Purpose: While a number of studies have evaluated the minimum number of axillary nodes that need to be examined to accurately determine nodal positivity or negativity, there is little information on the number of nodes which must be examined to determine the extent of nodal positivity. This study attempts to determine for patients with 1-3 positive nodes the probability that the number of positive nodes reported is the true number of positive nodes as well as the probability that 4 or more nodes could be positive based on primary tumor size and number of nodes examined.
Materials and methods: From 1979 to 1998, 1652 women with Stages I-II invasive breast cancer underwent an axillary dissection as part of their breast conservation therapy and had more than 10 lymph nodes examined. The mean and median number of nodes identified in the dissection was 19 and 17 (range, 11-75). The median age was 55 years. A total of 1155 women had T1 tumors and 497 had T2 tumors. Of the 459 node-positive women, 72% had 1-3 positive nodes, 18% had 4-9 positive nodes, and 10% had 10 or more positive nodes. A mathematical model based on tumor size and number of nodes examined was created using the hypergeometric distribution and Bayes Theorem. The resulting model was used to estimate the accuracy of the reported number of positive nodes and the probability of 4 or more positive nodes based on various observed sampling combinations.
Results: For patients with T1 tumors and 1, 2, or 3 positive nodes, the minimum number of nodes examined needed for a 90% probability of accuracy is 19, 20, and 20. For T2 tumors and 1, 2, or 3 positive nodes, a minimum of 20 nodes is required. The probability of 4 or more positive nodes increases as tumor size and the number of reported positive nodes increase and as the number of examined nodes decreases. For a 10% or less probability of 4 or more positive nodes, a patient with a T1 tumor and 1, 2, or 3 observed positive nodes would require a minimum of 8, 15, and 20 nodes removed. For a T2 tumor and 1, 2, or 3 observed positive nodes, the corresponding numbers are 10, 16, and 20.
Conclusion: The accuracy of the extent of axillary nodal positivity is influenced by the number of observed positive nodes, tumor size, and the number of nodes examined. Underestimation of the number of positive nodes will result in errors in the assessment of an individual's risk for locoregional recurrence, distant disease, and breast cancer death and will adversely impact on treatment recommendations. This model provides the clinician with a means for assessing the accuracy of the number of positive nodes reported in patients with 1-3 positive nodes.