SAN FRANCISCO—A speaker in a poster discussion session on sentinel nodes at the Breast Cancer Symposium here had some general criticism for research papers on breast cancer treatment: Too many studies do not recognize the importance of biology, said David W. Ollila, MD, Professor of Surgery in the Division of Surgical Oncology at the University of North Carolina. “Biology must dictate therapy. If we don't acknowledge that, then we are not advancing the science and we can't interpret different papers—different subtypes and different phenotypes exist, which need to be reflected in studies.”
Tumor Distance from Nipple Improves Nomogram Performance
One of the papers included in the session described an unlikely prognostic factor—namely, the distance of a breast cancer tumor from the nipple (Abstract 49).
That distance is associated with sentinel lymph node positivity, according to data from researchers from Medical College of Wisconsin and the Mayo Clinic, who reported that the addition of this measurement improves the prediction of node positivity in established nomograms. First author was Miraj C. Shah-Khan, MD, Assistant Professor of Surgery in the Division of Surgical Oncology at Medical College of Wisconsin, and the senior author was Judy C. Boughey MD, Associate Professor of Surgery at Mayo Clinic College of Medicine.
The team reviewed pre-biopsy ultrasound images of clinical T1/T2 tumors to measure the distance of the tumor from the nipple. The Memorial Sloan-Kettering Cancer Center and MD Anderson Cancer Center nomogram predictions were then calculated, and the AUC-ROC (area under the curve of a receiver operating characteristic) for each model was calculated—AUC-ROC is a statistical tool that compares the true-positive and false-positive rates at various threshold settings to select an optimal model.)
Included in the analysis were 401 tumors from 395 women; 79 of the 401 (19.7%) were sentinel node positive. In patients with positive nodes, the tumors were significantly closer to the nipple, the authors reported: 17 of 33 (51.5%) of tumors within 2 cm of the nipple were node-positive vs. 62 of 368 (16.8%) of tumors more than 2 cm from the nipple.
Both nomograms demonstrated good discrimination between node-positive and node-negative patients with AUC-ROC values of 0.71 and 0.74, respectively, Shah-Khan reported. When added to the Memorial nomogram, distance from the nipple of 2 cm or less contributed significantly to prediction of node positivity and improved the AUC-ROC to 0.73.
Similarly, distance from the nipple of 2 cm or less was significant when added to the MD Anderson nomogram and improved the AUC-ROC to 0.76.
“Tumor distance from the nipple should be factored in when considering the likelihood of nodal positivity for treatment planning,” Shah-Khan concluded.
In his discussion, Ollila said that this measurement is an additive for oncologists who use nomograms. “This [additional information] would allow them to counsel their patient on whether or not they should undergo a sentinel node procedure and whether or not that sentinel node will be positive or not,” he said.
He noted that this same research group had previously reported on using tumor distance from the nipple and from the skin to predict sentinel node positivity (Ann Surg Oncol 2011;18:3174-3180). “What they have done now is added that to the MSK and MD Anderson nomograms, and now have a more robust model for predicting positivity,” Ollila said. “So if you are a user of nomograms, or if you are a selective user for some patients who are still on the fence with an invasive breast cancer about whether or not they should undergo the sentinel node, this gives you a little more ability to predict what the positivity rate is.”
Ollila added that he himself is a selective user of nomograms: “The individual patient in front of me is not a statistic—she is a ‘yes’ or a ‘no’ for sentinel node biopsy, so using a nomogram for me is only for those on the fence; it is not a routine part of my practice.”
Triple-Negative Breast Cancer: No Higher Risk of Nodal Metastases
Patients with triple-negative breast cancers tend to be younger and have larger tumors which are more frequently high-grade, noted Alexandra Gangi, MD, a resident at Cedars-Sinai Medical Center, who presented a study about the likelihood of nodal metastases in patients with triple-negative breast cancer (Abstract 50).
The data showed that despite larger size and higher grade, patients with triple-negative breast cancer do not have an increased likelihood of lymph node positivity compared with those with non-triple negative breast cancer.
“The triple-negative phenotype is not a predictor of positive lymph node status nor a predictor of a higher number of positive nodes,” Gangi said.
The researchers looked specifically at nodal metastases independently of known predictors of increasing nodal involvement, and the triple-negative cohort did not have an increased rate of recurrence.
“It seemed to be actually predictive—if you have patients who have lower-grade tumors or are older with smaller tumors, then the ultimate treatment modality and the need for radiation therapy might be different for those patients than for others with triple-negative disease,” she said.
A database review identified 2,967 women with invasive breast cancer who underwent mastectomy or breast-conserving surgery between January 2000 and May 2012. Only patients who had sentinel node biopsy, axillary lymph node dissection, or both were included. Those receiving neoadjuvant therapy were excluded.
Patient and tumor characteristics evaluated included age; race; tumor size; grade; stage; histologic subtype; presence of lymphovascular invasion; and, ER, PR, and HER2 status.
Breast-conserving surgery was performed in 1,889 patients and mastectomy in 1,078 patients. Among those 2,967 patients, sentinel lymph node biopsy was performed in 1,094 (37%), axillary lymph node dissection in 756 (25%), and 1,117 (38%) patients had both.
Lymph node metastases were detected in 1,050 patients (35%). The lymph node positivity rate varied across subtypes with 734/2,201 (33%) in Luminal A; 143/344 (42%) in Luminal B; 108/278 (39%) in triple-negative breast cancer; and 65/144 (45%) in HER-2. On multivariable analysis, however, there was no difference in lymph node positivity among subtypes.
“Only age less than 50, grade 2 or 3 tumors, size greater than 2 cm, and presence of lymphovascular invasion were significant predictors of lymph node positivity,” Gangi said.
Four or more involved nodes were seen most commonly in the HER2 and Luminal B subtypes (19% and 14%, respectively), but not in triple-negative breast cancer or Luminal A (9% of each).
“If you have an older patient with a smaller tumor of a lower grade, you don't necessarily have to think that because they have a triple-negative breast cancer they have an increased likelihood of having positive lymph nodes,” she said.
In his Discussant remarks, Ollila said this is a controversial subject. “In the literature, estimates of the positivity rate of nodal metastases in triple-negative patients are ‘all over the map.’”
Ollila said that these authors, from a very large data set, conclude that individuals with triple-negative breast cancer do not have a statistically significant difference in the rate of positive sentinel lymph nodes compared with those with non-triple negative disease.
Ollila noted that the senior author on the study, Armando E. Giuliano, MD, Professor of Surgery and Chair of Surgical Oncology at Cedars-Sinai, is the senior author on the ACOSOG Z0011 study, which showed no overall survival benefit for axillary node dissection in patients with limited nodal disease.
“Can we really apply Z11 to triple-negative patients? I would say no,” Ollila said. “The reason goes back to biology: Triple-negative breast cancers are relatively radio-insensitive, and have an increased risk of locoregional recurrence after breast preservation.
“There are presentations [at the Breast Cancer Symposium] that suggest that the ports from the radiation are why we had such good results in Z11, or it may be that all those women got anti-estrogen hormonal blockade that actually works in hormone-positive breast cancer,” he said. But it doesn't work in triple-negative disease, so it's hard for me to say triple negative is the same as hormone-sensitive breast cancer.”
He added that more work needs to be done in this area.
Regional Recurrence Rare in Early-Stage BrCA with Negative SLNB
Also discussed in the session, a Japanese study evaluating regional recurrence rates and predictors of regional recurrence showed that regional recurrence is rare in early-stage breast cancer patients with negative sentinel lymph node biopsy (SLNB) (Abstract 51).
And lymphovascular invasion and nuclear grade can be used as predictive factors of regional recurrences after negative SLNB, concluded Akiko Matsumoto, MD, of the Department of Surgery at Keio University Hospital in Tokyo, who presented the study.
A regional recurrence rate of only 1.1 percent was seen in patients with negative SLNB. Between January 2001 and December 2012, of 1,536 patients with T1NM0 invasive breast cancer who underwent SLNB at that hospital, 1,185 (77%) had negative sentinel lymph nodes.
Matsumoto and colleagues explained that SLNs were detected using a combined method of blue dye and small-sized technetium 99m-labeled tin colloid. Intraoperative frozen examination was performed with H&E staining. SLNs, fixed and embedded in paraffin, were additionally diagnosed with H&E staining and immunohistochemical analysis.
At a median follow-up of 54.8 months there were 28 local (2.3%), 13 regional (1.1%), and 34 distant recurrences (2.9%). The median disease-free intervals of regional and distant recurrences were 32.6 and 22.7 months, respectively.
A higher nuclear grade was found to significantly correlate with regional and distant recurrences. The rate of lymphovascular invasion was significantly higher in patients with regional recurrences compared with patients without recurrences (58.3% vs. 27.6%), but lymphovascular invasion was not a significant predictor of distant recurrences, the researchers found.
Also, ER negativity significantly correlated with distant recurrences, but it was not associated with regional recurrences.
Ollila applauded the Tokyo group for its study design and the findings, but particularly for the Japanese surgeons' low SLNB false-negative rate.
“They are presenting a false-negative rate of one percent, with almost equivalent follow-up to the ACOSOG Z0011 study,” Ollila said. “They are presenting us with the standard that needs to be looked at, a benchmark to shoot for.”