2MB Sizes 0 Downloads 46 Views


0033-8389100 $15.00

+ .OO


Advances in technology have greatly improved the imaging capabilities of ultrasound (US). This article summarizes the current applications of US as a diagnostic tool in chest diseases. By scanning through the acoustic window, US is a very reliable and efficient tool for evaluating lesions of the chest wall, pleural cavity, peridiaphragm, mediastinum, hilum, and peripheral lungs. A precise puncture transducer can be used to perform USguided transthoracic needle biopsy (TNB) with real-time visualization of the biopsy needle and the lesion. The accuracy of US-guided TNB for peripheral pulmonary nodules, chest wall lesions, and mediastinal tumors is 88% to 100%. US-guided TNB is also useful for histologic diagnosis of tumors causing superior vena cava (SVC) syndrome, Pancoast's tumors, pulmonary consolidation of unknown etiology, and tumors with obstructive pneumonitis. Moreover, transthoracic needle aspiration under US guidance can provide adequate specimens for microbiologic diagnosis of lung abscesses, necrotizing pneumonia, and parapneumonic effusions. Color Doppler imaging further extends the diagnostic spectrum of US, allowing the hemodynamics and neovascularization of a pulmonary lesion to be assessed noninvasively. Pulmonary arteriovenous malformations, pulmonary sequestration, and pulmonary infarcti~


ons can be diagnosed easily with color Doppler US. The color Doppler US puncture guiding device can improve the safety of USguided TNB by simultaneously displaying blood vessel information, the needle shaft, and the puncture route. US examination and US-guided TNB have become indispensable diagnostic techniques for various chest diseases. TNB with imaging guidance is a well-established technique for the diagnosis of focal pulmonary lesions.6,22 CT and biplane fluoroscopy are the most common imaging modalities used to guide TNB. With advances in imaging capabilities, biopsy techniques, and cytopathology, CT- and fluoroscopy-guided TNB can now provide extremely high diagnostic yields (sensitivity 80% to 95%) and are relatively safe." 21 These diagnostic modalities may be time-consuming, however, especially for small peripheral pulmonary lesions, and carry the risk of excessive exposure to radiation. Recently, rapid advances in transducer design, signal processing, and Doppler technology have greatly improved the imaging quality of US. With the development of a precise puncture-guiding device, US has proved to be a reliable, efficient, and informative imaging modality for evaluation of a wide variety of complicated clinical problems associated with ~

From the Department of Internal Medicine, National Taiwan University Hospital; and The Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan, Republic of China





chest diseases.12, 32, 33, 3541, It can also be used. effectively to guide TNB and other interventional procedures in the thorax.', 2, 5, s, 13, 19, ~ - 3 1 , HThis article presents an overview of the state-of-art applications of US-guided transthoracic biopsy in chest diseases including indications and contraindications, imaging techniques, biopsy procedures, diagnostic spectrum and sensitivity, potential complications, and advantages and limitations. THE ULTRASONIC WINDOW: A PREREQUISITE FOR US-GUIDED TNB

US evaluation of chest lesions is inherently difficult for two reasons. First, the air-containing lung parenchyma is a poor US transmitter because it reflects most of the US beam. Structures located deep within normally aerated lung may not be visualized with US. Second, the lungs are well concealed by the ribs, scapulae, and spine. The concept of the US window, however, has greatly expanded the applications of US in the diagnosis of chest diseases.35,38 The US window is created by consolidation of lung parenchyma or pleural effusion interposed between the lesion and the chest wall, which allows the US beam to penetrate and visualize lesions deep within the lung parenchyma. The range of chest diseases for which US may yield useful diagnostic information has expanded to include not only chest wall,' 14,17, 25 pleural,', 9, 15,24, 36, 43 and peripheral pulmonary lesions,1z,13, 19, 30, 38, 40, 47 but also mediastinal tumors,2O, 29, 31, pulmonary ~onsolidation,2~, 30, 35 lung tumors with obstructive pneumonia,'O, 39 and peridiaphragmatic lesion^.^ Only patients with lesions that have an US window, however, are suitable for US-guided TNB. INDICATIONS FOR CHEST US AND US-GUIDED TNB

US may be used to guide TNB for either fine-needle aspiration for cytologic and microbiologic analyses, or large-bore Tru-cut needle biopsy for histologic analyses. The indications for US examination and US-guided TNB of the chest are summarized below: 1. Identify chest wall tumors and guide biopsy 2. Determine the nature of pleural densities

3. Localize minimal or loculated pleural effusion for thoracentesis 4. Guide chest tube insertion for drainage of loculated empyema 5. Localize pleural tumors or pleural thickenings for guiding pleural biopsy 6. Evaluate the nature of peripheral pulmonary nodules and guide TNB 7. Evaluate pulmonary consolidation and identify parapneumonic effusion, necrotizing pneumonia, and tumors with obstructive pneumonitis 8. Localize tumors with obstructive pneumonia and guide a TNB 9. Localize the cavity of lung abscesses or necrotizing pneumonia and guide transthoracic needle aspiration for microbiologic diagnosis 10. Evaluate the nature of mediastinal lesions and guide TNB for mediastinal tumors 11. Evaluate the cause of SVC obstruction 12. Assist in the assessment of cervical lymphadenopathy in lung cancer patients 13. Assist in the imaging diagnosis and management of critically ill patients 14. Assess the vascularity and angiogenesis of pulmonary tumors by color Doppler US 15. Evaluate the hemodynamics of pulmonary lesions by color Doppler US 16. Guide and improve the safety of TNB by color Doppler US

US is particularly useful for localizing chest wall and rib tumors, guiding percutaneous biopsy,14,17, 25 determining the nature of pleural opacities, identifying minimal or loculated effusions for thoracentesis,l, 24, 36 and providing safe and accurate guidance for chest tube drainage of loculated empyema.32US evaluation and US-guided TNB are helpful for determining the origin of peripheral pulmonary tumors and mediastinal masses? 12, 13, 31, 47 and examining focal pleural thickening and pleural tumors.' US is also useful for evaluation of pulmonary consolidation of unknown 35 and for guiding needle aspiration of the cavity in necrotizing pneumonia or lung abscess for microbiologic diagnosis.37 For patients with central tumors producing obstructive pneumonitis, US-guided TNB of the lesion is a good alternative when fiberoptic bronchoscopy is not possible.10,39 US can also assist in the evaluation of cervical lymph node involvement in lung cancer patients? 293


The introduction of color Doppler imaging has further extended the diagnostic potential of US in chest 27 By combining high-resolution tissue imaging with simultaneous display of flow information and conventional Doppler spectral analysis, color Doppler US provides detailed, noninvasive assessment of morphology and function of lesions, as reflected by their blood supply and perfusion. The applications of color Doppler US have been expanded to the diagnosis of pulmonary arteriovenous malformations and pulmonary sequestration.ls.45 The "fluid color sign" can be used to identify minimal effusion that is amenable to a ~ p i r a t i o n .Color ~~ Doppler US can also be used to assist in the imaging diagnosis of pulmonary consolidations and to assess regional hemodynamic 46 changes within the consolidated Spectral analysis of the pulsatility index, the resistive index, the acceleration time, and the peak systolic velocity is helpful in understanding the pathophysiology of pulmonary c o n ~ o l i d a t i o n s .Color ~~ Doppler US also makes assessment of perfusion and reperfusion status of pulmonary infarctions possible:6 and can be applied to assess neovascularization, or angiogenesis, of pulmonary tumors?, Because low resistance flow signals and A-V shunting are common in malignant tumors, spectral analysis may be helpful in differentiating a malignant lung tumor from a benign one.44 CONTRAINDICATIONS FOR CHEST US AND US-GUIDED TNB The primary contraindications for USguided TNB of the chest include (1)uncooperative patients who are unable to remain still and control breathing or cough on command, (2) patients who have uncorrectable abnormalities of coagulopathy, (3)lesions suspected to be of a vascular nature, and (4) inability to visualize the lesion with chest US. Other relative contraindications include severe pulmonary hypertension, severe pulmonary emphysema, respiratory failure requiring mechanical ventilation, and recent myocardial infarction or unstable angina. Before conducting a TNB informed consent should be obtained and a routine blood coagulation profile should be available. Minor coagulation abnormalities are acceptable for aspiration biopsy of pleural or chest wall lesions where local compression can be applied, but coagu-


lopathy must be corrected before cutting biopsy of a mediastinal mass or pulmonary inflammatory lesion. EQUIPMENT AND TECHNIQUES FOR CHEST US EXAMINATION The US scanners suitable for thoracic US imaging include the Aloka SSD-630, SSD-650, and SSD-2000 (Tokyo, Japan); the Toshiba SSA-270 (Tokyo, Japan); the Diasonic VST Series (Botnell, CA); the ATL HDI series (Advanced Technology Laboratories, Washington); and equivalent US units equipped with 3.5-, 5-, 7.5-, or 10-MHz linear, convex, and sector transducers. The ATL HDI series, Diasonic VST series, and Aloka SSD-2000 units also provide color Doppler US and amplitude US angiography software that can be used to detect blood flow signals of lesions. Patients are scanned in a supine or prone position by an intercostal approach. A watersoluble US transmission medium (ultraphonic conductivity gel, Pharmaceutical Innovations, Inc., Newark, NJ) is applied to the skin as a coupling medium. The gray-scale, real-time US scan is usually performed first to locate the lesion. For examination of near lesions, such as pleural or chest wall tumors, a higher frequency (5 or 7.5 MHz) linear or convex transducer provides better resolution of the lesion. A linear or convex transducer usually has a broad view of the near field, and is better than a sector scanner for evaluating pleural lesions. A 3.5-MHz transducer is more suitable for visualization of deeper lesions. For lesions with a small US window or very narrow intercostal space, a sector transducer is preferred. The liver and fasted gallbladder are used as tissue-texture references for solid and fluid-containing regions: the echogenicity of the lesion is compared with that of the liver and defined as hypoechoic, isoechoic, or hyperechoic. Normal areas are scanned for control comparisons. The US images, recorded on Polaroid film (Polaroid, Cambridge, MA), are analyzed to determine the size, shape, echogenicity, and location of the lesion, and the presence of air or fluid bronchogram, abscess cavity, pleural effusion, pleural tumor, pleural thickening, and respiratory movement of lesions. US-GUIDED TNB TECHNIQUE There are three kinds of approaches for US to guide percutaneous biopsy31: (1) sono-



graphic transducers with built-in needle slots within the central or side portion of the transducer that direct the needle at a predetermined angle within the plane of view of the transducers; (2) attachable stretcher guides that can be fitted to existing transducers, directing the needle to various depths from the transducer, depending on the angle set by the operator; and (3) the needle can be inserted through the skin directly into the plane of view of the transducer by a freehand approach, without a guide. For guiding percutaneous TNB, the transducer with built-in needle slots and adjustable needle-guiding channel is preferred, because it can precisely guide the needle to the target. Small lesions and deeply seated lung tumors can be reached easily with this technique. The attachable stretcher guides and the freehand approach are only suitable for superficial or large thoracic masses. For small tumors with small US windows, the needle may have to go through the normal aerated lung during the biopsy, and the needle tip might not be visualized on US. These factors may increase the complication rate. Patients referred for chest US study are carefully assessed for the location, sonographic pattern, and blood flow information of the tumor and exclusion of vascularization of the lesion by gray-scale US and Doppler US. Before TNB, the skin is prepared and a local anesthetic agent is applied. We routinely use a sterile puncture transducer (UST-507 BP Aloka, Tokyo, Japan) with a preset puncture area and a guiding channel is used to localize the lesion. The puncture area can also be displayed on the monitor. The biopsy route may be adjusted at a fixed angle from -5 to 25 degrees. An aspiration needle with an outer sheath and inner stylet is connected to an aspirator (Fig. l), and the needle is then inserted through the guiding channel and advanced to the lesion under real-time image monitoring. The needle tip appears as a bright spot in real-time, gray-scale imaging. For safety reasons, the aspiration routes are chosen to avoid penetration of the aerated lung, great vessels, and major bronchi. Patients are asked to hold their breath during the procedure-usually 15 to 20 seconds for transthoracic needle aspiration or biopsy of a parenchymal lesion. After transthoracic needle aspiration, the patient is observed for 1 hour and routine chest radiographs are taken the next day to assess potential complications.


Transducer Chest Wall

Pleural Line

Figure 1. US-guided transthoracic needle aspiration biopsy for a peripheral lung cancer. The tumor appears as a well-defined, hypoechoic nodule (arrowheads). The pleural line is interrupted. The needle is inserted through a guiding channel of the puncture probe at fixed angle, and the needle tip is directed to the tumor part under US guidance. The needle tip may be seen as a whitish spot on the real-time monitoring screen. A 20-mL syringe with needle holder or aspirator is attached to the needle and aspiration biopsy performed. L = lung; T = tumor.

We routinely perform fine-needle aspiration first with a 20-, 21-, or 22-gauge needle. These aspiration needles contain an outer sheath and an inner stylet. The aspirated materials are sprayed onto two sets of glass slides: one set of slides is air dried and the other is fixed in 95% alcohol. Papanicolaou’s stain is used on alcohol-fixed slides. Liu’s stain1*and Gram’s stain are used on air-dried slides. Liu’s staining for cytologic examination is a very convenient method that takes only 3 minutes to complete and can be done at bedside.30A cytopathologist can immediately assess the adequacy of the specimen and decide if a second pass or cutting biopsy is necessary. If the specimens are judged inade-


quate or inconclusive by a cytopathologist, Tru-cut biopsy with a 16-, 18-, or 20-gauge needle (Top Surgical, Tokyo, Japan) is performed to obtain a tissue core for histologic examination. The specimen obtained by Tmcut needle biopsy is about 10 to 15 X 2 to 3 mm, which is usually adequate for conventional histologic examination. COLOR DOPPLER IMAGING AND TNB USING A COLOR PUNCTURE GUIDING DEVICE

Color Doppler US is a technique for displaying blood flow information, which is superimposed in color on the two-dimensional, 27 Color real-time, gray-scale US image.26* Doppler imaging is a measurement of mean frequency shift and is angle-dependent. The Doppler shift is greatest when the reflector is moving in parallel with the transducer. The Doppler shift is positive when the reflector is moving toward the transducer, and negative when the reflector is moving away from the transducer. When the reflector is moving perpendicular to the beam axis, there is no Doppler shift. Because color Doppler US uses a pulsed-Doppler technique, it must use pulserepetition frequency. If the speed of the moving blood is faster than half of the pulserepetition frequency, a Doppler artifact called aliasing may occur. The Doppler spectral analysis data can be autocorrected for angles and converted to velocities, and provide information on blood flow and the real-time, twodimensional blood vessel distribution. When a vascular lesion is highly suspected or a lesion is close to the great vessels (e.g., a tumor located near the mediastinum or causing SVC obstruction), color Doppler imaging or amplitude US angiography is used to detect blood flow signals within or surrounding the lesion before TNB is attempted (Fig. 2).8, 19, 26 The color Doppler setting of the US unit is adjusted to a higher sensitivity range to detect low-velocity blood flow (low-velocity scale, 0.26 m/s for a Doppler angle of 0 to 180 degrees). The wall filter is set to a level that minimizes the rejection of small frequency shifts (low velocity flow), and still avoids interference from respiratory movement and the heartbeat (wall filter range, 70 to 120 kHz). The color Doppler gain is increased until the background noise appears as a colored “snowstorm” across the image, and then is backed off until only a few ran-


dom speckles remain visible. The steering angle of the color Doppler window is adjusted to - 45 or 45 degrees to avoid falsenegative results, which frequently occur when the US beam is perpendicular to the flow signal. The blood flow in a vessel appears as a persistent area of color signal with a tubular, curvilinear, or branching distribution on real-time images. Spectral wave analysis with pulsed Doppler can be used to confirm the vascularity of the lesion. A new color puncture transducer (Aloka UST 5045 P-3.5, Aloka) can be used to guide TNB of highly vascular peripheral lung lethis new color Doppler US s i o n ~By . ~ using ~ puncture device, vascular structures surrounding or within the target lesion can be verified clearly in real-time during biopsy. Visualization of the needle shaft or tip is also better than with the conventional gray-scale puncture-guiding device.



Diagnosis of Chest Wall Tumors Sonography of the chest wall usually shows soft tissue echogenicity with multiple layers of muscle and fascia (Fig. 3). When a transducer is oriented perpendicular to the intercostal spaces, normal ribs may appear as rounded, echogenic interfaces with prominent acoustic shadowing. Lesions of the chest wall may originate in the soft tissue or bony thorax. Malignant tumors, benign tumors, and infections can all appear as chest wall masses. Because there is no intervening gas, US is very useful for assessing such lesions.17,25 US can demonstrate bone lesions in which soft tissue or fluid replaces bone. Chest wall tumors usually appear as welldefined, hypoechoic masses within the soft tissue layer of the chest wall. Scattered areas of remaining destructive bone may be seen within the lesions (Fig. 4).14,15, 17* 25 Higherfrequency US (7.5 or 10 MHz) can help distinguish subtle differences in acoustic impedance, allow soft tissue tumors and osteolytic bone lesions to be recognized easily, and delineate the relationships of tumors to ribs, pleura, and surrounding structures. US with a high-frequency probe can also easily differentiate chest wall tumors from pleural lesions On US, an osteolytic bone lesion may appear as an eccentric hyperechoic platelike shadow



Figure 2. Color Doppler US image of a 17-year-oldwith anterior mediastinal tumor causing SVC syndrome. The tumor appears hypoechoic in density and with cystic change (arrowheads), locating just anterior to SVC and right atrium. Color Doppler US clearly can demonstrate the vascular structures surrounding the mediastinal tumor. SVC = superior vena cava. RA = right atrium.

Figure 8. A and B, See page 334. C,The cytology obtained from USguided fine-needle aspiration revealed adenocarcinoma (Liu’s stain, original magnification x 400).


Figure 3. Sonographic images of normal pleura and chest wall by 5- to 13-MHz linear array transducer. A, A transverse image through intercostal space reveals multiple layers of muscles and fascia. The pleural lines appear as an echogenic bright line. The visceral pleura and parietal pleura line may be seen to glide with each other during respiration. Pp = parietal pleura; Pv = visceral pleura. 6, Longitudinal image across the rib of chest wall. A normal rib seen as a hyperechoic, curved linear structure (arrowheads), with prominent acoustic shadow beneath the rib. L = lung.

Figure 4. A 66-year-old man with lung cancer metastasis to the rib. A, CT scan shows a rib tumor (arrowhead) with bony destruction and calcification. 6, Sonographic image with 7.5-MHz linear scanner reveals a hypoechoic mass with ossification within the lesion, which causes irregular acoustic shadow (arrowheads) distal to the calcifications. US guided transthoracic needle biopsy showed metastatic adenocarcinoma. P = pleural line; T = tumor; L = lung.




inside a hypoechoic tumor, a round hyperechoic shadow or ring in the center of a hypoechoic tumor, or a fusiform hypoechoic tumor (Fig. 4). The sensitivity of US-guided needle biopsy for diagnosis of chest wall lesions is 87.5% to loo%, and complications are minima1.25 Evaluation of Pleural Lesions: Guided Thoracentesis and Pleural Biopsy

US can accurately demonstrate pleural lesions. The parietal pleura lining the bony thorax and the visceral pleura covering the lung are seen as two thin, bright echogenic pleural lines just beneath the chest wall (Fig. 3). The pleural lines may glide with each other during respiratory movement. The underlying air-filled lung is a highly reflective interface that may block transmission of US into the lung parenchyma. The US image displays a pattern of repeated bright echoes caused by an acoustic reverberation artifact. These echoes are bright but formless, and diminish rapidly in intensity with increasing distance from the transducer. The US image of pleural effusion is characterized by an echo-free space between visceral and parietal pleura; this space may change shape with respiration (Fig. 5).%Pleural thickenings are defined as focal echogenic lesions greater than 3 rnm in diameter with irregular margins, arising from visceral or parietal pleura. Pleural tumors are

well-defined, hypoechoic, or echogenic solid nodular lesions located in the parietal or visceral pleura. There are four basic patterns of pleural effusion (Fig. 6): (1) anechoic, (2) complex nonseptated, (3) complex septated, and (4) homogeneously e ~ h o g e n i c .The ~~ echogenic materials or fibrin strands may float inside the pleural effusion. The value of US for evaluation of pleural diseases is well documented.', 28, 36, 40* US is very helpful for determining the nature of pleural opacity, identifying minimal and loculated effusion, and guiding thoracente~is.~, 24, 27 It is particularly useful when interpretation of chest radiographs is difficult; when there is question of whether fluid is subpulmonic or subphrenic; and when the patient is bedridden, unconscious, or unable to sit in an upright position. US can also identify the largest and most accessible area of fluid accumulation for thoracentesis. Thoracentesis can be carried out either (1) by making a mark on the skin and measuring the depth for the needle to penetrate, or (2) under direct vision with real-time US guidance. US-guided thoracentesis has a high success rate (about 97%).32It is particularly helpful for critically ill patients when tedious radiographic study is not possible and safe thoracentesis is mandat~ry.~' The sonographic pattern of effusion is also helpful for differentiating transudate from exudate: transudate is usually anechoic, whereas exudate can be anechoic, complex, or echogenic. Echogenic,

Figure 5. US image of a patient with minimal subpulmonic effusion seen as an echo-free space between visceral pleura, parietal pleura, and diaphragm. This echo-free space may change shape with respiration. D = diaphragm; L = lung; E = effusion.



Figure 6. Four basic US patterns of pleural effusion. A, Anechoic. B, Complex nonsepated. The hyperechoic spots (arrowheads) may move with respiration. These hyperechoic spots may indicate tissue debris or blood cell aggregates within the effusion. C, Complex sepated. Arrowheads indicate the fibrin septa. D, Homogeneously echogenic. E = effusion; D = diaphragm; L = lung.

complex septated, or complex nonseptated effusion are usually Although US is very powerful in the evaluation of pleural opacities, it has some limitations. First, differentiation of minimal effusion from pleural thickening may be difficult at times. Because both lesions appear anechoic on gray-scale US, predicting whether an echo-free or complex-appearing lesion is amenable to thoracentesis is not always possible. Although the criteria suggested by Laing and Filly9 (i.e., that effusion changes shape with respiration and has movable septa) are useful in predicting the presence of pleural fluid, some pleural lesions do not change shape with respiration or have movable septa, but are still amenable to aspiration. We recently observed that true fluid in cases of minimal effusion may generate a pseudo

color flow pattern during respiratory and cardiac cycles, and may display a turbulent color signal on Doppler imaging. This fluid color sign is a useful aid in predicting the presence of fluid in an echo-free space.z4 Thoracic US has another advantage in guiding pleural biopsy of focal pleural lesions, in that it is particularly helpful for patients with minimal pleural effusion or even without pleural effusion. Conventional closed pleural biopsy with the Cope or Abram needle must be carried out in the presence of pleural effusion or pneumothorax, and thickened pleura or pleural tumors can be clearly identified if US is used to guide the biopsy. A pleural biopsy can pinpoint the area with US abnormalities, and the needle does not overpenetrate the underlying lung parenchyma with real-time US monitoring. Because



the pleural disease involvement may be focal, US guidance of the needle biopsy can also inc2ease the chances of obtaining tissue with significant pathologic changes.' US Evaluation of Peridiaphragmatic Lesions

Radiographic peridiaphragmatic lesions and elevation of the hemidiaphragm of unknown etiology are usually diagnostic problems for chest physicians and radiologists. Three main types of lesions (supradiaphragmatic, diaphragmatic, and infradiaphragmatic) can lead to such abnormalities, and these lesions have multiple causes in many patients. In a prospective study of 56 patients,' we demonstrated that chest US and US-guided TNB are very useful imaging tools for diagnosis of peridiaphragmatic lesions. The pathology around the diaphragm can be clearly elucidated by using the liver and spleen as US windows. By defining the position of the diaphragm, differentiation of subphrenic fluid collection and pleural effusion is relatively easy. The real-time visualization of diaphragm motion allows differentiation between diaphragm palsy and eventration. US can also identify pulmonary tumors, subphrenic abscesses, and hepatic or splenic masses located in the peridiaphragmatic area. Diagnosis of Peripheral Lung Tumors

Peripheral lung tumors appear on US as well-defined, hypoechoic, or echogenic homogeneous densities with posterior echo enhancement (Fig. 7)." 9, l7 The pleural line may be interrupted if the tumor extends to pleura, and the echogenicity of the tumor increases with tumor size. US can help determine the nature of pe-

ripheral lung tumors and guide TNB to establish an etiologic diagn~sis.~', 34, 38, 47 Although TNB under fluoroscopy or CT guidance is still the standard diagnostic approach for peripheral lung lesions not accessible by fiberoptic bronchoscopy, small peripheral lesions in the low lung field may not be assessed easily with these techniques. These lesions move easily during respiration and may be obscured by the rib cage, making biopsy difficult without real-time image guidance. US can provide real-time image monitoring and biopsy can be carried out in the respiratory phase in which the nodule is most accessible and e~ident.4~ The needle tip and sometimes the needle shaft can be seen as an echogenic white spot on the monitor. The route of the needle can be selected to avoid puncturing the aerating lung or vital great vessels, and the depth of penetration can be selected and monitored with strict control. Our previous study showed that the sensitivity for diagnosis of peripherally located lung cancer was 96.8% and the accuracy was 97.5%.9Lung tumors that do not attach to the pleura but have accessible US windows can also undergo biopsy under US guidance.34 The location, depth, and size of the tumor do not influence the results of the needle bi0psy.4~Locating a tumor as small as 1 cm in diameter is possible, and an adequate specimen can be obtained easily with this approach (Fig. 8). TNB of large lung masses is not difficult with fluoroscopic guidance. Large lung masses, however, frequently have areas of central necrosis so extensive that only a small part of the tumor remains viable. A biopsy specimen from a central necrotic part of a malignant lung mass may yield a false-negative result. US can improve the sensitivity of needle biopsy of malignant lung masses with central necr~sis,'~ by demonstrating the viable portion of the wall and directing the needle to the mural part of the tumor.

Figure 7. A, A 75-year-old man has multiple lung nodules on chest radiography (arrowhead). Metastatic lung cancer was suspected. Fiberoptic bronchoscopy failed to detect these lesions. B, CT scan of chest reveals two small lung nodules (arrowheads). C, US examination of the posterior left chest shows a small hypoechoic subpleural nodule (arrowheads) with posterior echo enhancement representing the peripheral lesion in the superior segment of the left lower lobe. D, US-guided transthoracic needle biopsy obtained a 2-mm x 8-mm specimen; histology revealed a granuloma with multinucleatedgiant cells and fibrosis (hematoxyn 13eosion, original magnification x 200). The acid-faststain was positive and patient respondedwell to anti-tuberculoustreatment. PEE = posterior echo enhancement; P = pleural line.


Figure 7. See legend on opposite page




Figure 8. An 80-year-old man had multiple small lung nodules and weight loss. A, CT scan of chest shows two small subpleural densities at right costophrenic area (arrowhead). B, US examination shows a small subpleural nodule (arrowheads) with visceral pleural interruption and posterior echo enhancement. The hyperechoic densities within the nodule may indicate air densities or a cavity within the nodule. The fiberoptic bronchoscopy of this patient was negative. C,See page 328. The cytology obtained from US-guidedfine-needle aspiration revealed adenocarcinoma (Liu’s stain, original magnification x 400). E = posterior echo enhancement; P = pleural line.

Lung Tumor Causing Obstructive Pneumonitis

Obstructive pneumonitis is a common finding in lung cancer. Although the Golden S sign is a useful radiographic indicator of lung carcinoma as the cause of obstructive pneumonitis, differentiation between tumor tissue and the consolidated lung on the basis of the conventional chest radiographs is not always possible. Fiberoptic bronchoscopy is the standard diagnostic approach for patients with lung tumors associated with obstructive pneumonitis. If the obstructing lesions are intraluminal or intramural and biopsy can be performed directly by fiberoptic bronchoscopy the diagnostic success rate is high. If the obstructing lesions are extraluminal and fiberoptic bronchoscopy shows only external compression, however, satisfactory diagnostic material may not be obtained with bronchoscopic biopsy or transbronchial needle aspiration. In such cases, CT-guided aspiration biopsy or thoracotomy may be necessary to establish the histologic diagnosis. Contrastenhanced CT can delineate the obstructing tumor and consolidated lung for guiding transthoracic biopsy. Lung tumors causing obstruction, however, may not be in the crosssectional plane of the CT scan. In CT-guided transthoracic aspiration biopsy, the needle

may sometimes have to pass through aerated lung, increasing the risk of complications. This difficulty is not encountered with US, which permits a tangential scan and biopsy in any direction under real-time image guida n ~ e The . ~ ~tumor margin and the consolidated lung can be demonstrated clearly. Although some obstructing tumors are deeply seated and located near the hilum, specimens sufficient for histologic diagnosis can still be obtained easily. The diagnostic yield of USguided needle biopsy of tumors associated with obstructive pneumonitis is 94.5% and the accuracy is 95.20h.31, 39

Diagnosis of Pancoast’s Tumor

Pancoast’s tumor is a special type of peripherally located lung carcinoma that is confined to the superior pulmonary sulcus and causes a characteristic syndrome by local invasion to adjacent structures, such as the brachial plexus, ribs, sympathetic chains, and vertebrae. Because of their peripheral and apical location, these lesions cannot be visualized with fiberoptic bronchoscopy; CT-guided needle biopsy or mediastinoscopy is usually needed to obtain specimens for histologic diagnosis. US has recently been found useful for evaluating the local extent of tumor


involvement and guiding the needle directly to the apical lung tumor for histologic diagnosis.16Pleural and extrapleural extension can also be delineated clearly by US. A sector scan through the supraclavicular approach with a 3.5- or 5-MHz probe is recommended for detailed evaluation of these lesions. Because great vessels may be present in the surrounding area, Doppler US or color Doppler US provides more useful vascular information prior to needle biopsy. The diagnostic yield of US-guided TNB for this particular type of lung tumor is 91%.33 Histologic Diagnosis of SVC Syndrome

SVC syndrome has been considered an acute or subacute oncologic emergency in which diagnostic procedures may carry a high risk. Palliative radiotherapy for symptomatic relief is often initiated even before a definite histologic diagnosis is established. Improvements in chemotherapy and surgical techniques have led to prolonged survival among patients with SVC syndrome. The prognosis is greatly influenced by a definite histologic diagnosis of the tumor causing the obstruction. Because of the extensive collateral circulation, conventional invasive diagnostic approaches, such as bronchoscopic biopsy, mediastinoscopic biopsy, and TNB under fluoroscopic or CT guidance, may carry high risks of bleeding, respiratory distress, and airway compression. In a recent study of 40 patients, we demonstrated that US with color Doppler imaging is a very useful diagnostic tool for patients with SVC syndrome.8 Great vessels, collaterals, and tumor vessels can be demonstrated clearly, and the biopsy route can be preselected to avoid puncturing great vessels. US-guided TNB allows easy, safe, and rapid histologic diagnosis. The diagnostic yield in our previous study was 83.3% and the mean duration from admission to histologic diagnosis was 2.1 days.8 Specific therapy could be initiated without delay. Assessment of Pleural and Chest Wall Invasion by Lung Cancer

Evaluation of the extent of tumor invasion to the pleura and chest wall is important and may influence the subsequent treatment of patients with lung cancer. CT scan is recom-


mended for determining the extent of pleural and chest wall invasion in lung cancer patients. Recently, high-resolution, real-time US, particularly with higher frequency (7.5 to 10 MHz) scanning probes, has been found useful for evaluating tumor invasion of the pleura and chest wall. Sugama et all5 defined the US criteria for pleura or chest wall invasion as follows (Fig. 9): ultrasound pattern (UP) 0 (UPO) indicates that the tumor has not involved the visceral pleura and cannot be visualized; UP1 indicates that the tumor is in contact with the visceral pleura, but the visceral and parietal pleural surfaces are intact, and smooth respiratory movement of the tumor is visible; UP2 indicates that the tumor extends beyond the visceral pleura and is in contact with the parietal pleura, and the respiratory movement of the tumor is disturbed; UP3 indicates that the tumor has extended to the chest wall through the visceral and parietal pleurae. The parietal pleural line is interrupted and there is no respiratory movement. US is also more accurate than CT scan for determining tumor invasion of the chest wa11.16 US-Guided TNB of Mediastinal Lesions

Accurate histologic diagnosis of a mediastinal mass is a cornerstone for planning appropriate treatment. Fluoroscopy and CT are usually used to guide TNB of mediastinal tumors, with high diagnostic yields. Large mediastinal arteries, however, may be accidentally punctured during fluoroscopyguided TNB. Since 1980, CT has largely replaced fluoroscopy for guiding mediastinal biopsy. CT has the advantage of accurate localization of the lesion and, with contrast enhancement, can show the relationship of the lesion to the great vessels. The hazard of puncturing great vessels in the mediastinum is avoided. Recently, US was also proved to be a reliable tool for evaluating mediastinal lesions. TNB under US guidance is a safe means of obtaining adequate specimens for 29, 34, 4~ definite histologic diagnosis (Fig. The color Doppler puncture device can further improve the safety of mediastinal biopsy by demonstrating the great vessels in realtime during the b i ~ p s y . ' ~ Mediastinal lesions can be assessed by means of US through a suprasternal or parasternal approach. A suprasternal or supraclavicular approach through the paratracheal



Figure 9. US pattern (UP) of peripheral lung tumor with pleura or chest wall invasion. A, UP0 indicates that subpleural tumor has not invaded the visceral pleura. The visceral pleural line is not interrupted. The posterior echo enhancement is still visible. B, UP1 indicates that the tumor is in contact with the visceral pleura, but the visceral and parietal pleural surfaces are intact, and smooth respiratory movement of the tumor still is visible. C, UP2 indicates that the tumor extends beyond the visceral pleura and is in contact with the parietal pleura. The respiratory movement of the tumor is disturbed. D,UP3 indicates that the tumor has extended to the chest wall through the visceral and parietal pleura (arrowheads). The parietal pleural line is interrupted and there is no respiratory movement. T = tumor; P = pleural line; Pp = parietal pleura; Pv = visceral pleura; E = posterior echo enhancement.



Figure 10. A, A 74-year-old woman with anterior mediastinal tumor (arrowhead) as shown on CT. 6, US examination through a parasternal approach reveals a hypoechoic lobulated mass (arrowheads) just beside the manubrium. Histology of US-guided transthoracic needle biopsy showed malignant thymoma. T = tumor: M = manubrium.

soft tissue space can be used to detect tumors located in the upper mediastinum.20Usually, patients lay supine with a pillow below the scapulas to keep the neck maximally extended. The US transducer scans downward through the supraclavicular fossa or suprasternal notch, along the paratracheal soft tissue space. The supra-aortic and parasternal regions and the aortopulmonary window may be seen. Mediastinal lesions, such as thymic tumors, lymphomas, dermoid cysts, germ cell tumors, and aneurysms of great vessels, can be detected with this approach. Most thymomas are well defined and hypoechoic, and the capsule can be well demonstrated. In patients with invasive thymoma, disruption of the capsule or invasion to the heart or pericardium can be identified with US. Lymphoma

or germ cell tumors may exhibit cystic change and heterogeneous echogenicity.Several studies in the last decade on US-guided TNB of mediastinal tumors have demonstrated high diagnostic yields and safety with both cytologic and histologic procedures. The diagnostic yield of core needle biopsy (84% to 100%) is superior to that of fine-needle aspiration biopsy (45% to 78%).12,20, 29, 34, 42 Etiologic Diagnosis of Pulmonary Consolidation

Pulmonary consolidation has many causes. Although pneumonia is the most common, noninfectious diseases, such as lymphoma, infarction, bronchioloalveolar carcinoma, and



vasculitis, may also present as pulmonary consolidation. Conventional chest radiographs offer little information on the underlying etiology or morphologic details of pulmonary consolidation, but US has proved useful? Lung consolidation in US is defined as an isoechoic or hypoechoic area with a triangular shape that moves with respiration (Fig. ll).” An air bronchogram is identified as bifurcating hyperechoic lines arising from the hilurn?” 35 and the hyperechoic air densities can move with respiration. A fluid bronchogram is defined as a branching, hyperechoic, tubular structure with an anechoic interior lumen and no blood flow demonstrated on Doppler US.1o,39 Fluid bronchograms are usually observed within areas of lung consolidation caused by central airway obstruction. Parapneumonic effusion, necrotizing pneumonia, and tumors with obstructive pneumonia can be clearly identified, and major bronchi and pulmonary vessels are easily differentiated with high-resolution, real-time, gray-scale US and Doppler US. US also provides safe and successful thoracentesis, guided needle aspiration of parapneumonic effusions for microbiologic study in patients with necrotizing pneumonia, and guidance for cutting biopsy diagnosis of pulmonary The maconsolidation of unknown jor bronchi and vessels can be identified with high-resolution US in conjunction with Dopp-

ler US, and the biopsy route can be selected to avoid injury to these structures. In a study of 30 patients with pulmonary consolidation of unknown etiology, a diagnostic yield of 93%was obtained with Tru-cut biopsy.3O Only two patients had minor complications (pneumothorax and mild hemoptysis). This technique may be particularly useful for immunocompromized patients with pulmonary consolidation in whom conventional diagnostic approaches have failed. Microbiologic Diagnosis of Lung Abscesses

Microbiologic diagnosis of lung abscesses is difficult. Expectorated sputum and bronchoscopic aspirates are inappropriate specimens for bacteriologic culture of lung abscesses. Polymicrobial infection is very common in lung abscesses, however, and accurate microbiologic diagnosis would optimize the selection of antibiotics and influence treatment response. US is useful for evaluating lung abscesses and guiding transthoracic needle aspiration, which is the best way of obtaining reliable, uncontaminated abscess specimens for microbiologic culture and diagn ~ s i s An . ~ ~abscess cavity appears on US as an ovoid or an irregularly shaped hypoechoic or anechoic lesion with irregular outer mar-

Figure 11. US image of pulmonary consolidation appears as a wedgeshaped hypoechoic lesion with bifurcating hyperechoic lines (arrowheads) arising from the hilum. The hyperechoic lines are air densities within the airways of consolidated lung (air-bronchogram). These hyperechoic lines can move with the respiration.


gins37;the internal echotextures are heterogeneous. Air-fluid level can be demonstrated if the patient is scanned in the sitting position. The air portion is depicted as a hyperechoic area in the upper part of the lesion and casts a characteristic acoustic shadow below the lesion, whereas the fluid portion is heterogeneous and echogenic. High-resolution, realtime US can identify the area of visceral pleural involvement and the area where the visceral pleura adheres to the parietal pleura. Transthoracic needle aspiration can be performed safely through this area and prevents pneumothorax and spillage of abscess material into the pleural cavity. US can also provide precise guidance of the needle tip to the fluid portion of an abscess cavity and ensure aspiration of an adequate amount of fluid for bacterial culture. In our previous study, the success rate was 94% and transthoracic aspirates were found to provide far better material for culture than sputum, blood, or bronchoalveolar lavage. A total of 65 pathogens were isolated from 31 aspirates; only 3% of pathogens could be recovered from blood culture, 11%from sputum culture, and 3% from bronchoalveolar lavage.37US-guided transthoracic needle aspiration enables accurate bacteriologic diagnosis of lung abscesses, which is helpful in directing treatment, especially in immunocompromized patients with polymicrobial infections.


teristically presents with US features of dilated and irregular fluid-filled bronchogram, and some lesions show cystic changes and microabscess formation. The presence of a dilated, irregular fluid bronchogram may indicate a distended airway (bronchiectasis with infected intraluminal secretions), from which puslike needle aspirates can usually be obtained. Culture from transthoracic needle aspirates can identify clinically significant pathogens in 78% of febrile patients with obstructive pneumonitis.'O

US Guidance of lnterventional Procedure for Critically 111 Patients

US is useful for diagnosis of chest diseases in critically ill patients. Portable chest radiographs often fail to depict thoracic diseases clearly, whereas US provides clear images of pleural and peripheral lung lesions. Because US can be performed at the bedside, patients can be examined anywhere in the ward without the need for transportation of life support devices. Interventional procedures including thoracentesis; drainage of effusion, empyema, or abscess; and TNB can be performed accurately and safely at bedside. Our previous study indicated that thoracic US is a useful adjunct to portable chest film in making a correct diagnosis, and has significant influence on treatment planning for seriously ill patients in the intensive care unit.41

Microbiologic Diagnosis of Obstructive Pneumonitis

US has also proved useful for microbiologic evaluation of obstructive pneumonitis,'O the radiologic opacity that develops distal to an obstructing endobronchial lesion. Obstructive pneumonitis is actually a combination of atelectasis, bronchiectasis with mucous plugging, and true parenchymal inflammation. The pathogens causing obstructive pneumonitis are very heterogeneous and polymicrobial infection is common. The presence of infection usually cannot be confirmed on the basis of radiographic findings alone, and microbiologic diagnosis is also difficult because infectious lesions distal to the obstructive pneumonitis may not have direct communication with the major airways. Sputum cultures, transtracheal aspirates, and protected sheath brushings do not reliably yield clinically significant pathogens. Infectious obstructive pneumonitis charac-

Endoluminal US and US-Guided Transesophageal or Transbronchial Biopsy Endoluminal US through an esophageal or transbronchial approach allows evaluation of centrally located pulmonary lesions without accessible US windows, and those that might not be detected with transthoracic US. This technique is also useful for guiding needle aspiration biopsy of such le~ions.~, 23 Transesophageal US is a powerful tool for investigating aortopulmonary window lesions and lesions adjacent to the heart,23and provides valuable information on centrally located tumors with cardiac invasion or mediastinal involvement. Transesophageal US can be used to identify mediastinal lymph nodes and masses and to guide TNB. The use of endobronchial US with a catheter-based, miniaturized US transducer in-



serted through the biopsy channel of a bronchoscope was first described by Hurter and Hanrath4 in 1990. A high-frequency transducer of up to 20 MHz provides excellent imaging of the bronchial wall and lesions adjacent to the airway. Endobronchial US can also be used to localize peripheral tumors (Fig. 12) and evaluate endobronchial obstruction caused by extraluminal lesions, and may help identify large vessels near tumors. Endobronchial US can also guide transbronchial biopsy of peripheral lung tumors that cannot be visualized on bronchoscopy. RELIABILITY OF US-GUIDED TRANSTHORACIC NEEDLE ASPIRATION OR BIOPSY The reliability of fine-needle aspiration specimens for determining cell type in lung

cancer is of major importance for patient treatment. Most studies of US-guided transthoracic needle aspiration have used fine needles (smaller than 19 gauge). Although fineneedle aspiration is safe and can achieve a high diagnostic yield, several problems arise. Some investigators have suggested that the small tissue fragments obtained with fineneedle aspiration are adequate for both cytologic and histologic examinations but are not sufficient for detailed histologic studies. The reported agreement rates between the results of fine-needle aspiration, cytologic examination, and final histologic examination range from 60% to 90%.22In most cases, cytologic examination allows differentiation between small cell and non-small cell carcinoma. Fine-needle aspiration samples, however, are inadequate for cell type determi-

Figure 12. A, A catheter-based miniaturized US transducer is inserted through the biopsy channel of bronchoscopefor endobronchiol US examination. B, A 70year-old man with a tumor about 2 cm x 3 cm in size in right upper lobe of lung. Fiberoptic bronchoscopy failed to detect this lesion. Endobronchial US examination shows an irregular hypoechoic tumor (arrowheads) at one of the subsegmental bronchi of right upper lobe. The catheter-based US transducer was within a bronchus at the center of the tumor. The walls of the bronchus are not visible because of the tumor invasion. The endobronchial, US-guided, bronchoscopic biopsy at this level reveals adenocarcinoma. P = US transducer probe; T = tumor.


nation in 20% to 40% of cases. Another concern is the reliability of nonspecific, benign, or negative results. Although multiple passes may improve the diagnostic yield of malignant lesions by 35% to 45%, a negative result does not exclude malignancy.21In a comparative study,34we demonstrated that transthoracic large-bore Tru-cut biopsy under US guidance can be as safe as US-guided fineneedle aspiration, and that the diagnostic accuracy is significantly higher than that of fineneedle aspiration. Large-bore Tru-cut biopsy is particularly useful for confirming the histologic diagnosis of benign lesions. The diagnostic sensitivity of fine-needle aspiration for benign lesions is only 33%, whereas that of large-bore Tru-cut biopsy can be as high as 85%. Fine-needle aspiration cytologic examination, however, may be sufficient for clinical management in cases of primary lung cancer. Fine-needle aspiration yields an accurate cytologic diagnosis in 88% of cases, and the correct histologic cell type in 70%.34Most patients with discrepancies between the cytologic and histologic diagnosis have disagreement between the cell types of non-small cell lung cancer, which may not significantly influence clinical management decisions. COMPLICATIONS OF US-GUIDED TNB

Because US scanning can reliably distinguish vascular and bronchial structures from tumors and lung consolidation, the route of the needle can be planned to prevent damage to the great vessels and major bronchi. USguided transthoracic needle aspiration and biopsy are very safe: the overall complication rate of US-guided TNB is about 1%,and minimal pneumothorax and mild hemoptysis account for most of the complications. Only rarely do patients require chest tube drainage for significant pneumothorax or blood transfusion for hemoptysis. In our series of more than 3000 cases of US-guided TNB in the past 20 years, no mortalities occurred.28The location, size, and depth of lesion do not inIf the lesion fluence the risk of c~mplication.~~ has a US window, TNB can be performed safely even for lesions deeply seated near the hilum. We also have not observed any cases of tract metastasis at the needle insertion site. Our complication rate is lower than those reported for fluoroscopyand CTguided (35%)6,22 TNB, although comparing


the complication rate of US-guided TNB and fluoroscopy- and CT-guided TNB is problematic because the patient selection may have differed between studies. IMPROVING THE SAFETY OF US-GUIDED TNB

Improvements in (1)the visualization of the needle shaft or tip and (2) the identification of blood vessels surrounding and within the target tumor can enhance the safety of USguided puncture procedures. For gray-scale, real-time US-guidance, needle visualization can be aided by several methods, including roughening or scoring the outer needle or inner stylet and placement of a guidewire through the needle. Injection of a tiny amount of saline or air as a contrast marker also improves visualization of the tip. Although we routinely score the Chiba needle to improve its echogenicity, the needle tip was visible on the gray-scale image in only 17% of cases in our previous study.I9 The needle shaft was also not visible. The time available for transthoracic aspiration biopsy makes the injection of saline or air impractical, because the patient might breathe and hence increase the risk of a complication. Visualization of the needle tip is particularly difficult if a narrowangle technique is used or the needle is inserted parallel to the US beam. Color Doppler US provides another solution to this problem. Any moving object emits or reflects US, and thereby creates a Doppler shift. This can be modulated into a color signal that mimics flow. When a needle is inserted, the motion caused by manipulation produces the same color signal (motion marking). The needle shaft or tip is much more easily identified by color Doppler imaging than by gray-scale imaging.'9 The color Doppler US puncture-guiding device described previously not only verifies vascular structures surrounding or within a target lesion in real-time, but also provides better visualization of the needle shaft or tip. ADVANTAGES AND LIMITATIONS OF US-GUIDED TNB

There are several advantages of using US for TNB guidance. US is relatively inexpensive, has no radiation exposure, does not restrict image direction, is easy to handle, and



can be performed at bedside for critically ill patients. Most importantly, US can provide real-time dynamic images even during transthoracic biopsy, and enables the operator to be more sure of exactly what the needle is targeting or has passed through. The flexibility and short examination time also make repeated examination possible. Chest US examination is easily mastered, and the equipment required is similar to that used for evaluating hepatobiliary, renal, or gynecologic diseases. The facilities are available to most physicians and can be shared between different departments. US also has some disadvantages. First, because the US wave is easily hindered by air, as in an aerated lung or pneumothorax, a US window is a prerequisite for chest US examination. In addition, US provides poor visualization of the mediastinum as compared with CT scan, cannot visualize airways, has a restricted field of vision, and is operator dependent. SUMMARY

Recent studies have confirmed that US is a very useful diagnostic tool for various diseases of the chest. The image information provided by US is helpful for etiologic diagnosis and clinical management. US-guided needle biopsy provides a precise and safe approach for transthoracic tissue sampling of lesions. The diagnostic yield is high, and the procedure is relatively easy and very safe. Color Doppler US and amplitude US angiography further extend the diagnostic potential and safety of this invasive procedure. Vascular information can be obtained and the needle shaft can be visualized clearly while conducting a biopsy. US examination and US-guided needle aspiration biopsy have now become indispensable diagnostic tools for various chest diseases. References 1. Chang DB, Yang PC, Luh KT, et al: Ultrasoundguided pleural biopsy with Tru-cut needle. Chest 100:1328-1333, 1991 2. Chang DB, Yang PC, Yu CJ, et al: Ultrasonography and ultrasonographically guided fine needle biopsy of impalpable cervical lymphnodes in patients with non-small cell lung cancer. Cancer 70:1111-1114, 1992 3. Chang DB, Yuan A, Yu CJ, et al: Differentiation of benign and malignant cervical lymph nodes with

color Doppler sonography. AJR Am J Roentgen01 162:965-968, 1994 4. Hurter TH, Hanrath P: Endobronchial sonography feasibility and preliminary results. Thorax 47565567, 1992 5. Izumi S, Tamaki S, Natori H, et al: Ultrasonically guided aspiration needle biopsy in disease of chest. Am Rev Respir Dis 125:460-464, 1982 6. Klein JS, Zarka MA: Transthoracic needle biopsy: An overview. J Thorac Imaging 12232-249, 1997 7. KO JC, Yang PC, Chang DB, et al: Ultrasonographic evaluation of peridiaphragmatic lesions: A prospective study. J Med Ultrasound 284-92, 1994 8. KO JC, Yang PC, Yuan A, et al: Superior vena cava syndrome: Rapid histologic diagnosis by ultrasoundguided transthoracic needle aspiration biopsy. Am J Respir Crit Care Med 93:250-252, 1994 9. Laing FC, Filly RA: Problems in the application of ultrasonography for the evaluation of pleural opacities. Radiology 126211-214, 1978 10. Liaw YS, Yang I'C, Wu ZG, et al: The bacteriology of obstructive pneumonitis: A prospective study using ultrasound-guided transthoracic needle aspiration. Am J Respir Crit Care Med 149:1648-1653, 1994 11. Liu C H A study of staining blood film (Romanowsky system). J Niigata Med Assoc 70:635-12643, 1956 12. Mathis G: Thorax sonography-part 11: Peripheral pulmonary consolidation. Ultrasound Med Biol 23:1141-1153, 1997 13. Pan JF, Yang PC, Chang DB, et al: Needle aspiration biopsy of malignant lung masses with necrotic centers: Improved sensitivity with ultrasonic guidance. Chest 103:1452-1456, 1993 14. Saito T, Kobayashi H, Kitamura S: Ultrasonographic approach to diagnosing chest wall tumors. Chest 94:1271-1275, 1988 15. Sugama Y, Tamaki S, Kitamura S, et al: Ultrasonographic evaluation of pleura and chest wall invasion. Chest 93:275-279, 1988 16. Suzuki N, Saitoh T, Kitamura S: Tumor invasion of the chest wall in lung cancer: Diagnosis with US. Radiology 18739-42, 1993 17. Targhetta R. Balmes P, Martv-Double C. et al: Ultrasonkally guided aspiration Giopsy in osteolytic bone lesions of the chest wall. Chest 103:1403-1408, 1993 18 Wang HC, Kuo PH, Liaw YS, et al: Diagnosis of pulmonaIy arteriovenous malformations by colour Doppler ultrasound and amplitude ultrasound angiography. Thorax 53:372-376, 1998 19. Wang HC, Yang PC, Yu CJ, et al: Transthoracic needle biopsy of thoracic tumor by color Doppler ultrasound puncture guiding device. Thorax 50:12581263, 1995 20. Wemecke K, Vassallo P, Peters PE, et al: Mediastinal tumor biopsy under US guidance. Radiology 172473-476, 1989 21. Westcott J: Direct percutaneous needle aspiration of localized pulmonary lesions: Results in 422 patients. Radiology 13731-35, 1980 22. Westcott J L Percutaneous transthoracic needle biopsy Radiology 169:593-601, 1988 23. Wiersema MJ, Kochmun ML, Cramer HM, et al: Preoperative staging of non-small cell lung cancer: Transesophageal US-guided fine needle aspiration biopsy of mediastinal lymph nodes. Radiology 190:239-242, 1994 24. Wu RG, Yuan A, Liaw YS, et al: Image comparison of real-time gray-scale ultrasound and color Doppler ultrasound for use in diagnosis of minimal pleural

ULTRASOUND-GUIDED TRANSTHORACIC BIOPSY OF THE CHEST effusion. Am J Respir Crit Care Med 150:510-514, 1994 25. Yang GG, Wu HD, Yang PC, et al: Sonographic patterns of ribs with tumor involvement. J Formos Med ASOC 90:749-754, 1991 26. Yang PC: Applications of colour Doppler ultrasound in the diagnosis of chest diseases. Respirology 2:231238, 1997 27. Yang PC: Color Doppler ultrasound of pulmonary consolidation. Eur J Ultrasound 3:169-178, 1996 28. Yang PC: Ultrasound-guided transthoracic biopsy of peripheral lung, pleural, and chest wall lesions. J Thorac Imaging 12:272-284, 1997 29. Yang PC, Chang DB, Lee YC, et al: Mediastinal malignancy: Ultrasound guided biopsy through supraclavicular approach. Thorax 47:377-380, 1992 30. Yang PC, Chang DB, Yu CJ, et al: Ultrasound-guided percutaneous cutting biopsy for the diagnosis of pulmonary consolidations of unknown etiology. Thorax 47457460, 1992 31. Yang PC, Chang DB, Yu CJ, et al: Ultrasound-guided core biopsy of thoracic tumors. Am Rev Respir Dis 146:763-767, 1992 32. Yang PC, Kuo SH, Luh KT Ultrasonography and ultrasound guided needle biopsy of chest diseases: Indications, techniques, diagnostic yields and complications. J Med Ultrasound 253-63, 1993 33. Yang PC, Lee LN, Luh KT, et al: Ultrasonography of Pancoast tumor. Chest 94:124-128, 1988 34. Yang PC, Lee YC, Yu CJ, et al: Ultrasonographically guided biopsy of thoracic tumors: A comparison of large-bore cutting biopsy with fine-needle aspiration. Cancer 692553-2560, 1992 35. Yang PC, Luh KT, Chang DB, et al: Ultrasonographic evaluation of pulmonary consolidation. Am Rev Respir Dis 146:757-762, 1992 36. Yang PC, Luh KT, Chang DB, et al: Value of sonography to determine the nature of pleural effusion: Anal-


ysis of 320 cases. AJR Am J Roentgenol 15929-33, 1992 37. Yang PC, Luh KT, Lee YC, et al: Lung abscessesUltrasonography and ultrasound-guided transthoracic aspiration. Radiology 180:171-75, 1991 38. Yang PC, Luh KT, Sheu JC, et al: Peripheral pulmonary lesions: Ultrasonography and ultrasonically guided aspiration biopsy. Radiology 155:451-456, 1985 39. Yang PC, Luh KT, Wu HD, et al: Lung tumors associated with obstructive pneumonitis: US studies. Radiology 174:717-720, 1990 40. Yang PC, Sheu JC, Luh KT, et al: Clinical application of real time ultrasonography in pleural and subpleural lesion. J Formos Med Assoc 83:64&657, 1984 41. Yu CJ, Yang PC, Chang DB, et al: Diagnostic and therapeutic use of chest sonography: Value in critically ill patients. AJR Am J Roentgenol 159:695-701, 1992 42. Yu CJ, Yang PC, Chang DB, et al: Evaluation of ultrasound-guided biopsies of mediastinal masses. Chest 100:399405, 1991 43. Yu CJ, Yang PC, Chang DB, et al: Ultrasound study in unilateral hemothorax opacification: Image comparison with computed tomography. Am Rev Respir Dis 147:430434, 1993 44. Yuan A, Chang DB, Yu CJ, et al: Color Doppler sonography of benign and malignant pulmonary masses. AJR Am J Roentgenol 163:545-549, 1994 45. Yuan A, Yang PC, Chang DB, et al: Lung sequestration: Diagnosis with ultrasound and triplex Doppler technique in an adults. Chest 102:1880-1882, 1992 46. Yuan A, Yang PC, Chang DB: Pulmonary infarction: Diagnosis and assessment of re-perfusion by ultrasound with color Doppler mapping. AJR Am J Roentgenol 160:419420, 1993 47. Yuan A, Yang PC, Chang DB, et al: Ultrasoundguided aspiration biopsy of small peripheral pulmonary nodules. Chest 101:926-930, 1992

Address reprint requests to Pan-Chyr Yang, MD, PhD Department of Internal Medicine National Taiwan University Hospital Number 7, Chung-Shan South Road Taipei 100, Taiwan, Republic of China e-mail: [email protected]