Morphometric analysis of differentiation in human breast carcinoma. Tumor grading

F. E. Sharkey, R. J. Pavlak, A. S. Greiner

Research output: Contribution to journalArticlepeer-review

14 Scopus citations


Subjective grading systems suffer from limited reproducibility and do not assign a sufficiently large proportion of patients to the most predictive grades (1 and 3). In 50 cases of stage 1 or stage 2 mammary ductal carcinoma having at least five-year follow-up, we quantitated the proportion of neoplastic cells showing glandular differentiation (DI) or remaining within ducts (IN). Patients who experienced recurrence had significantly less of both DI and IN, and the point at which 10% of neoplastic cells was composed of DI + IN appeared to be predictive for recurrence; ie, 24 recurrences appeared among the 33 patients who had tumors with DI + IN < 10%, whereas none developed among the 14 patients having tumors with DI + IN > 10%. Only three patients could not be classified with this marker. Tumor stage and grade were independent variables. Reproducibility among three observers averaged 95%. The morphometric evaluation can be performed by a technician under the direction of a pathologist.

Original languageEnglish (US)
Pages (from-to)406-410
Number of pages5
JournalArchives of Pathology and Laboratory Medicine
Issue number8
StatePublished - 1983
Externally publishedYes

ASJC Scopus subject areas

  • Pathology and Forensic Medicine
  • Medical Laboratory Technology


Dive into the research topics of 'Morphometric analysis of differentiation in human breast carcinoma. Tumor grading'. Together they form a unique fingerprint.

Cite this