The purpose of this study is to determine whether there is a difference of diagnostic accuracy between cover slides and base slides from EUS-FNA cytopathologic sample in patients who were previously suspected of pancreatic cancer in radiologic test.

1. background

   1. Pancreatic cancer is the challenging malignancy in diagnosis and treatment.

   2. Recently, EUS-FNA has played a important role in diagnosing pancreatic cancer.

   3. Samples from EUS-FNA are sent to following three types of preparations.

      • cytologic smear (by pull apart method)
      • cell block
      • histologic biopsy

   4. In cytologic smear method, endoscopists tend to send as many as slides, while pathologists want minimal slides if they warrant diagnostic accuracy.

   5. Therefore we are going to find a point of contact between high diagnostic yield and minimal overloading of pathologist.

2. Method

   1. We are going to enrolled the patients who were suspected of pancreatic cancer in previous radiologic test.

   2. One identical and skilled endoscopist performs EUS-FNA with multiple punctures.

   3. We get total 8 pairs of slides (= 8 covers + 8 bases).

   4. All the cover slides are collected to the bottle-A, all the base slide to the bottle-B.

   5. One identical and skilled pathologist inspects these slide in blind method. (She cannot show the label of bottle so that she does not know what bottle is A or B.)

   6. She record the positive or negative in each bottle. (If there is only one slide is positive in malignant cell, the bottle is positive.)

   7. In these manner, we collect the all subjects' cytopathologic results as a pair of outcome. All the possible combinations are following four examples:

      • (cover/base) = (positive/positive) = (+/+)
      • (cover/base) = (positive/negative) = (+/-)
      • (cover/base) = (negative/positive) = (-/+)
      • (cover/base) = (negative/negative) = (-/-)

   8. Then we are going to take a two-sided non-inferiority test of proportion using McNemar's test.

Inclusion Criteria:

• patients who were suspected of pancreatic cancer in previous radiologic test such as computed tomography, magnetic resonance imaging or Positron Emission Tomography.

Exclusion Criteria:

• failed smear due to inexperienced assistant
• loss of slide
• patients who were diagnosed with other lesion far from pancreatic cancer in prospective confirmative tests.
• patients who refuse to participate in this study
• report of other pathologist who are not a previously-assigned pathologist

The patients who visited to tertiary medical center(Seoul National University Bundang Hospital) because of suspicious pancreatic cancer in radiologic study