Serum S100 concentrations are not useful in predicting micrometastatic disease in cutaneous malignant melanoma

Background S100 protein is an acidic calcium binding protein that is expressed by melanoma cells. Elevated serum values of S100 have been described in metastatic disease and it has been suggested that it may be used as an adjunct to staging and monitoring of treatment. Micrometastatic disease in the sentinel lymph node can be demonstrated by sentinel node biopsy (SNB) and the sentinel node status is known to be the most important predictor of relapse. Objectives To determine whether serum S100 concentrations could predict the presence of micrometastatic disease. Methods Thirty-one patients with primary cutaneous melanoma > 1 mm were recruited from referrals to the Melanoma clinic. All patients had serum S100 concentrations evaluated prior to undergoing SNB. Serum S100 concentrations were established using an immunoluminometric method. Sentinel nodes were identified using a dual technique with both radiolabelled colloid (residual from preoperative lymphoscintigraphy) and blue dye according to the MD Anderson Cancer Center protocol. Results Nine of these 31 patients had evidence of micrometastatic disease on SNB. The mean serum S100 concentration of those with positive SNBs was 0.027 mug L-1 compared with 0.045 mug L-1 in patients with negative SNBs (normal < 0.14 mu g L-1). No patient in the study demonstrated raised concentrations of serum S100. Conclusions We conclude that serum S100 concentrations do not predict the presence of micrometastatic melanoma in sentinel nodes in primary cutaneous melanoma.