Defining the optimal surgeon experience for breast cancer sentinel lymph node biopsy: a model for implementation of new surgical techniques

Abstract

Objective: To determine the optimal experience required to minimize the false-negative rate of sentinel lymph node (SLN) biopsy for breast cancer.

Summary background data: Before abandoning routine axillary dissection in favor of SLN biopsy for breast cancer, each surgeon and institution must document acceptable SLN identification and false-negative rates. Although some studies have examined the impact of individual surgeon experience on the SLN identification rate, minimal data exist to determine the optimal experience required to minimize the more crucial false-negative rate.

Methods: Analysis was performed of a large prospective multiinstitutional study involving 226 surgeons. SLN biopsy was performed using blue dye, radioactive colloid, or both. SLN biopsy was performed with completion axillary LN dissection in all patients. The impact of surgeon experience on the SLN identification and false-negative rates was examined. Logistic regression analysis was performed to evaluate independent factors in addition to surgeon experience associated with these outcomes.

Results: A total of 2,148 patients were enrolled in the study. Improvement in the SLN identification and false-negative rates was found after 20 cases had been performed. Multivariate analysis revealed that patient age, nonpalpable tumors, and injection of blue dye alone for SLN biopsy were independently associated with decreased SLN identification rates, whereas upper outer quadrant tumor location was the only factor associated with an increased false-negative rate.

Conclusions: Surgeons should perform at least 20 SLN cases with acceptable results before abandoning routine axillary dissection. This study provides a model for surgeon training and experience that may be applicable to the implementation of other new surgical technologies.

Figure 1. The learning curve: effect of increasing surgeon experience on the sentinel lymph node identification rate and false-negative rate.

Figure 2. Learning curves associated with various injection techniques used for sentinel lymph node mapping. (A) Sentinel lymph node identification rate. (B) False-negative rate. RC, radioactive colloid.