Current Status of Sentinel Node Surgery in Breast Cancer

There are two compelling reasons for performing sentinel node surgery in breast cancer patients: the increased accuracy of identifying lymph nodes that contain breast cancer (even outside the axilla) and the decreased morbidity compared to conventional axillary lymph node resection. Improved accuracy results from the conversion of an essentially blind one-size-fits-all protocol (axillary resection) to an individualized search for the nodes most likely to receive cancer cells via lymphatics leading from the tumor. Although longterm comparisons that quantitate the difference in morbidity between axillary node resection and sentinel node resection are not yet available, it is certainly expected that less morbidity will be associated with the excision of 3 lymph nodes as opposed to 15. Clinical Trials of Sentinel Node Surgery Sentinel node surgery remains experimental because long-term randomized trials comparing the survival outcome of this procedure with that of conventional axillary node resection have not been completed. Accrual to the node-negative trial (B-32) conducted jointly by the National Surgical Adjuvant Breast and Bowel Project (NSABP) and the University of Vermont is nearly complete, but outcome results will not be available for some time. Accrual to the node-positive trial conducted by the American College of Surgeons Oncology Group (ACOSOG Z0011) is further away from meeting target accrual. Survival End Point
These trials are important to ensure that sentinel node surgery, which leaves behind the majority of lymph nodes, does not result in lower survival than conventional axillary node resection. The impact of axillary node resection on survival has been evaluated in several trials. Data from each of six randomized clinical trials confirmed a higher survival rate in the group that underwent axillary dissection (Figure 1).[1-7] This includes NSABP B-04, which reported a 4% higher survival in the axillary dissection group. This trial, however, was underpowered to detect a 4% survival difference, and the investigators appropriately concluded that the trial data showed no significant difference in survival between the axillary dissection and observation groups. It has since been generalized that axillary dissection has no impact on survival. However, combined data from the available randomized trials indicate a statistically significant survival benefit of about 5% associated with axillary dissection. Now we need to know what impact sentinel node surgery will have on survival. This is an important issue because the major rationale for performing sentinel node surgery is to decrease morbidity. Is increased mortality- even a small increase-a fair price to pay for decreasing the morbidity of lymph node resection? The primary end point of both NSABP B-32 and ACOSOG Z0011 is to determine whether sentinel node surgery results in a survival rate as good as that of axillary node resection. These trials will accrue a sufficient number of patients to determine whether small differences in survival are associated with the two protocols. This is an important point because these trials are based on a careful review of the available literature on the impact of nodal surgery on survival. Hundreds of surgeons and thousands of breast cancer patients are involved in what are arguably the largest surgical trials in breast cancer ever conducted. These trials should answer the survival question with sufficient statistical power to satisfy nearly everyone. Ten Years of Progress
After a decade of research into the technical methods of sentinel node surgery and with results on the survival question incomplete, what is the status of the procedure now? In these 10 years, research on the methods and outcomes of sentinel node surgery have been applied to several major tumor types in which regional node resection is performed to coincide with surgery on the primary tumor. It is apparent that in all areas of investigation, exceptions to the conventional understanding of lymph node drainage patterns are frequently observed. Three major international sentinel node symposia have been held, and an International Sentinel Node Society has been founded. It can be safely said that surgeons in the United States who are not performing any type of sentinel node surgery for breast cancer (with or without completion axillary node resection) are in the minority. The remainder can be roughly categorized according to whether they are participating in one of the two randomized clinical trials, are in some sort of self-defined validation phase, or are simply performing sentinel node surgery without axillary dissection as their standard practice. Methods of Sentinel Node Resection A variety of methods are being used to perform sentinel node resection, but the technique has not been standardized. In the United States, a combination of tracers that label lymph nodes are injected into the breast. Blue dye facilitates visualization of lymphatic channels, and even after one has performed a large number of these procedures, it is still impressive to see a brilliant blue channel leading to a sentinel node. Radioactive tracer allows detection of nodes with a handheld gamma probe. The advantage of the gamma probe is that the location of the sentinel node can be determined prior to surgical exposure of tissues, allowing the surgeon to perform the smallest possible surgical procedure. In addition, this technique enables the detection of sentinel nodes located outside the level I axillary region. Blue dye can be used to tattoo the skin but, more importantly, can result in serious anaphylactic reactions. These reactions may be dramatic, with complete or near complete loss of vital signs, and usually occur within 15 to 30 minutes of exposure[8] in 1 of every 200 to 500 cases. Survival of the patient depends on immediate recognition of the reaction and aggressive management. Radioactive tracers appear to have a better safety profile, with serious allergic reactions occurring in 1 of every 1,000 or 10,000 doses. A drawback of each method, particularly radioactive tracers, is diffusion from the injection site. This can make detection of nearby sentinel nodes difficult or impossible by masking the node with excessive blue stain or radioactive tracer. Clearly, the ideal sentinel node detection agent has yet to be developed. Injection Sites
The ideal method of injection also remains to be determined. Incorrect localization of lymph nodes may result in a clean "miss," which could leave a cancer-containing lymph node undetected. The injection site is either deep (ie, into or around the primary tumor) or superficial (ie, into the skin over the cancer or at the edge of the areola). Injecting deep into or around the primary cancer has the obvious advantage that the tracer is most likely to be absorbed into the same lymphatic channels that cancer cells enter. The problem with this method is that the location of the injected tracer will vary from patient to patient according to the location of the primary tumor. Typically, less than 1% of the injected dose accumulates in the sentinel lymph nodes. A gamma detector cannot detect a radiolabeled sentinel node if the distance between the primary tumor and node is short; too much "shine through" from the injected tracer masks the nearby sentinel node. A primary cancer in the upper outer quadrant leads to the accumulation of tracer very close to the axillary nodes. A primary tumor in the inner quadrants can mask the presence of internal mammary sentinel lymph nodes. An alternative strategy is to inject a relatively small amount of tracer into the skin overlying the tumor or at the edge of the areola. A smaller amount of tracer can be used because the lymphatics below the skin absorb more tracer. Although the injected material is no longer in immediate proximity to the primary tumor, the pathways from the skin are frequently similar to the underlying gland. The biggest problem is that not all studies that have evaluated the concordance of different injection locations show a 100% node-labeling match.[9-11] The second problem is that extra-axillary lymph nodes do not appear to be labeled by injections into the skin. An additional issue is related to the observation that injections in the periareolar area are not reported to identify pathologically positive sentinel lymph nodes at the same rate. Skin injections over the tumor and deep injections into or around the tumor reveal pathologically positive sentinel nodes at an average rate of 34% to 37%, but the rate for periareolar injections is only 26%. The 10% split between rates may be due to patient selection or other factors, but the difference is enough to warrant a comparative clinical trial. Internal Mammary Nodes There is no reason to conclude that the clinical significance of a sentinel node containing cancer will be different when it resides in different locations. Metastases to the internal mammary nodes appear to confer the same prognosis as metastases to nodes in the axillary basin.[12] However, breast cancer metastasizes to internal mammary nodes less frequently than to axillary nodes, and the information obtained from routine resection of the internal mammary nodes has not been considered worth the risk of morbidity. With sentinel node surgery, the physician can determine if a patient has drainage from the cancer to the internal mammary nodes, allowing individualized selection of patients for biopsy. Importantly, the surgeon is guided to the precise location of the internal mammary nodes. Small incisions can be made with very low morbidity and few cosmetic consequences. Several centers are performing internal mammary node biopsies when the radioactive tracer indicates direct drainage from the primary cancer, and a small percentage of patients with metastases exclusively to the internal mammary nodes are being identified. This is being accomplished without the morbidity historically associated with internal mammary node resection and by limiting the procedure to the subset of patients with documented internal mammary sentinel nodes. Training and Certification The process of documenting the competency of sentinel node surgery is no different than that of other surgical procedures-all require some sort of training and credentialing process. However, widely accepted criteria that define the competency of performing sentinel node procedures have not been established. It is important that the end points for measuring proficiency are distinguished from the end points for establishing the validity of the technique. An assessment of the technical proficiency of an individual surgeon assumes that the procedure has already been validated and that the only question remaining is whether the surgeon is complying with the protocol. Most surgical training is readily accomplished through mentoring. This allows case-by-case evaluation and immediate feedback, with relatively few cases required. It has been demonstrated that as few as five cases are sufficient for a group of mentored surgeons to collectively achieve successful sentinel node resection and false-negative rates on a par with the standards reported in the world's literature.[13] Surrogate End Points
Establishing competency outside the mentor-student relationship is more problematic. Surrogate end points for direct observation are the sentinel node detection rate and the false-negative rate. Although these end points are attractive because they are somewhat objective, the number of cases required to establish an accurate success rate for sentinel node removal (as well as a false-negative rate) is surprisingly high. For example, it would take 77 cases to establish with a power of 80% that a surgeon's sentinel node resection rate was 80% rather than the par of 90%. Even more striking is the number of cases required to establish the false-negative rate of an individual surgeon. It would take about 90 pathologically node-positive cases to establish a false-negative rate of 5% (95% confidence interval limits ± 5%). An additional problem with the use of these rates as end points of competency is that this approach does not allow for immediate feedback if a deficient rate is identified; there would need to be problem-solving, and again, a long period of time to reestablish the rate. As the surgical community deals with issues related to establishing competency, it is important to separate measures of surgical competency from measures of the accuracy and efficacy of any sentinel node resection technique. Pathology Evaluation of Sentinel Lymph Nodes Since at least 1947, we have known that more intensive evaluation of lymph nodes increases the rate of detecting cancer in nodes.[14] However, it has not been considered practical to intensively evaluate every lymph node in an axillary resection specimen. This means that our commonly used breast cancer survival tables are based on information from cases that, in reality, were understaged. In the same way that sentinel node surgery is causing surgeons to reevaluate the management of regional nodes, pathologists are reevaluating protocols for analysis of lymph nodes. Fewer nodes to evaluate make intensive evaluation of nodes more practical. A variety of methods now being studied are based on similar strategies for increased sectioning of the node and the use of immunohistochemistry to make small clusters or individual cancer cells more apparent. The NSABP B-32 trial and the ACOSOG Z0011 study are designed to determine whether such small deposits of occult cancer in sentinel nodes are associated with decreased survival. Until these studies are complete and demonstrate definitively the prognostic value of occult metastases, the consensus of the College of American Pathology is that nodes with cancer deposits limited to micrometastases should be considered node-negative.[15] Conclusions While sentinel node surgery is not expected to increase the cure rate of breast cancer patients, a substantial reduction in the incidence of permanent side effects associated with axillary node resection will be a considerable advance. The completion of clinical trials establishing that no meaningful reduction in survival is associated with the decrease in side effects is important; indeed, it is a credit to the hundreds of clinicians and thousands of breast cancer patients participating in these trials.