Biopsy of Mediastinal Tumors: Needle Biopsy versus Mediastinoscopy

Abstract

Mediastinal masses include a myriad of nonneoplastic and neoplastic conditions that may be benign or malignant, primary or metastatic. Although the clinical data, the location within the mediastinum, and the radiologic findings all help in narrowing the differential diagnosis, obtaining tissue for pathologic evaluation is often mandatory for choosing the appropriate treatment regimen. Tissue diagnosis is especially desirable for lesions that do not require surgical management. Transthoracic needle biopsy is a useful method for obtaining tissue diagnosis from mediastinal masses. Needle biopsy has several advantages compared with mediastinoscopy and mediastinotomy. Needle biopsy is faster and can be performed as an outpatient procedure, using local anesthesia, thereby avoiding the complications of surgery and general anesthesia. Additionally, needle biopsy is less costly compared with surgical procedures. The depth of the lesion does not influence sampling notably. All regions of the mediastinum are potentially accessible to needle biopsy, whereas certain mediastinal lymph node stations (e.g., anterior, posteroinferior subcarinal, para-aortic, and aortopulmonary nodes) are inaccessible to mediastinoscopy, sometimes requiring additional invasive diagnostic procedures such as mediastinotomy or thoracoscopy. We discuss the entities accounting for the majority of mediastinal masses and the merits of needle biopsy in their evaluation. Metastatic carcinoma to mediastinal or hilar lymph nodes accounts for 49 to 72% of mediastinal lesions, and is either the result of lung cancer or an extrathoracic malignancy. Malignant lymphoma and thymoma are the most frequent primary neoplasms in the mediastinum. Germ cell tumors, neurogenic tumors, cysts, benign lymphadenopathy, and inflammation are much less common. 1–3 The diagnostic accuracy of needle biopsy in these various lesions differs substantially and thus they are discussed separately. STAGING OF PRIMARY LUNG CANCER Metastatic carcinoma to the mediastinal lymph nodes is the most prevalent pathology within the mediastinum. Establishing the TNM staging of nonsmall cell lung cancer determines whether the cancer is operable. The status of mediastinal lymph nodes, however, cannot be assessed accurately by means of diagnostic imaging alone. According to Shields, 4 only 10 to 20% of patients with lung cancer have resectable mediastinal N2 lymph nodes (ipsilateral mediastinal and subcarinal lymph nodes). Patients with N3 disease (metastasis to contralateral mediastinal or hilar lymph nodes and to scalene and supraclavicular lymph nodes) are defined as having inoperable disease. It follows that the most common application for mediastinal transthoracic needle biopsy (or any other mediastinal staging procedure) is to prevent unnecessary surgery in patients with unresectable mediastinal lymph node metastases. Several imaging criteria that are used frequently to indicate “unresectability” include “bulky lymphadenopathy,” involvement of two or more lymph node stations, high paratracheal N2 disease, and superior vena caval obstruction. 5 Instead of relying on these nonspecific guidelines, mediastinal transthoracic needle biopsy, with or without biopsy of the primary lung cancer, can be used for definite staging (Fig. 1). High sensitivity and specificity of needle biopsy for the diagnosis of metastatic cancer to the mediastinum has been reported. Morrissey et al. 6 reported a sensitivity of 90% for metastatic carcinoma to the mediastinum using fine-needle aspiration (FNA), and a slightly higher sensitivity (96%) using core biopsy. Protopapas and Westcott 1 reported a sensitivity of 98% for the diagnosis of metastatic carcinoma in slightly to moderately enlarged lymph nodes. Weisbrod et al. 7 reported an accuracy of 90.3% in diagnosing mediastinal metastatic disease from lung and extrathoracic sites. Other investigators have also reported a high accuracy of needle biopsy for the diagnosis of metastatic mediastinal cancer. 2,3 Based on these results, several authors have recommended needle biopsy as the initial method for biopsy of suspected mediastinal cancer, manifested either as enlarged lymph nodes or as suspected mediastinal invasion.FIG. 1.: (A) Computed tomographic–guided needle biopsy of a subcarinal lymphadenopathy. The patient is in the prone position. The tip of the needle is seen within the subcarinal mass. (B) The aspiration biopsy of the subcarinal mass shows large clusters of markedly atypical epithelial cells with irregular nuclear membranes and prominent nucleoli (arrow), diagnostic of adenocarcinoma (Papanicolaou stain, original magnification ×1200).Figure 1: ContinuedDespite its many advantages, it should be emphasized that a negative needle biopsy of enlarged mediastinal lymph nodes should not be considered definitive evidence for benignity. A negative biopsy may be the result of sampling error resulting from incomplete neoplastic replacement of the sampled lymph node. A repeat biopsy may yield carcinoma in biopsies that were initially nondiagnostic. Protopapas and Westcott 8 addressed this issue recently in a study of 48 patients who underwent needle biopsy of enlarged mediastinal lymph nodes. In eight of these patients, a repeat biopsy was required, and five of the eight were positive for carcinoma on the second attempt. ANTERIOR MEDIASTINAL MASSES Establishing a diagnosis that distinguishes between lymphoma, thymoma, and germ cell tumor has major importance. Lymphoma is treated according to the type and stage of the disease with radiotherapy, chemotherapy, or a combination thereof in accordance with the precise diagnosis. Conversely, the primary treatment of thymoma and germ cell tumor is surgical resection. When diagnosing anterior mediastinal lesions with needle biopsy, the differential diagnosis depends on whether the predominant cell type is epithelial or lymphoid. When aspirates show a predominantly lymphoid cell population, the primary considerations are lymphoma (non-Hodgkin's lymphoma [NHL] and Hodgkin's disease [HD]) and thymic hyperplasia. When there is evidence for epithelial cell predominance the possibilities include epithelial thymoma, thymic carcinoma, germ cell tumor, metastatic carcinoma, and poorly differentiated carcinoma extending from the lung (Fig. 2). Lymphoma and thymoma are the most prevalent primary tumors of the mediastinum.FIG. 2.: (A) Computed tomographic–guided needle biopsy of an anterior mediastinal mass. A direct mediastinal approach has been used, avoiding transgression of the aerated lung. (B) The aspiration biopsy of this anterior mediastinal mass shows a biphasic cellular pattern, with small lymphocytes, scattered among prominent groups of bland benign epithelial cells, diagnostic of thymoma (Papanicolaou stain, original magnification ×600).Figure 2: ContinuedLymphoma The reported sensitivity of transthoracic FNA biopsy for the diagnosis of lymphoma ranges from 42 to 82%. Herman et al. 9 reported an overall sensitivity for diagnosing lymphoma of 42% in 26 patients. Their sensitivity for diagnosing HD was only 20%. 9 Weisbrod et al. 7 reported the accuracy of cytologic examination for lymphoma to be 66.7%, using FNA. Shaab et al., 10 who have also used FNA, have identified correctly six of seven cases of HD based on the presence of Reed–Sternberg cells in the specimen, and five of six cases of NHL based on the presence of lymphoid cells. FNA has a much lower accuracy for the diagnosis of mediastinal lymphoma compared with mediastinal carcinoma. Additionally, the combined number of reported FNA biopsies of mediastinal lymphoma is considerably lower compared with biopsies for carcinoma of the mediastinum. 1–3,8 Several investigators have reported a lower sensitivity of mediastinal needle biopsy for the diagnosis of HD, particularly the nodular sclerosis subtype, compared with NHL. The low yield of FNA in HD has been attributed to the paucity of Reed–Sternberg cells in the cellular aspirate. Subclassification of NHL into low and intermediate grades using FNA is also difficult or even impossible because it requires a histologic specimen to characterize the tissue architecture. 1,3 Silverman et al. 11 have reported their experience with transthoracic needle biopsy in 102 patients with lymphoma, which included 93 patients with NHL and 9 patients with HD. They used both FNA and core needle biopsy. Treatment was initiated in 72% of patients based on the results of image-guided needle biopsy alone. In patients with a prior diagnosis of lymphoma, treatment was initiated on the sole basis of needle biopsy in 91% of patients in their series. Ben-Yehuda et al. 12 reported their results with core needle biopsies in 100 patients with lymphoma, 79 of whom had NHL and 21 of whom had HD. A total of 86% of their patients were treated solely on the basis of results obtained by needle biopsy. The use of core needle biopsy in lymphoma allows sufficient tissue to be sampled for accurate histologic and immunohistochemical evaluation. It enhances the positive diagnostic rate as well as the percentage of patients treated on the basis of needle biopsy alone. Moulton and Moore 13 compared the use of aspiration versus core needle biopsy in 22 patients with lymphoma (21 NHL and 1 HD). FNA was positive for lymphoma in 68% of patients, but only in 53% was the information sufficient to institute therapy. In comparison, a correct diagnosis was established in 91% using core needle biopsy, allowing treatment in all of these patients. 13 Zinzani et al. 14 reported a sensitivity of 81% in diagnosing lymphoma using core biopsy. The diagnosed patients included 45 patients with NHL and 22 patients with HD. Specific classification of the histologic subtypes of NHL and HD was possible by means of immunohistochemical techniques. 14 It is clear that FNA has a much lower accuracy for lymphoma than for carcinoma. The use of core needle biopsies and immunohistochemical staining improves substantially the diagnostic accuracy of needle biopsy for the diagnosis of lymphoma (Fig. 3).FIG. 3.: Computed tomographic–guided needle biopsy of a large anterior mediastinal mass. The aspiration biopsy showed a rich population of large atypical cells. Immunocytostaining was positive for leukocyte common antigen and negative for cytokeratin, thereby establishing the diagnosis of large cell lymphoma.Thymoma Patients with thymoma are generally treated surgically, with radiotherapy reserved for patients with local invasion. In a series of 29 patients, Weisbrod 15 reported a sensitivity of 61% for the diagnosis of thymoma using FNA biopsy. Three cases of thymoma in this series were misdiagnosed by FNA, two as lymphoma and one as cystic thymoma. Herman et al. 9 reported a sensitivity of 71% for FNA biopsy in a series of 28 patients with thymoma. In an additional study by Weisbrod et al., 7 an accuracy of 82.4% for the cytologic diagnosis of thymoma was reported. It should be noted, however, that the cytologic evaluation was limited to the general diagnosis of thymoma, and did not differentiate between benign and malignant lesions. It appears that core needle biopsy can be beneficial in the diagnosis of thymoma, 6 but reports are limited, and even core needle biopsy cannot distinguish between benign and locally invasive disease. It is reasonable to conclude that, currently, the role of mediastinal needle biopsy for patients with thymoma is mainly to exclude other diagnoses for which treatment is nonsurgical, such as lymphoma and metastatic carcinoma. Patients with thymoma require surgery for an accurate diagnosis as well as for treatment. Nevertheless, a preoperative diagnosis may be of value when the tumor is large or when it is suspicious of being malignant. In these situations, a preoperative course of radiotherapy or chemotherapy may be useful. Mediastinal Germ Cell Tumors Mediastinal germ cell tumors account for 15% of anterior mediastinal masses. 16 Most series dealing with transthoracic needle biopsy for mediastinal germ cell tumors comprise only a small number of patients. The sensitivity for needle biopsy of mediastinal germ cell tumors has varied from 91% in a series of 11 patients described by Herman et al. 9 to 57% in a series of 10 patients described by Weisbrod. 15 Mature teratomas are predominantly cystic and can be very large. They may be calcified and are usually filled primarily with grumous material and hair. Most commonly, an aspiration biopsy yielding grumous material and squamous cells is sufficient to make the diagnosis. Seminoma is the most common malignant germ cell tumor in the mediastinum and it is sensitive to radiotherapy. It is therefore imperative to distinguish this germ cell tumor from those that require surgical excision or chemotherapy. Distinction from large cell lymphoma or HD can be difficult. Embryonal carcinoma may resemble poorly differentiated adenocarcinoma or large cell undifferentiated carcinoma. Core biopsies or cell-block preparations are superior to smears in making this distinction. NEUROGENIC TUMORS The most common cause of a posterior mediastinal mass is a neurogenic tumor. Weisbrod et al. 7 identified cells suggestive or consistent with neurogenic tumors in only 8 of 16 biopsies of posterior mediastinal masses. They found that some of these tumors underwent cystic changes as well as bleeding and fibrosis. Those associated features complicated their cytologic evaluation. Dahlgren and Ovenfors 17 described 16 biopsies of neurogenic mediastinal tumors, and concluded that such tumors, even when small, may be diagnosed relatively simply with combined radiographic features and aspiration biopsy. Neurogenic tumors in adults are usually of peripheral nerve sheath origin, most commonly schwannomas. 18 In children, most neurogenic tumors are malignant and arise from the sympathetic nervous system. Neurogenic tumors in children, which are typically highly cellular, include neuroblastoma and ganglioneuroblastoma, and are usually accompanied by increased levels of metabolites of catecholamines in the urine. Because needle biopsy in children poses difficulties that often make it impractical, these tumors in children are most easily diagnosed (and treated) using surgery. IMAGE GUIDANCE To perform a mediastinal needle biopsy, it is generally helpful to obtain a contrast enhanced computed tomographic (CT) scan before the procedure. The size of the mass as well as its characteristics, location, and proximity to mediastinal and vascular structures can be evaluated. This information can help avoid hitting a major vascular structure, and can assist in planning the appropriate needle path and patient positioning. Fluoroscopy is often inadequate to guide mediastinal biopsies because some lesions are poorly visualized and a safe needle path, avoiding vital cardiovascular structures, cannot be ensured. CT guidance is the preferred imaging used by most operators. 19 CT–fluoroscopy provides rapid image display and reconstruction allowing real-time visualization of the biopsy process. Procedures performed under standard CT guidance take longer than those performed using CT–fluoroscopy. White et al. 20 reported a decrease of 2 to 4 minutes per needle adjustment using CT–fluoroscopy. This method allows imaging of the needle as it advances, real-time visualization of positional changes resulting from respiration, and rapid detection of complications. Sonographic guidance is also used for mediastinal biopsies. 21–23 Sonography allows real-time imaging, with no exposure to ionizing radiation. Because an adequate sonographic window is required to achieve satisfactory sonographic imaging, this technique is suitable mainly for lesions abutting the chest wall. 23 Gupta et al. 24 report success with biopsies from right paratracheal and pretracheal lesions using suprasternal sonography. In this way, the soft tissue around the trachea and the great vessels provides an adequate sonographic window as well as anatomic guidance of the nearby structures. 24 A prebiopsy CT scan can be used to select patients in whom ultrasonic guidance for mediastinal biopsies is feasible. 25 In addition to transthoracic ultrasound, endosonographic guidance can also be used to perform mediastinal FNA. 26,27 DIFFERENT APPROACHES TO DIFFERENT LOCATIONS Direct Mediastinal Approach Using the direct mediastinal approach, the needle should be placed medial to the lung to avoid penetration of a pleural surface. This can be done easily when the mass in the anterior mediastinum is large enough and projects lateral to the sternum. Positioning the patient supine in an oblique lateral decubitus position may move the mass enough to enable direct mediastinal biopsy. With lateral decubitus positioning, the internal mammary artery and vein may be very close to the needle path. These vessels are clearly identified with CT and can be avoided by choosing a needle path that is immediately adjacent to the sternum. Bressler and Kirkham 28 reported two patients in whom the internal mammary vessels were punctured inadvertently, resulting in small, self-limiting extrapleural hematomas. Transpulmonary Approach The transpulmonary approach 28 for biopsy of mediastinal masses has been used especially under fluoroscopic guidance. This method has the attendant risk of pneumothorax as a common complication, which occurs in 20 to 30% of patients 2 because the needle must transgress two visceral pleural surfaces. This risk can be minimized by positioning the patients in the postbiopsy dependent position. 29,30 Suprasternal Approach The suprasternal approach is useful for lesions in the anterior 28 and the middle mediastinum. 31 Care should be taken when using this approach for middle mediastinal lesions to avoid puncturing the great vessels. Middle mediastinal lesions usually cause substantial anatomic distortion, which facilitates using the suprasternal approach and thus avoids the great vessels. Transsternal Approach The transsternal approach 32 enables sampling of most anteriorly located masses without passing lung parenchyma, while avoiding the hazard of puncturing the internal mammary vessels. The system includes a coaxial length-matched bone biopsy system, comprising an outer cannula and an inner eccentric drill bit. The sole complication we have encountered using this technique was minor patient discomfort while penetrating the sternum, which could be overcome by additional local anesthesia. Pleural Space Approach The pleural space approach 28 enables placement of the needle into the pleural space, which is widened by an effusion or a pneumothorax, avoiding puncture of the visceral pleura during the procedure. If the effusion is not big enough, patient positioning can help manipulate even a small amount of effusion to avoid visceral pleural puncture. If a pneumothorax is present as a complication of a failed prior biopsy attempt, subsequent attempts should be made using the existing pneumothorax to avoid further lung parenchymal puncture. In patients, with chest tubes ipsilateral to the lesion, an iatrogenic pneumothorax can be created, thus expanding the pleural space for a mediastinal mass biopsy. At the end of the procedure the pneumothorax is evacuated and the lung is reinflated. This approach should not be used if the patient develops dyspnea. BIOPSY NEEDLES Needles that are used for biopsy of mediastinal masses can generally be divided into two main categories: aspiration needles and cutting needles. Some needles yield tissue fragments that are suitable for histologic diagnosis as well as a cellular aspirate. Examples of such needles are the Westcott needle (Becton–Dickinson, Rutherford, NJ, USA) and the Turner needle (Cook Catheter, Bloomington, NJ, USA). In these two needle designs, modifications in the cutting edge have been made to improve the yield of tissue fragments. The Westcott needle has a beveled tip and a slotted opening adjacent to its tip, whereas the Turner needle has a sharpened 45° cutting edge along the circumference. 33,34 One of two techniques is generally used to obtain the biopsy specimen: the single-needle technique or the co-axial technique. In the latter, a thinner inner needle is inserted through a larger outer needle. The tip of the outer needle is placed at the edge of the lesion or within it, and multiple samples can be obtained using a single puncture. With the single-needle technique, multiple punctures are required to obtain multiple samples, but thinner, less traumatic needles can be used to obtain the same amount of specimen relative to the co-axial technique. 35 High diagnostic accuracy has been reported for aspiration biopsy of mediastinal metastatic carcinoma using thin aspiration needles that are 19 gauge or smaller. The reported sensitivity for mediastinal cancer ranges between 84 and 100%, and is comparable with needle biopsy results for pulmonary lesions. 1 However, the diagnostic accuracy of aspiration biopsy for noncarcinomatous lesions such as lymphoma, thymoma, and neurogenic tumors, is considerably lower. 1,2 In cases in which small tissue fragments with disorganized histology are insufficient (such as in noncarcinomatous and nonneoplastic masses), larger tissue fragments can be obtained using co-axial automated cutting needle devices. As stated earlier, in a study comparing aspiration versus core biopsy for the diagnosis of lymphoma, Moulton and Moore 13 reported an increase in sensitivity from 68 to 91%. Other studies have also shown superiority of Tru-Cut-type biopsies over FNA biopsies, particularly for the diagnosis of noncancerous malignant or benign mediastinal processes. 6,36 In automated devices, the core biopsy needle can be 18 or 20 gauge, but needles as large as 14 gauge have been used. 37 The larger the needle, the higher the risk for complication, but the ability to obtain sufficient tissue for diagnosis is higher as well. Balanced judgment regarding the biopsy needle to be used should be considered on a case-to-case basis, and should take into account the radiologic characteristics of the lesion. Biopsies of lesions with imaging suggestive of a carcinoma could be started with smaller gauge needles, whereas those of lesions suggestive of lymphoma or thymoma should be started with larger gauge needles. COMPLICATIONS The most frequently encountered complications of needle biopsy of the mediastinum are pneumothorax and mild mediastinal hemorrhage. The reported incidence of pneumothorax in several large series of needle biopsies of mediastinal lesions ranges from 0 to 34%, and is comparable with the reported rate in transthoracic biopsy of pulmonary lesions. 1–3,6,8,9 The risk factors for pneumothorax in transthoracic mediastinal biopsies include traversing the lung, the presence of emphysema, the depth of the lesion, the number of pleural surfaces traversed, and the number of needle passes. Therefore, the direct mediastinal, suprasternal, transsternal, and pleural space approaches, as described earlier, are preferred to the transpulmonary approach. The rate of hemorrhage after needle biopsy of the mediastinum ranges from 0 to 10%. 1 Even when mediastinal bleeding does occur, short hospitalization for observation is sufficient. 1,28 The incidence of serious complications resulting from needle biopsy of the mediastinum is very low, 1,28 even though major intrathoracic hemorrhage causing cardiac tamponade and death have been reported. 38,39 SUMMARY Image-guided needle biopsy of the mediastinum is rapid, safe, and inexpensive, and does not require general anesthesia. All regions of the mediastinum are potentially accessible using this technique. The depth of the lesion does not limit the feasibility of the biopsy procedure. Unnecessary surgery can often be avoided, thereby reducing patient morbidity and shortening hospitalization. The importance of establishing a diagnosis in a nonsurgical way in patients who are not surgical candidates is clear. Many mediastinal biopsies are performed for lung cancer staging. Needle biopsy, even when using thin aspiration needles, is highly accurate in diagnosing mediastinal carcinoma. Advances in biopsy needle technology, namely the use of core biopsy needles, have enabled the yield of large tissue fragments with minimally distorted histology, facilitating the diagnosis of lymphoma and helping to establish a diagnosis in patients in whom fine-needle biopsy was nondiagnostic. Advances in immunohistochemical staining have also contributed to the diagnostic accuracy of needle biopsy in the diagnosis of lymphoma. Innovations in CT imaging, particularly CT–fluoroscopy, which is becoming more and more available, shorten the duration of the CT-guided needle biopsy procedure and enhance its accuracy. In cases of thymoma, the role of mediastinal needle biopsy is mainly to exclude other diagnoses for which treatment is nonsurgical, such as lymphoma and metastatic carcinoma. In our opinion, needle biopsy of the mediastinum should be the procedure of choice to obtain tissue diagnosis for a variety of mediastinal pathologies, particularly when the established diagnosis can obviate surgery.