In this study, we found that the incidence of sentinel lymph node metastasis in cases of pure DCIS was 0.39%. This incidence was significantly lower than that in cases of IDC-predominant invasive tumors (0.39% vs. 6.2%; p < 0.001). Therefore, our data suggest that sentinel lymph node biopsy can be avoided in cases of pure DCIS.
Many publications concerning this issue have reported only the rate of sentinel lymph node metastasis in pure DCIS. We also calculated the rate of metastasis in IDC-predominant invasive lesions. We believe that the relevance of the metastasis rate in pure DCIS is supported by comparing data concerning IDC-predominant invasive lesions. Furthermore, we can estimate the rate of sentinel lymph node metastasis in lesions mimicking DCIS clinically.
The issue of pure DCIS and sentinel node biopsy is associated with two major problems: one is that preoperative diagnosis of pure DCIS is difficult, and the other is that postoperative definitive diagnosis of pure DCIS is also difficult.
It is well known that preoperative diagnoses of DCIS based on core needle biopsy are likely to be underestimated. Rates of diagnosis range from 8.3% to 43.6% [8, 13, 14]. Preoperative core needle biopsy does not guarantee that the entire lesion is without stromal invasion. Furthermore, having less than 0.5 cm of stromal invasion increases the incidence of sentinel lymph node metastases [15, 16]. As a result, many investigators insist that sentinel lymph node biopsy should be encouraged when DCIS-like tumors are large enough to be palpable or when tumors require total mastectomy.
Furthermore, the postoperative pathological diagnosis of pure DCIS does not always guarantee the absence of lymph node metastasis. For many years, it has been believed that DCIS is associated with the absence of lymph node metastasis, that axillary dissection in DCIS could be omitted, and that cases of lymph node metastasis in DCIS are associated with invasive lesions that are too small to be detected by the usual pathological examination. However, in regular clinical practice the detection of minimal stromal invasion is quite difficult. Although sentinel lymph node biopsy is effective in DCIS, we suggest that the application of sentinel node biopsy to all DCIS cases should be avoided. That is because, although sentinel node biopsy is less morbid than axillary dissection, the procedure is not completely free from morbidity .
We believe that Moore et al., who encouraged the use of sentinel lymph node biopsy in pure DCIS, does not argue that sentinel lymph node biopsy should be carried out in all cases of pure DCIS . In their literature, only 22% of all DCIS cases had sentinel lymph node biopsy. The relatively high rate of axillary lymph node metastases in their study can be associated with this selection.
In our series there was one case of pure DCIS with positive sentinel nodes. This case underwent a partial mastectomy, and the surgical margin was slightly positive. Preoperative mammography, ultrasonography, and MRI did not reveal any other abnormal lesions besides the main tumor. However two sentinel nodes were positive for cancer and both metastases were larger than 2 mm. We think that this was an extremely rare case. Although some authors encourage the preservation of axillary nodes in cases of pure DCIS with positive sentinel nodes , an axillary dissection was performed in this patient.
There is much debate concerning the association between preoperative invasive procedures for diagnosis and the likelihood of lymph node metastases. Displacement of cancer cells around the main tumor is common, and frequencies from 28% to 32% have been reported previously [18, 19]. Moreover, there is the possibility that displacement can cause the migration of cancer cells to lymph nodes. However, the prognostic significance of this migration is uncertain. Previous studies show that large gauge needle biopsy does not affect the survival risk [21, 22]. Much more discussion and careful studies on this issue are necessary.
Our study has a considerable limitation. Our series could miss cases of micrometastases or isolated tumor cells (ICT) in sentinel nodes. In order to avoid this problem, the sentinel nodes should be sectioned at intervals of at least 0.15 mm and immunohistochemistry should be applied to sections at different levels. These analyses should be performed on permanent paraffin sections. Although the clinical significance of micrometastases and ICT in DCIS has been unclear , the latest report has shown that micrometastases or ICT may decrease the probability of survival in invasive breast cancers .
In conclusion, we found that the incidence of sentinel lymph node metastasis in cases of pure DCIS was 0.39%. This incidence was lower than that in IDC-predominant invasive lesions. Therefore, we believe that sentinel lymph node biopsy in pure DCIS can be safely omitted.