- Email: [email protected]
Predictive value of intra-abdominal lymph nodes in pancreatic endoscopic ultrasonography–guided fine-needle aspiration biopsy
Journal of the American Society of Cytopathology (2014) xx, 1e5
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.jascyto.org/
Predictive value of intra-abdominal lymph nodes in pancreatic endoscopic ultrasonographyeguided fine-needle aspiration biopsy Brian T. Collins, MD*, Cory T. Bernadt, MD, PhD, Laura J. Adhikari, MD, Jeff F. Wang, MD Cytopathology Section, Department of Pathology and Immunology, Washington University in St. Louis School of Medicine, Campus Box 8118, 660 South Euclid Avenue, St. Louis, Missouri Received 13 February 2014; received in revised form 28 March 2014; accepted 31 March 2014
KEYWORDS Fine-needle aspiration; Pancreas; Lymph node; Endoscopic ultrasonography; Adenocarcinoma
Introduction Endoscopic ultrasonography (EUS)-guided fine-needle aspiration (FNA) biopsy is a commonly used method for the evaluation of pancreatic lesions. EUS-guided FNA of the intra-abdominal lymph nodes (LNs) can provide critical diagnostic information that is important for clinical management and tumor staging. This study examines the predictive value of intra-abdominal LN EUS-guided FNA biopsy associated with pancreatic lesions. Materials and methods Over a 10-year period, the pathology database was searched for patients with concurrent pancreas and intra-abdominal LN EUS-guided FNA biopsy. The corresponding reports were reviewed, and clinical information and diagnostic results were recorded. Results There were 252 cases where both a pancreas lesion and intra-abdominal LN were biopsied. Of this group, 182 LNs were classified as negative (72%), 47 as positive (19%), and 23 as atypical (9%). Within the negative LN cohort, the pancreas FNAs fell into the following diagnostic categories: benign (47%), malignant (30%), and atypical/suspicious (23%). Within the positive LN cohort, the pancreas lesion correlated with the following diagnostic categories: malignant (89%), atypical (4%), and suspicious (6%). A positive LN EUS-guided FNA biopsy had a 98% positive predictive value for malignancy. Within the atypical LN cohort, the pancreas correlated with the following diagnostic categories: malignant (57%), atypical/suspicious (26%), and benign (17%). Conclusions An atypical LN diagnostic category is strongly associated with a malignant pancreas lesion. A positive LN EUS-guided FNA biopsy has a 98% positive predictive value for pancreatic malignancy.
This was presented as an abstract at the 103rd Annual Meeting of the United States and Canadian Association of Pathologists (USCAP), March 1 to 7, 2014, San Diego, California. *Corresponding author: Brian T. Collins, MD, Associate Professor of Pathology and Immunology, Department of Pathology and Immunology, Campus Box 8118, 660 South Euclid Avenue, St. Louis, MO 63110-1093; Tel.: 314 747-8159; Fax: 314 747-2663. E-mail address: [email protected] (B.T. Collins). 2213-2945/$36 Ó 2014 American Society of Cytopathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jasc.2014.03.011
2
B.T. Collins et al. A positive diagnostic category for an intra-abdominal LN can provide strong predictive evidence of a corresponding malignancy of the pancreas. Ó 2014 American Society of Cytopathology. Published by Elsevier Inc. All rights reserved.
Introduction Endoscopic ultrasonography (EUS)-guided fine-needle aspiration (FNA) is the standard method for evaluation of masses of the pancreas. Most of these patients present with painless jaundice or weight loss and have a high clinical suspicion for malignancy, primarily adenocarcinoma. The efficacy and effectiveness of the procedure for the diagnosis of malignancy has been well established.1 Due to the nature of the disease process, many patients who initially present will not be resectable due to a variety of factors, including imaging findings of local tumor invasion and presumed regional lymph node (LN) and liver metastases. Therefore, the primary diagnosis is often established by EUS-guided FNA biopsy. In some instances, the regional LN can be abnormal or enlarged. A variety of ultrasonography imaging characteristics can raise the possibility of a pathologic process in the LN, and these include such features as increased size and abnormal shape of the LN with loss of normal architectural configuration.2 Patients with a pancreatic abnormality and abnormal intraabdominal LNs can have them concurrently evaluated during the EUS-guided FNA biopsy procedure. There are select clinical scenarios where the pancreas lesion will not yield diagnostic material (as judged by a rapid on-site evaluation [ROSE] or on the final interpretation); and the abnormal intraabdominal LNs will be evaluated along with the pancreas during a single interventional procedure to assist in providing a diagnosis. Some primary pancreatic adenocarcinomas are nondiagnostic or below the threshold of definitive diagnosis (atypical to suspicious); this can be for a variety of reasons including among other contributing factors: extensive tumor desmoplasia, which makes obtaining a sufficient sample difficult; anatomic location, which can be more technically difficult to biopsy; lesional hemorrhage with hemodilution of the FNA biopsy; and processing or fixation problems. In these circumstances, the EUS-guided FNA biopsy of abnormal intra-abdominal LNs has the potential to provide a definitive diagnosis and therefore prevent the patient from having a delay in diagnosis, repeat interventional EUS-guided FNA biopsy procedure at a later date, or other invasive procedure (such as surgery or laparoscopy) to establish the diagnosis. The purpose of the study was to determine the predictive value of intra-abdominal LN EUS-guided FNA biopsy and pancreatic EUS-guided FNA biopsy in patients with pancreatic lesions.
Materials and methods The pathology database at Barnes-Jewish Hospital was retrospectively searched for EUS-guided FNA biopsy cases where biopsy of both the pancreas and intra-abdominal LNs
were performed. This occurred over a consecutive 10-year period from 2001 to 2011. These were limited to single procedures where both the pancreas and intra-abdominal LNs were sampled concomitantly. They did not include separate procedures where other sites or 1 site alone was biopsied. LN sites were limited to regional and intraabdominal locations. The cytopathology reports were collected, and diagnostic terminology provided on the reports was reviewed and categorized. When present, corresponding or subsequent pertinent surgical pathology reports were collected. All pathology reports were reviewed, and clinical information and diagnostic results were recorded. Where appropriate, the subsequent clinical course and follow-up was ascertained and the clinical demographics of the patients were reviewed. For purposes of categorization, the final results were placed in general categories, which included malignancy, suspicious for malignancy, atypical, and negative/benign. From a results standpoint, this patient cohort is primarily examined through a filter of the EUSguided FNA biopsy category for the intra-abdominal LN results and not primarily the pancreas. EUS-guided FNA biopsy performance included direct aspirate smears that were both air-dried and alcohol-fixed. Alcohol-fixed smears were stained by a Papanicolaou method and air-dried smears were stained by a modified Wright-Giemsa method. Needle rinse material was used for either cytospin slides or standard cell block method. The study was approved by the Institutional Review Board.
Results There were 252 cases where a pancreas lesion and intraabdominal LN were sampled together over the 10-year consecutive search. The majority of patients were male (62%), with media age of 60.5 years and an age range of 22 to 88 years. The majority of pancreas lesions were located in the head (57.5%) and the majority were solid (90.5%). The patient demographics are presented in Table 1. All the EUSguided FNA biopsies were performed without ROSE. This was performed in a large academic tertiary referral-based health care environment with an experienced group of active endoscopists and numerous individual pathologists with varying degrees of experience. Of the total 252-case cohort group, 182 LNs were classified as negative (72%), 47 as positive (19%), and 23 as atypical (9%) (Table 2). Within the negative LN cohort (182 cases), the pancreas lesion FNA correlated with the following diagnostic categories: 84 cases were benign (47%); 55 cases were malignant (30%); and 43 cases were atypical/suspicious (23%) (Table 3).
Predictive Value of LN EUS FNA Table 1 features.
3
Patient demographics and pancreatic lesion
Characteristics
Totals
Total cases Male Female Median age, yrs Age range, yrs Pancreas mass location Head Body Tail Uncinate Neck Not otherwise specified Solid Cystic
252 156 96 60.5 22-88 145 37 20 13 12 25 228 24
Percentage 62 38
Table 2
EUS FNA biopsy categories for LNs.
Category
Total
Percentage
Negative Positive Atypical Total
182 47 23 252
72 19 9 100
Abbreviations: EUS, endoscopic ultrasonography; FNA, fine-needle aspiration; LN, lymph node.
57.5 14.7 7.9 5.2 4.8 9.9 90.5 9.5
Within the positive LN cohort (47 cases), the pancreas lesion EUS-guided FNA correlated with the following diagnostic categories: 42 cases were malignant (89%); 3 cases were suspicious (6%); and 2 cases were atypical (4%). Including surgical follow-up and clinical outcome, a positive LN EUS-guided FNA biopsy had a 98% positive predictive value (PPV) for malignancy in the pancreas (Table 4). The 1 patient without primary pancreatic malignancy had a known hepatic cholangiocarcinoma with regional metastatic disease. Within the atypical LN cohort (23 cases), the pancreas lesion EUS-guided FNA correlated with the following diagnostic categories: 13 cases were malignant (57%); 6 cases were atypical/suspicious (26%); and 4 cases were benign (17%). Including surgical follow-up and clinical outcome (18 with malignancy and 5 without malignancy), an atypical LN EUSguided FNA biopsy had a 78% PPV for malignancy in the pancreas (Table 5). Cases with an “atypical” designation were generally instances where the EUS-guided FNA LN biopsy did not provide definitive diagnostic material; and this was typically related to a scant to minimal epithelial population with some variable cellular epithelial atypia. The qualitative and quantitative changes did not permit a definitive diagnosis of malignancy, and therefore these were classified as atypical.
Discussion EUS-guided FNA biopsy of the pancreas is the standard method for evaluating pancreatic lesions where there is a concern for neoplasia. When the endoscopist examines a patient, the ultrasonographic review can disclose a variety of regional imaging findings, including abnormal or enlarged LNs. In various clinical circumstances, a EUS-guided FNA biopsy of the LN can provide diagnostic or additional information that is helpful in patient management. Two of the more common clinical questions are the ability of the regional lymph node to provide a definitive diagnosis of
malignancy where the pancreatic lesion is negative to suspicious, and definitive documentation of a combined pancreatic and regional lymph node malignancy, which has the ability to impact the option for resection or further interventional procedures. The necessity for regional LN biopsy rests with the performing endoscopist and can be impacted by a variety of factors including the image appearance such as local or diffuse lymphadenopathy, imaging features of the pancreas lesion, preliminary FNA biopsy results for the pancreas lesion, and contributing clinical findings such as other concomitant malignancies.3,4 This study exclusively focused on patients with concurrent EUS-guided FNA biopsy of the pancreas and regional lymph nodes and the predictive value of the LN result in relation to the pancreas biopsy findings. No previous studies have focused exclusively on this correlation with a large series of pancreas and intra-abdominal lymph nodes sampled by EUS-guided FNA.5 Sampling of intraabdominal LN by EUS-guided FNA can provide a diagnosis but is not always diagnostic.6 For the 252 cases in this series, 62% were men with a median age of 60.5 years and the majority of the pancreas lesions present in the head (57%). This is in line with the typical pancreatic EUSguided FNA biopsy patient profile. Most EUS-guided FNA LN biopsies were negative (182 of 252) (72%). When examining the corresponding pancreatic EUS-guided FNA biopsy results for these patients, almost one half were also negative (84 of 182) (47%). About one third were definitively diagnostic for malignancy (30%) and the remaining 23% fell within the atypical/suspicious category. If the positive category and the atypical/ suspicious category were grouped together, they compared about evenly with the negative category. Therefore, a patient with a negative EUS-guided FNA intra-abdominal LN biopsy has about an equal chance for a negative or abnormal Table 3 Negative EUS FNA biopsy intra-abdominal LN category with corresponding pancreas result. Pancreas category
Total
Percentage
Negative Malignant Atypical/suspicious Total
84 55 43 182
46 30 24 100
Abbreviations as in Table 2.
4
B.T. Collins et al. Table 4 Comparison of pancreatic diagnostic categories among patients with positive LN by EUS FNA. Pancreas category
Total
Percentage
Malignant Suspicious Atypical Total PPV
42 3 2 47 98%
89.3 6.4 4.3 100
Abbreviations: PPV, positive predictive value; other abbreviations as in Table 2.
EUS-guided FNA pancreas biopsy result. And in almost one half of the cases, a negative EUS-guided FNA LN biopsy will support the negative categorization provided by the concurrent pancreatic biopsy. Therefore, when faced with a negative pancreatic EUS-guided FNA diagnosis, there is a very small chance of the LN being abnormal or providing a diagnosis. Overall, definitive positive EUS-guided FNA LN biopsies accounted for 19% of the total cases. And within this group of 47 cases, the majority of pancreas biopsies were definitively diagnostic for malignancy (89%) with the remaining cases in the atypical/suspicious category (4%/6%). There were no benign or negative pancreas biopsy results where the EUS-guided FNA LN was positive. When including the subsequent surgical findings and clinical follow-up for the group, there was a PPV of 98% for pancreatic malignancy in cases with a positive EUS-guided FNA biopsy of intraabdominal LN. The 1 positive LN EUS-guided FNA without pancreatic malignancy (by EUS-guided FNA biopsy, surgical confirmation, or subsequent clinical follow-up) involved a patient with known cholangiocarcinoma. A cyst in the uncinate process was evaluated by EUS-guided FNA biopsy and categorized as atypical with features of pancreatitis. The pancreatic cystic lesion was clinically suspicious for intraductal papillary mucinous tumor. The corresponding celiac LN was diagnosed as poorly differentiated carcinoma, which represented metastatic cholangiocarcinoma. Positive LNs with metastatic carcinoma certainly can represent spread from nonpancreas sources, and this consideration should be kept in mind when examining and reporting such cases. Additionally, metastatic processes that do not originate in the pancreas can provide a clinical picture of a pancreas mass and lymphadenopathy. An awareness of any Table 5
Atypical LN category with pancreatic EUS FNA.
Pancreas category
Total
Percentage
Malignant Atypical/suspicious Benign Total PPV
13 6 4 23 78%
57 26 17 100
Abbreviations as in Tables 2 and 4.
previous malignancy diagnoses and special attention to the morphologic and immunohistochemistry differentiation of any malignancy can be important. Within the EUS-guided FNA LN biopsy group, there were 23 cases (9%) in the atypical category. This cohort is most commonly a consequence of insufficiency of necessary quantitative and qualitative features for definitive benign or malignant categorization. As experienced pathologists are aware, sometimes there is minimal sampling of a LN or if there is partial involvement of a LN by a malignant process, there can be too few groups of concern to permit definitive categorization. Mucosal epithelial sampling also often occurs when intra-abdominal LNs are sampled by EUS-guided FNA and these groups, though usually obviously benign or small intestinal in origin, can raise questions about their precise origin. Nevertheless, within this atypical LN category, the majority of pancreas biopsies were malignant (57%) with atypical/suspicious accounting for about a quarter of cases (26%). Only 4 cases (17%) had a corresponding benign pancreas biopsy. When including the subsequent surgical findings and clinical follow-up, there was a PPV of 78% for pancreatic malignancy in cases with an atypical category EUS-guided FNA biopsy of intraabdominal LNs. The order of biopsy sites where both pancreas and LN are sampled can vary, and ROSE is provided, this has the potential to impact the disposition of the procedure. In some circumstances, it can be easier to initially biopsy the intraabdominal LN before the pancreatic lesion. If ROSE is provided, preliminary interpretative observations can be provided to the performing clinician about the LN FNA biopsy that can then help further to guide the procedure. If the LN biopsy provides a positive category, there might not be a need to perform a biopsy on the pancreas mass because a definitive malignant EUS-guided FNA LN biopsy has the ability to provide both a diagnosis and staging information for the patient. Similarly, when ROSE is provided and the pancreatic lesion is not diagnostic for malignancy, the interventional endoscopist can choose to sample regional, abnormal LNs in an attempt to obtain a definitive diagnosis with the knowledge that there is some uncertainty about the final pancreatic diagnostic category.
Conclusions In this study, most EUS-guided FNA LN biopsies were negative and supported a concurrent negative pancreatic biopsy in about one half the cases (47%). For the 20% of cases with a positive diagnostic EUS-guided FNA LN biopsy, they all had corresponding abnormal EUS-guided FNA pancreas biopsies and 89% were definitively categorized as malignant. There was a PPV of 78% for pancreatic malignancy in cases with an atypical category EUS-guided FNA biopsy of intra-abdominal LNs. For cases with a positive category EUS-guided FNA biopsy of intra-
Predictive Value of LN EUS FNA abdominal LN, there was a PPV of 98% for pancreatic malignancy. Therefore, a positive diagnostic category for an intra-abdominal LN can provide strong predictive evidence of a corresponding malignancy of the pancreas.
Funding sources
5
2.
3.
This work had no specific funding.
Conflict of interest disclosures
4.
The authors have no financial disclosures to report.
5.
References 6. 1. Raut CP, Grau AM, Staerkel GA, et al. Diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration in patients with
presumed pancreatic cancer. J Gastrointest Surg. 2003;7:118e126. discussion 127-128. Korenblit J, Anantharaman A, Loren DE, Kowalski TE, Siddiqui AA. The role of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) for the diagnosis of intra-abdominal lymphadenopathy of unknown origin. J Intervent Gastroenterol. 2012;2: 172e176. Kumon RE, Pollack MJ, Faulx AL, et al. Characterization of pancreatic cancer and intra-abdominal lymph node malignancy using spectrum analysis of endoscopic ultrasound imaging. Conf Proc IEEE Eng Med Biol Soc. 2009;2009:1949e1952. Kumon RE, Pollack MJ, Faulx AL, et al. In vivo characterization of pancreatic and lymph node tissue by using EUS spectrum analysis: a validation study. Gastrointest Endosc. 2010;71:53e63. Peng HQ, Greenwald BD, Tavora FR, et al. Evaluation of performance of EUS-FNA in preoperative lymph node staging of cancers of esophagus, lung, and pancreas. Diagn Cytopathol. 2008;36: 290e296. Bluen BE, Lachter J, Khamaysi I, et al. Accuracy and quality assessment of EUS-FNA: a single-center large cohort of biopsies. Diagn Ther Endosc. 2012;2012:139563.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.jascyto.org/
Predictive value of intra-abdominal lymph nodes in pancreatic endoscopic ultrasonographyeguided fine-needle aspiration biopsy Brian T. Collins, MD*, Cory T. Bernadt, MD, PhD, Laura J. Adhikari, MD, Jeff F. Wang, MD Cytopathology Section, Department of Pathology and Immunology, Washington University in St. Louis School of Medicine, Campus Box 8118, 660 South Euclid Avenue, St. Louis, Missouri Received 13 February 2014; received in revised form 28 March 2014; accepted 31 March 2014
KEYWORDS Fine-needle aspiration; Pancreas; Lymph node; Endoscopic ultrasonography; Adenocarcinoma
Introduction Endoscopic ultrasonography (EUS)-guided fine-needle aspiration (FNA) biopsy is a commonly used method for the evaluation of pancreatic lesions. EUS-guided FNA of the intra-abdominal lymph nodes (LNs) can provide critical diagnostic information that is important for clinical management and tumor staging. This study examines the predictive value of intra-abdominal LN EUS-guided FNA biopsy associated with pancreatic lesions. Materials and methods Over a 10-year period, the pathology database was searched for patients with concurrent pancreas and intra-abdominal LN EUS-guided FNA biopsy. The corresponding reports were reviewed, and clinical information and diagnostic results were recorded. Results There were 252 cases where both a pancreas lesion and intra-abdominal LN were biopsied. Of this group, 182 LNs were classified as negative (72%), 47 as positive (19%), and 23 as atypical (9%). Within the negative LN cohort, the pancreas FNAs fell into the following diagnostic categories: benign (47%), malignant (30%), and atypical/suspicious (23%). Within the positive LN cohort, the pancreas lesion correlated with the following diagnostic categories: malignant (89%), atypical (4%), and suspicious (6%). A positive LN EUS-guided FNA biopsy had a 98% positive predictive value for malignancy. Within the atypical LN cohort, the pancreas correlated with the following diagnostic categories: malignant (57%), atypical/suspicious (26%), and benign (17%). Conclusions An atypical LN diagnostic category is strongly associated with a malignant pancreas lesion. A positive LN EUS-guided FNA biopsy has a 98% positive predictive value for pancreatic malignancy.
This was presented as an abstract at the 103rd Annual Meeting of the United States and Canadian Association of Pathologists (USCAP), March 1 to 7, 2014, San Diego, California. *Corresponding author: Brian T. Collins, MD, Associate Professor of Pathology and Immunology, Department of Pathology and Immunology, Campus Box 8118, 660 South Euclid Avenue, St. Louis, MO 63110-1093; Tel.: 314 747-8159; Fax: 314 747-2663. E-mail address: [email protected] (B.T. Collins). 2213-2945/$36 Ó 2014 American Society of Cytopathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jasc.2014.03.011
2
B.T. Collins et al. A positive diagnostic category for an intra-abdominal LN can provide strong predictive evidence of a corresponding malignancy of the pancreas. Ó 2014 American Society of Cytopathology. Published by Elsevier Inc. All rights reserved.
Introduction Endoscopic ultrasonography (EUS)-guided fine-needle aspiration (FNA) is the standard method for evaluation of masses of the pancreas. Most of these patients present with painless jaundice or weight loss and have a high clinical suspicion for malignancy, primarily adenocarcinoma. The efficacy and effectiveness of the procedure for the diagnosis of malignancy has been well established.1 Due to the nature of the disease process, many patients who initially present will not be resectable due to a variety of factors, including imaging findings of local tumor invasion and presumed regional lymph node (LN) and liver metastases. Therefore, the primary diagnosis is often established by EUS-guided FNA biopsy. In some instances, the regional LN can be abnormal or enlarged. A variety of ultrasonography imaging characteristics can raise the possibility of a pathologic process in the LN, and these include such features as increased size and abnormal shape of the LN with loss of normal architectural configuration.2 Patients with a pancreatic abnormality and abnormal intraabdominal LNs can have them concurrently evaluated during the EUS-guided FNA biopsy procedure. There are select clinical scenarios where the pancreas lesion will not yield diagnostic material (as judged by a rapid on-site evaluation [ROSE] or on the final interpretation); and the abnormal intraabdominal LNs will be evaluated along with the pancreas during a single interventional procedure to assist in providing a diagnosis. Some primary pancreatic adenocarcinomas are nondiagnostic or below the threshold of definitive diagnosis (atypical to suspicious); this can be for a variety of reasons including among other contributing factors: extensive tumor desmoplasia, which makes obtaining a sufficient sample difficult; anatomic location, which can be more technically difficult to biopsy; lesional hemorrhage with hemodilution of the FNA biopsy; and processing or fixation problems. In these circumstances, the EUS-guided FNA biopsy of abnormal intra-abdominal LNs has the potential to provide a definitive diagnosis and therefore prevent the patient from having a delay in diagnosis, repeat interventional EUS-guided FNA biopsy procedure at a later date, or other invasive procedure (such as surgery or laparoscopy) to establish the diagnosis. The purpose of the study was to determine the predictive value of intra-abdominal LN EUS-guided FNA biopsy and pancreatic EUS-guided FNA biopsy in patients with pancreatic lesions.
Materials and methods The pathology database at Barnes-Jewish Hospital was retrospectively searched for EUS-guided FNA biopsy cases where biopsy of both the pancreas and intra-abdominal LNs
were performed. This occurred over a consecutive 10-year period from 2001 to 2011. These were limited to single procedures where both the pancreas and intra-abdominal LNs were sampled concomitantly. They did not include separate procedures where other sites or 1 site alone was biopsied. LN sites were limited to regional and intraabdominal locations. The cytopathology reports were collected, and diagnostic terminology provided on the reports was reviewed and categorized. When present, corresponding or subsequent pertinent surgical pathology reports were collected. All pathology reports were reviewed, and clinical information and diagnostic results were recorded. Where appropriate, the subsequent clinical course and follow-up was ascertained and the clinical demographics of the patients were reviewed. For purposes of categorization, the final results were placed in general categories, which included malignancy, suspicious for malignancy, atypical, and negative/benign. From a results standpoint, this patient cohort is primarily examined through a filter of the EUSguided FNA biopsy category for the intra-abdominal LN results and not primarily the pancreas. EUS-guided FNA biopsy performance included direct aspirate smears that were both air-dried and alcohol-fixed. Alcohol-fixed smears were stained by a Papanicolaou method and air-dried smears were stained by a modified Wright-Giemsa method. Needle rinse material was used for either cytospin slides or standard cell block method. The study was approved by the Institutional Review Board.
Results There were 252 cases where a pancreas lesion and intraabdominal LN were sampled together over the 10-year consecutive search. The majority of patients were male (62%), with media age of 60.5 years and an age range of 22 to 88 years. The majority of pancreas lesions were located in the head (57.5%) and the majority were solid (90.5%). The patient demographics are presented in Table 1. All the EUSguided FNA biopsies were performed without ROSE. This was performed in a large academic tertiary referral-based health care environment with an experienced group of active endoscopists and numerous individual pathologists with varying degrees of experience. Of the total 252-case cohort group, 182 LNs were classified as negative (72%), 47 as positive (19%), and 23 as atypical (9%) (Table 2). Within the negative LN cohort (182 cases), the pancreas lesion FNA correlated with the following diagnostic categories: 84 cases were benign (47%); 55 cases were malignant (30%); and 43 cases were atypical/suspicious (23%) (Table 3).
Predictive Value of LN EUS FNA Table 1 features.
3
Patient demographics and pancreatic lesion
Characteristics
Totals
Total cases Male Female Median age, yrs Age range, yrs Pancreas mass location Head Body Tail Uncinate Neck Not otherwise specified Solid Cystic
252 156 96 60.5 22-88 145 37 20 13 12 25 228 24
Percentage 62 38
Table 2
EUS FNA biopsy categories for LNs.
Category
Total
Percentage
Negative Positive Atypical Total
182 47 23 252
72 19 9 100
Abbreviations: EUS, endoscopic ultrasonography; FNA, fine-needle aspiration; LN, lymph node.
57.5 14.7 7.9 5.2 4.8 9.9 90.5 9.5
Within the positive LN cohort (47 cases), the pancreas lesion EUS-guided FNA correlated with the following diagnostic categories: 42 cases were malignant (89%); 3 cases were suspicious (6%); and 2 cases were atypical (4%). Including surgical follow-up and clinical outcome, a positive LN EUS-guided FNA biopsy had a 98% positive predictive value (PPV) for malignancy in the pancreas (Table 4). The 1 patient without primary pancreatic malignancy had a known hepatic cholangiocarcinoma with regional metastatic disease. Within the atypical LN cohort (23 cases), the pancreas lesion EUS-guided FNA correlated with the following diagnostic categories: 13 cases were malignant (57%); 6 cases were atypical/suspicious (26%); and 4 cases were benign (17%). Including surgical follow-up and clinical outcome (18 with malignancy and 5 without malignancy), an atypical LN EUSguided FNA biopsy had a 78% PPV for malignancy in the pancreas (Table 5). Cases with an “atypical” designation were generally instances where the EUS-guided FNA LN biopsy did not provide definitive diagnostic material; and this was typically related to a scant to minimal epithelial population with some variable cellular epithelial atypia. The qualitative and quantitative changes did not permit a definitive diagnosis of malignancy, and therefore these were classified as atypical.
Discussion EUS-guided FNA biopsy of the pancreas is the standard method for evaluating pancreatic lesions where there is a concern for neoplasia. When the endoscopist examines a patient, the ultrasonographic review can disclose a variety of regional imaging findings, including abnormal or enlarged LNs. In various clinical circumstances, a EUS-guided FNA biopsy of the LN can provide diagnostic or additional information that is helpful in patient management. Two of the more common clinical questions are the ability of the regional lymph node to provide a definitive diagnosis of
malignancy where the pancreatic lesion is negative to suspicious, and definitive documentation of a combined pancreatic and regional lymph node malignancy, which has the ability to impact the option for resection or further interventional procedures. The necessity for regional LN biopsy rests with the performing endoscopist and can be impacted by a variety of factors including the image appearance such as local or diffuse lymphadenopathy, imaging features of the pancreas lesion, preliminary FNA biopsy results for the pancreas lesion, and contributing clinical findings such as other concomitant malignancies.3,4 This study exclusively focused on patients with concurrent EUS-guided FNA biopsy of the pancreas and regional lymph nodes and the predictive value of the LN result in relation to the pancreas biopsy findings. No previous studies have focused exclusively on this correlation with a large series of pancreas and intra-abdominal lymph nodes sampled by EUS-guided FNA.5 Sampling of intraabdominal LN by EUS-guided FNA can provide a diagnosis but is not always diagnostic.6 For the 252 cases in this series, 62% were men with a median age of 60.5 years and the majority of the pancreas lesions present in the head (57%). This is in line with the typical pancreatic EUSguided FNA biopsy patient profile. Most EUS-guided FNA LN biopsies were negative (182 of 252) (72%). When examining the corresponding pancreatic EUS-guided FNA biopsy results for these patients, almost one half were also negative (84 of 182) (47%). About one third were definitively diagnostic for malignancy (30%) and the remaining 23% fell within the atypical/suspicious category. If the positive category and the atypical/ suspicious category were grouped together, they compared about evenly with the negative category. Therefore, a patient with a negative EUS-guided FNA intra-abdominal LN biopsy has about an equal chance for a negative or abnormal Table 3 Negative EUS FNA biopsy intra-abdominal LN category with corresponding pancreas result. Pancreas category
Total
Percentage
Negative Malignant Atypical/suspicious Total
84 55 43 182
46 30 24 100
Abbreviations as in Table 2.
4
B.T. Collins et al. Table 4 Comparison of pancreatic diagnostic categories among patients with positive LN by EUS FNA. Pancreas category
Total
Percentage
Malignant Suspicious Atypical Total PPV
42 3 2 47 98%
89.3 6.4 4.3 100
Abbreviations: PPV, positive predictive value; other abbreviations as in Table 2.
EUS-guided FNA pancreas biopsy result. And in almost one half of the cases, a negative EUS-guided FNA LN biopsy will support the negative categorization provided by the concurrent pancreatic biopsy. Therefore, when faced with a negative pancreatic EUS-guided FNA diagnosis, there is a very small chance of the LN being abnormal or providing a diagnosis. Overall, definitive positive EUS-guided FNA LN biopsies accounted for 19% of the total cases. And within this group of 47 cases, the majority of pancreas biopsies were definitively diagnostic for malignancy (89%) with the remaining cases in the atypical/suspicious category (4%/6%). There were no benign or negative pancreas biopsy results where the EUS-guided FNA LN was positive. When including the subsequent surgical findings and clinical follow-up for the group, there was a PPV of 98% for pancreatic malignancy in cases with a positive EUS-guided FNA biopsy of intraabdominal LN. The 1 positive LN EUS-guided FNA without pancreatic malignancy (by EUS-guided FNA biopsy, surgical confirmation, or subsequent clinical follow-up) involved a patient with known cholangiocarcinoma. A cyst in the uncinate process was evaluated by EUS-guided FNA biopsy and categorized as atypical with features of pancreatitis. The pancreatic cystic lesion was clinically suspicious for intraductal papillary mucinous tumor. The corresponding celiac LN was diagnosed as poorly differentiated carcinoma, which represented metastatic cholangiocarcinoma. Positive LNs with metastatic carcinoma certainly can represent spread from nonpancreas sources, and this consideration should be kept in mind when examining and reporting such cases. Additionally, metastatic processes that do not originate in the pancreas can provide a clinical picture of a pancreas mass and lymphadenopathy. An awareness of any Table 5
Atypical LN category with pancreatic EUS FNA.
Pancreas category
Total
Percentage
Malignant Atypical/suspicious Benign Total PPV
13 6 4 23 78%
57 26 17 100
Abbreviations as in Tables 2 and 4.
previous malignancy diagnoses and special attention to the morphologic and immunohistochemistry differentiation of any malignancy can be important. Within the EUS-guided FNA LN biopsy group, there were 23 cases (9%) in the atypical category. This cohort is most commonly a consequence of insufficiency of necessary quantitative and qualitative features for definitive benign or malignant categorization. As experienced pathologists are aware, sometimes there is minimal sampling of a LN or if there is partial involvement of a LN by a malignant process, there can be too few groups of concern to permit definitive categorization. Mucosal epithelial sampling also often occurs when intra-abdominal LNs are sampled by EUS-guided FNA and these groups, though usually obviously benign or small intestinal in origin, can raise questions about their precise origin. Nevertheless, within this atypical LN category, the majority of pancreas biopsies were malignant (57%) with atypical/suspicious accounting for about a quarter of cases (26%). Only 4 cases (17%) had a corresponding benign pancreas biopsy. When including the subsequent surgical findings and clinical follow-up, there was a PPV of 78% for pancreatic malignancy in cases with an atypical category EUS-guided FNA biopsy of intraabdominal LNs. The order of biopsy sites where both pancreas and LN are sampled can vary, and ROSE is provided, this has the potential to impact the disposition of the procedure. In some circumstances, it can be easier to initially biopsy the intraabdominal LN before the pancreatic lesion. If ROSE is provided, preliminary interpretative observations can be provided to the performing clinician about the LN FNA biopsy that can then help further to guide the procedure. If the LN biopsy provides a positive category, there might not be a need to perform a biopsy on the pancreas mass because a definitive malignant EUS-guided FNA LN biopsy has the ability to provide both a diagnosis and staging information for the patient. Similarly, when ROSE is provided and the pancreatic lesion is not diagnostic for malignancy, the interventional endoscopist can choose to sample regional, abnormal LNs in an attempt to obtain a definitive diagnosis with the knowledge that there is some uncertainty about the final pancreatic diagnostic category.
Conclusions In this study, most EUS-guided FNA LN biopsies were negative and supported a concurrent negative pancreatic biopsy in about one half the cases (47%). For the 20% of cases with a positive diagnostic EUS-guided FNA LN biopsy, they all had corresponding abnormal EUS-guided FNA pancreas biopsies and 89% were definitively categorized as malignant. There was a PPV of 78% for pancreatic malignancy in cases with an atypical category EUS-guided FNA biopsy of intra-abdominal LNs. For cases with a positive category EUS-guided FNA biopsy of intra-
Predictive Value of LN EUS FNA abdominal LN, there was a PPV of 98% for pancreatic malignancy. Therefore, a positive diagnostic category for an intra-abdominal LN can provide strong predictive evidence of a corresponding malignancy of the pancreas.
Funding sources
5
2.
3.
This work had no specific funding.
Conflict of interest disclosures
4.
The authors have no financial disclosures to report.
5.
References 6. 1. Raut CP, Grau AM, Staerkel GA, et al. Diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration in patients with
presumed pancreatic cancer. J Gastrointest Surg. 2003;7:118e126. discussion 127-128. Korenblit J, Anantharaman A, Loren DE, Kowalski TE, Siddiqui AA. The role of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) for the diagnosis of intra-abdominal lymphadenopathy of unknown origin. J Intervent Gastroenterol. 2012;2: 172e176. Kumon RE, Pollack MJ, Faulx AL, et al. Characterization of pancreatic cancer and intra-abdominal lymph node malignancy using spectrum analysis of endoscopic ultrasound imaging. Conf Proc IEEE Eng Med Biol Soc. 2009;2009:1949e1952. Kumon RE, Pollack MJ, Faulx AL, et al. In vivo characterization of pancreatic and lymph node tissue by using EUS spectrum analysis: a validation study. Gastrointest Endosc. 2010;71:53e63. Peng HQ, Greenwald BD, Tavora FR, et al. Evaluation of performance of EUS-FNA in preoperative lymph node staging of cancers of esophagus, lung, and pancreas. Diagn Cytopathol. 2008;36: 290e296. Bluen BE, Lachter J, Khamaysi I, et al. Accuracy and quality assessment of EUS-FNA: a single-center large cohort of biopsies. Diagn Ther Endosc. 2012;2012:139563.