Abstract PD06-09: Sentinel Node Biopsy Versus Axillary Lymph Node Dissection in Early-Stage Breast Cancer: A Meta-Analysis of Randomized Controlled Trials
Lyman, GH; Kuderer, NM; Marcom, PK; Olson, Jr, JA
Published in: Cancer Research
Background: Lymphatic mapping with sentinel node biopsy (SNB) has become a widely used technology for reducing morbidity associated with breast cancer staging despite limited data from randomized controlled trials (RCTs). A previous review of cohort studies conducted for ASCO guidelines prior to results from RCTs found considerable variation in SNB staging accuracy. (Kim et al, Cancer 2005; Lyman GH et al J Clin Oncol 2005). Early results from several RCTs of SNB have now been reported. Methods: A systematic review of RCTs of SNB for breast cancer staging was conducted utilizing electronic databases. Study eligibility was limited to RCTs comparing SNB alone in SNB negative patients versus axillary lymph node dissection (ALND) with or without prior SNB. SNB staging accuracy was assessed in study arms where patients were randomized to immediate ALND. Data abstraction was conducted by two independent reviewers. Rates of axillary, locoregional and all recurrence were evaluated along with all-cause and breast cancer-specific mortality. Heterogeneity was assessed by Cochran's Q-statistic and the Inconsistency Index (I2). Weighted summary measures of relative risk (RR) and absolute risk difference (ARD) with 95% CIs were estimated using the method of Mantel-Haenszel.Results: Seven eligible RCTs were identified of which six, involving 9,389 patients, reported either SNB performance and/or rates of recurrence and mortality. Eligibility in these studies was restricted to patients with clinically negative nodes with four limiting accrual to tumors <2-3 cm. Technical success of mapping was reported in five trials including 4,184 (97.1%) and 4,180 (96.8%) patients randomized to SNB versus SNB with immediate ALND, respectively. SNB was positive in 26.5% and 25.4% in those randomized to the SNB and SNB plus ALND arms, respectively. The false negative rate with SNB ranged from 5.5% to 22.9% with a summary estimate of 11.3% [7.5%-16.8%]. The median duration of follow-up across trials ranged from 12 to 95 months with an overall median of 65 months. The risk of axillary recurrence ranged from 0 - 0.8% in SNB patients and 0-0.9% in ALND patients, respectively with RR for axillary recurrence for SNB versus ALND patients of 1.59 [0.67-3.75; P=.292] and ARD of 0.2% [-0.01%-0.4%; P=.082]. The risk of locoregional recurrence ranged from 0.2%-4.6% in SNB patients and 0.8%-3.5% in ALND patients with RR of 1.13 [0.53-2.33; P=.776] and ARD of 0.3% [-1.0%-1.5%; P=.679]. Mortality from any cause has been reported in 222 and 196 patients in the SNB and ALND arms, respectively, with RR for mortality of 1.13 [0.94-1.36; P=.202] and ARD of 0.6% [-0.3%-1.6%; P=.201]. Breast cancer-specific mortality has been reported in 31 and 34 patients in the SNB and ALND arms, respectively, with a RR of 0.90 [0.56-1.45;P=.676] and ARD of 0.3% [-1.6%-1.0%; P=.676].Conclusions: The overall false negative rate with SNB greater than 10% and limited follow-up continue to raise concerns. While no increase in risk of recurrence or mortality in low risk patients managed primarily with SNB is evident to date, continued observation in these studies will be important in addition to more specific data on the accuracy and safety of SNB in higher risk patients.Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr PD06-09.
