Assay has high sensitivity for hard-to-detect SLN mets

SAN ANTONIO—A rapid molecular assay outperforms frozen sections when it comes to detecting breast cancer metastases in sentinel lymph nodes (SLNs), new data show. Moreover, the assay tends to perform better in cases where detection of metastases by frozen section is known to be difficult, such as cases of lobular carcinoma.

"In this era of sentinel node biopsy, we are not removing all of the axillary nodes, and histology varies all over the country, if not all over the world, in terms of its sensitivity and ability to pick up disease," Peter W. Blumencranz, MD, said at the 2007 San Antonio Breast Cancer Symposium (abstract 53). "So if we can come up with a more reproducible, standardized program to assess nodal involvement, I think it will behoove us—surgeons and pathologists—as the patients' advocates, to be able to tell them if there is disease in those nodes."

The molecular assay studied, the GeneSearch Breast Lymph Node (BLN) Assay, which was FDA approved in July 2007, is based on detection of two markers—cytokeratin 19 and mammaglobin—in sentinel nodes by reverse transcription-polymerase chain reaction (RT-PCR), said Dr. Blumencranz, a breast surgeon at Morton Plant Mease Health Care, Clearwater, Florida, and medical director of Comprehensive Breast Health Services.

The assay can evaluate up to half of the node, and the result is reported as positive if either or both markers are detected, Dr. Blumencranz said. "This is a rapid test. It can be done in less than 40 minutes, so the surgeon can get this information in the operating room if he wishes to make a decision to proceed with an axillary node dissection," he noted.

Multicenter study

In the prospective, multicenter trial, 490 women with invasive breast cancer underwent axillary lymph node dissection with sentinel lymph node sectioning and evaluation of adjacent sections for metastases by the molecular assay, by frozen section, and by permanent paraffin section (the comparator).

Many of the women had features known to make it difficult to assess for metastases by frozen section: 11% had lobular carcinoma, 45% had T1 tumors, 42% had grade 1 tumors, and 2% had received neoadjuvant therapy.

In the trial population overall, the assay had better sensitivity than frozen section for detecting sentinel node metastases (91% vs 82%), along with comparable specificity (95% vs 97%), positive predictive value (87% vs 93%), and negative predictive value (96% vs 93%), Dr. Blumencranz reported.

In analyses stratified according to the size of the metastases, Dr. Blumencranz said, the assay was more sensitive both in the 117 women with macrometastases (larger than 2.0 mm) (97% vs 91%) and in the 22 women with micrometastases (greater than 0.2 up to 2.0 mm) (68% vs 41%).

The difficult cases

Although numbers of patients were limited, the assay also appeared to be more sensitive than frozen sections for detecting sentinel node metastases in the difficult cases, Dr. Blumencranz said (see Table on page 46).

When these groups were further stratified by size of metastases, the difference in sensitivity in favor of the molecular assay was most marked in the patients with micrometastases, while specificity was comparable for the assay and frozen section.

In the micrometastases patients, for example, the assay found 3 of 4 lobular cancers, compared with 0 of 4 for frozen section.

"This assay would appear to be a viable replacement or an adjunct for frozen section and reduce the need for second surgeries," Dr. Blumencranz asserted.

Future studies

Future investigation using the assay might include larger studies of patients who have received neoadjuvant therapy and studies evaluating the need for completion axillary node dissection, he proposed.

"There might be a point in the future where perhaps taking the whole node and doing 100% analysis by molecular assay and combining that with the genetic profile of the tumor may be basically all we need," Dr.Blumencranz said.

Finally, he said, the assay could help determine how extensively the node is involved.

"Hopefully, by next year we can bring you back some correlation between the actual quantitative RT-PCR and the measured disease in the node, that is, the tumor burden in that histologic half of the node that has been preserved," he concluded.