Lymphoscintigraphy in the sentinel lymph node technique for breast tumor: value of early and late images for the learning curve.

As the performance of early (H+1 to H+4) and late (D1) lymphoscintigraphic images raises organizational problems in outpatient surgery for breast cancer, only early images are generally obtained. The present study evaluated whether two series of images are better than one and defined the advantages of both methodologies. One hundred and eighteen patients with infiltrating breast carcinoma (T(0), T(1) and T(2)) were included in the study: 87 in group A (early and late images) and 31 in group B (only early images). All patients received two peritumoral injections of (99m)Tc-sulfur colloid, 15-18 MBq (group A) and <15 MBq (group B). During the operation, the patent blue bye technique was associated with radioactivity detection. The two groups were comparable for histological type and tumor size and localization. Successful localization of sentinel nodes on early lymphoscintigraphic images was significantly greater for group B. The identification of a sentinel node focus on early lymphoscintigraphy increased by 10% during the study. Sentinel node detection by the isotopic method alone, or the two methods combined, was comparable for both groups. In radioactivity detection, the count rate for sentinel nodes versus background (contralateral breast) was similar for the two groups. During the learning phase, two series of images gave a definite advantage. Subsequently, lymphoscintigraphy performed at +2 h was sufficient (the results for the two groups became indistinguishable).