Factors affecting visualization rates of internal mammary sentinel nodes during lymphoscintigraphy

Abstract

There is great variation in the reported frequency of internal mammary (IM) sentinel node (SN) visualization. We observed a marked increase in our IM SN detection rate after 2 factors were changed simultaneously: depth of perilesional injection and dose.

Methods: A retrospective review of 82 consecutive patients (group 1) was compared with 61 consecutive patients (group 2) after changing the depth of perilesional injections and dose. Both groups had perilesional injections of (99m)Tc-sulfur colloid followed by intradermal injections at the areolar cutaneous junction. For group 2, activity was increased in all patients scheduled for next-day surgery. Group 2 had perilesional injections on top of, beside, and just below the estimated level of the tumor in an infiltrative manner, versus injections just on top of and beside the tumor as performed for group 1.

Results: The rates of IM SN visualization were 4.9% (4/82) for group 1 and 23.0% (14/61) for group 2 (P < 0.003). IM SNs were hotter in group 2 than in group 1. The total number of IM SNs detected per patient was also higher for group 2 than for group 1: 2.1 and 1.2, respectively. In group 2, patients with small breasts had an IM SN visualization rate of 46.2%; those with medium breasts, 21.1%; and those with large breasts, 0% (P < 0.017). In group 2, primary lesions located medially had a higher rate of IM SN visualization than did lesions located laterally: 38.9% (7/18) and 16.2% (6/37), respectively (P = 0.066). Dose was not a statistically significant factor within group 2 or group 1 when comparing IM SN visualization rate for doses above or below the mean or median.

Conclusion: Modification of just these 2 factors resulted in a striking change in our IM SN detection rates. The injection depth was the most important factor. Breast size had a marked effect on the probability of detecting IM SNs. This suggests that the variation in detection rates reported in the literature could be at least partly dependent on variations in these factors, among others. Many surgeons do not routinely harvest IM SNs, but information about their presence can potentially alter treatment decisions.