Fine-needle aspiration cytology is preferred for axillary lymph node metastasis with low costs and minimal risks. To improve diagnostic performance by incorporating clinical-radiological-pathological parameters, a large cohort pre-operative aspirates in were reviewed for parameters affecting adequacy rate and accuracy. Axillary nodal aspirates from three institutions with histologic correlation were retrieved. Case notes were reviewed for parameters pertaining to the primary tumor, nodal status, histologic and cytologic diagnoses. Totally 1361 specimens were included. The risk of malignancy for C1-C5 categories were 53.39%, 27.45%, 70.97%, 83.33% and 88.00%, increasing to 75.86%, 94.59% and 99.28% for C3/C4/C5 categories excluding cases with neoadjuvant therapy. Node size (p < 0.001) and histologic grade (p = 0.003) of primary tumor positively correlated with specimen adequacy. Presence of in situ component trended towards inadequacy (p = 0.069). Lymph node size remained a strong predictor of concordant cytologic diagnosis (p < 0.001). A higher percentage of involved node (p = 0.006) and HER2 overexpressed breast cancers (p = 0.027) increased concordance. Cases with ≥ 4 (up to ≥ 10) positive nodes were more likely to be concordant (p = 0.009- < 0.001), with improvements of 8.27%-12.37%. For size, cut-offs of ≥ 5 and ≥ 10mm were significant (p = 0.006- < 0.001). It is critical that clinical-radiological-pathological findings be interpreted together with cytology. Aspirates from smaller nodes are more likely to be non-informative, irrespective of the total number of suspicious nodes, or a high-grade primary. In axillae with less than 4 suspicious nodes and/or a target node of less than 5-10mm, the diagnostic accuracy of aspiration cytology decreases and should be interpreted cautiously.
Read full abstract