reviews were performed at a dedicated workstation using interactive axial cine mode at individual window settings and multiplanar reformatting according to previously published standards [4, 16]. Each scan was assessed for the extent to which the main, lobar, segmental, and subsegmental arteries could be analyzed. Respiration- or pulsation-related movement artifacts were scored at three levels (main pulmonary artery bifurcation at the level of the carina, upper lobe segmental arteries at the level of the aortic arch, and lower lobe segmental arteries at the level of the left atrium) using a 3-point scale (0 = absent, 1 = mild, 2 = severe). Scans that could not be analyzed at the level of the segmental pulmonary arteries were excluded. Mean pulmonary artery density was measured at the same levels. The criterion for diagnosis of PE was the presence of partly to totally obliterating low-attenuation material within the pulmonary artery tree [4, 16] (Figs. 1A, 1B, 1C, 1D, and 1E).

CT Severity Assessments
The first PE severity score was modified for CT requirements from the original angiographic obstruction score of Miller et al. [12] and ranged from 0 to 16 as described by Bankier et al. [13]. This scoring system rates the presence of embolic material using a 2-point scale (0 = absent, 1 = present) within a total of 16 segmental arteries and is limited to the "objective" part of the original angiographic score [12]. Only nine segmental arteries are scored in the right lung and seven in the left [12, 13]. Emboli at a more proximal arterial level are given a score equal to the number of segmental branches arising distally. Subsegmental emboli are not scored. This PE severity score is hereinafter referred to as the "Miller I" score (Table 1). The second obstruction score was modified from the same angiographic score containing 16 segmental arteries as outlined by Qanadli et al. [11]. This score includes 20 segmental pulmonary artery branches. Again, emboli at a more proximal arterial level are given a score equal to the number of segmental branches arising distally. Each segmental score receives an additional weighting factor for the degree of luminal obstruction (0 = no embolic material, 1 = partial branch obstruction, 2 = branch occlusion). Subsegmental artery findings are assigned a value of 1 [11]. This severity score is hereinafter referred to as the "Miller II" score (Table 1). The third severity score was obtained as outlined by Mastora and coworkers [15]. This scoring system includes the five mediastinal, six lobar, and 20 segmental arteries, each scored for the degree of luminal obliteration on a scale from 0 to 5 (0 = 0%, 1 = 1-24%, 2 = 25-49%, 3 = 50-74%, 4 = 75-99%, 5 = 100%). The sum of mediastinal, lobar, and segmental artery scores leads to a global obstruction score with a maximum of 155. In our study, subsegmental emboli in the absence of proximal segmental embolic material were weighted with a factor of 0.5 and included in the segmental score. This severity score is hereinafter referred to as the "Mastora" score (Table 1). The percentage obliteration (obstruction index) of the pulmonary artery circulation for each score was calculated by dividing the observed CT severity score by the maximum obstruction score.

Assessment of Clinical and Transthoracic Echocardiography Data
Information about substantial coexisting cardiorespiratory morbidity (defined as New York Heart Association class 3 or partial pressure of carbon dioxide 42 mm Hg), systemic anticoagulant therapy, and short-term 30-day patient outcome was obtained from the clinical files, coroner's reports, and postmortem reports. Therapeutic anticoagulation was defined as IV heparin therapy (with a prothrombin time 60 sec), later replaced by oral warfarin (international normalized ratio 2.0). Prophylactic anticoagulation was defined as body weight-adapted subcutaneous therapy using various heparins with prothrombin times within the normal range. Systemic thrombolysis was defined as Goldhabert and coworkers [17] have defined it.

Two-dimensional transthoracic echocardiography was performed on a Sonos 5500 imager (Hewlett-Packard Systems), using 2.5-MHz transducers. All examinations were performed by experienced, dedicated echocardiographers. Data from transthoracic echocardiography were available for 41 patients. Echocardiographically severe acute PE was defined as the presence of signs of acute cor pulmonale, including paradoxic movement of the interventricular septum, hypokinesis of the free wall of the right ventricle, and systolic pulmonary hypertension (which was defined as a pressure greater than 30 mm Hg) as assessed by quantification of tricuspid regurgitation velocities corrected for central venous pressure [18].

Statistical Analysis
Statistical analysis was performed using spreadsheet-based statistical software (StatsDirect, release 2.3.8, CamCode). Interobserver agreement was tested by weighted kappa statistics [weighted by 1 - abs(i - j) / (1 - k)]. All other analyses were performed using the mean of the scores of the two radiologists for each obstruction index. The variance between the three severity indexes was assessed by one-way analysis of variance with Tukey pair comparisons [19]. PE severity was compared between those patients with and those without acute cor pulmonale, and between those patients censored alive and those who died early, using the Mann-Whitney U test. The correlation between severity scores and pulmonary artery pressures was determined using linear regression, and the correlation between PE severity scores and the occurrence of cor pulmonale or death was determined using multivariate backward stepwise logistic regression. This procedure selects the best predictors until all remaining variables of the tested model are significant. The odds ratios (ORs) in logistic regression pertain to a 1% increase in obstruction severity index. In all outcome analyses, multivariate stepwise and univariate logistic regressions were stratified for preexisting cardiorespiratory failure and weighted for the amount of anticoagulant therapy instituted. The function fits and analyzes conditional logistic models for binary outcome or response data with one or more predictors, where observations are not independent but are matched or grouped in some way. The regression is fitted by maximization of the natural logarithm of the conditional likelihood function using Newton-Raphson iteration [20]. The logits of the response data are fitted using an iteratively reweighted least-squares method to find maximum likelihood estimates for the parameters in the logistic model [21]. For the severity scores identified as significantly predicting the occurrence of cor pulmonale or death, univariate analysis was added using the chi-square test, with the PE severity threshold at the logistic regression mean representing the best cutoff for prediction of cor pulmonale and death. The chi-square test was supplemented by overall ORs (pertaining to occurrence of the event) and by positive predictive value and negative predictive value. A p value of less than 0.05 was considered to indicate statistical significance.

Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Patients
From a total of 2,320 consecutive scans, 95 patients (4.1%) were identified as having PE. The scan quality was considered insufficient for 49 patients, who were excluded from the study. The clinical data review was complete in 89 of the patients with PE, who constituted the final study population (51 men and 38 women; age range, 23-83 years; median, 63.3 years). In total, PE occurred most frequently in patients referred for pulmonary CT angiography and for thin-collimation CT evaluation of esophageal cancer (33 [22.9%] of the 144 patients referred for pulmonary CT angiography, three [3%] of the 100 patients referred for aortic CT angiography, 19 [6.2%] of the 306 patients referred for esophageal CT, and 34 [1.9%] of the 1,770 patients referred for other types of chest CT). Twenty-six of the 89 patients had evidence of coexisting chronic cardiorespiratory failure. Further concurrent morbidity included malignancy (n = 59); obesity (n = 13); stroke (n = 7); thrombophilia (n = 3); trauma, surgery, or immobilization 14 days before MDCT (n = 14, 23, and 14, respectively); and imaging evidence of deep venous thrombosis in 11 patients, who were treated with therapeutic anticoagulation (n = 4) or low-molecular-weight heparin (n = 2) at the time of MDCT. The five remaining patients with evidence of deep venous thrombosis had advanced malignant disease contraindicating anticoagulant therapy. The amount of anticoagulant therapy instituted after CT varied: therapeutic doses in 46 patients (four patients with acute cor pulmonale who received systemic thrombolysis and 42 who received nonthrombolytic anticoagulant therapy), prophylactic anticoagulant doses in 19 patients, and no treatment in 24 patients who had contraindications such as active malignancy or who were without clinically suspected PE and whose diagnoses were missed on the initial routine assessment (n = 42).

Scan Quality and PE Severity
Artifacts significantly increased in frequency and severity from the upper third to the lower third of the lung (average score, 0-0.61; SD, 0-0.5; analysis of variance, p < 0.0001; pairwise comparisons between all thirds, p < 0.01). The average median arterial density was 257.2 ± 118.33 H. Contrast enhancement was homogeneous and did not vary significantly between the upper, middle, and lower thirds of the pulmonary artery tree (analysis of variance, p = 0.3; Tukey multiple comparisons, p = 0.21-0.85). Arterial density was adequate for analysis of segmental pulmonary emboli in all 89 patients and for diagnosis of subsegmental emboli in 55 patients. The interobserver agreement of PE severity was good to very good and did not differ significantly between the indexes (for the Miller I, Miller II, and Mastora obstruction indexes: = 0.89, 0.82, and 0.78, respectively; 95% confidence interval [CI], 0.76-1.02, 0.69-0.95, and 0.65-0.91, respectively). The overall PE severity was moderate (medians of 25%, 12.5%, and 7.1% and ranges of 6.25-100%, 3.1-65.6%, and 0.65-65.8% for the Miller I, Miller II, and Mastora obstruction indexes, respectively) but varied significantly between the three indexes (analysis of variance, p < 0.0001; Tukey multiple comparisons, p < 0.0001-0.028).

Cor Pulmonale
Transthoracic echocardiography was performed on 41 of the 89 patients within 6 hr of the MDCT scan. Acute cor pulmonale was confirmed in 15 patients and correlated positively with early death (p = 0.003; OR, 1.17 per 1 mm Hg increase [95% CI, 1.05-1.30]). The pressure gradient medians across the tricuspid valve were 33 mm Hg (range, 30-45 mm Hg) and 25 mm Hg (range, 17-28 mm Hg) among patients with and without pulmonary hypertension, respectively. The PE severity of all three obstruction indexes was significantly higher in patients with pulmonary hypertension (p < 0.0001 for the Miller I, Miller II, and Mastora indexes; Table 2). Among patients with normal pulmonary artery pressures on transthoracic echocardiography (n = 26), the Mastora obstruction index was less than 21%, whereas these patients received relatively high Miller I and Miller II indexes of up to 68.7% and 46.8%, respectively. The pulmonary artery pressures increased significantly with Mastora index values above 21.3% and again increased significantly for those above 50% (mean pressures of 36.0 ± 6.4 mm Hg and 26.2 ± 6.4 mm Hg for Mastora indexes above and below 21.3%, respectively [p = 0.002]; mean pressures of 39.4 ± 6.8 mm Hg and 27.9 ± 6.0 mm Hg for Mastora indexes above and below 50%, respectively [p = 0.004]).

Linear regression between the tricuspid pressure gradients and PE severity was better for the Mastora obstruction index (r = 0.69; p = 0.0016; power for 5% significance, 89.5%) than for the Miller I and Miller II indexes (r = 0.60 and 0.58, respectively; p = 0.01 and 0.01, respectively; power for 5% significance, 75.84% and 71.33%, respectively). When the 41 patients were entered into multivariate weighted retrograde stepwise logistic regression, the Mastora scoring system showed a significant correlation with elevated tricuspid pressure gradients (Table 3), whereas the Miller I and Miller II indexes were dropped as nonsignificant from the model. Therefore, an elevated Mastora index was the strongest multivariate predictor for the occurrence of acute cor pulmonale (p = 0.005; OR = 1.20 per percentage score increase [95% CI, 1.05-1.35]) with the regression mean at 21.84%. This finding was confirmed on univariate analysis, which evidenced the Mastora index as a highly significant predictor of cor pulmonale (using an index value equal to or greater than 21.3%: p = 0.0001; positive predictive value, 1; negative predictive value, 0.87; Table 4).

Short-Term Outcome
Eight patients died within 30 days after PE was found on CT. PE was confirmed at autopsy in one patient and was determined clinically to have been the major cause of death in six, whereas intracranial hemorrhage was the major cause of death in the remaining patient. Seven of these patients had received anticoagulants in therapeutic doses, whereas anticoagulant therapy was withheld from one patient with cerebral metastatic disease from lung cancer. PE was more severe in patients with a fatal outcome, with a trend to statistical significance on a Mann-Whitney U test comparison of the Mastora obstruction index values (p = 0.054; Table 2). However, this test could not be stratified for coexisting morbidity or weighted for anticoagulant treatment. When entered into multivariate weighted retrograde stepwise logistic regression, the Mastora obstruction index showed a significant correlation with early death, whereas the other two PE obstruction indexes (Miller I and II) were again dropped as nonsignificant from the model (Table 5). Equally, on univariate weighted logistic regression, neither of the two Miller obstruction indexes was a significant predictor of early death (p > 0.05). Therefore, an elevated Mastora index was the strongest multivariate and univariate predictor of early death (p = 0.018; OR, 1.03 per percentage score increase [95% CI, 1.00-1.06]; Table 5). This finding again was confirmed on chi-square analysis using an index value equal to or greater than 21.3% (p = 0.018; overall OR, 6.77; positive predictive value, 0.24; negative predictive value, 0.96; Table 4). Similar results were obtained with the Mastora obstruction index when the analysis was restricted to those 41 patients with available transthoracic echocardiography data: On multivariate logistic regression analysis, the two Miller obstruction indexes were dropped as nonsignificant—the Mastora index was the only significant predictor of early death (p = 0.017; OR, 1.04 per percentage score increase [95% CI, 1.01-1.07])—whereas on chi-square analysis using a Mastora index value equal to or greater than 21.3%, a significant association with early death again was found (p = 0.0006). In this group, none of the patients who died early had a Mastora obstruction index less than 21.3%, and all with a fatal outcome showed evidence of acute cor pulmonale. As a result, acute cor pulmonale was strongly associated with early death (p = 0.003; OR, 1.17 per mm Hg increase [95% CI, 1.05-1.30]).

Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Quantification of pulmonary artery obstruction in the diagnosis of acute PE on MDCT has potential value for the indirect assessment of hemodynamic compromise [13-15]. By comparing three scoring systems of PE severity, one of which—the Mastora index—was designed specifically for CT, we were investigating reproducible tools that could be used to assess the degree of vascular obstruction, stratify the patient's risk of death, and determine whether treatment should be altered [11]. In addition, the application of thrombolytic therapy for PE [22, 23] may benefit from noninvasive CT monitoring of thrombolytic efficacy, which is inversely related to the reduction of the pulmonary embolus burden.

Bankier et al. [13] have used two angiographic indexes, the Walsh score [24] and the Miller index, to quantify the severity of embolic pulmonary artery obstruction on helical CT. The more commonly used of these two scores in helical CT is the Miller index, because it is relatively simple to extrapolate to CT requirements [12]. However, to be translatable to helical CT, this scoring system was simplified by eliminating the functional part of the original angiographic index. Thus, no information about residual perfusion of the lung was taken into account. In this system, the presence of nonobstructive clots in the main pulmonary arteries corresponds to a 100% obstruction, which is not necessarily in keeping with the clinical severity of PE. By contrast, differentiation between complete and partial obstruction caused by the proximal clot may add relevant information about residual pulmonary perfusion [11]. Therefore, Qanadli et al. [11] addressed this issue by using a separate semiquantitative score, which was included in the main obstruction index. In our study, the severity indexes that were derived from the Miller score (with the Miller I representing the score applied by Bankier and coworkers and the Miller II representing the score applied by Qanadli et al.) were compared with a CT-derived severity index developed by Mastora et al., to assess their relative values in the indirect detection of cor pulmonale and the prediction of short-term patient outcome [11-13, 15]. The Mastora score is the most advanced of the three severity indexes because it does not extrapolate embolus burden from the proximal pulmonary arteries to the pulmonary periphery. This difference is important, because a proximal vascular obstruction does not necessarily impair flow in the distal vasculature. Therefore, each central pulmonary artery branch is scored individually. Moreover, this index applies comparatively discriminative obstruction grading to individual arterial branches by use of a hemodynamically "sensible" 5-point scale [15].

Our data show that all three indexes are simple to use and reproducible, with good to excellent agreement between our two investigators. However, the three scores varied significantly. The mean percentage of vascular obstruction as calculated by the Mastora index was less than the mean percentages expressed by the Miller scores. This variation can be explained by differences in the design of the obstruction indexes. Compared with the Miller I index, the weighting systems used in the other two indexes reduced the percentages of obstruction, especially in proximal emboli, which rarely caused total obstruction. In fact, a proximal partially occlusive embolus with a weighting factor of 1 on the Miller II index or 50% obstruction on the Mastora index could be associated with more distal occlusive emboli that alter parenchymal perfusion [11]. Our agreement analysis confirmed that this additional information was highly reproducible—there was no significant decrease in reviewer confidence from the Miller I to the Miller II and Mastora obstruction indexes. By contrast, subjective evaluation of the residual perfusion of the lung as part of the original conventional angiographic Miller index is likely to increase interobserver variability, as was observed in a study comparing the Miller II score with the results of pulmonary angiography [11]. In our study, the Mastora CT obstruction index, being the most discriminative scoring system with respect to residual perfusion of the lung, was the strongest predictor of the presence of acute cor pulmonale and correlated best with mean pulmonary artery pressures. The positive predictive value indicated the presence of acute cor pulmonale in all patients with a score greater than or equal to 21.3%. Alternatively, the presence of cor pulmonale would be unlikely in patients with minor PE and a Mastora score lower than this value. These results are consistent with previously reported data on selective pulmonary angiography [24] and are in keeping with the data of Mastora et al. [15], who found pulmonary artery pressures greater than 30 mm Hg in patients with index values less than 30% and a significant increase in pressures when index values were greater than 50%. Likewise, our study showed a significant increase in pulmonary artery pressures in patients with Mastora index values greater than 21.3% and a further significant increase for values greater than 50%.

In acute PE, embolic obstruction of the pulmonary vascular tree is the most important factor for increased pulmonary vascular resistance, resulting in pulmonary hypertension with a potentially poor prognosis [25-27]. The significance of reflexive vasoconstriction, which accompanies mechanical obstruction, in the pathogenesis of pulmonary hypertension and its influence on patient prognosis remain unclear [11]. However, the hemodynamic profile may change with the presence or absence of preexisting cardiac or pulmonary disease, which, along with an acute increase in vascular resistance and the extent and duration of anticoagulant therapy, affects patient outcome [28]. Our results are in keeping with those of McIntyre and Sasahara [29], who observed improved correlation between the mean pulmonary artery pressure and the degree of morphologic obstruction in preselected patients without underlying cardiopulmonary disease, when compared with unselected patients [30]. When stratifying for the presence of impaired cardiorespiratory reserve and weighting the response for the amount of anticoagulant therapy, the Mastora index was a significant predictor (p = 0.017), with a 6.7-fold increased risk of early death for patients with an index value greater than or equal to 21.3%. In contrast, below this threshold was found a high probability of survival with adequate anticoagulation therapy (negative predictive value, 0.96).

Our study was limited by the fact that in only 41 patients was pulmonary artery pressure assessed echocardiographically. Only a few patients with clinical signs of severe PE requiring thrombolytic therapy were included, because CT was rarely performed on patients with this condition. Therefore, to improve knowledge about the relationship between coexisting cardiorespiratory morbidity, morphologic embolus burden, anticoagulant therapy, acute cor pulmonale, and early death, this CT obstruction index should be compared with echocardiography in a more homogeneous prospective patient cohort to weigh their relative merits [22, 31].

In conclusion, our data suggest that it is important to enhance the concept of "detection of PE" by the evaluation of the degree of pulmonary artery obstruction. The Mastora index is simple and reproducible, correlates strongly with the presence of cor pulmonale, and may allow the identification of patients with a significantly increased risk of death for stratification of anticoagulant therapy.

Acknowledgments
 
We thank Martin Riedel from the Department of Cardiology, Deutsches Herzzentrum München, and Peter Manstein from our center for their help and advice in the clinical file review.

References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Schultz DJ, Brasel KJ, Washington L, et al. Incidence of asymptomatic pulmonary embolism in moderately to severely injured patients. J Trauma 2004; 56:727 -731[Medline]
2. Baile EM, King GG, Muller NL, et al. Spiral computed tomography is comparable to angiography for the diagnosis of pulmonary embolism. Am J Respir Crit Care Med 2000;161 : 1010-1015[Abstract/Free Full Text]
3. Goldhaber SZ, Elliott CG. Acute pulmonary embolism. I. Epidemiology, pathophysiology, and diagnosis. Circulation 2003;108 : 2726-2729[Free Full Text]
4. Remy-Jardin M, Mastora I, Remy J. Pulmonary embolus imaging with multislice CT. Radiol Clin North Am 2003;41 : 507-519[CrossRef][Medline]
5. Quinn MF, Lundell CJ, Klotz TA, Finck EJ, Pentecost M, McGehee WG. Reliability of selective pulmonary angiography in the diagnosis of pulmonary embolism. Am J Radiol 1987;149 : 469-471[Abstract/Free Full Text]
6. Qanadli SD, El Hajjam M, Mesurolle B, et al. Prospective evaluation of dual-slice helical CT in pulmonary embolism: comparison with selective pulmonary angiography in 157 patients. Radiology2000; 217:447 -455[Abstract/Free Full Text]
7. Goodman LR, Curtin JJ, Mewissen MW, et al. Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. AJR 1995;164 : 1369-1374[Abstract/Free Full Text]
8. Remy-Jardin M, Remy J, Baghaie F, Fribourg M, Artaud D, Duhamel A. Clinical value of thin-collimation CT in the diagnostic workup of pulmonary embolism. AJR 2000;175 : 407-411[Abstract/Free Full Text]
9. Ghaye B, Szapiro D, Mastora I, et al. Peripheral pulmonary arteries: how far in the lung does multidetector-row spiral CT allow analysis? Radiology 2001;219 : 629-636[Abstract/Free Full Text]
10. van Rossum AB, Pattynama PM, Ton ER, et al. Pulmonary embolism: validation of spiral CT angiography in 149 patients. Radiology 1996;201 : 467-470[Abstract/Free Full Text]
11. Qanadli SD, Hajjam ME, Vieillard-Baron A, et al. New CT index to quantify arterial obstruction in pulmonary embolism: comparison with angiographic index of echocardiography. AJR2001; 176:1415 -1420[Abstract/Free Full Text]
12. Miller GA, Sutton GC, Kerr IH, Gibson RV, Honey M. Comparison of streptokinase and heparin in treatment of isolated acute massive pulmonary embolism. Br Med J 1971;2 : 681-684[Medline]
13. Bankier AA, Janata K, Fleischmann D, et al. Severity assessment of acute pulmonary embolism with spiral CT: evaluation of two modified angiographic scores and comparison with clinical data. J Thorac Imaging 1997; 12:150 -158[Medline]
14. Collomb D, Paramelle PJ, Calaque O, et al. Severity assessment of acute pulmonary embolism: evaluation using helical CT. Eur Radiol 2003; 13:1508 -1514[CrossRef][Medline]
15. Mastora I, Remy-Jardin M, Masson P, et al. Severity of acute pulmonary embolism: evaluation of a new spiral CT angiographic score in correlation with echocardiographic data. Eur Radiol2003; 13:29 -35[Medline]
16. Remy-Jardin M, Remy J, Cauvain O, Petyt L, Wannbroucq J, Beregi JP. Diagnosis of central pulmonary embolism with helical CT: role of two-dimensional multiplanar reformations. AJR1995; 165:1131 -1138[Abstract/Free Full Text]
17. Goldhabert SZ, Kessler CM, Heit J, et al. Recombinant tissue-type plasminogen activator versus a novel dosing regime of urokinase in acute pulmonary embolism: a randomized controlled multicenter trial. J Am Coll Cardiol 1992; 20:24 -30[Abstract]
18. Currie PJ, Seward JB, Chan KL, et al. Continuous wave Doppler determination of right ventricular pressure: a simultaneous Doppler-catheterization study in 127 patients. J Am Coll Cardiol 1985; 6:750 -756[Abstract]
19. Armitage P, Berry G, Matthews JNS. Statistical methods in medical research, 4th ed. Oxford, England: Blackwell Science,2002
20. Howard S. Remark on the paper by Cox DR (1972): regression methods and life tables. J R Stat Soc [Ser B]1972; 34:187 -220
21. Pregibon D. Logistic regression diagnostics. Ann Stat 1981; 9:705 -724
22. Goldhaber SZ. Contemporary pulmonary embolism thrombolysis. Int J Cardiol 1998;65 [suppl 1]:S91 -S93
23. Goldhaber SZ. Treatment of pulmonary thromboembolism. Intern Med 1999;38 : 620-625[Medline]
24. Walsh PN, Greenspan RH, Simon M, et al. An angiographic severity index for pulmonary embolism. Circulation1973; 47[suppl 2]:101 -107[Abstract/Free Full Text]
25. Kasper W, Meinertz T, Henkel B, et al. Echocardiographic findings in patients with proved pulmonary embolism. Am Heart J1986; 112:1284 -1290[CrossRef][Medline]
26. Tobise K, Tosaka S, Onodera S. Time course of pulmonary vascular response to an acutely repetitive pulmonary microembolism in dogs: an analysis using pulmonary vascular impedance. Jpn Circ J1992; 56:432 -440[Medline]
27. Elliott CG. Pulmonary physiology during pulmonary embolism. Chest 1992;101 [suppl]:163S -171S[Abstract/Free Full Text]
28. Come PC. Echocardiographic evaluation of pulmonary embolism and its response to therapeutic interventions. Chest1992; 101[suppl]:151S -162S[Abstract/Free Full Text]
29. McIntyre KM, Sasahara AA. The hemodynamic response to pulmonary embolism in patients without prior cardiopulmonary disease. Am J Cardiol 1971; 28:288 -294[CrossRef][Medline]
30. McIntyre KM, Sasahara AA. The ratio of pulmonary arterial pressure to pulmonary vascular obstruction: index of preembolic cardiopulmonary status. Chest 1977; 71:692 -697[Abstract/Free Full Text]
31. Wolfe MW, Lee RT, Feldstein ML, Parker JA, Come PC, Goldhaber SZ. Prognostic significance of right ventricular hypokinesis and perfusion lung scan defects in pulmonary embolism. Am Heart J1994; 127:1371 -1375[CrossRef][Medline]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				R. M. Subramaniam, J. Mandrekar, C. Chang, D. Blair, K. Gilbert, P. J. Peller, J. Sleigh, and N. Karalus Pulmonary Embolism Outcome: A Prospective Evaluation of CT Pulmonary Angiographic Clot Burden Score and ECG Score Am. J. Roentgenol., June 1, 2008; 190(6): 1599 - 1604. [Abstract] [Full Text] [PDF]

				M. Remy-Jardin, M. Pistolesi, L. R. Goodman, W. B. Gefter, A. Gottschalk, J. R. Mayo, and H. D. Sostman Management of Suspected Acute Pulmonary Embolism in the Era of CT Angiography: A Statement from the Fleischner Society Radiology, November 1, 2007; 245(2): 315 - 329. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Engelke, C. Articles by Marten, K. Search for Related Content PubMed PubMed Citation Articles by Engelke, C. Articles by Marten, K. Social Bookmarking                      What's this?