Concordance and validation study of sentinel lymph node biopsy for breast cancer using subareolar injection of blue dye and technetium 99m sulfur colloid.

BACKGROUND

We have previously demonstrated the utility, accuracy, and advantages of a subareolar (SA) site of injection for blue dye compared with an intraparenchymal site. In later studies we advocated the additional use of preoperative SA-injected technetium 99m-labeled sulfur colloid as a directional aid in finding blue-stained sentinel lymph nodes (SLNs). Paramount to the usefulness of this dual-tracer, same-site technique is the degree to which SA-injected blue dye and SA-injected radiocolloid migrate concordantly and are deposited within the same sentinel nodes. The purpose of this study was to document the correlation and accuracy of SLN biopsy using blue dye and radiocolloid when both nodal markers are injected by the same SA route.

STUDY DESIGN

Between September 1999 and February 2002 (29 months), 185 consecutive patients with 187 operable breast cancers underwent 187 attempted SLN biopsies by a dual-tracer, same-site injection technique using the SA approach for both agents. Unfiltered technetium 99m-labeled sulfur colloid (1 mCi [37 MBq]) was SA-injected 30 to 45 minutes preoperatively; and just after anesthetic induction, 3 mL of 1% isosulfan blue dye was injected by the same SA route. SLN biopsies or complete axillary dissections were carried out, and SLNs identified during these procedures were classified as containing both blue dye and radioactivity ("blue-hot" nodes), radioactivity alone ("hot-only" nodes), or blue dye alone ("blue-only" nodes). Cases were categorized and tabulated based on the presence or absence of these three types of SLNs.

RESULTS

Of the 187 procedures, a SLN was identified successfully in 184 cases, indicating an SLN identification rate of 98.4% (95% confidence interval, 96.6% to 100.2%). In these 184 cases, a blue-hot node was present in 94.5% (n = 174 of 184). An SLN was positive in 50 cases, or 27.2% of the total group (n = 50 of 184). A blue-hot node was the only positive SLN in 43 of these 50 cases, or 86% of the node-positive cases. There were no false negatives in 20 confirmatory axillary node dissections carried out to document the findings of a negative SLN. A correlation analysis revealed that in 98.9% of cases (174 of 176), blue nodes were also radioactive ("blue-hot" case concordance = 98.9%). In 95.1% of cases (174 of 183), hot nodes had also taken up blue dye ("hot-blue" case concordance = 95.1%).

CONCLUSIONS

Using SA injections of both blue dye and radiocolloid, we achieved an SLN identification rate of 98.4% (184 of 187 cases), a false-negative rate of 0% (0 of 20 cases), and an accuracy in predicting the malignant status of the axilla of 100% (70 of 70 cases). The case concordance rate ranged between 98.9% ("blue-hot concordance") and 95.1% ("hot-blue concordance"). The present study is the first to evaluate dual-tracer, same-site SA injections of blue dye and radiocolloid. By demonstrating a high case concordance rate, a high SLN identification rate, and a 0% false-negative rate, this study adds further support to the validity and accuracy of same-site SA injections of both blue dye and radiocolloid during SLN biopsy in breast cancer.