Acute Pulmonary Embolism: Correlation of CT ...

Chest Imaging • Original Rese arch

Downloaded from www.ajronline.org by 54.152.109.166 on 11/06/15 from IP address 54.152.109.166. Copyright ARRS. For personal use only; all rights reserved

Metafratzi et al. CT of Pulmonary Embolism

A

C E N T U R Y

MEDICAL

O F

IMAGING

Zafiria M. Metafratzi1 Miltos P. Vassiliou2 George C. Maglaras2 Froso G. Katzioti1 Stauros H. Constantopoulos1 Afroditi Katsaraki3 Stauros C. Efremidis1

Keywords: angiography, CT; chest; embolism; lungs; pulmonary artery obstruction index DOI:10.2214/AJR.04.1320 Received August 23, 2004; accepted after revision January 17, 2005. 1Department of Radiology, University Hospital of Ioannina,

Panepistimiou Ave., Ioannina 45500, Greece. Address correspondence to Z. Metafratzi ([email protected]). 2Department

of Pneumonology, University Hospital of Ioannina, Ioannina 45500, Greece.

3Department

of Statistics, University Hospital of Ioannina, Ioannina 45500, Greece.

AJR 2006; 186:213–219 0361–803X/06/1861–213 © American Roentgen Ray Society

AJR:186, January 2006

Acute Pulmonary Embolism: Correlation of CT Pulmonary Artery Obstruction Index with Blood Gas Values OBJECTIVE. The purpose of our study was to investigate the relation between the pulmonary artery obstruction index assessed with helical CT and impairment in blood gases in patients with acute pulmonary embolism. SUBJECTS AND METHODS. Helical CT pulmonary angiography was performed in 78 patients who were suspected of having acute pulmonary embolism and selected as being free of underlying cardiopulmonary disease. Findings consistent with acute pulmonary embolism were observed in 34 patients. The severity was assessed by the pulmonary artery obstruction index, defined as Σ (n × d), where n is the number of segmental arteries occluded and d is the degree of obstruction. Spearman’s rank correlation coefficients were used to assess the correlation between the index of arterial obstruction and arterial partial pressure of oxygen (PaO2); alveolar–arterial difference in partial pressure of oxygen (PAO2 – PaO2); arterial partial pressure of carbon dioxide (PaCO2); and arterial oxygen saturation (SaO2). The statistical difference of the arterial blood gas values between the two groups of patients (those with and those without pulmonary embolism) was evaluated using the Mann-Whitney U test. Blood gases were comparatively evaluated below and above different index values (from 40% to 70%) and different PaCO2 values (25, 30, and 35 mm Hg) as possible indexes of embolism severity using the same test. The level of significance was set at 95% (p = 0.05). RESULTS. The values of PaO2, SaO2, and PaCO2 were significantly lower (p = 0.024, p = 0.0062, and p = 0.000075, respectively) and the values of PAO2 – PaO2 were significantly higher (p = 0.0169) in the pulmonary embolism group than in the no-pulmonary-embolism group. A significant correlation was observed between the obstruction index and PaO2 (r = −0.33, p = 0.05), PaCO2 (r = −0.34, p = 0.05), PAO2 – PaO2 (r = 0.39, p = 0.02), and SaO2 (r = −0.35, p = 0.04). Using cutoff values for the pulmonary artery obstruction index of 40%, 50%, 60%, and 70%, we observed that PaCO2 and PAO2 – PaO2 differed significantly between above and below the 40% (p = 0.018 and p = 0.03), 50% (p = 0.0087 and p = 0.029), and 60% (p = 0.005 and p = 0.003) cutoffs. PaO2 differed significantly for the cutoff values of 60% (p = 0.03) and 70% (p = 0.004). The same was observed for SaO2 at 60% (p = 0.05) and 70% (p = 0.03). Comparisons for PaCO2 showed that a value of 30 mm Hg significantly separates levels of the pulmonary artery obstruction index (p = 0.002), with 78% sensitivity and 82% specificity indicating a pulmonary artery obstruction index greater than 50%. CONCLUSION. In patients with acute pulmonary embolism but no other underlying cardiopulmonary disease, the severity of the pulmonary arterial tree obstruction assessed using the CT obstruction index is significantly correlated to the blood gas values. The strongest correlation was observed between the index and the PAO2 – PaO2. Furthermore, a PaCO2 value of 30 mm Hg or less is highly suggestive of an obstruction index of more than 50% of the arterial bed. elical CT of the pulmonary arteries (CTPA) has become the first-line technique for the detection of emboli in the large and segmental vessels [1–3]. More recently, the development of MDCT enables scanning of the entire thorax during the peak enhancement of the pulmonary arterial tree using submillimeter collimation,

H

thus increasing the sensitivity and specificity of the method [4–6]. However, when reporting positive CTPA findings in routine clinical practice, attention is mainly focused on the presence of pulmonary emboli, a description of their location (main pulmonary artery, lobar, segmental, and subsegmental branches), and a gross estimate of their extent (massive, extensive, or

213


Metafratzi et al. isolated). More precise and objective estimation of the severity of obstruction of the arterial bed relevant to the patient’s treatment and prognosis has not been established [7]. A few studies have addressed this issue [8–11] and, to our knowledge, in only one was the severity of obstruction as assessed on CTPA quantified using the obstruction index of Qanadli et al. [9], which is related to clinical data and, more specifically, to the clinical outcome [12]. We designed this study to investigate whether the quantification of acute pulmonary embolism using the pulmonary artery obstruction index as proposed by Qanadli et al. [9] correlates with the functional lung impairment of acute pulmonary embolism and, more specifically, with hypoxemia, alveolar–arterial gradient of oxygen (PAO2 – PaO2), hypocapnia, and hemoglobin desaturation. Subjects and Methods Seventy-eight patients (36 men and 42 women) with a high clinical suspicion of pulmonary embolism and with no underlying cardiopulmonary disease participated in this study. The mean age of the population was 62 ± 16 years (range, 30−87 years), and all were in stable hemodynamic condition. Underlying cardiopulmonary disease was excluded in all patients by history, physical examination, CT findings, and ECG. Cardiac echo was performed in a minority of patients (n = 5). Prior cardiopulmonary disease was defined as valvular, myocardial, or pericardial disease; myocardial infarction; prior left or right heart failure; chronic obstructive disease or other parenchymal and pleural disease; or chronic thromboembolic disease and concurrent pneumonia. The CT criteria used to exclude patients with cardiopulmonary disease were signs of emphysema; interstitial lung disease (according to the Nomenclature Committee of the Fleischner Society [13]); other chronic parenchymal, pleural, or thromboembolic disease; pulmonary edema; and chronic pericardial disease. All patients underwent CTPA within 12 hr of admission. Fifty-three patients were scanned on a single-detector CT scanner (Secura, Philips Medical Systems) and 25 patients on a 16-MDCT scanner (MX IDT 8000, Philips Medical Systems). The imaging parameters used for each type of scanner were as follows: for single-detector CT, slice thickness of 3 mm with a pitch of 1, reconstruction thickness of 2 mm, and gantry rotation of 0.75/1 sec; for 16-MDCT, collimation of 16 × 0.75 with a pitch of 1.238, slice thickness of 0.8 mm, reconstruction thickness of 0.4 mm, and gantry rotation of 0.5 sec. The lungs were scanned in a craniocaudal direction from the level of the lower pulmonary veins to 3 cm above the aortic arch in 25–28 sec for single-detector CT, whereas the scanning time for the whole chest

214

(from lung bases to the apex) was 10–12 sec for 16MDCT. The patients were scanned in suspended inspiration or during quiet breathing if they were unable to breath-hold. A total of 60–130 mL of low-osmolar contrast material (iopromide, 300 mg I/mL [Ultravist, Schering]) was administered IV with an automated injector at a rate of 3 mL/sec for singledetector CT and 4 mL/sec for 16-MDCT. To optimize the pulmonary arterial tree opacification, a timing bolus technique was used. The studies were interpreted on a workstation (Brilliance, Philips Medical Systems) by two radiologists experienced in thoracic imaging (> 5 years of experience) in a combined interpretation. Both observers were blinded to the severity of the patient’s clinical condition and the blood gas values. All axial images were viewed on the workstation using the cine stack technique and standard mediastinal windows (center, 50 H; level, 350 H). However, the observers were free to review multiplanar reconstructions and to change the window and level settings in real time to optimize visualization of the opacified vessels. The criterion used for the diagnosis of pulmonary emboli was the presence of an endoluminal filling defect on CTPA. The degree of obstruction was quantified according to obstruction index of Qanadli et al. [9], defined as the product Σ (n × d) expressed in percentage of vascular obstruction, where n is the value of the proximal clot site that equals the number of segmental branches arising distally, and d is the degree of obstruction where partial obstruction is scored as 1 and complete obstruction as 2. Values for n range from a minimum of 1 (one segment obstructed) to a maximum of 20 (obstruction of both right and left pulmonary arteries). The total Σ (n × d) product for each patient was divided by the maximum total score (20 segments × 2) and multiplied by 100 according to the formula ([Σ (n × d) / 40] × 100) to obtain to a percentage of vascular obstruction [9]. Arterial blood gases, including arterial partial pressure of oxygen (PaO2); arterial partial pressure of carbon dioxide (PaCO2); and arterial oxygen saturation (SaO2). were measured on admission to the emergency department with the patient breathing room air. Furthermore, the alveolar–arterial gradient for oxygen (PAO2 – PaO2) was calculated according to the formula: PAO2 – PaO2 gradient (mm Hg) = 150 – 1.25 PaCO2 – PaO2 [14]. The study was approved by the ethics committee of our institution; informed consent was not required.

Statistical Analysis Results are expressed as mean ± SD and median. Spearman’s rank correlation coefficients were used

to assess the correlation between the pulmonary artery obstruction index and PaO2, PAO2 – PaO2, PaCO2, and SaO2. The Mann-Whitney U test was applied to discriminate differences of the blood gas values between the two patient groups (those with positive and those with negative CTPA findings). The same test was used to evaluate differences of the same blood gas values above and below certain pulmonary artery obstruction index levels (40%, 50%, 60%, and 70%) and to compare different PaCO2 levels (35, 30, or 25 mm Hg) with the pulmonary artery obstruction index as possible cutoff values. The level of significance was set at 95% (p = 0.05).

Results Six CTPA images sets (6/78, 8%) were excluded from the study because of poor image quality: three caused by suboptimal vessel opacification and three because of motion artifacts. Thus, 72 patients remained for the analysis. Pulmonary embolism was identified in 34 patients (34/72, 47%). The mean percentage of obstruction of the pulmonary arterial tree (the pulmonary artery obstruction index) and the blood gases profile of the two patient groups (positive and negative CTPA findings for pulmonary embolism) are presented in Table 1. The values of PaO2, SaO2, and PaCO2 were significantly lower (p = 0.024, p = 0.0062, and p = 0.000075, respectively), and PAO2 – PaO2 significantly higher (p = 0.0169), in the pulmonary embolism group than in the no-pulmonary-embolism group. Furthermore, correlations between the pulmonary artery obstruction index and all blood gas values were found to be statistically significant. More specifically, the pulmonary artery obstruction index was significantly correlated with PaO2 (r = −0.33, p = 0.05), PAO2 – PaO2 (r = 0.39, p = 0.02), PaCO2 (r = −0.34, p = 0.05), and SaO2 (r = −0.35, p = 0.04). These correlations are graphically depicted in the x–y scatter diagrams of Figure 1. Of the 34 patients with pulmonary embolism, the pulmonary artery obstruction index exceeded 50% in 13 patients (13/34, 38%) and 60% in 11 patients (11/34, 32%) (Fig. 2). The pulmonary artery obstruction index ranged between 40% and 49% in two (2/34, 6%), between 30% and 39% in three (3/34, 9%), and between 2.5% and 29% in 16 patients (16/34, 47%) (Fig. 3). Using as cutoff values for the pulmonary artery obstruction index 40%, 50%, 60%, and 70%, we observed that PaCO2 and PAO2 – PaO2 differed significantly between above and below the 40% (p = 0.018 and p = 0.03), 50% (p = 0.0087 and p = 0.029), 60% (p = 0.005 and p =


CT of Pulmonary Embolism TABLE 1: Blood Gas Profile of Patients With and Without Pulmonary Embolism and Percentage of Pulmonary Artery Obstruction in the Former Pulmonary Artery Obstruction Index (%)

PaCO2 (mm Hg)

PaO2 (mm Hg)

PAO2 – PaO2 (mm Hg)

SaO2 (%)

No.

Mean ± SD

Median

Mean ± SD

Median

Mean ± SD

Median

Mean ± SD

Median

Mean ± SD

Median

Pulmonary embolism

34

39.8 ± 27.3

32.5

32.8 ± 5.2

33.9

63.9 ± 14.1

64.2

45.01 ± 16.4

44.8

90.9 ± 8

93

No pulmonary embolism

38

—

—

51 ± 26

38

74.7 ± 18.6

70

35.5 ± 11.9

37.07

97.3 ± 7

95

pb

—

0.000075

0.024

0.016

0.0062

Note—Dash (—) indicates not applicable. PaCO2 = arterial partial pressure of carbon dioxide, PaO2 = arterial partial pressure of oxygen, PAO2 –PaO2 = alveolar–arterial difference in partial pressure of oxygen, SaO2 = arterial oxygen saturation. a In pulmonary embolism group, patients’ CT pulmonary angiography examinations were positive for acute pulmonary embolism; in the non-pulmonary-embolism group, patients’ examinations were negative. b Using the nonparametric Mann-Whitney U test.

44

100 90

40

80 36

PaCO2

PaO2

70

32

60 50

28

40 24

30 20 –10

10

30

50

70

90

20 –10

110

10

30

50

70

90

110

Pulmonary Artery Obstruction Index

Pulmonary Artery Obstruction Index A 105

B

100 90

95

PAO2–PaO2

80 85

SaO2


Groupa

75

70 60 50 40

65

30

55

20 45 –10

10

30

50

70

90

110


10 –10

10

30

50

70

90

110

Pulmonary Artery Obstruction Index C

D

Fig. 1—Scatter diagrams show negative correlation with pulmonary artery obstruction index for 34 studied patients with acute pulmonary embolism. Note negative correlation between pulmonary artery obstruction index and blood gas values. A, PaO2 (arterial partial pressure of oxygen) (r = −0.33, p = 0.057). B, PaCO2 (arterial partial pressure of carbon dioxide) (r = −0.34, p = 0.05). C, SaO2 (arterial oxygen saturation) (r = −0.35, p = 0.045). D, PAO2 – PaO2 (alveolar–arterial difference in partial pressure of oxygen) (r = 0.39, p = 0.02).

0.003), and 70% (p = 0.00005 and p = 0.003) cutoff values. PaO2 differed significantly for the cutoff value of 60% (p = 0.03) and 70% (p = 0.004). The same was observed for SaO2


at 60% (p = 0.05) and 70% (p = 0.03) (Tables 2–5). Comparisons for PaCO2 showed that a value of 30 mm Hg separates pulmonary artery obstruction index levels significantly

(p = 0.002) (Table 6). Moreover, this PaCO2 value showed a 78% sensitivity and 82% specificity for indicating a pulmonary artery obstruction index greater than 50% (Fig. 4).

215


Metafratzi et al.

A

B

Fig. 2—76-year-old man with multiple large emboli. A and B, Coronal multiplanar reconstructions of CT pulmonary angiography performed with 16-MDCT scanner show endoluminal defects (arrows). Patient received pulmonary artery obstruction index of 87.5%, and his blood gas values were PaO2 (arterial partial pressure of oxygen), 43.4 mm Hg; PaCO2 (arterial partial pressure of carbon dioxide), 25.5 mm Hg; SaO2 (arterial oxygen saturation), 82.5%; and PAO2 – PaO2 (alveolar–arterial difference in partial pressure of oxygen), 74.7 mm Hg.

Fig. 3—23-year-old woman with chest pain and dyspnea. A and B, CT pulmonary angiography performed with 16-MDCT in sagittal multiplanar reconstruction (A) and axial projection (B) shows endoluminal defects (arrows). Patient had segmental clot receiving pulmonary artery obstruction index of 7.5%. Blood gas values were PaO2 (arterial partial pressure of oxygen), 91.8 mm Hg; PaCO2 (arterial partial pressure of carbon dioxide), 33.8 mm Hg; SaO2 (arterial oxygen saturation), 97%; and PAO2 – PaO2 (alveolar–arterial difference in partial pressure of oxygen), 15.95 mm Hg.

A

Discussion CTPA has dramatically improved the quality of imaging of the pulmonary vasculature in the last decade and has changed the diagnostic approach to suspected acute pulmonary embolism. The average sensitivity and specificity for detecting pulmonary embolism has been shown to be 88% [6] and 92–96%, respectively [2, 15–18]. The introduction of MDCT has made it possible to image the pulmonary arterial tree faster with submillimeter

216

collimation, thus improving the overall quality and increasing the sensitivity for isolated subsegmental pulmonary embolism [4–6]. However, the interpretation of CTPA is mainly focused on the presence or absence of emboli and a rough estimation of their severity (massive, extensive, or isolated). A few published articles confront the challenge of quantifying pulmonary arterial tree obstruction [8–11] by developing an obstruction index. In our study, the obstruction index of Qa-

B

nadli et al. [9] was used because of the simplicity of applying it in routine clinical practice. This index is based on the number of segmental arteries occluded and those arising distally to an occluded proximal branch, with the integration of a weighting factor of 1 for partial and 2 for complete obstruction. Six of 78 CTPA image sets were inconclusive because of inadequate imaging quality, which is in accordance with the published data in a recent review article [16, 17]. The


CT of Pulmonary Embolism TABLE 2: Levels of Pulmonary Artery Obstruction Index and Significance of Corresponding Arterial Partial Pressure of Carbon Dioxide (PaCO2)


Index (%) Index Level (%)

Mean ± SD

> 40

67.53 ± 13.3

< 40 > 50

PaCO2 (mm Hg) Median

No.

Mean ± SD

Median

70

15

30.1 ± 5.6

29

18 ± 9.5

15

19

34.9 ± 3.9

34

71 ± 10.5

72.5

13

29.4 ± 5.6

28

< 50

20.6 ± 12

17.5

21

34.9 ± 3.8

34

> 60

74.5 ± 7

72

11

28.72 ± 5.6

< 60

23.3 ± 14.8

20

23

34.8 ± 3.8

34

> 70

78.2 ± 5.9

75

7

29.7 ± 4.7

28

< 70

29.9 ± 21

25

27

33.4 ± 5.2

34

a Using nonparametric Mann-Whitney

26.8

pa 0.018 0.008 0.005 0.00005

U test.

TABLE 3: Levels of Pulmonary Artery Obstruction Index and Significance of Corresponding Arterial Partial Pressure of Oxygen (PaO2) Index (%) Index Level (%)

Mean ± SD

> 40

67.53 ± 13.3

PaO2 (mm Hg) No.

Mean ± SD

Median

pa

70

15

59.7 ± 14.6

58.5

NS

Median

< 40

18 ± 9.5

15

19

67.3 ± 13.2

67.5

> 50

71 ± 10.5

72.5

13

59.3 ± 14.7

58.5

< 50

20.6 ± 12

17.5

21

66.7 ± 13.4

67.5

> 60

74.5 ± 7

72

11

56.1 ± 13.6

57

< 60

23.3 ± 14.8

20

23

67.6 ± 13

68

> 70

78.2 ± 5.9

75

7

49.7 ± 11.7

49

< 70

29.9 ± 21

25

27

67.6 ± 12.4

68

NS 0.03 0.004

Note—NS = not significant (p > 0.05). a Using nonparametric Mann-Whitney U test.

TABLE 4: Levels of Pulmonary Artery Obstruction Index and Significance of Corresponding Alveolar–Arterial Difference in Partial Pressure of Oxygen (PAO2 – PaO2) Index (%) Index Level (%)

Mean ± SD

> 40

67.53 ± 13.3

< 40 > 50

PAO2 – PaO2 (mm Hg) Median

No.

Mean ± SD

Median

pa 0.03

70

15

52.6 ± 18.4

53.5

18 ± 9.5

15

19

38.9 ± 11.8

41.3

71 ± 10.5

72.5

13

53.83 ± 18.5

53.5

< 50

20.6 ± 12

17.5

21

39.5 ± 12.6

41.2

> 60

74.5 ± 7

72

11

57.9 ± 16.7

54.8

< 60

23.3 ± 14.8

20

23

38.8 ± 12.4

37.5

> 70

78.2 ± 5.9

75

7

63.2 ± 14.7

64.7

< 70

29.9 ± 21

25

27

40.3 ± 13.4

41.2

a Using nonparametric Mann-Whitney

0.003 0.003

U test.

mean value of the pulmonary artery obstruction index in patients was 39% ± 27%, which is higher than that described by other investigators [9]. This difference could be explained by the improved sensitivity of 16-MDCT in


0.029

assessing pulmonary embolism as a result of overall improved image quality. The index of Qanadli et al. [9] for pulmonary artery obstruction has been validated by the Miller index [19] and has shown a strong correlation

(r = 0.867, p < 0.0001). A significant correlation also exists between the index of Qanadli et al. and right ventricle dilatation [9]. However, to our knowledge no attempt has been made to relate the severity of obstruction expressed by this index to clinical parameters such as arterial blood gas measurements. Gas exchange abnormalities observed in patients with acute pulmonary embolism are related to the size of the emboli, the presence of underlying cardiopulmonary disease, the degree of obstruction, and the time since embolization [20]. Mechanisms responsible for hypoxemia in patients with acute pulmonary embolism are the degree of ventilation–perfusion (V/Q) mismatch, the amount of right-toleft shunt, the level of cardiac output, and diffusion impairment, with V/Q abnormalities accounting for most of the hypoxemia and diffusion impairment playing a limited role [20–24]. Moreover, PAO2 – PaO2 is mainly disturbed by the V/Q mismatch and intrapulmonary shunt, whereas oxygen saturation of hemoglobin falls in parallel with the PaO2 reduction. Hypocapnia is the result of hyperventilation triggered by hypoxemia and reflexes initiated in the lung parenchyma [20, 22]. Therefore, the degree of hypocapnia is closely and inversely related to the amount of excess ventilation. Blood gas analysis is a useful part of the diagnostic workup and follow-up of patients with suspected or documented pulmonary embolism because it is performed at the bedside. The patients included in this study had no other underlying cardiopulmonary disease. The blood gas values and the CTPA findings were assessed on admission, with the patient breathing room air before anticoagulant therapy, assuming that the blood gas levels measured at that time expressed purely the effects of the acute pulmonary embolism and the specific compensatory responses. The pulmonary artery obstruction index correlated significantly with all blood gas values. However, the strongest correlation was observed between the pulmonary artery obstruction index and PAO2 – PaO2. This finding is in accordance with the study by McFarlane and Imperiale [25], in which the PAO2 – PaO2 gradient showed a linear correlation with the severity of pulmonary embolism, assessed by the pulmonary artery mean pressure gradient and by the number of mismatch vascular–perfusion defects on V/Q scans. In addition, our results are in agreement with those of McIntyre and Sasahara [26], who reported a linear relationship between pulmonary embolism

217

severity and PaO2 levels (p < 0.05), although in their study pulmonary embolism was proven by selective angiography, a different obstruction index was used for the quantification, and they did not include correlation with the PaCO2 and PAO2 – PaO2 values.

Ventilation and PaCO2 are closely and inversely related, and our results suggest that the extent of pulmonary artery obstruction regulates the level of hyperventilation response. This might explain the higher sensitivity shown by PaCO2 for revealing the severity of

TABLE 5: Levels of Pulmonary Artery Obstruction Index and Significance of Corresponding Arterial Oxygen Saturation (SaO2) Index (%) Index Level (%)

Mean ± SD

> 40

67.53 ± 13.3

SaO2 (%) Median

No.

Mean ± SD

Median

pa NS

70

15

88.8 ± 11

91

< 40

18 ± 9.5

15

19

92.6 ± 4.3

95

> 50

71 ± 10.5

72.5

13

88.4 ± 11.7

91

< 50

20.6 ± 12

17.5

21

92.5 ± 4.3

95

> 60

74.5 ± 7

72

11

87.2 ± 12.3

89

< 60

23.3 ± 14.8

20

23

92.8 ± 4.2

95

> 70

78.2 ± 5.9

75

7

< 70

29.9 ± 21

25

27

84 ± 14.8

88

92.7 ± 4.4

NS 0.05 0.036

95


TABLE 6: Levels of PaCO2 and Significance of Pulmonary Artery Obstruction Index PaCO2 (mm Hg) Median

pa for Index

38.4

NS

PaCO2 Levels (mm Hg)

No.

Mean ± SD

≥ 35

12

38.3 ± 1.87

≤ 35

22

29.8 ± 3.9

30.5

≥ 30

23

35.9 ± 3.01

35.5

≤ 30

11

26.4 ± 2.4

26.8

≥ 25

31

33.7 ± 4.54

34

≤ 25

3

23.3 ± 1.25

23.5

0.002 NS




Metafratzi et al.

100

Fig. 4—Scatter diagram of pulmonary artery obstruction index versus PaCO2 (arterial partial pressure of carbon dioxide) illustrates sensitivity (78%) and specificity (82%) of PaCO2 = 30 mm Hg (dotted line) for indicating severe pulmonary embolism (pulmonary artery obstruction index > 50%).

218

90 80 70 60 50 40 30 20 10 0 20

25

30

35

PaCO2 (mm Hg)

40

45

acute pulmonary embolism in our study. Among our patients with a pulmonary artery obstruction index greater than 50%, only two patients showed PaO2 of less than 55 mm Hg. Conversely, 12 of 13 patients had a PaCO2 of less than 30 mm Hg. The high sensitivity and specificity of PaCO2 of 30 mm Hg for detecting severe acute pulmonary embolism (pulmonary artery obstruction index > 50%) offers additional confirmation of its clinical usefulness. PaCO2 and PAO2 – PaO2 were significantly different between above and below the four tested values of the pulmonary artery obstruction index, whereas PaO2 and SaO2 were significantly different only for the 60% and 70% values of the pulmonary artery obstruction index. Although overlapping gas values were present in all comparisons, PaCO2 was clearly more sensitive for reflecting the degree of pulmonary artery obstruction. We consider our results to be strongly indicative; however, their validity should be reinforced with further studies with larger number of patients. A limitation of our study was that two different scanning protocols were used with different collimations and scanning times. However, this is not expected to influence the results much, not only because thin collimation was used with the single-detector CT scanner (2–3 mm) but also because the pulmonary artery obstruction index depends on the presence of emboli on main, lobar, and segmental arteries that can be accurately documented with thin-collimation helical CT. Another limitation of our study was the lack of evaluation of the degree of agreement between the two observers because CTPA images were interpreted by consensus interpretation between the two radiologists. However, Qanadli et al. [9], who first described the CT obstruction index, observed a high correlation between scores obtained by two observers. Finally, the lack of baseline arterial blood gas measurements should be considered another limitation of the study. In summary, the CT obstruction index used in this study showed a strong correlation between the severity of arterial bed obstruction and the blood gas values in patients with acute pulmonary embolism and no underlying cardiopulmonary disease. The strongest correlation was observed between the index and PAO2 – PaO2. Moreover, an obstruction index of greater than 50% of the arterial bed may be expected for PaCO2 of 30 mm Hg or less. This preliminary observation suggests that PAO2 – PaO2 and PaCO2, which are routinely available in the initial evaluation of suspected pulmonary embolism, in combination with other


CT of Pulmonary Embolism


clinical parameters, can assist the primary care physician in suspecting severe acute pulmonary embolism and referring the patient for further evaluation.

9.

Acknowledgment We thank Dr. M. Remy-Jardin for her review of the manuscript, her support, and her advice. 10.

References 1. 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 2. Remy-Jardin M, Remy J, Deschildre F, et al. Diagnosis of pulmonary embolism with spiral CT: comparison with pulmonary angiography and scintigraphy. Radiology 1996; 200:699–706 3. Senac J, Vernhet H, Bousquet C, et al. Embolie pulmonaire: apport de la tomodensitometrie helicoidale [in French]. J Radiol 1995; 76:339–345 4. Ghaye B, Szapiro D, Mastora I, et al. Peripheral pulmonary arteries: how far in the lung does multi-detector row spiral CT allow analysis? Radiology 2001; 219:629–636 5. Fleischmann D. Present and future trends in MDCT applications: CT angiography. Eur Radiol 2002; 12[suppl 2]:S11–S16 6. Patel S, Kazerooni E, Cascade P. Pulmonary embolism: optimization of small pulmonary artery visualization at multi-detector row CT. Radiology 2003; 227:455–460 7. Carson JL, Kelley MA, Duff A, et al. The clinical course of pulmonary embolism. N Engl J Med 1992; 326:1240–1245 8. Bankier AA, Janata K, Fleischmann D, et al. Severity assessment of acute pulmonary embolism with


11.

12.

13.

14.

15.

16.

spiral CT: evaluation of two modified angiographic scores and comparison with clinical data. J Thorac Imaging 1997; 12:150–158 Qanadli SD, El Hajjam M, Vieillard-Baron A, et al. New CT index to quantify arterial obstruction in pulmonary embolism: comparison with angiographic index and echocardiography. AJR 2001; 176:1415–1420 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 Radiol 2003; 13:29–35 Collomb D, Paramelle PJ, Calaque O, et al. Severity assessment of acute pulmonary embolism: evaluation using helical CT. Eur Radiol 2003; 13:1508–1514 Wu AS, Pezzullo BAJ, Cronan JJ, Hou DD, MayoSmith WW. CT pulmonary angiography: quantification of pulmonary embolus as a predictor of patient outcome—initial experience. Radiology 2004; 230:831–835 Austin JH, Muller NL, Friedman PJ, et al. Glossary of terms for CT of the lungs: recommendations of the Nomenclature Committee of the Fleischner Society. Radiology 1996; 200:327–331 Stein DP, Goldhaber ZS, Henry WJ, Miller CA. Arterial blood gas analysis in the assessment of suspected acute pulmonary embolism. Chest 1996; 109:78–81 Remy-Jardin M, Remy J, Wattinne I, Giraud F. Central pulmonary thromboembolism: diagnosis with spiral volumetric CT with the single-breath-hold technique—comparison with pulmonary angiography. Radiology 1992; 185:381–387 Gottsaeter A, Berg A, Centergard J, Frennby B, Norhov N, Nyman U. Clinically suspected pulmonary embolism: is it safe to withhold anticoagulation after a negative spiral CT? Eur Radiol 2001;

11:65–72 17. van Beek EJ, Brouwers EM, Song B, Bongaerts AH, Oudkerk M. Lung scintigraphy and helical computed tomography for the diagnosis of pulmonary embolism: a meta-analysis. Clin Appl Thromb Hemost 2001; 7:87–92 18. Mayo JR, Remy-Jardin M, Muller NL, et al. Pulmonary embolism: prospective comparison of spiral CT with ventilation-perfusion scintigraphy. Radiology 1997; 205:447–452 19. 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 20. D’Alonzo GE, Dantzker DR. Gas exchange alterations following pulmonary thromboembolism. Clin Chest Med 1984; 5:411–419 21. D’Alonzo GE, Bower JS, DeHart P, Dantzker DR. The mechanisms of abnormal gas exchange in acute massive pulmonary embolism. Am Rev Respir Dis 1983; 128:170–172 22. Wilson JE III, Pierce AK, Johnson RL Jr, et al. Hypoxemia in pulmonary embolism: a clinical study. J Clin Invest 1971; 50:481–491 23. Dantzker DR, Wagner PD, Tornabene VW, Alazraki NP, West JB. Gas exchange after pulmonary thromboembolism in dogs. Circ Res 1978; 42:92–103 24. Manier G, Castaing Y, Guenard H. Determinants of hypoxemia during the acute phase of pulmonary embolism in humans. Am Rev Respir Dis 1985; 132:332–338 25. McFarlane MJ, Imperiale TF. Use of the alveolararterial oxygen gradient in the diagnosis of pulmonary embolism. Am J Med 1994; 96:57–62 26. McIntyre KM, Sasahara AA. The hemodynamic response to pulmonary embolism in patients without prior cardiopulmonary disease. Am J Cardiol 1971; 28:288–294

219


This article has been cited by: 1. Yunfeng Zhang, Qixing Zhou, Ying Zou, Xiaolian Song, Shuanshuan Xie, Min Tan, Guoliang Zhang, Changhui Wang. 2015. Risk factors for pulmonary embolism in patients preliminarily diagnosed with community-acquired pneumonia: a prospective cohort study. Journal of Thrombosis and Thrombolysis . [CrossRef] 2. V. Vukadinovic, J.D. Chiou, D.L. Morris. 2015. Clinical features of pulmonary emboli in patients following cytoreductive surgery (peritonectomy) and hyperthermic intraperitoneal chemotherapy (hipec), a single centre experience. European Journal of Surgical Oncology (EJSO) . [CrossRef] 3. B.S. Praveen Kumar, D. Rajasekhar, V. Vanajakshamma. 2014. Study of clinical, radiological and echocardiographic features and correlation of Qanadli CT index with RV dysfunction and outcomes in pulmonary embolism. Indian Heart Journal 66, 629-634. [CrossRef] 4. Xavier Bobbia, Romain Genre Grandpierre, Pierre-Géraud Claret, Alexandre Moreau, Stéphane Pommet, Jean-Marie Bonnec, Rémi Perrin Bayard, Jean-Yves Lefrant, Laurent Muller, Jean-Emmanuel de La Coussaye. 2013. Ultrasound guidance for radial arterial puncture: a randomized controlled trial. The American Journal of Emergency Medicine 31, 810-815. [CrossRef] 5. Savas Ozsu, Yasin Abul, Ismail Yilmaz, Asiye Ozsu, Funda Oztuna, Yilmaz Bulbul, Tevfik Ozlu. 2012. Prognostic significance of PaO2/PaCO2 ratio in normotensive patients with pulmonary embolism. The Clinical Respiratory Journal 6:10.1111/ crj.2012.6.issue-2, 104-111. [CrossRef] 6. Sven F. Thieme, Nima Ashoori, Fabian Bamberg, Wieland H. Sommer, Thorsten R. C. Johnson, Hanno Leuchte, Alexander Becker, Daniel Maxien, Andreas D. Helck, Jürgen Behr, Maximilian F Reiser, Konstantin Nikolaou. 2012. Severity assessment of pulmonary embolism using dual energy CT – correlation of a pulmonary perfusion defect score with clinical and morphological parameters of blood oxygenation and right ventricular failure. European Radiology 22, 269-278. [CrossRef] 7. Gregory GladishPulmonary Embolic Disease and Cardiac Tumors 707-719. [CrossRef] 8. Tongfu Yu, Mei Yuan, Qingbo Zhang, Haibing Shi, Dehang Wang. 2011. Evaluation of computed tomography obstruction index in guiding therapeutic decisions and monitoring percutanous catheter fragmentation in massive pulmonary embolism. Journal of Biomedical Research 25, 431-437. [CrossRef] 9. Naim Ceylan, Sezai Tasbakan, Selen Bayraktaroglu, Gursel Cok, Tarik Simsek, Soner Duman, Recep Savaş. 2011. Predictors of Clinical Outcome in Acute Pulmonary Embolism. Academic Radiology 18, 47-53. [CrossRef] 10. Kei Nakada, Takemichi Okada, Hisato Osada, Norinari Honda. 2010. Relation between pulmonary embolus volume quantified by multidetector computed tomography and clinical status and outcome for patients with acute pulmonary embolism. Japanese Journal of Radiology 28, 34-42. [CrossRef] 11. Hamid Shokoohi, Robert Shesser, Jeffrey P. Smith, Michael C. Hill, Robert Hirsch. 2009. The correlation between elevated cardiac troponin I and pulmonary artery obstruction index in ED patients with pulmonary embolism. The American Journal of Emergency Medicine 27, 449-453. [CrossRef] 12. B GHAYE, V WILLEMS, A NCHIMI, L KOUOKAM, C NOUKOUA, V DE MAERTELAER, P A GEVENOIS, R F DONDELINGER. 2009. Relationship between the extent of deep venous thrombosis and the extent of acute pulmonary embolism as assessed by CT angiography. The British Journal of Radiology 82, 198-203. [CrossRef] 13. Dolores Nauffal, Liliana Chacón, Rosalía Domènech. 2008. Utilidad de las pruebas básicas en el paciente con embolia pulmonar. Medicina Clínica 131, 48-53. [CrossRef] 14. Mary M. Costantino, Geneva Randall, Marc Gosselin, Marissa Brandt, Kristopher Spinning, C. David Vegas. 2008. CT Angiography in the Evaluation of Acute Pulmonary Embolus. American Journal of Roentgenology 191:2, 471-474. [Abstract] [Full Text] [PDF] [PDF Plus] 15. Ehab M. Kamel, Sabine Schmidt, Francesco Doenz, Ghazal Adler-Etechami, Pierre Schnyder, Salah D. Qanadli. 2008. Computed Tomographic Angiography in Acute Pulmonary Embolism. Journal of Computer Assisted Tomography 32, 438-443. [CrossRef]