CYTOLOGIC NUCLEAR GRADE OF MALIGNANT BREAST ASPIRATES AS A PREDICTOR OF HISTOLOGIC GRADE - LIGHT-MICROSCOPY AND IMAGE-ANALYSIS CHARACTERISTICS

Citation
K. Moroz et al., CYTOLOGIC NUCLEAR GRADE OF MALIGNANT BREAST ASPIRATES AS A PREDICTOR OF HISTOLOGIC GRADE - LIGHT-MICROSCOPY AND IMAGE-ANALYSIS CHARACTERISTICS, Acta cytologica, 41(4), 1997, pp. 1107-1111
Citations number
22
Categorie Soggetti
Cell Biology
Journal title
ISSN journal
00015547
Volume
41
Issue
4
Year of publication
1997
Pages
1107 - 1111
Database
ISI
SICI code
0001-5547(1997)41:4<1107:CNGOMB>2.0.ZU;2-D
Abstract
OBJECTIVE: To determine if cytologic nuclear grade characteristics com bined with image analysis assessment of morphometric nuclear parameter s (1) correlate with the modified Scarff-Bloom-Richardson grading syst em and (2) discriminate between low and high nuclear grades of invasiv e ductal carcinoma. STUDY DESIGN: Fifty-four fine needle aspiration bi opsies (FNABs) of breast carcinoma were evaluated for five morphologic nuclear grade characteristics. In addition, four morphometric, standa rdized object measurements were analyzed by an image analysis system. Corresponding biopsies of invasive ductal carcinoma (46 cases) were in dependently evaluated with the Scarff-Bloom-Richardson grading system, modified into low (scores 3-6) and high (scores 7-9) grades. RESULTS: An overall agreement of 82% was reached by three of four cytopatholog ists for each of five morphologic characteristics. There was a strong correlation (r = .8059, P <.0001) between cytologic nuclear grade and modified histologic grade. Only pleomorphism, nucleoli and sum optical density retained their statistical significance in distinguishing low from high grade ductal carcinomas, These three characteristics also h ad the strongest correlation with cytologic nuclear grade. CONCLUSION: Cytologic nuclear grade from aspirates of ductal carcinoma can be a p redictor of the modified histologic grades of Scarff, Bloom and Richar dson. Nuclear morphology reinforced by image morphometry may separate these tumors into low and high nuclear grade categories.