Kathleen Louden

September 26, 2013

CHICAGO — A novel risk-stratification system for endoscopic ultrasonography-guided fine-needle aspiration cytology results can help determine whether a pancreatic lesion is malignant, according to a new study.

"It's our hope that when a pathologist gets a fine-needle aspiration result for a suspicious pancreatic mass and is not sure if it is definitively malignant, he or she can use this set of criteria as a guide," said study coauthor Brandon Huffman, a third-year medical student at the University of Missouri School of Medicine in Columbia.

Although the cytologic characteristics of pancreatic ductal adenocarcinoma have been described, Huffman said he believes this study is the first to stratify the morphologic features of malignancy in this type of cancer.

He presented the research here at the American Society for Clinical Pathology 2013 Annual Meeting.

The proposed system evaluates 3 morphologic criteria: anisonucleosis, defined as a size variation in the cell nuclei of at least 4 times; single atypical epithelial cells; and mucinous metaplasia.

When endoscopic ultrasonography fine-needle aspiration cytology findings meet none of the criteria, the risk for pancreatic ductal adenocarcinoma is low, when they meet 1 of the criteria, the risk is moderate (odds ratio [OR], 8.8), and when they meet 2 or 3 of the criteria, the risk is high (OR, 51.9).

In their study, Huffman and his colleagues retrospectively reviewed pathology records from the University of Missouri Health System. They found results from more than 450 pancreatic fine-needle aspirations from 350 patients.

After the histologic analysis of 27 of these biopsies from 24 patients, they confirmed the diagnosis of pancreatic ductal adenocarcinoma.

Results from another 27 of these biopsies were considered benign (negative for malignancy) because the patient had a normal surgical biopsy specimen or was alive and well 2 years after the fine-needle aspiration, Huffman reported.

Three pathologists blinded to the cytologic diagnoses independently evaluated the 54 slide preparations for the presence or absence of 16 criteria found in previous studies to increase the risk for pancreatic malignancy.

The researchers identified 3 characteristics with the greatest discriminatory power.

Table 1. Morphologic Characteristics Associated With Pancreatic Malignancy

Characteristic Odds Ratio 95% Confidence Interval P Value
Anisonucleosis 16.6 7.1–39.2 <.001
Single atypical epithelial cells 14.9 5.8–38.0 <.001
Mucinous metaplasia 4.5 2.2–9.2 <.001


Using a mixed-effects regression analysis of the 162 observations made by the 3 pathologists, the researchers stratified the cases into low-, moderate-, and high-risk categories, assigning 1 point for each morphologic feature.

For the 3 morphologic characteristics used in the stratification, interobserver reliability (κ statistic) ranged from 0.39 to 0.52, Huffman reported.

Other diagnostic features associated with a significantly increased (P < .001) risk for malignancy included nuclear membrane irregularity, nuclear enlargement (more than 2 red blood cells), macronucleoli, necrosis, nuclear crowding, overlapping or 3-dimensionality, hyperchromasia, and chromatin clearing.

Huffman said his team is continuing to study this proposed risk-stratification system in collaboration with the University of Utah and ARUP Laboratories, both in Salt Lake City. He said his hope is that the system will help pathologists in the diagnosis of cases that are not definitively benign or malignant.

A pathologist reviewing fine-needle aspiration results for a pancreatic mass that has a risk score of 1 (moderate) or 2 (high) could interpret the result as atypical or suspicious, Huffman explained. "That might guide the clinician to get another biopsy or monitor the patient more closely," he said.

"When we try to make a malignant diagnosis in cytology, it's always good to use more concrete criteria," said Antoinette Sperelakis, MD, medical director of pathology at Palos Community Hospital in Palos Heights, Illinois, who was asked by Medscape Medical News to comment on the study.

Dr. Sperelakis, who was not involved with the study, said the proposed proposed-risk stratification system needs further study. She said she welcomes the development of uniform guidelines for reporting pancreatic cytopathologic findings, as exist in the Bethesda system for thyroid and cervical cancers.

New Guidelines

The Papanicolaou Society of Cytopathology is currently developing guidelines for pancreatic, biliary cytopathology reporting. Immediate past president of the society, Lester Layfield, MD, reported that they will likely be released in February or March 2014. Dr. Layfield, who is professor and chair of pathology and anatomical sciences at the University of Missouri School of Medicine, was one of the researchers on this risk-stratification study.

Another option when the results of a pancreatic endoscopic ultrasonography-guided fine-needle aspiration are not clear is to perform a second biopsy.

A repeated pancreas fine-needle aspiration biopsy can clarify and provide a definitive diagnosis of malignancy, according to Brian Collins, MD, who presented results from a related study here at the meeting. Dr. Collins is section head of cytopathology at Washington University Medical Center in St. Louis, Missouri.

He and his colleagues retrospectively studied the diagnostic utility of a second pancreatic biopsy in 109 cases.

They found a large increase in the percentage of biopsy results categorized as malignant between the first and the second biopsy report.

Table 2. Diagnostic Yield of Repeat Pancreatic Biopsies (n = 109)

Diagnostic Category Initial Biopsy, % Second Biopsy, %
Benign 50.5 44.0
Atypical 33.9 21.1
Suspicious (for malignancy) 11.0 4.6
Malignant 2.8 30.3*
Unsatisfactory 1.8 0.0

*Includes neuroendocrine tumors (3.7%).

There was no change in the way the biopsy procedures were performed. However, the second biopsy result was evaluated by a different pathologist in 75% of cases, and the result of the first biopsy was known in all second biopsies, Dr. Collins reported.

He explained that there is no pathologist in the biopsy room for on-site evaluation at his center, and agreed with an audience member that rapid on-site evaluation might have reduced the need for a second biopsy.

However, a second biopsy can be of value when the pathologic diagnosis is unclear and a definitive diagnosis is needed, such as before starting chemotherapy or radiation therapy, Dr. Collins told Medscape Medical News.

"Repeating the fine-needle aspiration biopsy could be less expensive than doing an open surgical procedure," he added.

Mr. Huffman, Dr. Sperelakis, Dr. Layfield, and Dr. Collins have disclosed no relevant financial relationships.

American Society for Clinical Pathology (ASCP) 2013 Annual Meeting: Scientific posters 1-7 and 2-5. Presented September 20, 2013.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.