Using Computerized Cytomorphometry to Distinguish between Benign and Malignant Cases in Thyroid Fine‐Needle Aspiration Cytology

Abstract

Computerized cytomorphometry is useful in distinguishing between benign and malignant lesions in thyroid FNAC. The measurement of cytologic nuclear parameters in cases suggestive of AUS may be useful for the probable classification of cases as benign or malignant. Although further studies are needed, in nuclear morphometric assessment of thyroid FNAC, Giemsa staining may be more useful and valuable than the Pap stain because of its association with various cytologic nuclear parameters. Diagn. Cytopathol. 2016;44:902-911. © 2016 Wiley Periodicals, Inc.

Digital image‐assisted quantitative nuclear analysis improves diagnostic accuracy of thyroid fine‐needle aspiration cytology

“…However, nearly all studies have used nuclear area as one of their indicators, and several have used some sort of measure of roundness/ovalness. In general, virtually all studies have found statistically significant differences in nuclear area (and many other nuclear features evaluated) between benign and PTC nuclei (Table ) . Many have reported sufficient overlap to limit the utility in clinical practice, but others have found that nuclear area, alone or in combination with other morphometric features, can achieve accuracy of 70% or greater .…”

“…In addition, only some studies used air‐dried smears; air drying is likely to accentuate the variation in nuclear size/shape, and as such may be a better preparation for this purpose. However, 2 studies that evaluated both air‐dried Giemsa stains and alcohol‐fixed Papanicolaou‐stained smears found relatively similar effectiveness for both . Finally, many studies grouped cases of PTC, follicular carcinoma, and in some cases other carcinomas into a heterogeneous “malignant” group, which potentially enhances overlap between categories in a way that does not necessarily represent the pathologist’s differential diagnosis in individual cases.…”

“…Our study is relatively unique in evaluating the performance in morphometry in cases with indeterminate (Bethesda categories AUS/FLUS and SFM) cytology findings, the cases in which the pathologist is likely to need help. The vast majority of previous studies did not address the cytologic diagnosis of the slides chosen for the study, or separately evaluate performance among cases with indeterminate cytology; only 2 recent studies specifically included cases with indeterminate Bethesda categorization on cytology . Celik et al found that area, perimeter, and diameter all significantly distinguished benign from malignant, even in indeterminate categories, but could not establish a specific clinically useful cutoff.…”

“…In general, virtually all studies have found statistically significant differences in nuclear area (and many other nuclear features evaluated) between benign and PTC nuclei (Table 6). [29][30][31][32][33][34][35][36][37][38][39][40][41][42] Many have reported sufficient overlap to limit the utility in clinical practice, but others have found that nuclear area, alone or in combination with other morphometric features, can achieve accuracy of 70% or greater. 31,32,34,35,41,42 The current study found relatively good distinction by nuclear area, albeit with limited overlap in cases with indefinite qualitative cytologic interpretations.…”

“…However, 2 studies that evaluated both air-dried Giemsa stains and alcohol-fixed Papanicolaou-stained smears found relatively similar effectiveness for both. 30,39 Finally, many studies grouped cases of PTC, follicular carcinoma, and in some cases other carcinomas into a heterogeneous "malignant" group, which potentially enhances overlap between categories in a way that does not necessarily represent the pathologist's differential diagnosis in individual cases. Similarly, none of the studies separated NIFTP from PTC (although a few did separate FVPTC from classical PTC 35,37 )-which, given the results of the present study, likely led to more overlap across categories in previous studies.…”