The value of level III clearance in patients with axillary and sentinel node positive breast cancer.

BACKGROUND

The value of level III axillary clearance is contentious, with great variance worldwide in the extent and levels of clearance performed.

OBJECTIVE

To determine rates of level III positivity in patients undergoing level I-III axillary clearance, and identify which patients are at highest risk of involved level III nodes.

METHODS

From a database of 2850 patients derived from symptomatic and population-based screening service, 1179 patients who underwent level I-III clearance between the years 1999-2007 were identified. The pathology, surgical details, and prior sentinel nodes biopsies of patients were recorded.

RESULTS

Eleven hundred seventy nine patients had level I-III axillary clearance. Of the patients, 63% (n = 747) were node positive. Of patients with node positive disease, 23% (n = 168) were level II positive and 19% (n = 141) were level III positive. Two hundred fifty patients had positive sentinel node biopsies prior to axillary clearance. Of these, 12% (n = 30) and 9% (n = 22) were level II and level III positive, respectively. On multivariate analysis, factors predictive of level III involvement in patients with node positive disease were tumor size (P < 0.001, OR = 1.36; 95% CI: 1.2-1.5), invasive lobular disease (P < 0.001, OR = 3.6; 95% CI: 1.9-6.95), extranodal extension (P < 0.001, OR = 0.27; 95% CI: 0.18-0.4), and lymphovascular invasion (P = 0.04, OR = 0.58; 95% CI: 0.35-1). Lobular invasive disease (P = 0.049, OR = 4.1; 95% CI: 1-16.8), extranodal spread (P = 0.003, OR = 0.18; 95% CI: 0.06-0.57), and having more than one positive sentinel node (P = 0.009, OR = 4.9; 95% CI: 1.5-16.1) were predictive of level III involvement in patients with sentinel node positive disease.

CONCLUSION

Level III clearance has a selective but definite role to play in patients who have node positive breast carcinoma. Pathological characteristics of the primary tumor are of particular use in identifying those who are at various risk of level III nodal involvement.