In the present study, we demonstrate that US-guided FNAC of axillary LN in breast cancer patients can reliably predict the presence of metastases and therefore refer the patients to the appropriate surgical treatment, avoiding the SLNB. In fact, in our series, US-guided FNAC presurgically screened 18% of invasive breast carcinoma operated in the same year and scored 49 out of 88 (55%) of the metastatic LN.
Since the early 1980s, US examination has been proposed by many authors as one of the most successful procedures for evaluating the status of the axilla in breast cancer patients before surgery (Bruneton et al, 1986
; Pamilo et al, 1989
; Mustonen et al, 1990
; Yang et al, 1996
; Strauss et al, 1998
;). Sonography can easily explore the different nodal chains and, when LN are found, specific signs may be searched to evaluate the presence of metastases (Feu et al, 1997
; Rizzatto, 2001
). We here reported a 16% false negative rate by US alone, since 28 out of 172 of US-nonsuspect LN were positive at histology. However, a great improvement of the diagnostic accuracy emerged when US examination was combined with FNAC of LN suspect to be metastatic, as demonstrated by the few available reports on the subject (Bonnema et al, 1997
; Verbanck et al, 1997
; de Kanter et al, 1999
). In fact, in our study and in previous ones that used a similar approach, the cytological diagnosis of metastases reached 100% specificity. On the other hand, Bonnema et al (1997)
reported 80% sensitivity and a 76% negative predictive value of the cytological method. In the present work, we observed an increase in both sensitivity (89%) and negative predictive value (84%), which has to be related to the additional use of ICC. The use of a cocktail of antiepithelial membrane antigen and antipancytokeratin antibodies for identifying scattered epithelial cancer cells has been proposed recently in bone marrow smears (Gebauer et al, 2001
). The possibility of using the same smears obtained for routine H&E staining for additional ICC analyses with the cocktail of antibodies recognising different antigens is an interesting approach, which reduces further sampling of the LN, and limits potential diagnostic pitfalls. In fact, cytokeratin-positive interstitial reticulum cells present in normal or reactive LN may complicate interpretation of the results (Linden and Zarbo, 2001
) and, on the other hand, the epithelial membrane antigen may enhance both sensitivity and specificity of the procedure, particularly for metastases of invasive lobular breast cancer. In the present series, in seven out of nine metastatic cases diagnosed by ICC, the primary tumour was an invasive lobular carcinoma with scattered neoplastic cells dispersed among lymphocytes and clearly outlined only by ICC.
In a recent multicentre study, de Kanter et al (1999)
demonstrated that, using US combined with FNAC in patients without palpable axillary nodes, SLNB could be avoided in 17% of cases since cytology had already diagnosed axillary metastases. Similarly, in our experience, 18% of metastatic invasive carcinomas were already diagnosed using US-guided FNAC, which presurgically scored 49 out of 88 (55%) of the metastatic LN.
In addition, the method could be particularly valuable in cases of extensive metastatic involvement, which is the cause of false negative SLNB procedures (Dequanter et al, 2001
). In fact, when a LN is completely replaced by the tumour, there is poor uptake of radioactivity. In our series, the whole group of pN2 LN were positive by cytological examination.
The requirement of neoadjuvant chemotherapy appears to be an additional indication for US-guided FNAC LN examination. Systemic chemotherapy has been broadened to include all invasive tumours larger than 1 cm regardless of axillary status (Fisher et al, 1998
). Nevertheless, to establish the metastatic status of LN can be useful for evaluating the response to chemotherapy and the prognostic work-up of patients who are not surgical candidates. In the present series, nine of the patients enrolled by the oncologist for neoadjuvant chemotherapy were diagnosed with LN metastases using the guided FNAC procedure.
Finally, in a quality control programme, the evaluation of the standard goal that range from 32 to 40% of metastases in the SLNB (de Kanter et al, 2000
; Rahusen et al, 2001
; Veronesi et al, 2001
) has to be re-evaluated taking into account that most of the metastatic LN can be already diagnosed with US-guided FNAC. In the present series, the selection of cases markedly reduces the percentage of metastatic SLNB with respect to the values accepted in the literature and, at the same time, increases the number of axillary dissection as first-choice treatment. Thus, US-guided FNAC of axillary LN could generate cost savings to the health care system by reducing the added cost incurred by subsequent axillary dissection for the patients who show metastatic SLN.
In conclusion, because of its low cost and high specificity, we pro-pose that examination by US of the axilla in patients with diagnosed breast cancer should be performed any time before surgery. FNAC of selected LN should be added to the diagnostic protocol, in order to directly schedule patients with a cytological diagnosis of LN metastases to the appropriate treatment, avoiding the SLNB.