SAN ANTONIO – The results of a sentinel node biopsy performed after neoadjuvant chemotherapy for breast cancer are often unreliable, new data suggest.
Researchers with the German SENTINA trial prospectively analyzed the impact of the timing of sentinel lymph node (SLN) biopsy on the accuracy and feasibility of the procedure in 1,737 patients with early breast cancer who received neoadjuvant chemotherapy.
Patients with initial clinically node-negative disease (cN0) underwent SLN biopsy before chemotherapy. If biopsy results were positive, they underwent biopsy along with an axillary dissection after chemotherapy. Patients with clinically node-positive disease (cN1) underwent chemotherapy followed by SLN biopsy plus axillary dissection if they had a downstaging to clinically node-negative disease.
The SLN biopsy procedure was standardized according to interdisciplinary consensus (Cancer 2005;103:451-61); radiocolloid tracer with lymphoscintigraphy was mandatory for identifying sentinel nodes, whereas blue dye was optional.
Study results showed that the false-negative rate was 51.6% in patients with cN0 disease who had an initial SLN biopsy showing nodal involvement and therefore went on to have another SLN biopsy after the chemotherapy.
SLN biopsy had a false-negative rate of 14.2% when performed after the neoadjuvant chemotherapy in patients with initial cN1 disease who had downstaging to cN0 disease, lead author Dr. Thorsten Kuehn reported at the San Antonio Breast Cancer Symposium.
This rate was up to twice as high as that seen historically in patients instead having the SLN biopsy at the time of primary surgery, which has been on the order of 7%-10%.
"The SLN biopsy as a diagnostic procedure is not a reliable tool in patients who convert on neoadjuvant chemotherapy from cN1 to cN0 compared to SLN biopsy in primary surgery," he said.
Within this downstaged group, data showed that the odds of a false-negative result were reduced by half when patients had more than one sentinel node examined.
"It strikes me that your conclusion should have been, if you take two or more sentinel nodes, this is a safe procedure," one attendee noted during a discussion session. "But if you take less than that, it’s not safe. Is that true?"
The study biopsies were adequate in that all patients had lymphoscintigraphy, contended Dr. Kuehn, who is head of the breast center and chief of the gynecology and obstetrics clinic at Klinikum Esslingen in Germany. "I don’t think you necessarily always have two sentinel nodes – you may have one sentinel node. And every institution or every surgeon had to prove that the sentinel nodes that were shown on lymphoscintigraphy had to be removed."
Another attendee wondered about the extent of residual axillary disease in the patients having false-positive SLN results.
"As a radiation oncologist, I am seeing a lot of patients who have had sentinel node biopsies done after neoadjuvant chemotherapy without an axillary dissection, and I’m being asked to irradiate the patient, which makes me uncomfortable because I don’t know what the volume of disease left is," he explained. "Do you have a sense of what the volume of disease is – micrometastases, macrometastases, number of nodes – in the patients with false-negative sentinel nodes, and do you also have a sense of what that volume is in the axillary dissection specimen outside the true-positive sentinel nodes?"
"We don’t have information on the size of the metastases. We have information on the number of involved non–sentinel nodes, and these numbers are from 1 to 11," Dr. Kuehn replied.
"What we want to say is SLN biopsy is a diagnostic tool, and this tool is worse after previous treatment. I can say I work with this tool, but it is not a good diagnostic tool," he added.
In the study, SLN detection rates, a measure of feasibility, also differed significantly across groups. Surgeons were able to detect these nodes in 99% of all SLN biopsies performed before any treatment, but in only 80% of those performed after a downstaging of disease with the neoadjuvant chemotherapy and in only 61% of the repeated SLN biopsies. The pattern of nodal uptake of the radiocolloid tracer mirrored these findings.
"Previous local and systemic treatment significantly impairs the tracer uptake and the detection rate," concluded Dr. Kuehn, who disclosed no conflicts of interest related to the research.
In a multivariate analysis in the downstaged group, none of a variety of tumor characteristics significantly predicted the ability to detect an SLN.