Clinical Trial of Sentinel Node Biopsy Versus Axillary Sampling in Women With Clinically Node Negative Operable Breast Cancer

Axillary node dissection in operable breast cancer provides an accurate way of staging the disease. With increased popularity of screening and earlier diagnosis, most women do not harbor axillary metastases, in which case, axillary lymph node dissection could be avoided. Sentinel node biopsy (SNB) gained rapid popularity because the technique could predict the presence of metastases in downstream lymph nodes, so that surgery and its associated morbidity could be safely avoided. SNB can be performed with injection of isosulphan blue subdermally and peritumorally alone or in combination with radio-guidance with technetium-99 labeled colloid injection. Radio-guided SNB is superior to the blue dye in identifying the SN, and a combined technique is still better. However, it failed to accomplish its objective in a proportion of cases where the sentinel node could not be identified even with expensive gadgetry like gamma probes. An overview of sentinel node studies showed that the node was found in only 85-90% of patients and the overall sensitivity in correctly predicting the negative axilla by sentinel node biopsy was only 94%. There was a false negative rate of 6.2% wherein the sentinel node was reported negative in presence of metastasis to non-sentinel nodes in axilla.

Axillary sampling was a forerunner to the targeted sentinel node biopsy. The main objective of axillary sampling at that time was to detect a positive axilla. More recently it is being tested as an accurate predictor of axillary status for absence of metastasis.

There is a definite learning curve in SNB with higher non-detection rates and false negative rate in the earlier phases of the study compared to the later part. This learning curve is absent in axillary sampling, which requires no specialized training or expensive equipment and has a 100% detection rate. The mean number of sentinel nodes that were identified, as reported in various SNB studies, was 2.13 (with a range of mean numbers of SN dissected being 1.3-3.1). The actual range of number of sentinel nodes dissected was 1 - 8, which is quite comparable to the average 4 -5 nodes that are required to be sampled for accurately predicting the axilla. Badwe and Mittra have concluded that sentinel node biopsy is expensive and is driven by lure for technology and fashion, and has little advantage over axillary node sampling in predicting a negative axilla.

There is a validation study of lower axillary sampling versus total axillary clearance reported by Steele et al in 1985 with a sensitivity of 100% and an accuracy of 99.5%. The more recent validation study of 5-node axillary sampling compared to level I-II dissection was reported from Stockholm on 416 operable breast cancer patients. The noteworthy finding was that in this study node sampling had a 100% yield, 97.3% sensitivity and a negative predictive value of 98.5% with a false negative rate of 2.7% in predicting the axilla. The comparative figures for SNB in the world overview are 94% sensitivity, 97% negative predictive value and a false negative rate of 6.5%.

Preliminary studies of SNB and axillary sampling at TMH Sentinel node biopsy by Isosulphan blue injection was performed as a validation pilot study at the Breast unit in TMH between April 1999 to November 2000 in 100 women with clinically node negative operable breast cancer. SN was found in 77% cases with a false negative rate of 16.6% and sensitivity of 76.1%. The procedure had a negative predictive value of 90.4% cases in predicting the rest of the axilla. The number of sentinel nodes dissected was 1-6 with an average of 1.4.