ANALYSIS OF THE CAUSES OF FALSE-NEGATIVE CYTOLOGY REPORTS ON BREAST-CANCER FINE-NEEDLE ASPIRATES

Forty-six (2.9%) false negative reports were recorded among adequate f ine needle aspirates from 1609 consecutive histologically proven carci nomas, observed from 1991 to 1993. False negatives were more frequent among younger women (<40 years=7.1, 40-49 years = 4.0; 50-59 years = 3 .6, 60-69 years = 1.7, > 69 years = 1.5%), lobular invasive subtypes ( 5.4%), and smaller tumours (pT1a=4.2; pT1b=7.6; pT1c=1.9, pT2=2.1; pT3 -4=1.5%). The latter finding is probably ascribable to better differen tiation and less precise sampling of small non-palpable tumours. No si gnificant association was found between the false negative rate and th e sampler's or reader's skill and experience. The former finding may b e ascribed to the wide use of sonography guided aspiration, even for p alpable masses, and the latter to the fact that readers were highly ex perienced and undergo periodic quality control of individual performan ce. False negatives were reclassified at reviews as true false negativ es, reading errors or inadequates in 27, 11, and eight cases, respecti vely. The observed findings suggest that in most cases cytological fau lts were due to the absence of cytological atypia in cells sampled fro m well differentiated tumours, rather than to misinterpretation or sam pling from adjacent normal tissues.