Prognostic Significance of True Occult Axillary Metastases

Abstract & Commentary

Synopsis: There is, as yet, no consensus in the literature regarding the prognostic significance and subsequent clinical management of patients with axillary nodal metastases transparent to H&E staining but detectable on immunohistochemical evaluation. This study reports that results shifted over extended follow-up and concludes that there is no significant difference in prognosis between node-negative patients in comparison to those with micrometastases identified on immunostaining.

Source: de Mascarel I, et al. Brit J Cancer. 2002;87: 70-74.

In this report, de mascarel and colleagues reported updated results on a group of postmastectomy/axillary dissection patients with longer follow-up. They previously published findings at 10 and 15 years, and now offered data from 20 years of follow-up.

Between 1965 and 1984, 2768 women underwent modified radical mastectomy and axillary node dissection at the Institut Bergonie in France. The mean number of lymph nodes examined at the time of surgery was 14 (range, 2-29). Each node was serially sectioned using 1-1.5 mm slices perpendicular to the long axis, with a mean of 4 slices per node (range, 1-9). All slices were evaluated on hematoxylin and eosin (H&E) staining. Fifty four patients received postoperative radiotherapy, and virtually none received chemotherapy. de Mascarel et al selected a subset of 218 patients whose axillary lymph nodes were negative on H&E staining for further analysis with immunohistochemical staining, including 129 with infiltrating ductal carcinoma (IDC), and 89 with infiltrating lobular carcinoma (ILC).

Restaining with a cocktail of 5 monoclonal antibodies directed against epithelial cell antigens was performed on the original diagnostic H&E slides that were negative on routine examination. Unequivocal metastases were found by immunostaining in 13 IDC patients (10%) and 37 ILC patients (41%). All IDC metastases were located in 1 lymph node, while 26% of ILC metastases were in 1 node, 6% were in 2 and 3 nodes each, and 3% were in 4 lymph nodes. The IDC metastases were characterized as small tumor clusters in the subcapsular sinuses up to 0.2 mm, and the ILC metastases were typically irregularly distributed single cells anywhere in the entire nodal section.

de Mascarel et al’s earlier results were notable for lower metastasis-free survival and overall survival for IDC patients with occult nodal metastases at 10-year follow-up. At 15-year follow-up, overall survival was not significantly different from patients without nodal metastases, but metastasis-free survival was still significantly better in the node-negative IDC group. For the ILC patients, there was never a difference in outcome for the node negative and occult node-positive groups in the previous publications. Now, at a median follow-up of 24 years for the IDC patients and 18 years for the ILC patients, neither actuarial metastasis-free nor overall survival was significantly different in comparison to the node-negative patients (P = .62 for ILC patients and P = .076 for IDC patients). There were 60/129 IDC and 19/89 ILC patients evaluable at the time of longest follow-up. Survival data were calculated from the time of surgery.

de Mascarel et al concluded that occult axillary nodal metastases detected on immunostaining do not have any prognostic significance at the time of extended follow-up for either infiltrating ductal or lobular carcinomas. They noted that their study was the only one to date that used serial macroscopic sectioning, provided extended follow-up, and differentiated between histologic types. They distinguished their technique of nodal analysis from those of other studies. Other researchers report that they most often recut blocks and find metastases with immunostains that might otherwise have been evident on H&E had more cuts been done. The French group emphasized that their findings were limited to patients with truly occult metastases. The fact that survival statistics lost their significance over time for IDC, while the statistics for ILC never showed significance, may point to a difference in the natural history of the 2 histologies. Further prospective work needs to be done to be sure that the results from this relatively small number of patients are applicable to the universe of node negative breast cancer patients.

Comment by Edward J. Kaplan, MD

This paper makes several good points. In the discussion section, de Mascarel et al reviewed 11 other studies and showed that most of them used the standard macroscopic technique of checking one node slice on H&E transected at the major axis of the lymph node. This is despite the fact that it is widely acknowledged that this style misses many metastases. They also felt that the high percentage of "occult metastases," ie, those visible on H&E but missed because of single slice analysis, may be the reason that "all women with breast cancer" seem to benefit from adjuvant chemotherapy, regardless of lymph node status.

In my opinion, before we can know whether occult metastases have prognostic significance, we must separate out 2 distinct entities which until now seem to have been lumped together. Those entities are macroscopic metastases noted on H&E staining using a multi-slice technique, and truly occult disease identified upon restaining of the same cuts with monoclonal antibodies. It is evident that, before a judgment can be rendered, we will need large patient numbers, longer follow-up, and meticulous technique. It should be easier to handle the caseload now that many axillary dissections are performed using a sentinel node technique, which means the number of slides to be checked is reasonable. As deMascarel et al described, all of the occult metastases in the IDC group were in one node, presumably corresponding to what would be the sentinel lymph node by today’s technique. It is less clear whether the same is true for ILC. de Mascarel et al never offered their selection criteria for the subset of node negative patients chosen for further study.

Dowlatshahi and colleagues, in a Medline review of papers examining micrometastases in axillary nodes, stated that the vast majority of lymph node metastases would be detected by taking 2 sections 0.3 mm apart and staining them with a single monoclonal antibody. This appears to be a practical and economically viable approach in the context of sentinel lymph node biopsies.1 In contradistinction to the French paper, Dowlatshahi et al noted that the more recent studies have consistently shown survival differences in patients with occult metastases. However, only about half of the papers included in their review used immunostaining. Many would argue that metastases found on H&E are not really occult by definition.

Tjan-Heijnen and associates in their paper on axillary micrometastases from the Netherlands, are of the opinion that survival outcomes are dependent on the size of the nodal metastases. This speaks to the issue alluded to above. They caution that the value of adjuvant therapy can be questioned for patients with micrometastases who otherwise have favorable prognostic factors.2

We may ultimately find that performing multi-slice H&E staining on sentinel lymph node tissue gives us all the prognostic information we need. Perhaps immunostaining will fall out of favor after we stop using it to find "occult metastases" that were there the whole time, in slices that we never checked with H&E. This seems like the type of study that would be tedious to perform, but could be done well by a group that maintains a large repository of relatively older sentinel node paraffin block specimens, where results can be correlated with follow-up data.

Dr. Kaplan is Acting Chairman, Department of Radiation Oncology, Cleveland Clinic Florida, Ft. Lauderdale; Medical Director, Boca Raton Radiation Therapy Regional Center, Deerfield Beach, FL.


1. Dowlatshahi K, et al. Cancer. 1997;80:1188-1197.

2. Tjan-Heijnen VC, et al. Breast Cancer Res Treat. 2001;70:81-88.