Analysis of cribriform morphology in prostatic neoplasia using antibody to high-molecular-weight cytokeratins

Mahul Amin, D. S. Schultz, R. J. Zarbo

Research output: Contribution to journalArticle

61 Citations (Scopus)

Abstract

Histologic review of 48 radical prostatectomy specimens containing both prostatic adenocarcinoma (PC) and high-grade prostatic intraepithelial neoplasia (PIN) resulted in 23 cases containing neoplastic cribriform glands (CGs) at the periphery or within PC fields. The histologic characteristics of CG PIN mimic those of CG PC in that a discernible basal cell layer defines CG PIN, while CG PC lacks a basal layer. To detect the presence of basal cells in CGs, step sections were immunostained with monoclonal antibody 34βE12 to high-molecular-weight cytokeratins (HMCK) found in basal cells, but not in PC cells. Optimal staining of formalin-fixed sections required pepsin predigestion followed by 14-hour monoclonal antibody incubation at 4°C. Of 436 CG foci identified, 239 (55%) were outlined by a circumferential HMCK- positive layer (identifying PIN); 182 (42%) totally lacked this layer (identifying PC), with appropriate internal controls; and 15 (3%) stained indeterminately. In an attempt to distinguish CG PIN from PC by routine histologic examination alone, CGs with open, round spaces were classified as classic (156 foci); nonclassic CGs (26% foci) featured irregular oblong to slitlike spaces. Cribriform gland PIN defined by HMCK outlining was more often nonclassic (193 CG foci) in histologic pattern, and CG PC was usually 'classic' (110 CG foci) (χ2 = 75; P < .001). We conclude that (1) more than half (55%) of the CGs in the PC tumors studied contain a basal cell layer fulfilling the definition of PIN; (2) focal and isolated HMCK positivity is found amid PC, and thus the overall pattern of staining together with morphological features is critical to correctly exclude carcinoma; (3) grading of PC on the basis of the presence of CGs may lead to erroneous results if PIN foci are misinterpreted as PC; and (4) since CG PIN is usually found in intimate anatomic association with PC, HMCK staining to detect basal cells in isolated CGs encountered in biopsy specimens may be a useful diagnostic discriminant.

Original languageEnglish (US)
Pages (from-to)260-264
Number of pages5
JournalArchives of Pathology and Laboratory Medicine
Volume118
Issue number3
StatePublished - 1994
Externally publishedYes

Fingerprint

Keratins
Prostatic Intraepithelial Neoplasia
Adenocarcinoma
Molecular Weight
Antibodies
Neoplasms
Staining and Labeling
Monoclonal Antibodies
Pepsin A
Prostatectomy
Formaldehyde

All Science Journal Classification (ASJC) codes

  • Pathology and Forensic Medicine
  • Medical Laboratory Technology

Cite this

Analysis of cribriform morphology in prostatic neoplasia using antibody to high-molecular-weight cytokeratins. / Amin, Mahul; Schultz, D. S.; Zarbo, R. J.

In: Archives of Pathology and Laboratory Medicine, Vol. 118, No. 3, 1994, p. 260-264.

Research output: Contribution to journalArticle

@article{1afdd0bee43041e0af59d7488948911a,
title = "Analysis of cribriform morphology in prostatic neoplasia using antibody to high-molecular-weight cytokeratins",
abstract = "Histologic review of 48 radical prostatectomy specimens containing both prostatic adenocarcinoma (PC) and high-grade prostatic intraepithelial neoplasia (PIN) resulted in 23 cases containing neoplastic cribriform glands (CGs) at the periphery or within PC fields. The histologic characteristics of CG PIN mimic those of CG PC in that a discernible basal cell layer defines CG PIN, while CG PC lacks a basal layer. To detect the presence of basal cells in CGs, step sections were immunostained with monoclonal antibody 34βE12 to high-molecular-weight cytokeratins (HMCK) found in basal cells, but not in PC cells. Optimal staining of formalin-fixed sections required pepsin predigestion followed by 14-hour monoclonal antibody incubation at 4°C. Of 436 CG foci identified, 239 (55{\%}) were outlined by a circumferential HMCK- positive layer (identifying PIN); 182 (42{\%}) totally lacked this layer (identifying PC), with appropriate internal controls; and 15 (3{\%}) stained indeterminately. In an attempt to distinguish CG PIN from PC by routine histologic examination alone, CGs with open, round spaces were classified as classic (156 foci); nonclassic CGs (26{\%} foci) featured irregular oblong to slitlike spaces. Cribriform gland PIN defined by HMCK outlining was more often nonclassic (193 CG foci) in histologic pattern, and CG PC was usually 'classic' (110 CG foci) (χ2 = 75; P < .001). We conclude that (1) more than half (55{\%}) of the CGs in the PC tumors studied contain a basal cell layer fulfilling the definition of PIN; (2) focal and isolated HMCK positivity is found amid PC, and thus the overall pattern of staining together with morphological features is critical to correctly exclude carcinoma; (3) grading of PC on the basis of the presence of CGs may lead to erroneous results if PIN foci are misinterpreted as PC; and (4) since CG PIN is usually found in intimate anatomic association with PC, HMCK staining to detect basal cells in isolated CGs encountered in biopsy specimens may be a useful diagnostic discriminant.",
author = "Mahul Amin and Schultz, {D. S.} and Zarbo, {R. J.}",
year = "1994",
language = "English (US)",
volume = "118",
pages = "260--264",
journal = "Archives of Pathology and Laboratory Medicine",
issn = "0003-9985",
publisher = "College of American Pathologists",
number = "3",

}

TY - JOUR

T1 - Analysis of cribriform morphology in prostatic neoplasia using antibody to high-molecular-weight cytokeratins

AU - Amin, Mahul

AU - Schultz, D. S.

AU - Zarbo, R. J.

PY - 1994

Y1 - 1994

N2 - Histologic review of 48 radical prostatectomy specimens containing both prostatic adenocarcinoma (PC) and high-grade prostatic intraepithelial neoplasia (PIN) resulted in 23 cases containing neoplastic cribriform glands (CGs) at the periphery or within PC fields. The histologic characteristics of CG PIN mimic those of CG PC in that a discernible basal cell layer defines CG PIN, while CG PC lacks a basal layer. To detect the presence of basal cells in CGs, step sections were immunostained with monoclonal antibody 34βE12 to high-molecular-weight cytokeratins (HMCK) found in basal cells, but not in PC cells. Optimal staining of formalin-fixed sections required pepsin predigestion followed by 14-hour monoclonal antibody incubation at 4°C. Of 436 CG foci identified, 239 (55%) were outlined by a circumferential HMCK- positive layer (identifying PIN); 182 (42%) totally lacked this layer (identifying PC), with appropriate internal controls; and 15 (3%) stained indeterminately. In an attempt to distinguish CG PIN from PC by routine histologic examination alone, CGs with open, round spaces were classified as classic (156 foci); nonclassic CGs (26% foci) featured irregular oblong to slitlike spaces. Cribriform gland PIN defined by HMCK outlining was more often nonclassic (193 CG foci) in histologic pattern, and CG PC was usually 'classic' (110 CG foci) (χ2 = 75; P < .001). We conclude that (1) more than half (55%) of the CGs in the PC tumors studied contain a basal cell layer fulfilling the definition of PIN; (2) focal and isolated HMCK positivity is found amid PC, and thus the overall pattern of staining together with morphological features is critical to correctly exclude carcinoma; (3) grading of PC on the basis of the presence of CGs may lead to erroneous results if PIN foci are misinterpreted as PC; and (4) since CG PIN is usually found in intimate anatomic association with PC, HMCK staining to detect basal cells in isolated CGs encountered in biopsy specimens may be a useful diagnostic discriminant.

AB - Histologic review of 48 radical prostatectomy specimens containing both prostatic adenocarcinoma (PC) and high-grade prostatic intraepithelial neoplasia (PIN) resulted in 23 cases containing neoplastic cribriform glands (CGs) at the periphery or within PC fields. The histologic characteristics of CG PIN mimic those of CG PC in that a discernible basal cell layer defines CG PIN, while CG PC lacks a basal layer. To detect the presence of basal cells in CGs, step sections were immunostained with monoclonal antibody 34βE12 to high-molecular-weight cytokeratins (HMCK) found in basal cells, but not in PC cells. Optimal staining of formalin-fixed sections required pepsin predigestion followed by 14-hour monoclonal antibody incubation at 4°C. Of 436 CG foci identified, 239 (55%) were outlined by a circumferential HMCK- positive layer (identifying PIN); 182 (42%) totally lacked this layer (identifying PC), with appropriate internal controls; and 15 (3%) stained indeterminately. In an attempt to distinguish CG PIN from PC by routine histologic examination alone, CGs with open, round spaces were classified as classic (156 foci); nonclassic CGs (26% foci) featured irregular oblong to slitlike spaces. Cribriform gland PIN defined by HMCK outlining was more often nonclassic (193 CG foci) in histologic pattern, and CG PC was usually 'classic' (110 CG foci) (χ2 = 75; P < .001). We conclude that (1) more than half (55%) of the CGs in the PC tumors studied contain a basal cell layer fulfilling the definition of PIN; (2) focal and isolated HMCK positivity is found amid PC, and thus the overall pattern of staining together with morphological features is critical to correctly exclude carcinoma; (3) grading of PC on the basis of the presence of CGs may lead to erroneous results if PIN foci are misinterpreted as PC; and (4) since CG PIN is usually found in intimate anatomic association with PC, HMCK staining to detect basal cells in isolated CGs encountered in biopsy specimens may be a useful diagnostic discriminant.

UR - http://www.scopus.com/inward/record.url?scp=0028344274&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028344274&partnerID=8YFLogxK

M3 - Article

VL - 118

SP - 260

EP - 264

JO - Archives of Pathology and Laboratory Medicine

JF - Archives of Pathology and Laboratory Medicine

SN - 0003-9985

IS - 3

ER -