characterize arterial supply of chest wall lesions using quantitative colour Doppler ... September 2002 and June 2003, 29 consecutive patients with chest wall ...
The British Journal of Radiology, 78 (2005), 303–307 DOI: 10.1259/bjr/28232950
E
2005 The British Institute of Radiology
Colour Doppler ultrasound mapping of chest wall lesions 1
¨ RG, MD, 1T BERT, MD, 1K GO ¨ RG, MD and 2M HEINZEL-GUTENBRUNNER, MD C GO
1
Department of Internal Medicine, Philipps-University, Baldingerstraße, Marburg and 2Institute of Medical Biometry and Epidemiology, Philipps-University, Bunsenstraße 3, Marburg, Germany
Abstract. Spectral curve-analysis of arterial flow signals (FS) in patients with pulmonary lesions is able to discriminate FS of bronchial arteries (BA) from FS of pulmonary arteries (PA). In patients with chest wall lesions a different FS from that of the BA/PA can be obtained. The aim of the study was to evaluate and characterize arterial supply of chest wall lesions using quantitative colour Doppler ultrasound (CDS). Between September 2002 and June 2003, 29 consecutive patients with chest wall lesions were examined by CDS. 16 lesions were located strictly to the chest wall (group I). 13 lesions had a chest wall lesion with pulmonary extension (group II). The following parameters were prospectively determined: (1) qualitative CDS (absence or evidence of vascularity); (2) quantitative CDS of intercostal or non-intercostal located arterial FS (resistive index (RI) and pulsatility index (PI)); (3) number of different arterial FS in one lesion using CDS-mapping. In a control group of 17 healthy volunteers quantitative measurement of RI and PI of the intercostal artery (ICA) was performed. 4 of 29 patients (14%) had no FS by CDS mapping. Quantitative CDS parameters of the control group were mean RI of ICA 0.88 (¡0.056); mean PI of ICA 2.88 (¡0.643); of group I mean RI of ICA 0.79 (¡0.127) mean PI of ICA 1.93 (¡0.641), and of group II mean RI of ICA 0.79 (¡0.144), mean PI of ICA 2.1 (¡1.015), mean RI of non-ICA 0.68 (¡0.675) mean PI of non-ICA 2.5 (¡2.506). Median RI as well as PIvalue obtained within the chest wall (ICA) do not differ between group I, group II, and the control group. Within group II impedance measurements discriminates intercostal from non-intercostal arterial supply. In 29 patients 37 different arterial FS were obtained. None of the 16 patients in group I and 8 of the 13 patients in group II had 2 or more different FS. Lesions strictly located to the chest wall had an arterial supply characteristic for ICA by quantitative CDS. Chest wall lesions with pulmonary extension demonstrate a complex arterial supply by quantitative CDS.
Colour Doppler ultrasound (CDS) can be used as a qualitative or a quantitative method to describe flow patterns in pleurally based pulmonary lesions. Qualitative CDS pattern show characteristic findings in various pulmonary lesions [1–6]. Quantitative CDS of arterial flow signals (FS) enables discrimination of FS of bronchial arteries (BA) characterized by a monophasic low impedance flow pattern from FS of pulmonary arteries (PA) characterized by a triphasic high impedance flow pattern [4–8]. In a recent study [9] of chest wall based pulmonary lesions a variable impedance, monophasic flow pattern could discrimate from FS of PA and FS of central BA by spectral analysis. The aim of the present study is to evaluate and characterize arterial supply of chest wall lesions with quantitative CDS.
Patients and methods Between September 2002 and June 2003, 29 consecutive patients with chest wall lesion were referred to our ultrasound laboratory and prospectively examined by B-mode ultrasound as well as quantitative and qualitative CDS. Diagnosis was histologically confirmed in all patients (Table 1). 16 lesions were strictly located in the chest wall and demarcated by pleural effusion from pulmonary tissue in eight cases. 13 lesions were predominantly located in the chest wall with pulmonary extension. 14 patients had a small size lesion with a largest diameter Received 26 November 2003 and in final form 6 October 2004, accepted 15 November 2004.
The British Journal of Radiology, April 2005
less than 4 cm, in 15 patients the lesion largest diameter was ¢4 cm. The range of lesion size was 2 cm to 14 cm. The age ranged from 18 years to 81 years. The following quantitative and qualitative CDS parameters were determined:
(1) Qualitative CDS of lesions was classified as absence of FS in all parts or evidence of FS in the lesion; (2) Quantitative measurement of resistive index (RI) and pulsatility index (PI) were performed in all FS studied; (3) Number of different arterial FS in a lesion were counted and classified as follows: Arterial vessels were assumed to be from the intercostal artery (ICA) by the following parameters:
Table 1. Diagnosis in 29 patients with chest wall lesions (group I) and chest wall lesions with pulmonary extension (group II)
Group I
Group II
Benign diseases
Malignant diseases
2 1 1 2
6 3 3 4 4 2 1
lipoma abscess pleural scar abscesses
metastases plasmocytoma malignant lymphoma metastases lung cancer PNET malignant lymphoma
PNET, primitive neuroectodermal tumour.
303
C Go¨rg, T Bert, K Go¨rg and M Heinzel-Gutenbrunner
(1) a strictly intercostal location; (2) a nearly horizontal flow direction, and; (3) a high impedance monophasic flow. Arterial vessels were assumed to be from the bronchial artery (BA) by
(1) a pulmonary location; (2) a variable flow direction; and (3) a low impedance monophasic flow. Arterial vessels were assumed to be from the pulmonary artery (PA) by
(1) a pulmonary location; (2) a centrifugal flow direction from the hilum towards the surface of the lung; and (3) a high impedance predominantly triphasic flow. Arterial vessels were assumed for tumour neoangiogenesis (TN) by a nearby constant flow that lacked systolic– diastolic variation [7]. In a control group of 17 healthy volunteers quantitative measurement of RI and PI of the intercostal artery was performed. Ultrasound was performed with a real time scanner with 3.5–8 MHz curved and linear array transducers (Acuson, Sequoia, Mountain View, CA) using tissue harmonic imaging. Ultrasound studies were performed by two authors experienced in CDS (CG and KG). Colour Doppler settings were optimized to achieve the greatest sensitivity to allow the detection of low flow. For evaluation of qualitative vascularity of pleural-based lesions power Doppler imaging was primarily used. For classification of flow direction colour Doppler imaging was performed. Colour gain setting was adjusted to a level just below the level creating artefacts. The colour box was designed as small as possible and the wall filter and pulse repetition frequency (PRF) were lowered. At least three similar, sequential Doppler waveforms were obtained while the subject suspended respiration. The single tracing obtained was selected for analysis. For statistical analysis the Kruskal-Wallis test was applied.
Results Qualitative CDS of lesions Four of 29 patients (14%) had no FS by CDS mapping: The underlying diseases were abscesses (n52) and lipoma (n52). The remaining 25 patients had FS which includes 13 patients with strictly chest wall lesion, and 12 patients with pulmonary extension.
Quantitative CDS of lesions Mean values of impedance indices of the ICA in the control group are shown in Figure 1. Impedance values were: control group mean RI 0.876 (¡0.056), mean PI 2.877 (¡0.643) (Figure 2); group I mean RI 0.793 (¡0.127), mean PI 1.929 (¡0.641) (Figure 3); group II mean ICA-RI 0.788 (¡0.144), mean non-ICA-RI 0.675 (¡0.446), mean ICA-PI 2.088 (¡1.015), mean non-ICAPI 2.520 (¡2.506) (Figure 4). In all 25 patients values of 304
Figure 1. Values of resistive-index (RI) and pulsatility-index (PI) (mean¡sd) of 37 spectral curves in 25 patients with chest wall lesions.
impedance indices of FS obtained within the chest wall did not differ from FS obtained in intercostal arteries of healthy volunteers (Figure 1). Within group II 12 non ICA-FS located within pulmonary tissue were obtained in 8 patients and could be discriminated from intercostal arterial supply by spectral curve analysis. Mean values of impedance indices of the ICA and non ICA in group II are shown in Figure 1.
Number of different FS in a lesion Four patients had no FS, 17 patients had one FS, 5 patients had two different FS, 2 patients had three different FS, 1 patient had four different FS (Figures 5– 7). In total 37 different FS were seen in 25 patients. 25 patients had ICA-FS, 12 patients had non-ICA-FS which included PA-FS (n55), BA-FS (n54) and TN-FS (n53).
Statistical analysis RI values with alpha correcture do not differ between the three groups (p50.037). The PI values do differ between the three groups (p50.002). The p-values for comparison of impedance measurements within group II were not significant (p50.55 for RI, p50.63 for PI).
Discussion Thoracic ultrasound is limited by complete sound reflection of the aerated lung. Nevertheless B-modeultrasound and qualitative CDS can provide additional information about the aetiology of based lesions diagnosed by chest radiography [1–6]. In the last decade several studies have shown a sensitivity and specificity of up to 95% for the differentiation of benign lesions from malignant lesions by impedance measurement of spectral curves [2, 4–6]. In these studies benign lesions such as pneumonia and atelectasis were characterized by a high impedance triphasic FS derived from PA. In malignant pulmonary lesions a low impedance monophasic FS was seen [2, 4–6]. Whereas some authors interpreted this FS as vessels of tumour neoangiogenesis (TN), a recent study identified the low impedant monophasic FS as BA by histological analysis [7]. Central BA form a vascular ring The British Journal of Radiology, April 2005
CDS of chest wall lesions
(a)
(b)
Figure 2. (a) Visualization of an intercostal artery in a healthy volunteer by power Doppler ultrasound. (b) A monophasic high impedant arterial flow pattern was seen suggesting an intercostal artery (ICA).
Figure 3. 56-year-old patient with lung
(a)
(a)
(b)
(b)
cancer and metastases within the chest wall. (a) On power Doppler ultrasound a large vessel was seen in the tumour which infiltrates the chest wall. The tumour involvement is bordered by a pleural effusion. (b) A monophasic high impedant arterial flow pattern was seen suggesting an intercostal artery (ICA).
(c)
Figure 4. 33-year-old patient with a primitive neuroectodermal tumour (PNET) of the chest wall with infiltration of pulmonary tissue. (a) On colour Doppler ultrasound a vessel was seen centrally in the tumour. (b) A monophasic high impedant arterial flow pattern was seen suggesting an intercostal artery (ICA). (c) In the periphery of the lesion a monophasic low impedant arterial flow pattern was seen suggesting a peripheral bronchial artery (BA) origin. The British Journal of Radiology, April 2005
305
C Go¨rg, T Bert, K Go¨rg and M Heinzel-Gutenbrunner
(a)
(b)
(c)
Figure 5. 34-year-old patient with testicular cancer and pleuropulmonary metastases, which infiltrate the chest wall. (a) On colour Doppler ultrasound mapping vessels were seen in the central region of the tumour. (b) A triphasic high impedant arterial flow pattern was seen suggesting a pulmonary artery (PA) origin. (c) Beneath the PA an additional flow signal (FS) characterized as a monophasic low impedant arterial flow pattern was seen suggesting a central bronchial artery (BA) origin.
Figure 6. (a) On colour Doppler ultra-
(a)
(b)
sound mapping small vessels were seen in the central region of the tumour (arrow). (b) A monophasic nearby constant flow pattern that lacks systolic– diastolic variation was seen suggesting tumour neoangiogenesis (TN).
Figure 7. (a) On colour Doppler ultra-
(a)
306
(b)
sound mapping small vessels originated in the chest wall were seen in the peripheral region of the tumour. (b) A monophasic high impedant arterial flow pattern was seen suggesting an intercostal artery (ICA) origin.
The British Journal of Radiology, April 2005
CDS of chest wall lesions
around the lung hilum from which branches centrifugally course along the bronchial tree. Peripheral BA lies within the interlobar and interlobular septa and supply the visceral pleura. In contrast to BA true vessels of TN are rarely found by CDS because of its slow flow [10]. They are characterized by a nearly constant flow that lacks systolic–diastolic variation [8]. In chest wall lesions this study characterized an additional arterial supply identified as ICA derived which could discriminate from PA, BA, and TN by spectral curve analysis. It should be emphasized that in almost 50% of pulmonary lesions various arterial FS can be identified [9]. Therefore a variable nutritional supply of the tumour is suspected corresponding to the complex arterial supply of the lung. In conclusion, quantitative CDS indicates a variable arterial supply of chest wall lesions which strongly depends on location of lesions. A single spectral analysis is not helpful for tissue characterization. Lesions strictly located to the chest wall had an ICA supply. Chest wall lesions with pulmonary extension demonstrated a complex arterial supply.
References 1. Gehmacher O, Mathis G, Kopf A, Scheier M. Ultrasound imaging of pneumonia. Ultrasound Med Biol 1995;21: 1119–22.
The British Journal of Radiology, April 2005
2. Yuan A, Chang DB, Yu CJ, Kuo SM, Luh KT, Yang PC. Color Doppler sonography of benign and malignant pulmonary masses. AJR Am J Roentgenol 1994;163:545–9. 3. Mathis G, Dirschmid K. Pulmonary infarction: sonographic appearance with pathologic correlation. Eur J Radiol 1993;17:170–4. 4. Yuan A, Yang PC, Lee L, Wu HD, Kuo SH, et al. Reactive pulmonary artery vasoconstriction in pulmonary consolidation by color Doppler ultrasonography. Ultrasound Med Biol 2000;26:49–56. 5. Civardi G, Fornari F, Cavanna L, Di Stasi M, Sbolli G, et al. Vascular signals from pleural-based lung lesions. Studied with pulsed Doppler ultrasonography. J Clin Ultrasound 1993; 21:617–22. 6. Hsu WH, Ikezoe J, Chen CY, Kean PC, Hsu CP, et al. Color Doppler ultrasound pulsatile flow signals of thoracic lesions: Comparison of lung cancer and benign lesions. Ultrasound Med Biol 1998;24:1087–95. 7. Hsu WH, Ikezoe J, Chen CY, Kean PC, Hsu CP, et al. Color Doppler ultrasound signals of thoracic lesions: Correlation with resected histologic specimens. Am J Respir Crit Care Med 1996;153:1938–51. 8. Babo VM, Mu¨ller KMG, Huzky A, et al. Die Bronchialarteriographie bei Erkrankungen der Lunge. Radiologie 1979; 19:506–13. 9. Go¨rg C, Seifart U, Go¨rg K, Zugmaier G. Color Doppler mapping of pulmonary lesions: evidence of dual arterial supply by spectralanalysis. J Ultrasound Med 2003;22:1033–9. 10. Harvey CJ, Albrecht T. Ultrasound of focal liver lesions. Eur Radiol 2001;11:1578–93.
307
