Use of thin‐layer preparations for gynecologic smears with emphasis on the cytomorphology of high‐grade intraepithelial lesions and carcinomas

1996 
Thin-layer (TL) technology can improve the detection rate for squamous lesions of the uterine cervix. Studies to date have under-represented high grade lesions and malignancies. The present study utilized a patient population at high risk for such lesions in order to analyze the performance of TL procedures in this group, and in addition, to assess the similarities and differences in morphologic appearances of specimens prepared by the two methods. Conventional (CS) and thin-layer smears (TLS) were made in parallel from the same specimen. Each slide was examined in a blinded fashion. Diagnoses were compared and morphologic observations made. Two hundred fifty-nine cases were included, of which 32 (12%) were high grade dysplasias (11) or carcinomas (21). Thirty five (14%) were atypical or low grade dysplasias. There was exact correlation between Bethesda classification in 231 cases (89%). Of the 21 carcinomas identified, 19 (91%) were present on each preparation. Two cases of endometrial adenocarcinoma were missed on unsatisfactory or negative TLS. One case of squamous cell carcinoma was called high grade squamous intraepithelial lesion (HGSIL) on TLS while the CS was unsatisfactory. Three cases called atypical glandular cells (AGCUS) on TLS, and negative on CS, showed HGSIL (1) or no lesion (2) on follow-up. Morphologic features of low grade lesions were virtually identical on both preparations. Distinct features were noted on TLS in the high grade lesions. These included smaller appearing nuclear areas, less distinct nuclear chromatin, thicker three-dimensional groupings, and more isolated cells. Such findings were most pronounced in the glandular lesions. With training and experience, these features were easily identified in TL preparations, further documenting the utility of this procedure for use in routine practice. Diagn Cytopathol 1996;14:201–211. © 1996 Wiley-Liss, Inc.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    7
    References
    47
    Citations
    NaN
    KQI
    []