Contemporary Evaluation of Breast Lymph Nodes in Anatomic Pathology

Beth T. Harrison, MD; Jane E. Brock, MBBS, PhD


Am J Clin Pathol. 2018;150(1):4-17. 

In This Article

Gross and Histopathologic Evaluation of the SLN

It is a common misconception that the goal of sentinel node evaluation is to detect all metastases, no matter how small. Our institution receives several thousand consults every year from hundreds of different hospitals across the country where sentinel nodes have been evaluated with multiple H&E levels and keratin immunohistochemistry (IHC), with the intent to identify isolated tumor cells (ITCs; deposits <200 cells or 0.2 mm). Even among pathologists at academic institutions, there is a belief that we need to detect all micrometastases (deposits 0.2–2 mm) in lymph nodes and that we can do this best with three H&E levels. In fact, it would require 10 H&E levels to detect all deposits greater than 0.2 mm in a node sectioned at 2 mm in thickness, which is cost prohibitive and far too time-consuming. Detecting all micrometastases in this manner is not necessary for therapeutic decision making, as the presence of micrometastases no longer inevitably results in additional treatment to the axilla, and the clinical significance appears to be minimal.

Over the past 40 years, more than 50 retrospective studies have investigated the prognostic significance of low-volume or occult disease in the axillary lymph nodes, the results of which have been conflicting.[16] A few large, population-based studies have suggested a negative impact on prognosis. Of note, a Surveillance, Epidemiology, and End Results (SEER)–based analysis of US women with early breast cancer diagnosed from 1992 to 2003 showed that micrometastatic disease (N1mi) carried a prognosis intermediate to node-negative (N0) and macrometastatic (N1) disease, even after adjusting for tumor and patient-related factors.[17] A subsequent analysis of SEER data from 2004 to 2011 found that patients with breast cancer with occult metastases detected via IHC had a worse overall survival; however, ITCs had no prognostic significance.[18] In addition, in a retrospective analysis of nationwide data from the Netherlands Cancer Registry, referred to as the MIRROR (Micrometastases and Isolated Tumor Cells: Relevant and Robust or Rubbish?) study, de Boer et al[19] investigated the prognostic significance of micrometastases in patients with early breast cancer who had SLN biopsy. Among women who did not receive adjuvant chemotherapy, ITCs or micrometastases in regional lymph nodes were associated with a reduced 5-year DFS rate (76.5 vs 85.7%, P < .001). DFS was improved in women with ITCs or micrometastases who received adjuvant chemotherapy (86.2% vs 76.5%, P < .001). In a systematic review of the literature prior to 2010 by the same authors, pooled analyses revealed worse survival in association with metastases 2 mm or less (hazard ratio of death, 1.44; 95% confidence interval [CI], 1.29–1.62) and occult metastases (relative risk [RR] of recurrence, 1.55; 95% CI, 1.32–1.82; RR of death, 1.45; 95% CI, 1.11–1.88).[16] However, the retrospective studies have had several limitations, including small sample size, lack of multivariable analysis, lack of standard pathologic assessment of lymph nodes, and lack of distinction of ITCs from micrometastases, among others.

Although small metastases appear to be associated with worse outcomes in retrospective studies, recent prospective studies have provided evidence to the contrary. In 2011, the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-32 RCT reported on the clinical impact of occult lymph node metastases detected on two deeper levels and keratin IHC in 3,887 patients who were tumor free on an initial H&E section. The standard protocol for the initial pathologic examination of the SLNs involved slicing them at approximately 2-mm intervals, submitting them entirely, and examining a single H&E section. As intended by the grossing protocol, most macrometastases (>99.5%) were detected on the initial H&E level. Occult metastases were identified on deeper levels in 16% of patients, of which 72% were isolated tumor cells, 28% were micrometastases, and only 0.02% were macrometastases. All patients were treated as node negative, and a very small but statistically significant difference in the OS, DFS, and distant disease-free interval was present between those with and without occult metastases, respectively. The 5-year Kaplan-Meier estimates of OS were 94.6% in the occult metastases-positive group vs 95.8% in the occult metastases-negative group. The authors concluded that the differences, although statistically significant, are minimal and do not justify changes in clinical management.[7]

At around the same time as the NSABP B-32 results, the American College of Surgeons Oncology Group (ACOSOG) published the results of the Z0010 study, a prospective observational study of occult metastases detected in one deeper level used for keratin IHC in 3,326 SLN specimens. Lymph nodes were processed in the same way, sliced at 2-mm intervals, and submitted entirely. Occult metastases were found in 10.5%. Size of metastases was not reported. There was no significant difference in 5-year OS (95%; P = .64) or DFS (P = .82) between patients with and without occult metastases.[8] Similar conclusions can be drawn from the results of Z0010 and B-32: multiple levels and keratin IHC are unnecessary.

It was initially difficult for pathologists to accept these results, and practice largely remained unchanged. Multiple levels and IHC were routinely performed despite national recommendations from the College of American Pathologists, the Association of Directors of Anatomic and Surgical Pathology, and the American Joint Committee on Cancer advising against this. All organizations agree that the goal is to detect all macrometastases (>2 mm). They recommend that the SLNs should be sliced no thicker than 2 mm and submitted entirely and that one H&E level is sufficient.[20–22]

It must be acknowledged that there are situations when keratin IHC on lymph nodes is useful and appropriate, most notably in the search for metastases of classical lobular carcinoma with low nuclear grade Image 1A and Image 1B. It can be very difficult to spot the dispersed pattern of tumor cells, and it is possible to miss an almost invisible significant tumor burden within the lymph nodes. Routine use of keratin IHC in the evaluation of the SLN is recommended in that setting. In addition, keratin IHC may be used to identify cells seen on H&E that are difficult to classify as epithelial.

Image 1.

Pitfalls in diagnosis of lymph node metastases. Well-differentiated metastatic lobular carcinoma may be invisible on H&E (A, ×10), but a macrometastasis is highlighted by a keratin AE1/AE3 immunostain (B, ×10). Endosalpingiosis in a lymph node capsule (C, H&E, ×40) often differs in morphologic appearance from a primary tumor or may appear in the setting of ductal carcinoma in situ only. The tubules are typically well formed and highlighted by PAX8 stain (D, ×40) and WT-1 stain (E, ×40). Aberrant staining with pankeratin of plasma cells and skin squamous epithelial cells, mimicking keratin-positive tumor cells (F, ×40). Nevus cell rests in the lymph node capsule (G, H&E, ×20) do not always contain melanin pigment, as in this case.