The most award winning
healthcare information source.
TRUSTED FOR FOUR DECADES.
abstracts & commentary
Synopsis: The National Cancer Institute (NCI) has developed a uniform reporting system for fine-needle aspiration (FNA) of the breast. In this series, a retrospective evaluation of archival FNAs was categorized using the new NCI system. Ultrasound-guided FNA, as reported by the new NCI guidelines, is both a sensitive and specific means to diagnose nonpalpable breast cancers.
Sources: Boerner S, et al. Cancer 1999;87:19-24; Masood S. Cancer 1999;87:1-4.
Because of substantial heterogeneity in technique, terminology, and clinical interpretation, the National Cancer Institute (NCI) supported the development of specific diagnostic categories for fine-needle aspiration (FNA).1,2 In this system, FNAs are classified into one of five categories: 1) benign; 2) indeterminate/atypical; 3) suspicious/probably malignant; 4) malignant; or 5) unsatisfactory. In this report from Boerner and colleagues, the NCI system was applied to a large number of archival samples and the interpretations were compared to subsequent biopsies and clinical outcomes.
Ultrasound-guided FNAs (US-FNAs) performed on nonpalpable breast lesions between the years 1988 and 1996 at M.D. Anderson Cancer Center were reexamined using the new NCI system. There were 1855 US-FNAs performed on 1639 patients during this period. The original FNA diagnoses were reclassified into the NCI categories, and the findings were correlated with the tissue specimens available in 851 cases or, with the clinical follow-up in those with FNA findings of benign lesions.
Of the 1885 cases, 56% were benign, 4.6% were atypical, 4.2% suspicious, probably malignant, 27% were malignant, and 8.5% unsatisfactory. Those that were benign included 45% cysts and 55% solid lesions. Of those categorized as benign, 3.7% (n = 13) were found by subsequent surgical specimen or clinical course to be malignant. Of those categorized as atypical, 53% were found to be malignant, whereas 76% of those categorized as suspicious were found to be malignant. Of the malignant FNAs, 1% of the lesions were found to be benign. Based upon the subsequent histologies and clinical follow-up, a sensitivity of 97.1% and a specificity of 99.1% were found for US-FNAs when definitive benign and malignant diagnoses were offered. The false-negative ratio of 3.7% was attributed to sampling error whereas the false-positive rate of 0.68% was thought secondary to interpretive error of proliferating lesions.
Boerner et al conclude that the application of the NCI-supported diagnostic categories to US-FNA of nonpalpable breast lesions is useful in stratifying aspirates with regard to the likelihood of indicating malignancy and thereby provides a rationale for the use of this approach, coupled with clinical exam and imaging studies for patients with possible breast cancer.
FNAs have received a lot of attention and increasing use, but inconsistencies in technique and interpretation have often left clinicians confused and patients misinformed. The new NCI guidelines may be of great value, but there have been limited large series published in which clinical outcomes have been correlated with their application. In this report, a special subset of patients, those with nonpalpable lesions, were reported. This reflects a difficult population and one in which precise and accurate sampling could save more invasive intervention. Indeed, as indicated in this retrospective series, over 50% of the FNAs resulted in a benign diagnosis and, presumably, clinical follow-up alone without biopsy would be sufficient management. Only 3.7% of patients with a nonmalignant diagnosis on FNA went on to develop cancer. In contrast, 26% of the FNAs resulted in an unequivocal malignant diagnosis. These patients would have also benefited since the FNA may well have spared them an open biopsy, and thereby allowed the direct initiation of definitive therapy. It has not been determined whether the 9% of patients within the intermediate categories of atypical or suspicious are actually benefited by that information. This study would suggest that about half of the 4.2% of patients in the atypical category, and about three-fourths of the 4.6% in the suspicious category will be found to have malignancy. Essentially all of these patients will need to proceed to biopsy for definitive diagnosis. However, the number of patients who must go on to biopsy is only about 17% of the total; the 9% with atypical or suspicious cytology and the 8% in whom the aspirate was technically inadequate.
Yet, for patients and clinicians, what is worrisome is the possibility of false negatives (and false positives) for diagnostic procedures such as FNA. The NCI strategy does help in this regard because it differentiates between the unequivocal benign and malignant categories, and about 75% in this series fell into one of those two categories. Even in expert hands, there were 13 patients (3.7%) who were incorrectly reassured that their breast lesions were benign, and an additional three patients were taken to surgery because they were told their lesions were malignant, only later to be informed that this was an error. With a disease such as breast cancer where fears and emotional tensions are inherently at a high pitch, even these small numbers are difficult to accept, and may have medical-legal consequences.
Nevertheless, the NCI guidelines are helpful, they allow better communication and uniformity of treatment between physician groups at different institutions. It is now the challenge for those in this field to further improve the reliability of minimally invasive diagnostic strategies, and to define the optimal management strategies for patients with nonpalpable breast masses who are diagnosed on the basis of FNA.
1. Anonymous. Acta Cytol 1996;40:1120-1126.
2. Anonymous. Breast J 1997;3:149-168.
a. Its value is diminished because close to 50% of interpretations fall into the "atypical" category for which appropriate management is uncertain.
b. In expert hands, a "malignant" interpretation is a reliable finding (close to 99% predictive value).
c. An "atypical" interpretation has a greater likelihood of reflecting a malignant lesion than a "suspicious" interpretation.
d. Even in expert hands, a "benign" interpretation has about a 10% chance of being an error.
e. Regardless of the result of FNA, a biopsy is necessary to confirm the result.