Micrometastases and Isolated Tumor Cells in the Lymph Nodes: What Do They Mean for Breast Cancer Patients?

• SAN ANTONIO—Interest has been growing for several years in determining the prognostic significance of micrometastases and isolated tumor cells (ITCs) in the lymph nodes of breast cancer patients. Several presentations at the 29th San Antonio Breast Cancer Symposium (SABCS) shed light on this issue—but did not reach a consensus. Micrometastases are defined as lesions 0.2 to 2.0 mm and are detected by step-sectioning H—E staining, while ITCs (sub/nanomicrometastases) are less than 0.2 mm and are usually found only by immunohistochemical staining (IHC).

In a Plenary Lecture, Emiel Rutgers, MD, PhD, a surgeon associated with the Netherlands Cancer Institute, Amsterdam, noted that few oncologists currently recommend adjuvant therapy for patients based solely on the presence of micrometastases. Nonetheless, many worry about possible disease progression in these patients.

Based on multivariate analyses in a number of studies, the finding of occult micrometastases in the lymph node is not a significant prognostic factor for poorer survival, Dr. Rutgers maintained. He drew on several studies, including most recently a study from SABCS in 2005 by Cox C et al, who reported that the chance of having "further involvement" in the nodes after initial sampling was 14%, equally divided between micrometastases and isolated tumor cells. After a mean follow-up of 2 years, for patients with T1 or T2 tumors, there was no difference in disease-free survival between patients with these minute lesions and patients who were completely node negative, Dr. Rutgers said.

In another study from the John Wayne Cancer Center, 5-year disease-free survival was approximately 98% for patients with either micrometastases or IHC-negative nodes, but only 73% for patients with macrometastases.

But micrometastases can indicate the presence of metastases in additional axillary nodes, he added. In a series of 2,150 patients from his institution, 650 (30%) were found to be sentinel node positive, 18% of whom had macrometastases. Another 7% had micrometastases, and 5% had ITCs, for a 12% rate of submacrometastases, which is in keeping with other series, he noted. Complete axillary dissection of these patients revealed micro- or macrometastases in additional nodes in 19% of patients, resulting in upstaging in 15% and prompting treatment changes in 7%. Isolated tumor cells heralded additional metastases in less than 8%, and prompted no treatment changes.

"In conclusion, the prognostic significance for survival when you have micrometastases only, with all axillary lymph nodes examined, is unclear and, at most, limited," Dr. Rutgers maintained. "Prognosis is related to size of the metastasis and the primary tumor characteristics."

Why Bother?

Why bother with this issue? he asked. "At most, with chemotherapy these patients would gain a 1% to 2% improvement in survival, and most of them receive adjuvant therapy anyway [based on other factors]," he said. Prognostic assessments might be better derived from other tools, he said, such as the 70-gene prognostic gene array (Mammaprint) developed at his institution, which can distinguish risk profiles even in patients with a single positive node.

"I think that pathologists should look for disease for which we know the clinical relevance," Dr. Rutgers said. "All the other things we can do—RT-PCR, multiple slicing, blending machines for lymph nodes—are the province of research, not routine clinical practice yet."

Recommendations

Based on the data at hand, Dr. Rutgers recommended the approach taken in the Netherlands:

• Bisect lymph nodes less than 10 mm in size.

• Slice lymph nodes 10 mm or larger. Take three levels per slice, and at 150-μm intervals, do H—E staining, with cytokeratin staining if possible.

• With micrometastases (0.2 to 2.0 mm), a lymph node dissection is generally advised. Patients with negative non-sentinel nodes should be considered node negative and treated with adjuvant therapy based on the primary tumor characteristics. Positive non-sentinel nodes are an adverse sign, and patients should be treated accordingly. Isolated tumor cells (less than 0.2 mm) have no reasonable clinical relevance, at this point, and these patients should be considered node negative, he said.

Italian Study

In another presentation (abstract 25), Saverio Alberti, MD, PhD, of the University of Chieti, Italy, said that current data on micrometastases are not informative enough. Since 50% of breast cancer recurrences are diagnosed more than 5 years after surgery for the primary tumor, studies initiated after 2002 to assess prognostic significance are immature, he said. Furthermore, he said, most studies of the topic have been underpowered to detect differences.

In a study involving 702 patients with 8.2 years of follow-up, investigators reevaluated node-negative patients for ITCs and their relationship to clinical outcomes. A total of 6,676 lymph nodes were removed from 377 patients with node-negative (pN0) disease. Using eight sectioning levels per node with a spacing of 100 μm between levels, pathologists provided a total of 250 sections per lymph node, all stained with H—E and anticytokeratin IHC, for a gross total of 90,000 data points.

The investigators detected 24 instances of ITCs (6.4%) and 25 (6.6%) of micrometastases. The frequency was highest in lobular carcinomas (mostly ITCs).

The relative risk for any adverse event was 2.51 in patients with ITCs vs those with no ITCs (P = .00019). A multivariate Cox model produced a relative risk of 2.16 (P = .0005). "The incidence of all adverse events in patients with micrometastases was higher than anticipated over time, compared with patients with ITCs," he said. Further multivariate analyses found the presence of ITCs accounted for 50% of the risk of relapse. "Up to 36 months, we saw nothing. Then the curves started to differentiate very much, producing a huge relative risk," he reported.

According to Dr. Alberti, these results suggest that isolated tumor cells are metastatic at very early stages of disease, as opposed to current models that propose a step-wise progression of tumor that requires additional mutations.