A Prospective, Randomized Trial of Sentinel Lymph Node Biopsy Versus No Additional Staging in Patients with Clinical T1-2 N0 M0 Invasive Breast Cancer and Negative Axillary Ultrasound
Amy E. Cyr, M.D.
Currently almost all women with breast cancer undergo sentinel lymph node biopsy (SLNB) as part of the staging process. The goal of SLNB is to determine if breast cancer has spread to the axillary lymph nodes. SLNB typically involves injection of a dye into the breast before surgery, and then removal of the 1-2 lymph nodes in the axilla that absorb the dye. For women undergoing breast conservation therapy, SLNB is a separate surgical procedure, requiring a second incision. There is no therapeutic benefit associated with SLNB; it is performed strictly to stage the axilla, predict prognosis, and help determine the most appropriate therapy. Unfortunately, SLNB is not a perfect strategy for staging the axilla. First, SLNB is an invasive surgical procedure, with potential complications including bleeding, infection, lymphedema (arm swelling) and sensory nerve injury. Second, SLNB fails to detect breast cancer spread to the axilla in up to 10% of cases. Given the significant limitations of SLNB, other strategies for staging the axilla should be considered as alternatives.
Ultrasound is a well-established technology, routinely used in combination with mammography to evaluate breast cancers. This technology has advanced considerably in the last ten years, and is routinely used at Washington University School of Medicine (WUSM) to evaluate axillary nodes for evidence of breast cancer spread. In a review of our initial experience with axillary ultrasound (AUS) at WUSM, we demonstrated that AUS has excellent performance characteristics for detecting breast cancer spread to the axilla. In our study, if the AUS was considered negative, very few patients had disease in the axilla. Of note, a focused review of cases at WUSM where AUS failed to detect breast cancer spread to the axilla suggests that if SLNB had been omitted following the relatively few false-negative AUS studies, treatment decision-making would not have been significantly altered in most cases.
Based on these analyses and our experience with AUS, the hypothesis of this application is that AUS represents a viable screening study for detection of clinically significant disease in the axilla. If the AUS is negative, SLNB can be safely omitted.
The study has gone well overall and there have been no unexpected adverse events. A total of 70 patients have been enrolled, and 2 patients were withdrawn (one was deemed ineligible and one patient chose to withdraw). Several postoperative complications (specifically seroma and wound infection) have been seen in the standard-of-care arm and were easily treated. We had one deviation from protocol in 2013 due to a change in location for patient consenting; this was reported to the IRB and the protocol amended to allow more flexibility for location for patient consenting. Additional subsequent protocol amendments have included broadening the inclusion/exclusion criteria and clarifying the reporting of adverse events.
Slow patient accrual has been an issue throughout this study. Reasons for this are multi-factorial and include originally very narrow inclusion criteria, patient buy-in (as seen in other surgical trials such as ACOSOG Z0011), and by competing clinical trials on campus. Patient accrual has been fairly steady, however, averaging 2-4 patients per month. This provides us with an expected rate of patient accrual for a larger, multi-institutional Phase II trial.
Two similar multi-institutional trials are now underway in Europe, but there has not yet been development of an American trial other than our pilot study. Data from this pilot study, funded by the Longer Life Foundation, has given us supporting data to apply for additional grant funding. In fact, a DOD grant was recently submitted in collaboration with other cancer centers including MD Anderson, University of Kansas, University of Pittsburgh, University of Wisconsin, Cleveland Clinic, and the University of Arkansas (awards have not yet been announced) and an R01 is being developed.
While completing the pilot phase of this trial, we updated our retrospective institutional data on AUS; this abstract (attached in the Appendix) was presented as a poster at the 2014 American Society of Breast Surgeons meeting and has subsequently been published. (Tucker NS, Cyr AE, Ademuyiwa FO, Tabchy A, George K, Sharma P, et al: Axillary ultrasound accurately excludes clinically significant lymph node disease in patients with early stage breast cancer. Annals of Surgery 2016, Epub ahead of print, PMID: 26779976.)
Results of the pilot study were presented as an oral presentation at the 2015 American Society of Breast Surgeons meeting and the manuscript has been accepted for publication and is in press (abstract attached in the Appendix). 70 patients were enrolled, and two patients were withdrawn as above. Arm A, the experimental arm (no SLNB) included 34 patients, and Arm B, the standard-of-care arm (SLNB) included 32 patients. The groups were well-matched in terms of patient and tumor characteristics and follow-up length. The median follow-up was 17 months (range, 1-32 months). Of the 32 patients randomized to Arm 2, three were found to have sentinel node disease missed with the AUS. For two of these patients, the missed disease was micrometastatic (< 2mm), which is considered clinically insignificant. One patient did have missed macrometastatic disease (> 2mm). This resulted in a negative predictive value (NPV) of 90.6% for AUS identifying any axillary disease, and a NPV of 96.9% for identifying clinically significant axillary disease. (Cyr AE, Tucker NS, Ademuyiwa FO, Margenthaler JA, Aft RL, Eberlein TJ, et al: Successful completion of the pilot phase of a randomized controlled trial comparing sentinel lymph node biopsy to no further axillary staging in patients with clinical T1-T2 N0 breast cancer and normal axillary ultrasound. Journal of the American College of Surgeons 2016, in press.)
Women with early stage breast cancer and with normal axillary (underarm) ultrasound (AUS) are randomized to the current standard of care, sentinel lymph node biopsy (SLNB), versus no surgical lymph node removal. Retrospective data suggest that AUS is accurate for detecting clinically significant lymph node disease (lymph node disease of large enough amount to alter adjuvant therapy recommendations), and prospective data suggest that removal of even metastatic lymph nodes does not improve disease control or survival in early stage breast cancer patients.
SLNB is the current standard of care for staging the axilla but it carries morbidities. Our study is a prospective evaluation of the performance characteristics of AUS. We believe prospective data will verify that AUS can identify clinically significant lymph node disease and that it may be able to replace SLNB as a means of staging the axilla without compromising outcomes.
Women with early stage breast cancer (T1-2) who are clinically node negative will undergo AUS. Women with abnormal or suspicious nodes on ultrasound are not eligible for the trial. If the AUS is normal, women who consent are randomized 1:1 to SLNB versus no axillary surgical staging. Women are treated as node negative for the purposes of adjuvant therapy (endocrine therapy, chemotherapy, and radiation therapy). Women will be followed for 5 years to monitor for local, regional, and distant recurrence.