点击显示 收起
1 From the Section of Vascular and Interventional Radiology, Division of Diagnostic Imaging, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd, Box 57, Houston, TX 77030-4009. Received March 15, 2001; revision requested April 24; revision received June 18; accepted July 5..G. Address
| ABSTRACT |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
MATERIALS AND METHODS: Medical records of 37 consecutive patients who underwent CT-guided transsternal biopsy of intrathoracic lesions were evaluated retrospectively. A coaxial needle technique was used in all patients; an 18-gauge needle was used for transsternal penetration, through which a 22-gauge needle was passed to obtain fine-needle aspirates. Five patients also underwent core-needle biopsy with a coaxially introduced 20-gauge needle. Medical records were reviewed for lesion size and location, needle path, number of needle penetrations, reasons for failure, biopsy results, and complications.
RESULTS: The transsternal approach was used in mediastinal (n = 32) or intrapulmonary (n = 5) lesions. Transsternal needle sampling of the target lesion was successful in 35 patients. In the remaining two, adequate angling of the transsternal needle could not be achieved. Extrapleural access to the mediastinal lesions was achieved in all but one patient in whom the 22-gauge needle traversed the lung. Major vessels were avoided in most patients; the 22-gauge needle was safely passed through the brachiocephalic vein in one patient with a retrotracheal mass. Thirty-two (91%) of the 35 biopsies yielded diagnostic specimens. No major complications were encountered. Minor complications were pneumothorax in one patient and mediastinal hematoma in another.
CONCLUSION: The CT-guided transsternal approach for coaxial core-needle biopsy allows safe access to masses in various locations in the mediastinum and anteromedial lung.
Index terms: Biopsies, technology • Mediastinum, biopsy, 67.126 • Mediastinum, CT, 67.1211 • Mediastinum, neoplasms, 67.315
| INTRODUCTION |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
Most mediastinal lesions require an angled transpulmonary approach with CT guidance (1,3,4,6,9). The needle path traverses both the lung and the pleura, which increases the risk of pneumothorax (6,9). The transsternal approach for mediastinal biopsy has been described in a few series of patients; the major advantage of this route is the reduction of the risks of pneumothorax and accidental injury to the internal mammary vessels (15–18). The purpose of our study was to report our experience with CT-guided coaxial needle biopsy of intrathoracic lesions by using the transsternal approach.
| MATERIALS AND METHODS |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
All biopsies were performed by using a helical CT scanner (CTi Smartview; GE Medical Systems, Milwaukee, Wis). Patients initially underwent imaging in the supine position with a section thickness of 5, 7, or 10 mm. Nonenhanced CT was performed in 35 of the procedures, while two patients underwent contrast-enhanced CT after intravenous administration of ioversol (Optiray 320; Mallinckrodt, St Louis, Mo) to define the major vessels in the vicinity of the lesions. Once the access route and needle entry site were chosen, the skin was prepared in a sterile fashion, and 1% lidocaine hydrochloride (Xylocaine; Astra USA, Westborough, Mass) was administered with a 25-gauge hypodermic needle (Becton Dickinson, Franklin Lakes, NJ) to anesthetize the skin, subcutaneous tissue, and periosteum of the anterior sternal cortex.
All biopsies were performed by using a coaxial core-needle biopsy technique. A variety of guide needles were used according to the preference of the radiologist performing the procedure; these included an 18-gauge, 1.5-inch hypodermic needle (Becton Dickinson; n = 17); an 18-gauge Hawkins needle (Medical Device Technologies, Gainesville, Fla; n = 6); an 18-gauge Chiba needle (Cook, Bloomington, Ind; n = 11); an 18-gauge spinal needle (Sherwood Medical, St Louis, Mo; n = 2); or a 14-gauge Ackerman bone-biopsy needle (Cook; n = 1). In each case, the guide needle was anchored into the anterior cortex of the sternum, with the needle alignment monitored by means of CT. A combination of steady pressure and a drilling motion were used to penetrate the anterior cortex. The alignment of the needle was reassessed with CT and adjusted if necessary. Once the needle angle was considered adequate, the guide needle was advanced to the posterior cortex. Before penetrating the posterior cortex, the posterior periosteum was anesthetized with 1% lidocaine instilled through the 18-gauge guide needle or through a 22-gauge Chiba needle passed coaxially through the outer guide needle.
CT was used to confirm the adequacy of the trajectory and the position of the needle tip when the guide needle traversed the posterior cortex into the anterior mediastinum. In patients with narrow mediastina and a small contact surface between the mediastinum and the sternum (n = 3), saline was injected through the needle into the anterior mediastinum in an attempt to widen the mediastinum and avoid crossing the lung or pleura. The distance from the tip of the guide needle to the lesion was measured; a 22-gauge Chiba needle of an appropriate length was advanced coaxially through the guide needle into the lesion.
In cases in which the trajectory of the guide needle was less than optimal and precluded a straight route to the lesion, a curved 22-gauge needle was used to compensate for the misalignment. The curved needle was created by using a hemostat to form a gentle curve in the tip of the 22-gauge needle. In some cases, the curved 22-gauge needles were also used to avoid major vessels or to sample different parts of the mass if the initial aspirates yielded necrotic or nondiagnostic material.
The aspirates were expressed and smeared onto glass slides that were then stained and examined immediately by the cytopathologist in attendance at the time of the biopsy to assess the adequacy of the specimen. Additional samples were obtained if required. One to four passes were required in each patient. In five patients, additional core-needle biopsies were also performed at the recommendation of the cytopathologist examining the slides after multiple aspiration samples had failed to yield adequate diagnostic material; these biopsies were performed with 20-gauge semiautomated cutting needles (Quick-Core; Cook) advanced coaxially through the guide needles.
Postprocedural CT scans were not routinely obtained to evaluate for complications. Expiratory chest radiographs were ordered at the discretion of the interventional radiologist performing the biopsy. The patients were observed for 1–3 hours after the procedure to ensure their hemodynamic stability and to monitor their respiratory status. The procedures were performed on an outpatient basis unless the patients were already hospitalized.
The medical records of all patients were reviewed jointly by two radiologists (S.G., M.J.W.), and data were collected concerning the size and location of the lesion, the needle path, the number of transsternal needle penetrations needed, the reasons for failure, the immediate and delayed complications, and the biopsy results. Follow-up clinical, radiologic, and pathologic information was obtained in patients in whom the biopsy specimens were nondiagnostic or negative for malignancy.
| RESULTS |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
|
|
There was a direct avascular path to the lesion through the mediastinal fat in 25 patients: 13 patients with lesions anterior to vessels (Fig 2), six patients with lesions lateral to the aortic arch or in the aortopulmonary window (Fig 3), five patients with pulmonary lesions, and one patient with a large pretracheal mass extending anteriorly up to the sternum with lateral displacement of the vessels. In nine patients, the pretracheal, paratracheal, or retrotracheal masses were posterior to the great vessels. In eight of these patients, the 22-gauge needle was safely passed between the major vascular structures to sample the lesion (Fig 4a, 4b). However, in one patient in whom a direct avascular route was not available, the 22-gauge needle was safely advanced through the left brachiocephalic vein to sample a retrotracheal lesion (Fig 4c, 4d).
|
|
|
|
|
|
|
|
|
|
The biopsy specimens were adequate for interpretation in 32 (91%) of 35 cases of successful transsternal biopsy. In the other three cases, the biopsy specimens were either inadequate (n = 2) or nonrepresentative (n = 1). One of these three patients had a history of treated Hodgkin disease and presented with new enlarged mediastinal lymph nodes anterior to the aortic arch, which was suggestive of recurrence; biopsy results revealed scant material with a few atypical lymphocytes. Subsequent surgical biopsy results revealed Hodgkin disease. In another patient, who presented with a soft-tissue mass lateral to the aortic arch, the biopsy specimen was nonrepresentative. The patient underwent thymectomy; results of histopathologic analysis showed thymic hyperplasia. In the third patient, the biopsy specimen from a 1-cm lymph node anterior to the aortic arch showed lymphoid cells suggestive of small cleaved cell lymphoma; the diagnosis was subsequently confirmed by means of CT-guided axillary lymph node biopsy.
In 22 patients, transsternal biopsy results revealed malignancy. The diagnoses were as follows: metastatic tumor consistent with the patient’s known primary malignancy (n = 13); primary bronchogenic carcinoma (n = 3); recurrent, residual, or new malignant lymphoma (n = 4); and thymoma (n = 2). In the remaining 10 patients, no malignancy was identified in the biopsy specimens. Five of these 10 patients underwent mediastinoscopy or open surgical biopsy, the results of which were used to confirm the initial transsternal biopsy results. In the other five patients, clinical and imaging follow-up findings were used to confirm that the disease process was benign.
No major complications were encountered. Two patients had minor complications. One patient who had renal cell carcinoma and underwent fine-needle aspiration biopsy of an aortopulmonary lymph node developed a small mediastinal hematoma after the first needle pass (Fig 3b, 3c). The patient remained clinically stable, and repeat CT 3 hours later showed that the hematoma was smaller; the patient was discharged after 4 hours of observation. In the other patient, who had a paracardiac mass, the 22-gauge needle had to traverse the lung, and the patient developed a small pneumothorax, which remained stable and did not require placement of a chest tube. This patient was discharged after 3 hours of observation in the radiology holding area. Minor pneumomediastinum resulting from air entering through the transsternal needle was noted in 10 patients during the course of the biopsy but did not lead to any symptoms or clinical sequelae. We did not prospectively record the level of pain experienced by the patients; however, judging from the opinion of the radiologists who performed these biopsies, all patients tolerated the procedure well without any complaints of undue pain or discomfort more than that experienced during any percutaneous biopsy.
| DISCUSSION |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
CT-guided parasternal and transpulmonary mediastinal biopsy also has limitations and risks. The parasternal approach is associated with a small but definite risk of hemorrhage from injury to the internal mammary vessels that run along the lateral border of the sternum (13,14). Even in cases in which these vascular structures are identified at CT, it may not be possible to find a safe window between the vessels and the sternum. In most patients, the lesion or the mediastinum is not in direct contact with the parasternal chest wall, which precludes extrapleural biopsy (1,3,6,9). Hence, most mediastinal tumors require an angled transpulmonary approach, with the needle traversing the lung and two layers (anterior and mediastinal) of the visceral pleura. The reported risk of pneumothorax with this approach ranges from 11% to 19.2% (6,9).
A limited number of cases (n = 34) in which patients underwent CT-guided transsternal mediastinal biopsy have been reported in the literature (15–18). Swanson and Wittich (18) reported successful transsternal biopsy in a mediastinal tumor in 1990. D’Agostino et al (16) described their early experience with the transsternal approach for performing biopsy in anterior mediastinal lesions in seven patients and found it to be safe, well-tolerated, and effective. Astrom et al (15) reported 10 transsternal biopsies in which a bone biopsy system was used to penetrate the sternum. Hagberg et al (17) used the same technique in their recent series involving 21 patients; however, their study group included five patients from the previous series reported by Astrom et al (15).
None of the authors of these series described the specific locations of the lesions in which biopsy was performed, but the transsternal approach was used primarily for mediastinal masses anterior to the great vessels (15–18). Results of our experience suggest that the transsternal approach is useful not only for the prevascular anterior mediastinal masses but also for lesions in other mediastinal compartments and in the anterior and medial aspects of the lungs. Safe extrapleural access to middle or posterior mediastinal masses behind the great vessels in the peritracheal compartments or in the aortopulmonary window can be achieved with this technique. Small-gauge (22-gauge) needles were advanced between vascular structures to reach these lesions in many of the patients in our study. No complication was encountered in the patient in whom the 22-gauge needle traversed the brachiocephalic vein en route to the target lesion. This latter transvenous biopsy technique has been reported previously (5,11,12) and has been shown not to increase the risk of hemorrhagic complications during mediastinal biopsy.
Patients with intrapulmonary lesions abutting and/or invading the anterior mediastinum or the retrosternal fat, the soft tissue, or both, were also included in our study. The transsternal approach avoids the traversal of aerated lungs, thus reducing the risk of pneumothorax. The results in a previous study (19) indicate that the incidence of pneumothorax after pulmonary biopsy is lower if no aerated lung is traversed. None of the five patients who had pulmonary lesions and underwent transsternal biopsy in our study developed pneumothorax.
Astrom et al (15) performed contrast-enhanced CT during biopsy in all patients and suggested that accurate planning of the needle trajectory requires intravenous administration of contrast material to help define the mediastinal vessels. We believe that this additional step is not warranted and that nonenhanced CT alone is sufficient for safe biopsy planning in most patients who have had previous diagnostic contrast-enhanced CT. Only two of the patients in our study received contrast material intravenously during the procedure to help differentiate the mediastinal structures.
Astrom et al (15) performed transsternal mediastinal biopsy by using a complicated procedure involving a bone biopsy system comprising two cannulas, long and short drill bits, a depth gauge, and a cutting biopsy needle. It is our experience that a variety of common 18-gauge needles are adequate to penetrate the sternum, negating the use of an elaborate and expensive bone biopsy system. As in the study by D’Agostino et al (16), we used a hypodermic needle to penetrate the sternum in many of the initial patients in our study. One disadvantage of the hypodermic needle is its limited inner-to-outer-diameter ratio, which precludes the coaxial insertion of a 20-gauge needle at core-needle biopsy to obtain larger specimens when needed. We, therefore, recommend the use of a thin-walled 18-gauge Hawkins or Chiba needle for transsternal biopsy.
Astrom et al (15) achieved the planned trajectory angle in all patients; in their series, the largest discrepancy between the calculated and final angles of the biopsy system was 5°. Although the desired trajectory was achieved with a single transsternal penetration in most of the patients in our study, incorrect angulation may warrant the removal and reinsertion of the transsternal needle, which occurred in seven patients in our study. It is important to perform CT to check the needle direction at small increments of insertion of the guide needle. This is particularly important after wedging the needle tip into the anterior cortex and before penetrating the posterior cortex, because this ultimately determines the final path. Patient motion must be avoided because it may change the mediastinal topography; hence, it is important to place the patient in a comfortable position and to minimize the examination time (15). A curved 22-gauge Chiba needle can also be used to compensate for minor discrepancies in the needle trajectory. Mediastinal movements from respiration and cardiac or aortic pulsations can also result in the lesion moving in and out of the biopsy plane; this is particularly true for small masses close to the aortic arch.
As did the authors of previous reports (15–18), we found that transsternal needle insertion after the patient received local anesthesia and intravenous sedation was a well-tolerated safe procedure. The infiltration of a local anesthetic into the periosteum of the anterior and posterior sternal cortices minimizes the discomfort associated with the procedure. No complications directly attributable to the transsternal approach were encountered in the patients in our study or in those in previous investigations (15–18). Hemorrhage, a potential complication of mediastinal biopsy, can be prevented by careful planning of the biopsy needle trajectory on the basis of the location of major vessels in relation to the lesion. A small mediastinal hematoma encountered in one patient in our series was attributed to hemorrhage from the hypervascular renal cell metastasis in which biopsy was performed rather than to accidental injury to the adjacent pulmonary artery.
The transsternal approach eliminates the risk of pneumothorax by providing an extrapleural route to the mediastinum in a majority of the patients. Injection of physiologic saline was used successfully in two patients to widen the mediastinum and to create an extrapleural window for biopsy in deep lesions. The pneumothorax seen in one patient resulted from pleural and pulmonary traversal by the 22-gauge needle because the saline injection did not sufficiently displace the mediastinal pleura to allow extrapleural access to a paracardiac lesion on the right side.
Our 91% success rate for diagnosis compares favorably with the rates reported in the previous studies (1–12,15–18) in which percutaneous mediastinal biopsy was described. Although some authors (15,17) stress the need for core-needle biopsy, we found that fine-needle aspiration biopsy provided adequate tissue for diagnosis in most patients. This is especially true if the cytopathologist is present to assess the adequacy of the samples at the time of the biopsy. If larger tissue specimens are required for diagnosis, the coaxial insertion of an 18-gauge thin-walled needle allows for use of a 20-gauge needle at core-needle biopsy.
In conclusion, the coaxial transsternal biopsy technique allows safe access to masses in various locations within the mediastinum and anteromedial lung. This approach does not limit biopsy to the anterior mediastinum alone but allows for the sampling of lesions in the posterior mediastinum. Routine biopsy needles can be used to penetrate the sternum and gain access to the mediastinum and to permit both multiple fine-needle aspirations and potential core sampling.
| REFERENCES |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|