[392] Utility of ProExC and IMP3 Immunocytochemical Staining of Atypical Glandular Cells of Undetermined Significance (AGUS) in Liquid-Based Cervical Cytology

Ricardo R Lastra, Joyce J Ou, Mary E Reilly, W Dwayne Lawrence, John S Brooks, Julieta E Barroeta. Pennsylvania Hospital of the University of Pennsylvania Health System, Philadelphia, PA; Women & Infants Hospital of Rhode Island, Providence, RI

Background: Studies have reported positive ProExC and IMP3 staining in neoplastic glandular lesions of the uterine cervix and corpus. Thirty percent of cases diagnosed as AGUS in liquid-based cervical cytology specimens have significant underlying pathology, while the remaining are related to reactive, reparative and metaplastic conditions. We address the utility of these markers in the evaluation of AGUS cases in liquid-based cervical cytology.
Design: Unstained Thin-Prep slides were prospectively collected on all cases diagnosed as AGUS from June 2007 to November 2009. Based on the follow-up biopsy findings in thirty-four cases (n=34), the study included cases of adenocarcinoma in situ (AIS, n=2), adenocarcinoma (AC, n=3), squamous cell carcinoma (SCC, n=1), LSIL (n=3), HSIL (n=4) and benign (n=21). ProExC (TriPath Imaging, prediluted) and IMP3 (Dako, dilution 1:100) immunocytochemical (ICC) stains were performed. The results were correlated with subsequent biopsy findings. Positivity was assessed by strong diffuse nuclear (ProExC) and granular cytoplasmic (IMP3) staining.
Results: The presence of AGUS cells on the ICC stained slides was confirmed in all cases. Of all glandular lesions, IMP3 was positive in 4/5 cases and negative in 27/29 non-glandular lesions/benign cases (sensitivity: 80%, specificity: 93%, negative predictive value - NPV: 96%); ProExC was positive in 3/5 glandular lesions and negative in 24/29 non-glandular lesions/benign cases (sensitivity: 60%, specificity: 82%, NPV: 96%). When used as a panel (ProExC + IMP3), at least one stain was positive in 5/5 glandular lesion cases and they were both negative in 24/29 non-glandular lesions/benign cases (sensitivity: 100%, specificity: 83%, NPV: 100%). Of the 5 cases of non-glandular lesions/benign cases in which at least one was positive, the follow-up diagnosis was SCC (n=1), HSIL (n=2), LSIL (n=1) and benign (n=1).
Conclusions: ICC staining for ProExC and IMP3, particularly when used as a panel, might serve as a predictor of glandular lesions on subsequent biopsies, as our study demonstrated positivity of at least one of the two markers in all AIS and AC lesions. Positivity also predicted the presence of squamous lesions in subsequent biopsies. Our findings suggest a role for IMP3 and ProExC ICC staining as an aid in the characterization and potential management of AGUS cases.
Category: Cytopathology

Tuesday, March 20, 2012 9:30 AM

Poster Session III # 42, Tuesday Morning


Close Window