Transcutaneous Fine-Needle Aspiration Biopsy of Pancreatic Cancer Kitai K i m , * Robert Booth, and Jonathan Myles Medical College of Ohio, Department of Pathology, Toledo, OH 43699
Summary Transcutaneous fine-needle aspiration (FNA) biopsies were performed on 30 patients with space-occupying iesi~ns in the pancreas. Patient selection for FNA was based on pancreatic lesions suspected of malignancy by clinical and/or radiolegic imaging methods. Patient~ with obstructive jaundice were excluded because of the necessity of laparotomy to relieve biliary obstruction. Of these 30 ca~es, 15 cases were positive for malignant cells. Of these 15 positive cases, 11 were adenocarcinomas of duct cell origin, one cystadenocarcinoma, one islet cell tumor, one lymphoma, and one metastatic malignant melanoma. The remaining 15 cases, negative for malignant cells, were ultimately shown to have the following lesions: five adenocarcinomas of the pancreas, one metastatic carcinoma from the breast, and nine nonneoplastic disease. Overall diagnostic accuracy was 80~ The sensitivity, specificity, and predictivity for positive FNA biopsy of pancreatic malignancy were "/9, 1t30, and 100%, respectively. Cytologic features with corresponding histologic types were presented. Immunocytochemical and electron microscopic studies were performed on some cases. Only one complication that was possibly related to FNA biopsy occurred. In conclusion, transcutaneous FNA biopsy of pancreatic malignancy is highly valuable in patient management.
Key Words: Pancreatic cancer; fine-needle aspiration biopsy. INTRODUCTION Epidemiotogic data indicate that incidence and mortality rates of carcinoma of the pancreas were steadily increasing up to 1970, and thereafter reached a plateau ( l ) . The epidemiology o f pancreatic cancer has been controversial. The most consistently documented etiologic factors are cigaret smoking, a high-fat diet, diabetes mel[itus, and an i~dustriat carcinogenic exposure (2-5). Almost all pancreatic cancers are duct ceil adenocarcinomas, with *Author to whom all correspondence and reprint requests should be addressed. International Journal of Pancreatology
61
9 1990 The Humana Press Inc,
62
Kim, Booth, and Myles
65070 occurring in the head of the pancreas, resulting in obstructive jaundice. The rest of the cases occur in the body and tail, and are in an advanced stage at the time of diagnosis. Unfortunately, no early-detection method is available. Ultrasound or computed tomography (CT) is the most common initial diagnostic approach. Traditionally, abnormal lesions detected by imaging were confirmed by open biopsy. Since the early 1970s, fine-needle aspiration (FNA) has made a significant contribution to the preoperative and intraoperative diagnosis of pancreatic cancer (6-7). This presentation includes indications for aspiration, aspiration techniques, and cytologic manifestations in cancer of the pancreas in fine-needle aspirates. MATERIALS AND METHODS Transcutaneous FNA biopsy was performed in 30 cases over the past 10-yr period, with three-quarters of the cases occurring in the last 5 yr. During the same period, 60 cases of cancer of the pancreas were diagnosed at the Medical College Hospital. All of the patients had a lesion in the pancreas that needed to be identified as neoplastic or nonneoplastic. Most of the lesions were fairly well delineated by radiologic imaging (CT or ultrasonography). Patients with obstructive jaundice were excluded from this study, because most underwent intraoperative FNA or core-needle biopsy during exploratory laparotomy, followed by bypass procedures to relieve the obstruction. Aspirations were performed by radiologists under CT or ultrasonography with a Chiba or Westcott 22-gage needle, according to the method of Zajicek (8). The aspirated material is expelled onto a glass slide and smeared by placing another glass slide on top of the expelled cellular material and pulling the two slides apart. The slides are fixed immediately in 95~ ethanol or with a spray fixative. One or two smears are stained with a Diff-Quick procedure, hematoxylin and eosin, or the Papanicalaou stain method. Smears may be saved for immunocytochemistry. Rinsing of the syringe was performed in 20 cases, from which smears, cell blocks, or both were prepared. Immunocytochemistrywas performed by the immunoperoxidase technique, using avidin-biotin complex and commercially available antisera. Papanicolaoustained smears were destained for immunocytochemistry if extra unstained smears were not available. Cellular material for electron microscopy (EM) was fixed in 2070 glutaraldehyde-paraformaldehyde for 4 h and centrifuged to give a sediment. The sediment was rinsed overnight in phosphate buffer and then postfixed in 2~ buffered osmium tetroxide for 2 h. The fixed specimen was dehydrated in graded alcohols and embedded in an epoxy resin. Sections for EM were stained with 6% uranyl acetate and 0.1~ lead citrate and studied with a Philips 410 transmission electron microscope. The final diagnosis of each case was based on histology in 22 cases and the clinical course, radiologic studies, or both in 8 cases. Three cases in the latter groups were lost in followup.
63
Transcutaneous FNA Biopsy Table 1 Correlation between Cytology and Histologic and/or Clinical Course Cytology (FNA) Positive for malignant cells (15)
Negative for malignant cells (15)
Histology/clinical course Adenocarcinoma (11) Cystadenocarcinoma (1) Islet cell tumor (1) Lymphoma (1) Metastatic melanoma (1) Adenocarcinoma (4)--false negative Chronic pancreatitis (4) Pancreatic pseudocyst (2) Nonneoplastic (3) Inadequate cell material (2) a
~ case was ruled as adenocarcinoma of the pancreas with metastasis to the liver based on positive for adenocarcinoma in liver aspirate. The other case was ruled as mammary carcinoma with metastasis to the peripancreatic lymph nodes and the lung, based on positive for adenocarcinoma in tung aspirate. RESULTS Twenty-one of the 30 patients were diagnosed as having malignant neoplasms of the pancreas: 16 adenocarcinomas of duct cell origin, one cystadenocarcinoma, one islet cell tumor, one large cell lymphoma, one metastatic melanoma, and one metastatic breast carcinoma. Correlation between the cytologic diagnosis and diagnosis based on histology, the clinical course, or both is shown in Table 1. All o f the 15 cases that were cytologically positive for malignant cells were shown to have a variety of malignant neoplasms. Fifteen patients with cytology that was negative for malignant cells were ultimately shown to have the following lesions: five adenocarcinomas o f the pancreas, one metastatic carcinoma from the breast, and nine nonneoplastic disease. The aspirate o f one case with metastatic carcinoma contained insufficient cell material for adequate evaluation. The diagnostic yield (overall accuracy), sensitivity, specificity, and predictive values of positive and o f negative results were as follows: 80, 79, 100, 100, and 70%, respectively.
Cytologic Features and Histologic Types Most smears obtained from neoplastic lesions contained adequate cellular material. In two cases with inadequate cellular material, neoplastic cells were identified in the cell-block sections. In two cases, the cellular sample was inadequate in both the smear and cell-block preparations. In one of the four false negative cases, the smears contained an abundance of normal acinar and ductal cells (Fig. 1), presumably from adjacent nonneoplastic pancreas. The cytologic features o f the 12 adenocarcinomas were similar, except for a smaller number o f cells with a lesser degree of atypia in
64
Kim, Booth, and Myles
well-differentiated adenocarcinomas (Fig. 2). The cytologic features of the mucinous cystadenocarcinoma were, except for prominent papillary fragments of neoplasm, not distinct from those of well-differentiated adenocarcinoma (Fig. 3). In poorly differentiated adenocarcinomas, relatively large numbers of cells occurred in clusters, and tended to be pleomorphic (Fig. 4). The cytologic features of melanoma metastatic to the pancreas were similar to those of poorly differentiated adenocarcinomas. The diagnosis of melanoma was confirmed by a positive reaction for S-100 protein by the immunoperoxidase method (Fig. 5). The patient had a previous history of malignant melanoma. The cytologic features of the neuendocrine tumor (islet cell tumor) were distinct from those of adenocarcinoma. The cells occurred singly and in loose aggregate. They were of uniform size and shape (round) with scanty cytoplasm, and their nuclei were uniformly round and hyperchromatic. Chromatin was evenly dispersed and finely granular (Fig. 6). Ceils of the large cell lymphoma of the pancreas were singly isolated, and have scanty cytoplasm and large, cleaved nuclei with prominent single macronucleoli. The cytologic features of lymphomas were distinct from those of other malignant neoplasms in that they were monomorphic and lacked cohesion (Fig. 7). COMMENTS FNA cytology of pancreatic tumors is of great value in terms of its diagnostic sensitivity and specificity, and its cost effectiveness, obviating exploratory taparotomy in inoperable cases. Intraoperative FNA is performed either directly from the lesion or transduodenally; it provides a reasonable diagnostic yield and the least chance of enzyme leakage. The technique is simple, using state-of-the-art imaging methods to delineate the lesion. Diagnostic accuracy has improved over the last 5 yr, with a positive yield in 850-/0 of pancreatic cancers (9,10). Success is closely related to the experience of the radiologist in obtaining good samples. Statistical evaluation of the accuracy of FNA of the pancreas based on our cases is summarized in Table 1. Three out of 4 false negative cases and 2 cases of inadequate cell material occurred in the first 5 yr, and there was a larger number of cases with a better diagnostic yield in the last 5-yr period. Most false negative results were mainly caused by inadequate sampling or by sampling from outside the target lesion. Syringe washing for smears and cell blocks are valuable, on the basis of this study. It is not unusual to see diagnostic cells in syringe-washing smears, cell blocks, or both (Fig. 8). In most studies, complications related to FNA were negligible, although we had one complication that resulted in septicemia. The patient, who died 3 days after the aspiration, underwent endoscopic retrograde choledochopancreatogram (ERCP) and FNA on the same day. Consequently, we are not sure whether this complication was related to the ERCP or to the FNA.
Transcutaneous F N A Biopsy
65
Fig. 1. Normal cells in FNA aspirate of the pancreas. Acinar ceils arranged in loose aggregates with gland pattern (upper left), and duct cells arranged in sheets with honey-comb pattern (center) (Papanicolaou stain, x 500).
Fig. 2. Well-differentiated adenocarcinoma of duct cell origin. (A) Atypical cell aggregates in FNA aspirate (Papanicolaou stain, x500). (B) Histologic features of adenocarcinoma (hematoxylin and eosin, x 300):
Fig. 3. Mucinous cystadenocarcinoma of the pancreas. (A) Papillary neoplastic cell cluster (Papanicolaou stain, x 500). (B) Corresponding histologic features (hematoxylin and eosin, x 125).
Fig. 4. Poorly differentiated adenocarcinoma of duct cell origin. (A) Pleomorphic neoplastic cells arranged in loose aggregates in FNA aspirate (Papanicolaou stain, x 500). (B) Histologic features of poorly differentiated adenocarcinoma (hematoxylin and eosin, x 300).
66
Fig. 5. Metastatic melanoma involving the pancreas. (A) Neoplastic cell clusters in FNA aspirate (Papanicolaou stain, x 500). (B) Neoplastic cells, positive for S-100 protein by immunoperoxidase method in the same FNA aspirate (immunoperoxidase stain using S-100 antisera, • 500).
Fig. 6. Islet cell tumor. (A) Uniform small ceils occurring in loose aggregates and singly isolated in FNA aspirate (-Papanicolaou stain, • 500). (B) Histologic features of the same tumor, showing cords and acini of neuroendocrine cells separated by stroma (positive stain for neuron-specific enolase by immunoperoxidase method using avidin biotin complex, x 80).
67
68
Kirn, Booth, and Myles
Fig. 7. Large cell lymphoma of the pancreas. (A) Singly isolated, large uniform lymphoid cells in FNA aspirate (Papanicolaou stain, • 500). (B) Histologic features of the same tumor, showing large lymphoid cells in sheets (hematoxylin and eosin stain, • (C) Neoplastic cells, positive for leukocyte common antigen ( i m m u n o p e r o x i d a s e method using avidin biotin complex, x 200).
Fig. 8. Well-differentiated adenocarcinoma of the pancreas. Neoplastic cell clusters in cell block section prepared from syringe washing of FNA aspirate (hematoxylin and eosin stain, x 300).
Transcutaneous FNA Biopsy
69
The data compiled by Smith in the United States indicated 47 complications in 63,108 percutaneous FNA biopsies in the abdomen, consisting of 27 cases of hemorrhage, four of peritonitis, and 16 of generalized infections (11). ACKNOWLEDGMENTS
The authors wish to thank Bernard Naylor for reviewing this manuscript, Carrie Greenway for technical assistance, and Marilyn Cline for typing. REFERENCES 1 Silverberg E. Cancer Statistics, 1988. CA 1988; 38: 5-22. 2 Lin RS and Kessler I. A multifactorial model for pancreatic cancer in man. Epidemiologic evidence. JAMA 1981; 245: 145-152. 3 Fraumeni JF Jr. Cancers of the pancreas and biliary tract: epidemiological considerations. Cancer Res. 1975; 35: 3437-3446. 4 Malagelada JR. Pancreatic cancer. An overview of epidemiology, clinical presentation, and diagnosis. Mayo Clin. Proc. 1979; 54: 459-467. 5 Wynder EL. An epidemiological evaluation of the causes of cancer of the pancreas. Cancer Res. 1975; 35: 2228-2233. 6 Christofferson P and Poll P. Preoperative pancreas aspiration biopsies. Acta Pathol. Microbiol. Scand. (Suppt) 1970; 212: 82-92. 7 Evander A, Ihse J, Lunderquist A, Tylen U, and Akerman M. Percutaneous cytodiagnosis of carcinoma of the pancreas and bile duct. Ann. Surg. 1978; 188: 90-92. 8 Stormby N. Pancreas. Monographs in Clinical Cytology, vol. 7, Aspiration Biopsy Cytology, part 2, Cytology of Infradiaphramatic Organs (Weld G, ed.), S. Karger, New York, 1979: 194-211. 9 Koss L, Stanislaw W, and Wlodzimierz O. Aspiration biopsy. Cytologic Interpretation and Histologic Bases, Igaku-Shoin, New York, 1984; 370-392. t0 Frias-Hidvegi D. Guides to Clinical Aspiration Biopsy, Liver and Pancreas, Igaku-Shoin, New York, 1988: 255-258. 11 Smith EH. Fine-needle aspiration biopsy: are there any risks? International Ultrasound, Holm HH, Kristensen JK eds. Thieme, New York, 1985; 169-177.
Summary Transcutaneous fine-needle aspiration (FNA) biopsies were performed on 30 patients with space-occupying iesi~ns in the pancreas. Patient selection for FNA was based on pancreatic lesions suspected of malignancy by clinical and/or radiolegic imaging methods. Patient~ with obstructive jaundice were excluded because of the necessity of laparotomy to relieve biliary obstruction. Of these 30 ca~es, 15 cases were positive for malignant cells. Of these 15 positive cases, 11 were adenocarcinomas of duct cell origin, one cystadenocarcinoma, one islet cell tumor, one lymphoma, and one metastatic malignant melanoma. The remaining 15 cases, negative for malignant cells, were ultimately shown to have the following lesions: five adenocarcinomas of the pancreas, one metastatic carcinoma from the breast, and nine nonneoplastic disease. Overall diagnostic accuracy was 80~ The sensitivity, specificity, and predictivity for positive FNA biopsy of pancreatic malignancy were "/9, 1t30, and 100%, respectively. Cytologic features with corresponding histologic types were presented. Immunocytochemical and electron microscopic studies were performed on some cases. Only one complication that was possibly related to FNA biopsy occurred. In conclusion, transcutaneous FNA biopsy of pancreatic malignancy is highly valuable in patient management.
Key Words: Pancreatic cancer; fine-needle aspiration biopsy. INTRODUCTION Epidemiotogic data indicate that incidence and mortality rates of carcinoma of the pancreas were steadily increasing up to 1970, and thereafter reached a plateau ( l ) . The epidemiology o f pancreatic cancer has been controversial. The most consistently documented etiologic factors are cigaret smoking, a high-fat diet, diabetes mel[itus, and an i~dustriat carcinogenic exposure (2-5). Almost all pancreatic cancers are duct ceil adenocarcinomas, with *Author to whom all correspondence and reprint requests should be addressed. International Journal of Pancreatology
61
9 1990 The Humana Press Inc,
62
Kim, Booth, and Myles
65070 occurring in the head of the pancreas, resulting in obstructive jaundice. The rest of the cases occur in the body and tail, and are in an advanced stage at the time of diagnosis. Unfortunately, no early-detection method is available. Ultrasound or computed tomography (CT) is the most common initial diagnostic approach. Traditionally, abnormal lesions detected by imaging were confirmed by open biopsy. Since the early 1970s, fine-needle aspiration (FNA) has made a significant contribution to the preoperative and intraoperative diagnosis of pancreatic cancer (6-7). This presentation includes indications for aspiration, aspiration techniques, and cytologic manifestations in cancer of the pancreas in fine-needle aspirates. MATERIALS AND METHODS Transcutaneous FNA biopsy was performed in 30 cases over the past 10-yr period, with three-quarters of the cases occurring in the last 5 yr. During the same period, 60 cases of cancer of the pancreas were diagnosed at the Medical College Hospital. All of the patients had a lesion in the pancreas that needed to be identified as neoplastic or nonneoplastic. Most of the lesions were fairly well delineated by radiologic imaging (CT or ultrasonography). Patients with obstructive jaundice were excluded from this study, because most underwent intraoperative FNA or core-needle biopsy during exploratory laparotomy, followed by bypass procedures to relieve the obstruction. Aspirations were performed by radiologists under CT or ultrasonography with a Chiba or Westcott 22-gage needle, according to the method of Zajicek (8). The aspirated material is expelled onto a glass slide and smeared by placing another glass slide on top of the expelled cellular material and pulling the two slides apart. The slides are fixed immediately in 95~ ethanol or with a spray fixative. One or two smears are stained with a Diff-Quick procedure, hematoxylin and eosin, or the Papanicalaou stain method. Smears may be saved for immunocytochemistry. Rinsing of the syringe was performed in 20 cases, from which smears, cell blocks, or both were prepared. Immunocytochemistrywas performed by the immunoperoxidase technique, using avidin-biotin complex and commercially available antisera. Papanicolaoustained smears were destained for immunocytochemistry if extra unstained smears were not available. Cellular material for electron microscopy (EM) was fixed in 2070 glutaraldehyde-paraformaldehyde for 4 h and centrifuged to give a sediment. The sediment was rinsed overnight in phosphate buffer and then postfixed in 2~ buffered osmium tetroxide for 2 h. The fixed specimen was dehydrated in graded alcohols and embedded in an epoxy resin. Sections for EM were stained with 6% uranyl acetate and 0.1~ lead citrate and studied with a Philips 410 transmission electron microscope. The final diagnosis of each case was based on histology in 22 cases and the clinical course, radiologic studies, or both in 8 cases. Three cases in the latter groups were lost in followup.
63
Transcutaneous FNA Biopsy Table 1 Correlation between Cytology and Histologic and/or Clinical Course Cytology (FNA) Positive for malignant cells (15)
Negative for malignant cells (15)
Histology/clinical course Adenocarcinoma (11) Cystadenocarcinoma (1) Islet cell tumor (1) Lymphoma (1) Metastatic melanoma (1) Adenocarcinoma (4)--false negative Chronic pancreatitis (4) Pancreatic pseudocyst (2) Nonneoplastic (3) Inadequate cell material (2) a
~ case was ruled as adenocarcinoma of the pancreas with metastasis to the liver based on positive for adenocarcinoma in liver aspirate. The other case was ruled as mammary carcinoma with metastasis to the peripancreatic lymph nodes and the lung, based on positive for adenocarcinoma in tung aspirate. RESULTS Twenty-one of the 30 patients were diagnosed as having malignant neoplasms of the pancreas: 16 adenocarcinomas of duct cell origin, one cystadenocarcinoma, one islet cell tumor, one large cell lymphoma, one metastatic melanoma, and one metastatic breast carcinoma. Correlation between the cytologic diagnosis and diagnosis based on histology, the clinical course, or both is shown in Table 1. All o f the 15 cases that were cytologically positive for malignant cells were shown to have a variety of malignant neoplasms. Fifteen patients with cytology that was negative for malignant cells were ultimately shown to have the following lesions: five adenocarcinomas o f the pancreas, one metastatic carcinoma from the breast, and nine nonneoplastic disease. The aspirate o f one case with metastatic carcinoma contained insufficient cell material for adequate evaluation. The diagnostic yield (overall accuracy), sensitivity, specificity, and predictive values of positive and o f negative results were as follows: 80, 79, 100, 100, and 70%, respectively.
Cytologic Features and Histologic Types Most smears obtained from neoplastic lesions contained adequate cellular material. In two cases with inadequate cellular material, neoplastic cells were identified in the cell-block sections. In two cases, the cellular sample was inadequate in both the smear and cell-block preparations. In one of the four false negative cases, the smears contained an abundance of normal acinar and ductal cells (Fig. 1), presumably from adjacent nonneoplastic pancreas. The cytologic features o f the 12 adenocarcinomas were similar, except for a smaller number o f cells with a lesser degree of atypia in
64
Kim, Booth, and Myles
well-differentiated adenocarcinomas (Fig. 2). The cytologic features of the mucinous cystadenocarcinoma were, except for prominent papillary fragments of neoplasm, not distinct from those of well-differentiated adenocarcinoma (Fig. 3). In poorly differentiated adenocarcinomas, relatively large numbers of cells occurred in clusters, and tended to be pleomorphic (Fig. 4). The cytologic features of melanoma metastatic to the pancreas were similar to those of poorly differentiated adenocarcinomas. The diagnosis of melanoma was confirmed by a positive reaction for S-100 protein by the immunoperoxidase method (Fig. 5). The patient had a previous history of malignant melanoma. The cytologic features of the neuendocrine tumor (islet cell tumor) were distinct from those of adenocarcinoma. The cells occurred singly and in loose aggregate. They were of uniform size and shape (round) with scanty cytoplasm, and their nuclei were uniformly round and hyperchromatic. Chromatin was evenly dispersed and finely granular (Fig. 6). Ceils of the large cell lymphoma of the pancreas were singly isolated, and have scanty cytoplasm and large, cleaved nuclei with prominent single macronucleoli. The cytologic features of lymphomas were distinct from those of other malignant neoplasms in that they were monomorphic and lacked cohesion (Fig. 7). COMMENTS FNA cytology of pancreatic tumors is of great value in terms of its diagnostic sensitivity and specificity, and its cost effectiveness, obviating exploratory taparotomy in inoperable cases. Intraoperative FNA is performed either directly from the lesion or transduodenally; it provides a reasonable diagnostic yield and the least chance of enzyme leakage. The technique is simple, using state-of-the-art imaging methods to delineate the lesion. Diagnostic accuracy has improved over the last 5 yr, with a positive yield in 850-/0 of pancreatic cancers (9,10). Success is closely related to the experience of the radiologist in obtaining good samples. Statistical evaluation of the accuracy of FNA of the pancreas based on our cases is summarized in Table 1. Three out of 4 false negative cases and 2 cases of inadequate cell material occurred in the first 5 yr, and there was a larger number of cases with a better diagnostic yield in the last 5-yr period. Most false negative results were mainly caused by inadequate sampling or by sampling from outside the target lesion. Syringe washing for smears and cell blocks are valuable, on the basis of this study. It is not unusual to see diagnostic cells in syringe-washing smears, cell blocks, or both (Fig. 8). In most studies, complications related to FNA were negligible, although we had one complication that resulted in septicemia. The patient, who died 3 days after the aspiration, underwent endoscopic retrograde choledochopancreatogram (ERCP) and FNA on the same day. Consequently, we are not sure whether this complication was related to the ERCP or to the FNA.
Transcutaneous F N A Biopsy
65
Fig. 1. Normal cells in FNA aspirate of the pancreas. Acinar ceils arranged in loose aggregates with gland pattern (upper left), and duct cells arranged in sheets with honey-comb pattern (center) (Papanicolaou stain, x 500).
Fig. 2. Well-differentiated adenocarcinoma of duct cell origin. (A) Atypical cell aggregates in FNA aspirate (Papanicolaou stain, x500). (B) Histologic features of adenocarcinoma (hematoxylin and eosin, x 300):
Fig. 3. Mucinous cystadenocarcinoma of the pancreas. (A) Papillary neoplastic cell cluster (Papanicolaou stain, x 500). (B) Corresponding histologic features (hematoxylin and eosin, x 125).
Fig. 4. Poorly differentiated adenocarcinoma of duct cell origin. (A) Pleomorphic neoplastic cells arranged in loose aggregates in FNA aspirate (Papanicolaou stain, x 500). (B) Histologic features of poorly differentiated adenocarcinoma (hematoxylin and eosin, x 300).
66
Fig. 5. Metastatic melanoma involving the pancreas. (A) Neoplastic cell clusters in FNA aspirate (Papanicolaou stain, x 500). (B) Neoplastic cells, positive for S-100 protein by immunoperoxidase method in the same FNA aspirate (immunoperoxidase stain using S-100 antisera, • 500).
Fig. 6. Islet cell tumor. (A) Uniform small ceils occurring in loose aggregates and singly isolated in FNA aspirate (-Papanicolaou stain, • 500). (B) Histologic features of the same tumor, showing cords and acini of neuroendocrine cells separated by stroma (positive stain for neuron-specific enolase by immunoperoxidase method using avidin biotin complex, x 80).
67
68
Kirn, Booth, and Myles
Fig. 7. Large cell lymphoma of the pancreas. (A) Singly isolated, large uniform lymphoid cells in FNA aspirate (Papanicolaou stain, • 500). (B) Histologic features of the same tumor, showing large lymphoid cells in sheets (hematoxylin and eosin stain, • (C) Neoplastic cells, positive for leukocyte common antigen ( i m m u n o p e r o x i d a s e method using avidin biotin complex, x 200).
Fig. 8. Well-differentiated adenocarcinoma of the pancreas. Neoplastic cell clusters in cell block section prepared from syringe washing of FNA aspirate (hematoxylin and eosin stain, x 300).
Transcutaneous FNA Biopsy
69
The data compiled by Smith in the United States indicated 47 complications in 63,108 percutaneous FNA biopsies in the abdomen, consisting of 27 cases of hemorrhage, four of peritonitis, and 16 of generalized infections (11). ACKNOWLEDGMENTS
The authors wish to thank Bernard Naylor for reviewing this manuscript, Carrie Greenway for technical assistance, and Marilyn Cline for typing. REFERENCES 1 Silverberg E. Cancer Statistics, 1988. CA 1988; 38: 5-22. 2 Lin RS and Kessler I. A multifactorial model for pancreatic cancer in man. Epidemiologic evidence. JAMA 1981; 245: 145-152. 3 Fraumeni JF Jr. Cancers of the pancreas and biliary tract: epidemiological considerations. Cancer Res. 1975; 35: 3437-3446. 4 Malagelada JR. Pancreatic cancer. An overview of epidemiology, clinical presentation, and diagnosis. Mayo Clin. Proc. 1979; 54: 459-467. 5 Wynder EL. An epidemiological evaluation of the causes of cancer of the pancreas. Cancer Res. 1975; 35: 2228-2233. 6 Christofferson P and Poll P. Preoperative pancreas aspiration biopsies. Acta Pathol. Microbiol. Scand. (Suppt) 1970; 212: 82-92. 7 Evander A, Ihse J, Lunderquist A, Tylen U, and Akerman M. Percutaneous cytodiagnosis of carcinoma of the pancreas and bile duct. Ann. Surg. 1978; 188: 90-92. 8 Stormby N. Pancreas. Monographs in Clinical Cytology, vol. 7, Aspiration Biopsy Cytology, part 2, Cytology of Infradiaphramatic Organs (Weld G, ed.), S. Karger, New York, 1979: 194-211. 9 Koss L, Stanislaw W, and Wlodzimierz O. Aspiration biopsy. Cytologic Interpretation and Histologic Bases, Igaku-Shoin, New York, 1984; 370-392. t0 Frias-Hidvegi D. Guides to Clinical Aspiration Biopsy, Liver and Pancreas, Igaku-Shoin, New York, 1988: 255-258. 11 Smith EH. Fine-needle aspiration biopsy: are there any risks? International Ultrasound, Holm HH, Kristensen JK eds. Thieme, New York, 1985; 169-177.
