Percutaneous Aspiration Biopsy of Lymph Nodes

Transcription

1 Percutaneous Aspiration Biopsy of Lymph Nodes WILLIAM L BETSILL, JR., M.D., AND STEVEN I. HAJDU, M.D. Betsill, William L., and Hajdu, Steven I.: Percutaneous aspiration biopsy of lymph nodes. Am J Clin Pathol 7: , 980. During a period of five years, patients each had cytologic examination of a specimen aspirated by percutaneous needle biopsy from a superficial lymph node. Fiftyfive aspirates were considered unsatisfactory owing to scanty cellularity, and were excluded because of lack of adequate follow-up. Of the remaining 84 aspirates, 8% were positive for malignant cells; 9% were considered negative. There were no cases with false-positive diagnoses. However, of negative aspirates, revealed tumor on subsequent excisional biopsy. Seventy-two percent of patients were admitted with prior history of malignancy. For both males and females, an apparent correlation was found between regional distribution of positive lymph nodes and the histologic types of primary tumors. Eighty-two percent of the positive cervical nodes in males and % in females were associated with epidermoid carcinoma. Most of the positive supraclavicular and axillary lymph nodes occurred in females and were associated with mammary carcinoma. Diagnosis of malignant neoplasms is feasible, in most instances, from nodal aspirates, but specific diagnosis of the histologic type of the tumor, e.g., malignant lymphoma, should be rendered only when the cytologic findings are supported by appropriate clinical and laboratory findings. In case of doubt, or if the aspirate contains many polymorphonuclear leukocytes, necrotic debris, or bizarre epithelioid cells, granulomatous lymphadenitis should be suspected and the diagnosis should be deferred until formal tissue biopsy can be obtained. (Key words: Aspiration; Lymph node; Biopsy.) ASPIRATION OF LYMPH NODES for diagnostic purposes was reported as early as 904 by Grieg and Gray", who used the procedure in the diagnosis of trypanosomiasis. Ward 4 and Chatard and Guthrie published reports on lymph-node aspiration in 94, the former in an attempt to diagnose neoplastic disease, and the latter for use in infectious disease cases. In 9 Guthrie first attempted to collate lymph-node aspiration cytology with various disease processes. Soon thereafter, at Memorial Hospital, aspiration biopsy became an often used method of obtaining diagnoses of superficial masses involving the breast, lymph nodes, and to a lesser extent, bone, lung, salivary glands, and prostate. 7_,l) reports regarding various aspects of this subject have been published; however, a detailed study specifically concerning lymph-node as- Received June 7, 979; received revised manuscript and accepted for publication July, 979. Address reprint requests to Dr. Hajdu: Memorial Sloan-Kettering Cancer Center, 7 York Avenue, New York, New York 00. Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, New York pirates is not apparent in the American literature. The purpose of this paper is to present the findings of a large series of lymph-node aspirations, specifically in an attempt to evaluate the feasibility and diagnostic reliability of this procedure. At Memorial Hospital, approximately 80 aspiration biopsies from all sites are performed annually. This diagnostic method has been used for more than fifty years. This study will analyze percutaneous superficial lymph-node aspirates obtained from various groups and body sites during a recent five-year period, and will relate appropriate clinico-pathologic data regarding these cases. Materials and Methods All percutaneous superficial lymph-node aspirations performed at Memorial Hospital during a recent fiveyear period (97-977) were reviewed and correlated with follow-up information. The patients ranged in age from 0 to over 80 years. Lymph-node aspiration biopsies were considered from four anatomic sites (the neck, the supraclavicular area, the axilla, and the groin). Intraoperative lymph-node aspiration biopsy specimens and aspirates of nonnodal lesions (e.g., lipoma) were omitted from this study. Not included in the 84 cases are eight cases of benign neoplasms that occurred in the anatomic sites mentioned above. Upon aspiration, no lymphoid tissue was obtained, indicating either soft tissue or salivary-gland lesions. The technic of aspiration biopsy employed at Memorial Hospital has been described in a number of publications 47 and will not be included herein. Results Aspiration smears obtained from superficial lymph nodes of patients were reviewed. Owing to inadequate follow-up data, cases were excluded from this study. An additional aspirates were judged unsatisfactory for diagnosis, owing to scanty cellularity of the smears (Table ). The remaining 84 cases /80/0400/047 $00.9 American Society of Clinical Pathologists 477 Downloaded from on March 08 V

2 47 BETSILL AND HAJDU A.J.C.P. April 980 Lymph Node Cervical Supraclavicular Axillary Inguinal Total Table. Anatomic Site and Cytologic Diagnosis of Aspirates Cytologic Diagnosis and Number of Cases Total No. of Cases Positive Negative (%) 47 (%) (%) J(ll%) Unsatisfactory 8 7 (8%) (%) (7%) formed the basis for this study. The patients were widely distributed with respect to age; the male to female ratio was approximately equal. Of a total of 9 patients who had lymph-node aspirations, underwent two aspiration procedures. No patient had more than two aspirations. Most male patients were older than 0 years of age (average male age was 0 years). The female population averaged years of age, with significantly more patients in the fifth decade of life (Fig. ). Of the 84 nodal aspirates with adequate material for diagnosis, % were obtained from the neck region. Sixteen percent were from supraclavicular sites; axillary and groin regions each represented % of the total number of aspirates. The greatest proportion of cytologically negative aspirates were obtained from the groin region (%), and the smallest percentage of noncontributory aspirates were from the neck area (4%). Whereas 4% of the neck aspirates were unsatisfactory for diagnosis, only 0% of the supraclavicular aspirates were found unsatisfactory for diagnosis. Table shows the relation between anatomic site and histologic type of tumors according to the sex of the patient. Fifty-six percent of the neck aspirates were obtained from males, and 44% were from females. Of the total number of malignant tumors diagnosed from aspirates of the neck region, 7% were epidermoid carcinoma. s had a higher incidence of adenocarcinoma than did males. Seventy-five percent of supraclavicular aspirates and 80% of axillary-region aspirates were obtained from females, and in these aspirates adenocarcinoma was the most often diagnosed tumor. An approximately equal proportion of groin node aspirates were obtained from males and females, and no predominant histologic type of tumor was evident. With respect to all sites, most malignant tumors were epidermoid carcinoma (% of the total number). Seventy-two percent of the total number of patients who had postitive aspirates had a previous diagnosis of a malignant tumor. Approximately 8% of the patients who had axillary and groin aspirates and 70% of the patients who had cervical and supraclavicular aspirates gave a prior history of malignancy (Table ). The correlation of each anatomic site with histologic type and site of the primary is shown in Table 4. Most positive neck aspirates from both males and females were associated with primary tumors from the oral cavity or upper respiratory tract. Among males, 4 tumors originated from primary sites in the oral cavity : Average Age: 0 : 49 Average Age: Table. Anatomic Site of Lymph Nodes and Histologic Type of Tumors According to Sex 80+ ]80EI Lymph Node Epidermoid Carcinoma Number of Cases Adenocarcinoma Miscellaneous Cancer WMnwrnm El 40 Ell EI Ml] i i i i i i i n i i i i i i i Number of patients FIG.. Age and sex distribution of 84 patients. Cervical Supraclavicular Axillary Inguinal Total 8 0 (%) (9%) 9 8 9(8%) Downloaded from on March 08

3 Vol. 7 No. 4 ASPIRATION OF LYMPH NODES 47 or upper respiratory tract. More specifically the larynx, tongue, lung, tonsil, soft palate, and buccal mucosa accounted for 4 of the 4 primary sites. Ten primary sites were occult at the time of presentation. Clinicopathologic work-up identified seven primary sites from the ten occult cases: lung, two; tongue, two; larynx, one; lymphoma, poorly differentiated lymphocytic type, one; lymphoma, histiocytic type, one. significant primaries associated with neck metastases in males were the lung (seven cases) and esophagus (four cases). Among females, 0 tumors were metastatic from primary sites in the oral cavity and the upper respiratory tract ( of the 0 cases originated in the tongue, larynx or floor of the mouth). sites included the cervix (four cases), lung (two cases) and thyroid (two cases). Seven tumors presented without prior history of carcinoma; clinical work-up detected five primary sites (lung, one; tonsil, one; tongue, one; lymphoma, histiocytic type, one; Hodgkin's disease, one). Metastatic adenocarcinoma to the cervical region among males was not associated with a predominating primary site (six cases from various sites, namely, the Table. Correlation of Prior History and Positive Cytologic Finding Lymph Node Cervical Supraclavicular Axillary Inguinal Total Prior History of Carcinoma Number of Patients No Prior History of Carcinoma _ 7 (7%) (8%) parotid, minor salivary gland of the tongue, kidney, prostate, colon and rectum). Among females, however, the breast was the primary in of positive neck aspirates diagnosed as adenocarcinoma. No occult adenocarcinoma was encountered in the neck area in either sex. Seventeen cases of miscellaneous tumors in males and females were diagnosed by aspirates from the neck; lymphoma comprised of the cases; melanoma (two cases) and sarcoma (four cases) were also present. Table 4. Correlation of Site of Cytologically Positive Lymph Nodes and Anatomic Site of Primaries Number of Cases Lymph Node Epidermoid Carcinoma Adenocarcinoma Miscellaneous Malignant Tumor Cervical Head and neck Unknown 4 0 Various Sarcoma Head and neck Unknown Breast 4 Sarcoma Axillary Lung " " Breast Lung 0 Sarcoma Supraclavicular Lung Prostate Colon Cervix Breast Inguinal Bladder Prostate Colon Sarcoma Vulvo-vaginal Anal Various Downloaded from on March 08

4 474 BETSILL AND HAJDU A.J.C.P. April 980 v N s Aspiration of Lymph Nodes Percent of Total Histoloc jic Type of Tumor Epid. Ca \ 8 Adenoca M sc.ca. 8 Fio.. Histologic types of the tumors according to anatomic site of the lymph nodes. "negative cases" by a long-term clinical follow-up. Thus, of the cases of malignancy (89%) were diagnosed correctly by the needle aspirate, but the remaining % required formal tissue biopsy for definitive diagnosis (Fig. ). Representative aspiration smears are illustrated in Figures 4-9. Epidermoid carcinomas showed a spectrum of differentiation ranging from poorly differentiated epidermoid elements to well-differentiated keratinizing squamous cells. In most instances there was an admixture of neoplastic epidermoid cells at various stages of maturation. However, we have seen buccal primaries in which the aspirates of the metastatic node were a pure culture of well-differentiated keratinizing squamous cells. In general, malignant epidermoid cells exhibited a remarkable tendency to form sheets and clusters. The cohesive tendency of epidermoid cells occasionally made it possible to be certain whether the neoplastic cells were epidermoid or glandular in origin (Fig. 4). Adenocarcinomas metastatic to lymph node showed considerable variation in size and shape of the tumor cells, but in general showed striking resemblance to the primary neoplasm. In the majority of cases, welldifferentiated metastatic adenocarcinomas diagnosed Sixteen of 0 axillary aspirates were from females, and ten of these cases were related to a previously known primary mammary carcinoma. Twenty-two of supraclavicular aspirates were obtained from females, and of these cases were related to known breast carcinoma. Owing to the small number of cases, the tumors diagnosed from aspirates of groin nodes did not reveal a predominating primary site (Fig. ). With respect to nonneoplastic lesions, eight of cases were diagnosed from aspirates from the neck region. Seven cases of granulomatous lymphadenitis were found. After special stains and appropriate cultures, three of these cases were subclassified as tuberculous lymphadenitis; four revealed no etiologic agent. Also, two cases of acute lymphadenitis (one with associated gram-positive cocci) and two cases of chronic lymphadenitis (one associated with sialadenitis) were encountered. Of the 84 nodal aspirates with sufficient material for diagnosis, (8%) were initially diagnosed as positive for malignant cells. Fifty-two aspirates were initially diagnosed as negative for malignant cells. Thirty-one of the same cases were found to reveal malignant tumor on follow-up aspiration or tissue biopsy; the remaining cases (8%) were confirmed Aspiration of Lymph Nodes Percent of Total Cytol. dx/follow-up +/+ -/ /+ FIG.. Correlation of cytologic diagnosis and follow-up information. (Positive finding = (+); negative finding = (-).) 4 8 Downloaded from on March 08

5 Vol. 7 No. 4 ASPIRATION OF LYMPH NODES 47 FIG. 4. Metastatic epidermoid carcinoma. (Left). Well-differentiated keratinizing epidermoid carcinoma of the tongue metastatic to a neck node. The tumor cells are large keratinizing squamous cells in a clean, noninflammatory, nonnecrotic background. Hematoxylin and eosin. x40. (Right). Poorly differentiated epidermoid carcinoma, so-called malignant lymphoepithelioma of the nasopharynx. The aspirate is from a neck node. The tumor cells are small pleomorphic cells in haphazardly arranged clusters. Note the lymphocytic background. Hematoxylin and eosin. x40. on the basis of cytomorphology of the tumor cells could be identified as to primary site of origin, e.g., the breast, colon, pancreas, or kidney. However, small cell adenocarcinomas such as adenocarcinoma of the endometrium, thyroid carcinoma, bronchiolar adenocarcinoma, and malignant carcinoids could not be differentiated with certainty (Fig. ). Oat cell carcinomas metastatic to lymph nodes, despite the resemblance of individual tumor cells to neoplastic or atypical lymphoreticular cells, were identified without difficulty, because of the typical cohesive cell clusters, uneven contour of the nuclei, and almost complete loss of the cytoplasm (Fig. ). Malignant lymphomas, in most instances, could not be diagnosed with certainty from lymph-node aspirates. In 8 cases, the diagnosis of malignant lymphorecticular neoplasm had been suggested, but in the remaining three cases, scanty cellularity and poor preservation of the smears precluded proper diagnostic assessment. The differential diagnosis of histiocytic lymphoma, Hodgkin's disease, atypical lymphoid hyperplasia, and granulomatous lymphadenitis could not be resolved with certainty in several cases. In these lesions, we found pleomorphic and mixed lymphoid elements with a large number of small but atypical lymphocytes, occasional eosinophils, and mono- or multi-nucleated "epithelioid cells." Lymphocytic lymphomas, on the other hand, showed a fairly monomorphic spread of atypical lymphocytes, which was rarely imitated by nonneoplastic lymphoreticular lesions (Fig. 7). We found aspirates from granulomatous lesions especially difficult to interpret cytologically. However, the necrotic background and bizarre pleomorphism of the cells in such lesions called our attention to lesions other than neoplastic lesions (Fig. 8). Metastatic soft tissue sarcomas were rare examples in this study (Table 4). On the other hand, we had the opportunity to see several aspirates from metastatic malignant melanoma. Of interest was the striking resemblance of the malignant melanocytes in aspiration smears to those in exfoliative smears (Fig. 9). Discussion This study presents the cytologic and clinicopathologic findings of a relatively large series of lymph- Downloaded from on March 08

6 47 BETSILL AND HAJDU A.J.C.P. April 980 FIG.. Mammary carcinoma metastatic to axillary lymph node. (Left). Clusters and single cells of duct carcinoma. The cytoplasm, as well as the nuclei, varies in size and shape. Hematoxylin and eosin. x40. (Right). A loosely arranged cluster of oval and round cells of medullary carcinoma. Note the intracytoplasmic granules and vacules. Hematoxylin and eosin. x40. node aspiration biopsies from four anatomic locations in males and females of ages 0 to over 80 years. Specific reports concerning the interpretation and feasibility of lymph-node aspiration cytology are relatively few. ',,8,94, In Europe, the technic is more widely but less discriminately used than in the United States. However, at Memorial Hospital, this procedure has been performed in selected cases for over fifty years. Hajdu and Melamed 4 have previously reported the efficacy of aspiration biopsies from numerous sites and have stated that the breast (4%), lymph nodes (%), and bone and soft tissue (9%) composed the majority of aspiration cases. However, few in-depth studies of lymph-node aspirates have been undertaken, and in the United States, studies on this subject based on a large series of cases have not been done. A brief review of the literature mentioned above points out problematic areas in the interpretation of cytologic smears prepared from aspirated material from lymph nodes. Our study notes three basic problems involved in the proper interpretation of lymph node aspirates. The first and most widely recognized problem concerns the diagnosis of primary lymphoid neoplasms from nodal aspirates. As early as 97, Forkner, 7 using supravital stains, reported relatively little difficulty in evaluating primary lymphoid neoplasms. However, in 9 Stewart, from this institution, voiced skepticism concerning the diagnosis of malignant lymphoma and reserved the procedure, for the most part, to the identification of metastatic disease. In 94 Loseke and Craver, also from Memorial Hospital, reported their experience using the procedure for the diagnosis of cases of Hodgkin's disease. Fourteen cases were considered positive cytologically, and required subsequent formal biopsy. In 9, Morrison and associates 0 reported the results of aspirations and briefly presented an oversimplified chart regarding diagnosis of normal, infectious, and primary neoplastic diseases in lymph nodes. Godwin 9 stated that the diagnostic effectiveness in malignant lymphomas is about 0%, whereas for metastatic carcinoma the accuracy is about 90% (a figure reflected in several other studies).,,4,0,, Cardoza, reporting a total series of, nodal aspirations, advocated confidence in the use of aspiration biopsy to diagnose primary lymphoid neoplasms. This series included 47 cases of lymphoma and a total of 70 cases of malignant disease. Downloaded from on March 08

7 «f -# iw* ^ 4f> m * VJ» «* % ****** c H * * % w A'*n. - ^jp fc i * ( IT f * " ^w $ f j^g < ). ^ IF FIG. (upper, left). Oat cell carcinoma metastatic to supraclavicular lymph node. The tumor cells are pyknotic and polygonal naked forms. Most of the tumor cells are in tight clusters. Hematoxylin and eosin. x40. FIG. 7 (upper, right). Malignant lymphoma, poorly differentiated lymphocytic type. This is almost a pure culture of poorly differentiated lymphocytes. Notice the numerous abnormal mitotic figures and occasional larger "histiocytic" cells. Hematoxylin and eosin. x40. FIG. 8 (lower, left). Aspiration smear of granulomatous lymphadenitis. A cluster of bizarre epithelioid cells with pale round nuclei and branching abundant cytoplasm are seen admixed with mature lymphocytes. Hematoxylin and eosin. x40. FIG. 9 (lower, right). Metastatic malignant melanoma. A pure culture of neoplastic cells ranging in size and shape from plasma cells to large polygonal forms. The nuclei, as well as the cytoplasm, are mostly round or oval. Note the giant nucleoli and occasional intranuclear vacoules. Hematoxylin and eosin. x40. Downloaded from on March 08

8 478 BETSILL AND HAJDU A.J.C.P. April 980 Hajdu and Melamed' 4 reaffirmed the idea that aspiration should not be routinely used for diagnosing primary lymphoid neoplasms. In their series, 9% of the nodal aspirates with diagnoses of malignancy represented metastatic disease. In 97 Frable 8 reported the results of 7 nodal aspirates, which included 97 metastatic tumors and lymphomas. He advocated the use of a metachromatic stain (Metachrome B) and Papanicolaou stain on aspirates suspect for primary lymphoid neoplasm. He stated that a characteristic empty appearance may be present in the lymphocyte nucleus in malignant cases, and that this appearance, together with a uniform cellular pattern and lack of phagocytosis, supports a diagnosis of lymphoma. Our series included cases of lymphoma. A review of these cases only confirmed the previously reported opinions of other authors from this institution. In several of these cases, the diagnosis from the initial aspirate reflected diagnostic suspicion, for example, "suspect lymphoma" or "suggest biopsy". In our experience, many cases which later proved to be histiocytic lymphoma or Hodgkin's disease could not with certainty be differentiated from some cases proven by biopsy to be tuberculous lymphadenopathy or infectious lymphadenopathies of other types. Several cases of granulomatous lymphadenitis or infectious lymphadenopathy in our study revealed very little associated necrosis in the aspirates. In most instances, the poorly differentiated lymphocytic lymphomas appeared to be more suggestive of malignant lymphoma than did the well-differentiated ones. Another problem concerns the identification of primary sites of the tumor by examining nodal aspirates. In 9 Stich discussed metastatic disease in nodal aspirates and expressed confidence in the diagnosis of thyroid carcinoma and bronchogenic carcinoma, but he could not accurately correlate the cytomorphology of aspirates and the histologic appearance of the tumors of other sites. Cardoza found that the primary tumor site could be determined "in a majority of the cases" by relying on clinical history and other laboratory findings. Our experience with this problem is similar; we were confident in determining the primary site of some welldifferentiated metastatic tumors. However, it is unfortunate that poorly differentiated tumors are more prone to metatasize than the well-differentiated ones, and one should take this into account when attempting to classify neoplasms by specific type. Our series includes too few cases of various adenocarcinomas to present a detailed description of the correlation between nodal cytology and follow-up identification of the primary tumor sites. The overwhelming majority of metastatic adenocarcinomas were of mammary origin, and many of these gave the characteristic pattern suggestive of infiltrating duct carcinoma. However, the patterns of other metastatic adenocarcinomas can appear identical to this pattern. Most of our cases of epidermoid carcinoma were from the oral, upper respiratory, or lower respiratory tracts, and could not be differentiated further. In nodal aspirates, oat cell carcinoma and metastatic melanoma reveal features that make specific diagnosis feasible. We think that great care should be taken in attempting to furnish to the clinician specific information from nodal aspirates in cases with no prior history of malignancy. One can be much more supportive in those cases of suspected recurrence or metastases of known tumors. The prior pathologic material should be obtained for comparison whenever possible. The diagnostic problem concerning nodal aspirates with associated necrosis, fibrosis, or extensive squamous debris has been discussed in many reports. Review of several smaller series ' 40 reveals that these findings resulted in the majority of false-positive or false-negative diagnoses. Engzell and Zajicek published their experience with the diagnosis of congenital cysts in the neck and metastatic squamous carcinomas. They found that % of smears from squamous carcinoma (00 total cases) revealed only cytologically benign cells and that, in certain cases, differential diagnosis could not be made based on cytologic examination alone. We advocate caution in diagnosing smears with marked necrosis, fibrosis, or extensive squamous debris. It would be in order to suggest formal biopsy in such cases if clinically indicated, rather than to attempt to make unwarranted diagnoses. In this study we presented the cytologic findings, with clinical and pathologic correlation, of a relatively large series of nodal aspirations and attempted to compare these findings with those which others have reported. We think that the procedure of nodal aspiration cytology is a valuable tool when used in the proper clinical setting, and when supported by appropriate clinical, historical, and other diagnostic data. We would like to emphasize the importance of good clinical and pathologic judgement in deciding whether the aspirate is representative of a specific lesion or whether diagnosis should be deferred. References. Block M: Comparative study of lymph node cytology by puncture and histopathology. Acta Cytol :9-44, 97. Cardoza PL: The cytologic diagnosis of lymph node puncture. Acta Cytol 8:94-0, 94. Chatard JA, Guthrie CG: Human trypanisomiasis; report of a Downloaded from on March 08

9 Vol. 7 No. 4 ASPIRATION OF LYMPH NODES 479 case observed in Baltimore. Am J Trop Dis Prevent Med :49-0, Chu EW, Hoye RC: The clinician and the cytopathologist evaluate fine needle aspiration cytology. Acta Cytol 7: 4-47, 97. Engzell U, Jakobsson PA, Sigurdson A, et al: Aspiration biopsy of metastatic carcinoma in lymph nodes of the neck a review of 0 consecutive cases. Acta Oto Laryng 7: 8-47, 97. Engzell U, Zajicek J: Aspiration biopsy of tumors of the neck. I. Aspiration biopsy and cytologic findings in 0 cases of congenital cysts. Acta Cytol 4:-7, Forkner CE: Material from lymph nodes of man. Arch Intern Med 40:47-0, Frable WJ: Thin needle aspiration biopsy a personal experience with 49 cases. Am J Clin Pathol :8-9, Godwin JT: Aspiration biopsy: technique and application. Ann NY Acad Sci :48-7, 9 0. Godwin JT: Cytologic diagnosis of aspiration biopsies of solid or cystic tumors. Acta Cytol 8:0-, 94. Grieg ED, Gray AC: Lymphatic glands in sleeping sickness. Br Med J :, 904. Guthrie CG: Gland puncture as a diagnostic measure. Bull Johns Hopkins Hosp :-9, 9. Hajdu SI, Hajdu EO: Cytopathology of Sarcomas and Nonepithelial Malignant Tumors. Philadelphia, W.B. Saunders, Hajdu SI, Melamed MR: The diagnostic value of aspiration smears. Am J Clin Pathol 9:0-, 97. Kline TA, Nael HS, Holroyde CP: Needle aspiration biopsy diagnosis of subcutaneous nodules and lymph nodes. JAMA :848-80, 97. Loseke L, Craver LF: The diagnosis of Hodgkin's disease by aspiration biopsy. Blood :7-8, Martin HE, Ellis EB: Aspiration biopsy. Ann Surg :9-8, Martin HE, Ellis EB: Aspiration biopsy. Surg Gynecol Obstet 9:78-89, Martin HE, Stewart FW: Advantages and limitations of aspiration biopsy. Am J Roent Rad Ther :4-47, 9 0. Morrison M, Sanwick AA, Rubenstein J, et al: Lymph node aspiration. Am J Clin Pathol :-, 9. Schour L, Chu EW: Fine needle aspiration in the management of patients with neoplastic disease. Acta Cytol 8:47-47, 974. Stewart, FW: The diagnosis of tumors by aspiration. Am J Pathol 9:80-8, 9. StichM: Lymph node aspiration. AmJ Med Sci 4:-, 9 4. Ward GR: Bedside Haematology Philadelphia, W.B. Saunders, 94 Downloaded from on March 08