The effect of changing gamma count threshold on sentinel lymph node accuracy

Purpose: Sentinel lymph node (SLN) biopsy has been increasingly accepted in many centers as an alternative to axillary lymph node dissection in the nodal staging of breast cancer. The goal of SLN biopsy is to accurately stage the axilla while minimizing postoperative morbidity. Theoretically, the continuing search for SLNs disrupts additional lymphatics and impacts on operative time. The gamma count threshold is a predefined threshold percentage of the ex vivo count of the "hottest" SLN, which when applied to each individually excised lymph node determines whether a given lymph node is the SLN or a non-SLN. The higher the threshold percentage, the less the number of lymph nodes will meet the criteria of being an SLN. This study examines the hypothesis that changing the gamma count threshold from 10% to 50% will not significantly affect accuracy or the false-negative rate. Methods: We retrospectively reviewed the charts of patients who underwent SLN biopsy with or without completion axillary lymph node dissection from March 1995 to January 2001 at Walter Reed Army Medical Center. Data were collected on gamma counts for each SLN and histopathology of each SLN. For each SLN ex vivo gamma count, percentage of the ex vivo gamma count of the "hottest" SLN was calculated. Results: The SLN identification success rate was 94% (163 out of 174 patients). On average, 2.07 SLNs were removed per patient and 58% of patients had more than 1 SLN removed (94 out of 163 patients). Only 10% had 4 or more SLNs removed (17 out of 163 patients). Sentinel lymph node metastasis was found in 21% of patients (35 of 163 patients). Of these 35 patients with positive SLNs, 8 patients had a negative "hottest" SLN when a less radioactive SLN was positive for metastasis. Changing the gamma count threshold from 10% to 50% lowers the extrapolated accuracy from 98% to 95% and increases the extrapolated false-negative rate from 8% to 21%. Conclusions: The accuracy and false-negative rate of SLN biopsy varies based on the lower limit gamma threshold. Maintaining our 10% gamma count threshold results in acceptable accuracy and false-negative rates comparable to reported literature.