| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, University of California, 200 W. Arbor Dr., San
Diego, CA 92103.
2
Present address: Department of Anatomy with Radiology, University of Auckland,
Park Rd., Auckland, New Zealand.
3
General Clinical Research Center, University of California, San Diego, CA
92103.
4
Department of Medicine, University of California, San Diego, CA 92103.
Received April 19, 1999;
accepted after revision July 1, 1999.
Address correspondence to C. J. Bergin.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Thirty-nine consecutive patients who underwent pulmonary angiography and thromboendarterectomy during 1995 and 1996 were included. Thirty-four underwent helical CT angiography. Measurements of postoperative pulmonary vascular resistance were compared with preoperative imaging features and preoperative pulmonary vascular resistance.
RESULTS. The best imaging indicators of a relatively high postoperative pulmonary vascular resistance were the extent of small vessel disease identified on CT angiograms as segments with abnormal perfusion but normal segmental arteries (p = 0.005) and the extent of central disease (p = 0.015). Combined with preoperative pulmonary vascular resistance, these features had a strong correlation with postoperative outcome (p = 0.0005). Segmental arterial disease seen on both conventional angiography and CT angiography correlated poorly with surgical outcome.
CONCLUSION. In patients with chronic thromboembolic pulmonary hypertension, CT angiographic evidence of extensive central vessel disease and limited small vessel involvement indicates a favorable surgical outcome.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although the extent of chronic thromboembolism traditionally has been revealed using conventional pulmonary angiograms, helical CT angiography recently has been used to delineate central and segmental disease accurately and noninvasively [4, 5, 6, 7]. Central disease may be depicted more accurately by using CT angiography rather than conventional angiography [5, 6]. The presence of thromboembolic material in central vessels has been recognized as a way to predict surgical accessibility [4], but the relationship between the extent and location of disease and surgical outcome has not yet been determined. Furthermore, clinicians attribute limited response in some patients to irreversible changes in subsegmental vessels that have not been detectable by any imaging technique to date. We examined preoperative imaging studies and cardiovascular responses to thromboendarterectomy in 39 patients in whom measurements by cardiac catheterization were taken before and after surgery to identify possible predictors of surgical success.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Measure of Outcome
Surgical outcome was determined by postoperative measurements of pulmonary
vascular resistance (PVR), defined by the following formula: PVR = (PAP - PCWP
/ CO) x 80 dynes·sec-1·cm-5. In this
formula, PAP = pulmonary artery pressure (mm Hg), PCWP = pulmonary capillary
wedge pressure (mm Hg), and CO = cardiac output (1/min).
The pulmonary artery pressure, capillary wedge pressure, and cardiac output
were determined by right heart catheterization. Patients were divided into
three groups on the basis of their postoperative pulmonary vascular resistance
values. Group I consisted of 22 patients in whom the postoperative pulmonary
vascular resistance was in the normal range (<200
dynes·sec-1·cm-5). This group was
considered to have the best surgical outcome. The remaining 17 patients all
had abnormal postoperative pulmonary vascular resistance measurements (
200
dynes·sec-1·cm-5) and were arbitrarily
divided into two groups. Group II consisted of the eight patients who were
considered to have an intermediate surgical outcome (pulmonary vascular
resistance 200 and 
250
dynes·sec-1·cm-5). Group III consisted of
the nine patients with the highest postoperative pulmonary vascular resistance
who were considered to have the worst surgical outcome (>250
dynes·sec-1·cm-5).
Helical CT Angiography
CT angiography was performed on a Signa 1.5-T scanner (General Electric
Medical Systems, Milwaukee, WI) using 5-mm collimation with reconstruction at
both 2-mm and 5-mm intervals. One hundred twenty-five milliliters of ioversol
(Optiray 320; Mallincrodt Medical, St. Louis, MO) was administered IV using a
power injector (CT 9000; Liebel-Flarsheim, Cincinatti, OH). For the CT
angiograms obtained early in the course of this study,a pitch of 1 was used
with an injection rate of 3 ml/sec. Scan delays varied between 10 and 20 sec
depending on cardiac function and location and size of the IV line. During the
latter part of the study, the injection rate was increased to 3.5 or 4.0
ml/sec with scan delays of 20 sec, and the pitch was increased to 1.7. From
experience, we realized that these modified parameters provided improved
opacification of pulmonary arteries with the least opacification of segmental
veins. With these protocols, one or two breath-hold helices were required for
coverage from above the aortic arch to the level of the diaphragm. Data were
reconstructed using 180° linear interpolation with a standard kernel for
mediastinal images and a lung kernel for lung images. Scans were then
photographed at both mediastinal (level, 40 H; width, 400 H) and lung (level,
600 H; width, 1200 H) settings.
CT angiograms were interpreted independently by two radiologists, one an experienced chest radiologist and the other a CT imaging fellow, neither of whom knew the other's interpretations or the results of other imaging and catheterization procedures of patients in this study. Evidence of thromboembolic disease within both central and segmental arteries was evaluated.
Central arteries were defined as vessels proximal to the segmental branches and were divided into four portions. These portions included the right and left main pulmonary arteries proximal to the upper lobe branches and the right and left descending portions of the central arteries between the upper lobes and the segmental branches [5]. Distinction between interlobe and truncus basalis vessels was not made because on axial CT angiograms these vessels are difficult to accurately distinguish. Disease within central vessels was identified by the presence of abnormal tissue lining the arterial wall or by irregularity of the intimal surface (Fig. 1). Central disease was quantitated by adding the number of abnormal central portions in each patient up to a maximum score of 4. If thromboembolic material was identified only in the descending portions of the central vessels distal to the upper lobe branches, the score was 2. If both descending portions were abnormal in addition to one of the central vessels proximal to the upper lobe branch, as in Figure 1, the score was 3. If thromboembolic material was identified in both the right and left pulmonary arteries both proximal and distal to the upper lobe branches, the score was 4.
|
Segmental vessels were defined according to the standard Boyden classification system, which assigns 10 segments to the right lung and eight to the left [8]. Segmental vessels were considered to be abnormal if intraluminal irregularities such as thrombi or webs were identified on mediastinal windows or if vessels were abnormally truncated, tortuous, or diminutive compared with their accompanying bronchi on lung windows (Fig. 2A, 2B). Using these criteria, both observers independently evaluated each segment in all patients and judged the segments as normal or abnormal. "Uncertain" or "indeterminate" were not used as classifications in part because if IV contrast enhancement was suboptimal, the observer's judgement was based on size and tortuosity of vessels on lung windows. Segmental vessel disease was quantitated by adding the number of abnormal segmental arteries in each patient.
|
|
Finally, independent of central and segmental vessel assessment, perfusion was evaluated on images photographed at lung windows. The perfusion in each pulmonary segment was assessed as being normal or abnormal using previously published criteria [4, 5, 9]. Briefly, segments were considered to have normal perfusion if the lung parenchyma in those segments was homogeneous and of relatively high attenuation and to have abnormal perfusion if the parenchyma was inhomogeneous or relatively hypoatenuating. In general, we found that segments with abnormal perfusion were supplied by segmental vessels that were also abnormal in appearance. However, in some patients, we noted areas of mosaic oligemia in pulmonary segments supplied by segmental arteries that were normal in contour and contrast enhancement (Fig. 3A, 3B, 3C). We hypothesized that segments with abnormal perfusion supplied by normally enhancing segmental arteries may represent the elusive small vessel disease that is not likely to be cured by pulmonary thromboendarterectomy. We quantitated the extent of small vessel disease by adding the number of segments in each patient in which perfusion was abnormal but in which the segmental artery was normal in contour and contrast enhancement.
|
|
|
Pulmonary Angiography
Biplanar rapid sequence cut-film pulmonary angiography was performed in the
anteroposterior and lateral projections using standard techniques and separate
injections of 55-60 ml of iohexol (Omnipaque 300; Nycomed, Princeton, NJ) in
each of the right and left main pulmonary arteries.
The pulmonary angiograms were evaluated independently by one of the two experienced pulmonologists who interpret all angiograms performed on patients referred for potential thromboendarterectomy at this institution (approximately 200/year). Segmental vessels in both lungs, as defined in the Boyden classification system [8], were evaluated. Intimal irregularities, bandlike narrowing with or without poststenotic vessel dilatation, "pouch" defects, and abrupt truncation or termination of vessels were considered angiographic evidence of chronic thromboembolic disease [10]. Central vessel disease was assessed on CT angiograms only because CT angiography depicts disease within central arteries more accurately than conventional angiography [5].
Surgical Technique
Pulmonary thromboendarterectomies were performed with patients on cardiac
bypass using a standard published procedure
[11].
Statistical Analysis
Results were analyzed three ways: Spearman's rank correlation, logistic
regression, and the Jonckheere test for ordered alternatives
[12,
13]. Spearman's correlation
and the Jonckheere test are nonparametric statistical methods that were used
because of the apparent violation of normality assumptions required for
Pearson's correlation and analysis of variance. Because nonparametric methods
use the relative rank of measurements rather than their absolute magnitudes,
the median, rather than the mean, is the measure of central tendency used in
the data descriptions.
Using Spearman's rank correlation, measurements of postoperative pulmonary vascular resistance were compared with preoperative pulmonary vascular resistance measurements and with the extent of central, segmental, and small-vessel disease quantitated on preoperative imaging studies. No adjustments were made for multiple testing.
To identify the variables that most strongly related to measurements of postoperative pulmonary vascular resistance, a logistic regression model was developed. The variables that were candidates for the final model included preoperative pulmonary vascular resistance, CT angiographic quantification of central and segmental vessel disease, CT angiographic estimation of small-vessel disease, and the extent of segmental vessel disease estimated from conventional angiograms. For each separate model developed, only subjects for whom data for all variables had been obtained were analyzed. Thus, the five patients in whom CT angiography was not performed were not included in any of the models that included CT angiographic estimates. Interobserver agreement for CT angiogram assessment of central, segmental, and small-vessel disease was first tested using Kendall's rank correlation coefficient for concordance [12]. Because there was significant interobserver agreement for all three CT angiographic variables, the mean of the measurement for the two observers was used for each of these three variables. Each of the specified analyzed variables was then entered by itself in separate logistic models together with the preoperative pulmonary vascular resistance to adjust for potential influence. From these logistic regression analyses, only the variables with p values for the likelihood ratio test for effect of less than 0.15 were included for further analysis. The significance of the complete model was tested using the difference in the log likelihood for the model with and without the covariates [14]. In addition, likelihood ratio tests for the effects of each covariate were used to confirm significance. The final model was constructed when the covariates included in the model as well as the overall model were significant at the level of p =.05.
Finally, the three groups of patients with different postoperative pulmonary vascular resistance measurements were compared. For each group, the median values for central and small vessel disease were calculated. Variation in CT angiogram quantification of central disease and of small vessel disease among the three was then assessed by the Jonckheere test for ordered alternatives [12, 13]. The Jonckheere test is a nonparametric test for equality used to determine the significance of differences in selected variables in a prespecified direction among the three groups of postoperative pulmonary vascular resistance measurements. The selection of variables tested by Jonckheere analysis (CT angiographic quantification of central disease and of small vessel disease) was determined by the results of the logistic regression analysis. For completeness, the mean values and standard deviations for central and small vessel quantification were also calculated, although they were not used in the statistical comparison of groups.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The pulmonary vascular resistance decreased in all patients after thromboendarterectomy; however, in 44% (17/39) the pulmonary vascular resistance remained elevated postoperatively; and in nine of the 39 patients studied, the postoperative pulmonary vascular resistance remained higher than 250 dynes·sec-1·cm-5. Using logistic regression, the best prognostic variables were CT angiographic quantification of the extent of central disease (p = 0.015) and of the extent of small vessel involvement (p = 0.005) after adjusting for the influence of preoperative pulmonary vascular resistance. In combination, these three features strongly influenced the postoperative outcome (p = 0.0005). Conversely, patient age and the extent of segmental vessel disease determined by both conventional angiography and by CT angiography, after adjusting for the effect of preoperative pulmonary vascular resistance, did not significantly influence postoperative outcome (p > 0.15). Interobserver agreement for CT angiographic assessment of central, segmental, and small vessel disease is shown in Table 1.
|
The group with the best outcome—namely, the group with postoperative pulmonary vascular resistance in the normal range (<200 dynes·sec-1·cm-5)—had the most central disease and the least small vessel involvement identified on preoperative CT angiograms (Tables 2 and 3). The two groups with abnormally high postoperative pulmonary vascular resistance values had more segments with abnormal perfusion in the presence of normal segmental arteries and less central disease. Although the median amount of small vessel disease increased progressively from group I to group III (Fig. 4), the median amount of central vessel disease decreased from group I to group II but increased from group II to group III (Table 2). However, the amount of central disease was less in both groups with postoperative pulmonary vascular resistance measurements that remained abnormal (>200 dynes·sec-1·cm-5) than in the group with postoperative pulmonary vascular resistance measurements in the normal range. When scores for both observers were combined, the tests for small vessel disease and for central disease were statistically significant by Jonckheere analysis (p < 0.005 and p < 0.02, respectively). Note that, unlike the median values used in the statistical analysis, the mean values for central disease decreased progressively from group I to group III (Table 3). Like the median values, the mean values for small vessel disease increased progressively (Table 3).
|
|
|
Using Spearman's rank correlation, the one preoperative feature that correlated significantly with postoperative pulmonary vascular resistance was small vessel disease, with correlation coefficients of 0.39 (observer 1, p = 0.02) and 0.64 (observer 2, p = < 0.0001) (Table 4). The next largest correlation coefficient (in magnitude) was for central disease identified on CT angiograms by observer 2, but this coefficient and all other preoperative measurements and imaging features lacked significant correlation.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Surgical selection of patients for thromboendarterectomy at this institution is a complex process that begins with establishing the diagnosis of chronic thromboembolic pulmonary hypertension. The decision to refer a patient for thromboendarterectomy is based primarily on the presence of pulmonary artery hypertension at rest (pulmonary vascular resistance <300 dynes·sec-1·cm-5), surgical accessibility of chronic thromboembolic obstruction determined at the time of angiography or angioscopy, and patient acceptance of the morbid and mortal risks of the procedure. The normal pulmonary vascular resistance is less than 200 dynes·sec-1·cm-5 with mean pulmonary artery pressures less than 20 mm Hg. When measurements of pulmonary vascular resistance exceed 300 dynes·sec-1·cm-5, the predominant clinical symptom of dyspnea increases markedly. Patients are sometimes referred for surgery with resting pulmonary vascular resistances of less than 300 dynes·sec-1·cm-5 if a significant ventilatory impairment is present resulting from increased dead space ventilation or if significant elevation of pulmonary artery pressure occurs during exercise. Age or the presence of a comorbid condition (with the exception of severe underlying parenchymal lung disease) generally is not used to exclude patients from surgical referral. Patients are not referred for pulmonary thromboendarterectomy if, in the opinion of the evaluating physician, insufficient quantities of proximal, surgically accessible thromboembolic material are present to assure that the patient will be weaned successfully from cardiopulmonary bypass at the time of the procedure.
The goal of pulmonary thromboendarterectomy is to decrease the extent of pulmonary vascular obstruction, thereby reducing pulmonary artery pressure and improving cardiac function. Early in the experience with thromboendarterectomy at this institution, postoperative improvement in pulmonary hemodynamics was thought to correlate positively with the extent of surgically accessible thromboembolic obstruction and negatively with the extent of surgically inaccessible obstruction. With increased experience in the examination and surgical treatment of patients with chronic thromboembolic pulmonary hypertension, it has become apparent that an additional component contributes to the total pulmonary vascular resistance and to the potential functional and hemodynamic improvement that can be observed postoperatively: a secondary, small vessel pulmonary arteriopathy that results from prolonged elevation in pulmonary vascular pressure or flow, or from the effect of an unidentified mediator. Considering the large cross-sectional area of subsegmental vessels compared with that of lobar and segmental vessels, it is not surprising that such a distal arteriopathy could significantly affect pulmonary vascular resistance. The ability to partition the disparate sources of elevated pulmonary vascular resistance in the preoperative period is essential because only the central component of thromboembolic obstruction is amenable to surgical correction. The myointimal thickening of arteriolar walls caused by prolonged elevation of pulmonary artery pressures [15] is the component of the arteriopathy that is unlikely to respond to thromboendarterectomy. This type of small vessel disease is associated with failure to lower pulmonary vascular resistance and reduced right ventricular workload that often results in an inability to wean the patient from cardiopulmonary bypass at the time of thromboendarterectomy [3, 16]. If the patient survives the surgical procedure, the results are usually residual pulmonary artery hypertension and a poor long-term outcome.
Before the advent of helical CT, pulmonary angiography and pulmonary angioscopy were the standard techniques used to delineate the surgically accessible and surgically inaccessible components of thromboembolic obstruction. The evaluative process using conventional angiography and angioscopy is a subjective process that requires considerable experience to correlate the extent of angiographic and angioscopic obstruction with the degree of hemodynamic impairment. Even in the setting of extensive experience, however, patients are encountered in whom the feasibility of surgical intervention is uncertain; that is, patients with pulmonary hypertension caused by chronic thromboembolism in whom the elevation in pulmonary vascular resistance is disproportionate to the apparent angiographic or angioscopic findings, suggesting the presence of a distal arteriopathy. Two options exist for these patients: thromboendarterectomy with the potential for an unfavorable perioperative outcome and without guarantee of acceptable hemodynamic improvement, or referral for lung transplantation. Neither option is associated with a long- or short-term outcome as desirable as that associated with successful thromboendarterectomy. More objective methods quantifying the contribution of central disease to the total pulmonary vascular resistance previously have not been available, and quantification of distal arteriopathy was not possible before CT angiography.
Mosaic attenuation in the lung parenchyma characteristic of patients with chronic thromboembolic pulmonary hypertension has been described as areas of increased attenuation in which segmental arteries are normal or increased in size, mixed with areas of decreased attenuation in which segmental arteries are relatively small [17]. However, we have noted that segmental vessels are not always diminished in size in areas of decreased attenuation in patients with chronic thromboembolic pulmonary hypertension [9]. From this observation, we hypothesized that mosaic attenuation in the presence of segmental arteries that are normal or increased in size may represent small vessel disease or those areas of lung parenchyma in which abnormal perfusion is unlikely to be corrected by thromboendarterectomy. Our results suggest that extensive mosaic attenuation in the presence of normal segmental arteries on CT angiograms corresponds with a less favorable outcome after thromboendarterectomy. This feature, which we have named the "small vessel factor," may help to define the contribution of distal arteriopathy to total pulmonary vascular resistance. Our quantification method probably underestimated the true extent of small vessel involvement because small vessel disease likely contributes to all areas of abnormal perfusion and not only to those with normal-appearing segmental arteries. Furthermore, our protocol using 5-mm thick slices was tailored to acquire maximum coverage during a single breath-hold in this population of patients, many of whom are profoundly dyspneic and unable to hold their breath for a long time. Five-millimeter collimation may be suboptimal for evaluating mosaic perfusion and may have caused underestimation of the number of abnormal segments. Despite these limitations, the small vessel factor identified on CT angiograms with the scanning parameters we used correlated more closely with postoperative measurements of pulmonary vascular resistance than did preoperative pulmonary vascular resistance, and correlated better than any other conventional CT or CT angiographic feature.
As we suspected, we also found that more extensive central disease, as quantitated by CT angiography, corresponded with a more favorable surgical outcome. The importance of disease within central vessels that is accessible to the surgeon performing the thromboendarterectomy has been appreciated for some time [4] but not previously quantitated. Our results show that more central disease was seen in patients whose postoperative pulmonary vascular resistance values were within the normal range (group I) than in patients in whom postoperative pulmonary vascular resistance remained above normal (groups II and III) and the decreasing trend was significant by Jonckheere analysis, despite the increase in median values for central disease between groups II and III. One explanation is the larger number of patients in group I. Alternatively, with Jonckheere analysis we tested for a decreasing trend in central disease so that these results could be explained if the amount of central disease in group I patients was considerably greater than that in patients in groups II or III, but the amount of central disease was not substantially different between groups II and III patients.
Ours was a retrospective study that involved only patients with surgically proven chronic thromboembolic disease; therefore, we did not assess central disease or the small vessel factor in patients who were not considered surgical candidates. Such candidates are usually advised against surgery for a variety of reasons that include comorbid conditions and the estimated extent of small vessel disease relative to central disease based on all imaging techniques. A prospective study involving more patients with chronic thromboembolic pulmonary hypertension who are potential candidates for thromboendarterectomy will help to further evaluate the usefulness of CT angiography in determining the influence of disease distribution on surgical outcome.
The clinical choice of pulmonary vascular resistance as a measure of surgical outcome is based on a multiplicity of factors that contribute to the postoperative state, including cardiac output, pulmonary capillary wedge pressure, and pulmonary artery pressure. Ultimately, clinical outcome as measured by morbidity, survival, and long-term quality of life may be more meaningful, but such an assessment will require a larger cohort of patients followed up for a longer time.
The extent of central disease and of small vessel involvement are only two of the factors that contribute to the likely success of thromboendarterectomy, but they are quantifiable, and our results suggest that they directly influence postoperative outcome. Although these features alone may not preclude a patient from surgery, when used in conjunction with other clinical and imaging features, they provide a noninvasive objective estimate of disease distribution that may help clinicians decide the feasibility of surgery in candidates who are borderline for other reasons.
In conclusion, our preliminary experience suggests that with helical CT angiography we may now have a method for distinguishing thromboembolism that is amenable to thromboendarterectomy from that which is unlikely to respond to surgical intervention.
Acknowledgments
We thank Gale Hurley and Barbara Cantle for their help in the preparation
of this manuscript.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J.-F. Paul, A. Khallil, A. Sigal-Cinqualbre, F. Leroy-Ladurie, J. Cerrina, E. Fadel, and P. Dartevelle Findings on Submillimeter MDCT Are Predictive of Operability in Chronic Thromboembolic Pulmonary Hypertension Am. J. Roentgenol., April 1, 2007; 188(4): 1059 - 1062. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. B. Pitton, G. Kemmerich, S. Herber, E. Mayer, M. Thelen, and C. Duber Hemodynamic effects of monomeric nonionic contrast media in pulmonary angiography in chronic thromboembolic pulmonary hypertension. Am. J. Roentgenol., July 1, 2006; 187(1): 128 - 134. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Castaner, X. Gallardo, J. Rimola, Y. Pallardo, J. M. Mata, J. Perendreu, C. Martin, and D. Gil Congenital and acquired pulmonary artery anomalies in the adult: radiologic overview. RadioGraphics, March 1, 2006; 26(2): 349 - 371. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Heinrich, M. Uder, D. Tscholl, A. Grgic, B. Kramann, and H.-J. Schafers CT Scan Findings in Chronic Thromboembolic Pulmonary Hypertension: Predictors of Hemodynamic Improvement After Pulmonary Thromboendarterectomy Chest, May 1, 2005; 127(5): 1606 - 1613. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Tanabe, A. Kimura, S. Amano, O. Okada, Y. Kasahara, K. Tatsumi, M. Takahashi, H. Shibata, M. Yasunami, and T. Kuriyama Association of clinical features with HLA in chronic pulmonary thromboembolism Eur. Respir. J., January 1, 2005; 25(1): 131 - 138. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. McGoon, D. Gutterman, V. Steen, R. Barst, D. C. McCrory, T. A. Fortin, and J. E. Loyd Screening, Early Detection, and Diagnosis of Pulmonary Arterial Hypertension: ACCP Evidence-Based Clinical Practice Guidelines Chest, July 1, 2004; 126(1_suppl): 14S - 34S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Engelke, C. Schaefer-Prokop, E. Schirg, J. Freihorst, S. Grubnic, and M. Prokop High-Resolution CT and CT Angiography of Peripheral Pulmonary Vascular Disorders RadioGraphics, July 1, 2002; 22(4): 739 - 764. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. D'Armini, B. Cattadori, C. Monterosso, C. Klersy, V. Emmi, F. Piovella, G. Minzioni, and M. Vigano Pulmonary thromboendarterectomy in patients with chronic thromboembolic pulmonary hypertension: hemodynamic characteristics and changes Eur. J. Cardiothorac. Surg., December 1, 2000; 18(6): 696 - 702. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
