Revisiting the cytologic features of autoimmune pancreatitis: An institutional experience

Document Type

Article

Publication Date

12-27-2022

Publication Title

Cancer Cytopathol

Abstract

BACKGROUND: Autoimmune pancreatitis (AIP) is a known mimicker of pancreatic ductal adenocarcinoma both clinically and radiologically. In this study, the authors present their institutional experience in diagnosing AIP on cytology and correlate results with the histologic findings.

METHODS: A 14-year computerized search for patients who had histologically confirmed AIP with concurrent or prior cytology was performed. Clinical data, cytology findings, and surgical pathology results were reviewed for analysis.

RESULTS: Eighteen patients were identified. The patients showed a male predominance, with a mean age of 59 years. Jaundice, weight loss, and abdominal pain were the most common clinical presentation. Five of 12 patients who were tested for serum immunoglobulin G4 had elevated levels. Cytologic findings of 16 cases that were available for review showed markedly inflamed fibrous stroma (54%) and cytologic atypia (50%). The final cytologic diagnoses were suspicious for adenocarcinoma (n = 1), atypical (n = 8), and benign/negative (n = 9). The corresponding surgical pathology diagnoses were classified as type 1 (n = 10), type 2 (n = 6), and AIP, not otherwise specified (n = 2). All type 2 AIP cases had at least atypical cytologic diagnoses, with one called suspicious for adenocarcinoma and another called adenocarcinoma at the time of rapid on-site evaluation. In contrast, eight of 10 type 1 AIP cases were negative/benign, and two of 10 were atypical. In these two atypical cases, the possibility of AIP was raised because of the presence of inflamed stroma.

CONCLUSION: AIP is a pitfall in cytology because moderate-to-marked atypia can be present, especially in type 2 AIP. Because atypia can be severe, the presence of cellular fibrous stroma with lymphocytic stromal infiltrates and the integration of serum immunoglobulin G4 levels could be helpful in avoiding diagnostic overcall in AIP.

PubMed ID

36574153

ePublication

ePub ahead of print

Share

COinS