Description

This is a non-inferiority randomized trial with the incidence of patients with metastatic SLN as primary endpoint. Based on recent literature, incidence of patients with metastatic SLN detected by ultrastaging has an average value of 11%. Assuming a -4% as the maximum allowable difference in detection rate to declare non-inferiority, a power of 80% and a significance level of 2.5% (one-side) a sample size of 1922 (961 per arm) is needed. Two interim analyses are planned, one after the first 640 and the second after 1280 patients. At the first interim analysis null hypothesis (inferiority of OSNA with respect to ultrastaging) will be rejected at a significance level \<0.00010 otherwise the study will go on and at the second analysis the null hypothesis will be rejected at a significance level of 0.0059. The final analysis will use 0.0189 as significance threshold. The O'Brien Fleming alpha spending function was used to define these values.

The primary endpoint is defined as the proportion of patients with positive SLN over the total of randomized patients. This proportion will be reported together with the 95% confidence interval to better identify the range of inferential values. Incidence of isolated tumor cells, micro- and macro-metastasis will be calculated.

Number of copies of CK19 mRNA to define volume of metastases will be adapted from previous reports.

To describe the sample, quantitative variables will be summarized using median and interquartile range while categorical items will be reported as absolute counts and percentages. Patients will be described according the arm they were randomized to. A first analysis will be performed on an Intent-To-Treat basis, considering all randomized patients. A secondary analysis will be performed on the Per-Protocol population.