Lymphatic Mapping And Sentinel Lymph Node Sampling In Breast Cancer: A Meta-Analysis
Reviewer: Roberto Santiago, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 18, 2002
Presenter: T. Kim
Presenter's Affiliation: Albany Medical Center
Type of Session: Scientific
- Sentinel lymph node sampling (SLNS) in breast cancer has become a common practice
- There is a lack of randomized clinical trials comparing SLNS to standard axillary lymph node dissection in the staging of breast cancer.
- This trial was designed to evaluate the efficacy of SLNS through a meta-analysis
Materials and Methods
- The literature was reviewed and studies with SLNS followed by full lymph node dissection in patients with breast carcinoma were identified.
- 69 trials reported between 1994 and 2000 met the study criteria
- 10,454 patients were represented of which 7,765 had a successful lymphatic mapping
- 40% (range, 17-74%) of patients were found to have positive lymph nodes
- Successful mapping was achieved in > 90% of patients in 28 (45%) trials
- The overall false negative rate (FNR) was 8.4%
- 11 of trials reported a FNR of 0% while 26 trials reported FNR > 10%
- The overall Negative Predictive Value (NPV) was 5.7%
- An inverse correlation was noted between the number of patients in the trial and the FNR (r = -0.42, p<0.01) as well as the NPV (r = -0.36, p<0.01)
- A direct correlation was noted between the NPV and the node positive rate (r = 0.39, p<0.01)
- An inverse correlation was noted between the proportion of patients successfully mapped and the FNR (r = -0.32, p=0.009)
- Trials were successful mapping was achieved in > 90% of patients yielded a lower FNR (6.3% vs. 11.1%, p=0.003)
- Lower FNR was noted in trials that included: Blinding (5.6% vs. 8.5%, p=0.03) Patient characteristics stratification (7.8% vs. 11.6%, p=0.009) Measures of test performance (7.0% vs. 10.3%, p=0.003) Measures of variability (6.2% vs. 9.0%, p=0.01) Use of both blue dye and radiocolloid (7.0% vs. 9.9%, p=0.07)
- Multivariate regression analysis revealed that the number of patients was the only parameters significantly related to FNR and NPV (p=0.004 for both parameters)
- Although widely used, more data is needed to establish SLNS as the standard of care
- An obvious learning curve effect can be appreciated from the SLNS data
- The data supporting SLNS as an equivalent to ALND is variable but promising overall
- Further testing of SLNS in newer treatment strategies (like neoadjuvant chemotherapy) is needed
- The use of blue dye in combination with radiocolloid appears to improve the accuracy of SLNS
- Despite the wide use of lymphatic mapping with SLNS, there is substantial inconsistency in the reported performance depending on volume, technique and measures of quality.
As with any new technique that is introduced into the clinical realm there is a learning curve with each surgeon who performs the procedure and a number of procedures must be perfomed before the surgeon is proficient and the results reliable. This has clearly been shown in a number of single institution studies where there is clear improvement in the FNR and NPV based on the number of cases over time. Through some complex statistics, this meta-analysis confirms the fact that experience matters.
Oncolink's ASCO Coverage made possible by an unrestricted Educational Grant from Bristol-Myers Squibb Oncology.